Professional Documents
Culture Documents
Co-editors
Jay Rusgrove, W1VD
George Woodward, W1 RN
Contributors
George Collins, AD0W
Doug DeMaw, W1 FB
Gerald Hall, KHD
Gerald Hull, VE1 BXC
John Lindholm, W1XX
John Montague, W0RUE
Bob Shriner, WA0UZO
Harold Steinman, K1 FHN
Ed Tilton, W1 HDQ
Glenn Williams, AFSC
1981
Fifty·Eighth Edition
/
Copyright © 1980 by .i
the American Radio Relay League, Inc.
.I
,Foreword
This 58th edition of The Radio Amateur's Handbook has
been extensively revised and expanded. The 64 additional
pages of this edition include many modern construction
projects, as well as updated tutorial sections. Much of the
theoretical discussion is inspired by the latest FCC
license examinations. ' .
The editors have drawn heavily from extensive profes-
sionar and amateur experience in creating and selecting
the contents of this volume. Therefore, we expect it to
have wide appeal in the electronics industry as well 'as in·
its traditional role as "the amateur's bible."
Newington, CT
November, 1980
/
Schematic Symbols Used in Circuit Diagrams
0
A-A ....ETER
v- VOLTMETER
eOR +
SINGIo!
CELL.
r+-.
CHASSIS
OR
[II
ASSEMBLY
MODtJLE
'fORjr
ANTENNA
60
HEADSET
-iOf-
CRYSTAL
QUARTZ
....
MA-MILL.IAMMETER
0 ~IF- ~
METERS MOTOR
IrIULTIC£LL EART"
GROUNDS
-=-
FUSE
~
MICROPHONE
a:J
SPEAKER
---.£'-0-
HAND
BATTERIES KEY
0==' ~
CONNECTIONS
I
I, -1;>4 4- ~~
~JACk PHONE PLUG SHIELDED WIRE OR CO"'XIAL CABLE
-t=J -J..::Y
D MULT,PLE.
FIXED ~
I
o==e= r----'
1>- EJ
I I
I I
GENERAL rL. ___ -'\
INVERTER
CONNECTORS ENCLOSURE
LOGIC
~SHIELOING·
~ NEON lAiC)
G.:@=
G.
N-CHANNELO
PH.OT
LAMPS
.2
=@=
••
DUAL.-GATE
S
MOSFET "UNCTION FET
E - G ) L 82
U.I"
.iG~E
BIPOL ...R
~
LINEAR INTEGRATED
MosrET TRANSISTORS CIRCUITS
-(J- ~
AIR CORE 'rYfY'
TAPPED
.L -::: >-
DEFLECTION PLATES
GRID PL. ... TE
O~
-EI-'
~ rrr' () r I • PHOTOEMISSIVE ZENEA
(LEO)
DIAC
(SCA)
-+--
FERRITE GAS IG)
BEAD
--- FILL.ED
rim ""*'"
COLD
RING HEATER
OR
INDIRECTLY
HEAlEO C...THODE wJ IT)
. ~
\TI
rrrn' ("(I(') FIL.AMENT C...THODE SEMICONDUCTOR
OIOOE IA[CTIFER TUNNEL TRIAC
CAPACITIVE
IVARACTOR}
RF CHOk£ IRON CORE
ADJUSTABLE
INDUCTORS ELECTRON TUBE ELEMENTS
DIODES
'~ ~
UJJJ SPST DPSl SPOT
("(I(')
AIR CORE
u.w
("(I(')
IRON CORE
ADJUSTABLE
INDUCTANCE
u.w
WITH LINK
m
ADJUSTABLE
COUPLING
®® ~
TRIODE
COMPLETE
---
PENTODE
TUBES
VOLTAGE
REGUL.ATOR
O-t..
-{}
<>-rL
<>-'-'t:....
=ll :[]
RELAYS
0-{
TRANSFORMERS
oOlo
-irt
ELECTR'OI..YTIC
T
.-if-
FIXED
-:*fooJ('
~
VARIABLE
-'\N'v-
~
FIUD
.TAPPIED
~
~
ADJUsrut.E
ifo
SPST
~SPOT
o~
OR +
CROSSING
CONDUCTORS
NOT JOINED ~
~
CONDUCTORS
FEEDTHROUGH SPLIT-STATOR TOGGLE MULTIPOINT OTERMINAL JOINED
-
-
,
CAPACITORS ftMSET RESISTORS SWITCHES WIRING
,
C.ontents
-~
1
2
Amateur Radio
Electrical Laws and Circuits ••
3
4
Radio Design Technique and Language
Solid-State Fundamentals
••
5
6
AC-Operated Power Supplies
H F Transmitting
••
7
8
VHF and UH~ Transmitting
Receiving Systems
••
9
10
VHF and UHF Receiving Techniques
Mobile, Portable and Emergency
Equipment
••
11
12
Code Transmission
....
Slng1e-Sideband Transmission
••
13
14
Frequency Modulation and Repeaters
Specialized Communications Systems
••
15 • Interference with Other Services
16 . Test Equipment and Measurements
••
17 Construction Practices and Data Tables
18 Wave Propagation
••
19
20
Transmission Lines
Antennas for High Frequency
••
21
22
VHF lind UHF Antennas
Operating a Station ••
23 Vacuum Tubes and Semiconductors
~
•
The Amateur's Code
ONE
The Amateur is Considerate • •• He never knowingly uses the air in such a
way as to lessen the pleasure of others.
TWO
The Amateur is Loyal . .• He offers his loyalty, encouragement and sup-
port to his fellow radio amateurs, his'local club and to the American Radio
Relay League, through which Amateur Radio is represented.
/
THREE
The Amateur is Progressive • •• He keeps his station abreast of science. It is
well-built and efficient. His operating practice is above reproach. .
FOUR
The Amateur is Friendly . .. Slow and patient sending when requested,
friendly advice and counsel to the beginner, kindly assistance, cooperation
and consideration for the interests of others; these are marks of the
amateur spirit.
.'
FIVE
The Amateur is Balanced • .• Radio is his hobby. He never allows it to in-
terfere with any of the duties he owes to his home, his job, his school, or his
community. . I
SIX
The Amateur is Patriotic. . . His knowledge and his station are always
, .:. !
ready for the service of his country and his community ..
- PAUL M. SEGAL
Ch'apter 1 \ .
Amateur Ra'dio
•
Amateur Radio. You've heard of it.
You probably know that Amateur Radio
operators are also called "hams."
(Nobody knows quite why!) But who are
these people and what do they do?
Every minute of every hour of every
day, 365 days a year, radio amateurs all
over the world communicate with each
other. It's a way of discovering new
friends while experimenting with different
and exciting new ways to advance the art
of their hobby. Ham radio is a global
fraternity of people with common and yet
widely,varying interests, able to' exchange
ideas and learn more about each other
with each new on-the-air contact. Because
of this Amateur Radio has the ability to
enhance international relations as does no
other hobby. How else is it possible to talk
to an engineer involved in a space pro-
gram, a Tokyo businessman, a U.S.
legislator, a Manhattan store owner, a
camper in a Canadian national park, the
head of state of a. Mediterranean-area
country, a student at a high ~chool radio
club in Wyoming, or a sailor on board a
ship in the middle of the Pacific? And all
without leaving your. home! Only with
Amateur Radio - that's how!
The way communication is accom-
plished is just as interesting as the people
you get to "meet." Signals can be sent
around the world using reflective layers of
the earth's ionosphere or beamed from
point to point from mountaintops by
relay stations. Orbiting satellites that
hams built are used to achieve ·com-
munication. Still other hams bounce their
signals off the moon! Possibilities are
almost unlimited. Not only do radio
amateurs use international Morse code
and voice for communication, but they
also use radioteletype, facsimile and
For many years, the name Hiram Percy Maxim was synonymous with Amateur Radio. The
cofounder and first Presi/Jent of ARRL was the first person to be inducted into the ARRL Hall of
various forms of television. Some hams
Fame. Aside from his pioneering work with radio, Mr. Maxim, who held the call sign W1AW, even have -computers hooked up to their
contributed to the development of the automobile and the movie~. equipment. As new techniques and modes
Amateur Radio 1 ·1
of communication are developed, .hams social as well as operational and techtiical
continue their long tradition of being , activities. The fun provid~ by Amateur
among the first to use them. , , Radio is greatly enhanced when hams get (
What's in the future? Digital voice- together so they can "eyeball" (see) each
encoding techniques? Three-dimensional other. It's a goop supplement to talking to
TV? One can only guess. But if there is each other over the radio. The swapping
ever such a thing as a Star Trek of tales (and sometimes equipment), and a
transporter unit, hams will probably have general feeling of-high spirits a4d a bit.of
them! spice to club meetings along with fechnical
Once radio amateurs make sure that matters on the agenda. Clubs offer many
their gear does work, they look for things people their first contact with Amateur
to do with the equipment and special skills Radio by setting up displays in shopping
they possess. Public service is a very large centers and at such events as county fairs,
and integral part of the whole. Amateur Scout jamborees and parades.
Radio Service. Hams continue this tradi- 'Nearly half of all U.S. amateurs belong
tion by becoming involved and sponsor- to a radio club. And nearly every amateur
ing various activities in their community. radio club is affiliated with the ARRL.
Field Day, just one of many public Club affiliation is available to most
service-type activities, is an annual event Hurtling through the airless reaches of space, organized Amateur Radio groups. The
occurring every June when amateurs take OSCAR 8 is a relay station for amateurs around benefits are many: Use of films and slide
their equipment into the great outdoors the world. It provides reliable vhf and uhf shows for club classes and meetings,
communications, and is the focus of a wide
(using electricity generated at the opera- range of experiments that advance the radio art.
rebate' on ARRL membership dues,
tion site) and test it for use in case of special publication offers, and other ser-
disaster. Not only do they test their vices. Complete information on the re"
equipment, but they make a contest out of quirements and privileges of affiliation is
the exercise and try to contact as ~any available from the Club and Training
other hams operating emergency-type The OSCAR (Orbiting Satellite Carry- Department, ARRL hq., Newington, CT
,stations as possible (along with "or- ing Amateur Radio) program is a relative- 06111.
dinary'~ types). Often they make Field ly new challenge for the Amateur Radio
Day a club social event while they are fraternity, Built by hams from many Getting Started in Amateur Radio?
operating. countries around the world, these, in- "All of this sounds very interesting and
. Traffic nets (networks) meet on the gen.ious devices hitch rides as secondary seems to be a lot of fun, but just how do f '
airwaves on a schedule for the purpose of payloads on space shots for commercial go about getting into this hobby? Don't
handling routine messages for people all and. government communications or you almost need a degree in electronics to
over the country and in other countries weather satellites. OSCAR satellites pass the test and get a license?"
where such third-party traffic is permit- receive signals from the ground on one Nothing could be further from the
ted. By doing sO,amateurs stay in practice frequency and convert those signals to truth. Although you are r~quired to have
for handling mbssages should any real another frequency.to be sent back down a license to operate a station, it only takes
emergency or disaster occur which would to earth. Vhf (very high frequency) and a minimal amount of study and'effort on
require operating skill'to move messages uhf (ultra-high freqwmcy) signals normal- your part to PFlSS the basic, entry-grade
efficiently. Nets also meet because the ly do not have a 'range much greater than exam and get on the air.
members often have common interests: the horizon, but when beamed to these "But what about the code? Don't I
similar jobs, interests in different lan- satellites, a vhf/uhf signal's effective have to know code to get a license?" Yes,
guages, different hobbies (yes, some range is greatly increased to make global you do. International agreementstequire
people have hobbies other than ham communication a possibility. These Amateur Radio operators to have the
radio!), and a whole barrelful of other OSCAR sateollites also send back telemetry ability to communicate in international
reasons. It is often a way to improve one's signals either in Morse or radioteleprinter Morse code. But the speed at which you
knowledge and to share experiences with (RTTY) code, constantly giving informa- are required to receive it is relatively low
other amateurs for the ,good of all tion on the condition of equipment so you should have- no dif.ficulty. Many
involved. aboard the satellite. grade-school students have passed their
OX (distance) contests are popular and Self~reliance has always been a tests and each month hundreds of people
awards are actively sought by many trademark of the radio amateur. This is from 8 to 8e} join the ever-growing
amateurs. This armchair travel is one of often best displayed by the man}" hams number of Amateur Radio operators
the more alluring activities of amateur who design and build their own equip- around the world. ,
radio, There are awards for Worked All ment. Many others prefer to build their Concerning the written exam: 1'0 get a
States (WAS), Worked All Provinces equipment from kits. The main point is license you need to know some basic elec-
(WAVE), Worked All Continents (WAC), that hams want to know how their equip- trical and radio principles and regulations
Worked 100 Countries (DXCC), and ment functions, what to do with it and governing the class of license applied-for.
many others. how to fix it if a malfunction should oc- The ARRL's basic beginner package,
Mobile operation (especially· on the cur. Repair shops aren't always open dur- Tune in the World with Ham Radio, tS
very high frequencies) holds a special ing hurricanes or floods and they aren't available for $7 from 10c.1 radio stores or
attraction to many hams. It's always fun always out in the middle of the Amazon by mail from ARRL.
to keep in touch with ham friends over the jungle, either. Hams often come up with
local repeater (devices w.hich receive your variations on a circuit design in common Finding Help
signal and retransmit it for better coverage use so that· they may achieve a special One of the first obstacles for a person
of the area) or finding new friends on function, or a totally original electronic seriously interested in Amateur Radi6 is
other frequencies while driving across the design may be brought out by a ham, all finding a local amateur to provide
country. Mobile units are often the vital in the interest of advancing the radio art. assistance. This volunteer amateur is
link in emergency communications, too, called an "Elmer." A nearby ham can
since they are usually first on the scene of Radio Clubs help a newcomer with technical advice,
an accident or disaste~ Amateur Radio clubs often provide putting up and testing antennas, advice on
1-2 Chapter 1
Whether it's trekking to the North Pole or flying high in the sky, where hams go Amateur Radio often goes as well. Japanese explorer Naomi Uemura,
JG10FW, used Amateur Radio for backup emergency communications during his solo adventure to the North Pole in 1978. Fred Hyde, K0L1S, was one of
four crew members on the DaVinci Trans America Balloon, which set a long-distance flight record for balloonists in the continental U.S .. before crash·
landing in Ohio because of a severe storm. Amateur Radio kept the 9rew in touch with hams on the ground.
buying that first radio or just some needed longer distances than commercial stations dropped (i.e., frequency increased)
encouragement. Also, nearly all would-be on the longer wavelengths. Even so, greater distances were achieved. The c.om-
amateurs attend an Amateur Radio class signals often had to be relayed by in- mercial staticns were nct ab.out to miss
for code, regulations and electronic termediate amateur stations to get a cut .on this opp.ortunity. They moved their
theory instruction. Where do you find this message to the proper destination_ stati.ons t.o 'the new shorter wavelengths
assistance? The ARRL Club and Training Because of this, the American Radio while the battle raged .over who had the
Department helps the prospective am- Relay League was organized tc establish right t.o transmit in this new area. Usually,
ateur in every possible way. It coordi- routes .of Amateur Radio communicaticn it turned out t.o be the station with the
nates the work of more than 5000 and serve the public interest thrcugh stronger signal, able to blot cut every.one
volunteer Amateur Radio instructors Amateur R.adio. But the dream of even- else. .
thoughout the United States and Canada tual transccntinental and even trans- Nati.onal and internati.onal c'.onferences
and provides a large variety of audio- oceanic Amateur Radic contact burned were .called in the twenties tc straighten
visual aids and refers inquiries on hot in the minds .of Radio Amateur ex- cut the tangle .of wavelength all.ocations.
Amateur Radio to one of the 5000 instruc- perimenters. Through'the efforts of ARRL o ffficials.
tors.If you are looking for an Amateur World War I broke cut and Amateur amateurs .obtained frequencies Aln varicus
Radio class or advice on how to get Radic, still in its infancy, was .ordered .out bands similar t.o what we have tcday: 160
started, write the ARRL Club and Train- .of existence until further notice. Many through 6 meters. When the amateur
ing Department for the name and address fcrmer Amateur Radio operators jcined cperat.ors moved t.o 20 meters, the dream
of the nearest Elmer. . the armed sevices and served with distinc- .of c.oast-tc-coast and transcceanic c.om-
tion as radio operatcrs, finding their skills munication without a relay staticn was
. LooJdng Back to be much needed. finally realized. (A more detailed hist.ory
How did Amateur Radio become the After the close .of the "War to End All .of the early days of Amateur Radio is ccn-
almost unlimi!ed hobby it is today? The Wars," Amateur Radic was still banned tained in the ARRL publication Two
peginnings are slightly obscure, but elec- by law; yet there were many hundreds of Hundred Meters and Down by Clint.on B.
trical ecXperimenters around the turn of fcrmerly licensed amateurs just itching tc DeS.ot.o.)
the' century, inspired by the experiments "get back .on the air." The gcvernment
of.Marconi and others of the time, began had tasted supreme authority over the Public Service
duplicating those experiments and at- radic services and was half inclined tc Amateur R~di.o is a grand and glori.ous
tempted to communicate among keep it. Hiram Percy Maxim, one of the hobby, but this fact alone w.ould hardly
.themselves. There were no regulatory founders .of the American Radio Relay merit the wholehearted supp.ort given it by
agencies at that time and much in- League, called the pre-war League's of- nearly all the w.orld's governments at in-
. terference was caused by these "amateur" ficers together and then ccntacted all the ternati.onal conferences. There are .other
experimenters to other stations until .old members who cculd be fcund in an at- reasons. One .of these is a thorough ap-
governments the world over stepped in tempt to re-establish Amateur Radio. preciation .of the value .of amateurs as
and established licensing, laws and regula- Maxim traveled tc Washingtcn, DC and s.ources of skilled radi.o personnel in time
tions to control the problems involved in after considerable effort (and untold red of war. Another aSset is best described as
this new technology. "Amateur" ex- tape) Amateur Radi.o was .opened up "public service."
perimenter stations were then restricted to again.on Oct.ober 1, 1919. The "public service" record .of the
the "useless" wavelengths of 200 meters Experiments on shorter wavelengths amateur is a brilliant tribute t.o his w.ork.
and below. Amateurs suddenly found that were then begun with enc.ouraging results, These activities can be roughly divided in-
they could achieve communication over It was fcund that as the wavelength to two classes, expeditions and
emergencies. Amateur ~ooperation with several explorations of,the Amlirctic being Sometimes'Mother Nature goes dn a'
expeditions began in 1923, when a League perhaps the best known. And this kind of rampage - with earthquakes such as",
member,~ Don Mix, ITS, accompanied work is not all in the distant past, either:' those in Alaska in 1964, Peru in 1970,
MacMillan to the Arctic on the schooner In 1978 Japanese explorer Naomi California in 1971, Guatemala and Italy in
Bowdoin with an amateur station. Uemura, JGIQFW, became the first per~ 1976; ,floods like those in Big Thompson
Amateurs in Canada and the U.S. pro- son to trek to the North Pole alone. Canyon, Colorado, in 1976, Kentucky,
vided the home contacts. The success of Amateur Radio, through member stations Virginia, West Virginia, and Johnstown,
tpis venture was so outstanding that other of the National Capitol DX Association Pennsylvania in 1977, Jackson, Miss. in
'explorers followed ,suit. During subse- and the 'Polar Amateur Radio Club, 1979; the big .forest fires of California,
quent years Amateur Radio assisted VE8RCS, at Alert, NWT, Canada, pro- particularly in 1977; tornadoes; hur-
perhaps 200 voyages and expeditions, the vided important backup communications. ricanes and typhoons, most anywhere,
Table 1
Canadian Amateur Bands'
. Band
(limlta· Frequency
tions) (MHz) Emissions
80 meters 3.500-3.725 Al, Fl 5650.000·5925.000 A0, A 1, A2, A3, peri mental Service may use such modulation ,
(1,3,4,5) 3.725 Al, A3, F3 A4, A5, Fl, F2, techniques or types of emission for packet '
40 meters 7.000-7.050 Al, Fl F3, F4, F5, P0, transmission as they may select by experimen-
(1,3,4,5) 7.050-7.100 Al; A3, Fl Pl, P2, P3, P4, tation on conditions that they do not exceed
7.100-7.150 Al, Fl P5, P9 the bandwidths established in 10, 11 and 12.
7.150-7.300 Al, A3, F3 10000.000·10500.000 A0, Al, A2, A3: 14) Only packet transmissions shall be used.
20 meters, 14.000-14.100 Al, Fl A4, A5, Fl, F2, 15) Final rf output power used for packet
(1,3,4,5) 14.100-14.350 Al, A3, F3 F3, F4, F5, P0, transmissions shall not exceed 100 watts peak
15 meters 21.000-21.100 Al, Fl Pl, P2, P3, P4, power and 10 watts average power.
(1,3,4,5) 21.000-21.450 Al, A3, F3 P5, P9
10 meters 28.000-28.100 Al, Fl (9, 13, 15) 24000.000-24010.000
(2,3,4,5) 28.100-29.700 Al, A3, F3 24010.000-24050.000 A0, Al, A2, A3, Operation in frequency band 1.800-2.000 MHz
6 meters 50.000-50.050 A1 A4, A5, Fl, F2, shall be limited to the area as indicated in the
(3,4) 50.050-51.000 Al, A2, A3, Fl, F3, F4, F5, P0, following table and shall be limited to the in-
F2, F3 Pl, P2, P3, P4, dicated maximum de power input to the anode
51.000-54.000 AfJ, Al, A2, A3, P5, P9 of the final radio frequency stage of the trans-
A4, Fl, F2, F3, 24050.000-24.050.000 A0, Al, A2, A3, mitter during day and night hours respectively;
F4 A4, A5, Fl, F2, for the purpose of this table "day" means the
2 meters 144.000-144.100 Al F3, F4, F5, P0, hours between sunrise and sunset, and "night"
(3,4) 144.100-145.500 A0, Al, A2, A3, P,l, P2, P3, P4, means the hours between sunset and sunrise.
A4, Fl, F2, F3, P5, P9 Al, A3 and F3emissions are permitted.
F4
(3,4,7) 144.500-145.800 Pfj, Pl, A0, Al,
A2, A3, A4, Fl, A BCDEFGH
F2, F3, F4
(3,4) 145.800-148.000 A0, Al, A2, A3, Limitations British Columbia 3' 3 3 1 000 0
A4, Fl, F2, F3, 1) Phone privileges are restricted to holders Alberta 3' 3 3 3 100 1
F4 of advanced Amateur Radio Operators Certifi- Saskatchewan 3' 3 3 3 3 1 1 3
(3,4) 220.000-220.100 A0, Al, A2, A3, cates, and of Commercial Certificates. ' . Manitoba 3' 2 2 2 222 3'
A4, Fl, F2, F3, 2) Phone privileges are restricted as in foot- Ontario 3 1 1 1 100 2
F4 note 1, and to holders of Amateur Radio Opera- North of 50' N.
(9, 10, 13, 15) 220.000-220.500 tors Certificates, whose certificates have been Ontario 3' 2 o 0 0 0
(9, 10, 13, 15) 220.500-221.000 ent;iorsed for operation on phone in these South of 50° N.
(10, 13, 14, 15) 221.000-223.000 bands. . Province of Quebec 0 0 100 2
(9, 12, 13, 15) 223.000-223.500 3) Amplitude modulation (A2, A3, A4) shall North of 52° N.
(3,4) 223.500-225.000 A0, A1, A2, A3, not exceed ± 3 kHz (6A3), Province of Quebec 3 2 0 00 0 0
A4, Fl, F2, F3, 4) Frequency modulation (F2, F3, F4) shall South of 52' N.
F4 not produce a carrier deviation exceeding New BrunSWick 3 2 1 o0 0 0 0
(4,6) 430.000-433.000 A0, A 1, A2, A3, ± 3 kHz; (6F3) except ttiat in the 52.54 MHz and
Nova' Scotia 3 2 1 o0 0 0 0
A4, A5, Fl, F2, 144.1-148 MHz bands and higher the carrier Prince Edward 'Island 3 2 1 o0 0 0 0
F3, F4, F5 deviation shall not exceed ± 15 kHz (30F3). Newfoundland (Island) 3 1 1 o0 0 0 0
112, 13, 14, 15) 433.000-434.000 5) Slow scan televiSion (A5), permitted by Newfoundland (Labrador) 2 o0 o 000 0
(3,4,8) 434.000-434.500 P0, Pl, P2, P3, special authorization, shall not exceed a band- Yukon Territory 3' 3 3 1 0 0 0 0
A0, Al, A2, A3, width greater than that occupied by a normal District of MacKenzie .3' 3 3 3 1 0 0 1
A5, Fl, F2, F3, District of Keewatin 3. 1 1 3 2 0 0 2
single-sideband voice transmission.
F4,F5 6) Television (A5), permitted by special District of Franklin 0 o0 0 1 0 0 1
434.5OO-45Q.000 A0, A1, A2, A3, authorization, shall employ a system of stan- 'The power levels 500 dayll00 night may be in;
A4, A5, Fl, F2, dard interlace and scanning with a bandwidth creased to 1000 day/200 night when authorized
F3, F4, F5 of not more than 4 MHz. by a Radio Inspector of the Department of Com-
902.000-928.000 A3, F3 7) Pulse modulation with any mode of trans- munications. '
1215.000-1300.000 A0, Al, A2, A3, mission shall not produce signals of'a band-
A4, A5, Fl, F2, width exceeding 15 kHz. Frequency Band
. F3, F4, F5 , 8) Pulse modulation with any mode of
A 1.800-1.825 MHz E 1.9O()-1.925 MHz
2300.009-2450.000 A0, A1, A2, A3, transmission shall not produce Signals of a B 1.825-1.850 MHz F 1.925-1.950 MHz
A4, A5, Fl, F2, bandwidth exceeding 30 kHz. C 1.850-1.875 MHz G 1.950-1.975 MHz
F3, F4; F5, P0, 9) Any mode may be used. D 1.875-1.900 MHz H 1.975-2.000 MHz
Pl, P2, P3, P4, 10) Packet transmissions shall not produce
P5, P9 signals exceeding 10 kHz.
Power Level - Watts
3300.000-3500.000 A0, A1, A2, A3, 11) Packet transmissions shall not produce
A4, A5, Fl, F2, Signals exceeding 25 kHz. o- Operation not permitted
F3, F4, F5, P0, 12) Packet transmissions 'sha" not produce 1- 25 night 125 day
Pl, P2, P3, P4, signals exceeding 100 kHz. 2·- 50 night 250 day
P5, P9 13) Licensees performing an Amateur Ex- 3 - 100 night 500 day
1 ·4 Chapter 1
any year, and the blizzards of 197~ and came along. Much later, transistors were them to industry with surprising 'results.
1980. When disaster strikes, amateurs are utilized; no.w integrated circuits are a part During World War II, thousands of
re~dy, with equipment not needing poWer of the everyday hardware in the Amateur skilled amateurs contributed their
from the electric company, to carrYon Radio shack. This is because the amateur knowledge to the development of secret
communications for police; fire depart- is constantly in the forefront of technical radio devices, both in government and
ments, and relief organizations. The abili- progress. His incessant curiosity and private laboratories. Equally as impor-
ty of radio amateurs to help the public in eagerness to try anything new are two tant, the prewar technical progress by
emergencies is one big reason Amateur reasons. Another is that ever-growing amateurs provided the keystone for the
Radio has surviyed and prospered. Amateur Radio continu~lIyoyercrowds its development of modern military com-
frequency assignments, spurring amateurs munications equipment.
Technical Developments to the development and adoption of new In the fifties, the Air Force was faced
Amateurs started the hobby with spark- techniques to. permit the accommodation with converting its long range com-
gap transmitters, which took up great of more stations. munications from Morse to voice; jet
hunks of frequency space. Then they Amateurs have come up with ideas in bombers had no. room for skilled radio
moved on to tubes when these devices their shacks while at home and then taken operators. At the time, amateurs had been
Table 2
U.S. Ama1eur Radio Frequency Allocations
(:requency Band Emissions Limitations Frequency Band Emissions Limitations Frequency Band Emissions Limitations
kHz
~8oo-2000 A1, A3 1,2 28.000-29.700 A1 A4, A5, F0, Fr,
3500-4000 A1 28.000-28.500 F1 F2, F3, F4, F5, P 5,8
3500-3775 F1 28.500-29.700 A3, F3,A5, F5 3300-3500 A0, A1, A2, A3,
3775-3890 A5, F5 50.0oo-54.QOO A1 A4, A5, F0, F1,
3775-4000 A3, F3 4 50.100-54.000 A2, A3, A4, A5 F2, F3, F4, F5, P 5,12
4383.8 A3J/A3A 13 F1, F2, F3, F5 5650-5925 A0, A1, A2, A3,
7000-7300 A1 3, 4 51. 000-54. 000 A0 A4, A5, F0, F1,
7000-7150 F1 3,4 144-148 A1 F2, F3, F4, F5, P 5, 9
7075-7100 A3, F3 11 144.100-148.000 A0, A2, A3,A4
7150-7225 A5, F5 3, 4 A5, F0, F1, F2, GHz
7150-7300 A3, F3 3,4 F3, F5
14{)00-14350 A1 220-225 A0, A1, A2, A3, 10.0-10.5. A0, A1, A2, A3, 5
14000-14200 Fl A4, A5, F0, Fl, A4, A5, F0, Fl,
14200-14275 A5, F5 F2, F3, F4, F5 5 24.0-24.25 F2, F3, F4, F5 5,10
14200-14350 A3, F3 420-450 A0, AI, A2, A3, 48-50, 71-76 A0, A1, A2, A3,
A4, A5, F0, F1, 165-170, 240-250 A4, A5, F0, F1,
MHz 5,7 F2, F3, 1<4, F5, P
F2, F3, F4, F5
21.000-21.450 A1 1215-1300 A0, A1, A2, A3,
21.000-21.250 F1 A4, A5, F0, F1, Above 300 A0, A1, A2, A3,
21.250-21.350 A5, F5 F2, F3, F4, F5 5 A4, A5, F0, F1,
21.250-21.450 A3, F3 2300-2450 A0, A1, A2, A3, F2, F3, F4, F5, P
Limitations priate Commission engineer in Charge and the borne operations will be permitted on this fre-
1) The use of frequencies in this band Is on a appropriate military area frequency coordinator. quency. Additionally, all stations aperating on
shared basts with the Loran-A radionavigation i) Those portions of Texas and New Mexico this frequency must be located in or within 50
system and Is subject to cancellation or revi- bounded by latitude 33°24' N., 33°53' N., and nautical miles of the state of Alaska.
sion, in whole or in part, by order of the Com· longitude 105°40' W .. and 106°40' W. 14) Ali amateur frequency bands above 29.5
mission, without hearing, whenever the Commis· Ii) The state of Florida, lI1cluding the Key MHz are available for repeater operation except
sion shall determine such action is necessary in West area and the areas enclosed within circles 50.0-52.0 MHz, 144.0-144.5 MHz, 145.5-146.0 MHz,
view of the priority of Loran-A radionavigation of 2OO-mile radius centered at 28°21' N, 80°43' 220.0-220.5 MHz, 431.0-433.0 MHz, and
system. The use of these frequencies by ama- W. and 30°30' N., 86°30' W. 435.0-438.0 MHz. Both the input (receiving) and
teur stations shall not cause harmful interfer- iii) The s~te of Arizona. output (transmitting) frequencies of a station in
ence to Loran-A system. If an amateur station iv) Those portions of California and Nevada repeater operation shall be frequencies available
causes such interference, operation on the fre- south of latitude 37°10' N. and the area within a for repeater oPeration.
quencies involved must cease if so directed by 2OO-mile radius of 34°09' N., 119°11' W. 15) All amateur frequency bands above 220.5
the Commission. 8) No protection in the band 2400-2450 MHz is MHz, except 431-433 MHz, and 435-438 MHz, are
2) Operation shall be limited to [the subbands afforded from interference due to the operation available for auxiliary operation.
and input powers in watts shown on page 6.) of industrial, scientific and medical devices on NOTE .
3) Where, in adjacent regions or subregions, a 2450 MHz. ' The types of emission referred to In the
band of frequencies is allocated to different ser· 9) No protection in the band 5725-5875 is af- amateur rules are as follows:
vices of the same category, the basic principle forded from interference due to the operation of fType A0 - Steady, unmodulated pure carrier.
is the equality of right to operate. Accordingly, industrial. soientific and medical devices on 5800 Type AI '- Telegraphy on pure continuous
the stations of each service in one region or MHz. waves.
subregion must operate so as not to cause ~O) No protection in the band 24.00-24.25 GHz Type A2 - Amplitude tone-modulated
harmful interlerence to services in the other is afforded from interference due to the opera- telegraphy. \
regions or subregions (No. 117,the Radio tion of industrial, scientific and medical devices Type A3-- A-m telephony including single and
Regulations, Geneva, 1959). on 24.125 GHz. double Sideband, with full, reduced or sup-
4) 3900-4000 kHz and 7100-7300 kHz are not 11) The use of A3 and F3 in this band is pressed carrier.
available in the following U.S. possessions: limited to amateur radio slat ions located out· Type A4 - Facsimile.
Baker, Canton, Enderbury, Guam, Howland, Jar- side Region 2. Type A5 - Television.
vis, the Northern Mariana Islands, Palmyra, 12) Amateur stations shall not cause inter- Type F0 - Steady, unmodulated pure carrier.
American Samoa and Wake Islands. ference to the Fixfjd-Satellite Service operating Type Fl - Carrier-shift telegraphy.
5) Aniateur stations shall not cause interfer· in the band 3400-3500 MHz. Type F2 - Audio frequency-shift telegraphy.
ence to th9' Government radio-location service. 13) The frequency 4383.8 kHz, maximum Type f3 - Frequency- or phase-modulated
6) (Reserved) power 150 watts, may be used by any station telephony.
7) In the following areas dc plate input power authorized under this part of communications Type F4 - Fm facsimile.
to the final transmitter stage shali not exceed with any other station authorized in the state of Type F5 - Fm television.
50 watts, except when authorized by the appro- Alaska for emergency communications. No air- Type P - Pulse emiSSions.
MONT.
N. DAK.
wyo.
ABClEFGH
43311002 NEBR.
COLO.
ABCDEFGH KANS.
43200002
N.HEX.
ABCDEFGH
31000002
HAWAII ()
ABCDEFGH ABCDEFGH
43310000 00002113
$egm~t. (kHz)
Input Power (Wlttl) A 1600-1625
o No operation. day or night B 1825-1850t
1 100 day. 25 night C 1850-1875
2 200 day. 50 night o 1875-1900
3 500 day. 100 night E 1900-1925 tThe range 1625-1830 kHz Is used by OX sta-
-4 1000 day. 200 night F 1925-1950 tions attempting to .contact North American
5 125 day. 25 night G' 1950-1975 stations. Use of this "OX Window" should be
6 250 day. 50 night H 1975-2000 avoided by stations on this continent.
using single sideband for about a decade, . example is the OSCAR series of satellites, OSCAR program as older spacecraft are
and were communicating by voice at great initially put together by amateurs who taken out of service. Write ARRL for
distances with both homemade' and worked in the aerospace industry. and more information.
commercially built equipment. Generals launched as secondary payloads with
LeMay and Griswold, both radio ama- other space shots. At this writing eight The American Radio Relay League
teurs, hatched an experiment in which Amateur Radio satellites have been Since its establishment in 1914 by
ham equipment was used to keep in touch launched. OSCARs 7 and 8. portions of Hiram Percy Maxim and Clarence Tuska.
with Strategic Ajr. Command head- which were built by amateurs of several the American Radio Relay. League has
quarters in Omaha, Nebraska, from an different countries. are currently in space been and is today not only the spokesman .
airplane travelling around the"world. The relaying the signals of amateurs. OSCARs for Amateur Radio in the U.S. and
system worked so well, the equipment 7 and 8 can be heard on almost any Canada, but the largest amateur organiza-
needed only slight modification to meet 29-MHz receiver. Development of third- tion in the world. It is strictly of. by and
Air Force needs, and the expense and time generation Phase III satellites- proceeds for amateurs. is noncommercial and has
of normal research and development pro- under the guidance of The Radio Amateur no stockholders. The members of the
cedures was saved. Satellite CorporatiOn (AMSA T) with the League are the owners of the ARRL and
Many youngsters build an early interest assistance of Project OSCAR. Inc., the QST. the monthly journal of Amateur
ill Amateuf- Radio into a career. Later, as original nonprofit grouP. both affiliated Radio published by the League.
professionals, they may run into ideas with ARRL. The new Phase III satellites The League is pledged to promote in-
which th<:;y tryout in ham radio. A good being built by AMSA T will continue the terest in two-way amateur communication
1·8 Chapter 1
. fUld experimentation. It is interested in the terest to a~ateursthe world ·over. ARRL World with Ham Radio is written for the
. relaying of messages by Amateur Radio. sponsors an Intruder Watch Program so person without. previous contact with
It is concerned with the advancement of that unauthorized use of the amateur Amateur Rapio. It is designed to assist the
the radio art. It stands for· the radio frequencies may be detected and ap- prospective amateur to get into the hobby
maintenance of- fraternalism and a high propriate action taken. At the head- in the shortest possible time. Tune in the
standard of conduct. It represents the quarters of the League in Newington, World comes complete with a code in-
amateur in legislative matters. Connecticut, is a well-equipped struction and practice tape. For the per-
One oLthe League's principal purposes laboratory to assist staff members in son seek,ing the General class or higher
is to keep amateur activities so well con- preparatio'n of technical material for QST license, there are the License Manual and
ducted that the amateur will continue t<:> and The Radio Amateur's Handbook. the ARRL Code' Kit. The ARRL also
justify his existence. Amateur Radio of- Among its other activities, the League publishes a s8ries of question and answer
fers its. followers~ountless pleasures and maintains a Communications Department manuals for each class of license. All are
unending satisfaction. It also calls for the concerned with the operating activities of available from eithe.r your local radio
shouldering of responsibilities - the League members. A large field organiza- store or the ARRL.
maintenance of high standards, a tion is headed by a section communica- Once you have studied, taken the test
cooperative. loyalty to the traditions of tions manager in each of the League's 73 and have received your • license, you will
Amateur Radio, a dedication to its ideals sections. There are appointments for find that there is no other thrill quite the
and principles - so that the institution of qualified members in various fields, as same as Amateur Radio. You may decide
Amateur Radio may continue to operate outlined in chapter 22. Special activities to operate on the lowest amateur band,
"in the public interest, convenience and and contests promote operating skill. A 160 meters (see map). Or you may prefer
necessity. " special place is reserved each month in to operate in the gigahertz bands (billions
. In addition to publishing QST, the QST for amateur news from every sec- of cycles per second), where the entire
ARRL maintains an active Amateur tion. future of communications may lie.
Radio station, W1AW; which conducts The ARRL publishes a library of infor- Whatever your interest, you are sure to
code practice and sends bulletins of in- mation on Amateur Radio. Tune in the find it in Amateur Radio.
"
.
\
Some of the manifestations, of elec- 'ferred from one part of the circuit to there are materials that fall in between the'
tricity and magnetism are familiar to another element by finding the numerical classifications of conductor and insulator,
everyone. The effects of static electricity values of entities called voltage and and might be labeled as semiconductors, '
on a dry, wintry day, an attraction by the current. the latter term is applied exclusively to
magnetic north pole to a compass needle, 'Finally, there is the consideration of the materials where the motion of electrons
and the propagation and reception of fundamental properties of the matter that and holes is important.
radio wavj!s are just a few examples. Less makes up the various circuit elements or
easily recognized as being electrical in devices. It is believed that all matter is Electrostatic Field and Potentials
nature perhaps, the radiation of light and made up of complex· structures called All electrical quantities can be' ex-
even radiant ,heat from a stove are atoms which in turn are composed of pressed in the fundamental dimensions of
governed by the same .physical laws that more or less unchangeable particles called (ime, /Qrce and length. In addition, the
describe a signal from a TV station or an . electrons, protons 'and neutrons. Con- quantity or dimension of charge is also
amateur transmitter. The ability to trans- struction of an atom will determine the required. The metric system of units
mit electrical energy through space with- chemical and electrical properties of (SI- - Systeme Imernational d'Unites) is
out any reliance on matter that might be matter composed of like atoms. The almost exclusively used now to specify
in that space (such as in a vacuum) or the periodic table of chemical elements is a such quantities, and the reader is' urged to
creation of a disturbance in space that can classification of such atoms. Electrons become familiar with this system. In the
produce a force are, topics that are play an important role in both the metric system, the basic unit of charge is
classified under the study of electromag- chemical and electrical properties of the coulomb. The smallest known charge
netic fields. Knowledge of the properties matter and elements where some of the is that of the electron' which is
and definitions of fields is important in electrons are relatively free to move .:... '1.6 X 1O·19coulombs. (The proton
understanding such devices as transmis- about. These materials are called conduc- has the same numerical charge except the
sion lines, antennas, and circuit-construction tors. On the other hand, elements where sign is positive.)
practices such as shielding. all of the electrons are tightly bonded in The concept of electrical charge is
Once a field I]roblem is solved, it is the atomic structure are called insulators. analogous to that of mass. It is the mass
often possible to use the results over and Metals such as copper, aluminum, and of an, object that determines the force
over again for other purposes. The field silver are very good conductors while of gravitation&! attraction between the
solution can be used to derive numerical glass, plastics, and rubber are good object and another one. A similar pheno-
formulas for such entities as resistance, insulators. menon occurs with two charged objects,
~ inductance and capacitance or the latter Although electrons play the principal If the charges can be considered to exist
quantities can be determined experi- role in the properties of both insulators at points in space, the force of attraction
mentally. These elements, then, form the and conductors, it is possible to construct (or repulsion ifthe charges have like signs)
building blocks for more' complex con- matter with an apparent charge of is given by the formula
figurations called networks or circuits. opposite sign to that of the electron.
Since there is no need to describe the Actually, the electron is still the charge
physical appearance of the individual carrier but it is the physical absence of an
elements, a pictorial representation is electron location that moves. However, it
often used and it· is ca.lled a schematic is convenient to consider that an actual
diagram. However, each 'element must be charge carrier is present 'and it has been where QI is the numerical value of ,one
assigned a numerical value, otherwise the labeled a hole. Materials in which the charge, Q2 is the other charge value, r is
schematic diagram is incomplete. If the motion of electrons and holes determine ' the distance in meters,c is the'permittivity
numerical values associated with the the electrical characteristics are called of the medium surrounding the charges,
sources of energy (such as batteries or semiconductors. Transistors, integrated and F is the force in newtons. In the case
generators) are also known, it is then circuits and similar solid-state devices are of free space of a vacuum, ~ has a value of
possible to determine the power trans- mad,e up from semiconductors. While 8.854 X 1O·~2 . and is the permittivity of
2,11 Chapter 2
,
free spac~:The product of relative permit- joules of energy to bring a 1:harge of 1 cou.
tivity and to (the permittivity offree space) lomb from a point of 0 voltage to any point
..,
; .... -- --..::
+1.Z5V
-+1.67V ......
gives the permittivity for a condition on the sphere (as indicated by the dotted
where matter is present. Permittivity is lines in Fig. 1). The direction of the force
// /;.---- ...... " also called the dielectric constant and rela- on a charged particle at the surface of the
I / _ .:!"2.~V"" ,
tive dielectric constants for p'lastics such surface of the sphere must be perpendicu-
/ ,I, / ' +5V ' " \,
I \ \ , as polyethylene and Teflon are '2.26 and lar to the surface. This is because charges
, ,,' • ,20 13. 14.
1 I I , I ,
2.1, respectively. (The relative dielectric are able to flow ~bout freely on the con-
\
, \
\ ....
.~.,
, /
I I , constant is also impo(tant in transmis- ductor but not off it. A force with a direc-
sion-line theory. The reciprocal of the tion other than a r-ight angle to the surface
.... ,
\ .... ,
r· .... \.· ..... _
.....
\ ..... ::-_ ....
"
-'"
~/
/
,/
I
square root of the dielectric constant of will have a component that is parallel to
~OV the material used to separate the conduc- the conductor and will cause the charges
----_
, ..... -----
.... .
......
.... ..,/ tors in a transmission line gives the velo- to move about. Eventually, an equilibrium
city factor of the line. The effect of velo- condition will be reached and any initial
city factor will be treated in later chap- field component parallel to the surface
ters.) will be zero. This motion of charge under
If instead of just two charges, a number the influence of an electric field is a very
•
Fig. 1- Field (solid lines) and potential (dotted
lines) lines surrounding a charged sphere.
of charged objects are present, the force important concept in electricity. The tate .
on anyone member is likely to be a com- at which charge flows past a reference
plicated function of the pOl'itions and point is defined as the current. A rate of I
magnitudes of the other charges. Conse- coulomb per second is defined as 1
quently, the concept of electric-field ampere.
strength is a useful one to introduce. The Because of the symmetry involved, the
field strength or field intensity is defined as direction of the electric force and electric
5 -- the force on a given charge (concentrated field can be represented by· the solid
at a point) divided by the numerical value straight lines in Fig. 1. The arrows
of \he charge. Thus, if a force of 1 newton indicate the direction of the force on a
4
existed on a test charge of 2 coulombs, the positive charge, At points away from the
iii field intensity would be 0.5 newtons/ . sphere, less energy will be required to
:; coulomb. , bring up a test charge from zero reference.
03
> Whenever a force exists on an object, it Consequently, a. serie~ of concentric
> will require an expenditure of energy to spherical shells indicated by the dashed
2 move the object against that force. In lines will define the equipotential surfaces
some instances, the mechanical energy around the sphere. From mathematical
may be recovered (such as in a compressed considerations (which will not be dis-
1 ------------- spring) or it may be converted to another cussed here), it can be shown that the
form of energy (such as heat produced by potential will vary as the inverse of the
o+---.-__ r-~--~--_r------
friction). As is the case for electric-field distance from the center of the sph&re.
o 0 30 20
40 50 intensity, it is convenient to express This felationship is indicated by the
DISTANCE SCALE energy + charge as the potential or voltage numbers in Fig. 1 and by the graph in Fig.
of a charged object at a point. For in- 2.
Fig. 2 - Variation of potential with distance for stance, if it took -the expenditure of 5 While the electric field gives the
the charged sphere of Fig. 1. newton-meters (5 joules) to move.a charge direction and magnitude of a force on a
of 2 coulombs from a point of zero energy charged object, it is also equal to the
to a given point, the voltage or potential negative .slope numerical value of the
at that point would be 2.5 joules/coulomb. curve in Fig. 2. The slope of a curve is the
Because of the frequency of problems of rate of change of some variable with
this' type, the dimension of joules/ distance and in this case, the. variable is
coulombs is given a special designation the potentiat This is why the electric field
and one joule/coulomb is defined. as 1 is sometimes called the potential gradient
volt. Notice that if the voltage is divided (gradient being equivalent to slope). In the
by length (meters), the dimensions of field case of a curve that varies as the inverse of
intensity will be obtained and a field the distance, the slope at any point is
strength of one newton/coulomb is also proportional to the inverse of the distance
defined as one volt/meter. The relation- squared .
. ship between field intensity and potential An examination of Fig. 1 would
is illustrated by the following example indicate that the potential variation is
shown in Fig. I. only dependent upon the shape of the
A conducting sphere receives a charge conductor and not its actual physical size.
until its surface is at a potential of 5 volts. That is, once the value of the radius a of
As charges are placed on the surface of a the sphere in Fig. 1 is specified, the
conductor, they tend to spread out into a potential at any other point a given
uniform distribution. Consequently, it will distance from the sphere is also known.
requi.re the same amount of ener,gy to bring Thus, Fig. 1 can be used for any number
a given amount of charge from a point of of spheres with different radii. When it is
O+---~---+---~I----r_--+_- zero reference to any point on the sphere. changed by a certain percentage, all the
20 30 40 ~.
DISTANCE SCALE The outside of the sphere is then said to other values would change by the· same
constitute an equipotential surface. Also, percentage too. However, the amount of
Fig. 3 - Variation of field strength with distance the amount of energy expended will be in- charge required to produce a given
around a sphere charged to 5 volts for spheres of dependent of the path traveled to get to voltage, or voltage change, does depend
different radii. the surface. For instance, it will require 5 upon the size of the conductor, its shape,
Electrical Laws and Circuits 2-2
.. '
,\~
'and its position in relation to other used on the tip of a whip antenna in order cross-sectional area. but for an important
conductors and insulators. For a given to lowedhe field strength under transmit- theoretical case this is assumed to be true
conductor configuration, the voltage is ting conditions. (Fig. 5). •
related to the required charge by the An examination of Fig. 3 reveals that A cylinder of a material with con-
formula the field strength is zero for distances less ductivity 0' is inserted between two end
than a which inclucles points inside the caps of infinite conductivity. The end caps
• V = g, sphere. The implication here is that the
effect of fields and charges cannot
are connected to a voltage source such as
a battery or generator. (A battery consists
penetrate the conducting surface and of a number of cells that convert chemical
where the entity C is defined as the disturb conditions inside the enclosure. energy to electrical energy and a generator
capacitance. Capacitance will be discussed The conducting sphere is said to form an· converts mechanical energy of motion to
in more detail in a later section. electrostatic shield around the contents of electrical energy.) The electric field is also
Since the electric-field intensity is the enclosure. However, the converse is considered to be constant along the
related to the change in potential with not true. That is, charges inside the sphere length, I, of the cylinder and,' as a
distance, like potential, the manner in will cause or induce a field on the outside consequence, the slope of· the potential
which it changes will be·uneffected by the surface. This is why it is very important variation along the ~Iinder will also be a
absolute physical size of the conductor that enclosures designed to confine the constant. This is indicated by the dashed
configuration. However, the exact numerical effects of charges be connected to a point lines in Fig. 5., Since the electric field is
value at any point does depend on the of zero potential. Such a point is often constant, the current density will also be
dimensions of the configuration. This is called a ground. constant. Therefore, the total current
illustrated in Fig. 3 for spheres with dif- entering the end caps will just 'be the
ferent radii. Note that for larger radii, the Fields and Currents product of the current density and the
numerical value of the field strength at the In the last section, the motion of cross-sectional area. The value of the
surface of'the ,sphere (distance equal to a) charged particles in the presence of an electric field will be the quotient of the
is less than it is for smaller radii. This electric field was mentioned in connection. total voltage and the length of the
effect is important 'in the design of with charges placed on a conducting cylinder. Combining the foregoing results
transmission lines and capacitors. (A sphere and the concept of current was and introducing twi:> new entities gives the
capacitor is a device for storing charge. In . introduced. It was assumed that charges following set of equations:
older terminology, it was sometimes. could move around unimpeded on the
and ~
called a condenser.) Even though the same surface of the sphere. In the case of actual
voltage is applied across the terminals of a conductors, this is not true. The charges J = 0' (T}ince J = O'E
transmission line or capacitor, the field appear to bump into atoms as they move '. O'AV
strength between the conductors is going through the conductor under the influence I = J(A) = -(-
to be higher for configurations of small of the electric field. This effect depends
physical size than it is for larger ones. If upon the kind of material used. Silver is a p = ~ and V = I(~)
the field strength becomes too high, the conductor with the least amount of
insulating material (including air) can
"break down." On the other hand, .the
opposition to the movement of charge
while carbon and certain alloys of iron are
R = p,1
"Aa?dV = IR'
effect can be used to advantage in spark rather poor conductors of charge flow. A where P = the resistivity of a conducting
gaps used to protect equipment conhected measure of how easily charge can flow material, R =the resistance. The final
to an antenna' which is subject to through a conductor is defined as the equation is a very basic one in' circuit
atmospheric electricity. The spark-gap conductivity and is denoted by 0'. theory and is called Ohm's Law. Con-
conductors or electrodes are filed to sharp The current density J, ina conductor is figurations similar· to' the one shown in
points. Because the needlepoints appear the rate of charge flow or current through Fig. 5 are very common ones in electrical
as conductors of very small radii, the field a given cross-sectional area. It is related to circuits and are called· resistors.
strength is going to be higher for the same the electric field and conductivity by the It will be shown in a later section that
applied potential than it would be for formula the power dissipated in a resistor is equal
blunt electrodes (Fig. 4). This means the to the product of the resistance and the
separation can be greater and the effect of . J =O'E square of the current. Quite often
the spark gap on normal circuit operation resistance is an undesirabk effect (such as
will not be as pronounced. However, a In general, the conductivity and electric in a wire carryirtg outtent from one
blunt electrode such as a: sphere is' often field will not be con st.ant over a large location to another one) and must be
reduced as much as pOSSible. This, cail be
accomplished by using a conductor with a
low resistivity such as silver (or copper
which ,is close to silver in resistivity, but is
s~
---- not as expensive) with a large cross-
_
- +
v
E,J
sectional area and as short a length as
possible. The current-carrying capability
decreases as the diameter of a conductor
size gets smaller.
~)'
called a network or circuit. It is evident resistors is in the same direction, all the
Y1
.that the energy consumed in a network potential drops have the same sign. The
lOY + must be equal to the energy produced. potential drop is the product of the
Applying this principle to the circuit current in amperes and the resistance I jn
a~ ______ ~~~
10
______ ~
shown in Fig. 6 gives an important ohms. The sums for the emfs and potential
extension of Ohm's Law. .. drops and the resulting current are given
Y6 +
In Fig. 6, a number of sources and by
resistances are connected in tandem or in
Fig. 6- A series circyit illustrating the effects of ~eries to form a circuit loop. It is desired to
emf anti. potential drop. Sum of emf = V1+ V4 = -10 + 5
determine the current I. The current can
be assumed to be flowing in either a = -5 volts
clockwise or counterclockwise direction.
If the assumption is not correct, the sign Sum of pot. drops = V2 + V3 + V5 + V6
mechanics and field theory would be of the current will be negative when the = 1(2 + 4 + 7 + 10) = 231
required to determine the characteristics network equations are solved and the
of an elec;:trical generator and an appiica- direction can be corrected accordingly. In
,tion of chemistry would be involved in de- order to solve the problem, it is necessary ':"5
= 2J = -0.217 ampere·
signing a chemical cell. However, i.t will be to find the sum of the emfs (which is
assumed that this problem has been proportional to the energy produced) and
solved and that the energy source can be to equate this sum to the sufn of the Because the sign of the current is negative,
replaced with a symbol such as that used potential drops (which is proportional to it is actually flowing in a counterclock-
in Fig. 5. the energy consumed). Assuming the wise direction. The physical significance
The term electromotive force (emf) is . current is flowing in a clockwise direction, of this phenomenon is that one source is
applied to describe a source of electrical the first element encountered at point a is being "charged!' For instance, the circuit
energy, and potential drop (or voltage an emf, VI, but it appears to be connected in Fig. 6 might represent a direct current
drop) is used for a device that consumes "backward." Therefore, it receives a (dc) generator and a battery.
Resistance
Given two conductors df the same size larger area will have the lower resistance. that no. 28 has a resistance of 66.17 ohms
and shape, but of different materials, the per thousand feet. Since the desired
amount of current that will flow when a Resistance 0/ Wires resistance is 3.5 ohms, the length of wire
given emf is applied will be found to vary The problem of determining the resis- required will be
with what is called the resistance of the . tance of a round wire of given diameter
material. The lower the resistance, the and· length - or its opposite, finding a
greater the current for a given' value of suitable size and leng~h of wire to supply a
6~:i 7 X 1000 = 52.89 feet.
emf. desired amount of resistance - can be
Resistance is measured in ohms. A easily solved with the help of the copper Or, suppose that the resistance of the wire
circuit has a resistance of I ohm when an wire table given in a later chapter. This in the circuit must not exceed 0.05 ohm
applied emf of I volt causes a current of 1 table gives the resistance, in ohms per and that the length of wire required for
ampere to flow. The resistivity of a thousand feet, of each standard wire size. making the connections totals 14 feet.
material is the resistance, in ohms, of a Example: Suppose a resistam;:e of 3.5 Then
cube of the material measuring one ohms is needed and some no. 28 wire·is on
centimeter on each edge. One of the best hand, The wire table in chapter 17 shows
14
conductors is copper, and it is frequently 1000 X R = 0.05 ohm
convenient, in making resistance cal-
culations, to compare the resistance of the Table 1
material under consideration with that of Relative Resistivity of Metals where R is the maximum allowable
a copper conductor of the same size and Resistivity resistance in ohms per thousand feet.
shape. Table I gives the ratio of the Materials Compared to Rearranging the formula gives
Copper
resistivity of various conductors to that of Aluminum (pure) 1.6
copper. Brass 3.7-4.9
The longer the path through which the Cadmium 4.4 0.05 X 1000
R = = 3.57 ohms/l000 ft.
current flows, the higher the resistiince of Chromium 1.8 14
Copper (hard-drawn) 1.03
that c;:onductor. For direct current and Copper (annealed) 1.00
low-frequency alternating currents (up to Gold 1.4 Reference to the wire table shows that no.
a few thQusand cycles per second) the iron (pure) 5.68 15 is the smallest size having a resistance
resistance is inversely proportional to the Lead 12.8 less than this value.
Nickel 5.1
cross-sectional Ilrea of the path the Phosphor Bronze 2.8-5.4 When the wire is not copper, the
current must travel; that is, given two Silver 0.94 resistance values given in the wire table
conductors of the same material and Steel 7.6-12.7 should be multiplied by the ratios given in
having the same length, but differing in Tin 6.7 Table 1 to obtaiq the resistance.
Zinc 3.4
cr6ss-sectional area, the one with the Example: If the wire in the first example
Electrical Laws and Circuits 2·4
we(e nickel insteac:l of copper, the length having large conductan~has low lesis~
required fo(3.5 ohms would be tan~e, and vice versa. In radio work the
term is used chiefly in connection with
vacuum-tube characteristics. The unit of
66.1 ~.~ 5.1 X 1000 = 1O.37feet conductance is the mho. A resistance of 1
ohm has a conductance. of I mho, a
resistance of 1000 ohms has a con-
'Temperature Effects ductance of 0.001 mho, and so on. A unit
The resistance of a conductor changes frequently used in connection with va-
with its temperature. Although it is cuum tubes is the micromho, or one-
seldom necessary to consider temperature millionth of a mho. It is the conductance
in making resistance calculations for Of a resistance of I MO.
amateur work, it is well to know that the
resistance of practically all metallic con- Ohm's Law
ductors increases with increasing tem- The simplest form of electric circuit is a
perature. Carbdn, however, acts in the battery with a resistance connected to its
opposite way; its ,resistance decreases terminals, as shown by the symbols in Fig.
when its temperature rises, The tem- 8. A complete circuit must have an Fig, 7 - Examples of various resistors, In the
perature effect is important when it is unbroken path so current can flow out of foreground are 1/4-',1/2- and 1-watt composition
necessary to maintain a constant resis- the battery, through the apparatus con- resistors, The three larger cylindrical
tance under all conditions. Special materials nected to it, and back into the battery. com ponents at the center are wirewound power
resistors, The remaining two parts are variable
that have little or no change in resistance The circuit is broken, or open, if a reSistors. pc-board mount althe lower left and
over a wide temperature range are used in connection is removed at any point. A panel mount at the upper center,
that case. switch is a device for making and breaking
cOnnections and thereby closing or open-
Resistors ing the circuit, either allowing current to
A "paokage" of resistance made up into flow or preventing it from flowing.
a single unit is called a resistor. Resistors The values of current, voltage and
having the same resistance value may be resistance in a circuit are by no means
considerably different in size and con- independent of each other. The relation-
struction (Fig. 7). The flow of current ship between them is known as Ohm's
through resistance causes the conductor Law. It can be stated as follows: The cur-
to become heated; the higher the resis- rent flowing in a circuit is directly propor-
tance and the larger the current, the tional to the applied emf and inversely
greater, the amount of heat developed. proportional to the resistance. Expressed Fig. 8 - A simple circuit consisting of a battery
Resistors intended for carrying large as an equation, it is and resistor.
currents must be physically large so the
heat can be radiated quickly to the
surrounding air. If the resistor does not
I (amperes) = ER (volts)
(ohms)
.
Table 2
get rid of the heat quickly it may reach a
Conversion Factors for Fractional and
temperature that will cause it to melt or The equation above gives the value of Multiple Units
burn. current when the voltage and resistance Change
are known. It may be transposed so that From To Divide by Multiply by
Skin Effect each of the three quantities may be found
The resistance of a conductor is not the
same for alternating current as it is for
when the other two are known:
Units Microunits,
Milliunits
Kilounits
.
1000
1,000,000'
1000
SOURCE
-- ! Rl
(total) = Rl + R2 + R3 + R4 + ... where
the dots indicate that as many resistors as
necessary may be added.
Example: Suppose that three resistors
OF EMF
are connected to a source of emf as shown
in Fig: IO~ The emf is 250 volts. R I is 5000
! !l2 ohms, R2 is 20,000 ohms, and R3 is 8000 where the dots again indicate that any
-- -
SERIES
-
R = RI +R2+ R3
= 5000 + 20,000 + 8000
= 33,000 ohms,
in parallel (a very common case), the
formula becomes
Rl X R2
SOURCE
OF EMF 1 R2
The current flowing in the circuit is then R-' = Rl. + R2
-- - = RE = . 250
33,000= 0.00757 amp.
= 7.57 rnA.
Example: If a 500~ohm' resistor is
paralleled with one of 1200 ohms, the
total resistance is
Fig, 9 - ReSistors connected in series and in (We need not carry calculat,ions beyond
parallel.
Rl R2 500 X 1200
three significant figures, and often two R = RI + R2 = 500+1200
will suffice because the accuracy of 600,000
measurements is seldom better than a few =
percent.) , 1700
Rl
= 3530hms
Voltage Drop
5000 Ohm's Law applies to any part of a It is probably easier to solve practical
circuit as well as to the whole circuit. problems by a different method than the
R2 Although the current is the same in all "reciprocal of reciprocals" formula. Sup-
20k
three of the resistances in the example, the pose the three resistors of the previous
total voltage divides among them. The example are connected in parallel as
R3
voltage appearing across each resistor (the shown in Fig. II. The same emf, 250 volts,
8000 voltage drop) can be found from Ohm's is applied to all three of the resistors. The
Law. current in each can be found from Ohm's
Fig, 10- An example of resistors in series, The Example: If the voltage across R 1 (Fig. Law as shown below, I I, being the current
solution of the circuit is worked out in the text. 10) is' call~d El, that across R2 is called through Rl, 12 the current through R2
Electrical Law. and Circuit. 2·6
and 13 the current through R3. The voltage drOps aCl'()SS R t and Req. are voltageappliedio ids 209 volts'! From the
For convenience, the resistance will be El =1 X RI = 23.3 X 5 = 117 volts equation
expressed in kilohms so the current will be E2 = I X Req = 23.3 X 5.71 = 133 volts _.!:.. _ (200)2 40,000
in milliamperes. with sufficient accuracy. These total 250 P- R - 4000 =40(}0
volts, thus checking the calculations so = 10 watts
far, because the sum of the voltage drops
II E 250
=Ri = T = SOmA must equal the applied voltage. Since E2 Or, suppose a current of 20 milliamperes
appears across both R2 and R3. flows through a 300-ohm resistor: Then
250
i2 = E
R2 =2i) = 12.5mA P = PR =(0.02)2 X 300
E2 133
12 = R2 = 20 = 6.65 rnA = 0.0004 X 300
13 E 250 = O.lZ'watt
= R3 = 8 = 31.25 rnA . E2 133
13 =
R3 = 8 = 16.6 rnA
Note that the current was changed from
milliamperes to amperes before sub-
The total current is stitution in the formula.
where 12 = current through R2 Electrical power in a resistance is.
13 = current through R3 turned into heat. The greater the power
== 11 + 12 + I3 The total is 23.25 rnA, which checks the more rapidly the heat is' generated.
= 50 + 12.5 + 31.25 closely enough with 23.3 rnA, the current Resistors for nidio work are made in
= 93.75mA through the whole circuit. ' many sizes, the smallest being rated to
"dissipate" (or carry safely) about 1/8
Power and Energy watt. The largest resistors commonly used
The total resistance of the circuit is Power - the rate of doing work - is in amateur equipment will dissipate about
therefore equal to voltage multiplied by current. 100 watts.
The unit of electrical power, called the
E 250 watt, is equal to one volt multiplied by one Generalized Definition of Resistance
R = T = 93.75 = 2.66 kilohms
ampere. The equation for power therefore Electrical power is not always turned
= 26600hms is into heat. The power used in running a
Resistors in Series-Parallel P = EI
An actual circuit may have resistances where
both in parallel and in series. To illustrate, P = power in watts
we use the same three resistances again,. E = emf in volts
but now connected as in Fig. 12. The 1 = current in amperes
method of solving a circuit such as Fig. 12 '> Rl5000 <' 20k
--:: =- E -250V '> ? R2
'2·7 Chapter 2
motor, for example, is converted to some other form than heat. Th~refore output; The efficiency iIi the above
mechanical motion. The power supplied power used in heating is considered to be example is 60 percent.
to a radio transmitter is largely converted a loss, because it is not the useful power ..
into radio ,waves. Power applied to a The efficiency of a device is the useful Energy
loudspeaker is changed into sound waves. power' output (in its, converted form) In residences, the power company's bill
But in every Cllse of this kind the p'ower is divided, by the power input to the device. is for electrical energy, not for power.
completely "used up" - it cannot "be In a vacuum-tube transmitter, for example, What you pay for is the work that
rec:;overed. Also, for proper operation of the object is to convert power from a dc electricity does for you, not the rate at
the device the power must be supplied at a source into ac power at some radio fre- which that work i!\ done. Electrical work
definite ratio of voltage to current. Both quency. The ratio of the rf power output is equal to power multiplied by time; the
these features are characteristics of resis- to the dc input is the efficiency of the tube. common unit is the watt-hour. which
tance; so it can be ~id that any device that That is, means that a power of I watt has been
dissipates power has a definite value of used for one hour. That is,
"resistance." This concept ofresistance as Po
something that absorbs power at a defi- Err. = Pi W=PT
nite voltage/current ratio is very useful,
since it permits substituting a simple resis- where W = energy lnwatt-hours
tance for the load or power-consuming where Eff. = efficiency (as a decimal) P = power in watts
part of the device receiving power, often Po = power output (watts) T = time in hours
'with considerable simplification of calcu- Pi = power input (watts)
lations. Of course, every electrical device
has some -resistance of its own in the more Example: If the dc input to the tube is Other energy units are the kilowatt-hour '
narrow sense, so a part of the power sup- 100 watts, and the rf power output is 60 and the watt-second. These units should be
plied to it is dissipated in that resistance watts, the efficiency is self-exp lanatory.
and hence appears as heat even though the Energy units are seldom used in
major part of the power m~y be converted Po 60 amateu~ practice, but it is obvious that a
to another form. Efr. = Pi = 100 = 0.6 small amount of power used for a long
time can eventually result in a "power"
Efficiency EffiCiency is usually expressed as a bill that is just as large as though a large
In devices such as motors and vacuum percentage; that is, it tells what percent of amount of power had been used for a very
tubes, the object is to obtain power in the input power will be available as useful short time.
Capacitance
Suppose two flat metal plates are placed, the current flows only during the time of insulation, is called the dielectic constant
close to each other (but not touching) and charge and discharge, and this time is of that particular insulating material. The
are connected ,to a battery through a usually very short. There can be no material itself is called a dielectric. The'
switch, as shown in Fig. 13. At the instant continuous flow of direct current "through" dielectric constants of a number of
the switch is closed, electrons will be a capacitor, but an alternating current can materials commonly used as dielectrics in
attracted from the upper plate to the pass through easily if the frequency is high capacitors are given in Table 3. If a sheet
positive terminal of the battery, and the enough. of polystyrene is substituted for air
same number will be repelled into the The charge or quantity of electricity between the plates of a capacitor, for
lower plate from the negative battery that can be placed on a capacitor is example, the capacitance will be increased
terminal. Enough electrons move into one proportional to the applied voltage aM to 2.6 times.
pJiate and out of the other to make the emf the capacitance of the capacitor. The
between them the same as the emf of the larger the plate area and the smaller the Units
battery. spacing between the plate the greater the, The fundamental unit of capacitance is
If the switch is opened after the plates capacitance. The capacitance also depends
have been charged in this way, the top upon the kind of insulating material be-
plate is left with a deficiency of electrons tween the plates; it is smallest with air Table 3
Dielectric Constants and Breakdown Voltages
and the bottom plate with an excess. The insulation, but substitution of other
Dielectric Puncture
plates remain charged despite the fact that insulating materials for air may increase Material Constant· Voltage··
the battery no longer is connected.. the capacitance many times. The ratio of Air \ 1.0 240
However, if a wire is touched between the the capacitance with some material other Alsimag 196 5.7 240
two plates (short-circuiting them) the than air between the plates, to the Bakelite' 4.4-5.4 300
excess electrons on the bottom plate will capacitance of the same capacitor with air Bakelite, mica-filled 4.7 325-375
Cellulose acetate 3.3-3.9 250-600
flow through the wire to the upper plate, Fiber 5-7,5 150-180
thus restoring electrical neutrality. The Formica 4.6-4.9 450
plates have then been discharged. Glass, window 7.6-8 200-250
The two plates constitute an electrical Glass, Pyrex 4.8 335
capacitor; a capacitor possesses the pro- Mica, ruby 5.4 3800-5600
Mycelex 7.4 250
perty of storing electricity. (The energy Paper, Royalgrey 3.0 200
actually is stored in the electric field Plexiglas 2,8 990
between the plates.) During the time the Polyethyle(le 2,31200
electrons are moving - that, is, while the Polystyrene 2,6 500-700
Porcelain 5.1-5,9 40-100
capacitor is being charged or discharged- METAL Quartz, fused 3.8 1000
PL.cATES
a current is flowing in the circuit even Steatite, low-loss 5.8 150-315
though the circuit is "broken" by the gap
between the capacitor plates. However, Fig. 13- A simple capacitor.
Teflon
• At 1 MHz '*. 2.1 1000-2000
In volts per mil (0.001 inch)
KA
C = 0.224 den - 1)
where C = capacitance in pF
K = dielectric constant of material
between plates
A = area of one side of one plate in (B)
square inches Fig. 15 - Fixed-value capacitors are seen at A. A large computer-grade unit is at the upper left. The
d = separation of plate surfaces in 40-j£F unit is an electrolytic capacitor. The smaller pieces are silver-mica. disk-ceramic. tantalum.
inches polystyrene and ceramic chip capacitors. The small black unit (cylindrical) is a pc-board-mount
electrolytiC. Variable capacitors are shown at B. A vacuum variable is !U the upper left.
n = number of plates
If the plates in one group do not have the
same area as the plates in the other, use unit. The solid dielectrics commonly used audio to several hundred megahertz.
the area of the smaller plates. are mica, paper and special ceramics. An
example of a liquid dielectric is mineral oil. Voltage Breakdown
· Capacitors in Radio The electrolytic capacitor uses aiuminum- When a high voltage is applied to the
The types of capacitors used in radio foil plates with a semiliquid conducting . plates of a capacitor, a considerable force
work differ considerably in physical size, chemical compound between them; the is exerted on the electrons and nuclei of
construction, and capacitance. Some rep- actual dielectric is a very thin film of insu- the dielectric. Because the dielectric is an
· resentative types are shown in the lating material that forms on one set of . insulator the electrons do not become
· photograph (Fig. 15). In variable capaci- plates through electrochemical action when detached from atoms the way they do in
tors (almost always constructed with air a dc voltage is applied to the capacitor. conductors. However, i(the force is great
for the dielectric) one set of plates is made The capacitance obtain~d with a given enough the dielectric wift "break down;"
movable with respect to the other set so plate area in an electrolytic capacitor is usually it will puncture and may char (if it
that the capacitance can be varied. Fixed very large, compared with capacitors hav- is solid) and permit current to flow. The
capacitors - that is, as~emblies having a ing other dielectrics, because the film is so breakdown voltage depends upon the kind
single, nonadjustable value of capacitance thin - much less than any thickness that and thickness of the dielectric, as shown in
- also can be made with metal plates is practicable with a solid dielectric. TaMe 3. It is not directly proportional to
and with air as the: dielectric, but usually The use of electrolytic and oil-filled the thickness; that is, doubling the
are constructed from plates of metal foil capacitors is confined to power-supply thickness does not quite double the
with a thin solid or Hquid dielectric sand- filtering and audio-bypass applications. breakdown voltage. If the dielectric is air
wiched in between, so that a relatively 'Mica and ceramic capacitors are used or any other gas, breakdown is evidenced
large capacitance can be secured in a small throughout the frequency range from by a spark or arc between the plates, but if
2·9 Chapter 2
the voltage is f(,moved the arc ceases,and The voltage across \each capaCitor, is
the capacitor is ready for use again.,
Breakdown will occur at a lower voltage
between pointed or sharp-edged surfaces SOURCE L LL proportional to the total capacitance
divided by the capacitance of the capacitor
i\1 question, so the voltage across CI is
I I T
OF EMF
than between rounded and polished
,surfaces; consequently, the breakdown
voltage between metal plates of given
f PARALLEL
EI _ 0.571
I'
X 2000 = 1142 volts
spacing in air can be increased by buffing
the edges of the plates.
Since the dielectric must be thick to Similarly, the voltages across C2 and C3
withstand high voltages, and since the are
thicker the diel~tric the smaller the
capacitance for a given plate area, a E2 = 0.571 X 2000 571 volts
high-voltage capacitor must have more 2
plate area than a low-voltage one of the SERIES
same cllpacitance. High-voltage, high-
capacitance capacitors are physically large. Fig, 16 - Capaci10rs in parallel and in series.
E3 = 0.571
4
X 2000 286 volts
Inductance
It is possible to show that the flow of This voltage "drop" (which has nothing constant associated with the circuit itself,
current through a conductor is accom- to do with the voltage drop .in any called the inductance of the circuit.
panied by magnetic effects; a compass resistance in the circuit) is the resj.llt of an Inductance depends on the physical
needle brought near' the conductor, for opposing voltage "induced" in the circuit characteristics of the conductor. If the
example, will be deflected from its normal while the field is building up to its final conductor is formed into a' coil, fol'
north-south position. The current, in value. When the field becomes constant example, Its inductance is increased. A
other words, sets up a magnetic field. the induced emf or back emf disappears; coil of many turns will have more
The transfer of energy to the magnetic since no further energy is being stored. inductance than one of few turns, if both
field represents work done by the source Since the induced emf opposes the emf coils are otherwise physically similar'.
ofemf. Power is required for doing work, of the, source, it tends to prevent the Also, if a coil is placed around an iron
and since power is equal to current current from rising rapidly when the core its inductance will be greater than it
mUltiplied by voltage, there must be a circuit is closed. The amplitude of the was without the magnetic core.
voltage drop in the circuit during the time induced emf is proportional to'the rate at The polarity of an induced emf is
in which energy is being stored in the field. which the current is changing and to a always such as to oppose any change in '
the current in the circuit. This means that though. the conductor is notfortn¢ into,.a·
when the current in the circuit is coil. The inductance of a .short length of
'. increasing, work is being done against the' straight wire is small, but it may not be
·.,.
induced emf. by storing energy in the negligible because if the current througb it
magnetic field. If the current in the circuit changes its intensity rapidly enough the
tends to decrease, the stored energy of the induced voltage may be appreciable. This
field returns to the circuit, and thus adds will be the case in even a few inches of
to the energy being supplied by the source wire when an alternating current having a
of emf. This tends to keep the current frequency of the order of 100 MHz, or
flowing. even though the applied emf may higher is flowing. However, at much lower
Fig. 18 - Assorted inductors. A rotary be decreasing or be removed entirely. frequencies the inductance of the same
(continuously variable) coil is at the upper left. The unit of inductance is the henry. wire could be ignored because the induced
Slug·tuned inductors are visible in the lower
foreground. An rf choke (three pi windings) is Values of inductance used in radio voltage would be negligibly small. .
seen at the lower right. equipment vary over a wide range.
Inductance of several henrys is required in Calculating Inductance
power-supply circuits (see chapter on .The. approximate inductance of single-
power supplies), and to obtain such values layer air-core coils may be calculated from
of inductance it is necessary to use coils of the simplified formula
many turns wound' on iron cores. In
radio-frequency circuits, the inductance- alnl
values used will be measured in milli-
L (uH) = 9a + lOb
henrys (a mH, one one-thousandth of a
where L = inductance in microhenrys
henry) at low frequencies, and in micro-
a = coil radius in inches
henrys ( uH, one one-millionth ofa henry)
b = coil length in inches
at medium frequencies and higher. Al- n = number of turns
though coils for radio frequencies may be
~------b------~ wound on special iron cores (ordinary The notation is explained in Fig. 19. This
iron is not suitable), most rf coils made formula is a close approximation for coils
and used by amateurs are of the having a length equal to or greater than
Fig. 19 - Coil dimensions used in the indue· "air-core" type; that is, wound on an 0.8a.
tance formula. The wire diameter does not insulating support consisting of non- Example: Assume a coil having 48 turns
enter into the formula. The spacing has been wound 32 turns per inch and a diameter of
exaggerated in this illustration for clarity. The magnetic material (Fig. 18).
formula is for closewound coils. Every conductor has inductance, even 3/4 inch. This a = 0.75/2 = 0.375, b
= 48/32 = 1.5, and n = 48. Substituting,
2. 0
L = .375 X.375X48X48 = 17.6 uH
V (9 X .375)+(10 X 1.5) .
.F
V V
,/ V To calculate the number of turns of a
single-layer coil for a required value of
i. 0
/V inductance;
B
n = V'-L=-(=9a-+-."""10""'6""')
a
O. 7 '/ ....-
,/
O. 6 ..,..-
/ Example: Suppose an inductance of 10
o. 5 1/ /'
/' uH is required. The form on which the coil
./ is to be wound has a diameter of one inch
:I: 0.4
~ A and isiong. enough to accommodate a coil
of 1-1/4 inches. Then a = 0.5, b = 1.25,
'"
~ 0.3 NO. 12 BARE WIRE and L = 10. Substituting, .
~ 8 TURNS PER INCH
u A INSIDE DIA. -1/2"
V 10 (4.5 + 12.5) V170
= ---o:s-
OJ
o
z 0.2
/
B INSIDE DIA. - 3/4
n = 0.5
= 26.1 turns
A 26~turn coil would be close enough in
practical work. Since the coil will be 1.25
/ inches long, the number of turns perinch
O. 1
I will be 26.1/1.25 = 20.8. Consulting the
1/ wire table, we find that'lllo. 17 enameled
0.07
wire (or anything smaller) can be used. The
0.06
/ proper inductance is obtained by' winding
/ , the required number of turns on the form
C.05 and then adjusting the spacing between
'" 0.04
the turns to make a uniformly spaced eoil
5 10 15 20 1.25 inches long. .
NUMeER OF TURNS
Inductance Cbarts
Fig. 20 - Measured inductance of coils wound with no. 12 bare wire, eight turns to the inch. Most inductance formulas"lose accuracy
The values include half-inch leads. Inches x 25.4 = mm., when applied to small coils (such as are
2·11 Chapter 2
TURNS TURNS
WIRE GAUGE PER WIREGAUGE PER
AWG OR a&S INCH AWG OR a&S INCH
8 7.6 24 46.3
9 8.6 25 51.7
10 9.6 26 58
11 10.7 27 64.9
12 12 28 72.7
13 13.5 29 81.6
14 15 30 90.5
15 16.8 31 • 101
16 18.9 32 113
17 21.2 33 127
18 23.6 34 • 143
19 26.4 35 158
20 29.4 36 175
21 33.1 37 198
22 37 38 224
23 41.3 39 248
1.0
0.9
1/
0.8
/
w
U>-
0.7
/ 1.0
I /
~CO
"'w
U..J
0.6 L 0.9
0.8 J /'
:::>CO
0«
/
W
U >-
Z CO 0.7
I V
!i'" ...«
>-~
..J
0.5
/
U ..J
w
:::> CO 0.6
1ft . . .V
Cl.W 0.4 0 « I /.
/8'
-
~..J
:::> V ~ I- 0.5
:::>«
:=; > 0.3
/ >- ~
..J 0.4
/.
/ Q. lJJ
=> V /
/ l-
..J ..J 0.3
0.2
V
:l
:=;">
«
0.2
V/
0.1
V
/
0.1 /.V
::V
° °
° LENGTH OF COIL IN INCHES
3 4
°
LENGTH OF COIL IN INCHES
• Fig. 21 - Factor to be applied fo the inductance of coils. listed in Fig. 22 - Factor to be ap,plied to the inductance of coils listed in
Table 4 for coil lengths up to five inches. Table 5, as a function of coil length. Use curve A for coils marked
A, and curve B for coils marked B.
9 k _ - :.....
.004
8k _ _ ~:-
.005
7k---lE.._ .006
.007
A
250 125 2S0
6k _ _ ~,....-
245
.008
120 240
.009 235
115 230
.01 225
110 220
5k _ _........ 215
105 210
205
100 20P
SCALE 8
95
90
4k 85
UJ
-1---.02 80
U
Z
:a; <l: 75
<l: I-
Q:
U
(!)
<l: 70
0 Q.
:a; <l: 65
0 u ...J
Z rJ)
::l
«
f-,
Q: rJ) 60
0 a: 3k SCALE A 0
I- UJ
f- SCALE C
U 110
::l >
0 > 105
U 100 50
~ Z
UJ .04 95
>
a: ::l 90 45
(J
~
Z UJ
UJ-
a: 40
I u. ./'" 80
::i .05 75
I- 35
.J
Z 70
:a; <l: 65
co Z 60
co 0 .06
rJ) 55
UJ 50
a:
.07
2k
.08
.09
.1
N
I- rJ)
a: 0
UJ <l:
I a:
<l:
> u.
U 0
Z a:
UJ u Nomograph for determining values of parall~1 resistances and inductances, and series capacitors.
::l
(J :a; The dashed line slllows that a total resistance of 50 ohms is obtained (8 scale)..when 85 ohms (A
UJ
tJ scale) is placed in parallel with 125 ohms (C scale). For greater quantities of R, C and L (250 or
a: greater), add the necessary number of zeros to the numbers of scales A, 8 and C.
u. u
I-
Z
Z <l:
<l: .2 t:
Z U
o
rJ)
<l:
Q.
UJ <l:
a: u
lk
2·13 Chapter 2
1.0 -
"/
NO. 34-.. V/ /
/
NO_ 40 k:I
0.1
~ 1/1/
(A)
:I:
w
~
// / /
W
IMPEDANCE
'"z o
« /
I-
o
:>
II /
~
/
a
/
/
IDEAL !: 0.01 /
/ INDUCTOR
~
V o· 0.5-IN. DIA ROD
~
~~
o·
CAPACITIVE
V~IA II
1.0 10
t FREQUENCY
SELF-RESONANCE WIRE LENGTH (INCHES)
(8)
\ ", - -......
,
__
'",
'......
_~~
-_/ ....
,..'"
",/
/ " I/
/
/
/
I
I
" ....... / //
are Completely use(ess at radio frequencies. " """--- _........ /"
' ....... , -------.,.....---------~- ....... /
T'~
E R
T = RC
U
if.« 60
\ l/DISCHARGE in seconds required for the current to
reach 63 percent of its final value. The
I~-
u
'"~ 40
formula is
where T = time constant in seconds '"
'"
' u
C = capacitance in farads
«
OJ 20
I
"-.
R = resistance in ohms '"
~
..J 0
r-- r-~
- R
g 0 RC 2RC 3RC
Example: The time constant of a 2-~F ~ TIME
where T = time constant in seconds
capacitor and a 250,000-ohm (0.25 M)
L = inductance in henrys
resistor is Fig. 29 - How the voltage across a capacitor R = resistance in ohms.
rises, with time, when charged through a
T = RC' = 0.25 X 2 = 0.5 .second r~sistor. The lower curve shows the way in
which the voltage decreases across the The resistance of the wire in a coil acts as
capacitor terminals on discharging through the
\ If the applied emf is 1000 volts, the same resistor ..
if it were in series with the inductance.
voltage between the capacitor plates will Example: A coil having an inductance
be 630 volts at the end of 1/2 second. of 20 henrys and a resistance of 100 ohms
If C is in microfarads and R in has a time constant of
megohms, the time constant also is in 100
../I--
~
seconds. These units usually are more ~ eo 20
convenient. OJ
;;:
--- --- V = 100 = 0.2 second
If a charged capacitor is discharged ~ 60
through a resistor, as indicated in Fig.
28B, the same time constant applies.· If
:I
"...z 40 / if there is no other resistance in the circuit.
there were no resistance, the capacitor
...
~ 20
/ If a dc emf of 10 volts is applied to such a
would discharge instantly when S was ::>
u
o
/ coil, the final current, by Ohm's Law, is
closed. However, since R limits the o L 2L 3L
current flow the capacitor voltage cannot R R R E 10
instantly go to zero, but it will decrease
TIME
R
=T= 100 = 0.1 A or 100 rnA
just as rapidly as the <.:apacitor can rid
itself of its charge through R. When the
capacitor is discharging through a resis-
tance, the time constant (calculated in
L
-=. E L
The current would rise from 0 to 63
millamperes in 0.2 second after. closing the
switch.
the same way as above) is the time, in An inductor cannot be "discharged" in
seconds, that it takes for the capacitor to the same way as a capaCitor, because the
lose 63 percent of its voltage; that is, for magnetic field disappears as soon as
Fig. 30 - Time constant of an LR circuit.
'the voltage to drop to 37 percent of its current flow ceases. Opening S does not
initial value. leave the inductor "charged." The energy
Example: If the capacitor of the resistance and inductance are in series. In stored in the magnetic field instantly
example above is charged to 1000 volts, it
Fig. 30, first consider L. to have no returns to the circuit when S is, opened.
resistance and also assume that R is zero.
will discharge to 370 volts in 112 second The rapid disappearance 'of the field
through the 250 kQ resistor. Then closing S would tend to send a causes a very large voltage to be induced
, current through the circuit. However, the in the coil - ordinarily many times larger
Inductance and Resistance instantaneous transition from no current than the voltage applied, because the
A comparable situation exists when to a finite value, however small, represents induced voltage is proportional to the
2·17 Chapter 2
speed wJtti which the fieldchan~es. The apPlications a resist.ance-ca~timce (RC)
common result of opening the sV(itch in a time cortstant is involved, and it is usually
100
circuit such as the one shown is that a necessary to know the voltage across the
spark or arc forms at the switch contacts
at the instant of opening. If the inductance ~
70 I" , capacitor at some time interval larger or
smaller than the actual time constant of
:> 50
is large and the current in the circuit is :;: 40 "- the circuit as given by the formula above.
X I"-.
high, a great deal of energy is released in a <t 30 Fig. 31 can be used for the -solution of
very short period of time. It is not at all
unusual for the switch contacts to burn or
:;:
IL
0 20
."'- such problems, since the curve gives the
voltage across the capacitor, in termS of
melt under such circumstances. The spark
or arc at the opened switch can be reduced
or suppressed by, connecting a suitable
capacitor and resi~tor in series across the
f-
Z
UJ
u
0:
UJ
Q.
10
7
" ~
"r--..
percentage of the initial charge, for
percentages between 5 and 100, at any
time after discharge begins.
Example: AO.Ol~",F capacitor is charged
contacts.
Time constants play an important part
in numerous devices, such as electronic
keys, timing and control circuits, and
5
,0.5 1.5
_1_
'RC
2 2,5
'"
3
to 150 volts and then allowed to discharge
through a O.l-megohm resistor. How long
will it take the voltage to fall to 10 volts?
In percentage, 10/150 = 6.7 percent.
maping of ,keying characteristics by From the chart, the factor corresponding
vacuum tubes. The time constants of Fig, 31 - Voltage across capacitor terminals to 6.7 percent is 2.7. The time constant
in a discharging RC circuit, in terms of the
circuits are also important in such of the circuit is equal to RC = 0.1 X 0.01
applications as automatic gain control
initial charged voltage. To..obtain time in
seconds, multiply the factor tlRC by the time = 0.001. The time is therefore 2.7 X 0,001
and noise limiters. In nearly all such constant of the circuit. = 0.0027 second, or 2':1 milliseconds.
'Alternating Currents
The term phase essentially means "time," phase measurements are made with re-
or the time interval between the in- spect to the lowest, or fundamental, fre-
stant when one' thing occurs and the quency.
J instant when a second related thing takes The time interval or "phase difference"
place. The later event is said to lag the under consideration usually. will be less
earlier, while the one that occurs first is than one cycle. Phase difference could be
said to lead. In ac circuits the current measured in decimal parts of a cycle, but
amplitude changes continuously, so the it is more convenient to divide the cycle
concept of phase or time becomes into 360 parts or degrees. A phase degree
important. Phase can be measured in the is therefore 1/36Q of a cycle. The reason Fig, 33 - When two waves 'of the same
frequency start their cycles at slightly different
ordinary time units, such as the second, for this choice is that with sine-wave times, the time difference or phase difference
but there is a more convenient method: alternating current the value of the is measured in degrees. In this drawing wave B
Since each ac cycle occupies exactly the current at any instant is proportional to starts 45 degrees (one-eighth cycle) later than
same amount of time as every other cycle the sine of the' angle that corresponds to wave 'A, and so lags 45 degrees behind A ..
of the same frequency, we can use the the number of degrees - that is, length of
cycle itself as the time unit. Using the time - from the instant the cycle began.
cycle as the time unit makes the specifica- There is no actual "angle" associated with
tion or measurement of phase indepen- an alternating current. Fig. 32 should help
dent of the frequency of the current, so make this method of measurement clear.
long as only one frequency is under con-
sideration at a time. When two or more Measuring Phase
frequencies are to be considered, as in the' The phase difference between two
case where harmonics are present, the currents of the same frequency is the time
or angle difference between corresponding
parts of cycles of the two currents. This is
shown in Fig. 33. The current labeled A
leads the one marked B by 45 degrees,
since A's cycles begin 45 degrees earlier in
time. It is equally correct to say that'B lags
A by 45 degrees.
Two important special cases are shown
in Fig. 34. In the upper drawing B lags 90
degrees behind A; that is, its cycle begins
just one-quarter cycle later than that of A. Fig. 34 - Two important special cases of
When one wave is passing through zero, phase difference. In the upper drawing, the
the other is just at its maximum point. phase difference between A and B is 90
In the lower drawing A and Bare 180 degrees; in the lower drawing the phase
degrees out of phase .. In this case it does difference is 180 degrees,
not matter which one is considered to lead
Fig. 32 - An ac cycle is divided off into 360
or lag. B is always positive while A is The waves shown in Figs. 34 and 35
degrees that are used as a measure of time or negative, and vice versa. The two waves could represent current, voltage,. or both.
phase. are thus completely out of phase.' A and B might be two currents in separate
Electrical Laws and Circuits 2·18
.'
repeat the events -,>f the.£rst and seCOnd,'
respectixely, with. this.differenc~ ..,.;....the
. polarity of the applied vohage;lJ;c.has
.... ""-',\ reversed, and the current chnges to
" CURRENT /
+ ',/ I , correspond. In other words, analterna-
I \
100 \ I \ ting current flows in the circuit because of
I \ the alternate charging and discharging of
92 I \
71
I' the capacitance. As shown by Fig. 35, the
I
I current starts its cycle 90 degrees before
38
I the voltage, so the current in a capacitor
I
I leads the applied voltage by 90 degrees.
w
o
:>
I-
,
I
I
Capacitive Reactance
C ~O~----~~~B-C~~~~~1------+------f------ The quantity of electric charge that can
Q. ,
be placed on a capacitor is proportional to
~ \
\ the applied emf and the capacitance; This
,
\
amount of charge moves back and forth in
\
\
the circuit once each cycle, and so the rate
\ of movement of charge - that is, the cur-
\
\ rent - is proportional to voltage, capa-
\
\
citance and frequency. If the effects of
\ , capacitance and frequency are lumped·
" / together, they form a quantity that plays a
'-" part similar to that of resistance in Ohm's
Law. This quantity is called reactance.
and the unit for it is the ohm,just as in the
~ase of resistance. The formula for it is
Fig. 35 - Voltage and current phase relationships when an alternating voltage is applied to a
capacitor. 1
Xc = 2:rfC
--
where Xc = capacitive reactance in ohms
f frequency in hertz
circuits, or A might represent voltage and current during AB is smaller than during
B current in the same circuit. If A and B ~A. In the third interval, BC, the voltage
C capacitance in farads
represent two currents in the same circuit rises from 71 to 92, an increase of 21 volts. 3.14
(or two voltages in the same circuit) the This is less than the voltage increase
total or resultant current (or voltage) also during AB, so the quantity of electricity Although the unit of reactance is the
is a sine wave, because adding any number added is less; in other words, the average .ohm, there is no power dissipation in
of sine waves of the same frequency current during interval BC is still smaller. reactance. The energy stored in the.
always gives a sine wave also of the same In the fourth interval, CB, the voltage capacitor in one quarter of the cycle is
frequency. increases only 8 volts; the chl\lrge added is simply returned to the circuit in the next.
smaller than in any preceding interval and The fundamental units '(cycles per
Phase in Resistive Circuits therefore the current also is smaller. second, farads) are too cumbersome for
When an alternating voltage is applied By dividing the first quarter cycle into a practical use in radio circuits. However; if
to a resistance, the current flows exactly, in very large number of intervals, it could be the capacitance is in microfarads V-£F)
step with the voltage. In other words, the shown that the current Charging the and the frequency is in megahertz (MHz),
voltage and current are in phase. This' is capacitor has the shape. of a sjne wave, the reactance will come out in ohms in the
true at any frequency if the resistance is just as the applied voltage does. The formula.
"pure" - that is, is free from the reactive current is largest at the beginning of the Example:. The reactance of a capacitor
effects discussed in the next section. cycle and becomes zero at the maximum of 470 pF (0.00047 uF) at a frequency of
Practically, it is often difficult to obtain a value of the voltage, so there is a phase 7150 kHz (7.15 MHz) is .
purely resistive circuit at radio frequencies, difference of 90 degrees between the
because the'reactive effects become more voltage and current. .During the first 1 1
pronounced as the frequenc~ is i.ncreased. quarter cycle the current is flowing in the X = 2:rfC = 6.28 X 7.15 X 0.00047
In a purely resistive circuit, or for normal direction through the circuit, since
purely resistive parts of circuits, Ohm's the capacitor is being charged. Hence the = 47.4 ohms
Law is just as valid for ac of any frequency current is positive, as indicated by the
as it is for dc. dashed line in Fig. 35. Inductive Reactance
In the second quarter cycle - that is, in When an alternating voltage is applied
Reactance: Alternating Current in the time from D to H, the voltage applied to a pure' inductance (one with no
Capacitance to the capacitdr decreases. During this resistance - all practical inductors have
In Fig. 35 a sine-wave ac voltage having time the capacitor loses its charge. resistahce) the current is again 90 degrees
a maximum value of 100 is applied to a Applying the same reasoning, it is plain out of phase with the applied voltage.
capacitor. In the period OA, the applied that the current is small in interval DE However, in this case the current lags 90
voltage increases from 0 to 38; at the end and continues to increase during each degrees behind the voltage - the opposite
of this period the capacitor is charged to succeeding interval. However, the current. of the capacitor current-voltage relation-
that voltage. In interval AB the voltage is flowing against the applied voltage ship. .
increases to 71; that is, 33 volts additionaL because the capacitor is discharging into The primary cause for this is the back
In this interval a sma/lel: quantity of the circuit. The current flows in the emf generated in the inductance, and since
charge has been added than in OA, negative direction during this quarter the amplitude of the back emf is
because the voltage rise during interval cycle. proportional to the rate at which the
AB is smaller. Consequently the average The third and fourth quarte~ cycles current· changes, and this in turn is
2-19 Chapter 2.
\
+ 1.0
~~~-----+~~
'"
'......
\
~~r~~gE
I
I
I
/-"',
/. \ \\
\ I
II
/
/-.. ,
\ \\
\
frequen'cies between 100 hertz and 100
megahertz. The approximate value of
reactance can be read from the chart or,
where more exact values are needed, the
0.92 CURRENT', / . \
~0~.7~1~____~~4-~V ,I , chart will serve as a check on the order of
\ I magnitude of reactances calculated from
0.38 \\ II the formulas given above, and thus avoid
\ I "decimal-point errors."
w \'
o 1/
~ ij Reactances in Serie; and Parallel
~Or-----~~~~C~~~D~E~~~-----+------~-----
When reactances of the same kind are
~ 1\ connected in series or parallel the resultant
« I\
reactance is that of the -resultant inductance
I NDUCED I
:
/ \
\
\
. or capacitance. This leads to the same rules
VOLTAGE'---...,,/ \ I
that are used when determining the
/ \ I resultant resistance when resistors are
/ \ / combined. That is, for series reactances of
/ \ /
/ \ I the same kind the resultant reactance is
I \ I
.,/
/ '... ,_// ,,_/ /
/
X = XI + X2 + X3 + X4
proportional to the frequency, the ampli- The resistance of the wire of which the x = 'I 1 1 1
tude of the current is inversely propor- coil is wound has no effect on the Xl + Xf + X3 + X4
tional to the applied frequency. Also, reactance, but simply acts as though it
since the back emf is proportional to were a separate resistor connected in
inductance for a given rate of current series with the coil. or for two in parallel,
change, the current flow is inversely
proportional to inductance for a given Ohm's Law for Reactance
X = XI X X2
'='X:7I---'+:-'--::-:X-=-2
applied voltage and frequency. (Another Ohm's Law for an ac circuit containing
way of saying this is that just enough only reactance is
current flows to generate an induced emf The situation is different when reac-
that equals and opposes the applied E tances of opposite kinds. are combined.
voltage.) =-X Since the current in a capacitance leads
The combined effect of inductance and the applied voltage by 90 degrees and the
frequency is called inductive reactance,
also expressed in ohms, and the formula E = IX current in an inductance lags the applied
voltage by 90 degrees, the voltages at the
for it is terminals of opposite types of reactance
are 180 degrees out of phase in a series
circuit (in which the current has to be the
same through all elements), and the
where XL = inductive reactance in ohms where E emf in volts = currents in reactances of opposite types
f = frequency in hertz amperes I = current in are 180 degrees out of phase in a parallel
L = inductance in henrys X '" reactance in ohms circuit (in which the same voltage is
r = 3.14 The reactance in the circuit may, of applied to all elements). The 180-degree
Example: The reactance of a coil having course, be either inductive or capacitive. phase relationship means that the currents
an inductance of 8 henrys; at a frequency Example: If a current of 2 amperes is or voltages are of opposite polarity, so in
of 120 hertz, is flowing through the capacitor of the the series circuit of Fig. 38A the voltage
XL = 2 rtl. = 6.28 X 120 X 8 = 6029 ohms earlier example (reactance = 47.4 ohms)' at EL across the inductive reactance XL is of
7150 kHz, the voltage drop across the opposite polarity to the voltage Ec across
In radio-frequency circuits' the in- capacitor is the capacitive reactance Xc. This if we
ductance values usually are small and the call XL "positive" and Xc "negative" (a
frequencies are large. If the inductance is . E = IX = 2 X 47.4 = 94.8 volts common convention) the applied voltage
expressed in millihenrys and the frequency E AC is EL - Ec. In the parallel circuit at
in kilohertz, the conversion factors for the If 420 volts at 120 hertz is applied to the B the total current, I, is equal to IL - I c ,
two units cancel, .and the formula for 8-henry inductor of the earlier example, since the currents are 180 degrees out of
reactance may be used without first the current through the coil will be phase.
converting to fundamental units. Similarly, E 420 In the series case, therefore, the resul-
no conversion is necessary if the in- I X 6029 =
.0697 A 69.7 = = = tant reactance of XL and Xc is
ductance is in micro henrys and the
frequenc1 is in megahertz. rnA
Example: The reactance of a 15-
microhenry coil at a frequency of 14 MHz' Reactance Chart and in the parallel case (Fig. 38B),
is The accompanying chart, Fig. 37,
shows the reactance of capacitances from X = -XLXC
Xl = 2rtl. = 6.28 X 14 X 15 = 13190hms I pF to 100 ~F, and the reactance of XL-XC
'" "" K
V
""
V L /
~ ~VV ~ ><
~
~/
/
",,'K V/ ",,'K
'"
/ /V- / V
10,000
/ /
~ ~+
(y~
7,000
.:! c,o -/--
5,000
/
/
,,0 /
"-1'\ / / ,, L
'0
~;(),o-
L
L
3,000
I'll / V V .,- lL. / II
~+/ II
/
, V
~ .L ~ !/
30
/ /
V
/ ~"",o-
V
L ~
, V
/ ~
I'
/
V , V ]\
/V V tV ~ L ~ /
"" KV ""
,,0
20
"
~
V'
""" K """ P<
~ ~K ~KlL K
V'
10 Iv V vV' V
.L
vV' V
~
.
V
V
7 -, ." "t
5
/ / '0
,"",0- / / /
/ / / /
3 V V IK it V V
V V V V
~ tv ~
,,"""",0- "" 0"
""
2
V /
V ~" .,/ k ~
vV'
~K .,L
vV'
~K .,/
V
~ 2iV ~" ~
10 0 0 0 0 0 N It) U')"'" 2 0 0 0 0 0 0 0 0 0 0 N If) 11').... 2 0 0 0 0 0
g ~ g ~ ~ ~ N tf) It)"" 2 2 ~ g.... ~ ~ N rt) .",.... g
Fig:37"": Inductive and capacitive reactance vs. frequency. Heavy lines represent multiples of 10, intermediate light lines multiples of five: e.g., the
light line between 10 I'H and 100 I'H represents 50 I'H; the light line between 0.11'F and 11'F represents 0.5I'F, etc. Intermediate values can be
estimated with the help of the interpolation scale. Reactances outside the range of the chart may be found by applying appropriate factors to values
within the chart range. For example, the reactance of 10 henrys at 60 Hz can be found by taking the reactance to 10 henrys at 600 Hz and dividing by
10 for the 10·times decrease in frequency.
Note that in the series circuit the total and infinitely large in the parallel circuit.
reactance is negative if Xc is larger than
XL; this indicates that the total reactance Reactive Power
is capacitive in such a case. The resultant In Fig. 38A the voltage drop across the
reactance in a ser.ies circuit 'is always inductor is larger than the voltage applied
smaller than the larger of the two in- to the circuit. This might seem to be an
dividual reactances. impossible condition, but it is not; the
In the parallel circuit, the resultant explanation is that while energy is being
reactance is negative (i.e., capacitive) if stored in the inductor's magnetic field,
XL is larger than Xc, and positive (induc- energy is being returned to the circuit
tive) if XL is smaller thanXc, but in every from the capacitor's electric field, and vice
case is always larger than the smaller of versa. This stored energy is responsible for (si
the two individual reactances. the fact that the voltages across reactances
In the special case where XL Xc, the = in series can be larger than the voltage Fig. 38 - Series and parallel circuits
total reactance is zero 'in the series circuit applied to them. containing opposite kinds of reactance.
2·21 Chapter 2
In a 'resistance the flow of current phase with the applied voltage while in a considered later in connection with tuned
causes heating and a power loss equal to reactance it is 90 degrees out of phase with circuits and transmission lines.
PRo The power in a reactance is equal to the voltage, the phase relationship be- A series RX circuit can be converted in- ,
PX, but is not a "loss"; it is simply tween current and voltage in the circuit as to its parallel equivalent by means of the
power that is transferred back and forth a whole may be anything between zero formula
between the field and the circuit but not and 90 degrees, depending on the relative
used up in .heating anything. To dis- amounts of resistance and reactance.
tinguish this "nondissipated" power from
the power which is actually consumed, the Series Circuits
unit of re'active power is called the When resistance and reactance are in
volt-ampere-reactive, or var, instead of the series, the imp~dance of the circuit is
watt. Reactive power is sometimes called
"wattless" power.' . Z=J Rl+Xl
where the subscripts p and s represent the
Impedance where Z = impedance in ohms parallel- and series-equivalent values,
When a circuit contains both resistance R = resistance in ohms respectively. If the parallel values are
and reactance the combined effect of the X = reactance in ohms known, the equivalent series circuit can be
two is called impedance, symbolized by the found from
The reactance may be either capacitive'~r
letter Z. (Impedance is thus a more gene- inductive. If there are two or more
ral term than either resistance or reac- reactances in the circuit they may be
tance, and is frequently used even for combined into a resultant by the rules
circuits that have only resistance or previously given, before substitution into
reactance although usually with a qualifi- the formula. above; similarly for ,resis-
cation - such as "resistive impedance" to tances.
indicate that. the circuit has only resis- RsRp
The "square root of the sum of the
tance, for example.) squares" rule for finding impedance in a Xp
The reactance and resistance com- series circuit arises from the fact that the Circuits containing reactance a'\d
prising an impedance may be connected voltage drops across the resistance and resistance in any series or parallel com-
either in series or in parallel, as shown in reactance are 90 degrees out of phase, and bination are called complex circuits. The
Fig. 39. In these circuits the reactance is so combine by the same rule that applies
shown as a box to indicate that it may be term "complex" means that the
in finding the hypotenuse of a right- numerical resistance and reactance values
either inductive or capacitive. In the series angled triangle when the base and altitude
circuit the current is the same in both can't be combined arithmetically because
are known. the voltages (in series circuits) and cur-
elements, with (generally) different vol-
tages appearing across the resistance and rents (in parallel circuits) are not in phase.
Parallel Circuits
Complex notation for a series RX circuit
reactance. In the parallel circuit the same
voltage is applied to both elements, but
With resistance and reactance in paral- has the form R ± jX, where j
lel, ..as in Fig. 39B, the impedance is
= v=-r:
different currents flow in the two branches. The reactive portion of the impedance is
Since in a resistance the current is in called the imaginary component, because
the square root of a negative number can
RX be represented only by a mathematical
z = ,J operator. This is so because squaring a
positive or negative number always pro-
duces a positive result.
where the symbols have the same meaning If the reactance is inductive, the sign of
10 as for series circuits. the j operator is positive; a negative sign
Just as in the case of series circuits, a indicates a capacitive reactance. The
EAC R number of reactances in parallel should be resistive part of the impedance is called
combined to find the resultant reactance . -the real component.
before substitution into the formula above; The magnitude of the impedance (in
similarly for a number of resistances in series-equivalent form) is represented by
(A) (8)
parallol. Z =VR2 + X2. Magnitude is simply a
numerical quantity expressing the ratio of
• Equivalent Series and P~rallel Circuits voltage to current at the -terminals of the
Fig. 39 - Series and parallel circuits The two circuits shown in Fig. 39 are complex circuit - it provides no infbrma-
containing resistance and reactance.
equivalent if the same current flows when tion about the type or amount of reac-
a given voltage of the same frequency is tance present.
applied, and if the phase angle between If parallel circuits are to be expressed in
voltage and current is the same in both R ± jX form, the R and X components
cases. It is in fact possible to "transform" must first be transformed into their series-
any given series circuit into an equivalent equivalent values. A useful complex nota-
~
SR=75 OHMS parallel circuit, and vice versa. tion for parallel circuits expresses the
<j Transformations of this type often lead components in terms of conductance and
tt> simplification in the solution of susceptance, the resultant being admit-
EAC ..... complicated circuits. However, from the tance. These concepts are treated in detail
....-J XL = 1000HMS standpoint of practical work the useful- in the section on radio frequency circuits.
-< ness of such transformations lies in the For a thorough explanation of complex
fact that the impedance of a circuit may be circuits with practical examples, see Hall,
modified by the addition of either series o r " A Simple Approach to Complex Cir-
parallel elements, depending on which cuits," July, 1977 QST.
Fig. 40 - Circuit used as an example for happens to be most convenient in the Another way to represent a complex im-
impedance calculations. particular case. Typical applications are pedance is to indicate the magnitude and
Electrical Laws and Circuits 2-22
the phase angle in the polar form Z L e. This current flows thr~ugh both the must be followed by the words "leading"
Given any two of the three quantities R, X n:sistance and reactance, so the voltage or "lagging" to identify tire phase of the
and Z, the phase angle can be determined drops are voltage with respect to the current. Speci-'
by fying the numerical power factor is not
ER = IR = 2 X 75 = 150 volts always sufficient. For example, many dc-
EXL = IX L = 2 X 100 = 200 volts
e = arctan XR or to-ac power inverters can safely operate
loads having a large net reactance of one
X The simple arithmetical sum of these two sign, by only a small reactance of the op-
e= arcsin Z or drops, 350 volts, is greater than the posite sign.
R applied voltage because the two voltages
e= arccos Z are 90 degrees out of phase. Their actual Reactance and Complex Waves
resultant, when phase is taken into It was pointed out earlier in this chapter
Inductive reactances are assigned positive account, is that a complex wave (a "nonsinusoidal"
wave) can be resolved into a fundamental
X values which lead to positive values of
e. Conversely, capacitive reactance causes ..J. (150)2+(200)2 = 250 volts frequency and a series of harmonic
frequencies. When such a complex voltage
e to be negative. Since the current in an in-
wave is applied to a circuit containing
ductor lags the applied voltage (voltage Power Factor
reactance, the current through the circuit
leads the current), defining e as the angle In the circuit of Fig. 40 an applied emf will not have the same wave shape as the
by which the voltage at the terminals of of 250 volts results in a current of 2
applied voltage. This is because the
the complex impedance leads the current amperes, giving an apparent power of 250 reactance of an inductor and capacitor
makes for consistency - a negative value X 2 = 500 watts. However, only the
depend upon the applied frequency. For
of e indicates the angle by which the resistance actually consumes power. The the second-harmonic component of a
voltage lags the current. Note that the arc power in the resistance is',
complex wave, the reactance of, the
cosine formula will not produce a sign for inductor is twice and the reactance of the
the phase angle unless the sign of the reac- P = PR .= (2)2 X 75 = 300 watts
capacitor one-half their respective values
tance is known. at the fundamental frequency; for the
The ratio of the power consumed to the third harmonic the inductor reactance is
Ohm's Law for Impedance apparent power is called the power factor three times and the capacitor reactance
Ohm's Law can be applied to circuits of. the circuit, and in this example the one-third, and so on. Thus the circuit
containing impedance just as readily as to . power factor would be 300/500 = 0.6. impedance is different for each harmonic
circuits having resistance or' reactance Power factor is frequently expressed as a component.
only. The formulas are percentage; in this case, it would be 60 Just what happens to the current wave
percent. shape depends upon the values of
E "Real" or dissipated power is measured
=7: resistance and reactance involved and
in watts; apparent power, to distinguish it how the circuit is arranged. In a simple
E = IZ from real power, is measured in volt- circuit' with resistance and inductive
amperes. It is simply the product of volts reactance in series, the amplitudes of the
E and amperes and has no direct relation-
Z = I" ship to the power actually used up or dis-
harmonic currents will be reduced because
. the inductive reactance increases in pro-
where E = emf in volts , sipated unless the power factor of the cir- portion to frequency. When capacitance
I = current in amperes cuit is known. The power factor of a pure- and resistance are in series, the harmonic
Z = impedance in ohms ly resistive circuit is 100 percent or 1, current is likely to be accentuated because
while the power factor of a pure reactance the capacitive reactance becomes. lower as
, Fig. 40 shows a simple circuit consisting is zero. In this illustration, the re~ctive the frequency is raised. When both
of a resistance of 75 ohms and a reactance power is inductive and capacitive reactance are
of 100 ohms in series. From the formula present the shape of the current wave can
previously given, the impedance is VAR = I'X = (2)' X 100 be altered in a variety of ways, depending
• = 400 volt-amperes upon the circuit and the "constants," or
Z = J Rl+XLl the relative values of L, C and R, selected.
An equivalent definition of power fac- This property of nonuniform behavior
= J (75)2+(100)2 = 125 tor is with respect to fundamental and har-
R monics is an extremely useful one. It is the
If the applied voltage is 250, then basis of "filtering," or the suppression of
Z
undesired frequencies in favor of a single
E 250 or cos e. Since power factor is always desired frequency or group of such
= Z = 125 = 2 amperes rendered as a positive number, the value frequencies.
Transformers
Two coils having IlJutual inductance without direct connection, and in the since no voltage will be induced in the
constitute a transformer. The coil con- process can be readily changed from one secondary if the magnetic field is not
nected to the source of energy is called the voltage level to another. Thus, if a device changing. If dc is applied to the primary
primary coil, and the other is called the to be operated requires, for example, 7 of a transformer, a voltage will be induced
secondary coil. volts ac and only a 44O-volt source is in the secondary only at the instant of
The usefulness of the transformer lies in available, a transformer .can be used to closing or opening the primary circuit,
the fact that electrical energy can be change the source voltage to that required. since it is only at these times that the field
transferred from one circuit to another A transformer can be used only with ac, is changing.
2·23 Chapter 2
Although the secondary voltage is higher
rs = ( :: ) Ep = 2,:> X 117 = 7 X 117 than the primary voltage, the secondary
IRON CORE
current is lower than the primary current,
= 819 volts and by the same ratio.
PRIMARY SECON-
DARY Power Relationships; Efficiency
Also, if an emf of 819 volts is applied to
the 28oo-turn winding (which then be- A transformer cannot create power; it
comes the primary) the output voltage can only transfer it and change the emf.
from the 4oo-turn winding will be 117 Hence, the power taken from the secon-
volts. dary cannot excee61 that taken by the pri-
Either winding of a transformer can be mary from the source of applied. emf.
T A
which are wound on ferromagnetic core
materials (ferrite and powdered iron). As
is the case with stacked cores made of
laminations in the classic I and E shapes,
Fig. 47 - A bc·band ferrite rod loop antenna is
the core material has a specific permeabi- at the top of the picture (J, W. Miller Co.) A
lity factor which determines the inductance blank ferrite rod is seen at the center and a
of the windings versus the number of wire flat bc·band ferrite loop antenna is in the lower
turns used. Both ferrite and powdered- foreground.
iron materials are manufactured with a
wide range of Ui (initial permeability)
LOAD characteristics. The value chosen by the
designer will depend upon the intended , type of core is the self-shielding charac-
Fig. 44 - The autotransformer is based on the operating frequency and the desired band- teristic. Another feature is the compact-
transformer principle, but uses only one width of a given broadband transformer. ness of a transformer or inductor, which
winding. The line and load currents in the is possible when using a toroidal format.
common winding (A) flow in opposite direc· Core-Types in Common Use Therefore, toroids are excellent not only
tions, so that the resultant current is the dif·
ference between them. The voltage across A is For use in radio-frequency circuits in dc-to~dc converters, but at audio and
proportional to the turns ratio. especially, a suitable core type must be radio frequencies up to at least 1000 MHz,
chosen to provide the Q required by the assuming the proper core material is selected
designer. The wrong core material destroys for the range of frequencies over which the
magnetic path as continuous as possible the Q of an rf type of inductor. device must operate. Toroid cores are
and thus reduce' flux leakage. Toroid cores are useful from a few available from micro-miniature sizes well
The number of turns required in the hundred hertz well into the uhf spectrum. up to several inches in diameter. The latter
primary for a given applied emf is deter- Tape-wound steel cores are employed in can be used, as one example, to build a 20-
mined by the size, shape and type of core some types. of power supplies - notably kW balun for use in antenna systems.
material used, and the frequency. The dc-to-dc converters. The toroid core is Another form taken in ferromagnetic
number of turns required is inversely pro- doughnut shaped, hence the name toroid transformers and inductors is the "pot-
portional to the cross-sectional area of the (Fig. 45). The principal advantage to this core" or "cup-core" device. Unlike the
Electrical Laws and Circuits 2-28
toroid, which has the' winding over the L1/L2 and L3/L41in8S must be 3 x R1
dt:~-z
outer surface of the core material, the pot- (30 ohms in this example).
core winding is inside the ferromagnetic
material (Fig. 46). There are two cup-
~~
shaped halves to the assembly, both made
of ferrite or powdered iron, which are
connected tightly together by means of a
screw which is passed through a 'Center
hole. The wire for the assembly is wound
on an insulating bobbin which fits inside
the two halves of the pot-core unit. The
·
advantage to this type of construction is
. that the core permeability can be chosen
to ensure a minimum number of wire TliPE BOARD
9:1 UNBALANCED TO
turns for a given value of inductance. This REMOVE
UNBALANCED TRAN.
88
COPPER
reduces the wire resistance and increases
the Q as opposed to an equivalent induc-
tance which is wound on a core that has
HI-Z
END IOr@1 !i 3
"'C) tJ-'
least four times the impedance the'
loop antennas found in portable radios winding is designed to look into. There-
and direction finders. fore, a transformer which has a 300-ohm
Core Size primary and a 50-ohm secondary load
should have winding reactances (XL) of at
The cross-sectional area of ferrom'agne- 2 least 1200 ohms and 200 ohms, respec-
tic core is chosen to prevent saturation CONVENTIONAL BROADBAND tively. The windings, for all practical pur-
from the load ~een by the transformer. TRANSFORMER
(A)
poses, can be regarded as rf chokes, and
This means that the proper thickness and the same rules apply. The permeability of
diameter are essential parameters to con- the core material plays a vital role in de-
sider. For a specific core the maximum
operational ac excitation can be dete-
mined by l L1
signing a good broadband transformer.
The performance of the transformer at the
low-frequency end of the operating range
depends on the permeability. That is, the
E rms X 10'
Bop (ac) = 4.44 fNpAe (gauss) J lAe (effective permeability) must be high
enough in value to'provide ample winding
where Ae= equivalent area of the mag- L1-TO OCCUpy ENTIRE CORE reactance at the low end of the operating
L2-WOUND OVER L1 WINtlING
netic path in cm' range. As the operating frequency is in-
(B)
E rms = applied voltage creased, . the effects of the core tend to
Np = number of core turns disappear progressively until there are
f = operating frequency in Hz Fig. 49 - Another conventional transformer. scarcely any core effects at the upper limit
Primary and secondary windings are wound . ofthe operating range. For this reason it is
Bop == flux density in gauss over the outer surface of a toroid core.
common to find a very low frequency core
The foregoing equation is applicable to material utilized in a transformer that is
inductors which do not have dc flowing in Types of Transformers contained in a broadband circuit which
the winding along with ac. When both ac The most common ferromagnetic trims- reaches well into the upper hf region, or
and dc flows • formers used in amateur radio work are even into the vhf spectrum. By way of
E rms X 10' the narrow-band, broadband, conven- simple explanation, at high frequency the
BOp(total) = + tional and transmission-line varieties. low-frequency core material becomes in-
4.444 fNJ?A e Narrow-band transformers are used when efficient and tends to vanish electrically.
where Ide the dccurrent through the selectivity is desired in a tuned circuit, This desirable trait makes possible the use
winding such as an audio peaking or notching of ferromagnetics in broadband applica-
AL the manufacturer's index circuit, a resonator in an rf filter, or a tions.
for the core b~ing used tuned circuit associated with an rf Conventional irans/ormers are those that
a~plifier. Broadband transformers are are wound in the same manner as a power
The latter can be obtained' for the core in employed in circuits which must have transformer. That is, each winding is
use by consulting the manufacturer's data uniform response over a substantial made from a separate length of wire, with
sheet. spread offrequency, as in a 2- to 30-MHz one winding being placed over the
2-27 Chapter 2
\,
Table e
Powdered.Jron Toroidal Corea - At. Valuea (uH/100 tuma)
41·Mlx 3·Mix 15·Mlx I·Mlx 2·Mix 6-Mlx ID-Mix 12·Mix D-Mix
Green Grey Rd& Wh Blbe Red Yellow Black Gn & Wh Tan
Core Il = 75 Il =
35 Il = 25 I' = 20 I' = 10 Il = 8 I' = 6 I' = 3 I' = 1
Size 0.05-0.5 MHz 0.1·2 MHz 0.5·5 MHz 1·30 MHz 10·90 MHz 60·150 MHz 100·200 MHz ~50·300 MHz
Wire Size T-200 T·130 T-I06 T-94 T·80 T·68 T·50 T-3l T-25 T·12
10 33 20 12 12 10 6 4 1
12 43 25 16 . 16 14 9 6 3
14 54 32 21 21 /18 13 8 5 1
16 69 41 28 28 24 17 13 7 2
18 88 53 37 37 32 23 18 10 4 1
20 111 67 47 47 41 29 23 14 6 1
22 140 86 60 60 53 38 30 19 9 2
24' 177 109 77 77 67 49 39 25 13 4
26 223 137 97 97 85 63 50 33 17 7
28 281 173 123 123 108 80 64 42 23 9
30 355 217 154 154 136 101 81 54 29 13
32 439 272 194 194 171 .127 103 68 38 17
34 557 346 247 247 218 162 132 88 49 23
36 683 424 304 304 268 199 162 108 62 30
38 875 544 389 389 344 256 209 140 80 39
40 1103 687 492 492 434 324 264 178 102 51
Phyaical Dimenaiona
Cross Cross
Outer Inner Sect. Mean Outer Inner Sect. Mean
Core Dia. Dia. Height Area Length Core Dia. Dia. Height Area Length
Size (in.) (in.) (in.) cm' cm Size (in.) (in.) (in.) cm' cm
T-2oo 2.000 1.250 0.550 1.330 12.97 T- 50 0.500 0.303 0.190 0.121 3.20
T-184 1.840 0.950 0.710 2.040 '11.12 T- 44 0.440 0.229 0.159 0.107 2.67
T-157 1.570 0.950 0.570 1.140 10.05 T· 37 0.375 0.205 0.128 0.070 2.32
T·l30 1.300 0.780 0.437 0.733 8.29 T- 30 0.307 0.151 0.128 0.065 1.83
T·l06 1.060 0.560 0.437 0.706 6.47 T- 25 0.255 0.120 0.096 0.042 1.50
T- 94 0.942 0.560 0.312 0.385 6.00 T~ 20 0.200 0.088 0.067 0.034 1.15
T- 80 0.795 0.495 '0.250 0.242 5.15 T- 16 0.160 0.078 0.060 0.016 0.75
T· 68 0.690 0.370 0.190 0.196 4.24 T- 12 0.125 0.062 0.050 0.010 0.74
Inches x 25.4 = mm. Courtesy of Amidon Assoc., N. Hollywood, CA 91607 and Micrometals, Inc.
previous one with suitable insulation in known as a triftlar one, and so forth (Fig. reference on the subject is by Ruthroff.2
between.(Figs. 48 and 49). A transmission- 50). It can be argued that a transmission-
line transformer is, conversely, one that line transformer is more efficient than a Ferrite Beads
uses windings which are configured to conventional one, but in practice it is diffi- Another form of toroidal inductor' is
simulate a piece of transmission line of a cult to observe a significant difference in the ferrite bead. This component is
specific impedance. This can be achieved the performance characteristics. An interest- available in various II, i values and sizes,
by twisting the wires together a given ing technical paper on the subject of toroi- but most beads are less than 0.25-jnch
number of times per inch, or by laying the dal broadband transformers was pub- (6.3-mm) diameter. Ferrite beads are used
wires on the core (adjacent to one another) lished by Sevick, W2FMI. 1 The classic principally as vhf/uhf parasitic sup-
at a distance apart which provides a two-
wir.e line impedance of a particular value.
In some applications these windings are 'Sevick, "Simple Broadband Matching Networks," 'Ruthroff, "Some Broadband Transformers," Proc.
called biftlar. A three-wire winding is QST, January 1976. IRE, Vol. 47, August 1959, p. 137.
/-Ii for that core" is 10. 10 - Inner diameter (inches) Ie - Effective magnetic path length (inches)
The required number of wire turns for a l'igt - Heigl:lt (inches) Ve - Effective magnetic volume (in)"
specified inductance on a given type of Aw - Total window' area (in)" . As - Surface area exposed for cooling (in)"
core can be determined by .
Inches X 25.4 = mm. Courtesy of Amidon Assoc .• N. Hollywood. CA 91607
Turns = 100 V desired L (~) + AL
where AL is obtained from Table 6. The
table also indicates how many turns of a
particular wire gauge can be close wound
Turns = 1000 V desired L (mH) + AL This will prevent the rough edges of the
core from abrading the enameled wire.
to fill a specified core. For example, a
where the AL for a specific core can be The inductance of a toroidal coil with
T-68 core will contain 49 turns of no. 24
enameled wire, 101 turns of no. 30 taken from Table 7. Thus, if one required known AL is
enameled wire, and so on. Generally a l-mH inductor and chose a no.
speaking, the larger the wire g;:tuge the
FT-82-43 toroid core, the number of turns
would be
L = AL(~)l
100
higher the unloaded Q of the toroidal
inductor. The inductance values are based' Land AL must be in the same units.
on the winding covering the entire
Turns = 1000 V I + 557 Tables 8 and 9 cross-reference the ferrite
circumference of the ,core. When there is
space between the turns of wire, some
= 1000 V 0.001795 toroidal cores offered by several sources.
control' over the net inductance can be = 1000 X 0.0424 = 42.4 turns Checking RF Toroidal Devices
effected by compressing the turns or The equations given previously will
spreading them. The ,inductance will For an FT-82 size core no. 22 enameled provide the number of wire turns needed
increase if compression is used and will wire would be suitable as indicated in for a particular inductance, plus or minus
decrease when the turns are spread farther Table 6 (using the T -80 core size as the 10 percent. However, slight variations in
apart. nearest one to an FT-82). If the toroid core permeability may exist from one
Table 7 contains data for ferrite cores. core has rough edges (un tumbled), it is production run to another .•Therefore, for
The number of turns for a specified suggested that insulating tape (3M glass circuits which require exact values of
inductance in mH versus the AL can be epoxy tape or Mylar tape) be wrapped inductance it is necessary to check the
determined by through the core before the wire is added. toroid winding by means of an RCL
2-29 ,Chapter 2
Table 8 L UNOER TEST '
I. !B·(t)i!
(inches)
00 10 Thickness Amidon Fair·Rite Indiana General Ferroxcube Magnetics, Inc. ~TWO TURN _ _ _ _ _--..,
0.100 0.050 0.050 40200TC / LINK ~
0.100 0.070 0.030 701 F426·1 ,
0.1550.088 - - 801 F2062·t 40502
0.1900.090 0.050 213T050
0.230 0.120 0.060 FT·23 101 F303·1 1041T060 40601
DIP METER
0.2300.1200.120 901
0.3000.1250.188 ~ F867·1 40705
0.375 0.187 0.125 FT·37 201 F625·9 266T125 41003 Fig. 51A - Method for checking the indue·
0.500 0.2810.188 FT·50 301 768T188 tance of a toroid winding by means of a dip
0.500 0.312 0.250 1101 F627·8 41306 meter, known capacitance value and a
0.500 0.312 0.500 1901 calibrated receiver. The self·shielding
0.825 0.520 0.250 FT-82 601 properties of a toroidal inductor prevent dip·
0.825 0.520 0.468 501 meter readings, when the instrument is coupled
0.870 0.500 0.250 401 directly to the toroid. Sampling is done by
0.870 0.540 0.250 1801 F624·19 846T250 42206 means of a coupling link as illustrated.
1.000 0.50Q 0.250 1501 F2070·1 42507
1.000 0.610 0.250 1301
1.1420.7480.295 FT·114 1001 K300502 42908 Fig. 516 - (A) Illustration of a homemade
1.225 0.750 0.312 1601 winding shuttle for toroids. The wire is stored
1.2500.7500.375 1701 F626·12 on the shuttle and the shuttle is pjlssed
1.417 0.905 0.591 K300501 through the center hole of tile toroid, again
1.417 0.905 0.394 K300500 and again, until the required number of turns is
1.5000.750 0.500 528T500 43813 in place. (6) It is best to leave a 30· gap
2.0001.2500.750 400T750 between the ends of the toroid winding. This
2.900 1.530 0.500 144T500 will reduce the distributed capacitance con·
3.375 1.925 0.500 F1707·15 siderably. (C) Edgewise view of a toroid core,
3.500 2.000 0.500 F1707·1 Illustrating the method for counting the turns
5.8352.50 0.625 F1824·1 accurately. (0) :The low'impedance winding of a
toroidal transformer is usually wound over the
large winding, as shown. For narrow·band
applications the link should be wound over the
Table 9 cold end of the main winding (see text).
SHUTTLE
fiT
"'0
16
Amidon Fair·Rite Indiana General
03
Ferroxcube Magnetics, Inc,c TOROID
20 68
40 FT -63 63,67 02 WIRE (Al
100 65
125 FT - 61 61 01 4C4
175 62
250 FT - 64 64
300 83
375 31
400 G
750 303 A
800 33
850 43 H
950 FT - 43 TC·3
1400 C END
1200 34 VIEW
1500 TC·7
1800 FT -77 77 369
2000 FT -72 72 TC·9 S,V,O
2200 05
2300 367 G
2500 FT -73 73 TC·12
2700 3E'(3C8)
3000 05P 3C5 F
4700 06
5000 FT - 75' 75 3E2A J
10,000 W
12,500 3E3
1 XL
Xc = 2ITfe and L{j.iH) = 2nf
bridge or an RX meter. If these in- illustration is necessary because the toroid
struments are not available, close approxi- has a self-shielding characteristic. The where Xc is the reactance of the known
mations can be 'had by, using a dip latter makes it difficult, and often capacitor value, f is in MHz and C is in
meter, standard capacitor (known value, impossible, to secure a dip in the meter /.£F. Using an example, where fis 3.5 MHz
stable type, such as a silver mica) and a reading when coupling the instrument (as noted on a dip meter) and C is 100 PF,
calibrated receiver against which to check directly to the toroidal inductor or L is determined by
the dipper frequency. Fig. 51A shows how transformer. The inductance can be
to couple a dip meter to a completed determined by XL since XL = Xc at 1
toroid for testing. The coupling link in the. resonance. Therefore, Xc = 6.28 X 3.5 X 0.0001 = 455 ohms
The Decibel
It is useful to appraise signal strengths in Common logarithms (base 10) are used. 20
terms of relative loudness as registered by 18
1/
the ear. For example, if a person estimates Voltage and Current Ratios V
16
that the signal is "twice as loud" when the Note that the decibel is based on power
transmitter power is increased from 10 ratios. Voltage or current ratios can be /
watts to 40 watts, he will also estimate that used, but only when the impedance is the
14
VOLTAGE
CURRENT
0V
'" 12
a 400-watt signal is twice as loup as a same for both values of voltage, or cur-
..J
"'
~10
/
100-watt signal: The human ear has a rent. The gain of an amplifier cannot be ""'o V PO VER
VV
8
logarithmic response. expressed correctly in dB if it is based on /
This fact is the basis for the use of the the ratio of the output voltage to the input 6
1/ ..-/
relative-power unit called the decibel (dB). voltage unless both voltages are measured 4
/ ... V
A change of one decibel in the power level across the same value of impedance. ../
V
2 /
is just detectable as a~·ange in loudness When the impedance at both points of oV
under ideal conditions. The number of measurement is the same, the following 1 1.~ 2 2.~ 3 4 ~ 6 7 8 910
RATIO
decibels corresponding to a given power formula may be used for voltage or cur-
ratio is given by the following formula: rent ratios: Fig. 52 - Decibel chart for power, voltage and
current ratios for power ratios of 1:1 and 10:1.
In determining decibels for current or voltage
P2 ratios the currents (or voltages) being com·
dB = 10 log ~ dB = 20 log VV2I 12
or 20 log -I-
I.
pared must be referred to the same value of
impedance.
2·31 Chapter 2
Table 10
Decibel equivalents E; I. and P ratios
where V =voltage relative-reading instrument on most Calibration of this kind is practical if the
I =
current amateur receivers. However, during WW receiver is designed for a single amateur
The S unit and dB are used as references II at least one receiver manufacturer used band, but multiband receivers exhibit dif-
on receiver signal-strength meters. No 50/AV for S9 and each S unit below S9 was ferent overall sensitivities for the various
particular standard has been adopted by supposed to be equivalent to 6 dB. The bands, rendering an S meter grossly inac-
the industry at this time, as an S meter is a meter divisions above S9 were in dB. curate except for a single band in which
Electrical Laws and Circuits 2-32
.
calibration may have been~ .attempted. , with a minus sign. Thus + 6 dB means ratios. For example, a power ratio of 2.5
that the power has been mUltiplied by 4, is 4 dB (from the chart). A power ratio of
Decibel Chart I while - 6 dB means that the power has ( 10 times 2.5, or 25, is 14 dB (10 + 4), and
The two formulas are shown graphical- been divided by 4. a power ratio of 100 times 2.5, or 250 is 24
ly in Fig. 52 for ratios from I to 10. Gains The chart may be used for other ratios \ dB (20 + 4). A voltage or current ratio of
(increases) expressed in decibels may be by adding (or subtracting, if a loss) 10 dB 4 is 12 dB, a voltage or current ratio of 40 ,
added arithmetically; losses (decreases) each time the ratio scale. is multiplied by is 32 dB (20 + 12), and one of 400 is 52
may be subtracted. A power decrease is 10, for power ratios; or by adding (or sub- dB (40 + 12). Table 10 provides an easy
indicated by prefixing the decibel figure tracting) 20 dB each time the scale is reference for· voltage power and current
mUltiplied by 10 for voltage ot current ratios versus dB.
--
0.2
6.28 X 13.23
£ ~
~~
"....,-
L The formula for resonant frequency is not o~
-.20 -10 0 +10 +20
affected by resistance in the circuit. PERCENT CHA~GE FROM
EAC RESONANT FREQUENCY
Resonance Curves
LJ
Fig. 53 - A series circuit containing L, C and
If a plot is drawn on the current flowing'
in the circuit of Fig. 53 as the frequency is
varied (the applied voltage being con-
stant) it would look like one of the curves
in Fig. 54. The shape of the resonance
curve at frequencies near resonance is
Fig. 54 - Current in a series-resonant circuit
with various values of series -resistance. The
valu~s are arbitrary and would not apply at all
circuits, but represent a typical case. It is
assumed that the reactances (at the resonant
frequency) are 1000 ohms. N.ote that at fre-
quencies m9re than plus or minus 10 percent
R is "resonant" at the applied frequency when away from the resonant frequency, the current
the reactance of C is equal to the reactance of determined by the ratio of reactance to is substantially unaffected by the resistance in
.L. resistance. the circuit.
2-33 Chapter 2
· '
w
0.8
,0-100
with these values the impedance does not An important application of the parallel-
"~O.6 have its maximum possible value. Another resonant circuit is as .an impedance-
.
o
w
!
...-0"50
set of values for Land C will make the
parallel impedance a maximum, but this
matching device in the output circuit of a
vacuum-tube rf power amplifier. There
wO.4
>
maximum value is not a pure resistance. is an optimum value of load resistance for
Ei Either condition could be called "reso- each type of tube or transistor and set of
..J ~'20
W
0: 0.2 I nance," so with low-Q circuits it is neces- operating conditions. However, the resis-
J/.. ~ sary to distinguish between maximum im-
pedance and resistive impedance parallel
tance of the load to which the active
device is to deliver power usually is
o I--- ~ ~ resonarice. The difference between these L considerably lower than the value required
-20 -10 0 +10 +20
PER CENT CHANGE FROM RESONANT FREOUENCY and C values and the equal reactances of a for proper device operation. To transform
series-resonant circuit is appreciable when the actual load resistance to the desired
Fig. 58- Relative impedance of parallel- the Q is in the vicinity of 5, and becomes value the load may be tapped across part
resonant circuits with different as. These curves more marked with still lower Q values. of the coil, as shown in Fig. 59B. This
are sim ilar to those in Fig. 55 for current in a
series-resonant circuit. The effect of a bn is equivalent to connecting a higher value
impedance is most marked near the resonant
Q 0/ Loaded Circuits of load resistance across the whole circuit,
frequency. In many applications of resonant and is similar in principle to impedance
circuits the only power lost is that transformation with an iron-core trans-
dissipated in the resistance of the circuit former. In high-frequency resonant cit-
itself. At frequencies below 30 MHz most cuits the impedance ratio does not vary
of this resistance is in the coil. Within exactly as the square of the turns ratio,
limits, increasing the number of turns in because all the magnetic flux lines do not
the coil increases the reactance faster than cut every turn of the coil. A desired re-
it raises the resistance, so coils for circuits flected impedance usually must be ob-
in which the Q must be high are made tained by experimental adjustment.
(A) (8) with relatively large inductance for the When the load resistance has a very low
frequency. value (say below 100 ohms) it may be
However, when the circuit delivers connected in series in the resonant circuit
Fig. 59 - The equivalent circuit of a resonant (as in Fig. 57A, for example), in which
circuit delivering power to a load. The resistor R
energy to 'a load (as in the case of the
represents the load resistance. At B the load is resonant circuits used in transmitters) the case it is transformed to an equivalent
tapped across part of L, by which transformer energy consumed in the circuit itself is parallel impedance as previously de-
action is equivalent to using a higher load usually . negligible compared with that scribed. If the Q is at least 10, the
resistance across ttie whole circuit. consumed by the load. The equivalent of equivalent parallel impedance is
such a circuit is shown in Fig. 59A, where
the parallel resistor represents the load to ZR =R
r
which power is delivered. If the power
At frequencies off resonance the impe- dissipated in the load is at least ten times
dance is no longer purely resistive because where ZR = resistive parallel impedance
as great as the power lost in the inductor at resonance
the inductive and capacitive currents are and capacitor, the parallel impedance of
not. equal. The off-resonant impedance X = reactance (in ohms) of either
the resonant circuit itself will be so high
- therefore is complex, and is lower than the compared with the resistance of the load the c9il or capacitor
resonant impedance for the reasons pre- that for all practical purposes the im- r = load resistance inserted in
viously outlined. series
pedance of the combined circuit is equal
The higher the circuit Q, the higher the to the load resistance. Under these If the Q is lower than 10. the reactance
parallel impedance. Curves showing the conditions the Q of a parallel resonant will have to be adjusted somewhat, for the
variation of impedance (with frequency) of circuit loaded by a resistive impedance is reasons given in the discussion of low-Q
a parallel circuit have just the same'shape circuits, to obtain a resistive impedance of
as the curves showing the variation of. the desired value.
R
current with frequency in a series circuit. Q=X While the circuit shown in Fig. 59B will
Fig. 58 is a set of such curves. A set of usually provide an impedance step-up as
curves showing the relative response as a with an iron-core transformer, the net-
function of the departure from the where R = parallel load resistance (ohms) work has some serious disadvantages for
resonant frequency would be similar to X = reactance (ohms) some applications. For instance, the
Fig. 55. The - 3 dB bandwidth (band- Example: A resistive load of 3000 ohms common connection provides no dc
width at 0.707 relative response) is given is connected across a resonant circuit in isolation and the common ground is
by which the inductive and capacitive reac- sometimes troublesome in regards to
tances are each 250 ohms. The circuit Q is ground-loop currents. Consequently, a
Bandwidth -3 dB = fo IQ then network in which only mutual magnetic
coupling is employed is usually preferable.
where fo is the resonant frequency and Q R 3000 H~ever, no impedance step-up will
the circuit Q. It is also called the Q = X = '250 = 12 result unless the two coils are coupled
"half-power" bandwidth, for ease of tightly enough. The equivalent resistance
recollection. The "effective" Q of a circuit loaded by seen at the input of the network will
. Parallel Resonance in Low-Q Circuits a parallel resistance becomes higher when always be lower regardless of the turns
the reactances are decreased: A circuit ratio employed. However, such networks
The preceding discussion is accurate for loaded with a relatively low resistance (a are still useful in impedance-transformation
Qs of 10 or more. When tile Q is b(:low 10, few thousand ohms) must have low- applications if the appropriate capacitive
resonance in a parallel circuit having reactance elements Jlarge capacitance and elements are used. A more detailed
2-35 Chapter 2
treatment of matching, networks and portant simplifications in computations COUPLING
similar devices will be taken up in the next involved in coupled networks. The as- NETWORK
~lOIOg(~~n)
that the network can be reduced to a source and load resistance, it is often
combination of series and shunt elements Attn= = possible to combine the processes of
consisting only of inductors and capaci"
f_ 4RinRp I
tors as indicated by the circuit shown in -10 log lSRp + Rin)2 + (Xp + X )2
in
J filtering and matching into one network.
BlI BC2 G2
RT" Rl + R2
Xro Xl. - (XC. + XC:t)
(A) (B)
Xl. GT' G1 + G2
ST- BCt + BC2-BL1
Fig. 62 - Application of conversion formulas
can be used to transform a shunt conductance
o>--___T..J XC2 and susceptance to a series equivalent circuit
A. The converse is illustrated at B.
Fig. 61 - Resistances and reactances add in series circuits while conductances and suscep·
tances add in paralle.1 circuits. (Formulas shown are for numerical values of X and B.).
Gr
R = GTl + BTl
constant with frequency. In the case of the formulas for the respective reciprocal
many circuits, this assumption is not true. entities are
However, if the value of Rs and Xs at any -BT
particular frequency is kno,wn, the at-
X = GTl + BTl
tenuation of the filter can be determined
even though it is improperly terminated.
Unfortunately, while the solution to These relations are illustrated in Fig. 62A
any ladder problem is possible from a Bc = 2:rfC and Fig. 62B respectively. While the
theoretical standpoint; practical difficul- derivation of the mathematical expres-
ties are encountered as the network com- and are defined as susceptances. In a sions will not be given,the importance of
plexity increases. Many computations parallel combination of conductances and the sign change cannot be stressed too
to a high degree of "accuracy may be susceptances, the total Gonductance is the highly. Solving network problems with a
tequired, making the process a: tedious sum of the individual conductances, and calculator is merely a matter of book-
one. Consequently, the availability of a the total susceptances is the sum of the keeping, and failure to take the sign
calculator or similar computing device is individual susceptances, taking the res-' change associated with the transformed
recommended. The approach used here is pective signs of the latter into account. A reactance and susceptance is the most
adapted readily to any calculating method comparison between the way resistance common source of error.
including the use of an inexpensive pocket and reactance add and the manner in
calculator. A Sample Problem
which conductance and susceptance add is
shown in the example of Fig. 61. An entity The following example illustrates the
Susceptance and Admittance called admittance can be defined in terms manner in which the foregoing theory can
The respective reactances of an in- of the total conductance and total be applied to a practical,problem. A filter
ductor and a capacitor are given by susceptance by the formula with the schematic diagram shown in Fig.
63A is supposed to have an insertion loss
at 6 MHz of 3 dB when connected
2:rtL between a 52-ohm load and a source with
a 52-ohm primary resistance (both Xp and
Xs are zero). Since this is a case where the
-1 effective attenuation is equal to the
Xc = 2:rfC
and is often denoted by the symbol Y. If insertion loss, the previous formula for
the impedance of a circuit is known, the effective attenuation applies. Therefore, it
i
In a simple series circuit, the total admittance is just the reciprocal. Like_ is required to find Rin and -"in.
resistance is just the sum of the individual wise, if the admittance of a circuit is Starting at the output, the values for the
resistances in the network and the total known, the impedance is the reciprocal ~f conductance and susceptance of the
reactance is the sum of the reactances. the admittance. However, conductance, parallel RC circuit must be determined
However, it is important to note the sign reactance, resistance and susceptance are . first. The conductance is just the recip-
of the reactance. Since capacitive reac- not so si.mply related. ·If the total rocal of 52 ohms and the previous formula
tance is negative and inductive reactance resistance and total reactance of a series for capacitive susceptance gives the value
is positive, it is possible that the sum of . circuit are known, the conductance and shown in parentheses in Fig. 63A. (The
the reactances might be zef(~ even though susceptance of the circuit are related to upside-down n is the symbol for mho:)
the individual reactances are not zero. In a the latter by the formulas The next step is to apply the formulas
series circuit, it will be recalled that the for resistance and reactance in terms
network is said to be resonant at the RT of the conductance and susceptance and
f requency wereh ·
the reactances canceI. G = RTl + XTl the results give a 26-01nn resistance in
A complementary condition exists in a series with a - 26-ohm capacitive reactance
parallel combination of circuit elements -XT as indicated in Fig. 63B. The reactance of
and it is convenient tQ introduce the B = the inductor can now be added to give a
concepts of admittance, conductance and total reactance of 78.01 ohms. The
susceptance. In the case of a simple conductance· and susceptance formulas
resistance; .the conductance is just the On the other hand, if the total con- can now be applied and the results of both
reciproca~. That is, the conductance of a ductance and total 'susceptance of a of these operations is shown in Fig. 63C.
50-ohm resistance is 1/50 or 2 X 10- 2 • parallel c,ombination are known, the Finally, adding the susceptance of the
The reciprocal unit of the ohm is the mho. equivalent resistance and reactance can be 51O.l-pF capacitor (Fig. 63D) gives the
For simple inductances and capacitances; found from the formulas circuit at Fig. 63A and applying the
2-37 Chapter 2
(104.f).) tables especially, to present all the circuit
2.759"H
components for a number of designs at
some convenient frequency. Translating
the design to some desired frequency is .
simply accomplished by multiplying all
the components by some constant factor.
The most common frequency used is the (Al
, (Al value of f such that 2nf is equal to 1.0.
78.01.f).
This is sometimes called a radian fre-
quency of 1.0 and corresponds to 0.1592
Hz. To change a "one-radian" filter to a
G-O.003845V
B-O.01154V new frequency fo (in Hz), all that is
necessary is to multiply the inductances
and capacitances by 0.1592/fo. te te
(81 (el In a similar manner, if one resistance
(or conductance) is multiplied by some (81
factor n, all the other resistances (or
rnru)
O.01923V O.003845V
RIN-51.98.f).
XIN- -104 f\.
conductances) and reactances (or sus- Fig. 64 - Ideal filter response curves are
ceptances) must be multiplied by the same shown at A and characteristics' of practical
factor in order to preserve the network filters are shown at B.
characteristics. For instance, if the secon-
dary resistance, Rs is multiplied by n,
all circuit inductances must be multiplied
by n and the circuit capacitances divided frequencies while passing all others.) And
(01 (El by n (since capacitive reactance varies as high-pass filters reject all frequencies be-
Rp
52f\. the inverse of C). If, in addition to low some cutoff frequency.
converting the filter of Fig. 63A to 7 MHz The attenuation shapes shown in Fig.
Ef1Al~:~4f\.
'\... RIN 3.01dB
from 6 MHz, it was also desired to change 64A are ideal and can only be approached
the impedance leYrel from 52 to 600 ohms, or approximated in practice. For instance,
51. 98 f\.
the inductance would have to be multi- if the filter in the preceding problem was
plied by (6/7)(600/52) imd the capac- used for low-pass purposes in an 80-meter
itances by (6/7)(52/600). transmitter to reject harmonics on 40
(Fl
meters, its performance would leave a lot to
Using Filter Tables be desired. While insertion loss at 3.5 MHz
Fig. 63 - Problem illustrating network reduc· In a previous example, it'was indicated was acceptable, it would likely be too high
tion to find·insertion loss. that the frequency' response of a filter at 4.0 MHz and rejection would probably
could be derived by solving for the be inadequate at 7.0 MHz.
insertion loss.of the ladder network for a Fortunately, design formulas exist for
formulas once more gives the value of Rin number of frequencies. The question this type of network and form a class called
and Xin (Fig. 63F). If the latter values are might be asked if the converse is possible. Butterworth filters. The name is' derived
substituted into the effective attenuation That is, given a desired frequency from the shape of the curve for insertion-
formula, the insertion loss and effective response, could a network be found that loss vs. frequency and is sometimes called
attenuation are 3.01 dB, which is very would have this response? The answer is a a maximally flat response. A. formula for
close to the value specified. The reader qualified yes and the technical nomen- the frequency response curve is given by
might verify that the insertion loss is clature for this sort of process is network
0.167,0.37 and 5.5 dB at 3.5, 4.0 and 7.0 synthesis. Frequency responses can be
MHz respectively. If a plot of insertion "cataloged" and, if a suitable one can be
loss versus frequency was constructed it found, the corresponding network ele-
would give the frequency response of the ments can be· determined from an asso-
• filter. ciated table. Filters derived by network where fc = the frequency for an insertion
synthesis and simiktr methods (such as loss of 3.01 dB
Frequency Scaling optimized computer designs) are often k = the number of circuit elements
and Normalized Impedance referred to as "modern filters" even The shape of a Butterworth low-pass filter
Quite often, it is desirable to be able to though the theory has been in existence is shown in the left-hand portion of Fig.
change a coupling network at one for years.' The term is useful in dis- 64B. (Another type that is similar in
frequency and impedance level to another tinguishing such designs from those of an nature, only one that allows some
one. For example, suppose it was desired older approximate method called image- "ripple" in the passband, is also shown in
to move the 3-dB point of the filter in the parameter theory. Fig. 64B. Here, a high-pass characteristic
preceding illustration from 6 to 7 MHz. illustrates a Chebyshev response.)
An examination of the reactance and Butterworth Filters As can be seen from the formula,
susceptance formulas reveals that multi- Filters can be grouped into four general increasing the number of elements will
plying the frequency by some cOQ.stant k categories as illustrateq in Fig. 64A. result in a filter that approaches the
and dividing both the inductance and Low-pass filters have zero insertion loss "ideal" low-pass shape. For instance, a
.capacitance by the same value of k leaves up to some critical frequency (re) or cutoff 20-element filter designed for a 3.ol-dB
the equations unchanged. Thus, if the frequency and then provide high rejection' cutoff frequency of 4.3 MHz, would have
capacitances and inductance in Fig. 63A above this frequency. (The latter condition an insertion loss at 4 MHz of 0.23 dB and
are multiplied by 6/7, all the reactances is indicated by the shaded lines in Fig; 64.) 84.7 dB at 7 MHz. However, practical
and susceptances in the new circuit will Band-pass filters have zero insertion loss difficulties would make such a filter very
now have the same value at 7 MHz that between two cutoff frequencies with high hard to construct. Therefore, some com-
the old one had at 6 MHz. rejection outside of the prescribed "band- promises are always require~ between a
It is common practice with many filter width." (Band-stop filters reject a band of theoretically perfect frequency response
Electrical Laws and Circuits 2-38
and e~se of construction. L2 L4
Element Values
Once ·the number of elements, k, is
determined, the next step is to find the
network configuration corresponding to
k. (Filter tables sometimes have sets of
curves that enable the user to select the
desired frequency response curve rather (A)
than 'use a formula. Once the curve with Ll L3 L5
the fewest number of elements for the
specified passband and stop-band in-
sertion loss is found, the filter is then
fabricated around the corresponding value
of k.) Table II gives normalized dement
values for values of k from I to 10. This
table is for I-ohm source and load
resistance (reactance zero) and a 3.01-dB
cutoff frequency of I radian/second
Fig. 65 - Schematic diagram of a Butterworth low-pass filter. (See Table 11 for element values.)
(0.1592 Hi). There are two possible circuit
configurations and these are shown in Fig.
65. Here, a five-element filter is given as
an example with either a shunt element Then Land C give the actual circuit-
next to the load (Fig. 65A) or a series element values in henrys and farads in
element next to the load (Fig. 65B). Either terms of the prototype element values
filter will have the same response. from Table II.
After the values for the I-ohm, 1- However, the usefulness of the low-pass
~ radian/second "prototype" filter are found, prototype does not end here. If the
the corresponding values for the actual following set of equations is applied to the
frequency/impedance level can be deter- prototype values, circuit elements for a
mined (see the section on frequency and high-pass filter can be obtained. The filter
irripedance scaling). The prototype in- (A)
is shown in Fig. 66A and Fig. 66B which
ductance and capacitance values are correspond to Fig. 65A and Fig. 65B in
C4 C2
multiplied by the ratio (0.1592/fc) where Table 9. The equations for the actual
fc is the actual 3.0I-dB cutoff frequency. high-pass circuit values in terms of the
Next, this number is multiplied by the low-pass prototype are given by
load resistance in the case of an inductor
and divided' by the load resistance if the
element is a capacitance. For instance, the
filter in the preceding example is for a
three-element design (k equal to 3) and the and the frequency response curve can be
reader might verify the values for the (B)
obtained from
compont;nts for an fc of 6 MHz and load
resistance of 52 ohms. Fig. 66 - Network configuration of a Butter-
worth high-pass' filter. The low-pass prototype
High-Pass Butterworth Filters can be transformed as described in the text.
The formulas for change of impedance
and frequency from the l-ohm,l-radianJ For instance, a high-pass filter with three
second prototype to some desired level elements, a 3.Q1-dB fc of 6 MHz and 52 through the use of Table II. Unfor-
can also be conveniently written as ohms, has a CI and C3 of 510 pF and an tunately, the process is not as straightfor-
L2 of 0.6897 uH_ The insertion loss at 3.5 ward as it is for low- and high-pass filters
R I
L = 2"fe Lpr010lype C = 2"f eR Cpr010lype and 7 ,MHz would be 14.21 and 1.45 dB if a practical design is to be obtained. In
essence, a low-pass filter is resonated to
respectively.
where R = the load resistance in ohms some "center frequency" with the 3.01-d.B
fc = the desired 3.Q1-dB· Butterworth Band-Pass Filters cutoff frequency being replaced by the
frequency in Hz Band-pass filte.rs can also be designed filter bandwidth. The ratio of the band-
Table 11
Prototype Butterworth Low·Pass Filters
Fig.65A Ct L2 C3 L4 C5 L6 C7 LB C9 LtO
Lt C2 L3 C4 L5 C6 L7 CB L9 CtO
Fig. 658
k
1 2.0000
2 1.4142 1.4142
3 1.0000 2.0000 1.0000
4 0.7654 1.8478 1.8478 0.7654
5 0.6180 1.6180 2.0000 1.6180 0.6180
6 0.5176 1.4142 1.9319 1.9319 1.4142 0.5176
.., 0.4450 1.2470 1.8019 2.0000 1.8019 1.2470 0.4450
8 0.3902 1.1111 1.6629 1.9616 1.9616 1.6629 1.1111 0.3902
0.3473 1.0000 1.5321 1.8794 2.0000 1.8794 1.5321 1.0000 0.3473
9
0.3129 0.9080 1.4142 1.7820 1.9754 1.9754 1.7820 1.4142 0.9080 0.3129
10
2·39 Chapter 2
width to center frequency must be rela- frequency."
tively large. otherwise component values If the response is plotted against a
tend to become unmanageable. logarithmic frequency scale. the symmetry
While there are many variations of will become apparent. Consequently.
specifying such filters. a most· useful using a logarithmic plot is helpful in
approach is to determine an upper and designing filters of this type.
lower frequency for a given attenuation. Examination of. the component values
The center frequency and bandwidth are reveals that while the filter is practical. it
then given by is a bit untidy from a construction (A)
standpoint. Rather than using a single
340.l-pF capacitor. paralleling a number
fo = ..rq; of smaller valued units would be ad-
visable. Encountering difficulty of this
BW = f2 - f1 sort is typical of most filter designs.
consequently. some tradeoffs between
If the. bandwidth specified is not the performance. complexity and ease of
3.01-dBbandwidth (BWe). the latter can construction are usually required.
be determined from
Coupled Resonators
BW A problem frequently encountered in rf
!O(MHz)
BWc= (8)
circuits is that of a coupled resonator.
Applications include simple filters. oscil-
lator tuned circuits. and even antennas. Fig. 67 -.A Butterworth band-pass filter.
(Capacitance values are in picofarads.).
The circuit shown in Fig. 68A is
in the case of a Butterworth response or illustrative of the basic principles in-
from tables of curves. A is the required volved. A series RLC circuit and the
attenuation at the cutoff frequencies. The external terminals ab are "coupled"
upper and lower cutoff frequencies (feu through a common capacitance. Cm.
Lr
~-r
and fel) are then given by Applying the formulas for conductance
and susceptance in terms of series
- BWc + ..J (BWe)2 + 4f02 reactance and resistance gives
rc1 =
em Rr
2 Rr
Gab = Rr 2+ X2
(A)
t
A somewhat more convenient method Rr
is to pick a 3.ol-dB bandwidth (the wider
the better) around some center frequency
and compute the attenuation at other The significance of these equations can
frequencies of interest by using the be seen with the aid of Fig. 68B. At some
transformation: point. the series inductive reactance will Bell'
cancel the series capacitive reactance (at a II)
[( rof - -f-
fO)foJ point slightly below fo where the con- <!>
BWc ductance curve reaches a peak). Depen-
I
ding upon the value of the coupling I
which can be ·substituted into the insertion- susceptance. Bm. it is possible that a point '\
\
I
I
loss formula or table of curves. can be found where the total input suscep- \ /1
As an example. suppose it is desired to tance is zero. The input conductance at '-'-
2Rr'
build a band-pass filter for the 15-meter this frequency. fo• is then Go' (B)
Novice band in order to eliminate the Since Go is less than the conductance at
possibility of radiation on the 14- and the peak of the curve. I/Go or Ro is going
28-MHz bands. For a starting choice. 16 to be greater than Rr. This effect can be Fig. 68 - A capacit(vely coupled resonator is
and 25 MHz will be picked as the 3.01-dB applied when it is desired to match a shown at A. See text for explanation of figure
shown at B.
points giving a 3-dB bandwidth of 9 MHz. low-value load resistance (such as found
For these two points. fo will be 20 MHz. It in a mobile whip antenna) to a more either a positive or negative reactance will
is common practice to equate the number practical value. Suppose Rr and Cr in Fig .. satisfy the equation for Gab. a positive
of branch elements or filter resonators to 68A are 10 ohms and 21 pF. respectively. value is required to tune out Bern. If the
certain mathematical entities called "poles" and represent the equivalent circuit of a coupling element was a shunt inductor.
and the number of poles is just the value mobile antenna. Find the value of Lr and the total reactance would have to be
of k for purposes of discussion here. For a Cm which will match this antenna to a 52- capacitive or negative in value.) Thus. the
three-pole filter (k of 3). the insertion loss ohm feed line at a frequency of 3900 kHz. required inductance value for Lr will be
will be 12.79 and 11.3 dB at 14 and 28 Substituting the:: foregoing values into the 80.1 uH. In order to obtain a perfect
MHz. respectively. formulas for input conductance gives. • , match. the input susceptance must be zero
Cl. C3 and L2 are then calculated for a and the value of Bern can be found from
1 10
9-MHz low-pass filter and the elements
52 = 102 + X2 20.49
for this filter are resonated to 20 MHz as o = Bern - 102 + (20.49)2
shown in Fig. 67A. The response shape is Solving for X (which is the total series
plotted in Fig. 67B and it appears to be reactance) gi"es a va,lue of 20.49 ohms.
unsymmetrical about fo. In spite of this The reactance of a 21-pF capacitor at giving a susceptance value of 0.04 mhos
fact. such filters are called symmetrical 3900 kHz is 1943.3 ohms so the inductive which corresponds to a capacitance of
band-pass filters and fo is the "center reactance must be 1963.7 ohms. While 1608 pF.
Electrical Laws and Circuits 2-40
Coefficient of CoupUng elements would seem to contradict the
idea that low-loss or high-Q circuits
If the solution to the mobile whip-
antenna problem is examined, it can be provide the best selectivity. However, this
seen that for a given frequency, Re, Lr, is actually done in some filter designs to
improve frequency response. In fact, the
and Cr, only one value of em results in an
filter with the added loss has identical
input load that appears as a pure resis-
characteristics to one with "pure" ele-
tance. While such a condition might be
ments. The method is called predistortion
defined as resonance, the resistance value'
and is very 'useful in designing filters'
obtained is not necessarily the one \.. ..:- where practical considerations require the
required for maximum transfer of power. o ~==;::::;::::===~ use of circuit elements with parasitic or
A definition that is helpful in deter- 52 I I
10 undesired resistance.
mining how to vary the circuit elements 100 1000
RIN As the frequency of operation is
in order to obtain the desired input
increased, discrete components become
resis.tance is called the coefficient of
Fig. 69 - Variation of k with input resistance smaller until a point is reaChed where
coupling. The coefficient of coupling is for circuit for Fig. 68. other forms of networks have to be used.
defined as the ratio of the common or
Here, entities such as k and Q are
mutual reactance and the square root of sometimes the only means of describing
the product of. two specially defined such networks. Another definition of Q
reactances. If the mutual reactance is the significance and even the merit of such
definitions as coefficient of coupling and that is quite useful in this instance is that it
capacitive, one of the special reactances is is equal to the ratio of 2n (energy stored
the sum of the series capacitive reactances Q. If the circuit element values are
known, and if the configuration can be per rf cycle)/(energy lost per rf cycle.)
of the primary mesh (with the resonator
disconnected) and the other one is the sum resolved into a ladder network, important
of the series capacitive reactances of the properties such as input impedance and Mutually Coupled Inductors
resonator (with the primary discon- attenuation can be computed directly for A number of very useful rf networks
nected). Applying this definition to the any frequency. On the other hand, circuit involve coupled inductors. In a previous
circuit of Fig. 68A, the coefficient of information might be obscured or even section, there was some discussion on
coupling, k, is given by lost by attempting to attach too much im- iron-core transformers which represent a
portance to an arbitrary definition. For special case of the coupled-inductance
example, the plot in Fig. 69 merely in- problem. The formulas presented apply to
dicates Cm and Cr are changing with ~!lstances where the coefficient of coupling
k = respect to one another. But it doesn't il- IS very close to 1.0. While it is possible to
lustrate how they are changing. Such in- approach this condition at frequencies in
formation is important in practical ap- the rf range, many practical circuits work
How meaningful the coefficient of plications and even a simple table of C at values of k that are considerably less
coupling will be depends upon the and Cr vs. Rin for a particular Rr would b~ than 1.0. A general solution is rather
particular circuit configuration under mu~h !TIore valuable than a plot of k. complex but many practical applications
SImIlar precautions have to be taken can often be simplified and solved
consideration and which elements are
being varied. For example, suppose the with .t~e interpretation of circuit Q. through use of the ladder-network method.
SelectlVlty and Q are simply related for In particular, the sign of the mutual in-
value of Lr in the mobile-whip antenna
single resonators and circuit components, ductance must be taken into account if
problem was fixed at 100 ",H and em and
but the situation rapidly deteriorates with there are a'number of coupled circuits or if
Cr were allowed to vary. (It will be complex configurations. For instance, the phase of the voltage between two
recalled that Cr is 21 pF and represents
adding loss or resistance to circuit coupled circuits is important.
the antenna capacitance. However, the
total resonator capacitance could be
changed by adding a series capacitor
between Cm and the antenna. Thus, Cr
could be varied from 21 pF to some lower
value but not a higher one.)
A calculated plot of k versus input resis-
tance, Rin , is shown in Fig. 69. Note the
unusually high change in k when going
from resistance values near 10 ohms to
slightly higher ones.
Similar networks can be designed to
work with any ratio of input resistance
and load resistance but it is evident small
ratios are going to pose difficulties. For
larger ratios, component tolerances are
more relaxed. For iJlstance, Cm might
consist of switchable fixed capacitors with
Cr being variable. With a given load
resistance, Cm essentially sets the value of
the reactance and thus the input resistance
while Cr and Lr provide the required
reactance for the conductance formula.
However, if Lr is varied, k varies also.
Generally speaking, higher values of Lr
(and consequently circuit Q) require lower Fig: 70 - ~wo .types of ma~netically coupled circuits. At A, only mutual magnetic coupling exists
values of k. while t~e clrcu!t at B contains a common inductance also. EqUivalents of both circuits are shown
at the right which permit the application of the ladder-network analysis discussed in this section
At this point, the question arises as to (If the sign of voltage is unimportant, T1 can be eliminated,) .
2-41 Chapter2
The latter consideration can be il- can also be analyzed with the aid of Fig.
lustrated with the aid of Fig. 70A. An ex- 68B. At some frequency (fl in Fig. 72), the
act circuit for the two mutually coupled series reactance is zero and Gab in the
coils on the left is shown on the right. T1 preceding formula will just be l/Rr . Typi-
is an "ideal" transformer that provides cal values for Rr ,range from 10 k-ohm and
the "isolation" between terminals ab and higher. However, the equivalent induc-
cd. If the polarity of the voltages between tance of the mechanical circuit is normally
these terminals can be neglected, the extremely high (over 10,000 henrys in the
transformer can be eliminated and just the case of some low-frequency units)which
circuit before terminals cd substituted. A results in a very high circuit Q (30,000).
second circuit is shown in Fig. 708. Here, Above fl> the real;tance is "inductive"
it is assumed that the winding sense and at f2, the susceptance of the series
doesn't change between LI and L2. If so, resonator is just equal to the susceptance
then the circuit on the right of Fig. 70B of the crystal holder, B,m' Here, the total
can be substituted for the tapped coil susceptance is' zero. Since Bon is usually
shown at the left. very small, the equivalent series suscep- Fig. 71 - Diagram illustrating how M can be
Coefficients of coupling for the circuits tance is also small. This means the value larger than one of the self inductances. This
represents the tranSition from lightly coupled
in Figs .. 70A and 70B are given by for X in the susceptance formula will be circuits to conventional transformers since an
very large and consequently Gah will be impedance step up is possible without the
M small, which corresponds to a high input addition of capacitive elements.
k = resistance. A plot of the magnitude of the
impedance is shown in Fig. 72. The dip at
Ll + M f\ is called the series-resonant mode and
k = the peak at f2 is referred to as the parallel-
resonant or "anti-resonant" mode. When
specifying crystals for oscillator applica-
If LI and L2 do not have the same value,
tions, the type of mode must be given
an interesting phenonemon takes place as
the coupling is increased. A point is along with external capacitance across the
reached where the mutual inductance ex- holder or type oscillator circuit to be used.
ceeds the inductance of the smaller coil. Otherwise, considerable difference in ac-
tual oscillator frequency will be observed.
The interpretation pf this effect can be il- Rr
lu~rated with the aid of Fig. 71. While all
The effect can be used to advantage and
the flux lines (as indicated by the dashed the frequency of a crystal oscillator can b~ II 12
lines) associated with LI also encircle "pulled" with an external reactive ele-
ment or even frequency modulated with a Fig. 72 - Frequency response of a quartz·
turns of L2, there are additional ones that
device· that converts voltage or current crystal resonator. The minimum value is only
encircle eXira turns of L2, also. Thus, approximate since holder capacitance is
fluctuations into changes in reactance.
there are more flux lines for M than there neglected.
are for L1. Consequently, M becomes
larger than L1. Normally, this condition is Matching Networks
difficult to .obtain with air-wound coils In addition to filters, ladder networks 73C is set to. some particular value. The
but the addition of ferrite material greatly are frequently used to match one im- parallel combination of C2 and R2 can
increases the coupling. As k increases so pedance value to another one. While there then be transformed to a series equivalent
that M is larger than LI (Fig. 71), the net- are many such circuits, a few of them offer (see Fig 74). Then, L could be found by
work begins to behave more like a particular advantages such as simplicity of breaking it down into two components, L'
transformer and for a k of 1, the design formulas or minimum number of and L". One component (L") would tune
equivalent circuit of Fig. 71A yields the elements. Some of the more popular ones out the remaining capacitive reactance of
transformer equations of a previous sec- are shown in Fig. 73. Shown at Fig. 73A the output series equivalent circuit. The
tion. On the other hand, for small values and 73B, are two variations of an "L" network is then reduced to the one shown
of k, the network becomes merely three network .. These networks are relatively in Fig 73A and the other component (L')
coils arrange,d in a "T" fashion. One ad- simple to design. of L along with the value for CI could be
vantage of the circuit of Fig. 70A is that The situation is somewhat more com- determined from formulas (Fig 73A). Add-
there is no direct connection between -the plicated for the circuits shown at 73C and ing the two inductive components would
two coils. This property is important from 73D. For a given value of input and give the actual inductive reactance required
an isolation standpoint and can be used to output resi~tance, there are many net- for match in the circuit of Fig. 73C.
suppress unwanted currents that are often works that satisfy the conditions for a As mentioned before, it is evident an
responsible for RFI difficulties. perfect match. The difficulty can be infinite number of networks of the form
resolv:ed by introducing the "dummy shown in Fig. 73C exist since C2 can be
Piezoelectric Crystals variable" labeled N. assigned any value. Either a set of tables
A somewhat different form of resonator From a practical standpoint, N should or a family of curves for CI and L in terms
consists. of a quartz crystal between two be selected in order to optimize circuit of C2 could then be determined from the
conducting plates. If a voltage is applied component values. Either values ofN that foregoing method and as illustrated in
to the plates, the resultant electric field are too .Jow or too high result in networks Fig. 74. However, similar data along with
causes a mechanical stress in the crystal. that are hard to construct. other information, can be obtained by
Depending upon the size and "cut" of the The reason for this coinplication is as approaching the problem somewhat dif-
crystal, a frequency will exist at which the follows. Only two reactive elements are ferently. Instead of setting one of tile
crystal begins to vibrate. The effect of this required to match any two resistances. element values arbitrarily and finding the
mechanical vibration is to simulate a Consequently, adding a third element other two, a' third variable is contrived
series RLC circuit as in Fig. 68A. There is introduces a redundancy. This means one and in the case of Fig 73C and Fig 73D is
a capacitance associated with the crystal element can be assigned a value arbitrarily labeled N. All three reactances are then
plates which appears across the terminals and the other two components can then be expressed in terms of the variable N.
(Cn in Fig. 68A).Consequently, this circuit found. For instance, suppose C2 in Fig. The manner in which the reactances
Electrlcel Lew. end Circuit. 2·42.
L
R, > R2
XL .J R,R 2 - R22 Rl-- R2
Xc =
(A)
(A)
XC= R,~~
~EQ
'+;'.
(8)
1
XC2 = .J
R2
R2/R,( I + nJ) - I ,,~ 1"rrrYI 1" f"
N + R 2IX C2
XL = R, Nl+I
(Cl
L'
Rt-
o-lCt
R2
XL = -
N R,
XC2 = J
R,(Nl + I) - I
R1-- C2
R2 L (Cl
R2
XCI = R2N
W+I
(1- -...!.L )
NX C2
(D)
change with variation in N for two with either one or the other or both of the a much lower resistance would deliver
representative circuits of the type shown remaining two elements variable. In many relatively little power. However, as the
in Fig. 73C is shown in Fig. 75. The solid amateur transmitters, it is the inductor resistance is lowered, increasing amounts
curve is for an RI of 3000 ohms and R2 that remains fixed (at least for a given of current will flow resulting in more..
equal to 52 ohms. The dashed curve is for band) while CI and C2 (Fig 73C) are made power output. Then, the source is said to
the same R2 (52 ohms) but with RI equal variable. While this system limits the be "load~d" more heavily.
to 75 ohms. For values of N very close to bandwidth and matching capability some- Similar considerations such as those
the minimum specified by the inequality what, it is still a very useful approach. For discussed for the network of Fig. 73C also
(Fig 73C), XC2 becomes infinite which instance, the plot shown in Fig. 76 exist for the circuit of Fig. 73D. Only the
means C2 approaches zero. As might be indicates the range of input resistance limiting L network for the latter is the. one
expected, the values of XL and XCI at this values that can be matched for an R2 of 52 shown in Fig. 73B. The circuit of Fig. 73C
point are approximately those of an L net- ohms. The graph is for an inductive reac': is usually called a pi network and as
work (Fig 73A) and could be determined tance of 219 ohms. XCI varies from 196 to pointed out, it is used extensively in the .
by means of the formulas in Fig 73A for 206 ohms over the entire range of RI (or output stage of transmitters. The circuit of
the corresponding values of RI and R2. approximately 20 percent). However, Xc2 Fig. 73D has never been given any special
The plots shown in Fig. 75 should give a varies from 15 to almost 100 ohms as can name, but it is quite popular iri both
general idea of the optimum range of be seen from the graph. antenna and transistor-matching applica-
component values. The region close to the Since C2 more or lc:ss sets the trans- tions.
left-hand portion should be avoided since formed resistance, it is often referred to as The plot shown in Fig. 75 is for fixed
there is little advantage to be gained over the "loading" control on transmitters input and output resistances with the
an L network, while art extra component using the network of Fig. 73C, with CI reactances variable. Similar figures can be
is required. For very high values ofN, the usually labeled "Tune." While the mean- plotted for other combinations of fixed
capacitance values become large without ing of the latter term should be clear, the and variable elements. An interesting case
producing any particular advantage either. idea of loading in a matching application is for XL and RI fixed with R2, XCI, and
A good design choice is an N a few per-- perhaps peeds some explanation. For XC2 variable. A lower limit for N also
cent above the minimum specified by the small values of Xc2 (very large C2), the exists for this plot only instead of an L
inequality. transformed resistance is very high. Con- network, the limiting circuit is a network
Quite often, one of the elements is fixed sequently, a source that was designed for of three equal reactances. A feature of this
2·43 Chapter 2
1000 circuit is that the output resistance is the
XC2
I
10 k
r ratio of the square of the reactance and
500 \1 the input resistance_ An analogous situa-
XL \ tion exists with a quarter-wavelength
~
7000
XCI
transmission-line transformer. The output
1\ resistance is the ratio of the square of the
X~1\
"- 5000
10
8
1.0
9
2.0
-- --
10
3.0
XCI
XC2
II
4.0
2000 1\ equal in. the circllit of Fig. 73C is the
lumped-constant analog of the quarter-
wavelength transformer. It has identical
phase shift (90 degrees) along with the
1000
10 30
""I".
XC,
50
~ '"--
70
same impedance-transforming properties.
Frequency Response
In many instances, a matching network
resistance of 3000 ohms and an output performs a dual role in transforming a
resistance of 52 ohms. The dashed curves are
for a similar network with an input and output Fig. 76 - Input resistance vs. output reac- resistance value while providing frequency
resistance of 75 and 52 ohms, respectively. tance for an output resistance of 52 ohms. The rejection. Usually, matching versatility,
Values of N from 1 to 4 are for the latter curve is for a fixed inductor of 219 ohms (Fig. component values, and number of ele-
curves. 73C). XcI varies from 196 to 206 ohms. ments are the most important considera-
tions. But a matching network might also
be able to provide sufficient selectivity for
0
- [I some application, thus eliminating the
'"
t-- I- need for a separate circuit such as a filter.
5
I---- 1\ It will be recalled that Q and selectivity
are closely related for simple RLC series
I
...............
i'-- r-- I-- V / \ and parallel circuits. Bandwidth and ,the
parameter N of Fig. 73 are approximately
10
iii
~ N-l0 \.01 related in this manner. For values of N
much greater than the minimum specified
Z by the inequality Nand Q can be
I-
15
~ considered to mean the same thing for all
\
20
\ 1\
practical purposes. However, the fre-
quency response of networks that are
more complex than simple RLC types is
\ \ usually more complicated also. Con-
sequently, some care is required in the
25
interpretation of N or Q in regard to
o.I 1.0
1\ 10 frequency rejection. For instance, a simple
2.,.,.f(Hz) circuit has a frequency response that
results in increasing attenuation for
Fig. 77 - Frequency response of the network of Fig. 73C for two values of N.
increasing excursions from resonance.
This is not true for the pi network as can
be seen from Fig. 77. For slight frequency
changes below resonance, the attenuation
--- - -
o increases as in the case of a simple RLC
/~ / ~ r-- I-1--,,,-
network. At lower frequencies, the at-
tenuation decreases and approaches ,2.55
5
,
V f\ dB. This plot is for a resistance ratio of
7 if ,
,5: 1, and the low-frequency loss is just
V,0
\
\ ,C caused by the mismatch in source and
\ load resistance. Thus, while increasing N
" "- ...
10 / J \
~ ~,
much farther away.
z \
V / //0
for the circuit of Fig. 73D. At frequencies
\
<t
\
\ far from resonance, either a series
/I
/I
1\ capacitance provides decoupling at the
/ I\. '
20
,, lower frequencies or a shunt capacitance
~ ,
I
3-1 Chapter 3
/'
K"---' , Rf" ....
I
/ ,-
\
ENERGY
/
/
" ,,
'
I ,\ //
I \ / ...-
I \ \ / /'
\ I /' ~ - - .,.. - -~::;::~:jJ
~~::::::;:~~K-"
\ .\ \
\ \
\ \
\' I
\ ' ..... - -" /
\
, (A)
"
I
ELECTROLYTIC
CAPACITOR LS·
~
~,
10
15V
__ JYYn __ ~_
1£
C1
eF'T
®
Oft
I
•
rh
Fig. 2 - A bypassing arrangement that affords some measure of isolation (with the equivalent circuit
shown in th, inset). Dashed lines indicate a mode of wave travel that permits rf energy to leak past the
bypass circuitry and should be taken into account when more stringent suppression requirements are
necessary. (La and As in the inset represent the equivalent circuit of the ferrite bead.)
<II
"
Z
I-
!i XEQ(CP) XEQ(CS)
0---1
REQUN) REQ(S)
'-0.05
r
occurs but the points where the trans- stances, where one high-impedance load is' type shown in Fig. 10 are widely used in
formed resistance is equal to the' desired matched to another one, Rs in Fig. 10 is radio circuits. Perhaps the most common
value, the reactances are not the same much greater than the reactance of Cs and example is the i-f transformer found in
numerically. Consequently, .there is -at- Cpo This simplifies the transformations a-m and fm bc sets. Many communica-
tenuation at peaks unlike the curve of Fig. and approximat,e relations are given by tions receivers have similar transformers
9. although the trend has been toward
Other Circuit Forms ~eq(S) == x~s somewhat different circuits. Instead of
achieving selectivity by means of i-f
While the coupled network shown in transformers (which may require a num-
Fig. 8 is the easiest to analyze, it is not ber of stages), a single filter with
commonly encountered in actual circuits. quartz-crystal resonators is used instead.
As the resistance levels are increased, the As an example, suppose it was desired (The subject of receivers is treated in a
corresponding reactances become very to match a 3000-ohm load to a SOOO-ohm later chapter.)
large also. In transmitting circuits, ex- source using a coupled inductor with a
tremely high voltages are then developed 2S0-ohm (reactance). primary and secon- Single- Tuned Circuits
across the coils and capacitors. For dary coil. Assume the coupling can be In the case of double-tuned circuits,
high-impedance circuits, the circuit shown varied. Determine the circuit configura- separate capacitors are used to tune out
in Fig. 10 is often used. Although the tion and the critical coefficient of coupling. the inductive components of the primary
frequency response is somewhat different Since the load and source resistance' and secondary windings. However, exami-
than the circuit of Fig. 8 (in fact, the have a much higher numerical value than nation of the equivalent circuit of the
out-of-band rejection is greater), a match- the reactance of the inductors, a parallel- coupled coil shown in Fig. 7 suggests an
ing network can be designed based upon tuned configuration must be used. In alternative. Instead of a separate capaci-
the previous analysis for the series circuit. order to tune out the inductive reactance, tor, why not "detune" a resonant circuit
This is accomplished by changing the the equivalent series capacitive' reactance slightly and "reflect" a reactance of the
parallel primary and secondary circuits to must be -2S0 ohms. Since both Rs and Rp proper sign into the primary in order to
series equivalents. (It should be em- are known, the exact formulas could be tune out the primary inductance. Since
phasized that this transformation is good solved for 'Y and Req. However, because the transformation function (shown in the
at one frequency only.) The equivalent the respective resistances are much greater box in Fig. 7) reverses the sign of the
circuit of the one shown in Fig. 10 is than the reactance, the simplified ap- secondary reactance, it is evident Xs must
illustrated in Fig. II where the new proximate formulas can be used. This be inductive in order to tune out the pri-
resistance and reactance of the secondary means the primary and secondary equiva- mary inductance.
are given by lent capacitive reactances are - 2S0 ohms. This might seem to be a strange result
The equivalent secondary resistance is but it can be explained with the following
(2S0)'/3OOO or 20.83 ohms, resulting in a reasoning. From a mathematical point of
sec~:mdary Q of 2S0120.83 or 12. (A view, the choice of the algebraic sign of
formula could be derived directly for the the transformed reactance is perfectly
Q from the approximate equations.) The arbitrary. That is, a set of solutions to the
equivalent primary resistance and Q are equations governing the coupled cih:uit is
12.S ohms and 20, respectively. Sub- possible assuming either a· positive or
stituting the values for Q into the formula negative sign for the transformed reac-
for the critical coefficient of coupling tance. However, if the positive sign is
A similar set of transformations exists for gives 11" (20)(12) or' 0.06S. chosen, the transformed resistance would
the primary circuit also. In most in- Double-tuned coupled circuits of the be negative. But from a physical point of
3-5 Chapter 3
Fig. 12 - A boil coupled magnetically to a -xs xs
"shorted" turn provides insight to coils near solid
shield walls.
Cs
/' Rs Fig. 17 - "Reflected" reactance into the primary of a single-tuned circuit places restraints on
3 resistances that can be matched. This gives rise to a general rule that high-Q secondary circuits
require a lower coefficient of coupling than low-Q ones; .
Fig. 13- "Link" coupling can be used to analyze view, this is a violation of the con- resonators can be coupled in this manner
a number of important circuits. servation of energysincl; it would imply although there may be considerable
the secondary resistance acts as a source separation (and no mutual coupling
of energy rather than an energy "sink." between- the larger coils) hence the term
Consequently, the solution with the nega- "link" coupling. While· this particular
tive resistance does not result in a physi- method is seldom used nowadays, the
cally realizable network. term is still applied to the basic con-
The foregoing phenomenon has im- figuration shown in Fig. 13. Applications
plications for circuits one might not would be antenna-matching networks,
normally expect to be related to coupled output stages for amplifiers and, es-
networks. For instance, consider coil I pe<:ially important, many circuits used at
(Fig. 12) in proximity to the one-turn' vhf that have no direct hf equivalent.
"shorted" coil 2. A Jime-varying current The cavity resonators used in repeater
in coil 1 will induce a current in coil 2. In duplexers are one form of vhf circuit that
turn, the induced current will set up a uses link coupling. A cross-sectional view
magnetic field of its own. The question is of a representative type is shown in Fig.
will the induced field aid or oppose the 14. Instead of ordinary coils and capaci-
primary field. Since the energy in a tors, a section of coaxial transmission
magnetic field is proportional to the line comprises the resonant circuit. The
square of the flux, the induced field must frequency of the resonator may be varied
oppose the primary field, otherwise the by adjusting the tuning screw which'
principle of the conservation of energy changes the value of the capacitor. Energy
would be violated as it was with the is coupled into and out of the resonator by
Fig. 14 - A vhf/uhf circuit which can be
approximated I)y a link-coupled network using
"negative" resistance. Consequently, the means of two small, one-turn loops.
"conventional" components. induced current must always be in a Current in the input loop causes a
direction such that the induced field magnetic field (shown by dashed lines). If
opposes changes in the generating field. the frequency. of the generating field is
This result is often referred to as Lenz's near one of the resonant "modes" of the
Law. configuration, an electric field will also be
If, instead of a one-turn loop, a solid generated (shown by solid lines). Finally,
shield wall was substituted, a similar energy may then be coupled out of the
CR phenomenon would occur. Since the total resonator by means of a second loop.
flux (for a given current) would be less A low-frequency equivalent circuit of
with the shield present than it would be in the resonator is shown in Fig. 15.
. the absence of the shield, the equivalent However, the circuit can only be used to
coil inductance is decreased. That is why it give an approximate idea of the actual
Fig. 15 - Equivalent low-frequency analog of the is important to use a shield around a coil frequency response of the cavity. At
circuit shown in Fig. 14. that is big enough to reduce the effect of frequencies not close to the resonant
such coupling. Also, a shield made from a frequency, the mathematical laws gover-
metal with a high conductivity such as ning resonant circuits are different from
copper or aluminum is advisable, other- those of "discrete" components used at
R'
wise a loss resistance WIll be coupled into hf. Over a limited frequency range, the
the coil as well. resonator can. be approximated by the
series LC circuit shown in Fig. 15.
Link Coupling Applying· the formulas for coupled
An example of a very important class of networks shown in Fig. 7 to the two-link
single-tuned circuits is shown in Fig. B. circuit of Fig~ 15, the output link and load
The primary inductor consists of a small can be transformed to an equivalent series
Fig. 16 - The network of Fig. 15 can be reduced coil either in close proximity or wound resistance and reactance as shown in Fig:
with the transformation shown in Fig. 7. over one end of a larger coil. Two 16. In most instances, the reactance, Xs , in
Radio De.lgn Technique and Language 3·6
present. A plot of the reflected 'reactance
as a function of Xs is shown in Fig. 17.
From mathematical considerations (which
RS will not be discussed) it can be shown that
the maximum and minimum of the curve
have a value equal to Xm2/2Rs. Con-
sequently, this value must be greater than
or equal to Xp in order that a value of Xs
Fig. 18- Single-tuned circuit with a parallel RC
exists such that the reflected reactanCe will , 0.85 0.9 0.95 1.05
1.0 1.1
secondary. cancel Xp. In the usual case where Xm2/ 2",.1 (Hz)
2Rs is greater than Xp, it is interesting to
note that two values of Xs exist where Xp
Fig. 20 - Input resistance of the Fig. 19 circuit as
and the reflected reactance cancel. This a function of frequency.
means there are two cases where the input
impedance is purely resistive and Rs could
be matched to either one of two source
resistances if so desired. The value ofXs at
these points is designated as Xsi and Xs2.
On the other hand, a high value of Rs ,
Fig. 19 - Text example of a single-tuned circuit. requires )b to be large also. This could \
\
be accomplished by increasing the coeffi- I 50
cient of coupling or by increasing the I
I
the formula is just the reactance, of the turns on the secondary coil. Increasing the I
output link. Since the. two-link network turns on the primary also will cause Xm to I 40
I
has been reduced to a single coupled be higher but Xp will increase also.. This I
circuit, the formulas can be applied again is somewhat self-defeating since Xm 2 is
to find the input resistance and reactance.' proportional to Xp.
An alternate approach is to use the paral-
Analysis of Single-Tuned Circuits lel configuration of Fig. 18. The approxi-
Single-tuned circuits are very easy to mate equivalent series resistance of the
construct and adjust experimentally. If parallel combination is then X(Cs)2!Rs
desired, the tuned circuit consisting of Ls, and the reactance is approximately
Cs, and perhaps the load, Rs, can be con- X(Cs). (See diagram and text for Fig. 11.)
structed first and tuned to the "naturaP' This approach is often used in multiband
. resonant frequency antenna. systems. On some frequencies,
0.9 1.1
I
the impedance at the input of the feed lipe
f0 = r-;--;::;-- is high so the circuit of Fig. 18 is
271' V Lees employed. This is, referred to as parallel
Then, the primary inductor, which may be tuning. If the impedance is very low, the
a link or a larger coil, is brought into circuit of Fig. 13 is used and is called
-20
proximity of the resonant circuit. The series tuning.
resonant frequency will usually shift As an example, suppose a single-tuned
upward. For instance, a coil and capacitor circuit is to be usedto match a I-ohm load Fig. 21 - Input reactance of the Fig. 19 network.
combination was tuned to resonance by to a50-.ohm source as shown in Fig. 19. It Note two "resonant" frequencies (where
means of a grid-dip oscillator (see the might be pointed' out at this juncture that reactance is zero).
chapter on measurements) at a frequency coupling networks using mutual magnetic
of 1.8 MHz. When a two-turn link was coupling can be sqaled in the same manner
wound over the coil, and coupled to the that filter networks ar~ scaled (as dis-
grid-dIp osc.illator the resonant frequency cussed in chapter 2). For instance, the cir-
had increased to 1.9 MHz. A three-turn cuit of Fig. 19 could be scaled in order to 0.9 0.95 1.0 1.05
link caused a change to 2 MHz. match a 50-ohm load to a 2500-ohm
Quite often an actual load may be an source merely by multiplying all the
unknown quantity, such as an antenna, reactances by a factor of 50.
and some insight into the effects of the The input resistance and reactance of
various elements is helpful in predicting the circuit of Fig. 19 are plotted in Figs. 20
single-tuned circuit operation. Usually, as and 21, respectively. As pointed out
in the case of most matching networks, Rs earlier, there are two possible points
(Fig. 7) and the input resistance are speci-. where the reactance is zero and this circuit
fied with the reactive components being could be used to match the I-ohm load to
'the variables. Unfortunately, the variables either a 50-ohm or IS5-ohm source.
in the case of mutually coupled networks Assuming a 50-ohm source was being
are not independent of each other which used, the attenuation plot as a function of
complicates matters somewhat. frequency would be given by the solid
Examination of the equivalent circuit curve in Fig. 22.
shown in Fig. 7 would' indicate the first With slight modification to include the
condition is that the reactance reflected effect of the source, the transformation of
from the secondary into the primary be Fig. 7 can be applied to the p~imary side
sufficient to tune out the primary reac- of the coupled circuit shown in Fig. 19.
tance. Otherwise, even though the proper This is illustrated in Fig, 23. The complete
resistance transformation is obtainable, a circuit is shown at Fig. 23A and the Fig. 22 - Response of the circuit shown in Fig.
reactive component would always be network with the transformed primary 19.
3· 7 Chapter 3
.' ,
resistance and reactance is shown in Fig. a I-ohm resistance (shown as a dashed circuit of Fig. 24A is
23B. line in Fig. 23B) must be th,e same as that
In a lossless transformer, the maximum delivered to a 50-ohm load }n Fig. 23A.
available power at the secondary must be This assumes that the rest of the circuit
the same as that of the original source on has been disconriected in either case. In
the primary side, neglecting the effects of order to fulfill this requirement, the
reactalfce. That is, the power delivered to original source voltage must be multiplied and the network shown in Fig. 24B in
by the square root of the ratio of the new terms of the coefficient of coupling is
and old source resistance. illustrated in Fig. 24C. For k equal to 1.0,
The single-mesh transformed network the input reactance is zero and the input
Ep
01h element values at resonance with the
circuit of Fig. 19. For comparison, the
response of such a circujt is shown in Fig.
22 as a dashed curve and it can be seen where N I and N2 are the number of turns
(A)
that it differs only slightly from the on coil 1 and 2, respectively. From
coupled-circuit curve. The reason for the maximum-power transfer considerations,
similarity is that even though the trans- such as those discussed for the circuit of
formation of the primary resistance and Fig. 23, the voltage transfer ratio becomes
If\.
reactance also changes with frequency, the
effect is not that great in the present case.
Broadband RF Transformers
(8) , The '''sensitivity'' of \he frequency
Sh characteristic of the transformation shown It will be recalled that the foregoing
in Fig. 7 depends mostly on the ratio of Xs equations occurred in the discussion of
toRs. However, ifXs is much greater than the "ideal transformer" approximation in
Rs, the transformed reactance can be ap- Chapter 2. ,It was assumed then that the
lh proximated by leakage reactance and magnetizing' cur-
rent were negligible. The effects on circuit
operation of these variables are shown in
Fig. 25. The curves were computed for
(C)
various load resistances (Rs) using the
exact equations shown in Fig. 7.
and 'the resistance becomes XI and X2 are assumed to be 100 and 10
Fig. 23 - The transformation of Fig, 7 applied to
the primary side of the circuit of Fig. 19. ohms, respectively, with the solid curves
for a k of 1.0 and the dashed reactance
curve for k equal to 0.99 (the resistance
curve for the latter value is the same as the
one for k equal to 1.0). The ideal-
X~{l Applying this .approximation to the transformer representation can be modi-
general coupled circuit shown in Fig. 24A fied slightly to' approximate the curve of
(A)
100r----------,-----------.----------,-----~--~~----~--~
XI
-XM2
X2
10r---------~----------~----~~--+_--------~~~--~--~
!';
x
k -0.99
z
Ii
(8)
. Fig. 24- Equivalent-circuit approximation of two Fig. 25- Input resistance and reactance as a function of output load resistance for XI and X2 E!.qual to
coupled coils. 100 ohms and 10 ohms respectively (Fig. 24). .
.
component as shown in Fig. 25. permeability much higher than that of air.
One approach is to use a core with a An example is illustrated in Fig. 28 and ~o
examples would be power transformers toroidal transformer. Since the flux is "\ : I
/
and similar types common to the af range. proportional .to . the product .of the .'-', \ 11 I I
f{owever, when an inductor configuration permeability and the magnetizing current; 1_~-I/IlL
contains materials of more than one the flux in the core shown in Fig. 28 will I ../-~--'
S \
permeability, the analysis relating to Fig~ be much greater than the coil con- 1'\~/4"
, '- I . ~ •
24C has to be modified somewhat. The figuration of Fig. 27. However, not all of I ~~_~, I COlLI
manner in which the core affects the the flux is confined to the core. As can be \\~z':
' .... _/ ; ~1
circuit is a bit complicated although even seen in Fig. 28, some of the flux lines I : I '-'-
I .
a qualitative idea of how such trans- never penetrate the core (see lines marked I
formers work is yery useful. a in Fig. 28) while others enclose all the I
I
First, consider the coupled coils shown windings of coil I but not coil 2 (see line I I
in Fig. 27. For a given current, II, a num- marked b). The significance of these
ber of "flux lines" are generated that link effects is as follows. The total,flux linkage
both coil I and coil 2. Note that in coil I, produced by the current, II, is'
not all of the flux lines are enclosed by all
~,b~
~/~~ COIL 2
3·9 Chapter 3
netizing reactance, the residual parasiti,c voltage-standing-waveratio (VSWR). twist the wires together. Either way, tbere
elements must still be made as low as These results are based on the exact are a number of advantages (and some
possible. This is especially important in equations and it can be seen that the disadvantages) to be gained. Referring to
matching applications as the following approximate relations shown in Fig. 29C Fig. 27, the fact that not all the flux lines
example illustrates. A transformer has a are valid up to I ohm or so. Curve A (Fig. linked all of the turns of a particular coil
primary and~econdary leakage reactance 30) only includes the effect of the meant the self inductance was lower than
of 1 oh,m and O.l ohm, respectively, with a secondary reactance and illustrates the if all the turns were linked. Since the
coefficient of coupling of 1.0 in the core. manner in which the reactance is trans- separation between turns of a particular
Xl and Xl are 1000 ohms and 100 ohms. formed. Curve B is the total input coil is quite large in the configuration of
A plot similar to the one of Fig. 25 is reactance which merely requires the Fig. 31, the flux linkage between turns is
shown in Fig. 30 along with a curve for addition of I, ohm. The VSWR curve quite low. This means the corresponding
includes the effect of the latter. Useful leakage inductance is reduced according-
range of the transfoTmer is between I and, ly. However, the coupling between both
10 ohms with rapid deterioration in coils is increased because of the bifilar
VSWR outside of these values. (The winding (flux line A) in Fig. 31 which also
::JrO~
VSWR curve is for a characteristic, tends to reduce the leakage inductance of
impedance equal to 10 times the secon- either coil.
dary resistance. For instance, the transfor- On the other hand, the capacitance
mer would be useful in matching as-ohm between windings is increased considerably
load to a 50-ohm line.) as indicated by B in Fig. 31. ,As a tesult,
(AI As mentioned previously, these dif- the coupling between windings is both
ficulties a're less pronounced at audio electrical and magnetic in nature. Generally
frequencies since tbe permeabilities nor- speaking, analysis of the problem is quite
mally encountered in iron-core trans- complicated. However, a phenomenon
formers are so high, the actual inductance usually associated with such coupling is
of the winding itself is small in com- that it tends to be directional. That is,
parison to the component represented by energy transferred from one winding to
the core. That is, a small number of turns another one propagates in a preferred
qf wire wound,on a core may actually be direction rather than splitting equally.
the equivalent of a very large coil.
However, materials suitable for rf ap- Directional Coupling
plications have much lower permeabilities
Two conductors are oriented side by
(81 and a narrower range of matching, values
side over a conducting plane as shown in
is likely to be the result (such as in the
example of Fig. 30). Therefore, other Fig. 32. A current I in conductor 1 will
induce a current 1m in conductor 2
means are required in ~eeping the
because of magnetic coupling. The actual
parasitic elements as low as possible.
Either that, or less conventional trans- value 'Of the current will depend upon the
external circuitry attached to the con-'
former designs are used.
One approach is shown in Fig. 31. ductors but it will be assumed that the two
of them extend to infinity in both
Instead of separating the windings on the
core as shown in Fig. 28, they are wound directions.
Since capacitive coupling exists also, a
in parallel fashion. This is called a bifilar
winding although a more common ap- second set of current components denoted
Fig, 29 - Effect of a high-permeability core on by Ie will also flow. The result is that a
transformer equiva~ent circuit. proach to achieve the same purpose is to
wave traveling toward the right in
conductor I will produce a wave traveling
100 r - - . - - - - - , - - - - - - - - , - - - - - - " 7 t - - - - - , - - - - - , 100'1 toward the left in conductor 2, Such
coupling is called contradirectional coup-
ling since the induced wave travels in the
opposite direction to the generating wave.
'This is ,the principle behind many
practical devices and ones that are quite
common in amateur applications. In
10 10'1
adjusting a load such as an antenna, it is
desirable to insure that energy is not
S II:
reflected back to the transmitter. Other-
)( ~
U>
wise, the impedance presented to the
0: > transmitter output may not be within
range of permissible values. A directional
1'0 coupler is useful in dC;!termining how much
power is reflected as indicated in Fig. 33.
Energy originating from the transmitter
and flowing to the right causes a voltage
to be produced across the resistor at the
left. On the other hand, a wave traveling
from the right to the left produces a
O~,I<..OI------O"",I:--------:-·II..-------':IO,-----~--JIOO voltage across the right-hand resistor. If
RS(nJ
both of these voltages are sampled, some
idea of the amount of power reflected can
be determined. (The subject of reflected-
Fig, 30..,... Curve for transformer problem discussed in the text. power'is taken up in more detail in the
Almost all the theory in previous output, it is convenient to consider certain element such as a resistor is said to be
sections has dealt with so-called passive elements as controllable sources of power. bilateral since it doesn't matter which way
components. Passive networks and _com- Such devices are callec;l amplifiers and are it is connected in a circuit. Semiconductor
ponents can be represented solely by part of a more general class of circuits and vacuum-tube devices such as triodes,
combinations of resistors, capacitors and called active networks. An active network diodes, transistors and integrated circuits
inductors. As a consequence, the power generally possesses characteristics that are (ICs) are all examples where the concept
output at one set of terminals in a passive different than those 'of simple RLC of a bilateral element breaks down. (For
-network cannot exceed the total power circuits although the goal in many readers with limited backgrounds in the
input from sources co'nnected to other instances is to attempt to represent them basic operation of vacuum tubes, recom-
terminals in the circuit. This assumes all in terms of passive elements and genera- mended study would be The Radio
the sources are at one frequency. Similar tors. Amateur's License Manual and Under-
considerations hold true for any network, standing Amateur Radio. Both publica-
however, it is possible for energy _to be Nonlineuity tions contain fundamental treatments of
converted from one frequency (including Two other important attributes Qf vacuum-tube principles and are available
dc) to other ones. While the total power passive RLC elements are that they are from T~e American Radio Relay League.)
input must still equal the total power linear and _bilateral. A two-ter~inal The manner in which the device is
3·13 Chapter 3
connected in a circuit and 'the polarity .of
the voltages involve(l are very important.
An implication of the failure to satisfy E2 SIN 3wt
the 6ilateral requirements is that such
devices are nonlinear in the strictest sense.
Linearity meaps that the amplitude of a EtSINwt
current), the term linear is applied to T /4. However, the fact that the curve is
distinguish the mixer from types without flattened somewhat means energy at the
this "quasi-linear" property. original sine-wave frequency has been
converted to other frequencies. It will b~
recalled that the sum of a number of sine
Harmonic-Frequency Generation waves at one frequency result in another
In a circuit with only linear com~ sine-wave at the same frequency. There- Eo
1'0 k
ponents, the only frequencies present are fore, it must be concluded that. the
+
those generated by the sources them- waveform of Fig. 40 has more thim one EIN
selves. However, this is not true with frequency component present since it is
nonlinear elements. One of the properties no longer a sinc;-wave.
of nonlinear networks mentioned earlier is One possible "model" for the new (B)
that energy at one frequency (including waveform is shownin Fig. 41A. Instead of
dc) may become converted to other one si....wave at a single frequency, there
frequencies. In effect, this is how devices are two generators in series with one·
such as transiStors--and vacuum tubes are generator at three times the "fundamen-
able to amplify radio signals: Energy from tal" ti"equencyw where wis 271' f(Hz). If the +
the dc power supply is converted to energy two sine waves are plotted point by point,
Ee RL -10 k EO
at the desired signal frequency. Therefore, the dashed curve of Fig. 41B results. While
a greater amount of signal power is this curve doesn't resemble the one of Fig.
available at the output of the network of 40 very closely, the. general symmetry is
C. active device than at the input. the !lame. It would take an infinite number (e)
O'n the other hand, such frequency of generators to represent the desired
generation may be undesirable. For curve exactly, but it is evident all the Fig. 42 - Basic triode amplifier and equivalent
instance, the output of a transmitter may frequencies must be odd multiples of the circuit.
fa o TUNING
SCREW
L1
,,-_....,-REPELLER
(A)
., C2 C4~
r--ir--'
I I
I I -v ..
I I
1'"-+-
I
.... - Fig. 51 - Microstripline 2-GHz oscillator circuit
I -uv using the RCA 2N54 70, .
I
I L C 1, C2 - 0.35-3.5 pF, Johanson type 4702 or
I
Fig, 50 - 1.6a-GHz oscillator using the RCA equiv.
*CI C3, C4 - 100 pF, Allen-Bradley type 5A5C or
I 2N510a.
I equiv. •
I
I
L1- Microstripline 0.70-inch (17.a-mm) long X
L----4-~~~----O
0.30-inch (7,6-mm) wide, mounted on 1/32-
OUTPUT
largely by solid-state devices. One device inch (o.a-mm) Teflon fiberglass board.
L2 _ Microstripline 0.43-inch (1 0.9-ml)"l) long X
, (S) which is popular with amateurs is the O.Oa-inch (2.0 mm) wide mounted on 1/32-inch
Gunn diode. When the dc voltage applied Teflon fiberglass board.
across Ii samplepf n-type gallium arsenide L3- 5 turns no, 33 wire, 1/16-inch (1.6-mm) 10
(GaAs) is increased, a point is reached ,X 3/16-inch (4.a-mm) long.
C2 L4 - 50-ohm miniature coaxial line, 1-1/2.inches
beyond which the current flow fluctuates
r--~~--' at a regular rate. the frequency of this
long.
I I
I
I variation is dependent upon the thickness
,.----+
I
of the GaAs material, becoming higher as (jaAI CHIP
I
the material is made thinner. Thinner
I
slabs require a lower threshold voltage to
*CI
I initiate the "Gunn effect."
I
I C5 In practice, a Gunn-diode package (see
L. - - - - - - --:H''-"1 Fig. 52) is mounted in a resonant cavity.
Dc bias is applied through a quarter-wave-
j.
work, meaning a system which operates
TUNING
with numbers represented by directly '.
measurable quantities (analog readout-
mechanical dial system. See digital).
attenuator - A passive network that re-
duces the power level of ,a signal with-
out introducing appreciable distortion.
balun - BaHlflced. to 1mbalanced-line
transformer.
bank wound - Pertaining to a coil (in-
• ductor) which has two or more layers
of wire, each being wound over the top
of the preceeding one.· (See solenoid.)
bandpass - A circuit or component
characteristic which permits the passage
Fig. 53- A simple Gunn-diode-oscillator cavity for direct mounting to waveguide or a waveguide horn. of a single band of frequency while at-
Coarse frequency adjustment is accomplished by means of the tuning screw. A smaller cavity which tenuating those fr.equencies which lie
functions as an rf choke is discussed in the text. above and below that frequency band.
band-reject - A circuit or device whiCh
rejects a specified frequency band while
passing those frequencies which lie
RFC1
above and below the rejected band (op-
CURRENT .L1MITER C2 posite of band-pass). Sometimes called
AND
VOLTAGE REGULATOR
~--~~--~~~=')-~
;:J:; , ,,..--- ---,,
"band-rejection." as applied to a filter.
bandwidth - The frequency width of
+C1
: D2 : CAVITY circuit or component, such as a band-
J~'
D1
+ I " pass filter or tuned circuit. Usually
, '
.... _ _ _ _ _ _ oJ
measured at the half-power points of
the response curve (- 3 dB points) ..
base loading - Applies to vertical anten-
nas for mobile and fixed-location use;
an inductance placed near the ground
end of a vertical radiator to change the
Fig. 54 - Block diagram of a basic Gunn-diode oscillator, showing the various components necessary electrical length. With variations the
for proper operation. Current lim iting is desirable to prevent damage to the supply in the event of Gunn
or Zener-diode failure.
il,lductor aids in impedance matching.
C1 - Electrolytic capacitor. voltage of Gunn diode. bifilar - Two conducting elements used
C2 - Feedthrough/bypass capacitor (part of D2- Gunn diode. in parallel; two parallel wires wound on
microwave cavity). RFC 1 - 1/4.-wavelength cavity. May be a coil form, as one example.
D1 - Zener diode. Voliage rating selected machined into qavitywall. A similar length of
slightly higher than maximum safe operating transmission line maybe used.
bilateral --' Having two symmetrical sides
or terminals; 'a filter (as one example)
which has a 50-ohm characteristic at
each port, with either port suitable as
. the input or output one.
bias - To influence current to flow in a
In addition to microwave oscillation, a additional current capacity is needed an specified direction by means of dc vol-
Gunn diode produces significant output external pass tran~istor may be added. tage; forward bias on a transistor stage,
at lower frequencies, This oscillation will or grid bias on a tube type of amplifier ..
modulate the bias voltage causing un- Glossary of Radio Terms binary - Relating to two logical ele-
desired extreme shifts in frequency. A It is not' unusual for an inexperienced ments; a system of numbers having two
10 u F electrolytic capacitor connected radio amateur to be confused by some of as its base ..
across the bias terminal will prevent such the terms which are taken for. granted by bit - An abbreviation of a binary digit;
oscillation, It is also desirable to protect those who have been involved with the a unit of storage capacity. Relates pri-
the Gunn diode from possible application pastime for many years. Many of the marily to computers.
of overvoltage or voltage of an incorrect "strange" words which are found in this blanker - A circuit or device which
polarity. A Zener diode will afford this handbook and other amateur publications momentarily removes a pulse qr signal
protection at low cost and should be are listed here for the convenience of the so that it is nQt passed to the next part
mounted on the bias tertninal as well'. The layman ip the radio field. Although this of a circuit; a noise blanker. Not to be
schematic of a basic Gunn oscillator and glossary is far from being all-inclusive, it confused with a clipper, which clips part
power supply is given in Fig. 54. does contain the most-used words which of a pulse or waveform.
Frequency modulation of a Gunn may cause confusion'. A complete dic- bridge - An electrical instrument used
oscillator may be accomplished by modu- tionary of modern electronics words and for measuring or comparing induc-
lation of the bias supply. Audio may also expressions is available to those wishing tance, impedance, capacitance or resis-
be applied to the bias terminal through a to complement the radio library - IEEE tance by comparing the ratio of two op-
coupling capacitor. Frequency deviation Standard Dictionary of Electrical and posing voltages to a known ratio; to
will be greater if the supply voltage is Electronics Terms by John Wiley & Sons, place one component in parallel with
directly modulated. Because oscillator Inc., New York, NY 10016 another; to join two conductors or
frequency is affected by supply voltage, components by electrical means.
the' bias regulation must be very good. The Terms broadband - A device or ·circuit that
Typical bias requirements for Gunn active - As use9 in active filter or active is broadband has the capability of being
diodes are from 10 to 14 volts. This device: A device or circuit which requires operated over a broad range of frequen-
voltage ~ay be obtained from a variable an operating voltage. (See passive.) cies. A broadband antenna is one example.
output regulator such as the 723. If . analog - A term used in computer byte - A sequence of adjacent binary
3-19 Chapter 3
digits operated upon as a unit - usual- ratio. Also, the number of decibels de- pulse or signal).
ly shorter than a word. . noting the ratio of two amounts of . encoder - A device for enabling a
cascade - One device or circuit which voltage being 20 tim.es the .logarithm to circuit; to express a' character or mes-
directly follews another; two or more , the base 10 of this' ratio.. sage by means of a code while using an
similar devices or cir:cuits in which the decoder - A devipe used for decoding an encoder. Using a tone or tones to acti-
,output of one is fed to the input of the encoded message. One such circuit vatea repeater, as one example, in
succeedins one (tandem). would be a decoder used for decoding which case a Touch-Tone pad could be
cascode - Cascaded stages which are the output signal of a Touch-Tone pad. the encoder. .
cathode-coupled. differential amplifier - An amplifier that excitation - Signal energy used to drive
chip - Slang term for: an integrated has an output signal which is propor- a transmitter st.age (see drive). Voltage
cir(:uit, meaning a chip of semiconduc- tional to the algebraic difference be- applied to a component to actuate it,
tor material upon which an IC is tween two input signals (sometimes such as the field coil of a relay.
formed. called a "difference amplifier"). Faraday rotation - Rotation of the plane
clamp - A circuit which maintains a digital - Relating' to data which is of polarization of an electromagnetic
predetermined characteristic of a wave renderetl in the form of digits; ,digital wave when traveling through a magne-
at each occurence so that the voltage or readout or display (see analog). tic field. In ~pace communications this
current is "clamped" or held at a speci- diplexer - The simultaneous transmis-' effect occurs when signals transverse
fied value. sionor reception of two signals while the ionosphere.
clipper -·A device or circuit which limits using a common antenna, made pos- feedback - A portion of the output
the instantaneous value of a wave form sible by using a "diplexer." Used in voltage being fed back to the input of
or pulse to a predetermined value, (see TV broadcasting to transmit visual and an amplifier. Description includes' ac
blimker). aural carriers by means of a single and dc voltage which can be used sepa-
closed loop '-- A signal path which in- antenna. rately or together, depending on the
cludes a forward route, a feedback discrete - A single device or circuit (a particular circuit.
path, and a summing point which pro- transistor as opposed to an IC) (see feedthrough - Energy passing through a
vides a closed circuit. In broad terms; composite). circuit or component, but not usually
an amplifying circuit which is providing dish - An antenna reflector for use at desired. A type of capacitor which
voltage or power gain while being ter- vhf and higher which has a concave can be mounted on a chassis or panel
minated correctly at the input and out- shape. For example, a part of a sphere wall to permit feeding through a
put ports, inclusive of feedback. 'p r paraboloid. dc voltage while bypassing it to ground
cold end - The circuit end of a com- Doppler - The phenomenon evidenced at ac or rf. Sometimes called a "coaxial
ponent which is connected to ground or by the change in the observed frequen- capacitor."
is bypassed for ac or rf voltage (the cy of a wave in a transmission system ferromagnetic - Material which has a
grounded end of a coil or capacitor). caused by a time rate of change in the relative permeability greater than unity
common-mode Signal - The instantane- effective length of the path of travel be- and requires a magnetizing force. (Fer-
ous algebraic average of two signals tween the source and the point of ob- rite and powdered-iron rods and to-
applied to a.balanced circuit, both sig- servation. roids). "-
nals referred to a common reference. drift - A change in componentor circuit finite - Having a definable quantity; a
composite - Made up of a collection of parameters over a period of time. finite value of resistance or other elec-
distinct components; a complete ("com- drive - Rf energy applied at the input trical measure.
posite") circuit rather than a discrete of an rf amplifier (rf driving power or flip-flop - An active circuit or device
part of an overall circuit. voltage). which can assume either of two stable
conversion loss/gain - Relating to a dummy load - A dissipative but essen- states at a given time, as dictated by the
mixer circuit from which less output tiaIiy nonradiating device having impe- nature of the input signal.
energy is taken than is supplied at the dance characteristics simulating those floating - A circuit or conductor which
input-signal port (loss); when a mixer of the substituted device. is above ac or dc ground for a particular
delivers greater signal output than is duplex.- Simultaneous two-way indepen- reason. Example: A floating ground bus
supplied to the input-signal port (gain). dent transmission and reception in both which is not common to the circuit
converter - A citcuit used to convert one directions. chassis.
frequency to another frequen~y.· In a re- duplexer - A device which permits simu- gate - A circuit or device, depending
ceiver the converter stage converts the ltaneous transmission and r~eption of upon the nature of the input signal,
incoming signals' to the imtermediate reilited signal energy while using a coQ)- which can permit the' passage or
frequency. mon antenna (se~ diplexer). blockage of a signal or dc voltage.
core - An element made of magnetic ma- dynamic range - Difference in dB or GDD":'" Abbreviation for a grid-dip
terial, serving as part of· a 'path for dBm between the overload level and oscillator (test instrument). Correct for
magnetic flux . minimum discernible signal level (MDS) only a tube-type of dip meter.
. dilmping - A progressive reduction in in a system, such as a receiver. Parame- ground loop _ A circuit-element con-
the amplitude of a wave with respect to ters include desensitization point and dition (pc-board conductor, metal chas-
time (usually referenced to microseconds distortion products as referenced to the sis or metal cabinet wall) which permits
or' milliseconds); a device or network receiver noise floor. the unwanted flow of ac current from
added to a circuit to "damp" unwanted EME - Earth-moon-earth. Communi- one circuit point to another.
oscillations. . cations carried on by bouncing signals ' half-power point - The two points on a:
decay time - The period of time during off the lunar surface. Commonly refer- response curve which are 3 dB lower in
which the stored energy or information red to as moonbounce. level than the peak Rower. Sometimes
"decays" to a specified value less than empirical - Not based on mathematical - called the "3 pB bandwidth."
its initial value, such as the discharge design procedures; experimental en- Hall effect - The change of .the electric
time of a timing network. deavor during design or modification conduction caused by the component of
decibel (dB) - One tenth ofa bel. The of a circuit. Founded on case-history the magnetic field vector normal to the
number .of decibels denotes the ratio of experience or intuition. current density vector, which instead of
two 'amounts of power being 10 times enabling - The preparation of a circuit being parallel to the electric field,
the logarithm to the base 10 of this for a subsequent function (enabling forms an angle with it.
Redlo Dnlgn Technique end Languege 3·20
high end - Refers generally to the "hot" tor for controlling a transmitter or version of Ii circuit design.
(rf or dc) end of a compontmt. or circuit; receiver frequency; Can be a VFO a
QL - Loaded Q of circuit.
the end opposite the grounded or by- (variable-frequency oscillator), VXO Qu - Unloaded Q of a circuit.
passed end (see cold end). (variable crystal oscillator), PTO (per- quagi - AI} antenna cOftsisting of both
high level - The part of a circuit which is meability-tuned oscil,lator), PLL' (phase- full-wavelength loops (quad) and Vagi
relatively high in power output and locked loop), LMO (linear master oscil- elements.
consumption as compared to the small- lator) or frequency.synthesizer. resonator - A general term for a high-Q
signal end of a circuit., Example: A modulatio" index - The ratio of the fre- resonant circuit, such as an element of
transmitter PA stage is the high-level quency deviation of the modulated a filter.
amplifier, as might be the driver also. wave to the frequency of the modu- return - . That portion of a circuit which
high-pass - Related mainly to filters or lating signal. ' permits the completion of current flow,
networks which are designed to pass narrowban,d ,- A device or circuit that usually to ground - a "ground retum."
energy above a specified frequency, but can be operated only over a narrow ringing - The generation of an audible
attenuate or block the passage of energy range of frequencies. Low-percentage or visual signal by means of oscillation
below that frequency. bandwidth. or pulsating current; the annoying
high-Z - The high-impedance part of a network - A group of components con- sound developed in some audio filters
circuit; a high-impedance microphone; nected together to form a circuit which when the Q is extremely high.
a high-impedance transformer winding. will conduct power, and in most ripple - Pulsating current. Also, the
hot end - see high end. examples effect an impedance match. gain depressions which exist in the flat
hybrid - A combination of two general- Example: An LC matching network portion of a' bandpass response curve
Iy unlike things; a circuit which con- between stages of a transistorized trans- (!lbove the - 3 dB points on the curve).
tains transistors and tubes, for example. mitter.! . Example: Passband ripple in the nose of
ideal - A theoretically perfect circuit or noise figure - Of a two-port transducer an i-f filter response curve.
component; a lossless transformer or the ratio of the total noise power to the rise time - The time required for a pulse
device that functions without any faults. input noise power, when the input ter- or waveform to reach the peak value
insertion loss - That portion of a signal, mination is at the standard temperature from some smaller specified value.
current or voltage which is lost as it of 290 K. rms - Root mean square. The square
passes through a circuit or device. The root of the mean of the square of the
loss of power through a filter or other nominal - A theoretical or designated voltage or current during a complete,
quantity which may not represent the
passive network. actual value. Sometimes referred to as cycle.
(nterpolate - To estimate a value be- the "ball-park value." rotor - A moving rotary component
tween two known values. within a rotation-control device. Not to
, leakage _ The flow' of signal energy op amp - Operational amplifier. A high- be confused with a rotator, which is the
gain, feedback-controlled amplifier, Per-
beyond a point at which it should not formance is controlled by external cir- total assembly.
be present. Example: Signal leakage saturation - A condition which exists
across a filter because of poor layout cuit elements. Most' op amps are used when a further change in input pro-
(stray coupling) or inadequate shielding. as dc amplifiers. duces no additional output (a saturated
open loop - A, -signal path which does
linear amplification - The process by not contain feedback (see closed loop). amplifier).
which a signal is amplified without parameter _ The characteristic behavior selectivity - A measure of circuit capa-
altering the characteristic of the input of a device or circuit. such as the bility to separate the desired signal from
waveform. Class A, AB and B ampli- operating characteristics of a 2N5109 those at other frequencies.
fiers are generally used for linear ampli- transistor. shunt - A device placed in parallel
, fication. with or a<!ross part of another device.
load _ A circuit or component that parametric amplifierJ..... Synonym for "re-
Examples: Meter shunts, shunt-fed ver-
receives power; the power which is actance amplifier." An inverting para- tical antennas and a capacitor placed
delivered to such a circuit or component. metric device for amplifying a signal (shunted) across another capacitor.
without frequency translation from
Example: A properly matched antenna input to output. , solenoidal - A single-layer coil of wire
is a load for a transmitter., . configured to form a long cylinder.
loaded - A circuit is said to be loaded I parasitic - Unwanted condition or
spectral purity - An emission which
when the desired power is being deli- quantity, such as parasitic oscillations
vered to a load. ' or parasitic capacitance; additional to contains essentially the desired signal
the desired characteristic. component, with all mixing products
logic - Decision-making circuitry of the , -, Operat'mg Wl'thou t an opera-
passIve and harmonics attenuated greatly.
type found in computers. ting voltage. Example: An LC filter standing-wave ratio - The ratio of the
lon!h~~eis~n~ :~~~~~~~~~ ~~r~r~~~:~~~ ~f~~~. contains no amplifiers, or a diode amplitude of a standing' wave at an
anti node to the amplitude at a node.
size. A long piece of wire does not (less PEP _ Peak envelope power; maximum The SWR of a uniform feed line is ex-
than one wavelength) qualify as a long amplitude that can be achieved with pressed as
wire. I
low end _ See cold end. any combination of signals.
permeability - A term used to express 1 +,p
low level - Low-power stage or stages of relationships between magnetic induc- t=P
a circuit as referenced to the higher- tion and magnetic force.
power stages (see high level). pill - Slang expression for a transistor where P is the reflection coefficient.
low pass - A circuit property which or an IC. strip - General term for two or more
permits the passage of frequencies be- PL - Private line, such as a repeater stages of a, circuit which in combina-
low a specified frequency, but atte- which is accessed by means of a speci- tion perform a particular function.
nuates or, blocks those frequencies fied tone. Examples: A local-oscillator strip, an
above that frequency (see high pass). PLL - Phase,locked loop type of oscil- audio strip or an i-f strip ..
low-Z - Low impedance (see high-Z). lator. subharmonic - A frequency that is an
mean - A value between two specified port - The input or output terminal of a i~!egral submultiple of a frequency to
values; an intermediate value. circuit or device. which it is referred. ~ misleading term
master oscillator - The primary oscilla- prototype - A first fuli-scale working which implies that subharmonic energy
3·21 Chapter 3
,,
can be created along with harmonic the transceiver to which it is connected unloaded - The opposite condition of
energy (not true). More aptly, a 3.5- as a basic signal source. ,Example: A- loaded.
MHz VFO driving a 4O-rneter transmit- .2-meter transverter used fn combina- varpctor - A two-terminal semiconduC-
ter, with 3.5-MHz leakage at the out- tion with an hf-bandtransceiver. tor device (diode) which exhibIts a
put, qualifying as a subharmonic. ' Transmatch·- An LC network used to voltage-dependent capacitance. Used
tank - A circuit consisting of inductance effect an iml?edance match between a primarily as a tuning device or fre-
and capacitance; capable of storing transmitter and a feed line to an anten- quency multiplier at vhf and uhf.
electrical energy over a band of fre- na. Not an "antenna 'tuner", or "an- veo - Voltage-controlled oscillator.
quencies continuously distributed about tenna coupler." Uses tuning diodes which have variable
a single frequency at which the circuit trap - A device consisting of Land C dc applied to, change their junction
is said to be resonant, or tuned. components which permits the blockage capacitances.
toroidal - Do"ghnut-Ilhaped physical of a specified frequency while allowing VSWR ~ Voltage standing-wave ratio.
format, such as a toroid core. the passage of other frequencies. (See SWR.)
transducer - A device which is used to Example: A wave trap or an antenna VU ...:.... Volume Unit.
transport energy from one .system trap. VXO - Variable crystal oscillator.
(electrical, mechanical or acoustical) to triftlar - Same as biftlar, but with three Zener diode - Named after the inventor.
another. Example: A .loudspeaker or . parallel conductors. - A diode used to regulate voltage or
phonograph pickup. trigger - To initiate action in a circuit function as a clamp or clipper.
transceiver - A combination transmit- 1>Y introducing an energy stimulus from Z - Symbol for a device or circuit
ter and receiver which uses some parts an external source, such as a scope which contains two or more compo-
of the circuit for both functions. trigger. nents. Example: A parasitic suppressor
transverter -:- A converter which permits U - Symbol for unrepairable assembly, \Yhich contains a resistor and an induc-
transmitting and reCeiving at a specified such as an integrated circuit. (UI, U2, tor in parallel (ZI, Z2, etc.). Z is also
~requency apart from the capability of etc.) the symbol for impedance.
e·
Solid~State Fundamentals
l:.e electrical characteristics of solid- quite important when an electron is either between the two circuits and the current
state devices such as diodes and transis- raised or falls between two different then consists of components at two slight-
tors are dependent upon phenomena that energy "states." For instance, when an ly different frequencies. The effect is simi-
take place at the atomic level. While semi- electron drops from one level to a lower lar to the splitting of electron energy levels
conductors can be employed without a. one, energy is emitted in the form of when two atoms are close enough to in-
complete knowledge of these effects, some electromagnetic radiation. This is the teract.
understanding is helpful in various appli- effect that gives the characteristic glow to
. cations. Electrons, which are the principal neon tubes, mercury-vapor rectifiers, and Conductors,Insulators and Semiconductors
charge carriers in both vacuum tubes and even light-emitting diodes. The frequency Solids are examples of large numbers of
semiconductors, behave much differently of the emitted radiation is given by the atoms in close proximity. As might be
in either of the two circumstances. In foregoing formula where E is the dif- expected, the splittirig of energy levels
free space, an electron can be consid~red ference in energy. However, if an elec- continues until a band structure is
as a small charged solid particle. On the tron receives enough energy such that reached. Depending upon the type of
other hand, the presence of matter affects it is torn from an atom, a process called atom, and the physical arrangemlJllt of the
this picture greatly. For instance, an elec- ionization is said to occur (although the component atoms in the solid, three basic
tron attached to an atom has many prop- term is also loosely applied to transitions conditions can exist. In Fig. 3A, the two
erties similar to those of rf energy·' in between any two levels). If the energy is discrete energy levels have split into two
tuned circuits. It has a frequency and divided by the charge of the electron (-1.6 bands. All the states in the lower band are
wavelength that depend upon atomic X 1019 coulombs), the equivalent in "occupied" by electrons while theones in
parameters just as the frequency associat- voltage is obtained. the higher energy' band are only partially
ed with electrical energy in a tuned circuit. A common way of illustr,ating these filled.
depends upon the values of inductance energy transitions is by means of the In order to impart motion to an
and capacitance. energy-level diagram shown in Fig. lAo It electron, the expenditure of energy is
A relation between the energy of an should be noted that unlike ordinary requited. This means an electron must
E21 .E2E2t;
" energy levels can be carried even further
in this case.
Consider the two identical circuits tliat
electron, it must be raised from the lower
level to the upper one. Since this requires
considerable energy, the material is an
+= 'Ef are coupled magnetically as shown in Fig. insulator. (The energy-level represen-
E1"
• I
2A. Normally, energy initially stored in tation gives an insight into the phenomena
, Cl would oscillate back and forth between of breakdown. It the force on an electron
L I and C 1 at a single frequency after the in an insulator becomes high enough
(A) (B) switch was closed. However, the presence because of an applied field, it'can acquire
of the second circuit consisting of L2 and enough energy to be raised to the upper
Fig. 1 - Energy-level diagram of a single atom is C2 (assume L1 equals L2 and Cl equals C2) level. When this happens, the material
shown at A. At B. the levels split when two atoms are results in the waveform shown in Fig. 2B. goes into a conducting state.)
in close proximity. Energy also oscillates back and forth A third condition is shown in Fig. 4. In
4-1 Chapter 4
,
~
..,
)-
...
II::
...z
;'
...
J== .,b
~ ~
Fig. 2 - EI$ctrical-circuit analog of coupled atoms.
(B)
Fig. 4- Semiconductor energy-I$vel representation.
N
•
CONDUCTION BAND
@§§~~--
Wd ~DONOR LEVEL
the material associated with this diagram, excess of mobile electrons is called an
the upper level is unoccupied but is very n-type semiconductor. WI~------------
4~3 Chapter 4
are fewer 'electrons inp-type material,
the.y are termed the minority carriers.
N VB P
Referring to Fig. lOA, there are some
holes in the n region (indicated by the + ~---x--.l N P
++
signs) because of the foregoing thermal +++++ I
10
effects. Those neat the' junction will
.experience a.force caused by' the electric
10-
-IT
Vf~------~--------~
I
field associated with the barrier voltage. ++' +++++
This field will produce a flow of holes into _! __ l~ __
the p region and the current is denoted by
IT. Such a current is called a drift current
as compared to the diffusion current ID.
VI
r--
r-r- Vx
-
VI
~
Under equilibrium conditions, the two 51
~:l
currents are equal and just cancel each .1
I
other. This is consistent with theassump- ,[+
tion that no current flows in the external 'I
circuit because of the fact that the Fermi (Al r; VX
I = Is(e q~x
where
q =1.6 X 10-" coulombs (the
fundamental charge of an electron),
k =1.38 X 10- 23 joules/Kelvin increased. Consequently, charge carriers as possible in a practical diode since it
(Boltzmann's constant), . must Overcome a large "potential hill" .would only degrade rectifier action. Also,
t =
junction temperature in Kelvins, and the diffusion current becomes very since it is the result of the generation of
e =
2.718 (natural logarithmic base) small. However, the drift currenf caused thermal carriers, it is quite temperature
Vx =applied emf, and Is =
reverse- by the thermally generated carriers returns sensitive which is important when the
bias saturation current. . to the value it had under equilibrium con- diode is part of a transistor. If the reverse
ditions. For large values of V x' the current voltage .is increased further, an effect
This equation is discussed in greater detail approaches Is, defined as the reverse called avalanche breakdown occurs as in-
in the section dealing with common silicon saturation current, Is is the sum of IT and dicated by the sudden increase in current
diodes. its counterpart in the lower or "valence" at Vb' In such an instance, the diode
band. Finally, the characteristic curves of might be damaged by excessive current.
The Reverse-Biased Diode the forward- and reversed-bias diode can However, the effect is also useful for
If the source, V x' is reversed as shown be constructed and are shown in Fig. IOD. regulator purposes and devices used for
in Fig. IOC, the barrier voltage is It is obvious that Is should be as small this purpose are called Zener diodes.
w",~.m'
troublesom~ to the designer. Where a reliable from rf small-signal applications
given junction type of diode may exhibit a to de power use,
capacitance of several picofarads, the
point-contact device will have an internal
TL~luPPORT
lAD
Silicon diodes are available in ratings of
1000 volts (PRV) or greater. Many of
CATWHISKER ~~~~~NIUM
capacitance of I pF or less. these diodes can accommodate de in ex-
(AI cess of 100 amperes. The primary rule
Selenium Diodes N-TYPE in preventing damage to any diode is to
Power rectifiers made frem selenium
WIRE
LEAD
r----~/........., MATERIAL . operate the device within the maximum
were in common use in ac power supplies
before 1965. Today they are manufac-
tured for replacement purposes only.
Selenium diodes ate characterized by high
" eli):,...
P-TYPE
MATERIAL
ratings specified 'by the manufacturer.
The device temperature is one of the im-
portant parameters.' Heat sinks arc used
with diodes that must handle large
forward voltage drop (which increases (B) amounts of power, thereby holding the
with age) and high reverse leakage cur- diode junction, temperature at a safe level.
rent. The voltage drop causes the device to . The behavior of junction diodes under
dissipate power, and a typical rectifier varying temperatures is of interest to
SYMBOL
stack has large cooling fins. An additional (e) designers of circuits that must perform
shortcoming of selenium rectifiers is that over some temperature range. The rela-
they sometimes emit toxic fumes when tionship between forward bias current,
they burn out. When replacing selenium Fig. 12 - A point-contact type of diode is seen forward bias voltage and temperature is
diodes with siticon units, be certilln that at A. A junction diode is depicted at B and the defined by the classic diode equation:
the filter capacitors (anq the' entire diode symbol is at C.
equipment) can withstand the higher. out-
put voltage. Some early solar cells were
made of selenium, but silicon devices have
If = IJe ~: -1)
taken over this area, too. .
~~
.. 50
40
/ . where q is the fundamental electronic
charge (1.6 X 10- 1' coulombs), V is the
Germanium Diodes
.. I-
~i:i
/ bias potential" k is Boltzmann's constant
The germanium diode is characterized o~
L1..a 20
30
/ (1.38 X 10- 23 joules/Kelvin), (Kelvin =
°Celsius +273), t is. the junction
by a relatively large current flow when BACK VOLTS to
I temperature in Kelvins, . Is is the reverse-
small amounts of voltage are applied in -6 o -!l0 -40 -30 -20 -to ./
the "forward" direction (Fig. II). Small bias saturation current, If is the forward-
currents will flow in the reverse. (back)
.,.,r I-' t 2
-tOO F~~~~~O·
3 4
bias current, and e is the natura!'
direction for much. larger applied /' logarithmic base (2.718). The ratio q/k is
'"'" ..
-200 l:!~
approximately 11,600, so the diode equa~
. voltages. A representative curve is shown / -300 ~o:"
in Fig. 13. The dynamic resistance in tion can be written:
either the forward or back direction is . 1I,600V )
determined by the change in current that Fig. 13 - Typical point-contact diode (ger-
manium) characteristic curve. Because the
If = Is ( e --t-·-· - I (Eq, I)
occurs, at any given point on the curve, back current is much smaller than the torward
when the appJie~ voltage is changed by a current, a different scale is used for back It is useful to have an expression for the
small amount. The forward resistance will voltage and current. voltage developed across the junction
4·5 Chapter 4
when the forward current is held constant. but it's a simple matter (with the aid of a easily by using diodes or transistors in
To obtain such an expression wc must pocket calci.llator) to extract the informa- , place ,of mechanical switches or relays.
solve the diodc equation for V. Expanding tion directly from Eq. 9. If the forward The technique is not a complicated one at
the right side Qf Eq. I yields: current is fixed at 1 milliampere, the diode dc and audio frequencies when large
11,600V drop at room temperature is 0.5955 volts. amounts' of power are being turned on
(Eq. 2) This potential decreases at an initial rate and off, or transferred from one circuit
of 2 millivolts per Kelvin. The point to another.
Adding Is to'both sides givcs: temperature coefficient gradually in- Examples of shunt and series diode
11,600V
creases to 3 millivolts per Kelvin at 340 switching are given in Fig. 14. The
Kelvins. While the temperature curve isn't illustration at A shows a IN914 rf-
If + I, = Isc t
(Eq.3) linear, it is gradual enough to be con- switching type of diode as a shunt on-off
Dividing through by Is produces: sidered linear over small intervals. When element between C I' and ground. When
the bias current is increased to 100 , + 12 volts are applied to DI through RI,
11,600V milliamperes, the room temperature junc- the diode saturates and effectively adds
tion potential increases to 0.7146 volts as C 1 to the oscillator tank circuit. R I
....!L
Is
+ I =e (Eq.4)
might be expected, but the temperature sho\lldbe no less than 2200 ohms in value
which implies coefficient stays well-behaved. The initial to prevent excessive current flow through
potential decrease is 1.6 millivolts per the diode junction.
Kelvin, and this value increases to 2.5 Series diode switching is seen in Fig.
1I,6ooV = ln (~ + I) (Eq.5) millivolts per Kelvin at 340 Kelvins. 14B. In this example the diode, DI, is
t 15 The significance of the very minor inserted in the audio signal path. When S 1
t dependence of temperature coefficient on is in the ON position the diode curren,t
Multiplying each term by -:-:~:-::-- bias current is that it isn't necessary to use path is to ground through R2, and the
11,600
an elaborate current regulator to bias diode saturates to become a closed switch.
leaves:
diodes used in temperature compensation When SI is in the OFF state Rl is
applications. The equations defining the grounded and + 12 volts are applied to the
(Eq.6) behavior of junction diodes are approxi- diode cathode. In this mode D 1 is back
mations. Some of the' voltages were ex- biased (cut off) to prevent audio voltage
pressed to five significant figures so the from reaching the transistor amplifier.
The undetermined quantity in Eq. 6 is Is, reader can verify his calculations, but this This technique is useful when several
the reverse saturation current. In ordinary much precision exceeds the accuracy of stages in a circuit, are controlled by a
silicon signal diodes this current approxi- the approximations. single mechanical switch or relay. Rf
mately doubles with each 4.5 Kelvin- circuits can also be controlled by means of
temperature increase. A mathematical ex-
pression for this behavior as a function of
.
Diodes lIS Switches series diode switching.' ,
A significant advantage to the use of
Solid-state switching is accomplished
temperature is:
15(1) = 21 5(1_4.5) (Eq.7)
At room temperature (300 Kelvins), the
reverse saturation current is the order on OSC.
of 10- 13 amperes. Eq. 7 describes a
1--'~-O+12V
phenomenon similar to radioactive decay,
where the 4.5-Kelvin current-doubling in-
terval is analogous to the half-life of a
radioactive substance. This equation with
C1
the given initial condition sets up an
initial-value problem, the solution of
which is:
(I - 3(0)102 Rl
4.5 (Eq.S) +12V
2200 01
lN914
Substituting this expression for Is into Eq.
6 produces the diode voltage drop as a SHUNT SWITCHING
function of temperature for a constant (Al
AF AMP.
current: 2.uF r ~.--_ _ _-+-_+--I~SIG.
SIG.~ r+_.......-I°*1_.---11-......_ _.---+-1 r-'-'OUT
t IN~
V(I) = ~:---,-:~
11,600
X lN914
'lnt IO- D e
(I - 3(0)102
4.5
+] (Eq.9) +12Vo---'4--o--
0-+-+-_.....
+12V
SERIES SWITCHING
(8l
The temperature coefficient of the junc-
tiQn potential Can be obtained from the Fig. 14 - A silicon·switching diode, 01, is used at A to place C1 in the circuil. At B is seen a
partial derivative of V with respect to t, series switch with 01 in the signal path.
DloUIlII'J'olt",e Re/MWtIcl8
Zener diodes are discussed later in this +9.1V(REG.)
chapter. They are used as voltage refer-
ences or regulators. Conventional junc- VFO
tion diodes can be used for the same pur-
poses by taking advantage of their barrier~
voltage characteristics. The greater the
voltage needed, the higher the number of
diodes used in series.. Some examples of
.....----+---1 ~OUTPUT.
...._ _-4
this technique are given in Fig. 16. At A
the diode (Dl) establishes a fixed value of
forward bias (0.7 V) for the transistor,
thereby functioning as a regulat,or. R I is
chosen to permit a safe amount of current
to flow throu~h the diode junction while
1· 10k
it is conducting at the barrier voltage.
The circuit of Fig. 16B shows two
diodes inserted in the emitter return of a
relay-driver transistor. pI and D2 set up a
cutoff voltage of approximately 1.4. This
reduces the static current of the transistor
when forward bias is not provided at the
transistor base. If too much static current
flows the relay may not drop out when the Fig. 17 - High-speed switching diodes of the 1N914 yarlety can be connected back to back and
forward bias decays across the timing used 'as tuning diodes. As the .reverse "oltage 'is' varied by means of R1, the Internal capaCitance
network. The more sensitive the relay the of the diodes will, change.
4.7 Chapter 4
VFO
2N4416
+9.1V (REG.)
47
~----------------~----;~OUTPUT
-4:r
(A)
SPEECH
AMP. v
MPFI02 AF AMP.
l"~~
1501 ,1---_--.....-+-1 02
TO
AF FILTER
lOOk
100
+12V
(8)
Fig. 18 - 01 serves as a bias stabilization device at A (see text). At 8, 01 and 02 are employed as clippers to flatten the positive and Aegative af
peaks. Clipping will occur at roughly 0.7 volt if silicon diodes,are used. Audio filtering is required after the'clipper to remove the harmonic currents
caused by the diode action.
greater the chance for such a problem. D 1 lowered. In the circuit shown here the The circuit of Fig. 18B shows how a
and D2 prevent relay dropout problems of capacitance will vary from roughly 5 pF pair of diodes can be connected in
this variety; D3 is used as a tr.ansient to 15 pF as Rl is adjusted. The diodes back-to-back fashion for the purpose' of
suppressor. A spike will occur when the used in circuits of this kind should have a clipping the negative and positive sine-
relay coil field collapses. If the amplitude high Q and excellent high-frequency wave peaks in an audio amplifier. If
of the spike is great enough, the transient, characteristics. Generally, tuning diodes germanium diodes are used at D 1 and D2
while following the de bus in a piece of are less stable than mechanical variable (lN34As or similar) the audio will limit at
equipment, can destroy transistors and capacitors are. This is because the diode roughly 0.3 V. With silicon diodes (lN914
diodes elsewhere in the circuit. In this junction capacitance will change as the or rectifier types) the voltage will not
application the diode (D3) can be ambient temperature varies. This circuit is exceed 0.7 V. Rl serves as the clipping-
regarded as a clamp. since it clamps the not well suited to mobile applications level control. An audio, gain contrQl is
spike at approximately 0.7 volt. because of the foregoing trait. normally used after the clipper: filter,
along with some additional gain stages .
. Using Diodes as Capacitors Diode CUppers an~ C1IImps The output of the clipper must be filtered
Later in this chapter there is a The previous mention of diode clamp- to restore the sine-wave if distortion is to
discussion about VVC (voltage-variable ing action (D3 in Fig. 16) suggests that be avoided. Diode clippers generate
capacitor) diodes. They are known also advantage can be taken of the charac- considerable harmonic currents, thereby
as tuning diodes and Varicap diodes. It is teristic barrier voltage of diodes tp clip requiring an RC or LC type of audio
possible, however, to use ordinary silicon or limit the amplitude of a sine-wave. filter.
diodes as voltage-variable capacitors. This Although there are numerous applications
is accomplished by taking advantage of in this general category, diode clippers are Diode Frequency Multipliers
the inherent changes in diode junction more familiar to the amateur In noise Designers of rf circuits use small-signal
capacitahce as the rev~rse bias applied to limiter, audio limiter and audio compres- diodes as frequency multipliers when they
them is changed. The primary limitation sor circuits. Fig. 18 illustrates some want to minimize the number of active
in using high-speed switching diodes of typical circuits which employ small-signal devices-(tubes or transistors) in a circuit.
the IN914 variety is a relatively10w maxi- diodes as clamps and clippers. D 1 in Fig. The primary disadvantage of diode'multi-
mum capacitance. At a sacrifice to low 18A functions as a bias ~lamp at the gate pliers is a loss in gain compared to that
minimum capacitance, diodes can be used of the FET., It limits the positive sine-wave which is available from an active multi-
in parallel to step up the maximum avail. swing at approximately 0.7 V. Not only plier. Fig. 19 contains examples of diode
able capacitance.· An example of two does the diode tend to regulate the bias frequency. multipliers. The circuit at A is
IN914 silicon diodes lb a diode tuning cir- voltage, it limits the transconductance of useful for obtaining odd or even multiples
cuit is given in Fig. 17. As R 1 is adjusted the FET during the positive half of the of the driving voltage. The efficiency of
to change the back bias on D 1 and D2, cycle. This action restricts changes in this circuit is not high, requiring that an
there will be a variation in the junction transistor junc,tion capacitance. As a amplifier be used after the diode multi-
capacitance. That change will alter the result, frequency stability of the oscillator plier in most applications. Resonator
VFO operating frequency. The junction is enhanced and the generation of LI/Cl must be tuned to the desired
capacitance increases as the back bias is harmonic currents is greatly minimized. output frequency. -
4-9 Chapter 4
Ni- SILICON WAFER
DI
GOLD-PLATED
PIG TAIL
CARRIER
• \
INPUT
(50 OHMS)
Rl
Dsa
OUTPUT
02 RFC
CONTACT
(SOLDERED)
EI!ITAXIAL
SILICON
CARRIER T2 Dsa
INPUT
(50 OHMS)
• Dl OUTPUT
(50 OHMS)
REJERSk
HORIZONTAL: '10V/D IV.
! ,II I
D2
04 VERTICAL: 5 mA/DIV. i HCD
i V
,!.
~I
I,
FORWARD.I
HORIZON1'AL:O.2V/DIV.
VE~TICAL:5mA/DIv.
/
il/ 1 PINT
AF
.d CONTACT
INPUT /' V
(B)
CI
HCD
DI
SIGNAL
INPUT
(50 OHMS)
\\,
L.OSS
!hI!
S '\
04
I •
DIODE
--
.....NOISE FIGI)fE
.... ~
/
Fig. 21 - The examples at A and B are for use in balanced modulators. The similarity between Fig. 24 - Noise figure and the conversion loss
these and balanced mixers is shown at C and D. of a typical HCD that has no bias applied.
and D for the purpose of illustrating the toroidal types. They provide a broadband at vhf and higher. Notable among the
similarity between balanced modulators circuit characteristic. good features of this type of diode are its
and mixers. It is evident that product higher operating frequency and lower
detectors, balanced modulators and mixers Hot-Carrier Diodes conduction voltage compared to ap-n
are of the same family. The diodes in all One of the more recent developments in junction diode such as the IN914.
examples can be hot-carrier types or the semiconductor field is the hot-carrier When compared to a point-contact
matched silicon switching diodes of the diode, or "HCD." It is a metal-to- diode, the HCD is mechanically and
IN914 class. semiconductor, majority-carrier cQnduc- electrically superior. It has lower noi~e,
CI and C2 in Fig. 2lC and D are u~ ting device with a siitgle rectifying $Teater conversion, efficiency, larger square-
for balancing purposes. They can be junction. The carriers are typically high- lqw capability, higher breakdown voltage,
employed in the same manner with the mobility electrons in an n type of and lower reverse current. The internal
circuits at A and B'. The transformers in semiconductor material. The HCD is capacitance of the HCD is markedly
each illustration are trifilar-wound particulatly useful in mixers and detectors lower than that of a p-n junction diode
24
9 \ '_
22
5
1\
7\
\ \ \5
plication Note APP-177 and in the
Hewlett Packard Application Note 907_
1\ \ \ \ 1\ 1\ Varactor Diodes
\ \ \\
20 Mention was made' earlier in this
In
" 18 1\ \ \ chapter of diodes being used as voltage-
variable capacitors, wherein the diode
junction capacitance can be changed by
'"::>
0::
C>
iL 16
1
1\ 1\ \ I \ \3
1\ varying the reverse bias applied to the
diode. Manufacturers have designed cer-
""~\ \ 1\ \ \ \ \
'-...
'"0z'" \ tain diodes for this application_ They are
called Varicaps (variable capacitor diodes)
\ \ \ \\
14
~ \ \
or varactor diodes (variable reactance
~ 1\ diodes). These diodes depend upon the
--- --~
12
\~ ~
~ \ \ \ l\ ,\ \ change in capacitance which occurs across
~ their depletion layers_ They are not used as
10 t--.
\ rectifiers.
K'- '\l---~ \ \ \
1
8
- ,
t---
......... r-.. \.
5<
Varactors are designed to provide
various capacitance ranges from a few
28
26
IFwoNF-3dB
24
.,
Vee -IV
LO - 900MHz
22
.~ ~
Fig. 25 ~ Curves for hot-carrier diode noise
\~
~ 9
figure versus local-oscillator drive power. The 20
bias currents are in mA as measured at pOint A
in the representative test circuit. ~~~
'" 18
:l'" ~~~ 11"
and it is less subject to temperature
z16
o
'"~14 VCC-3~
~" ~ "" i". 15\
'"
0:: 1\ 19
variations.
~
~~ ~ "- ~ \ \
Fig. 22 shows how the diode is z
'"
o 13\
u
structured internally. A typical set of
curves for an HCn and a p-n junction
diode are given in Fig. 23. The curves
show the forward and reverse charac-
12
10
~ ~ ~ :::--... ~ """ ~\ ~\ ~ l\ 1\
~
~ ~ ~ ~ ~ "- 1\1\
teristics of both diode types_
Fig_ 24 illustrates the noise figure and 8 .......
conversion loss of an HCD with no bias
~ 1'--
~ \.
~pplied. When forward bias is applied to
the diode, the noise figure will change 6
... " ""'\. ~
from that which is seen in Fig. 24. Curves
~
"" ~
for various bias amounts are seen inFig. 4
0.05 0.10 1.0 10.0 15.0
25. The numbers at the ends of the curves LOCAL OSCILLATOR DRIVE (!"W)
signify the amount of current (-in rnA)
flowing, into the test circuit at point A_ / Fig. 26 - local-oscillator drive power versus conversion loss for a specified bias amount.
A set of curves showing conversion loss Bias currents are in mA as measured at point A of the circuit in Fig. 25.
4-11 Chapter 4
200
fl.
100
70
r-
- .............
Fig. 28 - Representative circuit of a varactor
diode showing caser4tslstance, Junction
resistance and Junction capacitance.
............
,
............
50
....... 1'-. equivalent circuit the diode junction
......... consists of q Gunction capacitance) and
..............
'~ r- ~le7e
Rj junction resistance). The bulk resis-
tance is shown as Rs. For the most part
30
I"-....... ~ Rj can be neglected. The performance of
the 'diode junction at a particular fre-
~
r'-..... b-.,. ~~ quency is determined mainly by Cj andRs.
.e 20
As the operating frequency is increased,
III
u
Z
;!
- - ... ~
.........
r-.. ....... ~ r-..... r-
the diode performance degrades, owing to
the transit time established by q and Rs.
~
...............
An important characteristic of the
0
U
C
IL
C varactor diode is the Q, or figure of merit.
u ............... ........... The Q of a v~ractor diode is determined
.; 10
......... .........
by the ratio of its capacitive reactance (Xj)
...... ..........
and its bulk resistance, Rs, just as is true
.........
7
~
r- of other circuit elements, such as coils and
~le66
capacitors, where Q = X/Rs at a specified
TA- 25-C :-........... frequency. Fig. 27B characterizes the Q of
5 f-l.0MHz
.............. three Motorola varactor diodes (versus
4
............. reverse bias) at 50 MHz.
........... Present-day varactor diodes operate
~
into the microwave part of the spectrum.
3 They are quite efficient as frequency
multipliers at power levels as great as 25
watts. The efficiency of a correctly
2 designed va:ractor multiplier exceeds 50
percent in most instances. Fig. 29 il7
lustrates'the basic circuit of a frequency
multiplier which contains a varactor
diode. D 1 is a single-junction device
1
which serves as a frequency tripler in this
0.6 1.0 2.0 4.0 6.0 10.0 20.0 40.0 60.0 example. FLl is required in ordert.o
VR, REVERSE VOLTAGE (VOLTS) assure reasonable purity of the output
(A) energy. It is a high-Q strip-line resonator.
100 Without FLlin the circuit there would be
considerable output energy at 144, 288
80 and 864 MHz. Similar circuits are used as
./ /
MV1BII......
//
, doublers, quadruplers and higher.
40
~v / V A Motorola MVI04 tuning diode is
00 used in the circuit of Fig. 30. It contains
.! 20
MV1~72
V/ /" two varactor diodes in a back-to-back
~
I-
a: 1lI-2!1"C arrangement. The advantage in using two,
III
:I' f-!lO MHz
V'1:B78 diodes is .reduced signal distortion, as
"- 10 compared to a one-diode version of the
0
III same circuit. Reverse bias is applied
..
II:
:>
;;:
8
4 . / IL
L
./
,/
equally to the two diodes in the three-
terminal device. ,Rl functions as an rf
0
f'
L V" isolator for the tuned circuit. The reverse
bias is varied by means of R2 to shift the
operating frequency.. Regulated voltage is'
as important to the varactor ,as it is to the
FET oscillator if reasonable frequency
1
1 2 4 I 10 20 40 10 100 stability is to be assured. Varactor diodes
VII. REVERSE VOLTAGE IVOLTS) are often used to tune two or more circuit's
(8) at ·the sam,e time (receiver rf amplifier,
mixer and oscillator), using a single
potentiometer to control the capacitance
Fig. 27 - Reverse voltage respective to diode capacitance of three Motorolavaractor diodes (A). of the diodes. It is worth mentioning that
Reverse voltage versus diode Q for the varactors at A are shown at B. some Zener diodes and selected silicon
Solid State Fundam,ntall 4·12
FL1
~
432 MHz
. NO.2 LOW RESISTIVITY GoAl
o NO.1 n GaAI" .... o.5hcm
t-2,um
f •
144 MHz '"j' /t
7
8 LOW RESISTIVITY GoAl
""S,umFOR
16GHz
~
..... O.OO1Ji.- em ,"",Oum FOR
10 GHz
144-MHz
INPUT D1
Fig. 31- Cross-sectional illustration of Gunn diode.
(50 OHMS)
6.50 ~
S.8S g
f\
'\.
f'.",
4.55 ~
..
S.20 ;;;
..J
3.90 g
Fig. 29 - Typical circuit for a varactor-diode frequency tripler. I-.....
::!: i
8 10 12
FREQUENCY (GHZ)
---
14 16 f8
1.95
~
:3
I-
VFO
2N44l6 Fig. 32 - Active region thickness versus frequency
270 of a Gunn diode.
3.5 MHz
r - - -......- - ;
lOOk
"
Rt
toOk
R2
1 \ r - - - - - - - - - - - - - - - 4 I - - - - - O . +9V (REG.)
power-supply rectifier diodes will work solid-state devices it is possible to generate The composite ~afer of Fig. 31 is
effectively as va:ractors at frequencies as useful power up to' 35 G Hz. metalized on both sides to permit bonding
high as 144 MHz. If a Zener diode is used Amateurs have been using Gunn diodes into the diode package. This process of
in this manner it must be operated below at 10 GHz, but it is important to realize metalization also ensures a low electrical
its reverse breakdown voltage point. The that these diodes are available for other and thermal resistance. The completed
stud-m6unt variety of power supply microwave frequencies. An in-depth treat- chip is bonded to agQld-plated copper
diodes (with glass headers) are reported to ment of the technology and chemistry of pedestal, with layer no. 2 next to the heat
be the best candidates as· varactors, but Gunn diodes is provided in the Gunn sink. A metal ribbon is connected to the
not all diodes of this type 'will work. Diode Handbook by Microwave As- back side of the diode to provide for
effectively: Experimentation is necessary .. sociates, Incorporated. electrical contact.
Additional data concerning varactor diodes Fig. 31 shows a cross-sectional re- The curve in Fig. 32 shows the rela-
can be found in the Motorola booklet, presentation of a slice of the material from . tionship of the diode active-region thick-
Designer's Manual for VVC Tuning Diodes, which Gunn diodes are made. Layer no. 1 ness. to the frequency of operation. The
LIB 2 Rl. ( is the ,active region of the device. The curve illustrates an approximation be-
thickness of this layer depends on the cause the actual thickness of the active
Gunn Diodes chosen frequency, of operation! For 'the region depends on the applied bias voltage
Gunn diodes are named after the IO-GHz band it is approximately 10 urn and the particular circuit'used'-The input
developer. J. B. Gunn, who was studying (10,6 meters) thick. The threshold voltage power to the diode must be 20 to 50 times
carrier behavior at IBM Corp. in 1963. is roughly 3.3 volts. At 16 GHz the layer the desired output power. Thus the
Durini that period he discovered what is would be formed to a thickness of 8 urn, efficiency from dc to rf IS on the order of
known today as the "Gunn effect." and the threshold voltage would be about two to five percent.
Recently, semiconduetor devices of the 2.6 volts. The resonant frequency of the diode
"bulk effect" variety have become so Layer no. 2 is grown epitaxially and is assembly must be higher than the opera-
practical that in areas where advanced doped to provide low resistivity. This ting frequency to allow for parasitic C
technology is practiced they are com- layer is grown on ~he active region of the and L components which exist. Fig. 33
monplace. Among these newer devices are semiconductor, but it is not essential to shows the equivalent circuit of a packaged
the TDO (tunnel-diode oscillator), the the primary operation of the diode. It is Gunn diode. Assuming a diode natural
. ADO (avalanche-diode oscillator) and the, used to ensure good ohmic contact and to resonant frequency of 17 GHz, the
TEO (transferred-electron oscillator). With prevent metalization from damaging the following approximate values result: Lp =
some of the present-day techniques ·and n-layer of the diode. . 0.25 nH, Ca = 0.15 pFand Q, = 0.15 pF.
4-13 Chapter 4
.. GoAl CHIP
VARACTOR- IMPATT-DIODE
TUNED AMPLIFIER
GUNN-DIODE OSC.
AF INPUTo-~--I
(A)
I,MPATT-DIODE
FM I-F AMPLIFIER
AMP. uP-
CONVERTER FIL TER
4.15 Chapter 4
BIAS
ANT. 1 .A .AA 1
~A)
(A)
BIAS
ANT.
(8)
RFC
L
Fig. 40 - Two types of PIN diode .attenuator
circuits. The circuit at A is called a Bridged
Tee and the circuit at B is a pi type. Both
exhibit very broadband characteristics.
Fig. 39 - PIN diodes are used to transfer a common antenna to either a transmitter or receiver. A
voltage applied to the bias. terminal will switch the system to the transmit mode connecting the
output of the transmitter to the antenna. At'the same time the diode across the receiver input Is
forward biased to a low·resistance state to protectJhe input stage of the receiver. The quarter·'
wave line isolates the low resistance of the receiver diode from the common antenna connection.
At B the quarter·waveline Is replaced with a lumped·element equivalent
Emitted flux density is also defined as spread the flux over a wider viewing area Fig. 41 - A solar·energy diode cell. Electrons
radiation per unit area and is used to and hence, have much less point intensity flow when light strikes. the upper surface. The
describe light reflected from a surface. (luminance) than the point-source diodes. bottom of the cell is coated with foil to collect
This measure of reflectance determines current for the load, or for the succeeding cell
Solar-Electric Diodes in series·connected arrays of cells.
the total radiant or luminous emittance.
Source intensity defines the flux density Sunlight can be converted directly into
which will appear at a distant surface and electricity by a process known as photovol-
is expressed as lumens/steradian (photom- laic conversion. For this purpose a solar generate excess holes and electrons (one
etry) or watts/steradian (radiometry). cell is used. It relies on the photoelectric hole/electron pair fol' each photon absorb-
Luminance is a measure of photometric properties of a semiconductor. Practical- ed). When this occurs near the p-n
brightness and is obtained by dividing the ly, the solar cell is a large-area p-n junction, the electric fields in that region
luminous intensity at a given point by the junction diode. The greater the area of the will separate the holes from the electrons.
projected area of the source at the same cell, the higher the output current will be: This causes the holes to increase in the
point. Luminance is a very important A dc .voltage output of approximately 0.5 p-type material. At the same time the
rating in the evaluation of visible LEOs. is obtained from a single cell. Numerous electrons will build up in the n~type
While luminance is equated with cells can be connected in series to provide material. By making direct connection to
photometric brightness, it is inaccurate to 6, 9, 12, 24 or whatever low voltage is the p and n regions by means of wires,
equate 'luminance as a figure of merit for required. In a like manner, cells can be these excess charges generated ,by light
brightness. The only case where this rating connected in parallel to provide higher (and separated by the junction) will flow
is acceptable is when comparing physically output current, overall. into an external load to provide power.
identical LED:. Different. LEOs' are sub- The solar diode cell is built so that light Approximately 0.16 A can be secured
ject to more stringent examination. can penetrate into the region of the p-n from each square inch of solar-cell
Manufacturers do not use a set of consis- junction, Fig. 41. Most modern solar cells material exposed to bright sunlight. A
tent ratings for LEOs (such as optical use silicon material. Impurities (doping) 3-1/2 inch diameter cell can provide 1.5 A
flux, brightness and intensity). This is are introduced into the silicon material to of output current. The efficiency of a solar
because of the dramatic differences ·in op- establish excess positive or negative cell (maximum power delivered to a load
tical mea~urements between point- and charges which carry electric currents. versus total solar energy incident on the
area-source diodes. Point-source diodes. Phosphorous is used to produce n-type .cell) is typically 11 to 12 percent.
·are packaged in a clear epoxy or set within silicon. Boron is used as the dopant to Arrays of solar cells ·are manufactured
a transparent glass lens. Area-source produce p-type materiaL . for all manner of practical applications. A
diodes must employ a diffusing lens to Light is absorbed into' the silicon to storage battery is used as a buffer between
Zener Diodes
Zener diodes have, for the most part,
replaced the gaseous regulator tube. They
have been proved more reliable than tube
types of voltage regulators, are less
expensive and far smaller in size. - -t - - - - - - - - - - - .---....._1---- VALLEY POINT
These diodes fall into two primary I
I
.classifications: Voltage regulators and --L-r~r-~I-------------~-------Y
. voltage-reference diodes. When they are 30mV
360mV
used in power supplies as'regulators, they I I
provide a nearly constant dc output I I
voltage even though there may be large Vp I+-
I
changes in load re~stance or input I
voltage. As a reference element the Zener I
diode utilizes the voltage drop across its I
junction when a specified current passes I
through it in the reverse-breakdown ~---~--Vv------------J~'I
direction (sometimes called the Zener
direction). This "Zener voltage" is the
value established as a reference. There- Fig. 43 - Schematic symbol and current-vs.-voltage characteristic for a tunnel diode.
4-17 Chapter 4
\ 4!l0
+ R
~1~l
300
OSC.
If (mA)
1!l0 I LOAD
~ SIDE
REVERSE VOLTAGE
~-+-O +42V
30 Z,O .• 10 ./
~::J"
O.!l 1.0 1. 5
FORWARD VOLTAGE
l!l
IzlmA)
30
45
R.
~""",--""-'VV\.~+12 TOl4V
DC
91V ?,:-:)
400mW INP~ +O-~_-I
12V
Rs=
Ein (min.) - ED!
IL +O.li L
12 -9.1 2. 9
1
TRANSIENT PROTECTION
Rs = ---:O:-:I=-~OO~I
(D)
. A+. -
A --:orr
RF AMP
= 264 ohms (Cl 270 0.\
. and
PD(max.)-
_(Ein(max.) R- ED! -IL)EDI
s 0.1
IN 0--1
PD(max)=( .!±.2~49.1 - 0.01) 9.1
low capacitance and short storage time is amplify signal energy (current). The
the step-recovery diode (SRD), sometimes (A) (B)
inherent ,characteristic is one of "transfer-
called a "snap" diode. These diodes are ring current across a resistor." The
used as frequency multipliers well in.to the transistor was invented by Shockley,
Fig. 48 - At A. an n·channel JFET connected
lllicrowave spectrum. Switching the device Bardeen and Brattain at Bell Labs in 1947 as a constant·current source. At B. the
in and out of forward conduction is the and has become the standard amplifying schematic symbol for the circuit in A when it
multiplication mechanism, and the power device in electronic equipment. In rf and is packaged as a two-terminal device.
A
® ® B Power Amplification
Because the collector is biased in the EMITTER
back direction the collector-to-base resis- ALLOY-JUNCTION
TRANSISTOR
tance is high. On the other hand, the
emitter and collector currents are sub-
stantially equal, so the power in the
-«L
COLLECTOR
Fig. 49 - Illustration of a junction pnp tran· collector circuit is larger than the power in BASE
sistor. Capacitances Cbe and Cbc vary with the emitter circuit (P = PR, so the powers , EMITTER
changes in operating and signal voltage (see are proportional to the respective resis- NPN
text).
tances, if the currents are the same). In
--©=
COLLECTOR
practical transistors emitter resistance is BASE
of the order of a few hundred ohms while
how transistors function will help to the collector resistance is hundreds or
EMITTER
4-21 Chapter 4
+v
~
-% -% ~ ~,uF
+ f--<i
OUTPUT
INPUT ~,uF
0--} +
~ ,C
4700
Q1
INPUT
+9V
-v
S,uF
r
OUTPUT
Fig. 54 - Basic transistor amplifiers. Observe the input and o'utput phase relationships for the Blpolllr Transistor Dissipation
various configurations. Apart from the characteristics men-
tioned earlier, it is necessary to consider
the matters of collector dissipation,
collector voltage and current and emitter
collector resistance is some tens of and the cutoff frequency is the same as in current. Variations in these specifications
thousands of ohms, depending on the the grounded-emitter circuit. The output are denoted by specific parameter symbols
signal source impedance. The common- and input currents are in phase. which appear later in the chapter. The
emitter circuit has a lower cutoff frequen- maximum dissipation ratings of transis-
cy than does the common-base circuit, but Differential Amplifier Circuit
tors, as provided on the manufacturer's
it gives the highest power gain of the three An important variation of the fun- data sheets, tend to confuse some ama-
configurations. damental amplifier types is the differential teurs. An acceptable rule of thumb is to
In this circuit the phase of the output amplifier, drawn in Fig. 54A. The output select a transistor which has a maximum
(collector) current is opposite to that Qf voltage is proportional to the difference dissipation rating of approximately twice
the input (base) current so such feedback (with respect to ground) between the the dc input power of the circuit stage_
as occurs through the small emitter voltages applie<;i to the input terminals. That is, if a 5-watt dc input is con-
resistance is negative and the amplifier is With the proper choice of operating con- templated, choose a transistor with a
stable. ditions, several differential amplifier lO-watt or greater rating. When power
stages of the type shown can be cascaded levels in excess of a few hundred mWare
Common-Collector Circuit (Emitter- directly. Fig. 54A shows the circuit in its necessary there is a need for heat sinking.
Follower) classic balanced form, but many circuits A sink is a metal device which helps to
Like the vacuum-tube cathode use differential amplifiers in a single- keep the transistor cool by virtue of heat
follower, the common-collector transistor ended configuration. When only a single transfer from the transistor case to the
amplifier has high input impedance and input and output terminal is required, RI sink_ At power levels below. 5 watts it is
low output impedance. The latter is ap- could be a short circuit and the Q2 base common practice to employ clip-on heat
proximately equal to the impedance of the could be grounded. Under these cir- sinks of the crown variety. For powers
signal input source multiplied by (I - 0:). cumstances the differential amplifier can greater than 5 watts it is necessary to use
The input resistance depends on the load be understood as an emitter-follower driv- large-area heat sinks which are fashioned
resistance, being approximately equal to ing a common-base stage. The output is from extruded aluminum. These sinks
. the load resistance divided by (I - 0:). taken between the Q2 collector and have cooling fins on one or more of their
The fact that input resistance is directly ground. R3 establishes the current in QI surfaces to hasten the cooling process.
related to the load resistance is a disadvan- and Q2, which should be equal under Some high-power, solid-state amplifiers
tage of this type of amplifier if the load is static conditions. employ cooling fans from which the air
one whose resistance or impedaoce varies Differential amplifiers work best .when stream is directed on the metallic heat
with frequency. R3 is replaced by some type of constant- sink. Regardless of the power level or type
The current transfer ratio with this cir- current source. One type of current of heat sink used, silicone heat-transfer
cuit is regulator has been discussed in the diode compound should always be used between
section, and current sources made from the mating 'surfaces of the transistor and
bipolar transistors are covered later. the heat sink. Another rule of thumb is
1-0: With a current source biasing QI and offered: If the heat-sink-equipped transis-
Solid State Fundamentals 4-22
in the small-signal class carry dissipation
Ie
ratings of 500 mW or less. Power
transistors are normally classed as. 500-
mWand higher devices. The practical
applications for all of these semiconduc-
tors range from dc to the microwave spec-
trum.
Bipolar Transistor Switches
Our present-day technology includes
the use of solid-state switches as practical Fig. 57 - Circuit for a tranSistor switching cir·
alternatives to mechanical switches. When cuit (saturated).
a bipolar transistor is used in a switching
application it is either in an on or off
state. In the on state a forward bias is ap- OSC.
Fig. 55 - Typical characteristic for the collec·
plied to the transistor, sufficient in level to
saturate the device. The common-emitter
D~
tor of an npn transistor which shows the three Q2
primary regions involved during swit.ching. format is used for nearly all transistor
switches. Switching action is characterized
by large-signal nonlinear operation of the
EMITTER- COLLECTOR device. Fig. 55 shows typical output
BASE JUNCTION BASE JUNCTION characteristics for an npn switching tran-
EM1TTER BASE COLLECTOR sistor in the common-emitter mode. There
HOLES ELECTRONS 'HOLES
are three regions of operation - cutoff,
active and saturation. In the cutoff region
the emitter-base and collector-base junc-
tions are reverse biased. At this period the
2 AND 3
collector current is quite small and is com-
parable to the leakage current, leeo' leev or
DISTANCE
• lebo'
Fig. 56 illustrates the minority-carrier
concentration relative to an npn tran-
Fig. 56 - Illustration of the minority,carrier .
concentrations in an npntranslstor. No.1' sistor. During cutoff the coricentration is (AI
shows the cutoff region. No.2. is the active zero at both junctions because they are NPN
region at the threshold of the saturation reverse biased (curve no. I). SWITCH
region. No.3 is in the saturation region. The emitter-base junction is forward 01
2N2102
biased in the active region. At this time
the collector-base junction is reverse biased.
tor is too warm to touch with comfort, Fig. 55 shows a load line along which
the heat sink is not large enough in area. switching from the cutoff to the active
Excessive junction heat will destroy a region is done. The transit time (speed)
transistor. Prior to destruction the device through the active region is dependent
may go into thermal runaway. During this upon the transistor frequency-response (B)
condition the transistor becomes hotter characteristics. Thus, the higher the fre-
and its internal resistance lowers. This quency ratin.g of the device, the faster the Fig. 58 - Examples of practical switching cir·
causes an increase in emitter/collector switching time. 'Curve no. 2 in Fig. 56 cuits. A pnp switch is used to key an oscillator
and emitter/base current. This increased depicts the minority-clilrrier concentration at A. When R1 is grounded the switching tran·
current elevates the dissipation and of the active region. sistor is forward biased to !laturation, thereby
permitting current to flow from the dc supply
further lowers the internal resistance. In the saturation region the emitter- line to 02. The circuit at B shows 01 as a
These effects are cumulative: Eventually base and collector-base junctions are for- relay-driver npn switch. When + 12 volts Is ap-
the transistor will be destroyed. A heat ward biased. During this period the for- plied to the base of 01 it is forward biased to
sink of proper size will prevent this type of ward voltage drop across the emitter-base saturation, permitting current to 'flow through
the field coil of K1A. 01 and 02 are included to
problem. Excessive junction temperature junction VBE(sat) is larger than it is across reduce the static collector current of 01, which
will eventually cause the transistor to the collector-base junction. This results in in some instanCes could cause K1A to remain
become open. Checks with an ohmmeter a collector-emitter' voltage termed closed after forward bias was removed from
will indicate this condition after a failure. V CE(sat). Series resistances present in the
01. 03 serves as a spike suppressor when the
field of K1A collapses.
Excess collector voltage will also cause emitter and collector legs of the circuit
immediate device failure. The indication contribute to the determination of
of this type of failure, as noted by means V CE(sat). Since the collector in this state is base breakdown voltage, V(BR)CBO, .should
of an ohmmeter, is a shorted junction. forward biased, additional carriers are in- be no greater than Vee + VBE{ofO. Finally,
jected into the base. Some also reach the the minimum collector-to-emitter break-
Blpow TransIstor Applkations collector. Curve no. 3 of Fig. 56 shows down voltage, V(BR) CERL, must be greater
Silicon transistors are the most com- this minority-carrier concentration. Fig. than Vce. As is true in any transistor appli-
mon types in use today, although a few 57 contains the circuit for a basic cation, the junction temperature must be
germanium varieties are built for specific saturated-transistor switch. maintained at a safe value by whatever
applications. Collector voltages as great It is extremely important to make cer- means necessary.
, 'as 1500 can be accommodated by some of tain that none of the transistor voltage A transistor switch can be turned on by
the high-power silicon transistors avail- ratings is exceeded during the "off" means of a pulse (Fig. 57) or by applica-
able now. Most small~signal transistors period: The minimum emitter-base break- tion of a dc forward bias. Typical circuits
will safely handle collector voltages of 25 down voltage, V{BR)EBO, .must not exceed for the latter are given in Fig. '58. The
or greater. Generally speaking, transistors VBE(ofO' Also,' the minimum collector- circuit at A illustrates how a pnp iran-
4-23 Chapter 4
NPN MIC AMP.
AMPLIFIER 2IJF
2N2222A T5V
(--o~(l;. 2IJF
~~+
+
6v" + f-o H1 - Z
0-)+
MIC AMP.
LO-Z
2N2222A
..M
MIC 1~V
4700 M
+ f-oHI-Z
"-'1"t"
MIC T1
10k
(A). +~
12V +9V
PNP • (Al
AMPLIFIER
+ (-og~~ +9V
~~0-7 + (Sl
47.00
Fig. 60 - RC and transformer·coupled audio amplifiers suitable for high· and low·impedance
microphones. .
4-25 Chapter 4
,I
AMPLIFIER T3
FL1
1;4
T2
50-OHM
!NPUT
' - , _....----0 TO T3
+12V
(A)
(91
Fig. 64 - Broadband transformers are employed at A for impedance matching. FL1 suppresses harmonic currents at the amplifier output. In the
examples at B are feedback components Ctand R1 (see text). '
4-27 Chapter 4
AMPLIFIER
R2 Cl
FL1
0.1
+12V
Fig. 65 - Example of a fed-back, push-pull, rf power amplifier set up for broadband service from 1.8 to 30 MHz. The circuit is biased for linear
amplification.
56k
~LO-Z
~ OUTPUT
+12V +12V
COL PITTS OSCILLATOR PIERCE OSCILLATOR
(AI OVERTONE OSCILLATOR
(BI (CI
may be necessary with some crystals to means of the link shown. Alternatively, a use the smallest amount of feedback that
provide ample feedback to cause oscilla- capacitive divider can be placed across the will provide reliable oscillator, perfor-
tion_ The value of Cib will depend on the . inductor to provide a low-Z tap-off point. mance with the load connected.
operating frequency and the gain of the The trimmer should be retained in parallel Fig. 67B illustrates a series-tuned Col-
transistor. Typically for 1.8 to 20~MHz with the inductor to permit resonating the pitts oscillator, although this general cir-
crystals (fundamental mode) the circuit. cuit is often referred to as a "series-tuned
capacitance value ranges from 25 to 100 Some typical rf and audio oscillators Clapp" oscillator. It is very stable when
pF _ The higher values are typical at the are seen in Fig. 67. The circuit at A ob- polystyrene capacitors are used in the
lower end of the frequency range. In Fig. tains feedback by means of the emitter tap feedback and tuned circuits. Silver-mica
66C is an overtone oscillator. The collec- on the tuned circuit. Approximately 25 capacitors can be used as substitutes at a
tor tuned circuit must be able to resonate percent of the :oscillator rf power is used slight sacrifice in drift stability (long
slightly above the crystal overtone fre- as feedback. The tap point on this type of term). .
quency in order to ensure oscillation. oscillator is between 10 and 25 percent of A twin-T audio oscillator is shown at C
Low-impedance output can be had by the total coil turns. The. designer should in Fig. 67. It is a very stable type of circuit
,.+,O.t
I.B MHz 50
S.M.
f-oTO BUFFER
''''900Hz
3300
+t2V
3300
o---......--'\I'\,/\r--......- - - - ,
C~
0.01 O.Ot
2Nt925,
HEP254
Tl
330 330
c.t.
0.Q1 ~SPKR.
'--~~f-o~~fp:f#VE
C3 (DISTORTION:::: 5 %)
"L0.02
Ct,C2,C3-MYLAR
OR POLYSTYRENE
Fig. 67 - The circuits at A and Bare VFOs for use in transmitters or receivers. Audio oscillators are shown at C and D.
which delivers a clean sine-wave output. semiconductor manufacturers specify cer- with LO voltage applied to the emitter.
Mylar or polystyrene capacitors should be tain transistors for mixer service. Al- This technique requires slightly higher
used for best stability. though this does not mean that other levels of LO energy, but affords greater
A simple feedback circuit is effected by types of bipolar transistors can't be used LO isolation from the mixer input port.
means of TI in Fig. 67D. TI is a small for mixing, it is wise to select a device that A singly balanced bipolar-transistor
transistor output transformer with a is designed for that class of service. mixer is illustrated in Fig. 68 at C. RI is
center-tapped primary and an 8-ohm Fig. 68 contains examples of three basic adjusted to effect balance. This circuit
secondary. This circuit is excellent for use types of transistor mixers. At A is seen the could be modified for emitter injection by
as a code-practice or side-tone oscillator. most common one. It is found in simple changing RI to WOO ohms, replacing the
All of the rf oscillators described in circuits such as transistor a-m broadcast- 220-ohm resistors with l-mH rf chokes,
these examples should be followed by one band. receivers. As an aid to dynamic and injecting the LO output at the junc-
or more buffer stages to prevent frequency range, the mixers of Fig. 68 can be used tion of the two O.OI-",F capacitors. The
changes resulting. from load variations without rf amplifier stages ahead of them center tap of the input transformer (base
occuring after the oscillator chain. for frequencies up to and including 7 winding) would then be bypassed by
MHz: The noise in that range (ambient means of a O.OI-",F capacitor.
Transistor Mixers from the antenna) will exceed that of the
Much of the modern equipment used by mixer. Other Uses for Bipolar Transistors
amateurs contains mixers which utilize The primary limitation fn the perfor- It is possible to take advantage of the
FETs or diode rings. Good dynamic range mance of the mixer of Fig. 68A is that the junction characteristics of small-signal
is offered by those two circuits. However, local-oscillator voltage is injected at the transistors for applications which usually
there is no reason why a bipolar mixer base. This does not afford good employ diodes. One useful technique is
can't be used to obtain satisfactory LO/input-signal isolation. The un- that of employing transistors as voltage-
results if care is taken with the operating favorable result can be oscillator "pull- variable capacitors (varactors). This
parameters and the gain distribution in ing" with input load changes, and/or method is seen in Fig. 69. The collector-
the receiver or transmitter where they are radiation of the LO energy via the anten- base junction of QI and Q2 serve as
used. The bipolar transistors used irl na if the front-end selectivity is marginal diodes for tuning the VFO. In-this exam-
receiver mixers should be selected ac- or poor. The advantage of the circuit is ple the emitters are left floating. A single
cording to noise figure (low) and dynillllic that it requires less injection voltage than transistor could be used, but by connect-
range (high). The signal applied to it the one at B, where emitter injection is ing the pair in a back-to-backarrange-
should be kept as low as possible, used. ment they never conduct during any part
consistent with low-noise operation. Most At Fig. 68B is the same basic mixer, but of the rf cycle. This minimizes loading of
4-29 Chapter 4
ward bias applied to the base-emitter,
, junction, a superdiode results. If the col-
lector were left open, the base-emitter
junction would behave like an ordinary
diode. With the collector tied to the base,
the diode current rises much more rapidly
with applied voltage because of the
amplification provided by the transistor
action. Two cross-connected super-diodes
r~
form the basis fo .. a highly effective peak
clipper or hard limiter. Fig. 71 illustrates
the application. Npn transistors are,
shown, but pnp units will yield identical
5-MHz LO
(IV pk-pk) performance. .
+12V
Constant-Current Generators
(Al The curves in Figs. 51 to 53 show that
the collector current of a bipolar tran-
sistor is essentially independent of the
.collector-to-emitter· potential when the
MIXER device is biased in its active region. Fig.
MPS6547
72A illustrates a constant-current source
TO RF AMP.v--____
(or sink, if actual electron flow is con-
sidered) using a pnp transistor. A fairly
constant 1.2-volt potential drop is main-
tained across the diode string. The base-
emitter junction introduces a diode drop,
so the emf applied to the 62-{l resistor is a ,
constant 0.6 volt. A constant voltage
across a resistor forces a constant current.
This current flows in the emitter, and the
.1°. 01
+12V
high alpha causes the collector current to
be nearly the same.
The cin;:uit of Fig. 728 works, in a
5-MHz LO
(3V pk-pk) similar manner. RI biases QI into conduc-
(Bl
tion. When the emf developed by R2
reaches 0.6 volts, Q2 begins to conduct,
BAL. shunting base drive away from Q 1 and
MIXER
MPS6547 limiting its colIector current.
A device that passes an arbitrary cur-
rent independent of the applied voltage
presents an infinite dynamic impedance to
the driving signal. This feature makes the
constant-current generator valuable in
several applications. One use for the cir-
cuits of Fig. 72 is in the bias control circuit
of a differential amplifier. Either con-
figuration can be used to establish the
proper amplifier current while providing
the tightest possible coupling between the
emitters of the differential pair. Another
way to employ a constant-current circuit
is to use it as an active load for the collec-
tor of a transistor amplifier stage. The in-
finite dynamic impedance of the current
(el source causes the amplifier to exhibit very
high voltage gain. When the amplifier is
Fig. 68 - Some typical bipolar-transistor mixers. Their characteristics are discussed in the text. an npn transistor, the current source must
be a pnp device, and vice-versa.
Thyristors
Two complementary bipolar transistors
connected as in Fig. 73 form the solid-
state analog of the latching relay - a trig-
the oscillator. The junction capacitance is the reverse-breakdown characteristic of ger pulse applied to the base of Q2 will ini-
varied by adjusting the tuning control, QI to establish a fixed reference level. tiate current flow in both devices. This
RI. In this circuit the tuning range is ap- Most transistors provide Zener-diode ac- current is limited only by the external cir-
proximately 70 kHz. tion between 6 and 9 volts. The exact cuit resistance and continues independent
A bipolar-transistor junction can be value can be determined experimentally. of the trigger signal until the main source
used as a Zener diode in the manner When the base and collector of a is interrupted. Four-layer semiconductors
shown in Fig. 70. Advantage is taken of bipolar transistor are connected and a for- (pnpn or npnp) having this property are
Solid State Fundam,ntala 4·30
known as thyristors or silicon controlled
TUNING rectifiers (SCRs). SCRs find use in power
osc. supply overvoltage protection circuits
(crowbars), electronic ignition systems,'
alarms, solid-state commutating systems
for dc motors and a host of other applica-
tions. Two complementary SCRs
fabricated in parallel, with a common gate
terminal, form a triac. These are used to
220k 22k switch alternating currents. The most
common application of the triac is in in-
Q2 candescent light dimmers. Triacs have
2N3053
.--+-.---------1I_O+6.aV REG. sensitive gates, and prolonging the trigger
signal or injecting excessive gate current
can cause excessive. heating. In circuits
,L0m p. POLYSTYRENE operating on 117-volt ac, a diac is used to
trigger a triac. A diac is a bidirectional
current-limiting diode. Structurally, it can
be compared to a triac without a gate. A
Fig. 69 - Bipolar transistors serve as varactor tuning diodes in this circuit (01 and 02). motor speed control illustrating the use of
triacs and diacs is drawn in Fig. 74.
Unijunction Transistors
An unusual three-terminal semiconduc-
tor device is the unijunction transistor
(UJT), sometimes called a double-base
+9V REG.
OUTPUT
f---.---....---I!-o. diode. The elements of a UJT are base I,
REGULATOR base 2 and emitter. The single rectifying
Q1
2N3390
junction is between the emitter and the
+12V silicon substrate. The base terminals are
INPUT
ohmic contacts, meaning that the current
is a linear function of the applied voltage.
Current flowing between the bases sets up
a voltage gradient along the substrate. In
operation, ~he direction of flow causes the
Fig. 71 - A peak clipper circuit using bipolar emitter junction to be reverse biased. The
Fig. 70 - A bipolar transistor will function as transistors connected in the superdiode con· relaxation oscillator circuit (the most com-
a Zener diode when connected as shown here. figuration. mon UJT application) of Fig. 75 il-
lustrates the function of the UJT. When
the circuit is energized, the capacitor
charges through the resistor until the emit-
ter voltage overcomes the reverse bias. As
soon as current flows in the emitter. the
resistance of the base I region decreases
dramatically, discharging the capacitor.
+V ,.-----.---0 + v The decreased base I resistance alters the
voltage distribution along the substrate,
r
1.2 V
T
0.6 V
1
Rl
establishing a new bias point for the emit-
ter junction. As more and more emitter
current flows, the majority carrier injec-
tion builds a space charge in the base I
1k
·v
1
_----ANOOE - - - - _..
62 O.6V
j
10mA
t
(A) (B)
t
10mA
_CATHOOE_
Fig. 72 - Constant-current generators made with bipolar transistors. In A, the reference voltage
established by the diodes is corlvertad to a current by the emitter resistor. A two·transistor feed· Fig. 73 - An SCA and its discrete functional
back arrangement is employed at B. The functions of both circuits are explained in the text. near-equivalenl.
4·31 Chlpter 4
TOP
. SOURCE GATE DRAIN
BOTTOM GATE
# (A)
N-CHANNEL
~
F\AIN
GBOTTOM
(A)
Fig. 75 - A relaxation osciliator based on a unijunction transistor. The frequency of oscillation is
approximately 1500 Hz.
the capacitor and the cycle repeats. If the good dynamic range. These characteristics
resistor were replaced by a constant- apply to small-signal FETs. Power FETs,
current source, the output waveform which will be treated later, have different
would be a, linear ramp instead of a characteristics. Although some MOS-
sawtooth. The UJT schematic symbol FETs have but one gate, others have two
resembles that of an n-channe1 JFET - gates. Single-gate FETs· can be equated
the angled emitter distinguishes the uni- practically to a triode vacuum tube. The
junction transistor. gate represents the grid, the anode is
similar to the drain, and the cathode is (S)
Field-Effect Transistors like the source. The inpu~ impedance of
Field-effect transistors are assigned that FETs is a megohm or greater. The noise
name because the current flow in them is figure of an FET is quite low, making them Fig. 77 - Operation of a JFET Ulider applied
controlled by varying electric field which ideal as preamplifiers for audio and rfwell bias. A depletion region (light shading) is
formed, compressing the channel and Increas·
is brought about through the,application into the uhf region. Nearly all of the ing the resistance to current flow.
of a voltage that controls the electrode MOSFETs manufactured today have
known as the gate. The analogy for a built-in gate-protective Zener diodes.
bipolar transistor is that in the latter the Without this provision the gate insulation
current flow is controlled by the current can be perforated easily by small static
applied to the base electrode. charges on the user's hands or by the between source and drain. This current
There are two essential types of application of excessive voltages. The flow is made up of free electrons since the
field-effect transistors (FETs) in use protective diodes are connected between semiconductor is n-type in the channel, so
today. They are the junction FET and the the gate (or gates) and the source of the a positive voltage is applied at the drain.
MOSFET. The fox:mer is most commonly FET. This positive voltage attracts the negative-
called a JFET. It has no insulation ly charged free electrons and the current
between its elements, just as is the case The Junction FET flows (Fig. 77). The next step is to apply a
with bipolar transistors. The MOSFET As was stated earlier, field-effect tran- gate voltage of the polarity shown in Fig.
has a thin layer of oxide between the gate sistors are divided into two main groups: 77. Note that this reverse-biases the gates
or gates and the drain-source junction. Junction FETs and MOSFETs. The basic with respect to the source, channel, and
The term MOSFET is derived from JFET is shown in Fig. 76. drain. This reverse-bias gate voltage
metal-oxide silicon field-effect transistor .. The reason for the terminal names will causes a depletion layer to be formed
The basic characteristic of the two types become clear later. A dc operating condi- which takes up part of the channel, and
are similar - high input impedance and tion is set up \Jy starting a current flow since the electrons now have less volume
Solid State Fundamentale 4-32
in which to move the resistance is greater from the rest of the device by a layer of
and the current between source and drain very thin dielectric material, so this is not
is reduced. If a large gate voltage is ap- a p-n junction between the gate and the
plied the depletion regions meet, causing device - thus the name insulated gate. VOS '0
pinch all, and consequently the· source- When a negative gate polarity is applied,
drain current is reduced nearly to zero. positive-charged holes from the p-type
Since the large source-drain current substrate are attracted toward the gate. VGS '-1 VOLT
changes with a relatively small gate and the conducting channel is made more
voltage, the cjevice acts as an amplifier. In narrow; thus the source-drain current is VGS' -2 VOLTS
the operation of that JFET, the gate ter- reduced. When a positive gate voltage is
minal is never forward biased, because if connected, the holes in the substrate are VGS' -3 VOLTS
it were the source-drain current would all repelled, the conducting channel is made
be diverted through the forward-biased larger, and the source-drain current is in- vos
gate junction diode . creased. The MOSFET is more flexible (A)
The resistance between the gate ter- since either a positive or negative voltage
minal and the rest of the device is very can be applied to the gate. The resistance
high, since the gate terminal is always between the gate and the rest of the device
reverse biased, so the JFET has a very is extremely high because they are
high input resistance. The source terminal separated by a layer of thin dielectric.
is the source of current carriers, and they Thus the MOSFET has an extremely high
are drained out of the circuit at the drain. input impedance. In fact, since the
The gate opens and closes the amount of leakage through the insulating material is
channel current which flows in the generally much smaller than through the
pjnch-off region. Thus the operation of an reverse-biased p:n gate junction in the
FET closely resembles the operation of JFET, the MOSFET has a much higher
the vacuum tube with its high grid-input input impedance. Typical values of Rin for
impedance. . the MOSFET are over a million
megohms. There are both single-gate and
MOSFETs (Metal-Oxide Semiconductors) dual-gate MOSFETs available. The latter
The other large family which makes up has a signal gate, gate I, and a control
field-effect transistors is the insulated-gate gate, gate 2. The gates are effectively in (B)
FET, or MOSFET, which is pictured series making it an easy· matter to control
schematically in Fig. 78. In order to set up the dynamic range of the device by vary- Fig. 79 - At A are typical JFET characteristic
a dc operating condition, a positive ing the bias on gate 2. Dual-gate curves. The picture at B shows an actual
polarity is applied to the drain terminal. MOSFETs are widely used as agc- oscillograph of the family of curves produced
by a curve tracer.
The substrate is connected to the source, controlled rf and i-f amplifiers, as mixers
and both are at ground potential, so the and product detectors, and as variable at-
channel electrons are attracted to the tenuators. The isolation between the gates
VGS -+2 VOLTS
positive drain. In order to regulate this is relatively high in mixer service. This
source-drain current, voltage is appljed to reduces oscillator "pulling" and reduces
the gate contact. The gate is insulated oscillator radiation. The forward transad-
VGS-+1VOLT
mittance (transconductance, or gm) of
dual-gate MOSFETs is as high as 40,000
micromhos. VGS-O
SOURCE GATE DRAIN
INSULATING
GLASS Characteristic Curves
VGS - -1 VOLT
The characteristic curves for the FETs
described above are shown in Figs. 79 and VGS- -2 VOLTS
80. The drain-source current is plotted
against arain-source voltage for given gate vos
voltages.
SUBSTRATE
(A) Fig. 80 - Typical characteristic curves for a
ClIlsslflcations MOSFET.
Field-effect transistors are classed into
two main groupings for application in
circuits, enhancement mode and depletion
~
RAIN
mode. The enhancement-mode devices are the current to decrease. With the
GATE those specifically constructed so that they MOSFET we can apply a gate voltage of
SOURCE have no channel. They become useful only either polarity so the device can be
(B·) when a gate voltage is applied that causes depleted (current decreased) or enhanced
a channel to be formed. IGFETs (in- (current increased).
sulated gate FET) can be used as To sum up, a depletion-mode FET is
40673
enhancement-mode devices since both one which has a channel constructed; thus
polarities can be applied to the gate it has a current flow for zero gate voltage.
~
, DRAIN
without the gate becoming forward biased Enhancement-mode FETs are those which
.~ and conducting. have no channel, so no current flows with
G1 SOURCE/SUBS. A depletion-mode unit corresponds to zero gate voltage.
(e)
Figs. 76 and 77, shown earlier, where a
channel exists with no gate voltage ap- PowerFETs
plied. For the JFET we can apply a gate FETs capable of handling substantial
Fig. 78 - Profile and symbol for a MOSFET. voltage and deplete the channel, causing amounts 9f power are available for use
4-33 Chapter4
from dc through the vhf spectrum. They frequency performance improved in GaAs
seURCE seURCE
are known under more than one name - , FETs.· GaAs FETs, are classified as
vertical FETs, MOSPOWER FETs and depletion-mode junction devices. The gate
VMOS FETs. The power FET (MOS- is made of gold or aluminum, the latter
POWER TM FET) was introduced in 1976 type being susceptible to damage from
'by Siliconix, Inc. The device enabled static charges.
designers to switch a current of 1 ampere GaAs FETs are available for both
in less than four nanoseconds. The small-signal and power applications. The
transfer characteristic ofthe power FET is power devices have noise figures almost as
a linear one. It can be employed as a linear r~ ________ ________ T
N+SUBSTRATE
~ ~_ low as those specified for the small-signal
types, and naturally exhibit greater
power amplifier or a low-noise, small-
signal amplifier with high dynamic range.
With this kind ofFET there is no thermal
6
DRAIN
dynamic range and ruggedness. Several
semiconductor manufacturers throughout
runaway, as is the case with power types CROSS SECTION OF A VMOS CHANNEL the free world offer gallium-arsenide
of bipolar transistors. Furthermore, there field-effect transistors in various noise
is no secondary breakdown or minority- figure, frequency and power ratings. In
carrier storage time. The latter makes (A) the U.S., Hewlett-Packard and Micro-
them excellent for use in switching wave Semiconductor Corp. feature units
amplifiers (Class D service). Of particular usable up through Ku-band. Represen-
interest to amateurs is the immunity of tative type numbers are HFET-2201 and
power FETs to damage from a high SWR MSC HOOI, respectively. In Great Britain,
(open or short condition). These devices the Plessey GAT5 and GAT6 devices
can be operated in Class A, AB, B or C. feature low-noise performance up to 14
Zero bias results in Class C operation. GHz. The Nippon Electric Company of
Fig. 81 illustrates the manner in which a Japan is also competing strongly for
(8)
MOSPOWER FET is formed. These de- leadership in the GaAs FET market. A
vices operate in the enhancement 11)0de. practical GaAs FET preamplifier is, il-
The current travels vertically. The source Fig. 81 - Prefile and symbo.l fer a pewer FET lustrated in Fig. 86. For more background
is on top of the cnip but the drain is on the (VMOS enhancement·mede type). information on GaAs FETs, see Wade, .
backside of the chip. In this vertical struc- "Introduction to GaAs Field-Effect Tran-
ture there are four layers of material sistors," Ham Radio, January 1978, and
(N +, P, N - and N +). This device of- Wade and Katz, "Low-Noise GaAs FET
fers high current density, high source/ UHF Preamplifiers,;' QST, June 1978.
2.4
drain breakdown capability and low
gate/drain feedback capacitance. These
VDS= 24V Practical FET Circuits
_ 2.0. 8#11,1% I 1
features make the transistor ideal for hf VI DUTY CYCLE Small-signal FETs can be used in the
"- PULSE TEST
and vhf use. ::E
~ 1.6 V same general types of circuits given earlier
Fig. 82 depicts the drain current as l-
z I for bipolar transistors. The primary
being linearly proportional to the gate-to- "'0:0: 1.2
obstacle in some types of amplifier circuits
source voltage. The more conventional ..,:::> 'I is instability. Certain precautions are
JF.ET exhibits .a square-law response z
<i 0.8
/ • necessary to prevent unwanted self-
(drain current being proportional to the 0:
0
oscillations, but they do not differ
square of the gate-to-source voltage): e
I
0..4
/ . markedly from those techniques applied
As an example, the Siliconix VMP-4 / when working with triode tubes.
power FET can provide a power just short 0. V In Fig. 84 are examples of FET audio
0. 2 4 6 8 10. 12
of 20 watts (saturated) at 160 MHz. Fig. VGS-GATE-To.-So.URCE Vo.LTAGE (Vo.LTS)
amplifiers. The circuit at A shows'a simple
83 shows curves for this device respective RC coupled stage with a gain of 10 dB or
to saturated output power versus frequen- greater. The input and output impedances
Fig. 82 - Curve shewing relatienship between
cy. In this case both the input and output gate·seurce veltage and drain current ef a are set by the gate and drain resistors. The
impedances of the transistor are matched pewer FET. circuit at B in Fig. 84 is simiiaF to that at
conjugately. An advantage to this device A, except that a dual-gate MOSFET is
over the power bipolar transistor is that used as the active device (Ql). A positive
these impedances are barely affected by bias is supplied to gate 2 by means of a
the drive .levels applied. In wideband resistive divider. In the circuit of Fig. 84C,
amplifier service the MOSPOWER FET Ii!
l-
25 a pnp transistor is combined with a JFET
can be operated with complete stability. I- to provide a direct-coupled pair. This con~
In-depth data on these devices is given in ! 20. .... figuration provides high gain. The
the Siliconix application note, T A-76-1. 0:
"- ~DD'36V amount of gain is set by the ratio of Rl
:r"'
~
and R2. Again, the input and output im-
GaAs FEi's I-
15 pedances are determined for the most part
:::>
\
'"
For low-noise amplification at uhf and "-
I- by the values of the input and output
:::>
microwaves, the state of the art is defined 0 10. resistors, 1 megoh~ and 1000 ohms,
VDD'2~
~
0
by field-effect transistors fabricated from respectively.
gallium arsenide. Also used in LEDs and "'ti
0: :--........
5
RF and I·F Amplifiers
~
:::>
microwave diodes, gallium arsenide is tiVI
semiconductor compound, as opposed to 0.
Small-signal rf and i-f amplifiers which
silicon and germanium, which are 0. 10.0. 20.0. 300 40.0. 50.0. use FETs are capable of good dynamic
elements. This compound exhibits greater FREQUENCY (MHz) range and will exhibit a low noise figure.
carrier mobility (the electrons can move It is for these reasons that many designers
more freely) than silicon or germanium, Fig. 83 - Curves fer 24· and 36·velt eperatien prefer them to bipolar transistors. Fig. 85
hence the transit time is reduced and high- ef a. pewer FET. contains examples of FET rf or i-f
50-OHM
L4 eO-OHM
~OUTPUT ~OUTPUT
lOOk 330k
U257
L3
L2 G
50-(i)HM
INPUT
+20V
., MOSfET AMP.
+12V
CASCODE AMP.
(C) (0)
4-35 Chapter 4
IN PUT 0---;'1'1-
- .....-1 f---o OUTPUT
501\.
501\.
GAIN
;::;15 dB
NF < 3dB
Fig. 86 - A 1296·MHz preamplifier using a GaAs FET. Most of the Gurrently available GaAs devices will work well at 01.
C1, C2, C4 - 0.8· to 10·pF piston trimlner usable). wide by 0.9 in. (23 mm) long on 1/16.in.
(Johanson or JFO). Note: C1 may be reo 01 _ GaAs FET (see text). , (1.6·mm) thick double,sided G·10 printed.
placed by a fixed low·inductance capacitor RFC1 _ 3 turns, 1I16·in. (1.6·lnm) 10, in lead of circuit board. '
of 10 pF or more. . . resistor, spaced wire diameter. W2 - 50·0 microstripline, 0.105 in. (2.7 mm)
C3 - 0.3· to 3.5·pF piston tnmmer (Johanson RFC2 _ 5 turns no. 32 wire, 1/16.in. (1.6'mm) wide by 1.1·in. (28·mm) long on 1/16.in.
or JFO). . 10, spaced two wire diameters. (1.6·mm) thick double,sided G·10 printed·
01,02 - Zener diode, 5.6 V (4.7 to 6.2 V W1 - 50·0 microstripline, 0.105 in. (2.7 mm) circuit ,board.
if the one nearest to Voo has an loss 87 A. Optimum tradeoff between conver- the gate no. 2, isolation from the re-
higher than its mate. This ensures proper sion gain and IMD occurs near the point mainder of the electrodes. This mixer and
dc bias for cascode operation. Unmatched where the self-bias is 0.8 V. LO injection all other active FET mixers require a fairly
FETs require special forward-biasing voltage will be on the order of 1 (pk-pk) to low drain-load impedance in the interest
techniques and ac-coupling measures that provide good mixer performance. Con- of good IMD. If the drain tune~ circuit is
aren't seen in this circuit. version gain with this mixer will be ap- made high in terms of impedance (in an
Cascode amplifiers are noted for their proximately 10 dB. effort to improve conversion 'gain) the
high gain. good stability. and low noise Fig. ,87B illustrates a singly balanced drain-source peak signal swing will be
figure. With the circuit shown the noise JFET mixer. A broadband transformer high. This will lead to a change in junction
figure at 28 MHz is approximately 1.5 dB. (Tl) provides a low-impedance source for capacitance (varactor effect) and the
Short leads are necessary. and shielding the LO and supplies injection voltage in generation of harmonic currents. The
between the tuned circuits is recom- push-pull to the gates of QI and Q2. Th~ result is distortion. Of primary
mended in the interest of stability. Careful latter should be,matched FETs or a dual significance is the condition called' 'drain-
layout will permit the use of toroidal FET such as the U430 by Siliconix. This load distortion." This malady occurs
inductors at L2 and L3. These com- mixer provides between 10 and 20 dB of when excessive signal levels overload the
ponents should be spaced apart and isolation between the mixer ports. The drain circuit. The result is degraded IMD
mounted at right angles to one another in signal is applied in parallel across the and cross-modulation effects. RI in Fig.
order to reduce unwanted infringing sources of QI and Q2 by means of broad- 88 is used to decrease the drain-load im-
magnetic fields. Agc can be applied to this band transfor,mer T2.0utput at the i-f is pedance by means of swamping. A value
amplifier by routing the control voltage to taken from a balanced tuned circuit. to 10,000 ohms is suitable for a 40673
the gate of QIB. A doubly balanced FET mixer is shown MOSFET mixer. Some JFETs require a
at C in Fig. 87. Broadband transformers lower drain loaa for optimum perfor-
FET Mixers are used throughout, with FLl and FL2 mance. Values as low as 5000 ohms are
There are three types of FET mixers in providing low-pass selectivity at the mixer not unusual. This form of overloading is
common use today - single-ended, singly output: The filters also provide an im- more pronounced at low dc drain-voltage .
balanced and doubly balanced. In all pedance stepdown between the drains of levels, such as 6 or 8.
cases there is an advantage to using active QI and T4 (1700 ohms to 100 ohms). LO
devices in place of passive ones (diodes). injection is supplied to the gates and FET Crystal Oscillators
This assures a conversion gain which signal input is to"the sources. Port-to-port A group of crystal-controlled FET
helps minimize the number of gain stages isolation with this mixer is on the ord.er of oscillators is presented in Fig. 89. At A is
required in a given circuit. 30 dB or greater. Bandwidth is one oc- an overtone type. The tuned circuit in the
A single-ended JFET mixer requires 0 tave. In-depth information on this type of drain is resonated slightly higher than the
dbm of LO injection power. It can pro- circuit is given in the Siliconix application crystal frequency to assure reliable oscilla-
vide several decades of bandwidth and has note AN-73-4. tion. The circuit at B is a variation of the
a good IMD characteristic. The latter is 'Fig. 88 contains the circuit of a typical one at A, but performs the same function.
far superior to most bipolar single-ended dlJal-gate MOSFET single-ended mixer. A Pierce type of triode oscillator is seen in
mixers. The major shortcoming is very Its performance characteristics are similar Fig. 89 at C. It is suitable for use with fun-
poor isolation between the three mixer to those of the mixer at Fig. 87 A. The damental types of crystals. A Colpitts
ports (rf. LO and i-t)o A typical single- primary exception is that the port-to-port oscillator appears at D in Fig. 89. C tb in
ended mixer using a JFET is seen in Fig. isolation is somewhat better by virtue of the circuits at C and D are feedback
Ot
MPFt02 11+12
OR
12-11
s'-""'' '
It
SIGNAL INPUTo---....
100 ;J:0.Ol
LO IN
12 GAIN~10dB
(tV pk-pk) +12V
MIXER
(A)
Ot
t2 11+12
LOIN OR
(+3dBm) • 12 ;'11
l
X· 01
GAIN~5dB
+12V
U350 FLI
z-noo Z-100
11+12
f;\'"
OR
12 -11
T1 T2
,L T4
•
Z-50
FLI. FL2
OL -10
56
T3
GAINN3dB
+12V
12
LO IN
(+6dBm)
4-37 Chapter 4
Ql MOSFET MIXER
MPFI02
11+12
OR
12-11
Rl
10k
0.01
TO.01
100 r+-, 100 ,L0'O,
+12V
OVERTONE OSCILLATOR +12V
(A)
VFO
I-¥-..........---, 43 MHz
100
'---~I--Ir1--1'-....;:.GtJ~I-_~1"::~:""_-~-~-___- - - . - - - - -....----<'l+9V(REG.)
lOOk
0.0'
rL°.01
25
100 ~I--------------------~-----~~(-----OOUTPUT
+12V (Al
OVERTONE OSCILLATOR
(9)
VFO
33k lOOk
.--4~r..-ll""'i-i-i ,,",OUTPUT
L...r..3.5 MHz
RFC
2.5mH
~__-----------------~~----_1~---OOUTPUT
+12V
PIERCE OSCILLATOR
(C) .
VFO +9V
+12V
(REG.)
270
25
(---0 OUTPUT
COLPITTS OSCILLATOR
(0)
(e)
sO'' "'
parts of the circuit should be temperature-
stable. Polystyrene capacitors are recom-
mended, but dipped silver-mica capacitors
will serve adequately as a second choice.
"OW,
EFF.=85%
Preferably, Ll should be a rigid air- 0.1 •
~
wound inductor. A slug-tuned inductor
can be used if the coil Q is high. In such EXCEPT AS INDI~ATED, DECIMAL VAWES OF
cases the slug should occupy the least CAPACITANCE ARE IN MICROFARADS (JIF I ;
OTHERS ARE IN PICOFARADS (pF OR JlJIFI; GAIN"'8dB
amount of coil space possible: Tempera- RESISTANCES ARE IN OHMS; *HEAT SINK
ture changes have a marked effect on fer- k.IDDO,M·IODDOOO. CLASS CAMP
rite or powdered-iron slugs, which can (B)
change the coil inductance markedly. Cfb 270 0.1
of Fig. 90C are on the order of 1000 pF
each for 3.5-MHz operation. They are
proportionally lower in capacitance as the
operating frequency is increased, such as + 4V
REG.
680 pF at 7 MHz, and so on. INPUT "------.11--_ _
'\
......_ _ _*-;......
(0.27WI ~ 0.1
Power FET Examples
~. '::~:::::.
OUTPUT
(3.7WI
Fig. 91 contains examples of three 10K
amplifiers which employ power FETs.
The circuit at A is an audio amplifier
which can deliver 4 watts of output. At 3
0.001 II
watts of output the distortion is approxi-
mately two percent. Feedback is employed
20V
to aid the reduction of distortion. 10/2W o:4iV
IK
A Class C amplifier is seen at B in Fig. .--vV\~~~--------~--------~~~V\~---o+28V
01
91. The VN67 AJ is capable of a saturated
output near 15 watts at 30 MHz. In this.
circuit the power output is considerably BROAOBANO HF AMP.
less. A medium output power of 7 to 10 (C)
Q1
watts is suggested. The gain is approxi- 145MHz
VMP4
mately 8 dB over the frequency range
specified (with the appropriate drain
tank). If proper layout techniques are SIG.INo-;~........I
used this amplifier is unconditionally (0.4W)
stable.
A broadband hf linear amplifier is
sbown at C.
A narrow-band linear vhf power ampli- GAIN~'2dB
fier is shown at D in Fig. 91. Power output IMO· -30dB
is 5 wlltts PEP. IMD is - 30 dB. It is in-
teresting to realize that this same amplifier *HEAT SINK
is suitable as a high-dynamic-range
preamplifier for a vhf receiver. In. this ap-
+28 V
plication the noise figure is on the order of 200mA
2.5 dB and the gain is II dB. NARROW-BAND VHF AMPLI FIER
(D) ,
Other FET Uses
Fig. 92 contains illustrations of addi~ Fig, 91 - Examples of power FETs in three amplifier circuits.
4-39 Chapter 4
ponentf at the doubler output.
. . . - - - - - - - - - - ' -......--O+12V
'''''1
21 not be repeated here. Rather, the text will
provide data on practical applications of
ICs. Linear ICs are so-called because in
most applications where they are used the
performance mode is a linear one. This
does not mean, however, that they can't
be used in a nonlinear mode, such as Class
C. The biasing will determine the opera-
ting mode, Class A through and includ-
ing Class C.
B~R
gain and stray lead coupling set the stage brought out of the IC package by means
,C
for self-oscillation! of its own single pin. This enables the
designer to treat each transistor as a
~f--o CMOSICs discrete device, with the advantage that
E The term CMOS means that the IC isa each transistor has nearly identical elec-
complimentary metal-oxide silicon type of trical characteristics (fT, beta, dissipation
integrated circuit. Essentially, the internal rating, etc.). Some array ICs have fT
Fig. 93 - Pictorial and schematic representa-
tion of a simple IC. workings of the device are not unlike ratings as high as 1200 MHz, with maxi-
those of MOSFETs, the latter of which mum collector dissipation ratings as high
were treated earlier in this chapter: as I watt. Schematic illustrations of some
or more leads which connect to the MOSFETS are formed on the CMOS IC popular RCA array ICs are seen in Fig.
various elements on the chip. substrates. 95.
Some of the present-day ICs are called CMOS devices consume very low power
LSI chips. The term LSI means large-scale - an advantage in battery-operated Subsystems ICs
integration. Such devices may contain the equipment, especially. The transit time A branch of the linear-Ie family tree is
equivalent of several conventional ICs, (propagation delay) through the FET the subsystem IC. It is a conventional-
and can have dozens of dual-in-line gates of a CMOS IC is very short - ideal package integrated circuit, but contains
package (DIP) connector pins. LSI ICs in logic circuits. It ranges from 25 to 50 ns nearly as much circuitry as an LSI chip.
are used in electronic organs, digital in most devices. This does not imply that Some of these devices represent the entire
clocks, electronic calculators, and so on. CMOS ICs aren't useful in linear applica- active-device circuitry for an fm or a-m
Essentially, they are just super-size ICs. tions: Some are designed primarily for the radio receiver. The designer needs only to
linear amplification of audio and rf atld essential outboard components
Some Practical Considerations energy (CA3600E, for one). Another sali- (front-end tuned circuits, i-f
In circuits where slight extra lead ent feature of CMOS chips is low noise. transformers, tuning meter, and audio
lengths can be .tolerated, it is prudent to Because FETs are used in these les the in- power amplifier) to realize a composite
install the ICs in sockets rather than put impedance is high, making them more piece of equipment. Other subsystem les
solder them into the pc board directly. In suitable than bipolar ICs for interfacing· may contain only the i-f amplifiers, pro-
amateur work there is an occasional need with comparable impedance levels outside duct detectors, agc loops, and audio pre-
to replace an IC during circuit de- the IC package. Fig. 94 shows the diagram amplifiers. This style of IC is sold by such
velopment for a typical one-shot design. of a CA3600E CMOS IC along with a manufacturers as RCA, National
This is particularly pertinent when bargain- block-symbol circuit for its use as a high- Semiconductor and Plessey. Fig. 96 il-
house ICs are purchased: Many have de- gain audio amplifier. lustrates an example of this kind of device
fects, and the task of removing an IC that - the RCA CA3089E which is designed
is not in a socket is a task that no builder Array ICs expressly for uSe in wide- or narrow-band
finds delightful. One branch of the linear-IC family is fm receivers. It features a quadratlIre
2 11
JJ ;b 560
1 +
,"'OTt' wF 12V\-
~-
1
6 12
5
J 56
(--oAF
7 3 4 9
RCA 3600E COS/MOS ARRAY CASCADE iCC-dB AF AMPLIFIER
(GOOD TO 5 MHz) IB)
(A)
Fig. 94 - The diagram at A shows the Internal workings of a CMOS IC. A 100-<18 audio amplifier which employs the CA3600E is shown at B.
4~41 Chapter 4
11· 12 6
90--1-1' HI--....,-o() 4
B !I
,. 2
SUBS.
6 3
7
1
10
CA3018A
(A)
4
3
CA3026 OR
(8)
B
CA3049
9
SUBS.
B 13 14
12
4
L, :J c:
10
"J c:
2
3 6
7
!I
r
SUBS.
7 B 2
r' "J c:
1
12
, ~"".
CAY724G CA3019 CA3039
(e) (0) (E)
Fig. 95 - Various transistor and diode·array les. The configurations suggest a variety of amateur applications.
detector, and contains amplifiers, design work is started. It is beyond prac- cuit of Fig. 98A functions as a differential
limiters, squelch circuit, metering circuit ticality to include the schematic diagrams amplifier, as does the one at B. The basic
and an af amplifier. Those interested in of the ICs used in this book, but we will difference is that dc and ac balance are
compact portable amateur' receivers show the circuit of the RCA CA3028A, featured at A, whereas only dc balance is
should find these devices especially in- because it is used frequently in the follow-, effected at B. The gain of either stage is
teresting. ing section. Fig. 97 contains the block and approximately 40 dB. Pin 2 of VI is left
schematic representation of this IC. floating, but is used for LO injection
Practi~alExamples when the CA3028A is employed as a mixer
The main disadvantage in the use of IC RF and I·F Amplifiers. or product detector. A Motorola
symbols in circuit diagrams is that the in- Nearly every manufacturer of ICs pro- MCI550G is similar to the RC shown in .
ternal workings are not shown. This duces chips that are suitable for use as Fig. 98. A MC1590G is a more suitable IC
makes the designer work with a collection rfli-f amplifiers, mixers, detectors, for i-f amplification when' greater
of "magic boxes." Fortunately, IC oscillators and audio amplifiers. The cir- amounts of stage gain or agc control are
manufacturers publish data books which cuits of Fig. 98 are examples of CA3028A desired.
show the block symbols and pin ar- rf or i-f amplifiers to which agc is applied. An example of an MC1590G amplifier
rangements versus the schematic diagrams Maximum gain occurs when the agc volt- is given in Fig. 99A. It is shown with agc
of the active devices on the chips. This age (lC forward bias) is; at its highest applied to pin 2. The lower the agc voltage
permits the amateur to understand what potential. The IC is nearly cut off when the higher the stage gain. This is the op-
the circuit configuration is before the the agc level drops below 2 volts. The cir- posite condition from that of the
60
12k
+12V
(8)
1- F
AMPLIFIER
INPUT·o--~~-.,
I-F
IN~
0.01
0.01,L 10k
!560 AGe
(+~. TO +12V)
AGe. +2 TO +9V +12V
+12V (A)
RF OR I.-F AMPLI FI ER
(A)
I NPUTQ---.--, I-F
AMPLIFIER
MANUAL +12V
GAIN
MIN._MAX.
+12V
(8)
RF OR I-F AMPLIFIER
(8)
Fig. 98 - Rf and i-f amplifiers using the CA3028A IC. The example at A Is balanced for ac and dc, Fig. 99 - Circuit examples for Motorola IC i-f
whereas the circuit at 8 is balanced only for dc conditions. amplifiers.
4-43 Chapter 4
CA3028A of Fig. 98, where the gain in- BALAN,CEO'
creases with elevated agc voltage. The MIXER
MC1350Pof Fig. 99B is the low-cost ver-
sion of the MCi59OG. It is shown with
manual control of the gain (R1), but agc SIG.INPUTO-_-..,
voltage can be applied instead.
IC Mixers
Examples of IC active mixers are given
in Fig. 100. At A is seen a singly balanced
mixer formed by the differential transistor
pair in a CA3028A. A doubly balanced
mixer is illustrated at B in Fig. 100. The rL°. 1
1200
.------r--------~vv--------~
~01
MC1496G contains two differential tran-
sistor pairs to permit the doubly' balanced IO.00 1 060
configuration. This circuit does not exact-
ly follow the suggested one by Motorola. L.O
INPUT +12Y
It has been optimized for use as a trans- (1.0Y pk-pk)
mitting mixer by W7ZOI and KL 71AK
(Solid State Design for the Radio (Al
Amateur, 1st edition, page 204). There are
numerous other ICs which can be used as
mixers. Examples of many practical cir-
cuits are given in the ARRL book just DOUBLY
referenced. 0.1 BALANC.EO
The circuit arrangements for product SI G. INPUT 0-:--11-....----.., MIXER
1200 .
(e)
Fig. 102 - The standard negative·feedback op·amp circuits with their transfer equations. At A is a noninverting amplifier, at B, an inverting amplifier,
and a differential amplifier is shown at C.
input and the one marked "+" is the forms the output bHffer. A more com- The differential inputs allow for both in-
noninverting input. The next stage, the prehensive discussion of operational verting and noninverting circuits. Fig. 102
pnp transis~or, provides most of the amplifiers is given by Woodward in "A shows these configurations and gives their
voltage gain. High gain is realized through Beginner's Look at Op Amps," April and transfer equations. RL does not appear in
the use of a constant current source for June 1980 QST. the equations, implying that the output
the pnp collector load. The frequency The most common application for op impedance is zero. This condition results
response is determined by the collector-to- amps is in negative feedback circuits from the application of heavy negative
base capacitance of the pnp stage. This operating fr.om dc to perhaps a few hun- feedback. Most Ie op amps have built-in
capacitance is fixed internally in some Ie dred kHz. Provided the device has suffi- current limiting. This feature protects the
op amps and connected externally to cient open-loop gain, the amplifier Ie from damage caused by short circuits,
others. A pair of emitter followers in a transfer function is determined almost but also limits the values of load resistance
complementary symmetry arrangement solely by the external feedback network. for which the output impedance is zero.
1k
P.oLICE / FIRE
SCANNER
1k
0.22 1.ok
VHF/UHF AUDI.o .oUTPUT
FM TRANSCEIVER 1.ok
.0.22 1Dk
HF SSB
TRANSCEIVER
,r-
INPUT U2: LM383
MASTER VDLUME
C.oNTRDL
4-45 Chapter 4
Most op amps work· best with load re- 75k
sistances of at least 2 kO. 120k
10k
2'
Since the op amp magnifies the dif-
240k
ference between the voltages applied to its 22
inputs, applying negative feedback has the 10k 470k
effect of equalizing the input voltages. In 2'
the inverting· amplifier configuration the ~-'-4I........ -o OUTPUT
Ul: LF353N
+12V
1
The virtual ground at the inverting in-
put terminal of an inverting operational Fig. 104 - BCD D/A converter suitable for connection to a B·series CMOS driving source.
amplifier circuit allows several currents to
be summed without interaction. This prin-
ciple can be used to advantage by the
I~VU~--~
+v
amateur wishing to simplify his or her sta- +7V.
-v ________ _ OUTPUT
fier to restore the anti vox signal to the
proper phase. Fig. 104 shows another ap-
plication for a summing amplifier, a Df A -v
converter. An FET-input operational (A)
+1.6V
amplifier can operate with the high-value (8)
1°.,
amplifier must have input current to Although RC active filters can be built
POWER
AMPLIFIER
operate, hence it is' not a high-impedance with bipolar transistors, the. modern
device_ In the inverting"amplifier con- approach is to utilize operational ampli-
figuration the numerical voltage gain is fiers (op amps). The use of an op-amp IC,
Rf/Ri, but the noninverting input terminal such as a type 741, results in a compact
must be returned to the positive supply filter pole which will provide stable
through a resistance of 2Rf to equalize the operation. Only five connections are made
input currents. Any attempt to use this to the IC, and the gain of the filter section,
type of IC as a voltage follower is doomed plus the frequency characteristic, is de-
to failure - the input stage will be termined by the choice of components
*HEAT SINK destroyed by excessive current. The chief external to the IC. •
usefulness of Norton amplifiers is in Although there are numerous ap-
single-supply applications where the dc plications for RC active filters, the
level of the signal is very near ground. The principal use in amateur work is that of
ssb chapter of this Handbook features a establishing selectivity. at audio frequen-
4.47 Chapter 4
~--~--------.....- - - - - - - - . . . . . ; - - - . . . . . ; - - - , - - - . calledj1icker, and is a wldeband signal whose
fa =
900 Hz Av 1 CII 0 =
211' ro(HZ) C1 = amplitude varies inversely with frequency .
. For some analytical purposes, drift is con-
Q = 5 CI, C2 = 0.00068/.1.F sidered as a very low frequency noise compo-
5 nent. Op amps that have been optimized for
RI = 0) (6.28 X 900) (0.00068 X 10-6) offset, drift and noise are called instrumenta-
tion amplifiers. The latest instrumentation
= 1,300,948 ohms amplifier is the National Semiconductor
. 5 LMIO, designed by Robert Widlar, the
R2 = [2(25) -11 [6.28 X 900 (0.00068 X 10-6)1 = 26,550 ohms ....... ·\IV,,.......-u+'2V
acknowledged "father of the Ie op am·p."
The architecture of the LMIO is different
R3 = (6.28 X!}()() (~~OOO68 XlO-6) = 2,601,896 ohms p. POLYSTYREN~ from any other device, but the practical ap-
plications are the same.
The small-signal bandwidth of an op amp·
R4, R5 :::: 2,601,896 X 0.02 = 52,0\8 ohms is the frequency range over which the open-
loop voltage gain is at least unity. This
Fig. 108 - A design example based on the circuit of Fig. 107,
specification depends mostly on the frequen-
cy compensation scheme (for.example, the
120 capacitor in Fig. 101). Fig. 109 shows how the
lOR
maximum closed-loop gain varies with fre-
quency. The power bandwidth of an opera-
100
tional amplifier is a function of slew rate, and
INPUT is always less than the small-signal value. Slew
80 >-.......-() OUTPUT rate is a measurement of output voltage swing
..,
aJ
per unit time. Values from 0.8 to '13 volts per
60 microsecond are typical of modem devices.
~
<[ The hobbyist should maintain'a supply
(!)
of inexpensive 741 and 301 op amps for
UJ 40
(!) breadboarding, but should also be
<[
I- CLOSED LOOP prepared to use improved devices in the
..J
0 20 final design. In an active filter for exam-
> ple, a 741wiII demonstrate whether or not
0 the circuit is working, but a low-noise,'
wide-bandwidth device will give higher
-20
performance, esgecially in receiving ser-
10 100 lk 10k lOOk 1M 10M vice. An abbreviated table of operational
FREQUENCY, Hz
amplifier specifications is given in chapter
23. Most of the devices listed are available
Fig_ 109 - Open-loop gain and closed-loop gain as a function of frequency. The vertical distance from hobby electronics stores or the mail-
between the curves is the feedback or gain margin.". order firms listed in chapter 17.
0 0 0 0 0 1
0 , 0 0 1 1
1 0 0 , 0 1
, , 1 1 1 0
(A)
~COM
+vo-L-o-i ~~ NC '"-;;C
0 0 0 0 0 1
C = A+B 0 , 1 C=A+B 0 1 0
, 0 1 , 0 0
, 1 1 , 1 0
(B)
r------------------Oc
+V~
COM lNC
-t.
NO
NC=NORt.4ALLY CLOSED
NO~NORMALLY OPEN
{?-. NOT" ~
o
I
A
1
0
A
0
B
1
C
(INVERTER) XOR 1 0 1
(EXCLUSIVE OR)
(C) 1 1 0
c~
c~
1,.'''"'~
technology allows digital systems of
tremendous complexity to be built at a
small fraction of the cost of previous
+v 0 NC I NC
methods.
t NO
t NO NETWORK
Combinational Logic
The three logical operations are "and,"
"or" and "not." An AND gate may be
assembled with two relays as shown in
Fig. 1lOA. In order to have voltage at the
'7 ,A B C
•
ELECTRICAL
'7
output (C), we must energize A and B. If L L H TRUTH TABLE
we connect the contacts in parallel rather L H L H = HIGH VOLTAGE (A)
than in series, an OR gate results (Fig. H L L L= LOW VOLTAGE
1108). The "not," "complement" or
H H L
"inverse" function may be implemented
with a normally-closed relay contact as il-
lustrated in Fig. 11Oe. If we apply voltage
to A we will have no voltage at e, and vice
versa. With the proper system of AND and
NOT gates or OR' and NOT gates, any logical
or arithmetic function may be synthe-
A B C
sized. AND or OR gates are often combined NEGATIVE L.OGIC
A B C
0 POSITIVE LOGIC
with inverters In Ie packages and called 0 1
TRUTH TABLE 1 1 0
TRUTH TABLE
NAND and NOR gates. 0 1 0
H=1 H=O 1 0 1
A special combination of gates called 1 0 0 (B)
LoO L=1 0 1 1
an "exclusive OR" has an output only if
1 1 0 0 0 1
the two inputs are ·complementary. This
combination is used frequehtly enough to
be packaged specially and assigned a fun-
damental symbol. :=D--C=A+B :=O-C=AB
Logic systems have polarity. If the
highest voltage level represents a binary NOR NAND
one and the lowest level represents a zero,
the logic is said to be positive. If the op-
Fig. 112 - At A, combinational logic implemented with relays, shown with the electrical truth
posite representation is used, the logic is table. Assigning 1 and 0 to the electrjcal states as in B leads to two schematic symbols, one for
negative. positive logic and one for negative logic. The two symbols are electrically equivalent; depending
Since each input or output of a digital on the application, one may represent the logical operation being performed better than the other.
4·49 Chapter 4
binational circuit.
DEMORGAN'S THEORUM POSITIVE LOGIC NEGATIVE LOGIC
somewhat different from those of or- important results from Boolean alg~bra, I *= CHANGE
LEVEL
13' 13' t 0 0 I
13' 0 t
dinary algebra. The" + " symbol is used justifies the conversion from one logic (A) 13'= DON'r "
13' 13' t
1
t 1
0
1
to indicate the "or" function. "And" is convention to the other. An application of CARE
~
dicates that it has been inverted. Fig. III block representation of the Boolean equa-
shows the symbols for the common logic, tions clarifies the negative logic sym-
functions with their associated Boolean bology. '
equations and truth tables. Positive logic Finally, Fig. 113 gives detailed electrical
is assumed. With the exception of the "ex- truth tables showing identical output
clusive OR," all of the gates may be ex- states for any combination of inputs. A
panded to any number of inputs. There is complete chart of equivalent symbolic
no universally-accepted definition for an representations is given in Fig. 114.
exclusive OR gate with more than two in- A circuit made of the fundamental Fig. 115 - A 0 flip-flop. In A, set and reset
puts. gates and configured in such a way that ("jam") inputs are provided. Note that the func-
The Boolean algebra associated with the output is a function of the present tional truth table shows a and i5 both in the
high state for one combination of Rand S.
logic networks can sometimes be static input levels only is called a combina- While this appears contradictory, it is the stan-
simplified through the use of negative tionallogic circuit. Pulses and transitions dard way of defining the operation of this type
logic. Consider'h circuit having two inputs are not considered in the design of a com- of flip-flop.
•
I 1 -Switching theory is based on ideal
t 1
r~
I I
I I I 1
switches, and real electrical devices don't
I 1 19
I
1 always emulate ideal switches well enough
1 1
1
I If-
I
I to synthesize a logic design with switching
I I Ig '1
theory alone. Therefore, a digital designer
I
I 1 I'" I
I
1 I:;; 1 . must consider the electrical characteristics
1 I! 1
I 1 1 of the logic elements he's using. Propaga-
1 1 1 I
1 I tion delay and transmission line reflec-
1 1 I
1 ~--
CONTROL 1 I tions become significant factors as the
- I
f - - - - i - - - - - - .J speed of the logic system increases. The
--------1-- -- UNIT I
I ultimate application of the system also in-
I
1 1 fluences the design. For example, an at- .
L__________ ~ ______ ~
tempt to design a frequency synthesizer
.
- - - - - CONTROL PATH
strictly as a "numb.et crunc,her" is doom-
ed to failure. TIle reason is that any solid •
state digital device is also an analog device
- - - D A T A PATH having a finite transfer function. This
transfer function can cause a device to act
Fig. 117 - The basic parts of a {iigital computer. as an amplifier, multiplier or mixer as well
as a switch. A frequency synthesizer
designed only as a logical machine will be
rife with spurious outputs and noise that
There are programs called text editors, picture elements that differ from the can't be predicted from switching theory.
which translate invalid instructions into previous frame. This technique, known as Stray coupling between sections, small
the nearest valid ones, but even here the digital refreshment, is a sophisticated job transients on the power supply, and junc-
machine isn't thinking, because the text for a microcomputer system. The high- tion noise can pollute the final output
editor program was written by'a human. speed data communication made possible spectrum without affecting the logical
Amateur computing is a hobby quite by computers may allow a complete function.
distinct from Amateur Radio, but there vhf/uhf contact in a single meteor burst. Particular attention should be paid to
are some worthwhile computer applica- Existing practice with this propagation power supply decoupling. In general, a
tions in radio. Some of these are Morse mode sometimes requires hours to ex- few 0.01 ",F disc ceramic bypass capacitors'
code and RTTY encoding and decoding, change call signs and signal reports. will prevent the switching transients of
SSTV chracter generation, aural readout The current state of the art in amateur one IC from changing the state of
(any format) of digital displays and EME work requires many kilowatts of erp another. If the logic system is to be used
satellite commanding. Of course, a and ultra-low-noise receivers to obtain -with radio equipment, more extensive
microcomputer system can be used for barely perceptible lunar echos. By statis- measures may be necessary., For a logic
routine filing and record keeping. Some tical analysis of the receiver output, a system to have electromagnetic com-
hams keep their station logs by computer, inicrocomputer could possibly pull a lunar pa.tibility (EMC), it must not radiate
and can instantly retrieve information echo out of the noise, thereby easing the energ¥ into a radio receiver or be adverse-
about previous contacts. A computer can station gain requirements. ly affected by energy radiated from a
relieve the tedium of sorting tasks (for ex- Amateurs are becoming more interested nearby transmitter. The EMC problem is
ample, the index of this Handbook was in computer-aided design. This technique receiving increased attention as more
organized with the aid of a microcom, is especially useful for designs requiring radio amateurs acquire personal com-
puter system), but the most exciting ap- manY'iterative calculations, such as in- puters for their stations.
plications are to things that weren't possi- terstage matching networks in solid-state
ble before the personal computer era'. transmitters. The tables of filters appear- Logic Device Fabrication Technology
Some innovations that need to be ing elsewhere in this Handbook were Nearly all modern logic systems use in-
developed by radio/computer enthusiasts generated by a computer. tegrated circuits. The ICs are classified ac-
are video bandwidth compression QST publishes articles on nontrivial ap- cording to the complexity of the circuit on
(MSTV), high-speed data communication plications of microcomputers to Amateur the chip. ICs having 15 or more active
and weak-signal enhancement. One Radio. A bibliography of QST articles devices fall into the MSI or medium-scale
method for transmitting moving images in from 1975, to the present on micro- integration category. The simple gate and
a narrow bandwidth is to send only those processors and microcomputers is flip-flop circuits use MSI technology.
Also known as T'L (Lsquared L), this L._~ ___ ~_-* ___ ~GROUND NOR GATE
(NEGATIVE LOGIC)
family has a variety of circuit configura-
tions. Some devices have "open collec- (8l
tor" outputs, and these may be "wire
oRed." Open collector outputs are useful Fig. 118 - TIL circuits and their equivalent logic symbols (see ,text). Indicated resistor values are
for interfacing with other logic families or typical. Identification of transistors is for text reference only; these are not discrete components.
discrete components. Although + 5 volt~
is the recqmmended' power supply for
TTL, open collectors cim be connected to
a different voltage through the external
load resistor, within the liJ11its specified· at speeds up to 35 MHz. Two commonly is represented schematically in Fig. lISA.
for the device. Most TTL devices have used TTL devices are represented A LOW level applied to the input will c~use
"totem pole," or "active pull-up" output schematically in Fig. lIS. High-speed les 01 to conduct current. This will cause 02
stages, and these cannot be wired oRed. (50 MHz) are identified by 54Hoo- or to be near cutoff, in, turn biasing the
Typically, the'(JUtputs are capable of sink- 74Hoo-series numbers. These les "totem pole" arrangement oC 03 into
ingmore cur-rent than they can source. consume more power than their ordinary saturation and 04 near cutoff. As a result
This situation is of importance only when counterparts. The 54Loo- and the output level will be HIGH, about I volt
interfacing devices outside a particular 74Loo-series of devices are designed for below Vcc. If the signal at the input is
TTL subfamily. TTL devices have a lower power consumption that the stan- HIGH the conduction state of each tran-
fanout (number of inputs that can be dard types. These les typically dissipate sistor reverses and the output drops nearly
driven by a single output) of 10 within a one milliwatt per gate, or about one-tenth to ground potential (LOW). The' input
subfamily. If TTL inputs are left open, of that dissipated by standard TTL. diode protects the circuitry by clamping
they assume a "high" logic state,. but Operating speed is the tradeoff for the any negative potential to approximately
greater noise immunity will be realized if lower power, and the maximum speed for - 0.7 volt, limiting the current in 0 I to a
pull-up resistors are used. When operated this subfamily is 3 MHz. safe value. Note that this protection is ef-
with a + 5-volt supply, any input voltage The subfamilies discussed so far operate fective only against transients. The output
level between 2.0 and 5.5 is defined to be as saturated switches. The 54S00- and diode is required to ensure that 03 is cut
high. A voltage less than O.S is an input 74Soo-series have Schottky diode clamps off when the output is LOW.
low. TTL les output a minimum high that keep the transistors out of saturation. The circuit in Fig. IISB, one section of
level of 2.4 volts, and.a maximum low Some les of this series are useful up to a type 7400 gate is very similar to that of
level of 0.4 volts. The switching transients 125 MHz. The power dissipation is about A. The difference is that 05 is a multiple-
generated by TTL devices appear on the twice that of standard TTL. A commonly emitter transistor with one input to each
supply line and can cause false triggering used subfamily combining low power emitter. A LOW level at either input will
of other devices. For this reason, the dissipation with fairly high speed is the turn on 05, causing the output to go
power bus should have several bypass 54LSoo- and 74LSOO-series. The dissipa- HIGH.
capacitors per pc board. tion and speed for this series are 2 mW
"Plain" TTL les are identified by and 45 MHz, respectively. Emittrr-Coupled Logic - /fCL
5400- or 74oo-series numbers and operate One section of a type 7404 hex inverter EeL has the highest speed of any of the
4·53 Chapter 4
levels' swing to within a few millivolts of
the power supply voltages. The input
switching threshold is approximately one
half the power supply voltage (VDD -
Vss). This characteristic contributes to
high immunity to noise on the input signal
or power supply. 'CMOS input current
Q6 OUTPUTS
I
I.
drive requirements are miniscule, so the
Q4 I
I
t---ONOR fanout capability is tremendous, at least in
INPUTS
I
I
low-speed . systems. For high speed
Ql I
systems, the input capacitance increases
~ ~
I
I
the dynamic power dissipation and limits
I
I
:: RL RL the fanout. .
I
I
I
I I
I
Four subfamilies of CMOS logic ICs
I
I are being produced at present. The 4000A
~ ~
I
Rj
R
j I
I
series is the original commercial line and
I
I operates with power supplies from 3 to 12
I
L----4---~----~~---~---~-----OVEE volts. A subfamily having some improved
characteristics is the 4000B (for buffered)
series. The B series can be powered from.
supplies up to18 volts. This feature makes
the devices especially attractive for
automotive applications. The output im-
Fig. 119 - Circuit topology of the ECL family. The functions of the various components are ex· pedance of buffered ICs is independent of
plained in the text. Complementary outputs call for the modified logic symbol.
the input state. An unbuffered series,
designated 4000UB meets all the B series
specifications except that the logic outputs
are not buffered and the input logic levels
logic forms. Some ECL devices can these levels appear at the emitters of the must be within 20 percent of the power
operate at frequencies higher than 1.2 output buffers, Q5 and Q6. The circuitry supply terminal voltages. Several trade-
GHz. This family is different from the associated with Q4, Dl and D2 is a bias offs must be considered when choosing
other forms of bipolar logic in that the generator. The reference voltage between buffered and unbuffered ICs.
transistors operate in a non-saturating established at the base of Q3 determines The buffered devices have greater noise
mode that is analogous to that of some the input switching threshold. immunity and drive capability, but the
linear devices. The typical logic swing is speed is low compared to the uribuffered
only 800 mV. ECL devices are characterized Metal-Oxide Semicpnductors - MOS types. Some special-purpose 4000 series
for use with a - 5.2-volt power supply, The logic families using all n-channel or ICs have tri-state output circuits. The
but operation from other supplies is possi- all p-channel' field-effect transistors are third state is neither HIGH or LOW, but is a
ble. If the Vcc terminal is connected to used extensively in microprocessors, high-impedance condition which allows
+ 2.0 volts and the Vee terminal con- digital watches and calculators. Where en- several outputs to be paralleled for wire
nected to ....: 3.2 volts, the device can drive tire functions can be synthesized on a oRing or multiplexing. The 74COO series is
a 50-ohm load directly with respect to single chip, this technology is quite useful. designed to be a plug-in replacement for
ground. The power output obtained this Ordinary NMOS and PMOS gate pack- low-power TTL devices in some applica-
way is about 0 dBm. ECL ICs dissipate a ages are not very popular. Most general- tions. Some CMOS devices can function
great deal of power, and heat sinking is purpose logic networks are now made at speeds greater than 15 MHz.
sometimes necessary. The ECL family with complementary metal-oxide A simplified diagram for a CMOS logic
finds use in' uhf frequency synthesizers semiconductor (CMOS) ·ICs. This family inverter is given in Fig. 120. Some of the
and counters, as well as in computers. has p-channel and. n-channel transistors diodes in the input and output protection
Some highly specialized ICs have on the same chip. Only one of each com- circuits are inherently part of the
capacitively coupled inputs, and therefore plementary pair is turned on at any time, manufacturing process. Even with the
have minimum as well as maximum toggle so the power dissipation is negligible ex- protect\ion circuits, CMOS ICs are suscep-
rates. ' cept during logic transitions. A notable tible to damage from static charges.
There are several ECL subfamilies feature of CMOS devices is that the logic Certain precautions have become
being produced. Speed, power dissipa-
tion, and the ability to drive transmission
lines cannot be optimized simultaneously,
so different versions are offered to allow
the designer to choose the tradeoffs best
suited to his application. ECL subfamilies
are compatible, but only over a limited
temperature range. The differences be- 15-30.n.
tween subfamilies are mostly in resistance INPUT fJ----."""'N\r-+---O OUTPUT
values, and the presence or absence of in-
put and output pull-down resistors.
A significant feature of ECL gates is
that complementary output functions are
available from each circuit. The circuit of
Fig. 119, for example, is a NOR/OR gate.
Ql or Q2, together with Q3 forms a dif-
ferential amplifier. When the Q2 collector
goes HIGH, the Q3 cpllector goes LOW, and Fig. 120 - Internal structure of a CMOS gate.
Table 1
electrical Characteristics of the Common Ldglc Families
Standard TTL Schottky TTL High-speed TTL Low·power TTL Schottky low·power ECL 11/ ECL 10 k
(active pull·up) (745) (74H) (74L) TTL (74LS) (1600 series)
Vee = +5.0 V Vee = +5.0V Vee = +5.0 V Vee = +5.0 V Vee =. +5.0 V Vee = 0 V,
VEE = -5.2 V
minimum HIGH"
inpu\ voltage 2.0 V 2.0 V 2.0 V 2.0 V 2.0 V -1.095 V -1.105 V
maximum HIGH
input current 40~ 5Ol'A 5Ol'A 2Ol'A 2Ol'A
maximum LOW
input voltage 0.8 V 0.8 V 0.8 V 0.8 V 0.8 V -1.485 V -1:475 V
maximum LOW
input current 1.6 mA 2.0 mA 2.0 mA 400l'A 400l'A
minimum HIGH 2.4 V' 2.5 V 2.4 V 2.4 V 2.5 V -0.9 V -0.825 V
output voltage (30 mAl (30 mAl
maximum HIGH
output current 800~ 1.0 mA 1.0mA 4OOl'A 400I'A 40mA SOmA
maximum LOW
output voltage 0.4 V 0.5 V 0.4 V 0.3 V 0.5 V -1.75 V -1.725 V
maximum LOW Open emitter -
output current 16 mA 20mA 20 mA 4.0mA 8.0mA pull·down reo
sistor required
4·55 Chapter 4
satisfactory results, even if they have a
+ 15V +5V common power supply. The buzzword in-
terface is often' used to describe the in-
tegration of two types of logic into a com-
patible system. There are a number of ICs
intended especially for mating different
68k
. logic families. The CD4049UB and
CMOS LOW POWER
CD4050B hex buffers are designed to
OUT IN drive TTL gates trom CMOS input sig-
40008 OR TTL
74C SERIES lN270 nals. TTL-to-ECL and ECL-to-TTL con-
version can be implemented with the
NlO17 and N1068. integ~ated circuits
(A)
manufactured by Signetics and others.
Unfortunately, these com ponents aren't
always conveniently available to the small
investor, so logic interface must some-
times be accomplished by other means.
+5 TO 15V +5V A knowledge of the circuit topologies
and input/output characteristics will
allow the' designer to concoct reliable
digital interstage networks. Typical inter-
nal structures have been illustrated for
each common logic family. The input!
CMOS TTL
OUT IN output characteristics of the· common
(ANY TYPE) (ANY TYPE) logic families are listed in Table I. This in-
formation was compiled from various in-
220A
dustrial publications and is intended only
* GATE PROTECTED as a guide. Certain ICs may have charac-
teristics that vary from the values given.
The following section discusses some
(B) specific logic conversions. Often more
than one conversion scheme is possible,
Fig. 122 - CMOS·to·TTL interface circuits. When both devices operate from a + 5·volt supply, the
depending on whether the designer wishes
diode in A can be eliminated. The circuit in B exhibits maximum flexibility with respect to supply to optimize power consumption or speed.
voltages and logic subfamilies. Usually one must be traded off for the
other. Where an electrical connection be-
tween two logic systems isn't possible, an
For example, the highest frequency stages frequency range. The programming func- optical isolator can always be used.
in a uhf counter or frequency synthesizer tions have practically no speed con-
would use ECL. After the frequency has straints, so considerable power can be TTL-Driving CMOS
been divided down to less than 25 MHz, saved by using CMOS in that part of the A CMOS gate is easily driven by a TTL
the speed of ECL is unnecessary, and the system. device when both are powered by a + 5-
expense and power dissipation is un- Each of these logic families has its own volt source. The totem-pole output struc-
justified. TTL is the obvious choice for input voltage and current requirements, ture of most TTL ICs prevents a HIGH
the signal-processing operations in this so they can't be randomly intermixed with output level of sufficient potential to pro-
perly activate the CMOS input. A pull-up
. resistor connected from the interface
point to the power bus will remedy this
problem. The maximum usable value for
this component is 15 kO, but the circuit
capacitance will reduce the maximum
CMOS (4000A) CMOS (40008) CMOS (7400C) possible speed of the CMOS gate. Lower
Vee Vee Vee values will generate a more favorable RC
+5.0 V + 10.0 V +5.0 V + 10.0 V + 15.0 V + 5.0 V + 10.0 V + 15.0 V product at the expense of increased power
dissipation. A standard TTL gate can
drive a pull-up resistor of 330 0, but a
3.5 V 7.0 V 4.0 V 8.0 V 12.5 V 3.5 V 8.0 V low-power version is limited to 1.2 kO for
a minimum. The resistor pull-up tech-
1.0 fAA 1.0 fAA nique is illustrated in Fig. 121A.
When the CMOS device is operating on
1.5 V 3.0 V 1.0 V 2.0 V 2.5 V 1.5 V 2.0 V
a power supply other than + 5 volts, the
1.0 fAA 1.0 fAA TTL interface is more complex. The
4.95 V 9.95 V 4.95 V 9.95 V 14.95 V 2.4 V 9.0 V common-base level shifter of Fig. 121B
(..•......•....·......·..··..•·..no load ..............·................·) (360 fAA) (10~) will translate a TTL output signal to a
I
+ 15-volt CMOS signal while preserving
300 fAA ~50 ~ 1.6 mA 1.3 mA 3.4 mA 1.75 mA 8.0 mA the full noise immunity of both gates. An
(2.5 V) (9.5 V) (2.5 V) (9.5 V) (13.5 V) (0 V) (0 V) operational amplifier ,configured as a
0.05 V 0.05 V 0.05 V 0.05 V 0.05 V 0.4 V 1.0 V comparator, as in Fig. 121C, makes an ex-
(..·....···....·......··....···..no load .............•.......-..........) (360 fAA) (360 fAA)
cellent converter from TTL to CMOS
300 fAA 600 fAA 500 fAA 1.3 mA 3.4 mA 1.75 mA 8.0 mA
(0.4 V) (0.5 V) (0.4 V) (0.5 V) (1.5 V) (5 V) (10 V)
using dual power supplies. An FET op
amp is shown because the output voltage
Solid State Fundamentals 4·56
, of this type can usually swing closer to the
power rails than a bipolar unit. Where the
pulse rate is below 10 kHz or so, a 741 . - - - - - - - -.....--------<_--{) + 5.0 V
4·57 Chapter 4
.-----------------------~----------------------~~+5V
be coupled by the circuits drawn in Fig .
1238 and C. In 8, the TTL ga~e is di-
. vorced from the voltage divider network
OUT whell the output is HIGH. In this state the
2N4403
ECl 2N3906 junction of the 1.2-kO and 12-kO resistors
2N2907 .assumes a potential of nearly + 5 volts.
When the TTL output goes low, the
~EE IN anode end of the diode string is pulled
TTL
down to about + 2.5V. This 2.5-volt logic
swing is attenuated and shifted to the pro-
220
per non-saturatifllg ECL levels by the
resistor network.
An emitter follower stage is used in C.
(A) The - 1.8-volt potential at the ECL input
.-------------_----<) + 5 V established by the' resistor network
prevents the transistor from turning on
when the TTL output is LOW. A ger-
VCC
manium diode provides a stiff voltage
IN reference in the LOW state and prevents ex-
TTL
cessive conduction in the upper transistor
of the IC output structure. Th~ voltage
lN270 translation process is similar to that in
part 8 of the figure. Ret~rning the collec-
ECl tor to + 5V rather than ground keeps the
OUT
transistor well out of saturation.
tk 390k (B)
ECL Driving TTL
The complementary output of .ECL
-5.2V~--~~--------~--------~ gates can be used to advantage in con-
verting to TTL levels. Modern ECL ICs
have emitter-follower outputs that are
Fig. 124 - Bipolar transistors are used in these .ECl·to·TTl translators. ideal for switching the base-emitter junc-
tions of bipolar transistors. For coupling
logic systems having a common 5- or
5.2-volt power supply, the pnp transistor
+9V
, VDD
.. 3N187
and pull-up resistor combination of Fig.
124A may be used. A positive supply is
shown, but the system will also work with
negative supplies. The circuit in Fig. 1248
CMOS OUT will condition - 5.2-volt ECL signals to
drive + 5-volt TTL gates. Transposing the
out and OUt connections of the ECL device
Vss
will effect a logic inversion with the
translation. This technique can also be ap-
• GATE PROTECTED ECl plied to flip-flops, which have Q and Q
outputs.
-9V
VEE
CMOS Driving ECL
-5.2 V Speed is rarely a consideration when
(A) mating a relatively slow logic family to
+15V one that is very fast - the system cannot
be faster than the slowest logic element
v DD used. The speed of ECL ICs comes from
2N4403 keeping the transistor.s out of saturation,
2N3906
CMOS OUT and it idor this reason that the defined in-
2N2907
put logic swing is only about 400 mY.
However, . the input levels can be
Vss anywhere within the range of the power
15k supply without damaging the device.
VCC
Negligible input ~urrent is required for
IN
ECl either logic state, so when a common 5- or
.5.2-volt power supply is used, CMOS can
drive ECL directly.
10k VEE A variety of circuits can be used be-
tween CMOS and ECL systems having
-5.2 V
different power supplies. The scheme i1-
(B) luswtted in Fig. 125A I is useful when a
split power supply is used for the CMOS
Fig. 125 - Split·supply CMOS logic can drive ECl through a MOSFET. as in A. When the CMOS system logic. The advantage of using a MOSFET
is powered from a Single high·voltage supply. the bipolar transistor circuit of B can be used. Note that the converter is that the fanout (to other
logic is inverted when this method is used. CMOS devjces) is not compromised. Fig.
. Solid State Fundamentals 4·58
125B shows a + 15~volt CMOS system
+15V
driving a - 5.2-volt ECL gate through a
pnp transistor- Altering some of the
VCC
OUT
resistance values will make this circuit
work with split-supply CMOS as well.
·ECl This conversion method results in a logic
o,UT
inversion, but that problem can usually be
VEE remedied at the ECL output.
10k 10k· ECL-Driving CMOS
Some voltage amplification is required
-5.2 V
if an ECL gate is to drive CMOS. When
the ECL supply is negative and the CMOS
- (A) supply is positive, the circuit of Fig. 124B,
illustrated for ECL-to-TTL conversion
+5.2 V may be used. All of the resistors can be
made much larger with CMOS for re-
2N3906
duced power consumption.
2N4403
2N2907 The differential comparator arrange-
ment in Fig. 126A is another good transla-
VCC VOO
OUT
tion method. If the CMOS system has
IN
split power supplies, the - V terminal of
ECl CMOS the op amp should be returned to Vss.
OUT
With split CMOS supplies the op amp can
VEE VSS be connected directly to the CMOS input;
lOOk the resistor and diode are unnecessary. If
complementary ECL outputs aren't
available, one of the comparator inputs
should be biased to some potential be-
tween the two ECL logic levels.
Fig. 1268 shows a way to obtain the re-
(B) quired CMOS logic swing when both
families are powered from the same
Fig. 126 - Interface circuits for ECl-driving CMOS. A method useful when different power sup-
plies are used is illustrated in A. The diode prevents the - 5.2-volt LOW level from damaging the
source. This npn-pnp saturated amplifier.
CMOS device. When a common supply is available, the two-transistor amplifierltranslator of B will also work when a common negative
may be used. supply is used.
A - Voltage amplification Cibo - Input capacitance, open circuit ICED - Collector-cutoff current, base open
Ce - Intrinsic channel capacitance (common base) IE - Emittercurrent
Cds - Drain-to-source capacitance C ieo - Input capacitance, open circuit MAG - Maximum available amplifier gain
(includes approximately 1-pF drain-to- (common emitter) PCE - Total dc or average power input
case and interlead capacitance) C opo - Output capacitance, open circuit to collector (common emitter)
Cgd - ·Gate-to-drain capaci'~!nce (in- (common base) - large-signal output power
cludes O.1-pF interlead capacitance) C oeo - Output capacitance, open circuit ·(common emitter)
C gs - Gate-to-source interlead and case (common emitter) - Load resistance
capaCitance fe - Cutoff frequency - Source resistance
C iss - Small-signal input capacitance, iT - Gain-bandwidth product (frequency - Base-supply voltage
short circuit at which small-signal forward current- - Base-to-collector voltage
C rss - Small-signal reverse transfer transfer ratio, common emitter, is.unity - Base-to-emittervoltage
capacitance, short ci rcuit or1) - Collector-to-base voltage
gls - Forward transconductance gme - Small-signal transconductance - Collector-to-base (emitter open)
gis - Input conductance (common emitter) - Collector-supply voltage
gas - Output conductance hFB - Static forward-current transfer ratio - Collector-to-emitter voltage
10 - Dc drain current (common base) - Collector-to-emittervoltage (base
- Drain-to-source OFF current hlb - Small-signal forward-cwrrent open)
10S(OFF)
IGSS - Gate leakage current' transfer ratio, short circuit (common VCE(Sat) - Collector-to-emitter saturation
re - Effective gat8'series resistance base) voltage
rOS(ON) - Drain-to-source ON resistance hFE - Static forward-current - Emitter-to-base voltage
rgd - Gate-to-drain leakage resistance transfer ratio (common emitter) - Emitter-to-base voltage (collector
rgs - Gate-to-source leakage resistance hIe - Small-signal forward-current open)
IiOB - Drain-to-substrate voltage transfer ratio, short circuit (common ....:... Emitter!supply voltage
- Drain-to-source voltage emitter) - Forward transconductance
VOS
VGB - Dc gate-to-substrate voltage hiE - Static input resistance (common - Input admittance
VGB - Peak gate-to-substrate voltage emitter) - Output admittance
VGS - Dc gate-to-source voltage hie - Small-signal input impedance,
VGS .- Peak gate-to-source voltage short circuit (common emitter)
VGS(OFF) - Gate-to-source cutoff voltage Ib - Base current
Yls =
- Forward transadmittance 9fs Ie - Collector current
Yos - Output admittance ICBO - Collector-cutoff current, emitter
YL - load admittance open
4·59 Chapter 4
OPTICAL ELECTRONIC DEVICES
shaft-encoder is an array of open opto~ 117 V from' CMOS control signals are allows 'greater' precision. A digi tal display
couplers chopped, by a rotating wheel. manufactured by International Rectifier is an array of light sources that can be
When a shaft (mcoder is used to address a Company and others. energized in various combinations to form
frequency synthesizer, tht; operator can symbols. Some of the light source ar-
adjust the frequency in a manner that has Solid-State Displays rangements are illustrated in Fig. 129. The
the "feel" of an ordinary VFO. Qigital readout devices comprise an im- dot-matrix system is the most versatile,
An optoisolator packaged with a triac portant branch of optoelectronics. The but the seven-segment format is the one
forms a solid-state relay that can replace advantage of digital readout is that there most used by amateurs. It displays the
electromechanical units is most applica- is no interpretation error, as in an analog decimal digits with high readability_Each
tions. The advantages of this scheme in- readout such as a clock, speedometer or segment is identified by a letter, and Fig.
clude freedom from contact bounce, arc- ammeter. The digital readout can be no 130 shows the standard layout. In addi-
ing, mechanical wear and noise. Solid- more accurate than the circuitry driving it, tion to the segments, some displays
Solid State Fundamentals 4·60
ALPHANUMERIC
+5V
NUMERIC
•••
••• 1
IIUI ill
I:~I BC ::r.,. H,::t 16 EACH 150.1\.
;
~
5111 ARRAY 3ICS ARRAY 13 a ...
l-I
-IV'
..
......
17:~
6 12 P
MSB ().
V' , roo
1:=1 '"'
2 11
.Iv c ...
~
DATA
1 7447A 10 d
J--I B C I
I
2- =,-, LSB u-
'"'
~
7 9
"';\1\
-IV'v e .:...
14 BARS 7 BARS 15
" "V'
f ...
roo
.....
-
LAMP 3 14 9 ,
"-
TEST ......
Fig. 129 -
displays.
Various formats for symbolic
RIPPLE U-
BLANKING IN PUT
5_
- 8
... 4
-V RIPPLE
BLANKING
OlJTPUT COMMON - ANODE
DISPLAY
ri
A (AI
+5V
16 2N2907
G 2N4403
13
2N3906
E
7447A
c d f 9
o
Fig. 130 - Segment identification and layout
of a seven·segment readout device.
COMMON - CATHODE
LED Readouts b¥ Texas Instruments and' others. This is among several digits. The multiplexer
Displays made from light-emitting an· open-collector TTL device designed to logic, which is usually contained on a
diodes are the ones most commonly used pull down common-anode displays single IC, scans the input data lines and
because of their good readability and through external current-limiting re- sequentially routes the data for each digit
relative ease of Circuit integration. sistors. A 7447 A will also drive common- into the decoder. The displays are wired in
. Character heights of 0.3 and 0.6 inches are cathode displays if external transistors are parallel; that is, all of the "a" segments
standard, although larger units are used. Fig. 131 shows the connections for are connected together, and so on. The
available. LED readouts are manufac- both types of display.' common element or" each digit is enabled
tured with one element of each segment The dc illumination method shown is at the proper time by the multiplexing cir-
common. The character to be displayed is the easiest to implement, but higher light cuit that selects the input data. With this
usually encoded in BCD form, so it is output with lower energy consumption system, only one digit is energized at any
necessary to employ some combinational can be had by pulsing the display. A flash instant, a factor that greatly eases the
logic to illuminate the proper segments. rate of 100 Hz will be imperceptible power-supply requirements. In order to
The circuitry to accomplish this is called a because of the persistence of human vi- maintain the brightness of each digit, the
decoder. Various seven-segment decoders sion. As more digits are added to a dis- current to each LED segment must be in-
are manufactured to drive common~ play, using a decoder/driver for each digit creased. A typical display .should be
cathode and common-anode devices. becomes unfeasible from an economic operated at a peak current of 100 mA for
, Some of these devices contain advanced and pc-board real estate point of view. A each segment, with a pulse duration of 50
'features. The decoder IC most available technique called multiplexing allows a to 250 microseconds at a 20-percent duty
to hobbyists is the 7447 A, manufactured single decoder/driver IC to be time-shared cycle.
4·61 Chapter 4
Chapter 5
Ac-OperatedPower Supplies
Power-line voltages have been "stan- used in some applications (notably ac-dc being connected between one wire and the
dardized" throughout the u.s. at 117 and radios and some television receivers), neutral, while the other half of the load is
234 volts in residential areas wh«re a Supplies of this sort opetate directly from connected between the other wire and
single phase voltage is supplied. These the power line, making it necessary to neutral. Heavy appliances, such as electric
figures represent nominal voltages. "Nor- connect the chassis or common-return stoves and heaters are designed for
mal" line voltage in a particular area may point of the circuit directly to one side of 234-volt operation and therefore are
be between approximately 110 and 125 the ac line. This type of power supply connected across the two ungrounded
volts, but generally will be above 115 volts. represents a shock hazard when the wires. While both ungrounded wires
In many states the service is governed by a equipment is connected to other units in should be fused, a fuse should never be
•
pUC (public utilities commission). The the amateur. station or when the chassis is used in the neutral wire, nor should a
voltage average across the country is ap- exposed. For safety reasons, an isolation switch be used in this side of the line. The
proximately 117. transformer should be used with such reason for this is that opening the neutral
The ac-current capability of the· service equipment. wire does not disconnect the equipment. It
is a factor of line length from the dwelling simply leaves the equipment on one side
. to the nearest pole transformer\ plus the Power-Line Considerations: Connections of the 234-volt circuit in series with
conductor size of the line. Many older In most residential systems, three wires whatever load may be across. the other
homes are suppfied with a 6O-ampere . are brought in from the outside to the side of the circuit, as shown in Fig. lB.
service while most new homes have 100 distribution board, while in a few older Furthermore, with the neutral open, the
amperes. Houses equipped with electric systems there are only two wires. In the voltage will then be divided between the
heat will have services ranging from 150 to three-wire system, the third wire is the two sides in inverse proportion to the load
200 amperes. neutral, which is grounded. The voltage resistance, the voltage on one side
The electrical power required to operate between' the two wires ,normally is dropping below normal, while it soars on
Amateur Radio equipment is usually 234, while half of this voltage appears the other side, unless the loads happen to
taken from the ac lines when the equip- between each of these wires and neutral, be equal.
ment is operated where power is available. as indicated in Fig. lA. In systems of this The usual line running to ~aseboard
For mobile operation the source of power type the 117-voIt household load is outlets is rated at 15 amperes. Considering
is almost always the car storage battery. divided as evenly as possible between the the power consumed by filam«nts, lamps,
Dc voltages used in transmitters, recei- two sides of the circuit, half of the load transmitter, receiver and other auxiliary
vers and other related equipment are derived
from the commercial ac lines by using a
transformer-~ectifier-fi1ter system. The trans-
former changes the ac voltage to a suitable
value and the rectifier converts the ac to
pulsatingdc. A filter is u,sed to smooth out
~
these pulsations to an acceptably low level.
Essentially pure direct current is required
to prevent 60- or 120-Hz hum in most
pieces of amateur equipment. Transmit-
ters must be operated from a pure dc sup-
ply as dictated by federal regulations. If a
------ ------
nn
constant voltage is required under condi- ----- T1=====
S
tions of changing load or ac-line voltage, a
regulator is used following the filter.
When the prime power source is dc (a
battery), the dc is used directly or is first
(Al (e) (O)
changed to ac and is then followed by the
transformer-rectifier-filter combination.
The latter system has lost considerable
Fig. 1 - Three-wire power-line circuits. At A - Normal three-wire-line termination. No fuse should be used in
popularity with the advent of low-voltage the grounded (neutral) line. B - A switch in the neutral does not remove voltage from either side of the line, C
semiconductor devices. ' - Connections for both 117- and 234-volt transformers. D - Operating a 117-volt plate transformer
Transformerless power supplies are from the 234-volt line to avoid light blinking. T1 is a 2:1 step-down transformer.
5·2 Chapter 5
DIODE '
CONOUCT~n {\ {\
(j \ I \ I
~\,' \,'
DIODE / \j \j EPEAK·1,4E.IoI.
NON-CONDUCTION EAV - 0.45ERMS
(A) (8) RIPPLE .121 % IBOH.
T o
nn I I I
EPRV-i.4ERMS T
,~~~~:~~ 0
,,
,, \
' ,
,, , I' ,
: , \
\
EPRV- 2.8E.IO.
I
\ I \
1
REVERSE I
VOLTAGE \ I \ I , I
\
, II
..L
'-' '-' \~I
'-
,
]11El·. ·
(C) • (0)
Fig. 4 - Half-wave rectifier circuit. A illustrates the basic circuit and 8 displays the diode conduction
and nonconduction periods, The peak-reverse voltage impressed across the diode is shown at C and
o with asimple resistor load at C and a capacitor load at D. Eprv for the reSistor load is f4 Erms and
TO 2.8 Erms for the capacitor load, ,
EQUIP.
(C)
the main power switch, . or a similar The peak reverse voltage (PRV), the
arrangement at the bench, if the bench is voltage that the rectifier must withstand
Fig. 3 - Reliable arrangements for cutting off all located remotely from the transmitter. when it isn't conducting, varies with the
power to the transmitter. S is an enclosed A bleeder resistor with a power rating load. With a resistive load it is the peak ac
double-pole power switch, J a standard ac outlet,
P a shorted plug to fit the outlet and I a red lamp. which gives a considerable margin of voltage (1.4 Erms) but with a capacitor
A is for a two-wire 117 -volt line, B for a three- safety should,be used across the output of filter and a load drawing little or no
wire 234-volt system, and C a simplified all transmitter power supplies, so that the current it can rise to 2.8 Erms. The reason'
arrangement for low-power stations. filter capacitors will be discharged when for this is shown in Figs. 4C and
the high-voltage is turned off. 40. With a resistive load as shown at C
the amount of reverse voltage applied to
prominently ,in plain sight, and members Rectifier Circuits: Half-Wave the diode is that voltage on the lower side
of the household should be instructed in Fig. 4 shows a simple half-wave rectifier of the Zero-axis line or 1.4 Erms. A capaci-
its location and use, I is a red lamp located circuit. As pointed out in the semi- tor connected to the circuit (shown at D)
alongside the switch. Its purpose is not so • conductor chaptet a rectifier (in this cas,e a )Vill store the peak positive voltage when
much to serve as a warning that the power semiconductor diode) will conduct cur- the diode conducts on the positive pulse.
is on as it is to help in identifying and rent in one direction but not the other. H the circuit is not supplying any current
quickly locating the switch should it During one half of, the ac cycle the the voltage across the capacitor will re-
become necessary for someone else to cut rectifier will conduct and current will flow main at that same level. The peak reverse
the power off in an emergency. through the rectifier to the load (indicated voltage impressed across the diode is now
The outlet J should be placed in some by. the solid line in Fig. 4B). During the the sum of the voltage stored in the capa-
corner out of sight where it will not be a othor half cycle the rectifier is reverse citor plus the peak negative swing of volt-
temptation for children or others to play biased and no current will flow (indicated age from the transformer secondary. In
with. The shorting plug can be removed to by the dotted line in Fig. 4B) to the load. this case the PRY is 2.8 Er~s.
open the power circuit if there are others As shown, the output is in the form of
around, who might inadvertently throw pulsed dc and current always, flows in the FuU-Wave Center-Tap Rectifier
the switch while the operator is working same direction. A filter can be used to A commonly used rectifier circuit is
on the rig. If the operator takes the plug smooth out these variations and provide a shown in Fig. 5. Essentially an arrange-
with him, it will, prevent someone from higher average dc voltage from the circuit. ment in which the outputs of two half-
turning on the power in his absence and This idea will be covered in the next wave rectifiers are c'ombined, it makes use
either hurting himself or the equip- section on filters.' of both halves of the ac cycle. A .ransfor-
ment or perhaps starting a fire. Of utmost The average output voltage - the mer with a center-tapped secondary is re-
importance is the fact that the outlet J voltage read by a dc voltmeter - with this quired with the circuit.
must be placed in the ungrounded side of circuit (no filter connected) is 0.45 times The average output voltage is 0.9 times
the line. the rms value of tl)e ac voltage delivered the rms voltage of half the transformer
Those who are operating low power by the transformer secondary. Because the secondary; this is the maximum that can
and feel that the expense or complication 'frequency of the pulses is rather low (one be obtained with a suitable' choke-input'
of the switch isn't warranted can use the pulsation per cycle), considerable filtering filter. The peak output voltage is 1.4 times
shorted-plug idea as the main power is required to provide adequately smooth the rms voltage of half the transformer
switch. In this case, the outlet should be dc output. For this reason the circuit is secondary; this is the maximum voltage
located prominently and identified by a usually limited to applications where the that can be obtained from a capacitor-
signal light, as shown in Fig. 3C. current required is small, as in a input filter.
The test bench should be fed through transmitter bias supply .. As can be seen in Fig. 5C the PRV
I
in Figs. 6A and B, when the top. lead of
the transformer secondary is positive with , I, I, I
EPRY'2.8 ERMS ~
DIODE
...'
NON-CONDUCTION
Since the diodes are connected in series
each diode has 1.4 E rms as the reverse volt- (A) (B)
~
A.C
A·
.0
IB.DflA,C, ,B.D" _
1/1111"
\../ \.... '
2.8ERMS
A.C
(BOTH
I
EPRY·1.4ERMS
DIODES)
(EACH DIOOE)
(c)
Filtering
The pulsating dc waves from the Fig. 6 - Full-wave bridge rectifier circuit. The basic circuit is illustrated at A. Diode conduction and
non conduction times are shown at B. Diodes A and C conduct on one half of the input'cycle while
rectifiers are not sufficiently constant in diodes Band D conduct on the other. C displays the peak-reverse voltage for one-half cycle. Since
amplitude to prevent hum corresponding this circuit uses two diodes essentially in series, the 2.8 Erma is divided between two diodes, or, 1.4
to the pulsations. Filters are required be- Erms PRY for each diode. .
5-4 Chapter 5
must have a voltage rating safe for the component. Series chokes will readily pass the load resistance is high. Silicon recti-
highest value to which the voltage will \ dc but will impede the flow of the ac fiers have ~ higher allowable peak-to-dc
soar when the external load is removed. component. 'ratio than do thermionic rectifiers. This
A power supply will show more (higher) The alternating component is called permits the use of capacitor-jnput filters
regulation with long-term changes in load ripple. The effectiveness of the filter can be at ratios of.input capacitor to load resis-
resistance than with short temporary expressed in terms of. percent ripple, tance that would seriously shQrten the life
changes. The regulation with long-term which'is the ratio of the rms value of the of a thermionic rectifier system. When the
changes is often called the static regula- ripple to the dc value. in terms of series resistance through a rectifier and
tion, to distinguish it from the dynamic ' percentage. filte,r system is appreciable, las when high-
regulation (short temporary lo~ changes). vacuum rectifiers are used, the voltage re-
A load that varies at a syllabic or keyed l00E 1 gulation of a capacitor-input power sup-
Percent ripple (rms) = -E-
rate, as represented by some audio and rf 2 ply is poor.
amplifiers, usually requires good dynamic where The output voltage of a properly
regulation (15 percent or less) if distortion E\ = the rms value of ripple voltage designed choke-input power supply is less
products are to be held to a low level. The E2 = the steady dc voltage than would be obtained with a capacitor-
dynamic regulation of a power supply is input filter from the same transformer.
improved by increasing the value of the Any multiplier or -amplifier supply in a Generally speaking, a choke-input filter
output capacitor. code transmitter should have less than five will permit a higher load current to be
When essentially constant voltage re- . percent ripple. A linear amplifier can drawn from a, thermionic rectifier without
gardless of current variation is required tolerate about three percent ripple on the exceeding the peak rating of the rectifier.
(for stabilizing an oscillator, for example), plate voltage. Bias ·supplies for linear
special voltage-regulating circuits de- amplifiers, and modUlator and modulated- Capacitive-Input Filters
scribed later in this chapter are used. amplifier plate supplies, should have less Capacitive-input filter systems are shown
than one percent ripple. VFOs, speecp in Fig. 7. Disregarding voltage drops in
Bleeder amplifiers and receivers may require a the chokes, all have the same cqaracteris-
A bleeder resistor is a 'resistance ripple reduction to 0.01 percent. ·tics except in respect to ripple. Better
connected across the output terminals of Ripple frequency is the frequency of the ripple reduction will be obtained when LC
the power supply. Its functions are to pulsations in the rectifier output wave - sections are added as shown in Figs. 7B
discharge the filter capacitors as a safety the number of pulsations per second. The and C.
measure when the power is turned off and frequency of the ripple with half-wave
to improve voltage regulation by pro- rectifiers. is the same as the frequency of Output Voltage
viding a minimum load resistance. When the line supply - 60 Hz with 60-Hz To determine the approximate dc
voltage regulation is not of importance, supply. Since the output pulses ar"e voltage output when a capacitive-input
the resistance may be as high as 100 ohms doubled with a full-wave rectifier, the filter is used, the- graphs shown in Fig. 8
per volt. The resistance value to be used ripple frequency is doubled - to 120 Hz will be helpful. An example of how to use
for voltage-regulating purposes is discussed with a 60-Hz supply. the graph is given below.
in later sections. From the consideration The amount of filtering (values of Example:,
of safety, the power rating of .the resistor inductance and capacitance) required to ,Full-wave rectifier (use graph at B)
should be as conservative as possible, give adequate smoothing depends upon Transformer rms voltage = 350
since a burned-out bleeder resistor is more the ripple frequency, with more filtering Load resistance = 2000 ohms
dangerous than none at all! being required as the ripple frequency is Series resistance = 200 ohms
lowered. Input capacitance = 20 II- F
Ripple Frequency and Voltage
Pulsations at the output of the rectifier Type f!! Filter
can be considered to be the resultant of an Power-supply filters fall into two ..!. = 200 = OJ RC = 2000 X 20 =40
RS 2000, 1000 1000
alternating current superimposed on a classifications, capacitor input and choke
steady direct current. F,om this view- input. Capacitor-input filters are charac-
From curve 0.1 and RC = 40, the dc vol-
point, the filter may be considered to terized by relatively high output voltage in tage is (350 X 1.06) = 370.
consist of shunt capacitors which short- resp¢ to the transformer voltage. Advan- In many cas~s it is desirable to know
circuit the ac component while not tage of this can be taken when silicon rec-
the amount of capacitance required for a
interfering with the flow of the dc tifiers are used or with any rectifier when power supply' given certain performance
criteria. this is especially true when
designing a' power supply for an ap-
plication such as powering a solid-state
transceiver. The following example should
+o---~~---.----~+ +0----.'-' .......>---_---0+ give the builder a good' handle on how to
FROM DC FROM
arrive at circuit values for a power supply
DC
RECTiFIER OUTPUT RECTIFIER OUTPUT using a single capacitor filter.
Fig. 9 is, the circuit diagram of the
power supply to be used.
Requirements:
Output voltage =12.6
+0-----411---'
Output current = 1 ampere
Maximum ripple = 2 percent
FROM'
RECTIFIER
C2 DC Load regulation = 5 percent .
1
OUTPUT
The rms secondary voltage of Tl must
-O---4-----------~---~~--~~~O be the desired output voltage plus the volt-
(el' m age drops across D2 and D4 divided by
1.41. '
Fig. 7 - Capacitive-input filter circuits, At A is a simple capacitor filter, Band C are single· and _12.6 +1.4 -
ESEC- -993
.
double-section filters, respectively, 1.41 ,
C
(jlF)
ILt J 10·
= [ Erip(pk-pk). 0.1 0.3 0.5
FlC
iOOci
3
.
5 10
(A)
=. [IA X 8.3 X 10-
3
] 10.
0.7
= 11,857 #'
1.4F====tr=u..::.==:r:;nlTl TTI1:Ej;E::::::=t==+==f1Ti=m°,0005
0.005
whe~e IL is the current taken by the load.
The nearest standard capacitor value is
12,000 /-IF. It wiII be an acceptable one to
use, but since the tolerance of electrolytic
1.2
capacitors is rather loose, the builder may
elect to use the next larger standard value.
Diodes 01-04, inclusive, should have a
1.1
PR V rating of at least two times the i
. transformer secondary peak voltage. As- ---'-""-'-rt-r""7'"--r~-r-rj~,,:7'1,0.1
suming a transformer secondary rms 1.0~ .~_';"""'-;"";""-""""--"""T'-...,...."....,....,.~', '0.125
value of 10 volts, the PR V should be at _------:-;-------0.15
least 28 volts. Four 50-volt diodes will ! I I I I IIIII
0.4i;:,--
10 = 0.063 ohm
=0.05. (126)
Fig. 8 - Dc output voltages from a half- and full-wave rectifier circuit as a function of the filter
capacitance and load resistance (half-wave Shown at A and full-wave shown at. B). Rs includes
Therefore, the transformer secondary dc transformer winding resistance and rectifier forward resistance. For the ratio Rs/R, both resistances
resistance should be no greater than 0.063 are in ohms; forthe RC product, R is in ohms and C is in uF.
5-6 Chapter 5
ohm. The secondary current rating should Ft 02.
be equal to or greater than the IL = 1 am-
pere. 01
C 1 should have a minimum working H7
voltage of 1.4 times the output voltage. In VAC
the case of this power supply the capacitor
should be rated for at least 18 volts.
ON
Choke-Input Filters . ,.to.Ot
With, thermionic rectifiers better volt-
age regulation results when a choke-input Eo (no load) =Esee X 1.41 Cl (E min ) =>Eo X 1.41
filter, as shown in Fig. 10, is used. Choke Po =Eo X lu PI (A) = 2I1N (N = turns ratio)
input permits better utilization of the RL =Eo+IL Esee ~ Eo + 1.41
thermionic rectifier, since a higher load
current can be drawn without exceeding
the peak current rating of the rectifier. Fig. 9 - This figure illustrates how to design a simple unregulated power supply. See text for a
thorough discussion.
Minimum Choke Inducttlllce
A choke-input filter will tend to act as a
capacitive-input filter unless the input
choke has at least a certain minimum
value of inductance called the critical
value. Thill critical value is given by r
EO
L erit (henrys) -~ ~
E (volts)
__~
. I (rnA) i
where E = the supply output voltage
Fig. 10- Diagram showing various voltage drops that must be taken into consideration in determining
I = the current being drawn through the required transformer voltage to deliver the desired output voltage.
the filter.
If the choke has at least the critical load current of 200 rnA is drawn from the "F or more is usually adequate. When
value, the output voltage will be limited to supply, the inductance will drop to 5 H. the supply is used with a Class B amplifier
the average value of the rectified wave at The critical inductance for 200 rnA at (for modulation pr for ssb amplification)
the input to the choke when the current 1000 volts is 1000/200 = 5 H. Therefore or a cw transmitter, increasing the output
drawn from the supply is small. This is in the 5/25 H choke. maintains the critical capacitance will result· in improved dy-
contrast to the. capacitive-input filter in inductance at the full current rating of 200 namic regulation of tpe supply. However,
which the output voltage tends to soar rnA. At all load currents between 40 rnA a region of diminishing returns can be
toward the peak value of the rectified and 200 rnA, the choke will adjust its in- reached, and 20 to 30" F will usually
wave at light loads. ductance to the approximate critical value. suffice for any supply subjected to large
changes at a syllabic (or keying) rate.
Minimum-Load - Bkeder Resisttlllce Output Voltage
From the formula above for critical in- Provided the input-choke inductance is Resonance
ductance, it is obvious that if no current is , at least the critical value, the output Resonance effects in the series circuit
drawn from the supply, the critical induc- voltage may be calculated quite closely by: across the output of the rectifier, formed
tance will be infinite. So that a practical
Eo = O.9Et - (18 + IL ) X (Rl + R2) - Er by the first choke and first filter capacitor,
value of inductance may be used, some must be avoided, since the ripple voltage
current must be drawn from the supply at where would build up to large values. This not
all times the supply is in use. From the Eo = output voltage only is the opposite action to thai for
formula we find that this minimum value Et = rms voltage applied to the recti- which the filter is intended, but may also
of current is fier (rms voltage between center-tap cause excessive rectifier peak currents and
and one end of the secondary in the abnormally high peak-reverse voltages.
I (rnA) = E (volts) case of the center-tap rectifier)
L erit For full-wave rectification the ripple fre-
IB = bleeder current (A) quency will be 120 Hz for a 60-Hz supply,
In the majority of cases it will be most IL = load current (A) and resonance will occur when the pro-
convenient to adjust the bleeder resistance RJ = first filter choke resistance duct of choke inductance in henrys times
so that the bleeder will draw the required R2 = second filter choke resistance capacitor capacitance in microfarads is
minimum current. From the formula, it Er = voltage drop across the rectifier. equal to 1.77. At least twice this Prodllct'
may be seen that the value of critical in- of inductance and capaCitance should be
ductance becomes smaller as the load cur- The various' voltage drops !lre shown in used to ensure against resonance effects.
rent increases. Fig. 10. At no load IL is zero; hence the With a swinging' choke, the minimum
no-load voltage may be calculated on the rated inductance of the choke should be
Swinging Chokes basis of bleeder current only. The voltage used. If too high an LC filter product is
Less costly chokes are available that regulation may be determined from the used, the resonance may occur at the
will maintain at least the critical value of no-load and full-load voltages using the radio-t~legraph keying or voice syllabic
inductance over the range of current likely formulas previously given. rate, and large voltage excursions (filter
to be drawn from practical supplies. These bounce) may be experienced at that rate.
chokes are called swinging chokes. As an Output Capacitor
, example, a swinging choke may have an Whether the supply has a choke- or Ratings of Filter Components
inductance rating of 5/25 Handa current capacitor-input filter, if it is intended for In a power supply using a choke-input
rating of 200 rnA. If the supply delivers . use with a Class A af amplifier, the filter and properly designed choke and
1000 volts, the minimum load current !eactance of the output capacitor should bleeder resistor, the no-load voltage
should be 1000/25 = 40 rnA. When the full be low for the lowest audio frequency; 16 across the filter capacitors will be about
"
Ac· Operat.ed Power Supplies 5-7
+E2
'---t------O+E1
Fig. 11 - In most applications, the filter chokes may be placed in the negative instead of the positive
side of the circuit. This reduces the danger of a voltage breakdown between the choke winding and '
core.
Fig. 12 - The "economy" power.supply circuit is
a combination of the full-wave and bridge-
rectifier circuits.
NON-CONDUCTING
JIlOO
.....-T"'""VVV-"'" I-''+-~_--o +
the positive leads of plate power supplies.
:JII +
EDC
(s) CONDUCTING
This means that the insulation between
the choke winding and its core (which
should be grounded to chassis as a safety
measure) must be adequate to withstand
the output voltage of the supply. This
voltage requirement is removed if the
chokes are placed in the negative lead as
shown in Fig. 11. With this connection,
2.8 EPEAK-2.8ERMS the capacitance of the transformer secon-
EPRV a 2.8 EAIIS
dary to ground appears in parallel with
the filter chokes tending to bypass the
chokes. However, this effect will be
D negligible in practical application except
(O)
in cases where the output ripple must be
(e)
reduced to a very low figure. Such
applications are usually limited to low-
Fig. 13 - Illustrated at A is a half-wave voltage-doubler circuit. B displays how the first half cycle of voltage devices such as receivers, speech
input voltage charges C 1. During the next half cycle (shown at C) capacitor C2 is'charged with the amplifiers and VFOs where insulation is
transformer seeondary voltage plus that voltage stored in C 1 from the previous half cycle. D illustrates
the levels to which each capacitor is charged throughout the cycle.
no' problem and the chokes may be placed
in the positive side in the conventional
manner. In higher-voltage applications,
there is no reason why the filter chokes
should not be placed in the negative lead
nine-tenths of. the ac rms voltage. N ever- is used,' it is important that each of the to reduce insulation requirements. Choke
theless, it is advisable to use' capacitors capacitors be shunted with a resistor of terminals, negative capacitor terminals
rated for the peak transformer voltage. about 100 ohms per volt of supply voltage and the transformer center-tap terminal
This large safety factor is suggested applied to the individual capacitors, with should be well protected against acciden-
because the voltage across the capacitors an adequate power rating. These resistors tal contact, since these will assllme full
can reach this peak value if the bleeder may serve as all or part of the bleeder supply voltage to chassis should a choke
should burn out and there is no load on resistance. Capacitors with higher voltage burn out or the chassis connection fail.
the supply. ratings usually are made with a dielectric
In a cap active-input filter, the capaci- of thin paper impregnated with oil. The The "Economy" Power Supply
tors should have a working-voltage rating working voltage of a capacitor is the In many transmitters of the l00-watt
at least as high, and preferably somewhat voltage that it will withstand continuously. class, an excellent method for obtaining
higher, than the peak voltage from the plate and screen voltages without wasting
transformer. Thus, in the case of a center- FUter Chokes power in resistors is by the U8~ of the
tap rectifier having a transformer deliver- Filter chokes or inductances are wound "economy" power-supply circuit. Shown
ing 550 volts each side of the center tap, on imn cores, with a small gap in the core in Fig. 12, it is a combination of the full-
the minimum safe capacitor voltage rating to prevent magnetic saturation of the iron wave and bridge-rectifier circuits. The volt-
will be 550 X 1.41 or 775 volts. An 800-volt at high currents. When the iron becomes age at El is the normal voltage obtained
capacitor should be used, or preferably a saturated its permeability decreases, and with the full-wave circuit, and the voltage
WOO-volt unit. consequently the inductance also decreases. at E2 is that obtained with the bridge cir-
Despite the air' gap, the inductance of a cuit. The total dc power obtained from the
/i'liter ClIJHICltors In Series choke usually varies to some extent with transformer is, of course, the same as
Filter capacitors are made in several the direct current flowing in the winding; when the transformer is used in its normal
different types. Electrolytic capacitors, hence it iS,necessary to specify the induc- manner. In' cw and ssb applications, addi-
which are available for peak voltages up tance at the current which the choke is tional power can usually be drawn with-
to about 800, combine high capacitance intended to carry. Its inductance with little out excessive heating, especially if the
with small size, since tl}e dielectric is an or no direct current flowing in the winding transformer has a rect~fier filament wind-
extremely tI:tin film of oxide on aluminum will usually be considerably higher than - ing that isn't being used.
foil. Capacitors of this type may be the value when full load current is flow-
connected in series for higher voltages, ing. Half-Wave Voltage Dou~ler '
although the filtering capacitance will be Fig. 13 shows the circllit of half-wave
reduced to the resultant of the two Negadve-Lead Filtering voltage doubler. Figs. 13B, C and D
capacitances in series. If this arrangement For many years it has been almost illustrate the circuit operation. For clarity.
5-8 Chapter 5
assume the transformer voltage polarity at voltage, DAis cutoff and diode DB con- the capacitors will remain charged - €1
the moment the circuit is activated is that ducts charging capacitor C2. The amount to 1.4 E rms and C2 to 2.8 .E rms . When a
shown at B. During the first negative half of voltage delivered to C2 is the sum of load is connected to the output of the
cycle DA conducts (DB is in a noncon- peak secondary voltage of the transformer doubler, the voltage across C2 drops.
ductive state), charging Cl to the peak plus the voltage stored in Cl (1.4 Erms)' during the negative half cycle and is
rectified voltage (1.4 Erms). Cl is charged On the next negative half cycle, DB is recharged up to 2.8 Erms during the
with the polarity shown at B. During the nonconducting and C2 will discharge into positive half cycle.
positive half cycle of the secondary the load. If no load is connected across C2 The output· waveform across C2 re-
sembles that of a half-wave rectifier cir-
. cuit in that C2 is pulsed once every cycle.
The drawing at Fig. 13D illustrates the
R DA
levels to which the two capacitors are
+ charged throughout the cycle. In actual
:J
operation the capacitors will not dis-
+ C1 charge all the way to zero as shown.
LINE 1.4 ERMS
J
During the negative half cycle, as
shown at C, DB conducts charging
1.4 ERMS
capacitor C2 to 1.4 Erms while 0 A is
nonconducting. The output voltage is the
(01
sum of the two capacitor voltages which
+ C2
1.4 ERMS will be 2.8 Erms under no-load conditions.
CONDUCTING Fig. 14D illustrates that each capacitor
(el alternately receives a charge once per
cycle. The effective filter capacitance is
that of C I and C2 in series, which is less
Fig. 14 - A full-wave voltage doubler is displayed at A. One half cycle is 'shown at B and the next than the capacitance of either C I or C2
half cycle at C. Eaeh capacitor receives a charge during every cycle of input voltage. D illustrates
how each capacitor is alternately charged. alone.
Resistors R in Fig. 14A are used to limit
the surge 'cu.rrent through the rectifiers.
Their values are based on the transformer
voltage and the rectifier surge-current
rating, since at the instant the power sup-
2.8
. it 0.001
ply is turned on the filter capacitors look
like a short-circuited load. -Provided the
11
2.6
2.4
.JIIGJqJI1
• C
I
I
V
I
~
j.---- t-
i
! ,
0.01
limiting resistors can withstand the surge' -
current, their current-handling capacity is
based on the maximum load current from
the supply. Output voltages approaching
V I
twice the peak voltage of the transformer
2.2
Vv I, I 1 can be obtained with the voltage doubling
VV I circuit shown in Fig. 14. Fig. 15 shows
Vi :
I
0.03 how the voltage depends upon the ratio of
~
2.0
the series resistance to the load resistance,
~.J
iL .[L'V
~ , and the load resistance times the filter
~>
g:; 1.8
WV RS
R
capacitance. The peak reverse voltage
,,,,, / / across each diode is 2.8 E rms .
,
O~
1.6
/ I
Voltage TrlpHng and Quadrupling
'il
-
A voltage-tripling circuit is shown in
0.1
/ Fig. 16A. On one half of the ac cycle Cl
1.4
v/ and C3 are charged to the source voltage
through D 1, D2 and 03. On the opposite
1.2
V half of the cycle D2 conducts and C2 is
Iv charged to twice the source voltage,
3
VIV 10 30 ~o 100 300 roo 1000
because it Sees the transformer plus the
charge in Cl as its source. (Dl is cutoff
1~;O (R IN OHMS. C IN pF) during this half cycle.) At the same time,
03 conducts, and with the transformer
and the charge in C2 as the source, C3 is
Fig. 15 - Dc output voltages from a full-wave voltage-doubling circuit as a function of the filter
capacitances and load resistance. For the ratio As/R and for the RC product, resistances are in
charged to three times the transformer
ohms and capacitance is in mlcrofarads .. Equal resistance values for Rs and equal capacitance voltage.
values for C are assumed. ' The voltage-quadrupling circuit of
Ac - Operated· Power Supplies 5-9
Fig. 16B works in substantially similar This does not apply, of course, to
fashion. In either of the circuits of Fig. 16, amateur transmitter plate' transformers,
the output voltage will approach an exact which usually are rated for intermittent
multiple of the peak ac voltage when the service ..
output current drain is low and the
capacitance values are high. Rewinding Power Transformers
In the circuits shown, the negative leg
Although the home witiding of power
of the supply is common to one side of the
transformers isa task that few amateurs (A)
transformer. The positive leg can be made
undertake, the rewinding of a transformer
common to one side of the transformer by
secondary to give some desired voltage for
reversing .the diodes and capacitors.
powering filaments or a solid-state device
is not difficult. It involves a matter of only ~'V\"""'o--o~~I-.--o+EDC
Plate and Filament Transformers:
Volt-Ampere Rating a small number of turns and the wire is
The number of volt-amperes delivered large enough to be handled easily. Often a
by a transformer depends upon the type of receiver power transformer with a burned~
filter (capacitor or choke input) used, and out high-voltage winding, or the power
upon the type of rectifier used (full-wave transformer from a discarded TV set can
center tap, or full-wave bridge). With it be converted into an entirely satisfactory (8)
capacitive-input filter the heating effect in transformer without great effort and with
the secondary is higher because of the little expense. The average TV power Fig. 16 - Voltage-multiplying circuits with one
high ratio of peak-to-average current. The transformer for a 17-inch or larger set is sidE! of trans/1)rmer secOndary common. (A)
volt-amperes handled by the transformer capable of delivering. from 350 to 450 Voltage tripler; (8) voltage quadrupler.
may be several times the watts delivered to watts, continuous duty. If an amateur
Capacitances are typically 20 to 50,..F
depending upon output current demand. Dc
the load. With a choke-input filter, transmitter is being powered, the service is ratings of capacitors are related to
provided the input choke has at least the not continuous, so. the rati~gs can be Epeak (1.4 Eac)·
critical' inductance, the ~econdary volt- increased by a factor of 40 or 50 percent C1 - Greater than El1eak
amperes can be calculated quite closely by without danger of overloading the trans-
C2 - Greater than 2Epeak
former. C3 - Greater than 3E pea k
the equation: C4 - Greater than 4Epeak
The primary volt-ampere rating of the
(Full-wave et) Sec VA= O'i~I
transformer to be rewound, if known, can
be used to determine its power-handling
capability. The secondary volt-ampere
EI rating will be 10 to 20 percentJess than the
(Full-wave bridge) Sec V A = 'iOO'O primary rating. The power rating may
also be determined approximately from
where the cross-sectional area of the core which
E = total rms voltage of the secondary is inside the windings. Fig. 17 shows the
(between the outside ends in the case method of determining the area, and Fig.
of a center-tapped winding) 18 may be used to convert this in-
= dc output' current in milliamperes formation into a power rating.
(load current plus 'bleeder current) Before disconnecting the winding leads
from their terminals, each should be
The primary volt-amperes will be some- marked for identification. In removing the
what higher because of transformer losses. core laminations, care should be taken to
note the manner in which the core is
assembled, so that the reassembling will
Broadcast and Television Replacement be done in the same manner. Most
Transformers transformers have secondaries wound
Small power transformers of the type over the primary, while in some the order CROSS-SECTIONAL AREA.
sold for replacement in brqadcast. and is reversed. In,case the secondaries are on WIDTH X HEIGHT (WXH) OF CORE
television receivers are usually designed the inside, the turns can be pulled out
for service in terms of use for several from the center after slitting and removing Fig. 17 - Cross-sectional drawing of a typical
hours continuously with capacitor-input the fiber core. . power transformer. MuJtiplyirig the height (or
filters. In the usual type of amateur The turns removed from one of the thickness of the laminations) by the width of
transmitter service, where most of the original filament windings of known the central core area in ,Inches gives the value
to be applied to Fig. 18.
power is drawn intermittently for periods voltage should be carefully counted as the
of several minutes with equivalent inter- winding is removed. This will give the
vals in between, the published ratings can number of turns per volt and the same
be exceeded without excessive transformer figure should be used in dete~mining the amateur-service transformers is' 700 cmil/
heating. number of turns for the new secondary. A. The larger· the cmil/ A figure, the cooler
With a capacitor-i~put filter, it should For instance, if the old filament winding the transformer will run. The current
be safe to draw 20 to 30 percent more was rated at 5 volts ·and had 15 turns, this rating in amperes of various wire sizes is
current than the rated value. With a is 15/5 = 3 turns per volt. If the new shown in the copper-wire table in another
choke-input filter, an increase in current secondary is to deliver 1St volts, the chapter. If the transformer being rewound
of about 50 percent is -permissible. If a required number of turns on the new is a filament transformer, it may be
bridge rectifier is used, the output voltage winding will be 18 X 3 = 54 turns. necessary to choose the wire size carefully
will be approximately doubled. In this' In winding a transformer, the size of to fit the small available space. On the
case, it should be possible in amateur wire is an important factor in the heat other hand~ if the transformer is a power
transmitter service' to draw the rated developed in operation. A cross-sectional unit with the high-voltage winding re-
current, thus obtaining about tWice the area of 1000 circular mils per ampere is m,oved, there should be plenty of r.oom for
rated output power from the transformer. .conservative. A value commonly used in a size of wire that will conservatively
5-10 Chapter 5
.han~le the required current.· 4
After die first layer of turns is put on
duririg rewinding, secure the ends with
II)
3-3/4
~
'"J:
cellulose tape. Each layer should be
insulated from the next; ordinary house-
<.l
~ 3-1/2
V
~
hold waxed paper can be used for the '"
Q:
<I
3-114 /'"
purpose, a single layer being adequate. :>
0 ./
V
Sheets cut w size beforehand may be II) 3
secured over each layer with tape. Be sure ~
2-3/4
.;- . /
to bring all leads out the same side of the /
core so the covers will go in place when
<I
'"
Q:
<I
2 -112
.
the unit is.completed. When the last layer
of the winding is put on, use two sheets of '"
Q: 2-114 /'
waxed paper, and then cover those with
vinyl electrical tape, .keeping the tape as
.0
<.l
Z
0
2
/
/
taut as possible. This will add mechanical
strength to the assembly.
i=
<.l
'"
II)
1-3/4
1-1/2
/ .
The laminations and housing are
I.
II)
II)
0
assembled in just the opposite sequence to Q: 1-1/4
<.l
that followed in disassembly. Use a light
coating of shellac between each lamina- 1
50 100 150 200 250 300 350 400 450 500
tion. During reassembly, the lamination POWER RATING OF TRANSFORMER IN WATTS
stack may be compressed by clamping in a
vise. If the last few lamination strips can- Fig. 18- Power-handlinill capability of a transformer versus cross-sectional area of core.
not be replaced, it is better to omit them
than to force the unit together.
/
AND DIODE CURRENT
low operating temperature and high cur- (Al
D DIODE
rent-handling capability. Also, no fila- IPEAK CONDUCTION
ment transformers are required. FILTERED
:k-M
Silicon rectifi'ers are available' in a wide OUTPUT
. VOLTAGE
range of voltage and current ratings. In
peak reverse voltage ratings of 600 or less,
silicon rectifiers carry current ratings as
__
---... HALF -WAVE
--LU~L-_ ·-'-"'---'-.R"'E""C'OCTI""'F'IED SIGNAL
high as 400 amperes, and at 1000 PRV the
current ratings may be several amperes or (8l
so. The extreme compactness of silicon D
types makes feasible the stacking of
several units in series for higher voltages.
Standard stacks are available that will
handle up to 10,000 PRY at a dc load
current of 500 rnA, although the amateur
can do much better, economically, by
HALF- WAVE
stacking the rectifiers hims~lf. . (el RECTIFIED SIGNAL
Surge Protection
Each time the power supply is acti-
TO vated, assuming the input filter capacitor
RECTIFIER
has been discharged, the rectifiers must
look into what represents a dead short.
Some form of surge protection is usually
necessary to protect the diodes until the
(A) input capacitor becomes nearly charged.
Although the dc resistance of the trans-
former secondary can be relied upon in
some instances to provide ample surge-
current limiting, it is seldom enough on
high-voltage power supplies to be suit-
Rs able. Series resistors can be installed
between the secondary and the rectifier
strings, but are a deterrent to good voltage
THY RECTOR regulation. By installing a surge-limiting
117VAC DIODE TO
RECTIFIER d<;vice in the primary circuit of the plate
transformer, the need for series resistors
in the secondary circuit can be avoided. A
(8) practical method for primary-circuit surge
control is- shown in Fig. 20. The resistor,
Fig. 21 - Methods of suppressing line tran·sients. See text.
Rs introduces a voltage dro~ in the
primary feed to Tl until C is nearly
charged. Then, after C becomes partially
the shorter the diode conduction time and diode's capability can be made by using charged, the voltage drop across Rs les-
the higher the peak repetitive current(lREP). these rules of thumb for silicon diodes of sens and allows K 1 to pull in, thus apply-
When the supply is first turned on, the the type commonly used in amateur ing full primary power to Tl as KIA
discharged input capacitor looks like a power supplies: shorts out Rs. Rs is usually a 25-watt resis-
dead short; and the rectifier passes a very Rule 1) The maximum IREP rating can tor whose resistance is somewhere be-
heavy current. This is ISURGE. The maxi- be assumed to be approximately four tween 15 and 50 ohms, depending upon
mum ISURGE rating is usually for a dura- times the maximum 10 rating. the power supply characteri~tics.
tion of one cycle (at 60 Hz), or about 16.7 . Rule 2) The maximum IsURGE rating
milliseconds. can be !lssumed to be approximately 12 Transknt Problems
If a manufacturer's data sheet is. not times the maximum 10 rating. (This A common cause of trouble is transient
available, an educated guess about a should provide a reasonable safety factor. voltages on the ac power line. These are
5-12 Chapter 5
/ short spikes, mostly, that can temporarily should be placed across each diode. Fig. R R
increase the voltage seen by the rectifier to 22A shows the complete series-diode
values much higher than the normal circuit. The capacitors should be non-
transformer voltage. They come from inductive, ceramic disk, for example, and
distant lightning strokes, electric motors should be well matched. Use lO-percent-
turning on and off, and so on: Transients tolerance capacitors if possible.
eause unexpected, and often unexplained,
loss of silicon rectifiers.
It's always wise to suppress line
Diodes In Paralkl
Diodes can be placed in parallel to
GIa (A)
the PR V drops and to guard against Example: The plate of the tube in one Rl
transient voltage spikes, as shown in Fig. stage and the screens of the tubes in two +E.
22A. Even though the diodes are of the other stages require an operating voltage FROM
POWER R2
same type and have the same P~ V rating, of 250. The nearest available supply volt- SUPPLY
they may have widely .different back age· is 400 and the total of the rated plate + E2
resistances when they are cut off. The and screen currents. is 75 mAo The . R3
reverse voltage divi~es according to required resistance is
Ohm's Law, and the diode with the higher
back resistance will have the higher 400 - 250 150 '
R = .075 = .075 = 2000 ohms (e)
voltage developed' across it. The diode
may break down.
If, we put a swamping resistor across The power rating of the resistor is Fig. 23 - A - Series voltage-dropping resistor.
each diode, R as shown in Fig. 22A, the obtained from P (watts) = PR = (0.075)2 8 - Simple voltage divider.
resultant resistance across each diode will X (2000) = .11.2 watts. A. 20-watt resistor is
the nearest safe rating to be used. E1 E -E1
be almost the same, and the back voltage R2 = - R1 = - •
will divicle almost equally. A good rule of 12 11 +12
thumb for resistor size is this: Multiply the Voltage Dividers 12 must be assumed.
PR V rating of the diode by 500 ohms. For The regulation of the voltage obtained
example, a 5OO-PRV diode should be in this manner obviously is poor, since C- Multiple divider circuit.
shunted by 500 X 500, or 250,000 ohms. any change in current through the resistor
E1 -E2
The shift from forward conduction to will cause a directly proportional change R3 =£ R2 = "'i1"'+'i3
13
high back resistance does not take place in the voltage drop across the resistor. The
instantly in a silicon di9de. Some diodes regulation can be improved somewhat by
take longer than others to develop high E -E1
connecting a second resistor from. the R1
back resistance. To protect the "fast" 11 +12+13
low-voltage eodof the first to the negative
diodes in a series string until all the diodes power-supply terminal, as shown in Fig.
are properly cut off, a 0.01- uF capacitor 23B. Such an arrangement constitutes a . 13 must be assumed.
5-14 Chapter 5
02 25.3,,' , PO
t3V
Ot (REG)
+
117
VAC
ct RL
03 vo 26
OHM5
04
51 ON I
vsec (rllis F- 1.4 V 0 C2 (uF) == 0.5 Cl (uF) Vl =Vsec(rms) X 1.41
Cl (uF),- See section on capacitive filters Rp =V o X80 Po =V o XL
Cl (V) == 2V' VRI =V o +0.7 RL =Vo+IL
C2 (V min) > Vz Vo=Vo-Vz-0.7 Fl =I L X2
Fig. 26- Illustration of a power supply with regulation. A pass transistor. Q1. is used to extend the range of the Zener-diode regulator.
operated in a higher ambient temperature. single Zener diode. and also to give two (electronically). The average dc voltage
its power capability must be derated. A values of regulated voltage. The diodes available· from the regulator is propor-
typical I-watt diode can safely dissipate need not have equal breakdown voltages. tional to the duty cycle of the switching
only 1/2 watt at lOO°C .. because the arrangement is self equalizing. wave form, or the ratio of the on time to
However. the current-handling capability the total period of the switching cycle.
Limiting Resistance of each diode should be taken into ac- Switching frequencies of several kilohertz
The value of Rs in Fig. 25 is determined count. The limiting resistor may be calcu- are normally used to avoid the need for
by the load requirements. If Rs is too large lated as above. taking the sum of the extensive filtering to smooth the switching
the diode will be unable to regulate at diode voltages as Ez. and the sum of the frequency from the dc output.
large values of I L, the current through R L. load currents as IL. The above information pertains essen-
If Rs is too small. the diode dissipation tially to voltage regulators. A circuit can
rating may be exceeded at low values of Electronic Voltage Regulation ,also be constructed to provide current
IL. The optimum value for Rs can be cal- regulation. Such regulation is usually
Several circuits have been deveioped for
culated by: regulating the voltage output of a power obtained, in the form of current limitation
supply electr,onically. While more compli- - to a maximum value which is either .
cated than the VR-tube and Zener-diode preset or adjustable. depending on the
circuits. they will handle higher voltage circuit. Relatively simple circuits. such as
and current variations'. and the output described later. can be used to provide
When Rs is known. the maximum dis- voltage may be varied continuously over current limiting only. Current limiting
sipation of the diode. Po. may be deter- a wide range. . circuitry may also be used in conjunction
mined by Voltage regulators fall into two basic with voltage regulators.
types. In the type most commonly used by
EOC(max) - Ez ] amateurs, the dc supply delivers a voltage Discrete Component Regulators
Po = [ Rs . -IL(minl Ez
higher than that which is available at the The previous section outlines some of
output of the regulator, and the regulated the· limitations when using Zener diodes as
In the first equation, conditions are set voltage is obtained by dropping the volt- regulators. Greater current amounts can
up for the Zener diode to draw l/lO the age down to a lower value through a be accommodated if the Zener diode is
maximum load current. This assures dropping "resistor." Regulation is accom- used as a reference at low current.
diode regulation under maximum load. plished by varying' either the current permitting the bulk of the load c.urrent to
I Example: A 12-volt source is to supply through a fixed dropping resistance as flow through a series pass transistor(Qf
a circuit requiring 9 volts. The load cur- changes in input voltage or load currents of Fig. 26). An added benefit in using a
rent varies between 200 and 350 rnA. occur (as in the VR-tube and Zeller-diode pass transistor is that of reduced ripple onl
regulator circuits). or by varying the'equi- the output waveform. This technique is
Ez =9.1 V (nearest available value)
valent resistive value of the dropping ele- commonly referred to as "electronic
ment with such changes. This latter tech- filtering."
Rs = 12-9.1 2.9 75 h
1.1 XO.35 = 0.385 = . 0 ms nique. is used in electronic regulators Q1 of Fig. 26 can be thought of as a
where the voltag~-dropping element is a simple emitter-follower de amplifier. It
Po = r12 7.5-9.1_0.;1~
L
9.1 vacuum.tube or a transistor. rather than increases the load resistance seen by the
an actual resistor. By varying the dc volt- Zener diode by a factor of beta ( ~ ). In this
=0.185 X9.1 = 1.7 W age at the grid or current at the base of circuit arrangement D5 is required to
these elements. the conductivity of the supply only the base current for Q1. The
The nearest available dissipation rating device may be varied as necessary to hold net result is that the load regulation and
above 1.7 W is 5; therefore. a 9.1-V 5-W the output voltage constant. In solid-state ripple characterjstics are improved by a
Zener diode should be used. Such a rating, regulators the series-dropping element is factor of beta. Addition of C2 reduces the
it may be noted. will cause the diode to be called a pass transistor. Power transistors ripple even more. although many simple
in the safe dissipation range even though are available which will handle several supplies such as this do not make use of a
the load is completely disconnected amperes of current at several hundred capacitor in that part of the circuit.
[IL (min) = 0]. volts, but solid-state regulators of this The primary limitation of this circuit is
type are usually operated at potentials that Q 1 can be destroyed almost im-,
.' Obtaining Other Voltages below 100 volts. mediately if a severe overload occurs at
Fig. 25B shows how two Zener diodes The second type of regulator is a RL. The fuse cannot blow fast enough to
may be used in series to obtain regulated switching type. where the voltage from the protect Q1. In order to protect Q1 in case
voltages not normally obtainable from a dc source is rapidly switched on and off of an accidental short at the output. a
not be repeated here. In order to calculate Q 1 that has at least twice the rating
150
the value or Rs in Fig. 26 the base current calculated. In' this example a transistor
of Ql must be known. The base c:urrent is with a power dissipation rating 12 watts
approximately equal to the emitter cur- or more would be used.
rent of Ql in amperes divided by beta. The 0.01- uF capacitors at the primary 10K
The transistor beta can be found in the of T 1 serve two functions. They act as
manufacturer's data sheet, or measured transient suppressors and help prevent rf
with simple test equipment (beta = Ic/Ib). energy from entering the power-supply
. Since the beta spread for a particular type regulator. B
of transistor - 2N3055 for example, (Al
. where it is specified as 25 to 70 - is.a fair-
Current Limiting for Discrete-Component
ly unknown quantity, more precise calcu- Regulators
lations for Fig. 26 will result if the transis-Damage to Q 1 of Fig. 26 can occur
tor beta is tested before the calculations When the load current exceeds the safe 03
cn-_-, )r--..J\,iV\~-<l E
are done. A conservative approach is to amount.' Fig. 27 illustrates a simple
design for beta minimum of the transistor current-limiter circuit that will protect Q 1.
used. Calculating Ib: All of the load current is routed through
R2. A voltage difference will exist across
fb ::;::¥s= 0.02A = 20 rnA R2, the amount b~ing dependent upon the
- exact load current at a given time. When
As pointed out earlier, in order for 05 the load current exceeds a predetermined
to regulate properly it is necessary that a safe value, the voltage drop across R2 will
B
fair portion of the current flowing forward bias Q2 and cause it to conduct.
through Rs should be drawn by D5. The Since D6 is a silicon diode, and because (Bl
resistor will have 0.02 A flowing through Q2 is a silicon transistor, the combined
it as calculated above (base current of voltage drops through them (roughly 0.7
Q 1). A conservative amount of 10 mA will Veach) will be 1.4 V. Therefore the 05
be used for the Zener diode current voltage drop across R2 ~ust exceed 1.4 V cn-_-, ~----~~~OE
bringing the total current through Rs to beforeQ2 can turn on. This being the
0.03 A or 30 mAo From this, the value of case, R2 is chosen for a value that
Rs can be calculated as follows: provides a drop of 1.4 V when the
maximum safe load current is drawn. In
R - (V'- V z) _ (25.3 -14) _ 376 h this instance 1.4 volts will be seen when IL
s - IRs - 0.03 - 0 rns reaches 0.5A.
When Q2 turns on, some of the current
The nearest standard ohmic value for Rs through Rs flows through Q2, thereby de-
is 390. The wattage ratings for Rs and priving Q I of some of its base current.
D5 can be obtained with the aid of the This action, depending upon the amount
formulas given earlier for Zener-diode of Q 1 base current at a precise moment,
regulators. . cuts off Q 1 conduction to some degree,
The power rating for Q 1 will be calcu- thus limiting the flow of current through (Cl
lated next. The power dissipation of Q 1 is it.
equal to the emitter current times the Fig. 28 - At A, a Darlington-connected pair for
collector-to-emitter voltage. CalCUlate as High-Current-Output Regulators use as the pass element in a series-regulating
follows: . circuit. At Band C, the method of connecting two
When a single pass transistor is not or more transistors in parallel for high current
available to handle the current which may output. Resistances are in ohms. The circuit at A
PQ1 =I E XVCE
be required from a regulator, the current- may be used for load currents from 100 mA to 5
where handling capability may be increased by A, at B for currents from 6 to 10 A, and at C for
VCE = the desired V' - (V z - VBE), connecting two or more pass transistors in cu rrents from "9 to 15 A.
01 - Motorola MJE 340 or equivalent.
and VBE is approximately 0.7 V for a parallel. The circuits at Band C of Fig. 28 02-07, incl. - Power transistor such as 2N3055
silicon transistor. show the method of connection. The or 2N3772.
5-16 Chapter 5
resistances in the emitter leads of each where they protrude through the heat
transistor are necessary to equalize the. sink. Tantalum capacitors are usually
currents. recommended because of their excellent
bypass capabilities up into the vhf range.
Fixed-Voltage IC Regulators
The modern trend in regulators is Adjustable-Voltage IC Regulators (A)
toward the use of three-terminal devices Relatively new on the electronic scene
commonly referred to as three-terminal are high-current, adjustable voltage regu-
regulators. Inside each regulator is a hitors. These ICs require little more than
reference, a high-gain error amplifier, an external potentiometer for an ad-
sensing resistors and transistors, and a justable voltage range from 5 to 24 volts at
pass element. Some of the more sophis- up to 5 amperes. The unit price on these
ticated units have thermal shutdown, items is currently around $6 making them
over-voltage protection and current fold- ideal for a test bench power supply. An
back. Many of the regulators currently adjustable~voltage power supply using the
on the market are virtually destruction- Fairchild 78HG series of regulator is (B)
proof. Several supplies using these ICs are described in the construction section of
featured in the construction section of this this chapter. The same precautions should
chapter. be taken with these types of regulators as
Three-terminal regulators (a connection with the fixed-voltage units. Proper heat
for unregulated dc input, regulated dc sinking and lead bypassing is essential for r-----~h_----~~-----o+
output and ground) are available in a wide proper circuit operation.
range of voltage and .current ratings.
Fairchild, National and Motorola are A 12-Volt 3-Ampere Power Supply
perhaps the three largest suppliers of these Shown in Fig. 31 is a no-frills 12-volt
regulators at present. It is easy to see why supply capable of continuous operation at (e)
regulators of this sort are so popular when the 3-ampere level. Many low-power hf
one considers the low price and the transceivers and most vhf-fm ·transceivers
number of individual components they require voltages and currents on this
can replace. The regulators are available . order. Power supplies of this type
in several different package styles -:- purchased from the manufacturers can be
TO-3, TO-39, TO-66, TO-220 and dual quite costly. Described here is a very
in-line (DIP), to name just a few. simple to build and relatively inexpensive
Three-terminal regulators are available (around $20 using all new components) (0)
as positive or negative types. In most alternative.
cases, a positive regulator is used to The schematic diagram for the power
regulate a positive voltage and a negative supply is shown in Fig. 32. As can be seen, Fig. 29 - A and B illustrate thlf conventional
regulator a negative voltage. However, the circuit is simplicity itself. A trans- manner in which three-terminal regulators are
depending on the systems ground require- former, two diodes, three capacitors and a used. C and D show how one polarity regulator
can be used to regulate the opposite polarity
ments, each regulator type may be used regulator form the heart of the supply. voltage.
to regulate the "opposite" voltage. Binding posts, a pilot light, fuse and
Figs. 29A and B illustrate how the regu- on-off switch complete the design.
lators are used in the conventional mode. Ac from the mains is supplied to the
Several regulators can be used with a transformer-primary winding through the
common-input supply to deliver several fuse. in one leg, and the on-off switch in
voltages with a common ground. Negative the other. The secondary circuit feeds a
regulators may be used in the same full-wave rectifier circuit which is filtered
manner. If no other common supplies by Cl. This unregulated voltage is routed
operate off the input supply to the to the input terminal of the regulator IC
regulator, the circuits of Figs. 29C and D which is bypassed directly at the caSe with
may be used to regulate positive voltages .a 2-j.l.F tantalum capacitor. The case of the Fig. 30 - Three-terminal regulators require
careful bypassing directly at the case. Here, both
with a negative regulator and vice versa. IC is connected to ground. A 2-.uF the input and output leads are bypassed.
In these configurations the input supply is tantalum capacitor is also used at the
floated; neither side of the input is tied to output terminal of the regulator to
the system ground. prevent unwanted oscillat~on of the error
When choosing a three-terminal regu- amplifier inside the IC. A pilot light
lator for a given application the important' attached to the regulated output indicates
specifications to look for are maximum when the supply is in use.
output current, maximum output voltage, The regulator has built-in thermal shut
minimum and maximum input voltage, down and over-current protection. Short
line regulation, load regulation and power circuiting the output of the supply will
dissipation. cause no damage. A wide margin of
In use, these regulators require an conservative component rating was used
adequate heat sink since they may be in the design of this supply. It should be
called on to dissipate a fair amount of possible to run the supply for hours on
power. Also, since the chip contains a end at its maximum rating.
high-gain error amplifier, bypassing of the
input and output leads is essential to Construction
stable operation (See Fig. 30). Most Rather than using an expensive cabinet,
manufacturers recommend bypassing the the power supply is housed on an alu- Fig. 31 - Exterior view of the 12-volt, 3-ampere,
input and output directly at the leads minum chassis measuring 5 X 9-1/2 X 3 no-frills power supply .
5-18 Chapter 5
limits the maximum regulator· output cur- ing that the fins would be more efficientif transformers of different manufacture are
rent to about 300 rnA. This additional mounted vertically. A more conservative used it might be wise to select units having
prot~tion is not shown in the IC approach might be to mount an additional a higher current rating - just to be sure.
manufacturer's applications literature. heat sink in the rear panel area and use 9 S2 is included in the design so that
R3, R4 and R9 set the output voltage over or 12 pass transistors.' The emitter either half or all of the secondary voltage
an adjustment range of 12 to 14 V. The spreading resistors are mounted on the may be applied to V 1. This feature was in-
power supply includes a "crowbar" over- transistor heat sinks. Q4 should be cluded so that the dissipation of the pass
voltage protection citcuit consisting of separately heat sinked, although the sink transistor may be reduced when using the
Q5, Dl and the associated resistor net- shown is larger than necessary. A "top supply with low:voltage, high-current
work. RIO forms a divider with the hat" and chassis bracket similar to that of loads: The graph displayed in Fig. 41 can
cathode-to-gate r~sistor internal to most Q5 would be sufficient. be used as a guide in selecting the HI or
SCR~. If your device does not include The rectifier diode anodes are common LO mode of operation.
such a resistor, one can be installed exter- to the mounting studs. For this reason, The regulator consists of a pass tran-
nally. Should the output voltage exceed two of the studs must be sleeved and the sistor "wrapped around" an adjustable
the threshold of D 1, plus some level deter- cases and nuts must be insulated from the voltage regulator. Circuit operation can
mined by RIO, Q5 will fire, short- base plate with mica washers. If a suitably be understood by noting the values of R3
circuiting the output terminals. Q5 has a rated bridge assembly can be obtained at a and R2. The majority of the three-.
continuous current rating of 25 A. While reasonable price, it will simplify the terminal regulator current will flow
this may seeI1,1 underspecified, the compo- mounting while providing superior ther- through R3 and D1. The offset voltage in
nent can withstand 'a surge of many times mal conduction. In any case, be sure to DI is approximately equal to the emitter-
the continuous figure, and won't be called use mica washers with all semiconductors base potenti~l of Ql. Because of this, the
upon to pass the full current for more whose cases are above ground and apply voltage drop across R3 will be the same as
than a few milliseconds. As soon as Q5 heat sink compound to all mounting sur- that across R2. Since the ohmic value of
fires, it pulls the output voltage below the faces. R2 is 0.33 of R3; three times as much cur-
conduction threshold of Q3, which allows Liberal use of terminal strips simplifies rent will flow in QI 'as in V2. The net
the. input voltage to shut down the the wiring and troubleshooting, should it result is that the current capability of the
regulator. Without this feature, both the be necessary. Vse no. 10 or no. 12 wire for overall circuit is increased by a factor of
series pass transistors and the SCR would the high-current circuits. The electronic four. Also, the current-limiting
be destroyed in a short time from ex- components that don't generate ap- characteristics of the three-terminal
cessive dissipation. preciable heat are contained on a printed regulator are transferred directly to the
This supply has one otl\er feature - circuit card. The etching pattern and parts composite circuit.
remote sensing. When heavy loads are placement guide are presented in Fig. 37. MI and its associated shunt resistor are
connected through long cables, the cable Most of the components for this power placed at the input to the regulator circuit
voltage drop degrades the load regulation. supply were purchased from mail-order so that the voltage drop across the resistor
This problem can be circumvented by in- firms that advertise in the Amateur Radio will not adversely alter the supplY'voltage
cluding the cable within the regulator press. regulation. The relatively small current
feedback loop. To accomplish this, drawn by the regulator circuitry does not
remove the jumper. between the output A 1.2- to IS-Volt, S-Ampere Supply seriously affect the· meter accuracy. M2.
and sense terminals, and connect the sense The power supply shown in Figs. 38 to measures supply output voltage.
terminal directly to the load (the sense 42 is intended for general purpose, test-
lead wire size isn't critical).· This connec- bench applications. The output ·is ad- Construction
tion can't compensate for resistance in the justable from 1.2 to 17 volts at currents up The power supply is housed in a
return (ground) lead, but an extra chassis- to 6 amperes. Metering is provided for homemade enclosure that was fabricated
to-chassis connection can usually be made voltage levels up to 15 volts and current from sheet aluminum. Dimensions of the
to reduce the return resistance to an in- levels up to 5 amperes. Most of the com- enclosure are 5-1/2 X 6 X 8 inches (140
significant value. The pk-pk output ripple ponents used in this supply are of the X 152 X 203 mm), although any cabinet
at full load is 5 mV. A complete schematic junk-box variety with the possible excep- that will house the components may be
diagram is given in Fig. 35. tion of V2 j the three-terminal voltage. used. Circuit board pattern and layout in-
regulator. The circuit will tolerate fairly formation is given in Fig. 42. The com-
Mechanical Details wide component substitutions and still of- pleted circuit board is mounted vertically
One assembly method can be seen in fer good performance. The majority of to the chassis using spade lugs and no. 6
Figs. 34 and, 36, although this is by. no the Circuit components are mounted on a hardware. A small heat sink for the
means the only workable solution. The 2-3/4 X 4-1/2 (70 X 114 mm) circuit LM317K regulator was;made from a scrap
unit shown uses a 16-1/2 X 12 X 1/8-in. board. All controls, including the mains piece of aluminum.. A Radio Shack
(419 X 305 X 3 mm) aluminum plate for fuse are located on the front panel for 276-1364 sink designed for the TO-3
a foundatio'n. This is expensive material if easy access. package will work fine. The pass tran-
purchased new - a framework fashioned sistor is mounted to a larger heat sink
from angle stock may be an expedient The Circuit which is bolted to the rear panel of the
alternative. Transformers suitable for this Two power transformers are used in power supply. Here, a Motorola MS-IO
application are quite heavy and require a parallel to feed VI, the full-wave bridge was used. A suitable substitute would be
firm supporting structure. The pass tran- rectifier assembly. The transformers the Radio Shack 276-1361. Bear in mind
sistor heat sinks form the sides of the specified are rated at 2 amperes each. The that the transistor must be insulated from
enclosure. These sinks, like most of the prospective builder might question the the sink. Vse a small amount of heat-sink
components, are surplus items measuring, wisdom of using only 4 amperes worth of compound between the transistor and the
4 X 12 inches,(l02 X 305 mm) with fins transformer in a 5-ampere supply. This is sink for a good thermal bond.
protruding 112 in. from each flat surface. a valid- question. With a 5-ampere load Since the power supply can deliver up to
Asymmetrically spaced transistor connected to the output of the supply, the 6 amperes, fairly heavy wire should be
mounring holes explain the surplus status transformers deliver more than their rated used for those runs carrying the bulk of
of the sinks. .The dimensions quoted secondary voltage and do not become the .current .. No. 18 plastic wire was used
represent a'minimum, rather than maxi~ lInreasonably warm to the touch even in this unit and it ~peared to be ade-
mum radiation area, especially consider- after continuous-duty operation. If quate. .
R12 R19
0.1
2N~05!5
I
5W
I
I
I
I I
I
8A r-------..., Tf
1 51
ff7VAC
4
C2
SA
O.OOI~
MC1469
Ul
R5 6.8k
Rl
4.70
2N2222
\ . 2N3904
2N4401
5-20 Chapter 5
this unit as compared to a ready-made
. supply with the same features. Cost, 'using
all new componen~, will be on the order
of$75. The most expensive single item in
the supply is the digital panel meter,
which sells in single 10t quantities for
around $40 at present. As more com-
panies start manufacturing these items the
prices-- should drop significantly.
The digital readout, however, is not
much more expensive than two high-
quality meters. The prospective builder
should consider' this when choosing
between the digital panel meters and two
analog panel meters.' Voltage mellsure-
ments are read directly off the panel meter
in volts. Current is measured .in amperes
with a reading of 0.05 equal to 50 rnA.
(A)
2N3904
2N2222
_ _ _----' 2N4401
(8)
Fig. 37 - At A, the etching pattern for the control pc card. 8 is the parts-replacement guide. The
black lines are an x-ray view of the foil side.
5-21
\
Circuit Details remain unchanged. Notice that U2 is a
reason board pattern is not supplied.
placed to the lert or at the input side of The front and rear panels are spray
The circuit diagram of the power supply the regulator. This is so the current painted white and the cover is blue. Dymo
is shown in Fig. 44. T1 is a 36-volt, center- drawn by the readout will not affect labels are used on the front panel to
tapped transformer rated at 6 amperes. current readings taken at the load. identify each of the controls. Cable lacing
DI and D2 are used ina full-wave rectifier Sections A and B of S2 are used to switch of the various leads adds to the clean
providing dc output to the filter capacitor, the meter between the voltage and current appearance of the supply.
C3, a 34,OOO-IlF, 50-volt electrolytic of the sensors. S3C is used to switch the decimal
computer-grade variety. The unregulated point in the digital panel meter to read A "Sanitary" High-Voltage Supply
voltage is fed to Ul, a Fairchild 78HGKC correctly for both voltage and current. Power supplies aren't usually- noted for
regulator, the heart of the supply. This' As shown in the schematic, a single- a~ding decor to the shack. Most hams
chip is rated for 5-A continuous duty poin.t ground is used for the supply. Used wbuld rather hide them so that nonham
when used with an adequate heat sink. RI in many commercial supplies, this tech- visitors won't ask, "What's that ugly
and R2 form a voltage divider'which sets nique provides better voltage regulation looking thing?" However, an attempt was
the output voltage of the supply. RI is a and stabilization than the "ground it made to improve the appearance of this
ten-turn potentiometer. UI is bypassed anywhere" attitude. In this supply, the I model along with the function of pro-
with 2.2-IlF tantalum capacitors directly at single-ground point is at the front panel viding high voltage for general amplifier
the input and output pins. binding post labelled MINUS. All leads purposes. Not all the additions are
ZI, as outlined earlier, is a digital panel that are to be connected to ground should frivolous. For instance, the use of "rug
meter. CQnn~~H.ons to the meter are made go only to that point. runners" instead 'of the usual sharp
. through a s~;i}edge connector supplied corners on the bottom of the unit prevents
with the read6ut. U2 is used to supply a Construction gouging an easily damaged surface such as
regulated 5 volts for powering the digital a bench or floor.
panel meter. The input and ground leads The deluxe power supply is housed in a The diode bridge rectifier is mounted
of this regulator are attached to the input homemade enclosure that measures 9 X II on a separate pc board that can be
(non-regulated) side of UI. X 5 1/4 inches (229 X 279 X 133 mm). UI removed easily. Accidental contact is
. R4, R5 and R6 form a divider circuit to is mounted to a large heat sink (3 X 5 prevented by a Plexiglas sheet which also
supply the digital meter with an output X 2 inches; 76 X 127 X 51 mm) which is permits viewing of the circuit board while
voltage reading. R5 is made adjustable so attached to the rear apron of the supply. it is still in the power supply. Although a
that the meter can be calibrated. R3 is a The front panel sports the digital-panel sheet-metal cutter and bender were used
current-sensing resistor which is placed in meter, power switch, binding posts, fuse to fabricate the sides, a "cut-and-fiIe"
the rlegative lead of the supply. This holder, voltage-adjust potentiometer and method. could result in a similar job if the _
resistor is used on the input side of the meter-selector switch. Although a circuit builder was willing to spend the time.
regulator (U I) so as not to affect the board is shown in the photograph as Either that or angle brackets (such as
voltage regulation of the power supply at supporting R4, R5, R6, R7, DI, D2, CI those on the front of the unit) could be
high load currents. Any.voltage oropped and C2 these items could just as well be used inside of the top and bottom covers
across the resistor will be made up by the mounted on terminal strips. For this in order to form an overlap surface for the
regulator, so the output voltage will covers.
",1
Ul
S1
lA
~~----_T----_,
r----------, I
I
I 0-5
I
I Rl R2 Ql It
I
o~
+ 5W
R4 .1Q!!:...
Cl 1 3 3 0 0 P F
2W
35 V '
R3 Dl
0.33.n. 2.5A
S2 --sw- i'O'O'V
* HEAT SINK
Fig. 39 - Schematic diagram of the 5·ampere power supply. Component designations on the schematic diagram but not shown in the'parts list are
for text or Illacement-auide reference only. ' ( R e s i s t o r used only as a form for the wire).
C1 - 3300 "F, 35 volt,axial leads. Q1 - Silicon PNP power, Radio Shack R6 - 2500 ohms, 2 watts, panel mount.
J1, J2 - Binding posts. 276·2043 or equiv. S1 - Spst, toggle.
M1 - 0·50 mA, Calectro 01·914 or equiv. R1 - Meter shunt, 13 inches no. 22 enameled S2 - Opdt, toggle. Both sections connected
M2 - 0·15 volt, Calectro 01·920 or equiv. wire wound on a high·value, 1·watt resistor. in parallel.
5-22 Chapter 5
5
~ 1'---
4
3
"" ~ HI
Ul
IIJ
II:
IIJ
~ f'."
Q. 2
:::t
« LOW "" ~
~
~
Fig. 40 - Inside view of the power supply. 2 3 4 5 6 7 8 9 10 11
Component placement is not. at all critical,
however the layout shown here provides a neat
appearance. VOLTS
Fig. 41 - For voltage and current requirements that fall to the left of the diagonal line, the power'
supply may be operated in the LO mode. Pass-transistor dissipation will be reduced when the
supply is operated in this manner. .
(A) -
JI
0-15
U2 * M2
J2
2500.n.
~
(B)
C1
51
+ JI
3
T1r---~~-4-----1--~----~--~~~~
I
I 01
H7V
02 0-25V
5A
ON
+
C3
0.01
~
33000)lF
50VDC
H7 VAC
_J2
r+;<
IN MICROFARADS I jJF) ; OTHERS
ARE IN PICOFARA OS I pF OR jJjJF I;
RESISTANCES ARE IN OHMS;
k .tpOO. M·I,OOO.OOO
AMPS
S2
0 4
E 5
F 6
270
H 7
J B
Fig. 44 - Schematic diagr1lm of the deluxe power supply. All resistors are half-watt carbon types unless noted otherwise. CapaCitors are disc
ceramiC unless noted otherwise. Numbered components not appearing in the parts list are for text reference only.
Jl·J3, incl. - Binding post. '
C3 - Electrolytic capacitor, 34,000 I'F, 50 V. Rl - Potentiometer, 20·kO, linear, 10 turn. secondary 36 V ct, 6 A. Stancor P·8674 or
Sprague 36D343G050DF2A or equiv. Clarostat type 731A or equiv. equiv.
C4, C5 - Tantalum capaCitor, 2.2 I'F, 50 V. R2 - ReSistor, 0.10, circuit board mount. Ul - Regulator, Fairchild 78HGKC or equiv.
C6 - Tantalum capacitor, 4.7 I'F, 50 V. Sl - Toggle switch, dpst. U2 - Regulator, I' A7805 or equiv.
Dl, D2 - Silicon rectifier, 100 V, 12 A S2 - Toggle SWitch, 3pdt. 21 - Digital panel meter. Dalel DM3100N or
Fl- Fuse, 2 A. T1 - t-'ower transformer; primary 117 V, equiv.
5·24 Chapter 5
... Fig. 46 - Front view of the heavy-duty.
3400-volt power supply.
CIRCUIT Tl
BREAKER HAMMOND
234 V
I~f
AMF
W6BX2QI-2-10 1000
10 AMPS
Ml
2~OV
10M
lOOk
20W
!2Q!!.
20W
3400V
lAMP
\ lOOk
20W
lOOk
20W
H F Transmitting
6·1 Chapter 6
supply becomes rather expensive because \
OSC. HALF - WAVE
of complex regulator-circuit requirements.
For this reason it is the choice of most
amateurs to utilize vacuum tubes in high-
oH
3.5 MHz
1--+.-----.
0.01
CLASS C
AMP
HARMONIC
FILTER
FL1 ~ANT.
power hf and vhf amplifiers. The number
of power transistors required (plus com-
biners) to generate a l-kW signal may run
considerably higher in cost than a tube or
tubes for an amplifier of equivalent power.
The price of large heat sinks versus a cool-
ing fan may place the solid-state amplifier RFC
1mH
in a prohibitive class also.
The decision . to buy or build a
transmitter is founded on some basic +12 V
considerations: Cost compared to features;'
professional equipment appearance con-
trasted to that of homemade apparatus;
5600 --VnKEY
the knowledge and satisfaction gained
from building equipment, as weighed +12 vo--lJ
against buying store-bought gear and sim-
ply becoming an operator. The judgment Fig. 2- Circuit example of a simple, solid-state cw transmitter.
must fit the amateur's objectives and af-
fluence. Home-built transmitters are usual-
ly easier to service than commercial ones ANT.
because the builder knows the circuit lay-
out and how each stage functions. Further-
more, the cost of maintenance is markedly
lower' for homemade equipment than for
most factory-built gear. But the greatest
significance to home-built circuits .is the
knowledge gained from constructing a
project and the pride that goes with using
ANT.
•
it on the air!
Frequency Generation
The most basic type of transmitter is
one which contains a single stage, is
crystal controlled, and i~ designed for cw
ANT.
operation. A circuit example is given in
Fig. 1. This kind of transmitter is not
especially suitable for use on the air
because it is somewhat inefficient and is
, prone to generating a chirpy cw signal
unless loaded rather lightly. But, the same
circuit is entirely acceptable when fol-
lowed by an isolating stage (buffer/ampli- (Cl
fier) as shown in Fig. 2. The second stage
not only builds up the power level, but it ANT.
gives the oscillator a relatively constant
load to look into. The latter helps to pre-
vent oscillator pulling and attendant chirp-
ing of the cw note.
FREQUENCY
Fig. 3 illustrates the basic types of SYNTHESIZER
transmitters for cw and RTTY work. The
drawing at A represents the general circuit
given in Fig. 2. Illustration B is an (Dl
expansion of that circuit and includes a
frequency multiplier. A heterodyne type Fig. 3- Block diagrams of various transmitters which use specific frequency-generation methods.
of generator, which is currently popular
for multiband. transmitters and trans-
ceivers, is shown as the exciter section of a
transmitter in drawing C. A frequency The block diagram of Fig. 4 outlines the work. However, linear amplifiers are
synthesizer is shown as the rf generator at functional stage'S of a ssb transmitter. ZI .entirely satisfactory for any transmission
D. can be a simple VFO, a heterodyne mode at a sacrifice in efficiency. Once the
For operation on a-m, any of the generator (Fig. 3C) .or a frequency ssb signal is generated it can not be passed
lineups given in Fig. 3 are suitable, synthesizer. The essential· difference be- through a frequency multiplier. All post-
provided a molulator is added. It is used tween this type of transmitter and one that fIlter stages must operate straight through.
to mOQulate the operating voltage to the would be used for cw /R TTY is that the rf Class C amplifiers are generally used in fm
PA stage, or in some designs the operating amplifiers must operate in the Class A, transmitters as well as in cw and R TTY
voltage to the PA and the stage im- AB or B mode (linear) rather than the transmitters. Fm operators who are heard
mediately before it. Class C mode which is suitable for cw to say, "I'm running my transceiver into a
H F Transmitting 6·2
1.21.3'1. ETC.
ANT.
s:
OUTPUT
Vi
~CFB
~6.3V
Fig. 4 - Block diagram of a heterodyne type of ssb transmitter.
MODIFIED PIERCE
(A)
:s:
OUTPUT
r--00UTPUT
100 (---<>OUTPUT
tOO
6GK6
+12V
PIERCE
(A) o.O~
COLPITTS
• (B) 22k
21.31. ETC.
2N44t6 +225 V
OUTPUT
S
COlPI TTS
(B)
131.51.
OF FUNO.I Fig. 6 - Two com mon tube types of crystal
-r0 .01
r+7 100
oscillators.
6·3 Chapter 6
A Colpitts style of tube oscillator is series capacitor (Cl) is connected between
..L 'illustrated in Fig. 6 at B. The feedback is the low side of the crystal and ground•
c::J between the grid and cathode by means of The series hookup is used to help offset
a capacitive divider (Cl and C2). The the high input capacitance of the oscil-
plate tank can be tuned to the crystal lator. The input capacitance consists of
frequency or harmonics thereof. In the the series value of feedback capacitors C2
interest of good oscillator stability it is and C3 plus the input capacitance (C n) of
suggested that the supply voltage to the Ql. Conversely, the input capacitance of
circuits of Figs. 5 and 6 be regulated. This the circuit at B in Fig. 8 is quite low be-
is especially significant in the case of cause a triode tube is employed. In this
harmonic or overtone oscillators where kind of circuit the trimmer capacitor is
.' +vcc small amounts of drift are multiplied by used in a parallel manner as shown. The
(Al
the chosen harmonic factor. choice between series and parallel trim-
The usual cause of erratic oscillation, or ming will depend on the active device used
no oscillation at all, is excessive loading and the amount of input capacitance
on the oscillator output (succeeding stage present. This rule applies to tube oscil-
of circuit), insufficient feedback, or a lators as well as those which use
sluggish crystal. Concerning the latter, a transistors.
crystal which is not ground to a uniform
thickness and feathered carefully around Crystal Switching
the edges may be difficult to make Although several crystals for a single
oscillate. Attempts by inexperienced ama- oscillator can' be selected by mechanical
teurs to grind their own crystals may lead means, a switch must be contained in the
to this condition. rf path. This C;in impose severe restrictions
Some crystal oscillators develop un- on the layout1of a piece of equipment. Fur-
wanted vhf self-oscillations (parasitics) thermore, mechanical switches normally
even though the circuits may be function- require that they be operated fr'om the
ing normally otherwise. The result will be front panel of the transmitter or receiver.
(8l a vhf waveform superimposed on the de- That type of format complicates the re-
sired output waveform when the rfvoltage mote operation of such a unit. Also, the
Fig. 7 - Two methods for suppressing vhf and
is viewed by means of an oscilloscope. switch leads can introduce unwanted reac-
uhf parasitic oscillations. R1 at A damps the Parasitics can cause TVI and specific pro- tances in the crystal circuit. A better tech-
parasitics and Z1 at B (ferrite beads) serves that blems elsewhere in the circuit with which nique js illustrated in Fig. 9, where D 1 and
purpose. the oscillator is used. Two simple methods D2 - high-speed silicon switching diodes
for preventing vhf parasitics are shown in - are used to select one of two or more
Fig. 7. The technique at A calls for the in- crystals from some remote point. As'
sertion of a low-value resistance (Rl)in operating voltage is applied to one of the
osc.
the collector lead as close to the transistor diodes by means of S 1, it is forward biased
body as possible. Typical resistance values into "hard" conduction, thereby complet-
:lOo-O......' , , , , , _... , +vcc are 10 to 27 obms. The damping action of ing the circuit between the crystal trimmer
tbe resistor inhibits vhf oscillation. An al- and grol,lnd. Some schemes. actually call
ternative to the use of resistance for for reverse-biasing the unused diode or
swamping vhf oscillation is illustrated at B diodes when they are not activ;ited. This
in Fig. 7. One or two high-mu miniature ensures almost complete cutoff, which
ferrite beads (Ui = 950) are placed near the may not be easy to achieve in the circl,lit
transistor body in the lead to gate 1. The shown because of the existing rf voltage
.beads can be used in the drain lead when a on the anodes of D I and D2.
SERIES C tuned circuit or rf choke are used in that
part of the circuit. However, when the Variable-Frequency Oscillators
(Al
drain is at ac ground, as shown at B, it The theory and general application of
does' not constitute part of the feedback variable-frequency oscillators is treated in
circuit. Ferrite beads can be used in the chapter 8. The circuit principles are the
osc. OUTPUT base or collector lead of the circuit of Fig. same regardless of the VFO application.
~_----iHf-o 7A rather than employing Rl. Similarly, Some additional considerations pertain
Rl can be used at gate 1 of the oscillator. to the use of VFOs in transmitters as com-
It is necessary in some applications of. pared to a VFO contained in a receiver.
crystal oscillators to ensure spot accuracy Generally, heating of the interior of a
of the operating frequency. Various transmitter cabinet is greater than in a re-
reactances are present in most oscillator ceiver. This is because considerably more
circuits, causing the operating frequency power is being dissipated in the former.·
to differ somewhat from that for which Therefore, greater care must be given to
the crystal was manufactured. Addition of oscillator long-term stability. Tempera-
a trimmer capacitor·will permit "rub- ture-compensating capacitors are often
PARALLEL C bering" the crystal to a specified fre- needed in the frequency-determining portion
(8l quency within its range. This procedure is of the oscillator to level off the long-
sometimes referred to as "netting" a term stability factor. Some oscillators are
Fig. 8 - The crystal-oscillator operaiing crystal. designed for use with a temperature-
frequency can be shifted slightly by means of Fig. 8 shows two circuits in which a control oven for the purpose of main-
trimmer capacitors as shown at,A and B. A series
hookup (AI is used with transistors to help
trimmer capacitor might be used to . taining a relatively constant ambient
compensate for the relatively high input compensate for differences in the opera- temperatdre in the oscillator compart-
capacitance of the transistor. . ting frequency of the oscillator. At A the ment - even while the equipment
HF Transmitting 6-4
. being adjusted from the front panel of the
osc. equipment. Variable capacitors with plated
brass plates are preferred over those
which have aluminum plates. The alumi-
num is more subject to physical changes in
the presence of temperature variations
c:::J Y1 than is the case with brass. The VFO tun-
ing-capacitor rotor must be grounded at
both ends as a preventive measure 'against
RFC
instabili.ty. ,Some designers have found
that a 1/8- to 1/4-inch (3.2- to 6.4-mm)
thick piece of aluminum or steel plate
serves as ap excellent base for the VFO
assembly. It ,greatly reduces instability
To.0 1 which can be caused by stress on the main
rh 2200 chassis of the equipment. The VFO mod-
ule can be installed on shock mounts
~"<r--~ + 12 V to enhance stability during mobile opera-
Y2 tion.
6-5 Chapter 6
is generally true of C5, C6 and C7. In the
interest of stability, C5should be the VFO
smallest value that will permit reliable 01
oscillation. Feedback capacitors C6 and
\r---o+12 V
C7 are typically the same value and have
an Xc of roughly 60. Therefore, a suitable
value for a 1.9-MHz VFO would be 1500
pF.
. C8 of Fig. lOA should be the smallest
capacitance value practical with respect to
ample oscillator drive' to the succeeding
stage - generally a buffer or amplifier.
The smaller the value of C8, the less the
chance for oscillator' pulJing during load SERIES -TUNED COLPITTS
changes. D I is a gate-clamping diope for
(A)
controlling the bias of the FET. The
function of this stabilizing diode is treated VFO
in chapter four. Basically, it limits the 01
positive swing of the sine wave. This \r---D+12 v
action restricts the change in Q1 junction
capacitance to minimize harmonic genera-
tion and changes in the amount of Casso-
ciated with L 1.
The reactance of RFC can be on the
order of lO-kO. The drain bypass, C9,
should have a maximum Xc of 10 ohms to
ensure effective bypassing at the operating HARTLEY
frequency. Ideally, an X c of 1 ohm would (8)
be used (0.1 /.L F at 1.5 MHz). D2 is used to
VFO
provide 9.1 volts, regulated, at the drain
of QI. Lower operating voltages aid
stability through reduced rf-current heat-
ing, but at the expense of reduced
oscillator output. 1---"l~+-""'----4....--'W v---v+12V
C3
A Hartley oscillator is shown in Fig.
lOB. This circuit offers good stability
also, and is one of the better circuits to use
when the tank is parallel tuned. The tap
on LI is usually between 10 arid 25 percent
of the total coil turns, tapped above the
grounded end. This ensures adequate
feedback for reliable oscillation. The
higher the FET gm the lower the feedback PARALLEL-TUNED COLPITTS
needed. Only that amount of feedback (C)
which is necessary to provide oscillation
should be used: Excessive feedback will
cause instability and prohibitive rf heating Fig. '10 - Three common types of VFOs for use in receivers and transmitters.
of the components. Most of the rules for
the circuit of Fig. lOA apply to the
oscillator in Fig. lOB.
Parallel tuning of the kind used in Fig. A suitable transistor for QI of Fig. IOC polar transistors have a substantially
lOB and C are suitable for use below, say, is an RCA 40673. The Texas Instruments higher capacitance, which tends to com-
6 MHz, although the circuit at B can be 3N211 is also ideal, as it has an extremely plicate a VFO design for the higher
used successfully into the vhf region: high gm - approximately JO,OOO. Dual- operating frequencies. The uhf small-
However, the Colpitts oscillators of A and gate MOSFETs are suitable for the signal transistors, such as the 2N5179, are
C in Fig. 10 have large amounts of shunt circuits of Fig. lOA and B if biased as best suited to the circuits under discus-
capacitance caused by C6 and C7 of A, 'shown at C. Also, they can be used as sion.
and C5 and C6 of C. The smaller the single-gate FETs by simply connecting
coupling capacitor between L I and the gates I and 2 together. No external bias is Load Isolation for VFOs
gate, the less pronounced this effect is. required if this is done. Gate 2 of Q1 (Fig. Load changes after the oscillator have a
The net result is a relatively small value of IOC) should be bypassed with a low- pronounced effect on the operating
inductance at LI, especially with respect reactance capacitor (C4), as is the rule for frequency. Therefore, it is imperative to
to Fig. IOC, which lowers the tank the drain bypassing of all three examples provide some form of buffering (isolating
impedance and may prevent oscillation given in Fig. 10. stage or stages) between the oscillator and
(high C-to-L ratio). The series-tuned Bipolar transistors are satisfactory for the circuit to which it will be interfaced.
circuit of Fig. lOA solves' the shunt-C use in the three VFOs just discussed. The The net effect of load changes, however
problem nicely by requiring considerably, principal disadvantage attendant to the minor, is a change in reactance which
greater inductance at L I than would be use of bipolars in these circuits is the low causes phase shifts. The latter affects the
acceptable in the circuit of Fig. IOC. The base impedance and higher device input operating frequency to a considerable
circuit at A resembles the popular capacitance. Most FETs exhibit an input degree. Therefore, the more isolating
"series-tuned Clapp" of the early 1950s. C of approximately 5 pF, but many bi- stages which follow the oscillator (up to a
HF Transmitting 8-8
Q of 3. The transformation ratio is on the
EMITTER order of 20:1 (lOOO-ohm drain to 50-ohm
FOLLOWER
02 .
load). RI is placed across J.,I to further
BUFFER
01 r.~- ........-.JYVV-O+12V broaden the network response. The 50-
100 ohm output level is recommended in the
interest of immunity to load changes: The
higher the output impedance of a buffer
chain the greater the chance for oscillator
pulling with load changes. Pk-pk output
across C3 should be on the order of 3 volts
when using the oscillator of Fig. lOA.
'Other VFO Criteria
Apart from the stability considerations
+9V(REG.l just treated, purity of emissions from
(A)
VFOs is vital to most designers. It is
prudent to minimize the harmonic output
of a VFO chain and to ensure that vhf
parasitic energy is not being generated
within the LO system.
The pi-network output circuit of Fig.
lIB helps reduce harmonics because it is a
+9V(REG.) low-pass network. Additional filtering can
be added at the VFO-chain output by
inserting a half-wave filter with a loaded
SOURCE
Q of I (XL and Xc= 50 for a 50-ohm line).
FOLLOWER Vhf parasitics are not uncommon in the
oscillator or its buffer stages, especially
FROM CI when high fT transistors are employed.
VFO O--i t--.-t~ The best preventive measures are keeping
~..............~~50-f'l the signal leads as short as possible and
~OUTPUT
~C3 0.1 adding parasitic suppressors as required.
The parasific energy can be seen as a
superimposed sine wave riding on the
VFO output waveform when a high-
01,02, - MPFI02, 2N44t6,
frequency scope is used.
E300, SUITABLE . A low-value resistor (10 to 22 ohms)
can be placed directly at the gate or base
of the oscillator transistor to stop
parasitic oscillations. Alternatively, one
(B) or two ferrite beads (950Ui) can be slipped
over the gate or base lead to resolve the
problem. If vhf oscillations occur in the
Fig. 11 - VFO buffer·and buffer/amplifier sections which provide isolation between the oscillator and buffer stages, the same preventive mea-
the VFO-chain load. The circuit at B is recommended for most applications.
sures can be taken.
VFO noise should be minimized as
practical number, of course), the less Q2 of Fig. IIA operates as an much as possible. A high-Q oscillator
likelihood of load shifts being reflected emitter-follower. The rf-voltage output tank will normally limit the noise band-
back to the oscillator. will be approximately 0.9 of that which is width adequately. Resistances placed in
Buffer stages can perform double duty supplied to the base. In a typical VFO the signal path will often cause circuit
by affording a measure of ·rf amplifica- chain, using an oscillator such as the one noise. Therefore, it is best to avoid the
tion, as needed. But, care must be taken to in Fig. lOA, this buffer strip will deliver temptation to control the rf excitation to a
avoid introducing narrow-band networks approximately 1 volt. pk-pk across the' given LO stage by inserting a series
in the buffer/amplifier chain if consider- 470-ohm emitter resistor of Q2. resistor. The better method is to use
able frequency range is planned, e.g., 5.0 A somewhat better circuit is offered in small-value coupling capacitors.
to 5.5 MHz. If suitable broadband charac- Fig. llB. QI is a JFET which has a high
teristics are not inherent in the. design, the input impedance (I-MOor greater) by vir- A Practical VFO
oscillator-chain output will not be constant o tue of the FET-device characteristic. This The circuit of Fig. 12 is for a high
across the desired tuning range. This minimizes loading of the oscillator. RFCI stability 1.8- to 1.9-MHz VFO: A
could seriously affect the conversion gain is chosen to resonate broadly with the stray circuit-board pattern and layout is in-
and dynamic range of a receiver mixer, or circuit capacitance (roughly 10 pF) at the cluded. Although the VFO frequency is
lower the output of a transmitter in some midrange frequency of the LO chain. Al- for 160-meter operation, other tuning
parts of a given band. though this does not introduce significant ranges up to'lO MHz are possible with
Fig. IIA illustrates. a typical RC selectivity, it does provide a rising charac- this circuit. The design guidelines offered
coupled VFO buffer with broadband teristic in the rf-voltage level at the source in the previous section will be useful in
response. CI is selected for minimum of QI. altering the circuit to other frequencies. A
coupling to the oscillator, consistent with Q2 functions as a fed-back amplifier close approximation can be had by simply
adequate drive to Q1. QI and Q2 should with shunt feedback and source degenera- taking the XL and Xc values for the
have high fT and medium beta to ensure a tion. The feedback stabilizes the amplifier components specified in Fig. 12A and
slight rf-voltage gain. Devices such as the and makes it broadband. The drain tank determining from those reactances the
2N2222A and 2N5179 are suggested. is designed as a pi network with a loaded new values for the frequency of interest.
6-7 Chapter 6
/
1.8-1.9MHz
C12
0.1
osc. RFC3
2.5 mH
01
1.8-1.9 MHz
AMP.
C19 o:;m
C1 6 0.001 ., ?5U;P~:MS)
Ql.02 Q3
C14
0.1
00'
G ~
;J; . '0->.".
cn
0.0036
~D~C E
EXCEPT AS INDICATED. DECIMAL VALUES OF
CAPACITANCE ARE IN MICROFARADS 1 jJF );
OTHERS ARE IN PICOFARADS (pF OR jJjJF):
RESISTANCES ARE IN OHMS:
D·
O.
RMS VOLTAGE
DC VOLTAGE
• • • • ••
Fig. 12 - At A is the circuit of a proven VFO chain which has exceptional stabilty under fairly constant room temperature. Although it is shown for
operation from 1.8 to 1.9 MHz. it can be modified and used as high as 10 MHz. The pc·board pattern and layout at Band C are suitable for any
operating frequency. The etched side of the board is shown in both examples. L1 is a 25·581'H inductor ,(Miller 43A47CBI); L2 is a 10·18.7 I'H inductor
(Miller 23A155RPC).
The tuning range will be determined by a highly stable VFO is mixed with energy oscillator energy as high as the microwave
the capacitance value of Cl. Data on from a crystal-controlled oscillator. The region. The PLL has the advantage that
precise component values for other fre- frequencies of the two oscillators are no mixing stage is used in conjunction
quency ranges are not available from the chosen so that spurious outputs generated with the output oscillator, so the output
ARRL. Further information on VFO during the mixing process do not fall with- energy is quite "clean." The Kenwood
design and the general circuit of Fig. in the. desired output range. A band-pass TS-820, the Collins 651 S-I, and the
12-A was provided by DeMaw in June filter at the mixer output attenuates out- National HRO-600 currently use PLL
1976 Ham Radio. of-band spurious energy. The charts given high-frequency oscillator systems.
in chapter 8 can be used to choose oscil- The basic diagram of a PLL is shown in
Premixing lator combinations which will have a Fig. 13B. Output from a voltage-controlled
It is difficult to build a variable-' minimum of spurious outputs. oscillator (VCO) and a frequency 'stan-
frequency oscillator for operation above dard are fed to a phase detector which
10 MHz with drift of only a few Hz. A PLL produces an output voltage equal to the
scheme called premixing shown in Fig. Receivers and transmitters of advanced difference in frequency between the two
13A, may be used to obtain VFO output design are now using phase-locked loops signals. The error voltage is amplified, fil-
in the 10- to 50-MHz range. The output of (PLLs) to generate highly stable local tered, and applied to tpe VCO. The error
HF Transmitting 6-8
mabie frequency dividers included so that
OUTPUT OUTPUT the yeO output is always locked to a
crystal reference. The frequency is changed
by modifying the instructions to the divi-
ders; steps of 100 Hz are usually employed
for hf receivers while lO-kHz increments
are popular in vhf gear.
YFO Dials
One of the tasks facing an amateur
PREMIXING PHASE-LOCKED LOOP builder is the difficulty of finding a
(A) (8)
suitable dial and drive assembly for a
YFO. A dial should provide a sufficiently
slow rate of .tuning - 10 to 25 kHz per
Fig. 13 - Block diagrams of the (A) premixing and (B) phase-lock loop methods. knob revolution is considered optimum -
without backlash. Planetary drives are
popular because of their low cost;
however, they often develop objectional
voltage changes the frequency of the yeO ceiver. A typical circuit is given in Fig. 14. backlash after a short period of use.
until it is locked to the standard. The Complete construction details on this PLL Several types of two-speed drives are
bandwidth of the error-voltage filter de- were given in January 1972 QST. A sec- available. They are well suited to home-
termines the frequency range over which ond type of phase-locked loop uses a made amateur equipment. The Eddystone
the system will remain in phase lock. stable mf YFO as the standard which sta- 898 precision dial has long been a favorite
Three types of phase-locked loops are bilizes the frequency of an hf or vhf yeo. with amateurs, although the need to ele-
now in use. The simplest type uses har- This approach is used in the receiver de- vate the YFO far above the chassis intro-
monics of a crystal standard to phase- scribed by Fischer in March 1970 QST. duces some mechanical-stability problems.
lock an HFO, providing the injection for The other PLL system also uses acrystal- If a permeability-tuned oscillator (PTO) is
the first mixer in a double-conversion .re- controlled standard, but with program- used, one of the many types of turn
BUFFER
. AMP./FILTER
VCO
Q2
+
1'~OO)JF
r+-,25V
TO PIN 470
I ,
EXCEPT AS INDIC'ATED, DECIMAL VAWES OF
CAPACITANCE ARE IN MICROFARADS (JIF ) :
OTHERS ARE IN PICOFARADS (pF OR JlJIFI;
RESISTANCES ARE IN OHMS;
~ -1,000. \01-1,000.000.
Fig. 14 - A practical PLL for a crystal-controlled HFO. Y1 is chosen so the harmonic content is ample at the desired output frequency. A 200-kHz crystal is
fine to 40 MHz. a 500-kHz ore is suitable to 60 MHz and a 1-MHz crystal is good for use to 80 MHz. L1 and L3 are resonant at the outpat frequency.
6·9 Chapter 6
..
21,3f.4f
SINGLE-ENDED MULTIPLIER
100 KHz
tA)
• • 21
02
2N4416 h
sooirh
'-----"G+t>-l D'
hp!-----J
PU$H-PUSH DOUBLER
+VDD
<B)
01
2N22'22A
(e) +VCC
1.5: 1 reduction should be employed so
that only the center of the capacitor range
is used. Then, as ~ final adjustment, the Fig. 17 - Single-ended inultiplier (A), push-push doubler (8) and pusl'l-pull tripler (e).
plates of the capacitor must be filed until
linear readout is achieved. In a PTO, the
pitch ofthe oscillator coil winding may be
varied so that linear frequency change use neon-glow tubes called-Nixies (a trade should be chosen, because this type of
results from the travel of the tuning slug. name of the Burroughs Corp.), or a seven- high-accuracy crystal exhibits the best
Such a VFO was described in July 1964 segment display using incandescent lamps, temperature stability. The clock output
QST. A different approach was employed filament wires in a vacuum tube, or LCDs energy is divided in decade-counter ICs to
by Lee (November 1970 QS1), using a (liquid crystal display), or LEDs (light- provide the pulse which opens the input
variable-capacitance diode (Vaiicap) as emitting diodes). A typical LED display is gate of the counter for a preset time. The
the VFO tuning element. A meter which shown in Fig. 15. The use of MSI and LSI number of rf cycles which pass through
reads the voltage applied to the Varicap circuits, some containing as many as 200 the gate while it is open are counted and
was calibrated to indicate the VFO transistors on a single chip, reduces the stored. Storage is used so that the readout
frequency. size required for an electronic diaf to a few does not blink. At the end of each
square inches of circuit-board space. counting cycle the information that has
Electronic Dials A typical counter circuit is given in Fig. been stored activates the display.-LEDs,
An electronic dial consists of a sim- 16. The accuracy of the counter is which present the numbers counted until
plified frequency counter which reads determined by a crystal standard which is another count cycle is complete. A
either the VFO or operating frequency of often referred to as a clock. The output complete electronic dial arranged. to be
a transmitter 'or receiver. The advantage from a loo-kHz calibration oscillator, the combined with an existing transmitter or
of . an electronic dial is the excellent type often used in receivers and trans- receiver was described in October 1970
accuracy (to 1 hertz, if desired) and the ceivers, may be employed if accuracy of QST. Also, Macleish et al reported, an
fact that VFO tuning does not have to be 100 Hz is sufficient. For readout down to adapter which allows a commerc~aHy
linear. The readout section ofthe dial may I Hz, a l-tolO-MHz AT-cut crystal made frequency counter to be mated with
HF Transmitting 6·10
ham gear so that the counter performs as Table 1
an electronic dial (May 1971 QSn, Chebyshev High·Pass and Low-Pass Filters-Attenuation (dB)
No. poles, ripple VSWR 2 Ie 3 Ie 4 Ie 5 Ie 6 Ie 71e
Frequency Multiplication 3 pole, 1 dB 2.66 22.46 34.05 41.88 47.85 52.68 56.74
It may be necessary to use frequency 3 pOle, 0.1 dB 1.36 12.24 23.60 31.42 37.39 42.22 46.29
3 pole, 0.01 dB 1.10 4.08 13.73 21.41 27.35 32.18 36.24·
multipliers at some point after the VFO or 3 pole, 0.001 dB 1.03 0.63 5.13 11.68 17.42 22.20 26.25
other frequency generator in a transmit- 5 pole, 1 dB 2.66 45.31 64.67 77.73 87.67 95.72 102.50
ter. When this need is present, the circuits 5 pole, 0.1 dB 1.36 34.85 54.21 . 67.27 77.21 85.26 92.04
of Fig. 17 can be applied. Of course, 5 pole, 0.01 dB 1.10 24.82 44.16 57.22 67.17 75.22 82.00 .
vacuum-tube multipliers are entirely suit- 5 pole, 0.001 dB 1.03 14.94' 34.16 47.22 57.16 65.22 71.99
able if the design is not one which uses 7 pole, 1 dB 2.66 68.18 95.29 113.57 127.49 138.77 148.26
semiconductors. The fundamental prin- 77 pole, 0.1 dB
pole, 0.01 dB
1.36
1.10
57.,72
47.68
84.83
74.78
103.11
93.07
117.03
106.99
128.31
118.27
137.80
127.75
ciples for frequency. multiplication are 7 pole, 0.001 dB 1.03 37.68 64.78 83.06 96.98 108.26 117.75
applicable to tubes and transistors alike. 9 pole, 1 dB 2.66 91.06 125.91 149.42 167.32 181.82 194.01
The requisite is that of operating 'the 9 pole, 0.1 dB 1.36 80.60 115.45 138.96 156.86 171.36 183.55
. devices in Class G, Although a transistor 9 pole, 0.01 dB 1.10 70.56 . 105.41 128.91 146.81 161.31 173.51
9pole, 0.001 dB 1.03 60.55 95.40 118.91 136.91 151.31 163.50
circuit may be seen with forward bias
applied to a frequency multiplier, the Note: For high-pass filter configuration 2fc becomes fc/2, etc.
stage must be. driven hard enough to
override the bias and operate Class C.
Forward bias is sometimes used in a Table 2
Chebyshev Low-Pass Filter - T Configuration
multiplier stage (solid state) to lower the
excitation requirements. Negative voltage No. poles, ripple L 1 L2' L3 L4 L5 C1 C2 C3 C4
(reverse bias) is often used on the grid of a 3 pole, 1 dB 16.10 16.10 3164.3
vacuum-tube multiplier, but forward bias 3 pole, 0.1 dB 8.209 8.209 3652.3
3 pole, 0.01 dB 5.007 5.007 .3088.5
is not. .3 pole, 0.001 dB 3.253 3.253 2312.6
The circuit of Fig. 17A is probably the 5 pole, 1 dB 16.99 23.88 16.99 3473.1 3473.1
least suitable for frequency multiplication. 5 pole, 0.1 dB 9.126 15.72 9.126 4364.7 4364.7
Typically, the efficiency of a doubler of 5 pole, 0.01 dB 6.019 12.55 6.019 4153.7 4153.7
this type is 50 percent, a tripler is 33 5 pole, 0.001 dB 4.318 10.43 4.318 3571.1 3571.1
percent, and a quadrupler is 25 percent. 7 pole, 1 dB 17.24 24.62 24.62 17.24 3538.0 3735.4 3538.0
7 pole, 0.1 dB 9.400 16.68 16.68 9.400 4528.9 5008.3 4528.9
Additionally, harmonics other than the 7 pole, 0.01 dB 6.342 13.91 13.91 6.342 4432.2 5198.4 4432.2
one to which the output tank is tuned will 7 pole, 0.001 dB 4.690 12.19 12.19 4.690 3951.5 4924.1 3951.5
appear in the output unless effective 9 pole, 1 dB 17.35 24.84 25.26 24.84 17.35' 3562.5 3786:9 3786.9 3562.5
bandpass filtering is applied. The collector 9 pole, 0.1 dB 9.515 16.99 17.55 16.99 9.515 4591.9 5146.2 5146.2 4591.9
tap on L 1 of Fig. 17A is placed at a point 9 pole, 0.01 dB 6.481 14.36 15.17 14.36 6.481 4542.5 5451.2 5451.2 4542.5
9 pole, 0.001 dB 4.854 12.81 13.88 12.81 4.854 4108.2 5299.0 5299.0 4108.2
which offers a reasonable compromise
between power output and spectral purity: Component values normalized to 1 MHz and 50 ohms. Lin "H; and C in.pF.
The lower the tap with respect to Vee, the
lighter the collector loading on Ll and the
greater the filtering action of the tuned Table 3
Chebyshev Low-Pass Filter - PI Configuration
circuit. The tradeoff is, however, a
reduction in output power as the mis- No. poles, ripple C1 C2 C3 C4 C5 L1 L2 L3 L4
match of the collector to the load 3 pole, 1 dB 6441.3 6441.3 7.911
increases. 3 pole, 0.1 dB 3283.6 3283.6 9.131
3 pole, 0.01 dB 2002.7 2002.7 7.721
A push-push doubler is seen at Fig. 3 pole, 0.001 dB 1301.2 1301.2 5.781
17B. Because of the conduction angle of 5 pole, 1 dB 6795.5 9552.2 6795.5 8.683 8.683
this type of circuit the efficiency is similar 5 pole, 0.1 dB 3650.4 6286.6 3650.4 10.91 10.91
to that of a straight amplifier which 5 pole, 0.01 dB 2407.5 5020.7 2407.5 10.38 10.38
operates in Class C. Also, the driving 5 pole, 0.001 dB 1727.3 4170.5 1727.3 8.928 8.928
pole, 1 dB
frequency (f) will be well attenuated at the 77 pole, 0.1 dB
6896.4 9847.4 9847.4 6896.4 8.85 9.34 8.85
3759.8 6673.9 6673.9 3759.8 11.32 12.52 11.32
doubler output if electrical balance and 7 pole, 0.01 dB 2536.8 5564.5 5564.5 2536.8 11.08 13.00 11.08
component symmetry are ensured. A 7 pole, 0.001 dB 1875.7 4875.9 4875.9 1875.7 9.879 12.31 9.879
12AU7A tube will work nicely in this type 9 pole, 1 dB 6938.3 9935.8 10,105. 9935.8 6938.3 8.906 9.467 9.467 8.906
of circuit well into the vhf region. Tl in 9 pole, 0.1 dB 3805.9 6794.5 7019.9 6794.5 3805.9 11.48 12.87 12.87 11.48
this example is a trifilar-wound, broad- 9 ~ole, 0.01 dB 2592.5 5743.5 6066.3 5743.5 2592.5 11.36 13.63 13.63 11.36
band toroidal transformer. It drives the 9 pole, 0.001 dB 1941.7 5124.6 5553.2 5124.6 1941.7 10.27. 13.25 13.25 10.27
gates of Q 1 and Q2 in push pull (opposite Component values normalized to 1 MHz and 50 ohms. Lin "H; C in pF.
phase). The drains are in parallel and are
tuned to 2f. R 1 is used to establish
electrical balance between QI and Q2. RI what better than a tripler using the circuit . current FCC regulations, wherein all spu-
is set while the doubler is being fully of Fig. l7A. If vacuum tubes are used. in rious emissions from a transmitter must
driven. Diode doublers can be used in a the circuits of Fig. 17, the input. ports be 40 dB or greater below the peak power
similar circuit, but the suJ:>ject will not be should employ high-impedance tuned of the desired signal, filtering is important.
treated here (see chapter four). . circuits for best performance. The type of filter used - band-pass, notch,
A push-pull tripler is illustrated in Fig. low-pass or high-pass - will depend on
l7C. Once again the matter of electrical Output Filtering the application. Band-pass fifters afford
balance and symmetry is important to Output purity from oscillators, multi- protection against spurious responses
good operation. The circuit discriminates pliers and amplifiers is of paramount above and below the amateur band for
against even harmonics, thereby aiding importance to the performance of numerous which they have been designed. Low-pass •
spectral purity. The efficiency is some- circuits. In the interest of compliance with filters attenuate energy above the desired
8-11
Tlble 4
Chebyshev Hlgh,Plss 'filter - TConfigurltlon
No. poles, ripple C1 C2 C3 C4 C5 L1 L2 L3 L4,
3 pole, 1 dB 1573.0 1573.0 8.005
3 pole, 0.1 dB 3085.7 3085.7 6.935
3 pole, 0.01 dB 5059.1 5059.1 8.201
3 pole, 0.001 dB 7786.9 7786.9 10.95
5 pole, 1 dB 1491.0 1060.7 1491.0 7.293 7.293
5 pole, 0.1 dB 2775.6 1611.7 2775.6 5.803 5.803
5 pole, 0.Q1 dB 4208.6 2018.1 4208.6 60.98 6.098
5 pole, 0.001 dB 5865.7 2429.5 5865.7 7.093 7.093
7 pole, 1 dB 1469.2 1028.9 1028.9 1469.2 7.160 6.781 7.160
7 pole, 0.1 dB 2694.9 1518.2 1518.2 2694.9 5.593 5.058 5.593 Fig. 19 - Here is a photograph of a 7-pole low-
7 pole, 0.Q1 dB 3994.1 1820.9 1820.9 3994.1 5.715 4.873 5.715 pass filter designed witt:l the information con-
7 pole, 0.001 dB 5401.7 2078.0 2078.0 5401.7 6.410 5.144 6.410 tained in Table 3. The filter is housed in a
9 pole, 1 dB 1460.3 1019.8 1002.7 1019.8 1460.3 7.110 6.689 6.689 7.110 small aluminum Minibox.
9 pole, 0.1 dB 2662.2 1491.2 1443.3 1491.2 2662.2 5.516 4.922 4.922 5.516
9 pole, 0.01 dB 3908.2 1764.1 1670.2 1764.1 3908.2 5.576 4.647 4.647 5.576
9 pole, 0.001 dB 5218.3 1977.1 1824.6 1977.1 5218.3 6.657 4.780 4.780 . 6.657
Component values normalized to 1 MHz and 50 ohms. Lin I'H; and C in pF. output frequency, while high~pass filters
reduce energy below the band of interest.
It is common practice to inc1udell harmo-
nic filter at the output of a VFO chain to
ensure purity of the driving voltage to a
Tlble 5 mixer or amplifier stage. The filter
Chebyshev Hlgh,Plss Filter - PI Conflgurltlon bandwidth must be adequate for the
No. poles, ripple L1 L2 L3 L4 L5 C1 C2 C3 C4 tuning range of the VFO in order to
3 pOle, 1 dB 3.932 3.932 3201.7 prevent attenuation of the output energy
3 pole, 0.1 dB 7.714 7.714 2774.2 within the desired band. For this reason, a
3 pole, 0.Q1 dB 12.65 12.65 3280.5 low-pass type of filter is used in preference
3 pole, 0.001 dB 19.47 19.47 4381.4
to a bandpass one by some designers.
5 pole, 1 dB 3.727 2.652 3.727 2917.3 2917.3 The information contained in Figs.
5 pole, 0.1 dB 6.939 4.029 6.939 2321.4 2321.4
5 pole, 0.01 dB 10.52 5.045 10.52 2439.3 2439.3 18-20 and in Tables 1-5 will allow the
5 pole, 0.001 dB 1.466 6.074 1.466 ' 2837.3 2837.3 builder to select an appropriate
7 pole, 1 dB 7.159 5.014 5.014 7.159 1469.2 1391.6 1469.2 Chebyshev filter design to fulfill a par-
7 pole, 0.1 dB 6.737 3.795 3.795 6.737 2237.2 2023.1 2237.2 ticular need. Information is included for
7 pole, 0.Q1 dB 9.985 4.552 4.552 9.985 2286.0 1949.1 2286.0 both high-pass and low-pass filters with 1,
7 pole, 0.001 dB 13.50 5.195 5.195 13.50 2564.1 2057.7 2564.1
0.1, .0.01 and 0.001 db passband ripple.
9 pole, 1 dB 3.651 2.549 2.507 2.549 3.651 2844.1 2675.6 2675.6 2844.1
9 pole, 0.1 dB 6.656 3.728 3.608 3.728 6.656 2206.5 1968.9 1968.9 2206.5 These figures correspond to VSWRs of
9 pole, 0.Q1 dB 9.772 4.410 4.176 4.410 9.772 2230.5 1858.7 1858.7 2230.5 2.66, 1.36, 1.10 and 1.03 respectively. Ad-
9 pole, 0.001 dB 13.05 4.943 4.561 4.943 13.05 2466.3 1911.8 1911.8 2466.3 ditionally, information is provided for
Component values normalized to 1 MHz and 50 ohms. L in I'H; and C in pF. both "T" and "pi" types of filter con-
figurations. .
The filters are "normalized" to a fre-
quency of 1 MHz and an input and output
impedance of 50 ohms. In order to trans-
late the designs to other frequencies, all
that is necessary is to divide the compo-
0,.--_.
nent values by the new frequency in MHz.
(The I-MHz value represents a "cutoff"
'10
frequency. That is, the attenuation in-
creases rapidly above this frequency for
-20 / 5 POLE •.001 dB
the low-pass filter or below this frequency
/ /5 POLE,.Ol dB
5 POLE,.l dB for the high-pass filter. This effect should
-30 5 POLE, 1 dB not be confused with the variations in at-
tenuation in the passband.) For instance,
ATTENUATION -40 if it is desired to reduce harmonics from a
(dB)
VFO at frequencies above 5 MHz (the new
-50 cutoff frequency), the inductance and
capacitance values would be divided by
-60 5.0.
Other impedance levels can also be used
-70 by multiplying the inductor values by the,
ratio Zo/50 and the capacitor values by
-80 __ ____ ____ ____ ____ __ ______
~ ~~ ~ ~ ~ ~~ ~~~-L
Ll
'"~ I
rh
High·Pass Filter - Pi configuration
(See Table 5 for normalized values)
Cl C2 CN
o--1~~~--t--I~----o
I
I ~0.01
~'"
+275V
. BIAS
(-V) (A)
I ,
rh
DRIVER
High·Pass Filter - T configuration
Q1
(See Table 4 for normalized values) 2N4427 TO
r--.!........_-~ l -------f'o PA
~GRID
~mn_~
f 0.001 (HI-Z)
INPUT~I---_--f-~
(LO-Z) 0.1
Cl CNl
Low·Pass Filter - T configuration
(See Table 2 for normalized values)
8·13
/
AMP. AMP.
-Z= 6DDn
[UJ
~
VFD _ ~ \
INPUT ~H ....-t-{
(IDmWJ .0.1
15.0.0
(SMA) (4DMAJ
composite transmitter. Some designs con- important to use low-Q, low-inductance rf region, thereby making the stage suitable
tain a mixture (hybrid) of tubes and chokes and matching networks to dis- for ssb signal.amplification. A l.5-ohm
semiconductors, while other circuits have courage low-frequency tuned-base, tuned- resistor can be added between the emitter
no vacuum t).lbes at all. If tubes are used collector oscillations. The suppression and ground to help prevent thermal
in a hybrid circuit, they are generally concepts just discussed are illustrated in runaway and to introduce degeneration
restricted to the driver and PA sections of Fig. 21 at Band C. In the circuit at B there (feedback) for enhancing stability. No
the transmitter. There is no particular are two 950-mu ferrite beads added over bypass capacitor would be used from
reason why tubes should be used in the pigtail of RF~ 1 to swamp the Q of the emitter to ground if this were done. T 1 is a
preference to power transistors for output choke. Three bypass capacitors (0.001, narrow-band toroidal rf transformer that
powers up to, say, 150 watts, despite the 0,01 and 0.1 u F) are used with RFC2 of has a turns ratio suitable for transforming I
prevailing myth that tubes are more Fig. 20B to provide effective rf decoupling the collector impedance to the grid
rugged, operate more stably, and produce from vhf to mf. A 22-uF capacitor is used impedance (determined by the value of the
less spurious output. It is true that near RFC2 to bypass the +Vee line at low grid resistor of the PA) of the final
transistors are less tolerant than tubes to frequency and audio. This .method is amplifier. The secondary winding of T1 is
SWR levels in excess of 2:1, but a recommended for each high-gain solid- tuned to resonance at the operating
correctly designed transistor amplifier can state stage Hl a transmitter. frequency. Approximately 1 watt of
be operated safely if SWR-protection power output can be taken from Q 1 in the
circuitry is included. Furthermore, spec- Driver Circuits hf region when a 12-volt Vee is used. This
tral purity can be just as good from a The circuits of Fig. 2lA and 21B are is ample power for driving a pair of6146B
solid-state amplifier as it is from a tube typical of those which would be employed tubes in Class AB 1.
type of amplifier. A harmonic filter to excite a tube type of PA stage. The A broadband type of solid-state driver,
normally follows a solid-state power 6GK6 tube driver at A ca,n be biased for is shown in Fig. 21C. The tradeoff fori
stage, whereas this measure may not be Class C or Class AB operation, making it broadband operation (1.8 to 30 MHz in
required when tubes are used in the suitable for cw or ssb service. Of course, this example) is a reduction in maximum
amplifier. Amplifier IMD (third- and the AB mode would be suitable for cw and available gain (MAG). Therefore, the
fifth-order products) in solid-state power ssb, and would require considerably less output power from Ql of Fig. 21C will be
stages which operate linearly is fully as excitation power than would the same less than 1 watt. The stage operates Class
acceptable as that which is observed in stage operating in Class C. Other tubes A, making it linear. The emitter is
most tube types of linear amplifiers. that perform well in this circuit are the unbypassed to .provide emitter degenera-
Typically, if a design is correct, the IMD 6CL6, 12BY7A and 5763. The output tion. Shunt feedback is used between the
will be on the order of - 33 dB from the tank is designed for high impedance in base and collector to enhance stability and
reference power value. order to interface properly with the contribute to the broadband characteristic
The major area of concern when high-impedance grid of the PA. It may be of the circuit. T 1 is a broadband conven-
designing a solid-state driver or PA necessary to include a neutralization tional transformer wound on a toroid core.
section is to prevent low-frequency self- circuit with this type of amplifier, The turns ratio is adjusted to match the
oscillations. Such' parasitics tend to especially if careful layout is not used. The approximate 200-ohm collector impedance
modulate the carrier and appear as high transconductance of the 6GK6 series to the base impedance of the transistorPA
spurious responses within the amplifier encourages self-oscillation near the opera- stage. The latter is typically less than 5
passband. The low-frequency parasitics ting frequency. Z 1 is a parasitic choke ohms. Heat sinks are required for the tran-
occur as a result of the extremely high which should be included as a matter of sistors of Fig. 21B and C. The primary of
gain exhibited by hf and vhf transistors at course to prevent vhf parasitics. T1 should have a reactance of roughly
the low-frequency end of the spectrum. A transistor amplifier which is suitable four times the collector impedance. This is
The theoretical gain increase for a ·given for driving a Class C tube PA is presented related to the lowest proposed operating
transistor is6 dB per octave as the in Fig. 21B. Ql operates Class C, so it is frequency. Therefore, for 1.8 MHz the
operating frequency is lowered. The same not satisfactory for amplifying ssb energy. primary winding would be 70 u H (XL =
is not true of vacuum tubes. Therefore, it However, forward bias (approximately 800 ohms). This can be achieved easily by
is necessary to employ quality decoupling 0.7 volt) can be added to move the using an FT-50-43 Amidon core. The pri-
and bypassing in the circuit. It is similarly operating curve into the Class AB (linear) mary advantage to a broadband driver is
HF Transmitting 6-14
CAPACITIVE COUPLING BANDPASS COUf'LlNG
output range from 15 to 30 watts. Cl at the
emitter of Q I can be selected to provide
PA
the overall gain needed in this strip. The .
V2
PA value given at Cl proved suitable for the
V2 ARRL version of this amplifier. The final
value will depend on the gain of, the
individual transistors acquired for this
circuit.
Coupling Between Transmitter Stages
Correct impedance matching between a
stage and its load provides maximum
transfer of power. The load can be an
+v antenna or' a succeeding stage in a
+V
(A) (B)
transmitter. Thus, the output impedanre-
of astag~ must be matched to the input of
the following stage. Various forms of
TRANSFORMER COUPLING
DRIVER
CAPACITIVE DIVI DER COUPLING coupling networks are popular for use in
PA DRIVER tube or transistor circuits. The choice will
rn
Tl
Q2 Qt PA depend on a number of considerations -
available driving power versuS tolerable
mismatch, selectivity required and the
impedance levels being matched. When
working with transistors, the collector
rh impedance can be approximated by
V 2
Z=~
2Po
1 - Select QL
2 - XL1 = QL,RIN + Xc(OUT) +vee
o·tl WHEN RIN < RL,
3-XC2=ARL
1 - Select QL
4-XC1 = _B_ 2 - XL1 = XC(OUT)
QL-A 3 - Xc1 = QL RIN
Where A =/[RIN(1R~ 2
QL )] _ 1
Fig. 25 - Circuit and equations for network no. 2.
And B = RIN (1 + QL2)
Q2
Fig. 24 - Circuit and mathematical solution for
matching network no. 1.
6·17 Chapter 6
RF both circuit locations. Generally, the low-
AMP. Z1 est power port is best suited for the sup-
pression devices discussed - here. This
means that the resistor or ferrite beads
should be located at the base terminal of
the transistor.
Because of the rising gain characteristic
of bipolar transistors as the frequency is
lowered, shunt anp degenerative feedback
are often used to prevent instability. The
net effect is that in the regions where
low-frequency self-oscillations are most
likely to occur, the feedback increases by
nature of the feedback network. The
Fig. 28- Example of neutralization of a single-ended rf amplifier.
heavier the feedback, the lower the
amplifier gain. In the circuit of Fig. 30 C 1
and R3 provide negative feedback which
RF increases progressively as the frequency is
AMP lowered. The network has some effect at
~
the desired operating frequency - a
gain/stability tradeoff - but has a
pronounced effect at the lower frequencies.
The values for C 1 and R3 are chosen ex-
~
perimentally in most instances, the precise
values being dependent upon the operat-
+vcc ing frequency of the amplifier and the
(A) amount of feedback voltage available
from the tap-off point. C 1 will usually be
between 220pF and 0.0015uF for hf~band
RF amplifiers. R3 may be a value from 51 to
AMP.
5600 ohms. A rule of thumb that may
~
Q1 Z2 prove helpful is to use a network that re-
Z1
duces the stage gain by apprcximately 1.5
dB at the lowest operating frequency.
R2 of Fig. 30 provides emitter degene-
~ ration at low frequencies when the bypass
capacitor, C2, is chosen for adequate rf
+vcc bypassing at the intended operating fre-
(8) quency. Below the desired frequency, C2
becomes progressively less effective as the
Fig. 29 - SuppresSi.on methods for vhf and uhf parasitics in solid-state amplifiers. frequency is lowered, thereby increasing
the degenerative feedback caused .by R2.
This lowers the amplifier gain. R2 in a
power stage is seldom greater than 10
should be kept as short as possible in an suppression can be had by inserting a
ohms in value, and may be as low as 1
amplifier circuit, and likewise with the low-ohmage resistor (10 to 51 ohms) in
ohm. It is importa:nt to consider that
pigtails of bypass capacitors. This applies series with the tube input, near the tube
under some operating and layout condi-
to tube or transistor amplifiers. socket. This is illustrated by R 1 of Fig. 28.
tions R2 can cause instability. This form
Z1 of Fig. 28 is a vhf parasitic choke. Vhf or uhf parasitics can be detected by
of feedback should be used only in those
Such a network will damp self-oscillations means of a high-frequency scope, or by
circuits' where unconditional stability can
at vhf and uhf by acting' as a series probing the plate tank with a sensitive
impedance which breaks up the usual wavemeter which tunes from 30 MHz and be achieved.
. unwanted vhf/uhf circuit path. Z1 con- higher. Solid-state amplifiers that are built on
sists of a non-inductive resistor between Parasitic· oscillations can be prevented pc boards can be made stable (in addition
51 and 100 ohms. A coil is wound around in solid-state amplifiers by using a small to the foregoing measures) by utilizing
the resistor body to provide a broadband amount of resistance in the base or double-clad pc board material. The
rf choke which presents a high impedance collector lead of low-power amplifiers copper on the component side of the
at vhf and higher, but looks like a low (Fig. 29A). The value of R1 or R2 is . board is used as a ground-plane surface by
reactance in th,e hf region and lower. A typically between 10 and 22 ohms. removing the copper around each hole
typical parasitic suppressor for a power Resistors are seldom necessary at both where a component is to be mounted. This
level up to 150 watts contains 6 to 8 turns points in a circuit, but can be used ground plane is made electrically common
ot no. 20 wire wrapped around a 56-ohm, effectively at either point. R 1 or R2 to the ground elements on the etched side
..i-watt composition resistor. The coil ends should be located as close to the transistor of the board. Such a technique helps
are soldered to the resistor pigtails near as practical. ' prevent unwanted ac ground loops which
the body of the resistor. Z1 is then placed At power levels in excess of approxi- can cause feedback and instability. Fur-
as close to the tube plate pin or cap as mately 0.5 Watt, the method of parasitic thermore, the nonground etched elements
possible. For higher rf powers it is suppression shown in Fig. 29B is suggested. form low-capacitance bypass capacitors
practical to use a high-wattage Globar The voltage drop across a resistor would against the ground-plane surface of the
resistor, or a 25-watt noninductive (NIT) be prohibitive at the higher power levels, board. This aids in reducing the chance
power resistor around which a suitable so one or more ferrite beads can be substi- for vhf and uhf parasitic oscillations.
number of wire turns (no. 14 or 12 wire) tuted (Z1 and Z2). A permeability of 125 R 1 of Fig. 30 is useful in swamping the
have been wound. Additianal parasitic· will suffice. The beads need not be used at input of an amplifier. This reduces the
HF Transmitting 6·18
chance for low-frequency self~osbillations, RF
but has' a minor effect on the amplifier AMR
Ct R3
performance in the desired operating
range. Values from 3 to 27 ohms are
typical. When they are in shunt with the
normal (low) base impedance of a power
amplifier, they contribute only slightly to
the lowering of the device input im-
pedance. The rule of thumb is to use the
largest value of resistance that will ensure
stability. RI should be located as close to
the transistor base terminal as possible. +vcc
The pigtails must be kept short to pt:event
stray inductances from forming. It is
helpful to use two resistors in parallel to Fig. 30 - Illustration of shunt feedback in a transistor amplifier. Components C 1 and R3 comprise the
limit the amount of inductive reactance feedback network.
introduced by a single resistor.
C3 of Fig. 30 can be added to some capacitance of C3 must be absorbed into expressions "bal-oon" and "bal-um" are
power amplifiers to damp vhfluhf parasi- the network design in the same manner as not correct.
tic oscillations. The capacitor should be the Cout of the transistor. The broadband transformers illustrated
low in reactance at vhf and uhf, but must in Figs. 31, 32 and 33 are suitable for use
present a high reactance at the operating Broadband Transformers in solid-state circuits, as matching devices
frequency. The exact value selected will The usefulness of toroidal broadband between circuit modules and in antenna-
depend upon the collector impedance. A transformers is practically beyond de- matching networks. For low power levels
reasonable rule of thumb is to use an Xc of scription in this text. But, some of the the choice of core material is often ferrite.
10 times the collector impedance at the more popular transformer configurations Powdered-iron is more often the designers
operating frequency. Silver-mica or ceramic- are presented here for those who wish to preference when working with fairly high
chip capacitors are suggested for this ap- employ them in matching networks levels of power. The primary objection of
plication. For example, a 3.5-MHz ampli- associated with solid-state devices and some, respective to the use of ferrite at
fier with a lO-ohm collector impedance tubes. It is important to realize that high power, is damage to the core material
would use a capacitor with an Xc of 100 broadband transformers are best suited to during saturation and overheating. This
ohms. C 1 would be 454 pF under this rule. low-impedance applications, say, up to a can alter the permeability factor of the
At 150 MHz the same capacitor would few hundred ohms. they ~hould be core material permanently. Powered-iron
have an Xc of only 2.3 ohms, making it an thought of as devices which can transform is more tolerant in this regard.
effective vhf and higher bypass element. one impedance to another, in terms of the Fig. 31 shows two types of 4: 1
An additional advantage is seen in the by-. transformation ratio they make possible. transformers, plus a Imethod for con-
passing action for vhf and uhf harmonic They should not ·be regarded as devices necting two of them in series to effect a
energy in the collector circuit. C3 should which are built for some specified pair of 16: 1 transformation. The circuit at E is
be placed as close to the collector terminal impedances, such as 200 ohms to 50 ohms often used between a 50-ohm source and
as possible, using short leads. The effects in the case of a 4: 1 transformer. The term the base of an rf power transistor.
of C3 in a broadband amplifier are "balun," despite its misuse, pertains only Two styles of 9: 1 transformer are seen
relatively insignificant at the operating to a broadband transformer which con- in Fig. 32 at A and C. They are also found
frequency. However, when a narrow-band verts a balanced condition to one which is at the input to transistor amplifiers and
collector network is employed, the added unbalanced, or vice versa. The often-heard between the collector and the load. The
.---...,-------0 RL 20011.
BALANCED
.INDICATES PHASING
RL
4'1 UNBALANCED TO UNBALANCED
200
HI-Z
6·19 Chapter 6 .
9'.1 LmBALANCED TO
. UNBALANCED TRAN.
(AI
RIM
Fig. 32 - Circuit examples of 9:1' broadband transformers (A and C) and a variable-impedance transformer (El.
variable-ratio transformer of Fig. 32 C is the chapter on electrical laws and circuits. the tube must dissipate more heat. Other
excellent for obtaining a host of im- It is important to remember that true types of operation, such as cw or
pedance transformations. This trans- plate, . screen or biasing voltage· is the single-sideband phone are intermittent in
former was developed by W2FMI for use voltage between the particular electrode nature, resulting in less average heating
in matching ground-mounted vertical and filament or cathode. Only when the than in other modes where there is a
antennas. cathode is directly grounded to the chassis continuous power input to the tube during
Phase-reversai, 1: 1, balun and hybrid- may the electrode-~o-chassis v.oltage be transmissions. There are also different
combiner transformers are shown in Fig. taken as the true voltage. The required rf ratings for tubes used in transmitters that
33. The circuit at E of Fig. 33 is useful driving voltage is applied between grid are in almost constant use (CCS -
when it is necessary to feed two signals to and cathode. Continuous Commercial Service), and for
a single load. When the input signals are Plate power input is the dc power input tubes that are to be used in transmitters
on different frequencies the power is split to the plate circuit (dc plate voltage X dc that average only a few hours of daily
evenly between R3 and R4. When the platt current). Screen power input like- operation (ICAS - Intermittent Com-
input voltages are on the same frequency wise is the dc screen voltage X the dc screen mercial and Amateur Service). The latter
(as with two transistor amplifiers feeding current. . are the ratings used by amateurs who wish
a single load), with the amplitudes and Plate dissipation is the difference to obtain maximum output with reasonable
phase identical, all of the power is between the rf power delivered by the tube tube life.
delivered to R4. to its loaded plate tank circuit and the dc
RF Power Amplifier Circuitry plate power input. The screen, on the Maximum Tube Ratings
other hand, does not deliver any output Maximum ratings, where they differ
In addition to proper tank and output- power, and therefore its dissipation is the
coupling circuits, an rf amplifier must be from the values given un~er typical
same as the screen power input. operating values, are not normally of
provided with suitable operating voltages
Trtmsmitting-Tub, Ratings significance to the amateur except in
and an rf driving or excitation voltage. All
special applications. No single maximum
rf amplifier tubes require a voltage to Tube manufacturers specify the maxi- value should be used unless all other
operate the filament or heater (ac is mum values that should be applied to the ratings can simultaneously be held within
usually permissible), and a positive dc tubes they produce. They also publish sets
voltage between the plate and filament or the maximum values. As an example, a
of typical operating values that should tube may have a maximum plate-voltage
cathode (plate voltage). Most tubes also result in good efficiency and normal tube rating of 2000, a maximum plate-current
require a negative dc voltage (biasing life. I rating of 300 mA, and a maximum
voltage) between control grid (grid no. 1) The same transmitting tube may have plate-power-input rating of 400 watts.
and filament or cathode. Screen-grid different ratings, depending upon the Therefore, if the maximum plate voltage
tubes require in addition a positive manner in which the tube is to be of 2000 is used, the plate current should be
voltage (screen voltage or grid no. 2 operated and the service in which it is to limited to 200 mA (instead of 300 mA) to
voltage) between screen and filament or be used. These different ratings are based stay within the maximum power-input
cathode. primarily upon the heat that the tube can rating of 400 watts.
Biasing and plate voltages mllY be fed to safely dissipate. Some types of operation,
the tube either In series or parallel with the such as with grid or screen modulation, Maximum Transistor Ratings
associated rf tank circuit as discussed in are less efficient than others, meaning that Transistor data sheets specify a maxi-
HF Transmitting 6-20
.= R3
R1
50 Rl
UNBAL. 50
BAL .
. 1~
•PHASE-REVERSAL r--
UNBALANCED TRAN. ~c
(A)
• -INDICATES PHASING 1:1 BALANCED 51G. iI
L __
UNBALANCED TRAN.
(C)
SINGLE-ENDED
IE) HYBRID COMBINER
Rl
50
BAL.
(0)
mum operating voltage for several condi- frequency and is by no means all-inclusive facturers push the power margin a bit har-
tions. Of special interest to amateurs is the from, for example, medium frequency up der, utilizing a transistor which delivers a
Veeo specification (collector-to-emitter vol- to the vhf spectrum. The frequency at power output which is as great as 3/4 the
tage, with the base open). When a transis- which a particular gain figure applies is Po rating. So close a safety margin is
tor is called upon to handle an ac signal, stated on the transistor data sheet. Gene- somewhat risky for inexperienced builders.
the collector-to-emitter voltage can rise to rally, the gain will be higher below that Sources of Tube Electrode Voltages:
twice·the dc supply. Thus, if a 12-volt sup- frequency and it will decrease above that Filament or Heater Voltage
ply is used, the transistor should have a frequency.· Gain information is useful in
Veeo of 24 or greater to prevent damage. If predicting how much output power can be The heater voltage for the indirectly
that same transistor is amplitude-modulated obtained for a given input power; i.e., a heated cathode-type tubes found in low-
(as in the PA of an a-m transmitter), a 13-dB gain transistor delivering an output power classifications may vary.' 10 percent
collector-emitter voltage swing (theoreti- of 10 watts would require a driving power above or below rating without seriously
cal) as great as four times the supply vol- of 0.5 W (Gaifi(dB) = 10 log [P2/Pl]) reducing the life of the tube. But the vol-
tage can occur. A transistor chosen for Power dissipation for a transistor is ex- tage of the higher-power, filament-type
this application should have a Veeo of 48 pressed symbolically as Po. This maximum tubes should be held closely between the
or greater. rating is based on a case temperature of rated voltage as a minimum and five per-
The fT rating of a common-emitter 25°C. For example, a total device dissipa- cent above rating as a maximum. Make
transistor amplifier is based on the point tion of 30 watts might be specified at a sure that the plate power drawn from the
at which the transistor gain is unity (1) case temperature of 25°C. If greater tem- power line does not cause a drop in fila-
with respect to operating frequency. Inthe peratures were expected, the transistor ment voltage below the proper value when
interest of predictable performance and would· have to be derated in mW per de- plate power is applied.
amplifier stability it is best to select a gree C. A Motorola MRF215 would be , Thoriated-type filaments lose emission
transistor that was designed for a par- derated- 177 mW per additional degree C. when the tube is overloaded appreciably.
ticular frequency range. When this is not The effectiveness of the transistor heat If the overload has not been too pro-
practical, the. fT should be roughly 5 to 10 sink plays an important role in maximum longed, emission sometimes may be re-
times the operating frequency. Therefore, power utilization of a given device. It is stored by operating the filament 'at rated
a suitable transistor for use at 3.5 MHz not unusual·to see a cooling fan used in voltage with all other voltages removed
would have an fr between 17.5 and 35 combination with a large heat sink to aid for a period of 10 minutes, or at 20 per-
MHz. If a much higher fr were selected, in lowering the transistor case tempera- cent above rated voltage for a few minutes.
say, 250. MHz, the published rf perfor- ture: Heat is one of the worst enemies of Plate Voltage
mance curves for the .device would be power transistors.
quite inaccurate at 3.5 MHz, and the tran- A rule of thumb for selecting a Po rat- Dc plate voltage for the operation of rf
sistor gain would be extremely high com- ing which is suitable for a given rf power amplifiers is most often obtained from a
pared to the rated gain at the intended output amount is to choose a transistor transformer-rectifier-filter system (see
operating frequency of the device. which has amaximum dissipation of twice power-supply· chapter) designed to deliver
PowC!r transistor gain is normally the desired output power. Hence, a 20- the required plate voltage at the ~equired
specified as "typical" dB. This informa- watt transistor would be picked for use in current. However, batteries or other dc-
tion applies to some specified operating a lO-watt-output amplifier. Some manu- generating devices are 'Sometimes used in
·6·21 Chapter 6
j
certain types of operation '(see portable- operation (Fig. 3SA), a resistance of the
mobile chapter). appropriate value can be placed in the
emitter return as shown. Most transistors
Bias and Tube Protection will operate iIi Class C without adding
Several methods of obtaining bias are bias externally, but in some instances the
shown in Fig. 34. At A, bias is obtained by amplifier efficiency can be improved by
the voltage drop across a resistor (Rl) in means of emitter bias. Reverse bias
the grid dc return circuit when rectified supplied to the base of the Class C
grid current flows. The proper value of transistor should be avoided because it
resistance may be determined by dividing will lead to internal breakdown of the
the required biasing voltage by the dc grid device during peak drive periods. The
current at which the tube will be operated. destruction is frequently a cumulative
Then, so long as the rf driving voltage is phenomenon, leading to gradual destruc-
(Al
adjusted so that the dc grid current is the tion of the transistor junction.
recommended value, the biasing voltage A simple method for Class AB biasing
will be the proper value. The tube is is seen in Fig. 3SB. Dl is a silicon diode
biased only when excitation is applied, which acts as a bias clamp at approximately
since the voltage drop across the resistor 0.7 V. The forward bias establishes linear-
depends upon grid-current flow. When amplification conditions. That value of
excitation is removed, the bias falls to bias is not always optimum for a specified
zero. At zero bias most tubes draw power transistor in terms of IMD. Variable bias
far in excess of the 'plate-dissipation of the type illustrated in Fig. 3SC permits
rating. It is advisable to make provision for the designer sufficient variance to locate
protecting the tube when excitation fails by the best operating point respective to linea-
accident, or by intent as it does when a rity.
precendip.g stage in a cw transmitter is keyed.
If the,hlaximum cw ratings shown in the Screen Voltage for Tubes
tube tables are to be used, the input For cw and fm operation, and under
should be cut to- zero when the key is certain conditions of phone operation
open. Aside from this, it is not necessary (amplitude modulation) the screen may be
that plate current be cut off completely Fig, 34 - Various techniques for providing
operated from a power supply of the same operating bias with tube amplifiers,
but only to the point where the rated type used for plate supply, except that
dissipation is not exceeded. In this case voltage and current ratings should be
plate-modulated phone ratings should be appropriate for screen requirements. The
used for cw operation, however. screen may also be operated through a
With most tubes this protection, plus series resistor or voltage-divider' from a
the required operating. bias, can be source of higher voltage, such as the
INPUTo--_-->-+--f
HF Transmitting 6·22
·
An approxImate \
value for the screen- for each tube under various operating impedance step-up is necessary if the grid
v.oltage dropping resistor may be obtained conditions. These figures, however,do not is .to be fed from a, low~impedance
by dividing the voltage drop required from include ,circuit losses. In general, the transmission line.
the supply voltage (difference between the driver stage for any Class C amplifier
supply voltage and rated screen voltage) should be capable of supplying at least Cooling: Tubes
by the rated screen current in decimal . three times the driving power shown for Vacuum tubes must be operated within
parts of an ampere .. Some further ad- typical operating conditions at frequencies the temperature range specified by the
justment may be necessary, as mentioned up to 30 MHz and from three to 10 times at manufacturer if long tube life is to be
above, so an adjustable resistor with a higher frequencies. achieved. Tubes with glass envelopes
total resistance' above that calculated Since the dc grid current relative to the rated at up to 25 watts of plate dissipation
should be provided. biasing voltage is related to the peak may be run without forced-air cooling, if a
driving voltage, the dc grid current is moderate amount of cooling by con-
Protecting Screen-Grld Tubes commonly used as a convenient indicator vection can be arranged. If a perforated-
Considerably less grid bias is required of driving conditions. A driver adjustment metal enclosure is used, and a ring of
to cut off an amplifier that has a that results in rated de grid current when 1/4-inch diameter holes are placed around
fixed-voltage screen supply than one thilt the dc bias is at its rated value, indicates the tube socket, normal air flow can be
derives the screen voltage through a high proper excitation to the amplifier when it relied upon to remove excess heat at room
value of dropping resistor. When a "stiff' is fully loaded. temperatur.. .
screen voltage supply is. used, the neces- In coupling the grid input circuit of an For tubes with greater plate dissipation,
sary grid cutoff voltage may be determined amplifier to the output circuit of a driv~ng or those operated with plate currents 'in
from an inspection of the tube curves or stage the objective is to load the driver excess of the manufacturer's ratings
by experiment. plate circuit so that the desired amplifier (often the case with TV sweep tubes)
grid excitation is obtained without ex- forced air cooling with a fan or blower is
Feeding Excitation to the Grld ceedingthe plate-input ratings of the needed. Fans, especially those designed
The required rf driving voltage is driver tube. for cooling hi-fi cabinets, are preferred
supplied by an oscillator generating a Drlving Impedance because they operate quietly. However, all
voltage at the desired frequency, either fans lose their ability to move air when ex-
directly or through intermediate amplifiers, The grid-cummt flow that results when cessive back pressure exists. For applica-
mixers, or frequency multipliers. the grid is driven positive in respect to the tions where a stream of air must be
As explained in the chapter on vacuum- cathode over a portion of the excitation directed through a tube socket, a blower is
tube fundamentals, the grid of an. amplifier cycle represents an average resistance usually required.
operating under Class C conditions must across which the exciting voltage must be One method for directing a flow of air
have an exciting voltage whose peak value developed by the driver. In other words, around a tube envelope or through tube
exceeds the negative biasing voltage over a this is the load resistance into which the cooling fins involves the use of a'
portion of the excitation cycle. During driver plate circuit must be coupled. The pressurized chassis. This system is shown
this portion of the cycle, current will flow approximate grid input resistance is given in Fig. 36. A blower is attached to the
in the grid-cathode circuit as it does in a by chassis and forces air up through the tube
diode circuit when the plate of the diode is socket and around the tube. A chimney
positive in respect to the cathode. TI Input impedance (ohms) (not shown in this drawing) is used to
quires that the rf driver supply power. The driving power (watts) guide the air around the tube as it leaves
-:--......-:--.,....,...-.,...,""" X 620,()()O the socket. A chimney will prevent the air
power required to develop the required dc grid current (mA)l
peak driving voltage across the grid-cathode from being dispersed as it hits the
impedance of the amplifier is the rf driving For normal operation, the driving power envelope or cooling fins, concentrating
power. and grid current may be taken from the the flow for maximum cooling.
The tube tables give approximate tube tables. Since the grid input resistance Most manufacturers rate tube cooling
figures for the grid drivin~ power required is a matter of a few thousand ohms, an requirements for continuous-duty opera-
tion. The manufacturer's literature will in-
dicate the required cubic feet per minute
(CFM) of air flow at some particular back
EXHAUST pressure. Back pressure is the pressure
that is built up inside the airtight chassis
when the blower is operational. Forced air
entering the chassis from the blower can
escape only through the tube socket/tube/
~ANOMETER
chimney assembly.' Since this assembly
_____ AIR SYSTEM represents a certain amount of resistance
, SOCKET to the flow of air, an amount of pressure
is built up inside the chaSSIS. The exact
amount of pressure will depend on the
blower and the tube socket/tube/chimney
characteristics. Blowers vary in their abili-
ty' to work against back pressure so the
--
AIR TIGHT CHASSIS
matter of blower selection should not be
taken lightly.
Values of CFM and back pressure for
some of the more popular tubes, sockets
and chimneys are given in Table 6. Back
pressure is specified in inches of water and
Fig. 36 - Air is forced into the chassis by the blower and exits through the tube socket. The can be easily measured as indicated in
manometer is used to measure system back pressure, which is an important factor in determining' Figs. 36 and 37, by means of a
the proper size blower. manometer. A manometer is nothing
8-23 Chapter 8
the unit. As an example, assume that an
ATMOSPHERIC ATMOSPHERIC amplifier is to be built using a 3-1000Z
PRESSURE \
PRESSURE ' \ tube. A blower capable of supplying 25
CFM at a back pressure of 0.38 inches of
MANOMETER MANOMETER water is required. Referring to Table 7 it
appears that the second blower listed
would be suitable, although it may be
PRESSURE marginal since it can only supply 25 CFM
DIFFERENCE
"H 011 into a back pressure of 0.4 inches of
_L_2 water. The next larger size would provide
a margin of safety.
When a pair of tubes is used, the CFM
\CHASSIS
rating is doubled, but the back pressure
BLOWER "OFF" remains the same as 'that for one .tube. A
pair of 3-1000Z tubes, for example, would
require 50 CFM at a back pressure of 0.38
inches of water. In this case the fifth
(A) (8)
blower listed in the Table would be
suitable since it can supply 85 CFM at a
back pressure of 0.4 inches of water.
Fig. 37 - At A the blower is "off" and the water will seek its own level in the manometer. At B Always choose a blower that can supply at
the blower is "on" and the amount of back pressure in terms of inches of water can be measured least the required amount of air. Smaller
as indicated. ' blowers will almost certainly lead to
shortened tube life.
Table 6 Table 6 also contains the part numbers
Specifications of Some Popular Tubes, Sockets and Chimneys for air-system sockets and chimneys to be
used with the tubes that are listed. The
Tube CFM Back Pressure Socket Chimney
(inches)
builder should investigate which of the
3·400Z/8163 13 0.13 SK·400, SK·410 SK·416 sockets listed for the 4CX250R,
3·500Z 13 0.082 SK·400, SK·410 SK·406 4CX300A, 4CXlOOOA and 4CXI500A
3·1000Z/8164 25 0.38 SK·500, SK·510 SK·516 best fits the circuit needs. Some of the
3CX 1500/8877 35 0.41 SK·2200, SK·2210 SK·2216
4·250A/5D22 2 0.1 SK·400, SK·410
sockets have certain tube elements
SK·406
4·400A/8438 14 0.25 SK·400, SK·410 • SK·406 grounded internally through the socket.
4·1000A/8166 20 0.6 SK·500, SK·510 SK·506 Others have elements bypassed to ground
4CX250R/7850W 6.4 0.59 SK·600, SK·600A, SK·602A, SK·606 through capacitors that are integral parts
SK·610, SK·610A, SK·611, SK·626
SK·612, SK·620, SK·620A
of the sockets.
SK·621, SK·630 An efficient blower is required when
4CX300A/8167 7.2 0.58 SK·700, SK·710, SK·711A, SK·606 using the external-anode tubes, such' as
SK·712A, SK·740, SK·760 the 4CX250R. Such tubes represent a
SK·761, SK·770
4CX350Al8321 7.8 1.2 Same as 4CX250R
trade-off which allows high-power opera-
4CX1000A/8168 25 0.2 SK·800B, SK·810B, SK·890B SK·806 tion with a physically small device at the
4CX1500/8660 expense of increased complexity in the
8874 8.6 0.37 cooling system. Other types of external-
These values are for sea·level elevation. For locations well above sea·level (Denver, Colorado, for anode tubes are now being produced for
example), add an additional 20% to the figure listed.
conductive cooling. Electrical insulators
which are also excellent thermal conduc-
tors, such as AISiMg (aluminum-silicon-
Table 7 magnesium compound) and Be
Blower Performance Specifications (beryllium), couple the tube to a heat sink.
Requirements for the heat dissipator are
Wheel Wheel RPM Free Back Pressure (inches) Cutoff Srock calculated in the same way as for power
Dia. Width Air 0.1 0.2 0.3 0.4 0.5 f'Jo.
2" 1" 3160 15 13 4 0,22 2C782 transistors, as outlined below. Similar
3" 1·15/32" 3340 54 48 43 36 25 17 0.67 4C012 tubes are made with special anode struc-
3" 1·718" 3030 60 57 54 49 39 23 0.60 4C440 tures for water or vapor cooling, allowing
3" 1·7/8" 2880 76 70 63 56 45 8 0.55 4C004 high-power operation without producing
3·13/16" 1·718" 2870 100 98 95 90 85 80 0.80 4C443
3·13/16" 2·112" 3160 148 141 135 129 121 114 1.04 4C005 an objectionable noise level from the cool-
ing system.
Transistor Cooling
more than a piece of clear tubing, open at plications a standard ruler can be used for Some bipolar power transistors have
both ends and fashioned in the shape of a the measurement and the results will be the collector connected directly to the case
"U." The manometer is temporarily con- sufficiently accurate. of the device, as the collector must
nected to the chassis and is removed after Table 7 illustrates the performance dissipate most of the heat generated when
the meas'urements are completed. As specifications for one particular brand of the transistor is in operation. Others have
shown in the diagrams, a small amount of blowers. These are Dayton blowers which the emitter connected to the case. How-
water is placed in the tube. At Fig. 37 A, are available through W. W. Grainger ever, even the larger case designs cannot
the blower is "off" and the water will outlets throughout the' U.S. Blowers with . conduct heat away fast enough to keep the
seek its own level. At B, the blower is similar wheel dimensions of different operating temperature of the device
"on" (socket, tube and chimney in place) manufacture likely have similar functioning within the safe area, the
and the pressure difference, in terms of characteristics. If in doubt about maximum temperature that a device can
inches -of water, is measured. For most ap- specifications contact the manufacturer of , stand without damage. Safe area is usually
HF Transmitting 6-24
20
z
0
f= 10 '-...
<.)
w 8 ..........
z> 6
0
<.)
..J
5
4
""'" "-
r--..
'"::>
a:
3
\ r--
>-
'"a:
z
2 ~ Device Case 5W lOW 25W 50W l00W
0 TO-5 17.2 7.2 1.2 0.71
~
u. 0.35
~
'-
<.)
~ 1.0
~ 0.8
"',- TO-44 1.2 9.2
(B)
0~44 nla nla
0.1
1 2 3 4 5 8 10 20 30 50 80 100 200 300 500 1000
VOLUME IN CUBIC INCHES OF FINNED-TYPE HEAT SINK
(A)
Fig. 38 - Heat-sink thermal resistance versus size. The sink volume can be determined by multiplying the height and cross-sectional area. At Bare
numbers which show the approximate thermal resistance needed for operating at various power levels with adequate cooling. TO-5 and TO-44 case
styles are listed.
0.001
specified in a device data sheet, often in 2000V
;hOI
Fig. 39 - Circuit examples of typical single- +700V
ended, parallel and push-pull tube amplifiers.
8-25 Chapter 8
These figures are based on several excitation is keyed in a cw transmitter, Grid- (control or screen). modulated
assumptions, so they can be considered a . Class C operation of subsequent amplifiers output amplifiers in a-m operation run at
worst-case situation. Smaller heat sinks will, under certain conditions, introduce a carrier efficiency of 30 to 35 percent, and
may be usuable. key clicks not present on the keyed excita- a grid-modulated stage with 100 watts
The thermal design of solid-state cir- tion (see chapter on Code Transmission). input has a carrier output of 30 to 35
cuits has been covered in April 1972 QS'{. The peak envelope power (PEP) input or watts, (The PEP output, four times the
The surface contact between the transistor output of any cw (or fm) transmitter is the carrier output, is 120 to 140 watts.)
case and the heat sink is extremely "key-down" input or output, Running the legal input limit in the
important. To keep the sink from being United States, a plate-modulated output
"hot" with dc, a mica insulator is usually A-M stage can deliver a tarrier output of 650 to
employed between the transistor case and In an amplitude-modulated phone trans- 750 watts, while a screen- or control-
the heat dissipator. Newer types of mitter, plate modulation of a Class C out- grid-modulated output amplifier can deli-
transistors have a case mounting bolt put amplifier results in the highest output ver only a carrier of 300 to 350 watts.
insulated from tbe collector so that it may for a given input to the output stage. The
be connected directly to the heat' sink. efficiency is the same as for cw or fm with SSB
the same amplifier, from 65 to 75 , Only linear amplifiers can be used to
Output Power from Transmitters: percent. (In most cases the manufacturer amplify ssb signals without prohibitive
CWorFM rates the maximum allowable input on distortion, and this limits the choice of
In a cw or fm transmitter, any',class of plate-modulated phone at about 2/3 that output amplifier operation to Classes A,
amplifier can 'be used as an output or of cw or fm,) A plate-modulated stage AB" AB2 and B. The efficiency of
intermediate amplifier. (For reasonable running 100 watts input will deliver a operation of these amplifiers runs from
efficiency, a frequency multiplier must be carrier output from 65 to 75 watts, about 20 to 65 percent. In all but Class A
operated Class C.) Class C operation of depending upon the tube, frequency and operation the indicated (by plate-current
the amplifier gives the highest efficiency circuit factor. The PEP output of any a-m meter) input will vary with the signal, and
(65 to 75 percent), but it is likely to be signal .is four times the carrier output it is not possible to talk about relative
accompanied by appre~iable harmonics power, or 260 to 300 watts for the inputs and outputs as readily as it is with
and consequent TVI possibilities, If the lOO-watt input example, other modes. Therefore linear amplifiers
are rated by PEP (input or output) at a
given distortion level, which indicates not
only how much ssb signal they will deliver
27W VI
MAX. 6673
0.001
5000v
but also how effective they will be in
INPUT 500W
amplifying an a-m signal.
~01 Linear Amplifiers for A-M
• In considering the practicality of adding
a linear output amplifier to an existing
0.01
a-m tnmsmitter, it is necessary to know
the carrier output of the a-m transmitter
~ +2200V
and the PEP output rating of the linear
amplifier. Since the PEP output of an a-m
6.3V
t -_ _-o 3 . 2A signal is four times the carrier output, it is
Ip = 22 TO 500mA obvious that a linear with a PEP output
·~0.01 rating of only four times the carrier
CLASS ABi AMP
output of the a-m transmitter is no
(B) amplifier at all, If the linear amplifier has
a PEP output rating of eight times the a-m
transmitter carrier output, the output
0.001
C1 2000V power will be doubled and a 3-dB
I
~r-------61-~-6B--~~--~)~-~ improvement will be obtained. In most
cases a 3-dB change is just discernible by
the receiving operator.
By comparison, a linear amplifier with
OUTPUT a PEP output rating of four times an
i2QW existing ssb, cw or fm transmitter will
quadruple the output, a 6-dB improve-
ment, it should be noted that the linear
amplifier must be rated for the mode (ssb,
cw or fm) with which it is to be used.
Grounded-Grid Amplifiers
The preceding discussion applies to
vacuum-tube amplifiers connected in a
C2
grounded-cathode or grounded-grid cir-
SET FOR 50mA At'"
5 cuit. However, there are a few points that
NO-SIG. Ip NEUT.
V1 V2
apply only to grounded-grid amplifiers.
. O,O:2~ A tube operated in a given Class(AB" B,
~~ 6,3V +750V C) will require more driving power as a
grounded-grid amplifier than as a grounded-
cathode amplifier. This is not because the
CLASS ABI AMP
(0)
grid losses run. higher in the grounded-
grid configuration but because some of
HF Transmitting 6-26
the driving power is coupled directly ended amplifier which operates Class AB2 connected as shown to obtain a l-kW-dc
through the tube and appears in the plate in grounded-grid fashion. TI is a broad- input linear amplifier. Attention must be
load circuit. Provided enough driving band, trifilar transformer which keeps the paid to selecting a set of tubes with nearly
power is available, this increased require- cathode and filaments above ac ground to matched dynamic characteristics. If this is ,
ment is of no concern in cw or linear opera- provide a method for driving the cathode not done, one or more of the tubes may
tion. In a-m operation, however, the fed- of VI. Operating bias is developed by in- draw the major part of the current during
through power prevents the grounded-grid serting DI in the cathode return. ZI is a the driven period. This would cause them
amplifier from being fully modulated (100 vhf parasitic suppressor. RFC2 functions to operate in excess of their safe plate-
percent). as a safety device in the event the plate dissipation ratings and be destroyed.
blof,:king capacitors short and dc flows in- An example of a push-pull amplifier is
Amplifier Circuits: Parallel and to the load (antenna or Transmatch). The offered in Fig. 39D. Two 6146Bs are used
Push-Pull Amplifiers rf choke permits high dc current to flow to in a Class ABI format. CI and C2 are the'
The circuits for parallel-tube amplifiers ground,.blowing the power supply fuses neutralizing capacitors. Join points A-A
are the same as for a single tube, similar and destroying the choke. It should have and B-B to complete the circuit. CI and
terminals of tubes being connected to- an XL which is at least 10 times the load C2 are shown in series with O.OOI-j.lF
gether. The grid impedance of two tubes resistance. blocking capacitors. The latter need not
in parallel is half that ora single tube. This TV sweep tubes used in parallel are seen be used if access to the neutralizing
means that twice the grid tank capacitance at Fig. 39C. ,Each plate lead contains a capacitors can be had without shock
should be used for the same Q. • parasitic choke (ZI and Z2). DI is chosen hazard. The circuit concepts shown in Fig.,
The plate load resistance is halved so to provide the necessary idling current for 39 are applicable to arll types of transmit-'
that the plate-tank capacitance for a single the class of operation desired. The unique ting tubes and power levels. Specific types
tube also should be doubled. The total feature of this circuit is that the control of tubes are shown merely to provide
grid current will be doubled, so to grids of VI and V2 are tied together and practical examples during this treatment.
maintain the I>ame grid bias, the grid-leak driven across a 50-ohm resistor. This
resistance should be half that used for a method eliminates the need for a tuned Grounded-Grid Amplifiers
single tube. The required driving power is circuit or matching transformer at, the Fig. 38A shows the input circuit of a
doubled. The capacitance of a neutraliz- amplifier input. Additionally, by strap- grounded-grid triode amplifier. In con-
ing capacitor ~hould be doubled and the ping the grids to a low impedance (50 figuration it is similar to the conventional
value of the screen dropping resistor should ohms), amplifier stability can be realized grounded-cathode circuit except that the
be cut in half. without the need for neutralization. The grid, instead of the cathode, is at ground
In treating parasitic oscillation, it is 100-ohm resistors in the screen-grid leads potential. An' amplifier of this type is
often necessary to use a choke -in each are used to discourage vhf oscillations and characterized by a comparatively low in-
plate lead, rather than one in the common to help equalize the screen currents of the put impedance and a relatively high driver
lead. This avoids building in a push-pull tubes. Several sweep tubes can be parallel- power requirement. The additional driver
type of vhf resonance, which may cause
inefficient operation at higher frequen-
cies. See Fig. 39C and D.
Two or mtre transistors are often
operated in parallel to achieve high output 3-500Z
power, because several medium-power
. devices often cost less than a single
high-power type. When parallel operation
is used, precautions must be taken to
insure that equal drive is applied to each INPUTCOt--1
C01--1
INPUT
~O,Ot
transistor. Otherwise, one transistor may ~O.01
"hog" most of the drive and exceed its
safe ratings. II
In practice, it is not wise or necessary to
use transistors in parallel. A push-pull cir-
cuit, such as that of Fig. 43 is preferable
O.t=+, O.~
and it tends to cancel even harmonics - a +BIAS
FIL,V
benefit. Alternatively, single-ended
amplifiers can be joined to deliver power (A) (8)
to a single load by means of hybrid com-
3-500Z
biners. This technique was iJJustrated by
Granberg in April and May 1976 QST. He
~B~~~E I 1: ':,Y' . ~
50f\:n MATCHING
used combiners to parallel.the outputs of NETWORK
currents.
Fig. 39B shows the circuit of a single- Fig. 40 - Methods for driving grounded-grid amplifiers,
~
50n
O~ ~
0.1
RFC
3,uH
1.8 -2.0
0.5W
INPUT
•
f:]li
8,uH
10
all)plifiers in amateur work below 30 bifilar-wound filament choke (Fig. 40B four times the 5V-ohm source impedance.
MHz is in the case where the available and C). With this method, a double A 7-I.IH winding satisfies the need at
driving power far exceeds the power that solenoid (often wound on a ferrite core) is 7 MHz. The collector circuit employs a T
can be used in driving a conventional generally used, although separate chokes rietwork to transform the 5.6-ohm collec-
grounded-cathode amplifier. can be used, or a toroid core of large tor to a 50-ohm load. The collector rf
Screen-grid tubes are also used some- cross-sectional area can be used. A typical choke is followed by a second one, which
times in grounded-grid amplifiers. In filament choke is shown in Fig. 41. with the associated bypass capacitors
some cases, the screen is simply connected The input impedance of a grounded- decouples the amplifier from the 13.5-volt
in parallel with the grid and the tube grid power stage is usually between 30 and power supply.
operates as a high- '" triode. In other 150 ohms. A high-C, low-Q pi-section net- A push-pull, broadband, solid-state
cases, the screen is bypassed to ground work can be used to obtain an SWR of 1: 1 . amplifier circuit is seen in Fig. 43. As
and operated at the usual dc potential. between the exciter and the amplifier. shown, it is biased for Class C operation_
Since the screen is still in parallel with the This is shown in Fig'. 40C. The input net- However, if linear amplification was
HF Tran.miHlng 8·28
desired, the center tap of Tl could be
lifted from ground and forward bias ap-
c·
100 0
plied at that point. If that were done, the .-+-~
80 0
to-ohm, base-swamping resistors would ,
be returned to the transformer center tap
instead of being grounded as shown.
T2 of Fig. 43 is a phase-reversal choke
500
40
30 0
o "-
"-
im
,J
~'''''
m·1f'
~,
O>~
1-0
.J
. ~
which places the collectors of Ql and Q2
in the correct phase (180 degrees apart).
T3 is a conventional transformer which
matches the 44-ohm collector-to-collector
u. 20
0.
~
"~
z
15
o "-
0 ",-
1
'" K'4,.I.1 "K
~I1-0
~1-0
i'..
'""'-.
.--1
..•
-BIAS +S.G. +H.V. ;J::,
impedance to a 50-ohm harmonic filter, :;l
10 0
80
"'" '"'4,. j
FL!. The collector coupling capacitors are
pairs of O.I-",F capacitors in parallel. This
.
Q.
50
"- .;>
K'1"
'S-"
O>~ I,
Fig. 45 - Inductive·link output coupling cir·
cuits. .
method wiII pass more current with less
"
"
z 40
1"'"j(:
, 0> 1-0 C1 - Plate tank capaCitor - see text and Fig.
~ 44 for capacitance. .
capacitor heating than would be the case 0>-.&0 } "- L1 - To resonate at operating frequeney with
30
"'Ko>~
+
if only one capacitor were used at each
point in the circuit. Ceramic chip 20 1-
'" C1. See LC chart and inductance formula
in electrical·laws chapter, or use ARRL
capacitors are recommended. Dl and D2
may be added as protection against dc
voltage spikes on the 13.5-V line. Also, if
10
'"
1"'- 1"'-
lightning Calculator.
L2 - Reactance equal to line impedance. See
reactance chart and inductance formula in
electrical·laws chapter, or use ARRL light·
8
the amplifier should break into self- ning Calculator.
6 R - Representing load.
oscillation, the Zener diodes will limit the 1 2 3 4 5 678 10 15 20
performance, and they do not enhance the milliamperes. Select the vertical line cor·
respondlng to the answer obtained. Follow this Parallel-Resonant Tank
generatipn of harmonic currents. The vertical line to the diagonal line for the band in
reason is' that the diodes are not conduct- question, and thence horizontally to the left to The amount of C that wiII give a Q of
ing under normal conditions. They have read the capacitance. For a given ratio of plate to for various ratlos is shown in Fig. 44.
been proven effective as high as 30 MHz, voltage/plate current, doubling the capacitance For a given plate-voltage/plate-current
shown doubles the Q. When a split'stator
and may function satisfactorily into the capacitor is used in a balanced circuit, the ratio, the Q will vary directry as the tank
vhf region. Matching networks and their capacitance of each section may be one half capacitance, twice the capacitance
solutions can be found earlier in this the value given by the chart. doubles the Q, and so on. For the same Q,
chapter. the capacitance of each section of a split-
stator capacitor in a balanced circuit
RF Power-Amplifier Tanks and should be half the value shown.
Coupling for Tubes by the ratio of loaded Q to unloaded Q by These values of capacitance include the
the relationship: output capacitance of the amplifier tube,
Tank Q the input' capacitance of a following
Rf power-amplifiers used in amateur Err. =100 (1 - ~ amplifier tube if it is coupled capacitively,
transmitters are operated under Class C and all other stray capacitances. At the
or AB conditions (see chapter on tube where OL is the loaded Q and Qu is the higher plate-voltagejplate-current ratios,
fundamentals). The main objective, of unloaded Q. the chart may show values of capacitance,
course, is to deliver as much fundamental The Q is' determined (see chapter on for the higher frequencies, smaller than
power as possible into a load, R, without electrical laws and circuits) by the LjC those attainable in practice. In such a
exceeding the tube ratings. The load ratio and the load resistance at which the case, a tank Q higher than lOis
resistance, R, may be in the form of a tube is operated. The tube load resistance unavoidable.
transmission line to an antenna, or the is related, in approximation, to the ratio
Inductive-Link Coupling: Coupling
input circuit of another amplifier. A' of the dc plate voltage to dc plate current
to Flat Coaxial Lines
further objective is to minimize the at which the tube is operated and can be
harmonic energy (always generated by an. computed from When the load R in Fig. 45 is located
amplifier) fed into the load circuit. In for convenience at some distance from the
attaining these objectives, the Q of the amplifier, or when maximum ,harmonic
tank circuit is of importance. When a load Class-A Tube: reduction is desired, it is advisable to feed
is coupled inductively, the Q of the tank the power to the load through a low-
circuit will . have . an effect on the R _ Plate Volts impedance coaxial cable. The shielded
L - 1.3 X Plate Current
coefficient of coupling necessary for construction of the cable prevents radia-
proper loading of the amplifier. In respect tion and makes it possible to install the
Class-B Tube:
to all of these fact0rs, a tank Q of 10 to.20 line in any convenient manner without
is usually considered optimum. A much Plate Volts danger of unwanted coupling to other cir-
lower Q will result in less efficient opera- RL - 1.57 X Plate Current cuits.
tion of the amplifier tube, greater har- If the line is more than a small fraction
monic output, and greater difficulty in of a wavelength long, the load resistance
coupling inductiveiy to a load. A much Class-C Tube: at its output end should be adjusted, by a
higher Q will result in higher tank current matching circuit if necessary, to match the
with increased loss in the tank coil. _ Plate Volts impedance of the cable. This reduces los-
Efficiency of it tank circuit is determined RL - 2 X Plate Current ses in the cable and makes the coupling
8-29 Chapt.r 8
L1
....~~-_--OOUTPUT
+
(Al
+S.G. +H.V.
(A)
PI NETWORK
L1 L2
(B)
PI-L NETWORK
(el
Fig. 47 - Examples of pi (A) and pi·L (B) networks.
input powers up to 2-kW PEP when feed- 2500 3.50 244. 11.010 1115. 4.518
ing 52-75-ohm loads. In obtaining the lar- 2500 4.00 191. 11.010 827. 4.518
2500 7.00 114. 6.033 493. 2.476
ger capacitances required for the lower 2500 7.30 105. 6.033 453. 2.476
frequencies, it is common practice to 2500 14.00 56. 3.069 240. 1.259
switch one or more fixed capacitors in 2500 14.35 53. 3.069 230. 1.259
parallel with the variable air capacitor. 2500 21.00 38. 2.053 158. 0.843
While the voltage rating of a mica or 2500 21.45 036. 2.053 154. 0.843
2500 28.00 29. 1.483 127. 0.609
ceramic capacitor may not be exceeded in 2500 29.70 26. 1.483 111. 0.609
a particular case, capacitors of these types
/ are limited in current-carrying capacity. 3000 3.50 204. 12.903 1042. 4.518
Postage-stamp silver-mica capacitors should 3000 4.00 159. 12.903 777. 4.518
3000 7.00 94. 7.070 468. 2.476
be adequate for amplifier inputs over the 3000 7.30 87. 7.070 426. 2.476
range from about 70 watts at 28 MHz to 3000 14.00 47. 3.597 222. 1.259
400 watts at 14 MHz and lower. The 3000 14.35 44. 3.597 217. 1.259
larger mica capacitors (CM-45 case) hav- 3000 21.00 32. 2.406 146. 0.843
3000 21.45 30. 2.406 145. 0.843
ing voltage ratings of 1100 and 2500 volts 3000 28.00 24. 1.738 115. 0.609
are usually satisfactory for inputs varying 3000 29.70 21. 1.738 105. 0.609
from about 350 watts at 28 MHz to 1 kW
at 14 MHz and lower. Because of these 3500 3.50 174. 14.772 997. 4.518
.3500 4.00 136. 14.772 738. 4.518
current limitations, particularly at the 3500 7.00 81. 8.094 444. 2.476
higher frequencies, it is advisable to use as 3500 7.30 75. 8.094 404. 2.476
large an air capacitor as practicable, using 3500 14.00 40. 4.118 215. 1.259
the micas only at the lower frequencies.
8·31 Chapter 8
Broadcast-receiver replacement-type to emitter or cathode. In general, the best amplifier, it is usually necessary to 19ad
capacitors can be obtained reasonably. arrangement using a tube is one in which the grjd circuit, or to use a neutralizing
Theil: voltage insulation should be ade- the cathode connection to ground, and the circuit.
quate for inputs of 1000 watts or more. plate tank circuit are on the same side of The capacitive neutralizing system for
the chassis or other shielding. The "hot" screen-grid tubes is shown in Fig. 48A. C I
More About Stabilizing Amplifiers. lead from the input tank (or driver plate is the neutralizing capacitor. The capaci-
A straight' amplifier operates with its tank) should be brought to the socket tance should be chosen so that at some ad-
input and output circuits tuned to the through a hole in the shielding. Then justment of CI,
same frequency. Therefore, unless the when the grid tank capacitor or bypass is
coupling between these two circuits is grounded, a return path through the hole
CI Tube grid-plate cal?acitance (or Cgp )
brought to the necessary minimum, the to cathode will be encouraged, since
amplifier will osCillate as a tuned-plate, transmission-line characteristics are simu- C3 Tube input capacitance (or C 1N )
tuned-grid circuit. Care should be used in lated.
arranging components and wiring of the The g,id-cathode capacitance must in-
two circuits so that there will be negligible Screen-Grid Tube Neutralizing Circuits clude all strays directly across the tube
opportunity for coupling external to the The plate-grid capacitance of screen- capacitance, including the capacitance of
tube or transistor itself. Complete shiel- grid tubes is reduced to a fraction of a the tuning-capacitor stator to ground.
ding between input and output circuits picofarad by the. interposed grounded . This may amount to 5 to 20 pF. In the
usually is required. All rf leads should be screen. Nevertheless, the power sensitivity case of capacitance coupling, the output
kept as short as possible, and particular of these tubes is so great that only a very capacitance of the driver tube must be
attention should be paid to the rf return small amount of feedback is necessary to added to the grid-cathode capacitance of
paths from input and output tank circuits start oscillation. To assure a' stable the amplifier in arriving at the value of
CI.
Neutralizing a Screen-Grid
Amplifier Stage
There are two general procedures
lin Freq. CI L1 C2 L2 available for indicating neutralization in a
(Ohms) (MHz) (pF) (",H) (pF) (",H) screen-grid amplifier stage. If the screen-
3500 14.35 38. 4.118 206. 1.259 grid tube is operated with or without grid
3500 21.00 27. 2.755 136. 0.843 current, a sensitive output indicator can
3500 21.45 25. 2.755 138. 0.843
3500 28.00 21. 1.989 106.
be used. If the screen-grid tube is operated
0.609
3500 29.70 18. 1.989 99. 0.609 with grid current, the grid-current reading
can be used as an indication of neutraliza-
4000 3.50 153. 16.621 947. 4.518 tion. When the output indicator is used,
4000 4.00 119. 16.621 706. 4.518 both screen and plate voltages must be re-
4000 7.00 71. 9.107 418. 2.476
4000 7.30 65. 9.107 387. 2.476
moved from the tubes, but the de circuits
4000 14.00 35. 4.633 204. 1.259 from the plate and screen to cathode must
4000 14.35 33. 4.633 197. 1.259 be completed. If the grid-current reading
4000 21.00 23. 3.099 137.. 0.843 is used, the plate voltage may remain on .
4000 21.45 22. 3.099 132. 0.843
4000 28.00 18. 2.238 107. 0.609 but the screen voltage must be zero, with
4000 29.70 16. 2.238 95. 0.609 the de circuit completed between screen
and cathode.
5000 3.50 123. 20.272 872. 4.518
5000 4.00 95.
The immediate objective of the neutrali-
20.272 658. 4.518
5000 7.00 57. 11.108 387. 2.476 zing process is reducing to a minimum the
5000 7.30 52. 11.108 360. 2.476 rf-driver voltage fed from the input of the
5000 14.00 29. 5.651 186. 1.259 amplifier to its output circuit through the
5000 14.35 27. 5.651 183. 1.259 grid-plate capacitance of the tube. This is
5000 21.00 19. 3.780 125. 0.843
5000 21.45 18. 3.780 123. 0.843
done by adjusting carefully, bit by bit, the
5000 28.00 15. 2.730 95. 0.609 neutralizing capacitor or link coils until
5000 29.70 13. 2.730 89. 0.609 an rf indicator in the output Circuit reads
minimum, or the reaction.ofthe unloaded
6000 3.50 103. 23.873 829. 4.518
6000 4.00 80. 23.873 621. 4.518
plate-circuit tuning on the grid-current
6000 7.00 48. 13.081 368. 2.476 value is minimized.
6000 7.30 44. '13.081 340. 2.476 The wavemeter shown in the Measure-
6000 14.00 24. 6.655 172. 1.259 ments chapter makes a sensitive neutrali-
6000 14.35 22. 6.655 173. 1.259
6000 21.00 16. 4.452 117. 0.843
zirtg indicator. The wavemeter coil should
6000 21.45 15. 4.452 116. 0.843 be coupled to the output tank coil at the
6000 28.00 13. 3.215 87. 0.609 low-potential or "ground" point. Care
6000 29.70 11. 3.215 84. 0.609 should be taken to make sure that the
8000 3.50 30.967
coupling is loose enough at all times to
78. 747. 4.518
8000 4.00 60. 30.967 569. 4.518 prevent burning out the meter or the rec-
8000 7.00 36. 16.968 337. 2.476 tifier. The plate tank capacitor should be
8000 7.30 33. 16.968 312. 2.476 readjusted for maximum reading after
8000 14.00 18. 8.632 165. 1.259
8000 14.35 17. 8.632 159. 1.259
each change in neutralizing.
8000 21.00 12. 5.775 104. 0.843 When the grid-current meter is used as
8000 21,45 11. 5.775 106. 0.843 a neutralizing indicator, the screen should
8000 28.00 9. 4.171 86. 0.609 be grounded for rf and dc, as mentioned
8000 29.70 8. 4.171 71. 0.609 above. There will be a change in grid
Operating a - 12. Output load - 52 ohms. Computer data provided by Bill Imamura, JA6GW.
current as the unloaded plate tank circuit
is tuned through resonance. The neutrali-
HF TransmlHlng 8-32
zing capacitor (or inductor) should be ad-
justed until this deflection is brought to a
minimum. As a final adjustment, screen
voltage should be returned and the
neutralizing adjustment continued to the
point where minimum plate current,
maximum grid current and maximum
screen current occur simultaneously. An
increase in grid current when the plate
tank circuit is tuned slightly on the
high-frequency side of resonance indicates
that the neutralizing capacitance is too
small. If the increase is on the low-
frequency side, the neutralizing capaci-.
Cl
tance is too large. When neutralization is
complete, there should be a slight decrease
in grid current on either side of resonance.
Grid Loading
The use of a neutralizing circuit may -v C2
L ______________ ~- _" _________ _
. often be avoided by loading the grid
circuit if the driving st~e has some power
capability to spare. Loading by tapping +v
(A)
the grid down on the grid tank coil (or the
plate tank coil of the driver in the case of
capacitive coupling), or by a resistor from
grid to cathode is effective in stabilizing
an amplifier.
Low-Frequency Parasitic Oscillation
The screening of most transmitting
screen-grid tubes is sufficient to prevent
low-frequency parasitic oscillation caused
by resonant circuits set up by rf chokes in (Bl
grid and plate circuits. When rf chokes are
used in both grid and plate circuits of a
triode amplifier, the split-stator tank Fig. 48 - A neutralization circuit may use either C 1 or C2 to cancel the effect of the tube grid-plate
capacitance (A). The circuit at.B shows the usual vhf-parasitic circuit in bold lines.
capacitors combine with the rf chokes to
form a low-frequency parasitic circuit, un-
less the amplifier circuit is arranged to
Iprevent it. Often, a resistor is substituted
for the grid rf choke, which will produce
Table'11
the desired result. This resistance should
be a least 100 ohms. If any grid-leak resis- Typical Tank·Capacitor Plate Spacings
tance is used for biasing, it should be sub- Spacing Peak Spacing Peak Spacing Peak
stituted for the lOO-ohm resistor. Inches (mm) Voltage Inches (mm) . Voltage Inches (mm) Voltage
0.015 (0.4) 1000 0.D7 (1.8) 3000 0.175 (4.4) 7000
Component Ratings: Output· Tank 0.02 (0.5) 1200 0.08 (2.8) 3500 . 0.25 (6.3) 9000
Capacitor Voltage 0.03 (0.8) 1500 0.125 (3.0) 4500 0.35 (8.9) 11000
0.05 (1.3) 2000 . 0.15 (3.8) 6000 0.5 (12.7) 13000
In selecting a tank capacitor with a
spacing between plates sufficient to
prevent voltage breakdown, the peak rf
voltage across a tank circuit under load,
but without modulation, may be taken between plates. Typical plate spacings are capacitor shaft and the dial. The section of
conservatively as equal to the dc plate or shown in Table II. the shaft attached to the dial should be well
collector voltage. If the dc supply voltage Output tank capacitors should be grounded. This can be done conveniently
also appears across the tank capacitor, mounted as close to the tube as tem- through the use of panel shaft-bearing units.
this must be added to the peak rf voltage, perature considerations will permit, to
making the total peak voltage twice the dc make possible the shortest capacitive path Tank Coils
supply voltage. If the amplifier is to be from plate to cathode. Especially at the Tank coils should be mounted at least
plate-modulated, this last value must be higher frequencies, where minimum cir- their .diameter away from shielding to
doubled to make it four times the dc plate cuit capacitance becomes important" the prevent a marked loss. in Q. Except
voltage, because both dc and rf voltages capacitor should be mounted with its perhaps at 28 MHz it is not important
double with lOO-percent.amplitude modu- stator plates well spaced from the chassis that the coil be mounted quite close to the
lation. At the higher voltages, it is desirable or other shielding. In circuits where the tank capacitor. Leads up to 6 or 8 inches
to choose a tank circuit in which the dc rotor must be insulated froni ground, the are permissible. It is more important to
and modulation voltages do not appear capacitor should be mounted on ceramic keep the tank capacitor as well as other
across the tank capacitor, to permit the insulators of size commensurate with the components out of the immediate field of
use of a smaller capacitor with less plate plate voltage involved and - most the coil. For this reason, it is preferable to
spacing. important of all, from the viewpoint of mount the coil so that its axis is parallel to
Capacitor manufacturers usually rate safety to the operator - a well-insulated the capacitor shaft, either alongside the
their products in terms of the peak voltage coupling should be used between the capacitor or above it.
-8·33 Chapter 8
• often employed, as the small inductance . Circuit Description
values required for a lumped inductapce
Table 12
become difficult to fabricate. The schematic diagram of the transmit-
Wir. Sizes for Transmitting Coils for Tube ter is displayed at f'ig. 50. QI and
Transmitters RFChokes associated components comprise a
Power Colpitts variable-frequency crystal
Input (Watts) Band (MHz) Wire Size The characteristics of any rf choke will
vary with frequency, from characteristics oscillator. C I is used to adjust the fre-
1000 28-2 6
14-7 8 resembling those of a parallel-resonant quency of the oscillator and C2 is used to
3.5-1.8 10 circuit of high impedance, to those of a limit the span of the oscillator. If no limit
500 28-21 8 series-resonant circuit, where the impe- is provided the oscillator can operate "on
14-7 12 dance is lowest. In between these extremes, its own" and no longer be under the con-
3.5-1.8 14
the choke will show varying amounts of trol of the crystal - an undesirable cir-
150 28-21 12 cumstance. Supply voltage is fed to the
14-7 14
inductive or capacitive reactance.
3.5-1.8 . 18 In series-feed circuits, these characteris- oscillator only during transmit and spot
75 28-21 14 tics are of relatively small importance be- periods. This prevents the oscillator from
14-7 18 cause the rf voltage across the choke is interfering with received stations
3.5-1.8 2~ negligible. In a parallel-feed circuit, how- operating on the same frequency.
25 or less' 28-21 18 ever, the choke is shunted across the tank Output energy from the oscillator is
14-7 24 circuit, and is subject to the full tank rf routed to Q2, a grounded-base amplifier.
3.5-1.8 28
voltage. If the choke does not present a This stage provides some gain, but more
'Wire size limited principally by consideration of Q. sufficiently high impedance, enough power important, a high degree of isolation be-
will be absorbed by the choke to cause it tween the oscillator and the driver stage.
to burn out. Pulling and chirp are virtually non-
To avoid this, the choke must have a existent.
sufficiently high reactance to be effective The driver stage uses a broadband
There are many factors that must be at the lowest frequency, and yet have no amplifier that operates Class A. Keying of
taken into consideration in determining series resonances near the higher-frequency this stage is accomplished by ungrounding
the size of wire (see table 4) which should bands. the base and emitter resistors. CIO is used
be used in winding a tank coil. The con- to shape the waveform. Although the key-
siderations' of form factor and wire size A VXO-Controlled Transmitter for 3.5 to ing is rather hard there is no evidence of
21 MHz clicks.
which will produce a coil of minimum loss
are often of less importance in practice than The variable-crystal oscillator (VXO) Two MRF472 transistors are used in
the coil size that will fit into available represents a convenient method for pl;lrallel for the power amplifier. These
space or that will handle the required power generating a highly stable, adjustable- transistors were designed for the Citizen
without excessive heating. this is par- frequency signal. With the circuit shown Band service and work quite nicely at 14
ticularly true in the case of screen-grid in Fig. 50 the frequency spans shown in MHz. Each transistor is rated for 4-watts
tubes where the relatively small driving Table II can be realized. This circuit output which gives a margin of safety
power required can be easily obtained makes use of fundamental-type crystals when operated at the 6-watt output level.
even if the losses in the driver are quite only and is therefore limited to frequen- L2 is used as a dc ground for the bases
high. It may be cohsidered preferable to cies below roughly 25 MHz. Most crystals making the transistors operate Class C.
take the power loss if the physical size of produced for frequencies above 25 MHz The low output impedance at the collec-
the exciter can be kept down by making are overtone types and will not work in tors is stepped up by a broadband
the coils small. this circuit. Since the operator is apt to transformer to the 50-ohm impedance
Transistor output circuits operate at favor one portion of a band over another, level. A five-pole Chebyshev filter is used
relatively low impedances because the a few crystals are all that is necessary for to assure a dean output signal. This
current is quite high. Coils should be coverage of a part of the band. The transmitter exceeds current FCC specifica-
made of heavy wire or strap, with transmitter, as shown, is rated for 6 watts tions regarding spectral purity. D2is used
connections made for the lowest possible of output while running at the lO-watt dc to clamp the collector voltage waveform
resistance. At vhf, stripline techniques are input level. .should the transmitter be operated into an
Fig. 49 - Photographs of the completed 6-watt, VXO-controlled transmitter. Miniature coaxial cable (RG-174/U) is used for connections between the
circuit board, connectors and switches for all runs carrying rf energy. .
HF Transmitting 6-34
open circuit or high SWR antenna system. a dead short to a w.ide-open load will
The transmitter is designed to operate into cause no damage to a VMOS power FET.
a load that is close to 50 ohms resistive. Damage can occur as a result of excessive
SI is used as the transmit/receive operating voltages on the gate or drain,
switch. One section transfers the antenna just as when using too much voltage on
to an accompanying receiver or the output the grids or plate of a vacuum tube. The
of the transmitter strip. Another section is same rules apply to drain current (ex-
used to activate the VXO during transmit cessive dissipation) verSus plate current.
and the third section is provided for Finally, excessive heat will ruin a tran-
receiver muting purposes. sistor as it will when tubes are permitted to
run too hot.
Construction It is convenient to regard a VMOS
The majority of the circuit components device in the same fashion as we might
are mounted on a double-sided, printed- perceive a t1'iode vacuum tube. The major Fig. 51 - Spectral display of the VXO·
circuit boards. One side of the board is difference is that VMOS transistors use controlled transmitter. Here the transmitter is
etched with the circuit pattern and the lower operating voltages than tubes do, . operated in the 20·meter band. The second har·
monic is dOwn 56 dB relative to the fundamen·
other side is left as a ground plane. A and the input and output impedances are tal output. Similar presentations were obtained
small amount of copper is removed from low, comparatively speaking. One can on ea.ch of the other. bands. This transmitter
around each hole on the ground-plane even dip the drain current during tuneup, complies with the current FCC specifications '
side of the board to prevent the leads from and monitor the current by means of a regarding spectral purify.
shorting to ground. A scale etching pat- drain meter.
tern and parts layout guide is shown in
Fig. 52.
Affixed to the front panel are the
transmit/receive switch, spot switch, and
the tuning capacitor. The rear apron sup-
ports the antenna and mute jacks, key
jack and binding posts.
a homemade cabinet measuring 3 X 6
X 8-114 inches (76 X 150 X 210 mm)
was used in the construction of this
transmitter. The builder may elect to build
his or her own cabinet from sheet
aluminum or circuit-board material. The
layout is not critical except that the lead
R3
from the circuit board to C1 should be 220 R8
kept as short as possible - an inch or two ~ 3900
Dl/1W
(25 to 51 mm) is fine.
The final transistors are heat sinked to
the ground plane of the circuit board vxo C!!I AMP
using mica washers and silicone com-
R2
pound. With the normal transmit duty cy- .osth
cle this heat sink is sufficient. If for some
reason the prospective builder plans long
key-down periods, an additional heat sink
connected to the tops of the transistors
would be helpful.
The only adjustment needed is that of
setting the VXO limit capacitor. This can,
be done with the aid of a receiver. Using a
21-MHz fundamental-type crystal, adjust
the capacitor for a maximum frequency
spread of approximately 1O~ 15 kHz. If too
much frequency spread is available,. in-
crease the amount of capacitance. Make a
final check with the receiver by listening
to the keyed signal from the transmitter.
It should be steady and chirp free. The
spectral display of the transmitter is
shown in Fig. 51.
Low-Power VMOS Transmitter
for 3.5 to 28 MHz
Fig. 50 - Schematic diagram of the VXO-controlled transmitter. All resistors are 11~-watt carbon
Power FETs have a distinct advantage types unless noted otherwise. All resistors are mylar or disc ceramic unless. otherWise no~ed •.
over power bipolar transistors: They are Polarized capaCitors are electrolytiC or tantalum. 04 and 05 must be heat slnked to the CirCUit
virtually destruct-proof when handled board. .
correctly. They do not go into thermal ce,
C1, C2, C3, C4, C17 and C18 .... See Table 11. M1 - 0·1 mA meter, Calectro 01·91Z or equiv.
runaway, are not subject to secondary 01 - Zener diode, 9.1 V, 1 W. . 01,02,03 - transistor, 2N2222A or equiv.
breakdown, and are immune to the poten- 02 - Zener diode, 36 V, 1 W. 03 - TranSistor, 2N3866 or equiv.
J1, J2 - Binding post. 04,05 - Transistor, MRF·472 or equiv.
tially damaging effects of operating into J3 - Key jack. R14 . RH, incl. - Fixed resistor, 1.8 ohms,
incorrect loads. Therefore, anything from L1, L3, L4, L5 - See Table 11. 112 W:
8-35 Chapt... 8
POLY = POLYSTYRENE' K = CATHODE SM = SILVER MICA
Fig. 52 - Scale etching pattern and parts layout guide for the 6-watt transmitter. Gray areas represent unetched·copper. One side of the board is a
complete ground plane.
Transmitt,r Circuit
A block diagram of the transmitter is
provided in Fig. 54. It shows the switching
52
SPOT
method used to change from the transmit
to receive modes, plus the keying circuit.
J2
------------------------.-----.-----~--~JV~~----------~+12.5V Q5 and the related components are built
J3 DC as a separate module. In this model a
C14+ CURRENT ,.r:;<.- small piece of circuit board is used. It has
a group of 1/4-inch (6.3 mm) squares cut
rh2LF O-lmA
RX MUTE
~~
25V S38 into the copper-clad side by means of a
hacksaw. The pads serve as solder ter-
CI5 minals. The two resistors and three
rJj68 TX
capacitors in the base lead of Q5 are as
key-shaping components.
q
SIB
C T3
ANT. Fig. 55 show~ the interior of the
PA J5. transmitter. The keying transistor is at the
RX upper right, the P A module is just below
RX it and the oscillator/doubler pc board is at
ANT. the left of the first two. Although Fig. 54
J6
indicates that S4B is used as a receiver-
muting switch, it has not been wired into
the unit shown, and no muting jack has
been included on the back panel. The
V-shaped main chassis measures 5 X 7 X
2 inches (127 X 178 X 51 mm), the width
being the larger dimension. The chassis
R18 - Meter shunt, 13-1/2 inches no. 26 T1, T2 - Broadband transformer, 10 turns no.
enamel wire wound on a high value 1-watt 24 enameled wire. bifilar wound on an
resistor. FT-37-43 core.
S1 - Toggle switch, 3pdt. T3 - Broadband transformer, 10 turns no. 24 Fig. 53 - Exterior of the W1 FB VMOS
S2 -.Push-button switch, spst, normally open. enameled wire on an FT-50-43 core. transmitter. The basics of this unit were
S3 - Toggle switCh, dpdt. i Y1 - See Table 11. treated in May 1979 QST.
HF Tr.namHtlng 8·38
. mounted on the rear lip of the chassis.
Four enhancement-mode FETs are used
osc. in the transmitter. Ql, Q2 and Q3 (Fig.
56) are Siliconix VN66AK devices in
r-------~DrY~1----~
* - 3/4" DIA. SINK
TO-39 cases. Supertex VNOI06N-2 VMOS
r-----~DrY~2----~ RI- SELECT FOR METER USED
FETs are suitable as direct substitutes. I
PUSH - PUSH
ARROWS INDICATE DIRECTION
Crown heat sinks are required on all three
Y3 DBLR.
D OF SIGNAL AND VOLTAGE transistors (Thermalloy 2215B or equiv.).
To ensure operation as an oscillator, Ql
Dr-
Y4
51
Pierce oscillator is used at Ql, with one
winding of broadband transformer Tl
being the drain impedance. Cl, C2 and C3
are feedback capacitors. Source capacitor
Cs controls the amount of degenerative
feedback in the oscillator. It mayor may
not be required. This depends upon the
type of crystals used at YI-Y4 (crystal ac-
tivity). If low oscillator output is noted, or
RCVR.
RCA if a chirpy cw note is obtained, Cs should
~3
2N4037
be included. Use only that amount of
capacitance which will ensure proper
oscillation. Typical values range from 10
pF.to 100 pF.
Tl supplies drive to push-push doubler
52
OPR. Q2/Q3 at a phase difference of 180 0 • The
drains of the doubler are in parallel and
~TUNE tapped toward the B + end of Ll to effect
an impedance match between the doubler
ZERO BEA~ S5A
and the gate of the PA transistor. Fig. 57
shows the board pattern and parts layout
to scale.
Circuit details for the P A stage are
Fig. 54 - Block diagramof the VM05 transmitter. The arrows indicate the direction of dc and the given in Fig. 58. A Siliconix VN67 AJ or
signal. 05 is the'keylng switch. It may be eliminated by keying the B + line to 01 directly. J1·J4, Supertex VN1206N-l (TO-3 case) is
Incl. are single-hole mount phono jacks. J5 is a two-circuit phone jack. M1 is a 500·mA instru·
ment. A 0-1 mA meter can be used by adding a suitable shunt at R1. A 5impson no. 2121 (1-1/2 employed at Q4. It is insulated from its
inch -.38 mm) diameter meter is seen in Fig. 53. 52-55, incl. are miniature toggle switches. homemade U-shaped heat sink (l X 2 X
1-114 inch - 25 X 51 X 32 mm) by
means of a standard TO-3 size mica
wafer. The Q4 mounting screws must pass
through holes large enough to prevent
contact between the transistor case (com-
mon to the drain), the heat sink and cir-
cuit ground. Final checks should be made
with an ohmmeter before applying power.
Silicone grease is spread on both sides of
the mica washer before assembly, plus on
the bottom surface of the heat sink.
A half-wave low-pass filter is used at
the P A input to suppress harmonic cur-
rents in the drive to Q4. It is designed for
a bilateral impedance of 150 ohms and a
QL of 1. The 35 pF input capacitance
(Ciss ) of Q4 is absorbed into the network
at C7.
RFCI-RFC4 of Fig. 56 and RFC6 of
Fig. 58 are 950-mu miniature ferrite
beads. They are necessary to damp vhf
parasitic oscillations. Drive is applied to
the P A gate across a 150-ohm resistor if
Class C operation is desired. For Class AB
use, RI is added and the gate-to-ground
resistor is changed to 160 O.
8·37 CMpter 8
OSC. PUSH - PUSH DBLR
Cl EXCEPT AS INDICATED, DECIMAL VALUES OF
1.75-14.25 MHz CAPACITANCE ARE IN MICROFARADS ( JlF ) ; 3.5-29 MHz
VN66AK,
OTHERS ARE IR' PICOFARADS ( pF OR JlJIF);
VN0106N-2
150/S.M. RESISTANCES ARE IN OHMS; C4
560k k.1 DOD, M' I 000 000.
02
*
SI
s~
~D
18
S.M. rJ,0.OI
01,02,03
BOTT. VIEW
*- HEAT SINK
TO J5 • - PHASING +24V
(IBOmA)
S.M. - SILVER MICA
TO 83
Fig. 56 - Schematic diagram of the VMOS exciter strip. Fixed·value capacitors are disc ceramic unless otherwise noted. Resistors are 1/2·W com·
position unless indicated otherwise.
C1·C3, incl. - Silver·mica feedback capacitor. RFC1·RFC4, incl. - Miniature 950·mu ferrite
Cs - See text. bead by Amidon Assoc.
C4 - Miniature air variable, 100 pF (Hammar· RFC5 - 10 turns no. 20 enam. wire on Amidon Y1·Y4, incl - Fundamental crystal at one half
lund MAPC·100·B or equiv.) Arco 424 mica T50·43 (950 mu, 0.5·in. diam.) ferrite toroid. the desired operating frequency. Sockets are
trimmer can be used. If trimmer is used, S1 - Single·pole, 4·position, single·wafer F-605 pc mount. These and the crystals are
mount it on the pc board. phenolic switch. type GP, 30 pF load capacitance in HG-6/U
L 1, L2 - See Table 14. T1 - See Table 14. type cases.
Fig. 57 - Scale pattern and layout of the exciter board. Parts marked with an asterisk (0) are mounted on the etched side of the board.
The P A drain is tapped near the B + oscillograph is seen at Fig. 59. It was for the various component values are pro-
end of L5 to provi~e an impedance match taken while using an HP analyzer. Q4 was vided in Table 10.
to the load. FL2 removes excessive har- operating in Class AB, which provides Cl2 of Fig. 58 must be added between
monic currents from the transmitter out- somewhat better spectral purity than is the rf modules if Class AB operation is
put, making the spectral purity con- obtained under Class C conditions. planned. It prevents the gate bias from
siderably better than that required by pre- Ll, L2, L5, L6, FLl and FL2 must be being short-circuited to ground.
sent. FCC regulations. A spectral built for the band of operation. Details If a VN67 AJ is not readily available for
HF Transmitting 8·38
PA 3.5-29 MHz
FL2
L8 L9
Cl2
TO ,') ~"
L2 cS 0.01
O.SW
INPUT 0.001
ISO Il
L7
,Ii EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPAC ITANCE ARE
IN MICROFARADS (JIF l ; OTHERS
ARE IN' PICQFARADS (pF OR JlJIF);
0.1 TO-SSB RESISTANCES ARE I N OHMS;
i1
f---1t'---o+24 V k -IOOO.M-I 000000
, (330mAl
S.M. =SILVER MICA
GAIN:::::;12dB , 22:F * ON ETCHED SIDE OF BOARD
50V ** LA'RGE HEAT SINK
Fig. 58 - Schematic diagram of the VMOS power amplifier, Fixed·value capacitors are disc ceramic unless otherWise noted. Resistors are 1·watt
composition.
C8 - 50·pF miniature air variable (Hammar· board. See Table 10. L7 - 10 turns no. 20 enam. wire on Amidon
lund MAPC·50·B or equiv.). Arco 424 mica C5, C6, C7, C9, C10, C11 ~ See Table 10. FT·50·43 ferrite toroid (0.5·in, dia, 950 mu).
trimmer suitable. If used, mount on pc L3, L4, L5, L6, L8, L9 - See Table 10, RFC6 - Miniature 950·mu ferrite bead.
8-31 Chapter 8
The front and rear panels of the trans-
Table 13 mitter are sprayed with Aerosol-can
enamel paint of the builder's choice. The
Component Values for the VXO·Controlled, 6·Watt Transmitter
aluminum surfaces are first abraded with
C3, C17, L3, VXO sandpaper to aid the paint in adhering ,to
Band C1 C2 C4 C6 C18 L1 L5 L4 Range the metal surfaces. Soap and hot water
80 M 365 . 220 100 820 52 Turns
T·37·2
29 Turns
T·502
35 Turns
' T·502
3·5 kHz should be applied liberally to the metal
surfaces prior to painting. Moisture and
40 M 365 100 470 40 Turns 21 Turns 25 Turns 6·8 kHz
T·37·2 T·50·2 T·50·2
finger marks should not exist on the metal
30 M 150 68 50 330 30 Turns 18 Turns 22 Turns 8·10 kHz to be painted. Dymo tape labels are used
(10 MHz) T·37·2 T·50·2 T·50·2 to identify the controls. The tape should
20 M 50 10 50 50 240 35 Turns 16 Turns 19 Turns 10·12 kHz be of a color that matches the painted sur-
T·37·6 T·50·6 T·50-6
15 M 50 10 33 33 150 27 Turns 15 Turns 18 Turns 12·14 kHz
faces, thereby imparting a professional
T·37·6 T·50-6 T·50·6 decor.
The top cover can be made of solid
aluminum if desired. Metal L brackets can
be mounted on the right and left edges of
the chassis bottom to permit affixing the
cover at two points. No. 6 sheet-metal
serves as a ground plane to reduce un- vides clearance between the chassis and screws are suitable for the purpose.
wanted rf ground loops - an aid to. the etched sides of the boards. It also en-
stability. 2 sures a secure grounding method for the Operation
ground foils on the boards. Checkout should include .visual inspec-
Assembly Notes All if leads between the boards and tion for unwanted solder bridges on the pc
The pc boards are-supported above the their related panel controls are shielded. boards. Check also to make certain that
chassis on metal standoff posts. This pro- Miniature RG-174/U 50-0 coax cable was no poor solder joints exist. Finally, test
used in the version described here, The the B + lines by means of an ohmmeter to
'Boards, negatives and parts kits are available from
Circuit Board Specialists, P. O. Box 969, Pueblo, shield braids are grounded at each end of ensure that no short circuits are present.
CO 81002. Tel. 213·722·2000. each cable. A 24-volt regulated power supply is re-
Table 14
Coil and Capacitor Data for the 5·Band VMOS Power FET Transmitter
All cores other than for T1 are powdered iron. Cores and ferrite beads for this circuit were furnished via courtesy of Amidon Assoc., N. Hollywood,
CA 12033.
HF Trans,mlttlng 6·40
.t
qui red for .this circuit. It should be negative feedback voltage through a
capable of delivering at least 600 rnA of separate one-turn winding. The charac-
current continuously. It needs to be well teristic impedance of ac and bd (windings
filtered so that a T9 cw note will result. on T2) should be approximately equal to
Positive keying is used if a solid-state the collector-to-collector impedance of Q I
keyer is employed. A bug or straight key is and Q2, but it is not critical. For physical
suitable if a keyer is not desired. convenience a bifilar winding is used. The
Tuneup entails adjusting the doubler center-tap of T2 is really bc, but for stabi-
tank for a dip in drain current. Alter- lization purposes, band c are separated by
natively, C4 can adjust for maximum rf chokes which are bypassed individually.
drain current (optimum drive) at Q4.
Next, C8 is adjusted for a dip in drain cur-
Construction
rent at Q4. . A scale pc-board template is shown in
This transmitter is designed to work in- Fig. 64. Double-sided board is used, and
to a 50-0 load. If the antenna SWR is patterns are shown for each side of the
much above 1.5: 1, a Transmatch is re- board.
commended to assure that FL2 looks into All parts are mounted on the circuit
the required 50 O. A proper match will board. The MRF454s are soldered to the
result in maximum power transfer to the board, but the flanges are bolted tightly to
antenna. Output from this transmitter will Fig. 61 - The completed 140·watt amplifier the heat sink. Apply a thin layer of silicone
vary from 6 to 7 watts in Class AB (5 to 6 board and heat sink. 01 is mounted on the
underside of the circuit boar~ sandwiched be· grease to Ql, Q2 and Dl where they will
watts for Class C). Amplifier efficiency tween the board and the heat sink. contact the heat sink. This aids in efficient
for Class C operation is roughly 85 per- heat transfer. The board is attached to the
cent, whereas an efficiency of 79 percent heat sink by means of five screws. They
was observed during Class AB conditions. are tapped into the heat sink: Four are at
The standing drain current of Q4 is ap- the frequency range this amplifier is the board corners, and the fifth goes
proximately 40 rnA for Class AB service. designed for. An input frequency-correction through the board, through Dl, and into
for Class C use it is approximately zero. network is ·employed to give a gain flat- the heat sink.
ness response of better than I dB across A large.heat sink is required fo~ proper
A 140-W Solid-State Linear Amplifier the band. Rl, R2 and the associated 5600- dissipation of heat. A nine-inch (229-mm)
Solid-state technology has progressed pF shunt capacitors comprise the input- length of Thermalloy 6153 or a seven-inch
to the point where devices capable of correction network. With the negative (178-mm) length of Aavid Engineering
producing up to l00-watts output in the hf feedback applied via R3 and R4 through 60140 extrusion is suitable for l00-percent
. range are available at reasonable cost. L5, the equivalent of an attenuator is duty cycle operation (such as RITY or
This unit, designed by K7ES, applies the formed with frequency-sensitive charac- SSTV operation).
aforementioned technology to a broad- teristics. At 30 MHz the input power loss
band linear-amplifier design. Power out- is 1 to 2 dB, increasing to 10 to 12 dB at Tun, Up and Operation
put is 140 watts from 1.6 to 30 MHz. The 1.6 MHz. The gain variation of the rf tran- Since this is a broadband amplifier, no
amplifier described here originally ap- sistors is the reciprocal of this, resulting in tuning adjustments are necessary. One
peared in QST for June and July 1977. a gain flatness of approximately 1 dB. In- need only set the bias control so that the
put VSWR is better than 1.75 across the amplifier draws 200 rnA of quiescent
Th, Circuit band. current. A regulated power supply capable
The MRF454 transistor is specified for The input transformer, n, and the of delivering 13.6 volts dc at 25 A is re-
a continuous power output of 70 watts. output transformer, T3, are of the same _quired for amplifier operation.
The maximum continuous current allow- type. The low-impedance windings consist If direct operation into an antenna is
able is 15 A. Maximum dissipation is of two pieces of metal tubing which are expected, filtering of the output is neces-
limited to 250 watts. The minimum shorted electrically at one -end - the sary to meet FCC regulations for spectral
recommended collector idling current is opposite ends being the connections of the purity. The filters shown in Fig. 65 will
100 rnA per device. windings (Fig. 63). The multiturn, high- provide more than sufficient harmonic at-
Fig. 62 shows the basic circuit. The bias impedance windings are wound through tenuation with less than 1 dB of loss at the
supply uses active components rather than the tubing so that the low- and high- operating frequency.
the common diode clampmg scheme. impedance winding connections are at Collector efficiency is in the vicinity of
Some advantages of this system are opposite ends ofthe transformer. Alterna- 50 percent. IMD products are typically 32
line-voltage regulation capability, low tively, copper braid can be'substituted for to 34 dB below PEP. Power gain is
standby current, and a wide range of the metal tubing. The braid can be taken approximately IS dB, which means that
voltage adjustment. With the component from the coaxial cable of the proper 'approximately 5 watts will drive the
values shown the bias is adjustable from diameter. This is exemplified in Fig. 63B. amplifier to the rated output of 140 watts.
0.5 to 0.9 volt. The O.5-ohm resistor The primary and secondary coupling The spectral displays for harmonics and
between pins 1 and 10 of the Me 1723 sets coefficient is determined by the length- IMD are shown in Fig. 66.
the current-limiting point of the chip at to-diameter ratio of the metal tubing or
1.35A. Temperature-compensating diode braid, plus the gauge and insulation The Novice "1/4 Gallon"
D I is added for bias tracking as the power thickness of the wire used for the Some Novice class licensees own trans-
devices heat up. This circuit prevents high-impedance winding. A high coupling mitters that are not capable of running the
thermal runaway. The base-emitter junc- coefficient is necessary for good high- maximum legal-power input for that class
tion of a 2N5190 is used as this diode. frequency response. Both transformers license - 250 watts. This amplifier was
Physical dimensions of the 2N5190 allow are loaded with ferrite material to provide designed primarily with this in mind.
it to be used as the center standoff post of sufficient low-frequency inductance. Approximately 25 watts of power are
the circuit boarn, ensuring adequate ther- The collector choke (T2) provides an required to drive the amplifier to 165-
mal bonding of the diode junction to the artificial center-tap for T3. This produces watts output on 80 meters. If a 75-watt
heat sink. even-harmonic cancellation. Additionally, transmitter is used to drive the amplifier,
Device input impedances vary across T2 is a low-impedance point that supplies the transmitter output should be reduced
8-41 Chapter 8
BROADBAND AMPLIFIER
5600pF
R3
01
50-OHM
OUTPUT
TO FILTERS
T2
~"
50-OHM
INPUT
51
. . pF
R2
BIAS 5600pF
REGULATOR R4
03
+13.6V
+ 500pF
~3V
1000
18k
+13.6V
8200
Fig. 62 - Circuit diagram of the 140-watt amplifier. All capacitors except part of Cl, C2 and the electrolytic types are ceramic chips. Capacitors
with values higher than 82 pF are Union Carbide type 1225 or Varadyne size 14. Others are type 1813 or size 18, respectively.
Cl - 1760 pF (two 470-pF chip capacitors in 03 - 2N5989 or equiv. T2 - 6 turns no. 18 enameled wire, bifilar wound.
parallel with an 820-pF silver mica). Rl, R2 - Two 3.6 ohm, 1/2 W, in parallel. Ferrite core: Stackpole 57-9322, Indiana
C2 - l000-pF disc ceramic. R3, R4 - Two 5.6 ohm, 1/2 W, in parallel. General F627-8 01 or equiv.
01 - 2N5190 or equiv. Tl - Primary: 3 turns no. 22 insulated wire. T3 - Primary: Braid or tubing loop. Secondary: 4
L 1, L2 - Ferrite choke, Ferroxcube VK200 Secondary: Braid or tubing loop. Core material: turns no ..18 insulated wire. Core material: Two
19/4B. Stackpole 57-1845-24B, Fair-Rite Products Stackpole 57-3238 ferrite sleeves (70 material)
L3, L4- Two Fair-Rite Products ferrite beads 2873000201 ortwo·Fair-Rite Products 0.375- or a number of toroids with similar magnetic
(2673021801 or equiv.) on no. 16 wire. inch 00 X 0.2-inch 10 X 0.4 inch (9.5 X 5.1 X • characteristics and 0.175-inch sq. (113-mm
L5 -"- 1 turn through torroid of T2. 10.2 mm), Material·77 beads for type A sq.) total cross-sectional area.
01,02- MRF454. (Fig. 63A) transformer. Ul- Motorola MC1723G orequiv.
to a level just sufficient to drive the ampli-. from a transmitter that has more than The power supply uses an old TV -set
fier to its full-power input. This can usual- enough power to drive the amplifier than transformer that has three secondaries:
ly be accomplished by lowering the drive it is to boost the output from a transmitter 600, 6.3 and 5.0 volts. The 600-volt
to the transmitter output stage and re- that provides insufficient drive. Class B winding supplies the full-wave bridge rec-
loading. operation was chosen for that reason. tifier with ac energy. Dc output from the
The Circuit
The main ingredient in this amplifier is
a pair of 6KD6 television-sweep tubes (see
Fig. 68. Although the tubes are rated at 33
watts of plate dissipation, they can
handle temporary overloads of at least
100 watts without sustaining permanent
damage. These tubes were chosen over
811s or 572Bs because they can often be
"liberated" from old television sets or can
be purchased new from most TV service
shops.
D11, an 11-volt Zener diode, provides
cathode bias for. the tubes and establishes
the operation as Class B. That class of
amplifier requires less 'driving power than Fig. 63 - Shown are the two methods of constructing the transformers as outlined in the text. At
does a Class C stage for the same power the left, the one-turn loop is made from brass tubing; at the right, a piece of coaxial cable braid is
output. It is easier to reduce the output used for the loop. .
HF TransmItting 8-42
LOOPS CAN BE PROV.IOED FOR
CURRENT-PROBE MEASUREMENTS
BOTTOM
Fig. 64 - Actual-size layouts for the amplifier. Gray areas represent unetched copper. The pattern at the left is for the top side of the board and the
one at the right for the bottom of the board.
rectifier assembly is filtered by means of plate current. One should be careful if Construction
three 330-IJF capacitors. The 47ckO resis- working near the meter with the power
tors across each of the capacitors equalize on, as full plate potential will be present Perhaps the best way to classify this
the voltages across the capacitors and between each of the meter leads and amplifier would be to call it a "junker
drain the charge when the power supply ground. Caution: Turn off and unplug the type" amplifier. Every attempt was made
is turned off. Equalizing resistors are used amplifier before making any changes or to keep the amplifier as simple and easy to
across each of the diodes to ensure that adjustments. build as possible. The only critical values
. the reverse voltage will divide equally be-
tween the two diodes in each leg. The
capacitors across each diode offer spike
protection. Each 6KD6 draws· 2.85 am-
peres of filament current. Both filaments
are connected in parallel across the 6.3-
volt transformer winding. The 5-volt
winding is connected in series with the 6.3-
volt winding; the total (11.3 volts) is rec-
tified, filtered, and used to power the
relay. .
L2 L4
C5
1
'til Fig. 65 - This table
shows the values for
seven-pole Chebyshev
low-pass filters suitable
for use with the
amplifier. These filters
have been deSigned
around standard
capacitance values for
Band Fe (MHz) VSWR C1, C7 C3, C5 L2, L6 L4
A 0-50 IJA meter is used to measure the each of the capacitors.
160 2.02 1.09 1200 2700 5.42 6.41 Capacitance values are
plate voltage and current. The meter reads 80 4.11 1.07 560 1300 2.62 3.13 in pF and inductance
0-1000 in the plate-volts position and 40 7.98 1.08 300 680 1.37 1.62 values are in /lH. This in-
20 15.16 1.09 160 360 .72 .85
0-500 rnA in the plate-current position. A 15 21.69 1.06
formation was prepared
100 240 .49 .59 by Ed Wetherhold,
0.26-ohm shunt is placed in the high- 10 36.85 1.10 68 150 .30 .36 W3NQN.
voltage lead to facilitate metering of the
8-43 Chapter 8
Fig. 66 - Spectral photographs Glf the output from the 140;watt amplifier. The display at the left Fig. 67 - The completed quarter-kilowatt
was taken with the amplifier operating on the SO-meter band. At the right is a close look at the amplifier. .
IMO products generated by the amplifier. Third-order products are down some 32 to 34 dB below
PEP.
K18
L,
J1
J2
OUTPUT
INPUT
6KD6
3,11
RFCI
lW
.006
.01
rL 2kV
S.M.· SILVER MICA
39k
2W
39k
2W
• S3~ J3
RELAY
+ 0 PRo CONTROL
470...,F
~ KIA
Fig. 6S - Schematic diagram of the 1/4-kilowatt amplifier. Resistors are the 1/2-watt composition type unless otherwise specified. Fixed-value
capacitors are disc ceramic unless otherwise noted. Polarized capacitors are electrolytic.
Bl - 117-V ac blower. L 1 - '8 turns B & W 3022 coil. The entire coil is Amidon FT-82-72 core.
Cl- Variable capacitor, 340 pF maximum, used for SO-meter.operation and is tapped for RFC2, RFC5- Rf choke, 2.5 mH, 500 mAo
(Millen 19335 or equiv.) the other bands as follows: 7-1/2 turns for 40 RFC3, RFC4 - 15 turns no. 14·enam. wire on an
C2 - Variable capacitor, 1095 pF maximum meters; 13-1/2 turns for 20 meters; 14-1/2 turns Amidon FT-S2-72 core.
(surplus 3-section 365-pF variable). for 15 meters; 15-1/2 turns for 10 meters. Turns S1 - Spst toggle switch, 4 A.
01 - OS, incl. - Silicon diodes, 1000 Volt, 2 A. are measured from C2 end of coil. S2 - Opdt toggle switch, low current type.
09 - Silicon diode, 50 V; 1 A. Ml - Panel meter 0-50I£A, Calectro 01-910 or S3 - Spst toggle switch, low current type.
01 O~ Pilot lamp assembly, 12 V. equiv. S4 - Single-pole, 5-position ceramic rotary
011 - Zener diode, 11 V, 50 W. Rl - Meter shunt, 10 feet (3.048m) no. 24 enam. switch, single-wafer type.
Jl, J2 - Coaxial.connector, SO-239. wire wound of a large ohmic value 2-watt Tl - Television transformer (see text)..
J3 - Connector, phono type. composition resistor. Zl, Z2 - 5 turns no. 18 enam. wlre on a 47-ohm,
K 1 - Opdt relay, 12-V field, 2-A contacts. RFC 1 - 20 turns no. 24 enam. wire on an l-watt resistor.
HF Transmitting 8·44
noninductive dummy loan. Connect the!
relay control line to the transmitter or
external antenna-relay contacts. Then
plug in the line cord and turn the power
switch to the ON position. With the meter
switch in the PLATE VOLTS PQsition,
the reading on the meter should be
approximately 425, which corresponds to
850 volts. If the power transformer used
has a high-voltage, secondary other than
600 volts, the reading will vary according-
ly. If no plate voltage is indicated by the
meter, check your wiring for .possible er-
Fig. 69 - An inside view of the amplifier. C1 is
located at the lower right. C2 is mounted
rors or defective components. Next, place
below the chassis and is connected to L1 meter in the PLATE CURRENT posi- Fig. 70 - Underside of the amplifier. Compo,
using a feedthrough insulator. tion, the band switch to the 80-meter nent layout is not particularly critical.
band, and apply a small amount of drive
to the amplifier - enough to' make the
are those for the amplifier plate-tank meter read 50 rnA (5 on the meter scale). The Circuit
circuitry. Reasonable parts substitutions With the plate-tank loading control fully
elsewhere in the circuit should have little meshed, quickly adjust the plate tuning The power tube is used in the grounded-
or no effect on the performance of the capacitor for a dip in plate current. Apply cathode configuration (Fig. 72). If the
amplifier. For example, if 330-I'F filter more drive (enough. to' make the meter tube to be used is a tetrode, the screen and
capacitors are not available, 250- or 'read 100 rnA), advance the loading con- grid should be tied together to simulate a
, 300-j.lF units could be substituted. They trol approximately one-eighth turn and triode. The O.OI-j.lF disc capacitors from
should be rated at 450 volts or greater to readjust the plate tuning control for a dip each cathode terminal to ground provide
provide a margin of safety. If a O.OOI-I'F in the plate current. Continue this proce- a low-impedance rf path, effectively by-
plate-blocking capacitor is not on hand dure until the plate-current maximum dip passing the cathode to ground. Grid vol-
and a 0.OO5"l14iF unit is, use it. Builders is approximately 300 rnA. The final value tage is developed a£ross a low resistance,
often attempt to match parts exactly to of plate current at which the amplifier Rl, which coincidentally maintains a con-
the type specified in a schematic or parts should be run depends on what the plate stant load impedance for the exciter and
list. A few projects are this critical in voltage is under load. In our case this prevents instability.
nature, but the majority, including this value was 800 volts. Therefore, the amount The output circuit is a pi-network. The
one, are not. ' of current corresponding to 250 watts in- values for the plete-tank components
The transformer used in this amplifier put is approximately 310 rnA. (I = PIE, I depend on the plate load impedance of the
was garnered from an old TV set. Any =,250/800, I = 312.5 rnA.) The same tune-up tube. This is a function of the plate
hefty transformer with a high-voltage procedure should be followed for each of voltage and the plate current. To allow the
secondary between 550 and 700 volts the other bands. The amplifier efficiency builder flexibility in power-supply re-
should be adequate. Most of these on 80 through 20 meters is approximately quirements, Table 15 lists the pi-network
transformers will have multiple low- 65 percent, dropping to 60 percent on 15 values needed for different plate voltages.
voltage secondaries suitable for the tube meters. On 10 meters, efficiency is slightly The computed values assume the use of
filaments and relay requirements. less than 50 percent. Poor efficiency on appropriate plate current for 1 kilowatt
The chassis used to house the amplifier the higher bands is caused primarily by input at the plate voltage given. Observe
happened to be on hand and measured 3 X the high-output capacitance characteris- the maximum tube ratings when choosing
10 X 14 inches (76 X 254 X 356 mm). No tics of sweep tubes . a power-supply voltage.
. doubt the amplifier could be constructed High voltage goes to the plate through
on a smaller chassis. The beginner is RFC 1. The choke is wound in such a way
cautioned not to attempt to squeeze too A "Universal" Three-Band Linear as to reduce its distributed capacitance
much in too small a space. Amplifier over a conventionally wound choke.
The front, rear, side and top panels are The cost for a project can be as RFC3 prevents high voltage from ap-
constructed from sheet aluminum and important a goal as ultimate performance. pearing on the antenna should the
help to keep the amplifier "rf tight." Any The challenge of finding surplus or used plate-blocking capacitors (Cl and C2)
air-flow openings are "screened" with electronic components for a project may short; it provides a dc path to ground,
perforated aluminum stock. The front- be even greater than the challenge of thereby blowing the power-supply fuse.
panel meter opening is shielded by means actual construction. This amplifier was
of an aluminum enclosure (a small designed to make use of the many tubes
Minibox would serve quite nicely). The available through surplus channels and
on-off power switch, pilot light, meter flea markets for bargain prices. Some
switch, band switch, tuning and loading commercial users employ these tubes and
controls, and amplifier in-out switch are remove them from service after a pre-
all located on the front panel. On the rear scribed period as part of preventive
panel are the amplifier input and output maintenance. These "pulls" still offer
. connections, relay control jack and the thousands of hours of satisfactory use in
fuse holder. As can be seen from the amateur equipment.
photograph, a fan is located near the tube This amplifier is capable of running one
envelopes to keep them cool during kilowatt of dc input on the 160-, 80~ and
operation. 40-meter amateur bands. The high-
frequency usefulness of the 833A in the
Setup and Operation amplifier shown limits the usage to those
Attach the transmitter output to the three bands. Pi-network values are given
amplifier input connection. Then, join the for various surplus tubes: 4-400A,
output of the amplifier to a 50-ohm 4-250A, 304TH and 833A (Table 15). Fig. 71 - Front panel of the amplifier.
1-45 Chapter 1
~
~ - ~t-~-------B~ 40 80
160
INPUT 0.01
(SO!)') C5
1000
rJ.;.SkV
05
37V
50W EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPACITANCE ARE
IN MICROFARADS (~F) ; OTHERS
ARE IN PICOFARADS (pF OR ~~Fl;
RESISTANCES ARE IN OHMS;
k ,,\000. M'l,OOO,OOO
.c.
cij K2C
J4
RELAV
CONTACTS
Operating bias for the tube sets the 'a sealed box. All metal-to-metal bond'
Table 15
operation to Class AB2. D5 and D6 raise areas should be cleaned carefully to
the cathQde potential above ground to guarantee good electrical contact. An Pi·Network Values
establish the appropriate idling current of abundance of fastening screws hold these
160 meters
the tube. When in the standby mo.de, relay seams together, further "RFI-proofing"
contacts switch to a different bias level, the box. Air is blown across the tube by Plate Voltage C3 C4 L1
reducing plate current to zero. means of a muffin fan. Plenty of cooling
2000 490 pF 2987 pF 17.83 "H
air not only stabilizes the tuning of the 2500 305 pF 2098 pF 27.59 "H
Construction amplifier but also lengthens tube life. 3000 222 pF 1676 pF 37.29 "H
C.onstruction is straightforward. While Many variations in switching, metering 3500 193 pF 1676 pF 42.74 "H
the iayout shown here proved ideal for the and power supplies are possible with a
80 meters
components on hand, many builders will unit such as this. The methods shown here
. use parts other than the ones shown. are intended to be examples of the various 2000 252 pF 1536 pF 9.17 "H
Physical constraints may require different possibilities. Parts availability and in- 2500 157 pF 1079 pF 14.19 "H
layout arrangements. The main con- dividual desires of the builder dictate the 3000 114 pF 862 pF 19.18"H
3500 99 pF 862 pF 21.98 "H
siderations are to keep lead lengths to a final design.
minimum, and to ensure adequate grounds 40 meters
where needed. Tune-up and Operation
A homemade sealed box encloses the Because of the expected variations 2000 114 pF 670 pF 5.03 "H
2500 71 pF 451 pF 7.77 "H
circuitry. Three potential problem areas between different amplifiers, there are no 52 pF 341 pF 10.48 "H
3000
dictate that in this, and all rf power ampli- hard and fast rules for tune-up. For initial 3500 45 pF 341 pF 12.Q2 "H
fiers, a shielded box is necessary. Safety tune-up less than full plate voltage is
considerations, both d.ue to high dc and recommended. This permits a greater These values normalized for a dc power input
of 1 kW.
high rf voltages, RFI reduction and ade- - safety margin (both for the amplifier and
quate air cooling of the tube all necessitate the operator) if unexpected events occur.
HF Trananmtlng 8-48
7 1/4"(32mm)
KI is a plug-in type of relay with con-
tacts that are rated for 10 amperes of cur-
rent. This relay plugs into an octal socket
t J TUBE END that is located between the filament trans-
former and one of the 4-400A tubes. K2 is
POLYSTYRENE
OR
PLEXIGLAS
f·K::::=A-- a small spdt relay that was garnered from·
the junk box and is of unknown origin.
Any similar relay should work fine.
ROD I
Construction
e"
The amplifier was build on a 10 4 14 X
(203mm) :=-=-r~- 3-inch (254 X 356 X 76-mm) aluminum
chassis. A heavy duty front panel was
I 70 TURNS.
constructed from a discarded rack panel
114" (6.3m m) I and' measures 10 X 14.inches (254 X 356
SPACE BETWEEN i
WINDINGS mm) making the overall dimensions of the
amplifier 14 X 10 X 10 inches (356 X
254 X 254 mm).
Layout of the components is not
8+ END
RFC1 especially critical, however the builder
Fig. 74 - Interior view of the three·band should strive to keep the tank-circuitry
Fig. 73 - RFC1 is wound on 1·1/4 inch (32·mm) amplifier. The band switch is mounted to the leads as short as possible to avoid stray
diameter polystyrene rod as depicted. front panel inside one end of the tank coil. capacitances and inductances. The general
layout of this amplifier can be seen in the
accompanying photographs.
Increase drive slowly while making ad- fed through a heavy duty bifilar-wound A flange made from Plexiglass tubing
justments, this again providing a safety choke which uses a 950 mu ferrite rod 112 and flat 114-inch Plexiglass stock is used
margin. Remember, hazardous voltages are inch (13 mm) in diameter and 7 inches to connect to the defroster hose. Part of
present in this amplifier, and these should (178 mm) long of material. Power is fed this flange is visible in the photograph of
be respected at all times. directly into the filament without the aid the underneath of the chassis. An alter-
A wattmeter is helpful during tune-up. of an input matching' network. The input . native to building your own flange would
A goal of maximum output power VS~R is such that most transmitters be to purchase one from a supplier of cen-
consistent with high efficiency should be should have no difficulty in driving the tral vacuum cleaning systems. Of course,
sought. Efficiency on the order of 55 to 65 amplifier; there is no reason the builder couldn't
percent can be expected. In order to The 8.2-volt Zener diode develops bias mount the blower directly on the rear
increase tube longevity, the manufac- and allows the tubes to run in Class AB. panel of the amplifier.
turer's maximum ratings should never be The 50-kO/IO watt resistor in conjunction A perforated aluminum cover (not
exceeded. with K2 cut the amplifier off during shown in these photographs) is used to
receive or standby periods. A single 0-150 enclose the top of the amplifier. Good
An Economy l-kW Amplifier rnA panel meter is used to monitor either contact between this cover and the mating
The amplifier describ'ed here was built plate or grid current - selectable with a surfaces is important for effective
with mostly junk-box components. A pair front-panel switch, S2. For plate-current shielding. Make sure to remove any paint
of 4-400A tubes are run in grounded-grid measurements the meter reads 0-1.5 to ensure a good metal-to-metal cont~ct.
fashion and can develop 2-kW PEP input amperes (utilizing meter shunt R2) and
when driven from a 100-watt exciter. Ap- reads 0-450 rnA for grid-current measure- Operatian
proximately 40 watts of drive is required ments (using meter shunt RI). This amplifier is designed to operate at
for l-kW operation and 100 watts for A small power supply capable of the l-kW level for cw operation and at the
2-kW operation. The amplifier makes use operating KI and K2 is included in the 2-kW PEP level for ssb. This amplifier
of Eimac SK-41O air-system sockets and amplifier. An amplifier control jack was optimized for the 1500-watt level,
SK-406 chimneys. A blower, which is located on the rear apron of the amplifier thereby providing reasonable efficiency at
mounted external to the amplifier chassis, is grounded for amplifier operation. both the I and 2 kW levels without the
forces air through a length of automobile The pi network consists of two coils - need for switching the power-supply
defroster hose, approximately· 2 inches one wound from 3/16-inch (4.7 mm) cop-
(51 mm) in diameter and is flexible, through per tubing and one wound from large
the pressurized chassis and out the air Miiliductor stock. The coil made from
system sockets and chimneys. By mount- copper tubing is wound in two different
ing the blower away from the amplifier, diameters. It starts out from the 4-400A
and therefore the operating position, this plates at the smaller diameter (for IO and
source of noise can be greatly reduced, if )5 meters) and progresses to the larger
not eliminated. diameter (for 20 meters) all with one con-
tinuous piece of tubing. A portion of the
copper tubing is used on 40 meters along
The Circuit with a length of Miniductor stock. The
The schematic diagram of the amplifier builder may wish to end the copper tubing
is shown in Fig. 76. Relay KI is used to coil at the 20-meter tap position and con-
switch "around" the ~mplifier during tinue on from that point with the Mini-
receive periods or when it is desired' to ductor stock for 40 and 80 meters. The in-
bypass the amplifier. S4 allows the ductance values required for each band Fig. 75 - This is a photograph of the front
are given in Fig. 76 so that the builder may panel of the 4·440A amplifier. The amplifier is
operator to switch around the amplifier relatively compact, measuring 10 x 14 x 10
while leaving it in a standby conditon, adapt junk-box coils for the tank cir-' inches (254 x 356 x 254 mm). The PEP input
ready for operation. Filament voltage is cuitry. for this amplifier is 2 kW.
8-47 Chapter 8
tubes. The circuit configuration is ground-
ed grid and uses no tuned-input tank
components. When properly adjusted, the
amplifier is capable of IMD charac-
teristics which are better than can
J1
INPUT be achieved by a typical exciter, therefore
the added complexity of band switching a
tuned-input circuit was deemed unneces-
sary.
Construction
Building an amplifier such as this is
EXCEPT AS INDICATED, DECIMAL VALUES OF often an exercise in adapting readily
CAPACITANCE ARE IN MICROFARADS (jJF) ;
OTHERS A~E IN PICOFARADS (pF OR jJ)lF);
available components to a published
RESISTANCES ARE IN OHMS; circuit. For this reason, a blow-by-blow
k.IOOO, M=IOOO 000,
description ofthis phase ofthe project will
not be given. An effort was made,
+3400V
however, to use parts which are available
generally, and should the builder desire,
this model could be copied verbatim.
The most difficult constructional prob-
lem is that of aligning the tube sockets
correctly. It is imperative that the sockets
be aligned so that when the tubes are
mounted in place, the flat surfaces of the
anodes fit smoothly and snugly against
the thermal-link heat-transfer material.
AMP.
Any misalignment here could destroy the
J~ONTRck tubes (or tube) the first time full power is
applied. The mounting holes for the tube
'sockets are enlarged to allow final
positioning after the tubes are "socked" in
place with the clamping hardware. Pres-
sure must be applied to the anodes so that
Fig. 76 - Schematic diagram of the amplifier. A pair of 4·400A tetrodes are used.
they are always snug against the thermal
C1, C4 - Transmitting ceramic, 1000 pF. or equiv.
C2 - Variable, 150 pF, 4500 V. Johnson 154·15 R1 -Grid meter shunt, 0.33 ohms. Wind 2 feet
link. The hardware used to perform this
or equiv. (610 mm) no. 32 wire on a large value 1· or function must be nonconducting material
C3 - Variable, 1095 pF, receiving type, 3 sec· 2·watt resistor. capable of withstanding as much as
tion, 365 pF per section used here. R2 - Plate·meter shunt, 0.0733 ohms. Wind 1 250°C (482°F). The pressure bracket
J1, J2 - Coaxlal connector, builder's choice. foot,1 inch (330 mm) no. 32 wire on a large used here was fabricated from' several
J3 - Phono connector. value 1· or 2·watt resistor.
K1 - Relay, dpdt, 12·volt field. RFC1 - Filament choke, 20 bifilar turns on no. Millen jack-bar strips (metal clips removed)
K2 - Relay, spdt, 12·volt field. 12 enameled or plastic covered wire on a mounted in back-to,back fashion. The
L1 - Copper tubing, 3/16·inch (4.7 mm) 950 mu, 7 x 1/2 inch (178 x 13 mm) ferrite entire assembly is held in place by means
diameter. 6·1/2 turns at 2·inch (51 mm) dia· rod.
meter tapering into 6 turns at 3·inch (76 mm) RFC2 - Plate choke, B&W 800 or equiv.
of a long piece of no. 10 threaded brass
diameter. Tap at 5·1/4 turns for 10 meters, RFC3 - Choke, 2.5 mH, 300 mAo rod which passes through a small hole in
6·1/4 turns for 15 meters, 9·1/4 turns for 20 S1 - Band switch, 5 position. Millen 51001 or the center of the heat sink. An attempt to
meters. (10 meters, 2.3/AH; 15 meters 3.2/AH; equiv. give meaningful comments about how
20 meters, 4.9/AH. Note: stray inductances S2 - Toggle switch, spdt. tight the tubes should be pressured to the
must be subtracted from these figures.) S3, S4 - Toggle switch, spst.
L2 - 14·1/2 turns no. 12 wire, 6 turns T1 - Filament transformer, 5 volts, 30 'copper and aluminum sink will not be given.
per inch. B&W 3033 or equivalent. Tap amperes, 117·volt primary. Stancor P6492 or Suffice it to say that the tubes should fit
at 4·3/4 turns from l1 end for 40 equiv. flat and snugly against the thermal hard-
meters. Use entire coil for 80 meters. T2 - Transformer, 12.6 volts, 300 mAo ware. The heat sink was purchased from
(40 meters, 9.6/AH; 80 meters, 21,22 - Parasitic suppressor, 2 turns 5/16
17.5/AH. Note: stray inductances inch (8 mm) wide copper strap wound around Thermalloy and is connected to a Ij4-inch
must be subtracted from these figures.) three 1oo·ohm, 2·watt resistors in parallel. (6.3-mm) thick piece of ordinary copper
M1 - 0·150 mA panel meter. Simpson 06400 plate. The total cost for the copper and
the aluminum sink is somewhat more than
the price of a good centrifugal blower ($30)
but the savings offered by not having to
purchase special tube sockets and glass
chimneys overcomes the cost differential.
voltage. With 3400 volts on the plate, the dealing with high-power amplifiers is heat
1 kW plate current should be 294 rnA and and how to reduce it. The usual method TheRF Deck
the grid current should be roughly 100 has been to use a large fan or blower, but The two sections of the pi-L network
rnA. At 2 kW the plate current will be 588 this solution is generally noisy. By using are isolated from each other by placing
rnA and the grid current approximately the principles of heat transfer, a noiseless one of them lInder the chassis. Although
200 rnA. Idling plate current will run ap- amplifier can be made with the use of an not shown in the photograph, a shield was
proximately 120 rnA .. Efficiency on all adequate heat sink and conduction-cooled added to prevent rf energy from entering
bands should be roughfy 60 percent. tubes. the control section underneath the chassis.
The amplifier shown in the photo- The shield divides the chassis between the
A Conduction-Cooled 2-Kilowatt Amplifier graphs and schematically in Fig. 80 uses a tube, sockets and the inductors. The
One of the major concerns when pair of 8873 conduction-cooled triode loading capacitor is mounted directly
HF Transmitting &-48
Fig. 77 - A top view of the amplifier. The transmit/receive relay can be Fig. 78 - Photograph of the underside of the chassis. Component
seen between the filament transformer and one of the 4-400A tubes. placement is not critical. The builder should plan a layout that suits his
or her components.
8-49 Chapter 8
"
AMPLIFIER •
500
;hTk"V
tM 1
~ 5kV
~
J6
+2500V
tM
~:r
J~
RF
1M
OUTPUT
01
tM J2
INPUT
04 06
05 07
M1 M2
R1
03
rd.~t J4
RLY
S
J5
EXCEPT AS INDICATED, DECIMAL VALUES OF
ALe
OUT CAPACITANCE ARE IN MICROFARADS (JIF I ;
OTHERS ARE IN PICOFARADS (pF OR JlJIFI;
100V SW SW FIL. FIL. G
RESISTANCES ARE IN OHMS; 6.36.3
k.l 000, M.t,OOO, 000.
Fig. 80 - Circuit diagram for the 8873 conduction-cooled. amplifier. Component designations not listed below are for text reference. RFC1 and RFC2
are wound on the same ferrite rod in the same direction; three wires are wound tqgether (Amidon MU-125 kit). Tube sockets for V1 and V2 are E. F.
Johnson 124-0311-100. The thermal links are a,vailable from Eimac with the tubes. The heat sink is part number 2559-080-AooO from Astrodyne Inc.,
353 Middlesex Ave., Wilmington, MA 01887.
C1 - Transmitting air variable, 347pF, E. F. from connection point with L1, 2-1/2 Inches from 3.375 feet of no. 22 enam. wire wound
Johnson 154-0010-001. (63.5-mm) for 20 meters, 7-3/4 turns for 40 over the body of any 2-watt reSistor higher
C2 - Transmitting air variabie, 100 pF, E. F. meters. than 100 ohms in value.
Johnson 154-30. L3 - 11-1/2 turns, 2-inch (51-mm) diameter, 6 R2 - Meter shunt, 0.2 ohm made from five 1-
02-07, incl. - 1000 PRY, 2.5 A, Motorola tpi, Barker and Williamson 3025. ohm, 1-watt resistors connected in parallel.
HEP170. . L4 - 10 turns, 2-inch (51-mm) diameter, 6 tpi, RFC1, RFC5, RFC6 - 2.5 mH Millen
08 - 8.2-V, 50-W Zener diodes. with taps at 3 turns for 10 meters, 3-1/2 turns 34300-2500.
J1 - SO-239 chassis mounted coaxial for 15 meters, 4-3/4 turns for 20 meters, 6-3/4 RFC3 - Rf choke, Barker and Williamson
connector. turns for 40 meters, all taps made from Model 800 with 10 turns removed from the
J3, J4, J5 - Phono jack, panel mount. junction of L3, Barker and Williamson 3025. bottom end.
J6 - High-voltage connection, Millen 37001. M1 - 200 mA full scale, 0.5-ohm internal resis- RFC4 - 22"H, Millen 34300.
K1 - Enclosed, three-pole relay, 110-volt dc tance Simpson Electric Designer Series S1 - High-voltage band-selector style, double
coil, Potter and Brumfield KUP14015. model523. pole, six pOSition, James Millen 51001 style.
L1 - 4-3/4 turns of 1I4-inch (6.3-mm) copper M2 - 1 mA full scale, 43 ohms internal resis- Z1, Z2 - 2 turns 318-inch (9.5-mm) wide copper
tubing, 1-3/4-inch 10,2-114 inches long. tance (Simpson Electric, same series as M1). strap wound over three 1oo-ohm, 2 watt
L2- 12-1/2 turns, 1I4-inch (6.3-mm) copper R1 - Meter shunt 0.05555 ohms constructed resistors connected in parallel.
tubing, 2-3/4-inth (70 mm) 10, tap at one turn
HF Tran.mlttlng 6-50
Chapter) 7
T:e frequencies above 50 MHz were stage. While relatively simple to construct, pow~r levels involved, there is no reason
once a world apart from the rest of such transmitters can be a cause of much to consider transmitting mixers differently
amateur radio, in equipment required, in grief unless the builder takes precautions than their receiving counterparts. One
modes of operation and in results to prevent undesired multiples of the thing to keep in mind is that many
obtained. Today these worlds blend oscillator and the multiplier stages from deficiencies in the transmit mixer will
increasingly. Thus, if the reader does not being radiated. For frequencies below 450 show up on the air. Receiver-mixer
find what he needs in these pages to solve MHz the transmitting ~ixer is not troubles are your problem. Transmit-
a transmitter problem, it will be covered difficult to construct and is recommended mixer troubles become everyone's prob-
in the hf transmitting chapter. This for most applications. Spurious-signal lem!
chapter deals mainly with aspects of radiation is much easier to prevent with A trio of popular types of transmitting
transmitter design and operation that call the latter, although it does not lend itself mixers is shown in Fig. 2. The doubly
for different techniques in equipment for to compact fm equipment design. For balanced diode mixer at A may be built
50 MHz and up. operation on the higher amateur uhf using either discrete components, or the
bands, the oscillator-multjplier approach phase relationship between ports may be
SSB/CW vs. FM offers definite advantages and is recom- established using etched-circuit strip lines.
Whenever vhf operators gather, the mended at present. Fig. 1 shows how the Miniature DBMs are available at low cost
subject of fm vs. ssb and cw is bound to harmonics of a 144-MHz signal may be from several manufacturers. They offer an
come up. Because of their mode differences, multiplied to permit operation on ama- almost-foolproof method of generating
the two types of operation are segre- teur microwave bands. Stability at 144 vhf ssb. Another popular mixer uses a pair
gated on the lower four amateur vhf bands. MHz is easy to achieve with the current of FETs in a singly balanced configuration.
Actually, both forms of communications technology, making stable microwave If care is taken in construction and adjust-
have their advantages and disadvantages. signals simple to generate. Varactor ment, local-oscillator rejection will be ade-
They are better discussed SUbjectively diodes are llsed as frequency-multiplying quate with this circuit. To be safe, a series-
among operators. Here we are only in- devices. They are installed in resonant tuned trap, designed to attenuate the LO
terested in the different requirements each cavities constructed from double-sided . leakage even further, should follow this
mode places upon the transmitting equip- pc-board material. Operation will proba- stage. A typical FET balanced mixer is
ment used. In general, equipment used for bly be crystal controlled, as even the best shown in Fig. 2B.
fm is of the oscillator-multiplier type. transceiver/transverter. combination used Finally, we see a typical vacuum-tube
Because ssb cannot be passed satisfactori- to generate the 144-MHz signal may create mixer (Fig. 2C). Because it can handle
ly through a frequency-multiplication stage, problems when the output is multiplied in mQre power, the tube mixer has endured
generation of vhf ssb signals requires the frequency 40 times! A frequency synthe- at vhf. Its higher output, when compared
use of one or more mixer stages. Vhf cw sizer with a stable reference oscillator may . to most solid-state mixers, reduces the
may be generated by either method. Re- be used to generate the 144-MHz sign~l, number of subsequent amplifier stages
cently, manufacturers of synthesized as but its output should be well filtered to
well as crystal-controlled amateur fm eliminate noise.
transceivers have been using a combina- Although spurious outputs of the
X9 .1296 MHz
tion of both approaches. The multimode various multiplier stages may not cause
vhf transceiver, which· offers the operator harmful interference, that is no excuse for X16' 2304 MHz • t
a choice of cw, ssb, fm and often a-m, is a not removing them. In most cases, the Q
X24= 3456 MHz
reality. Here again we find both approach- of successive cavities will suffice. A
es to signal generation. band-pass filter may be used to filter the X40.5760 MHz
final mUltiplier stage. Construction details X72.10,368 MHZ
The Oscilllltor-Muitiplier Approach of a 432-to-1296 MHz frequency multi-
This type of transmitter, which may be plier using switching diodes are presented
used for fm or cw, generally starts with a later in this chapter. Fig. 1 - The harmonic relationships of most
microwave bands to the 2-meter band are
crystal oscillator operating in the hf range, diagrammed here. The 15-mm (24 GHz) band
followed by one or more frequency- Transmitting Mixers
bears no easily utilized integral relationship to
multiplier stages and at least one amplifier With the possible exception of the 144 MHz. .
7·1 Chapter 7
needed to reach a specific power level.
Apart from feeling more comfortable with
tubes, this is the only advantage available
from using them as mixers, at least
on the lower vhf bands.
LO
High-Level Tl'fIIISmittin, Mixers (LOW-LEVELl
(A)
When designing a transmitting conver-
ter for vhf, the tradeoffs between the ad-
vantages of mixing at a low power level, Voo
such as in a diode-ring mixer, and using
several stages of linear amplification must
be weighed against the cost of amplifying
devices. Linear uhf transistors are still
..
relatively expensive. On 432 MHz and
,~
above, it may be desirable to mix the i-f
and local-oscillator signals at a fairly high
level. This method makes it unnecessary
~. rOW-LEVELl
"
to use costly linear devices to reach the
same power level. High-level mixing
results in a slightly more distorted signal
than it is usually possible to obtain with
I-F
conventional methods, so it should be (LOW-LEVELl
used only when essential. Fig. 3 gives the
schematic diagram of a typical 432-MHz
high-level mixer. VI is the final amplifier
LO
tube of a retired commercial 450-MHz fm (LOW-LEVEll
transmitter. The oscillator and multiplier
•
(B)
stages now produce local-oscillator injec-
tion voltage, which is applied to the grid
as before. The major change is in the cath-
ode circuit. Instead of being directly at
ground, a parallel LC circuit is inserted and
tuned to the i-f. In this case a lO-meter i-f
was chosen. With the exception of 144
MHz any amateur band could serve as the
i-f. Two meters is unsuitable because the
1- Fo---:1 1-.......-1-
(HIGH-LEVEll
LO~I--"""
(HIGH-LEVELl
i,"'" -
~RF
''''co
third harmonic of the i-f would appear at
the output, where it would combine with
the desired signal. In fact, some additional
output filtering is needed with this circuit.
(e)
A simple strip-line filter, such as appears
in the ARRL Radio' Amateur's VHF
Manual, will do the job. The original crys- Fig. 2- A trio of commonly used vhf transmitting mixers. At A, perhaps the simplest, a commercially available
tal in the transmitter is replaced with one diode doubly balanced mixer. Rf output is low, requiring the use of several stages of amplification to reach a
yielding an output at the desired local- useful level. At B, a singly balanced mixer using FETs. Adjustment of this circuit is somewhat critical to prevent
oscillator frequency, then the intermediate the local-oscillator signal from leaking through. A mixer of this type can supply slightly more output than a
diode mixer. (3 dBm, as opposed to 1 dBm for the mixer shown at A). At C, a high-level mixer using a vacuum-
stages are retuned. tube triode is shown. V1 in this case might be a 2C39 or 7289. With the correct circuit constants this mixer
One disadvantage of the high-level could provide an ssb output of 15 watts on 1296 MHz. Power input would be about 100 watts! In addition, 10
mixer is the relatively large amount of watts of LO and 5 watts of i-f drive would be needed. Despite these requirements, such a circuit provides a
local-oscillator injection required. In most relatively low-cost means of generating high-level microwave ssb. Spurious outputs at the LO and image
cases it is simpler to mix at a lower level. frequencies
output.
will be quite strong. To attenuate them a strip-line or cavity filter should be used at the mixer
StabU/QUion
LO
404-MHz Most vhf amplifiers, other than the
grounded-grid variety, require neutrali-
zation if they are to be satisfactorily sta-
ble. This is particularly true of AB I ampli-
PLATE
SUPPLY
fiers, which are characterized by very
high power sensitivity. Conventional neu-
tralization is discussed in chapter 6. An
BIAS
example is shown in Fig. 4A,
28-MHZ~
1- F
A tetrode tube has some frequency
IN ' where it is inherently neutralized. This is
likely to be in the lower part of the vhf
28
MHz regi'on for tubes designed for hf service.
VI
Neutralization of the opposite sense may
Q£w 7
6.3V
;L'
5 !,.:h.005
be required in such amplifiers, as in the
example shown in Fig. 4B.
Conventional, screen bypassing methods
may be ineffective in the vhf range. Series-
tuning the screen to ground, as in 4C, may
Fig. 3- Partial schematic diagram of a 70-cm (432-MHz) mixer, built from a converted fm transmitter. The be useful in this situation. A critical com-
original oscillator-multiplier-driver stages of the unit now provide LO injection. A strip-line filter should bination of fixed capacitance and lead
be used at the 'output of the mixer to prevent radiation of spurious products. length may accomplish the same result.
Neutralization of transistorized amplifiers
is not generally practical, at least where
narrow segment of a band, which the efficiency of the AB I linear amplifier is bipolar transistors are used.
transceiver encourages, is less than ideal low in a-m service, this type of operation Parasitic oscillation can occur in vhf
use of a major asset of the vhf bands - makes switching modes a very simple amplifiers, and, as with hf circuits, the
spectrum space, Separate ssb exciters and matter. Moving toward the high efficiency oscillation is usually at a frequency
receivers, with separate vhf conversion of Class C from AB I for cw or fm service considerably higher than the operating
units for transmitting and receiving, tend is accomplished by merely raising the frequency, and it canno( be neutralized
to suit our purposes better than the drive frc;>m the low ABI level. In ABI out. Usually it is damped out by methods
transceiver-transverter combination, at service the efficiency is typically 30 to 35 illustrated in Fig. 5. Circuits A and Bare
least in home-station service. percent. No grid current is ever drawn. As commonly used in 6-meter transmitters.
the grid drive is increased, and grid Circuit A may absorb sufficient fundamen-
Amplifier Design and Operation current starts to flow, the efficiency rises tal energy to burn up in all but low-power
All amplifiers in vhf transmitters once rapidly. In a well-designed amplifier it transmitters. A better approach is to use
ran Class C, or as near thereto as available may reach 60 percent, with only a small the selective circuit illustrated at B. The
drive levels would permit. This was amount of grid current flowing. Unless circuit is coupled to the plate tank circuit
mainly for high-efficiency cw and quality the drive is run well into the Class C and tuned to the parasitic frequen~. Since a
high-level amplitude modulation. Class C region, the operating conditions in the minimum amount of the fundamental
is now used mostly fot' cw or fm, and in amplifier can be left unchanged, other energy will be absorbed by the trap, heat-
either of these modes the drive h:vel is than the sm~1l increasing of the drive, to ing should no longer be a problem.
completely uncritical, except as it affe~ts improve the efficiency available for cw or At 144 MHz and higher, it is difficult to
the operating efficiency. The influence of fm. No switching or major adjustments of construct a parasitic choke that will not be
ssb techniques is seen clearly in current any kind are required for near-optimum resonant at or near the operating fre-
amplifier trends. Today Class ABI is operation on ssb, a-m, fm or cw, if the quency. Should uhf parasitics occur, an
popular and most amplifiers are set up for amplifier is designed primarily for AB I effective cure can often be realized by
linear a'mplification, for ssb and - to a service. If high-level a-m were to be used, shunting a 56-ohm\ 2-watt resistor across
lesser extent - a-m. The latter is often there would have to be major operating- a small section of the plate end of the
used in connection with small amplitude- conditions changes, and very much higher tuned circuit as shown .in Fig. 5C. The
modulated vhf transmitters, having their available driving power. resistor should be attached as near the
own built-in audio equipment. Where a-m plate connector as practical. Such a trap
output is available from, the 8sb exciter, it Tank-Circuit Design can often be constructed by bridging the
is also useful with the Class AB I linear Except, in compact low-powered trans- resistor across a portion of the flexible
amplifier, for only a watt or two of driver mitters, conventional coil-and-capacitor strap-connector that is used in some
output is required. circuitry is seldom used in transmitter transmitters to join the anode fitting to
There is no essential circuit difference amplifiers for 144 MHz and higher the plate-tank inductor.
between the AB I linear amplifier and the frequencies. U-shaped loops of sheet InstabIlity in solid-state vhf and uhf
Class C amplifier; only the operating metal or copper tubing, or even copper- amplifiers can often be traced to oscilla-
conditions are changed for different laminated circuit board, generally give tions in the If and hf regions. Because the
classes of service. Though the plate higher Q and circuit efficiency at 144 and gain of the transistors is very high at the
7·3 Chapter 7
CapaCitive reactance of C is chosen to be
AMP. very low at the parasitic frequency. R then
appears as a swamping resistor, damping
the oscillation.
VHF TVI Causes and Cures
The principal causes of TVI from vhf
144 MHz C1 transmitters are:
1) Adjacent-channel interference in
channels 2 and 3 from 50 MHz.
2) Fourth harmonic of 50 MHz in
channels 11, 12 or 13, depending on the
operating frequency:
3) Radiation of unused harmonics of
;J;0.001 P F the oscillator or multiplier stages. Examples
are 9th harmonic of 6 MHz, and 7th har-
BIAS B+ monic of 8 MHz in channel 2; 10th har-
monic of 8 MHz in channel 6; 7th harmonic.
of 25-MHz stages in channel 7; 4th har-
(Al
monic of 48-MHz stages in channel 9 or
AMP. 10; and many other combinations. This
,------.---------------------,144MHz may include i-f pickup, as in the cases of
24-MHz interference in receivers having
~"""'
21-MHz i-f systems, and 48-MHz trouble
in 45-MHz i-fs.
4) Fundamental blocking effects, in-
cluding modulation bars, usually found
2
of the tank inductor, thus forming Z3 and Z4. BIAS
The circuit at 0 illustrates how bypassing ,...-PARASITIC NETWORK
for both the operating frequency and lower fre- AMP. r------------,
quencies is accomplished. Low-frequency
oscillation is discouraged by the addition of
the 0.1/-1F disc ceramic cap,acitors. RFC1 and
RFC2 are part of the decoupling network used 50 MHz
i
I
I
,
CI~~
Zrn i
I
to isolate the two stages. This technique is not L __________ ...:...lI
required in vacuum-tube circuits. At E, a
capacitor with low reactaf'lce at the parasitic
frequency is connected in series with a
1/2-watt carbon resistor. At 144 MHz, C is
typically 0.001/-1F. R may be between 470 and
2200 ohms.
' '"'1=3 RFC
500~OUTPUT
,! ~ rh
sections of the transmitter, should' these
steps become necessary. {Bl
Use coaxial line to feed the antenna BIAS B+
system, and locate the radiating portion of AMP.
the antenna as far as possible from TV
receivers and their antenna systems.
A complete discussion of the problems
and cures for interference is in the ARRL
publication, Radio Frequency Interference.
A Linear Transmitting Converter
for 432 MHz
Linear transmitting converters offer
several advantages over frequency multi-
plication schemes for the 70-cm band.
(el
7.5 Chapter 7
L102
O.OOllFT
I
I
rhi RFC3
L203
Ll04
O.OO1!FT
PLATE
4000 CURRENT
~~J7
lOW
RFC4
AMP.
J3
1""
'~;:i~_
~ r' ~ ;ooV~8
I
EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPACITANCE ARE
r-h IN MICROfARADS I pF I; OTHERS
ARE IN PICOfARA OS I pF OR ... pf);
L204
~_____________________0~.OrO_I~/>F_T__~_V_4______________rIri(I(lL-__________~~FT RESISTANCES ARE IN OHMS;
k -1000. M-I 000000
;J;
A~P~ _____ -':3~: ~ _-,
I I___ PLATE
I I ENCLOSURE
I I
I
J5
OUT
1A
0-100
mA
FT • FEEDTHROUGH
Fig. 7 - Bottom view showing stages for V1 through VS. Fig. 8 - Bottom ViJw of T-44 cavity.
•
7·7 Chapter 7
ANODE END NO. 32 HOLES FOR
vice. This leaves two options for owners the tube (roughly 100 pF). The tube socket
OF CAVITY FASTENING C10 of to-watt-output rigs. One solution is to and strays add an additional 20 pF. This
" r---- 4 7/8
- "-----j use a solid-state "brick" amplifier to
drive the grounded-grid amplifier. This
tuning method places the tube input
capacitance in series with ,the tuning
/ - HM"r- o
results in complex relay-switching capacitor, effectively reducing the circuit
o o
114"
o
systems, not to mention the cost of the. capacitance. A number of other networks
LIPS
~~------~--~
amplifier and a high-current supply to were tried, including one somewhat exotic
~2-1/4"~ power it. The other approach is to use a current-feed method, but none proved as
TOP COVER tetrode such as the 4CX1000A in la simple and effective as the one shown.
(INCHES. 25.4=mm) grounded-cathode amplifier. In the circuit Bias voltage is shunt fed to the grid
described, 10 watts of drive will provide 2 through Rl. This resistor also heavily
t- 1 - 3/4' , kilowatts of input power, thereby ob- swamps the grid circl,Jit and helps to pro-
fH/8"
NO. 32
HOLES
D --±.
0
0
C10
viating an intermediate amplifier stage.
Circuit Description
vide amplifier stability without the need
for neutralization. Should it be desired to
increase the gain of the amplifier for use
The schematic diagram of the amplifier with lower power rigs, the value of Rl can
Fig. 9 - Layout details for modified top cover is shown in Fig. 16. As the 4CXI000A is a be incre;ised and the network components
and brass capacitor flap. Inches x 25.4 =. high-mu tube, care must be taken in the changed as appropriate. The cathode is
mm ..
design and construction of the amplifier grounded through short, heavy leads at
to prevent instability. In the circuit the base of the SK-800B tube socket. A
shown, this was accomplished without the screen bypass capacitor is an integral part
need for neutralization - one of the of this socket. The capacitor is 1500 pF, is
drawbacks commonly associated with the rated for 400 volts and is of the mylar-film
use of these tubes. variety. Capacitors are connected between
the chassis. All the perforated holes above The input network is series tuned and a ground and each of the three filament-
the top of the chassis must be covered with link couples power from the exciter to the connection points.
masking tape to make the top portion of tube. Series tuning was chosen primarily The plate circuit used in this amplifier is
the cabinet airtight. Directly beneath the because of the large input capacitance of of the common pi-network variety, with
4CX250B tube socket, a large hole is
punched in the bottom of the cabinet for
an air entrance. The photograph of the
amplifier shows that the 4CX250B does
not have a chimney. It was later found
that the chimney must be used to provide
adequate cooling. A 4-inch diameter hole
is cut in the back panel of the cabinet and
a 5-inch Roton Whisper fan is mounted 0-500
over the hole to exhaust air from the
cabinet. EXCEPT AS INDICATED, DECIMAL
Air flow is through the bottom of the VALUES OF CAPAC ITANCE ARE
IN MICROFARADS (JlF); OTHE~S
cabinet, through the socket of the ARE IN PICOFARADS (pF OR JlJIF)j
4CX250B, through the chimney into the RESISTANCES ARE IN OHMS;
anode and out of the cabinet through the k '1000.104-1000000
Results
EXCITER
A suitable power supply is shown in
Fig. 14. With 2000 volts on the anode and
a plate current of 250 rnA, the power out- T-R
~
+3~
watts. This agrees closely with the tube
~9
specification sheets.
A 2·kW PEP Amplifier for 50 to 54 MHz
2 REci~LATEW
A number of manufacturers sell low-
power-output (10 watts) transceivers and
transverters for 6-meter operation. This
Fig. 10 - Schematic diagram of the 500·W amplifier for 6 meters. Unless otherwise specified, .
particular power-output level is not in line capacitors are disc ceramic and resistors are carbon composition. (Inches x 25.4 = mm).
with that required to drive the popular C1 - 50 pF, receive spaCing or aux. relay if only spst option available.
grounded-grid amplifiers. Even the high- C2 - 25 pF, 3 kV (surplus cap. in unit). L1 - 6 turns no. 14 solid wire, 1/2·inch dia,
mu triodes, such as those of the 8874 C3 - 140 pF, receive spacing. 1·1/4 inch long. Tap 1·1/2 turns from gnd end.
family y require at least 25 watts of drive C4 - Cer. cap. 2 paralleled 500 pF, 5 kV. L2 - 5 turns no. 10 solid wire, 1-3/88 dia, 2
K1 - Opdl' relay, 12-V coil (can have dclpower inch long (see text).
for 1 kilowatt of input. Over 50 watts of type contacts but rf design preferable). RFC1 _ 35 turns no. 22 enam. wire on 5/8·
drive may be required for 2-kilowatt ser- K2 - Opdt relay. Either T-R contacts in exciter inch dia cer. ins.
+324V
SCREEN
7·9 Chapter 7
._ _ • ___ . ~T~ ~~E~
50-54NHz
C4 *I
I
VALUES OF CAPAC ITANCE ARE
IN MICROFARADS (jlF) ; OTHERS
ARE IN PICOFARADS (pF OR ~jlFI;
r77 RESISTANCES AR£ IN OHMS;
k -1000. M-IOOOOOO
4CX1000A
V1
RFC'
-r C7 .Q,QQ!. * SEE PI\oRTS LIST
GRID ENCLOSURE""
r77 5kV
1•
~------------------
50-54 MHz
+--.....---,- - -,
222~ R1
J3
+HV
PLATE
CONTACT
AREA
RFC3 POWER
LINE ____ SI
~~N
rh lk~
IIiV
CONTROL
e NC
. .
eNC.
C25 O.OI/lkV
RS >o------.....
25.
-'V'I/'v-....----i
Tw
STANDBY
J5
- HV
J6
Fig. 16 - Schematic diagram of the 6·meter amplifier. Parts designations shown on the schematic but not called out in the parts list are for text
reference only.
Cl - Miniature variable, panel mount, 32 pF 014 - Zener diode, 30 volt, 1 watt. Ml - 0-1 mA, Simpson 15070 or equiv.
maximum. OSl - Neon indicator built into Sl. M2 - 0-1 A, Simpson 15101 or equiv.
C2 - Silver mica, 15 pF, mounted directly Fl - Fuse, 1 ampere. Ql - 2N2905A or equiv.
. across Cl. Jl - Coaxial connector, BNC chassis mount, R4 - Meter shunt, l' 7-1/2" (41.4 mm) no.
C4 - Tube internal plate capacitance. UG-625B/U or equiv. 34 enam. wire wound on a high-value, 1/2-
C5, C7 - Transmitting capacitor, 0.001 ,..F, J2 ~ Coaxial connector, type N chassis watt resistor.
5000 volt. Centralab 85SS-1ooo or equiv. mount, UG-5S/U or equiv. R5 - Potentiometer, 25k-ohm, 5 watt.
C6 - Transmitting variable, 250 pF maximum. J3 - High-voltage connector, Millen 37001 or R7 - Meter shunt, 4.3 ohm, 1/4 watt.
CS - Screen bypass built into SK-SooB socket. equiv. RFC1 - 36 turns no. 24 enam. wire on a 1"
C20, C24 - Electrolytic, 40 ,..F, 450 volt J4 - 4 conductor. (25.4 mm) diameter Teflon rod.
01,02,010-012, incl., 015, 016 - Silicon, 2.5 J5, J6 - Phono connector. RFC2, RFC3 - 6 turns no. 22 enam. wire on an
ampere, 1000 volt. L1 - 2 turns no. 24 enam. wire wound over L2. FT-50-43 core.
03 - Zener qiode, 45 volt, 1 watt. L2 - 10 turns no. 24 enam.wire on a Sl - Spst, rocker type with built-in neon
04-09,.incl. - Zener diode, stud mount, 56 T-50-12 core. indicator.
volt, 10 watt. L3 - See text and drawing. S2 - Rotary, 2 pole, 2 position.
013 - Zener diode, 200 volt, 5 watt. T1 - Stancor PS356 or equiv. See text.
* 3-1/2 TURNS
2.3/4 11 X 3-114" 114" COPPER TUBING
DELRIN BLOCK WOUND ON A 1- 3/4 II
11
2-3/4 X 2-3/4" 2-:3/4"X 3-1/4"
DERLIN BLOCK DELRIN BLOCK
CENTER HOLE PANEL BUSHING
TAPPED 1/4" X 28 CENTERED
Fig. 19 - Dimensional drawing of the inductor assembly. Delrin can be obtained from most
plastic supply houses. Check the Yellow Pages of your local telephone directory for dealers.
7·11 Chapter 7
0 0
0 0 a ~
0 0
, •
a
.r-~
Fig,22 - Full scale etching pattern for the printed-circuit board. Black areas represent unetched copper.
TO S2A
AND GRID
TO C18
04·09
TOC15
Fig. 23 - Parts-placement guide for the printed-circuit board as shown from the component side.
components and the control circuitry_ for the diodes is supplied primarily by the heater/defroster hose. A length of this
Many of the parts are contained on a system blower, as the diode stack is hose connects the pressurized chassis to
single-sided circuit board_ The board mounted in the air flow path. Cooling for the blower, which can be mounted at
etching pattern and parts-layout diagram the tube is by means of a blower that is some convenient location. For ultra-quiet
are given in Figs_ 22 and 23. Zener diodes mounted external to the chassis. A plastic operation the blower can be mounted in a
are mounted on an aluminum plate and flange (Newtone 366, used with central closet or in an adjoining room. Copper or
tJIe circuit board. While the aluminum vacuum cleaning systems) is suitable for brass ~creening is used to cover the flange
provides some heat-sink action, cooling use with 2-inch (51-mm) automobile opening, thus qIaintaining an rf-tight.
7·13 Chapter 7
144 MHz J3 J4 J5
FLI
,~.~~:'
OUTPUT
L5 L6
Jl
E~CEPT AS INDICATED, DECIMAL VALUES OF
5pF
CAPACITANCE ARE IN MICROFARADS I JlF ) ;
5'h"
OTHERS ARE IN PICOFARADS (pF OR JlJIF);
RESISTANCES ARE IN OHMS;
.~
•• '000, M"OOO 000.
VI C25
144MHz 4CX250
Ll
0.01
T1 RI-RI2-390k SCREEN
CURRENT
01- 012 -1000PRV/2.5A INDICATOR
SET
Cl- C12 - O.01)1F/l000V
II
0.02
013
1000PIV
T2.~
0.02
117 VAC
0.02
a-A FUSE
Fig. 26 - Schematic diagram and parts list of the 144·MHz amplifier. Inches x 25.4 = mm.
B1 - Blower, 15 ft'/min. . 014,015 - Light emitting diode (VF = 1.6 V, R13, R26 - 5kO, 2-W potentiometer.
C1-C12, incl. - 0.01 ,..F disc ceramiC, 100-Vdc. IF = 60 mAl. R14-R19, incl. - 39kO, 10-W.
C13 - 100 ,..F, 350 Vdc. 016 - 33-V, 5-W Zener diode. R2O-R25, incl. - 2OkO, 25-W.
C14·C19, incl. - 200,..F electrolytic, 450·Vdc. 017-022, incl. - 56-V, 10-W Zener diode. R27 - Meter shunt, 0.05555 ohms (3.375 feet
C20, C22 - 15-pF air variable, Hammarlund F1 - 10-A fuse. (1.03 m) no. 22 enam. wire wound o.n any
HF-15-X or equiv. J1, J3, J4, J5 - Type SO-239. large-value, 2-watt resistor). .
C21 - 5-pF air variable, Hammarlurid J2 - Phono jack, panel mOUnt. RFC1 - 20 turns no. 24 enam. wound on 100k,
HFA-25-B with 2 middle rotor plates and two L1, L2 - See text and Fig. 45. 1-W resistor.
end stator plates removed, or equiv. L5, L6 - 3 turns no. 18 tinned, 1I4-inch 10, S1 - Spst, 10A.
C23, C24 - 5-pF air variable, E. F. Johnson 3/8 inch long. T1 - Primary 117 V ac, secondary 1250 V ac,
160-0104-001 or equiv. L3 - 3-1/2 turns no. 10, 1-1/4 inch 10,2 inch 500 mA, Hammond 720 or equiv.
C25 - 0.001 ,..F, 4-kV feedthrough capacitor, long. T2 - Primary 117 V ac, secondary 125 V ac,
Erie 2498 or equiv. L4 - 1 turn no. 14 enameled, 1-inch ID. 50 mA; 6.3 V ac, 2.0 A, Stancor PA-8421 or
C26 - Screen bypass capacitor built into M1 - 0-1 mA meter. equiv.
Eimac SK-630A tube socket. R1, R12, incl. - 390kO, 1/2-W. V1 - Eimac 4CX250B.
01-013, incl. - 1000 PIV, 2.5-A silicon diodes.
Construction prevents air leakage. The grid circuit electrical contact to the screen. A stan-
should be installed so that no other circuit dard Eimac ceramic tube chimney is
Despite the compact design, few pre- is in close proximity to the tuned line, link mounted on a wooden standoff. The
cautions are necessary. 'Caution with or tuning capacitor. Cooling air is blown chimney fits over the tube anode· in an
respect to high-voltage leads is mandatory, into the plate compartment through a inverted style. The hole in the top cover is
bf course. The plate circuitry is entirely screened hole. Several screw, nut and screened in the same way as is the fan
enclosed by a shielded box, which also flat-washer combinations guarantee good inlet. Cool air enters through the fan hole
Chapter 7 7·14
Fig. 27 - Top view of the 2-meter amplifier.
Voltage-dropping resistors and Zener diodes
may be seen on the left-hand side of the
.chassis. The small transformer next to the
plate compartment is T2. Details of the wood REFLECTOMETER
spacer may be seen at right. Jl
144MHz
INPUT Ll
L2
0'
I
L--:- __________________
200 10k
lOW 25W
02
CANaOE
LZ (NO.IS ENAMELED WIRE) CURRENT
lN33n
(8) 12V --=---",--:1
E)(CEPT AS INIiHCATED, DECIMAL
50W TO vox RELAY
VALUES OF CAPACITANCE ARE CONTACTS
IN MICROFARADS (JiF 1 ; OTHERS
Fig. 28 - Formation details ,for L 1 and L2. ARE IN PICOFARADS (pF o~ joIjlF);
RESISTANCES ARE IN OHMS; B-
k -1000. M-' 000000
B.M.·~BUTTON MICA
Fig. 33 - Schematic diagram of the 144-MHz amplifier. Included is information for the input reflec-
tometer used as an aid to tuning the cathode circuit for low SWR. C7, C8 and C9 are fabricated as
described in the text and Fig. 32. Inches x 25.4 = mm.
B1 - Blower, Fasco 59752-IN or Dayton 2C610. 82-62 or equiv.
Wheel diameter is 3-13/16 inches. J5 - HV connector, James Millen 37001 or
C2 - 5- to 30-pF air variable, Hammarlund equiv. ,
HF-30-X or equiv. L1 - Double-sided pc board, 1-1/4 x 4-7/16
C3, C4, C5, C6 - 0.1 /IF, 600-V, 20-A inches.
feedthrough capacitor. Sprague 80P3 or L2 - 4-1/4 inches of no. 18 wire. L1 and L2 are
equiv. part of the input refiectometer circuit.
J1, J2, J6 - Type BNC. L3 - 6 turns no. 18 enam., 5/8-in. long on
J3-TypeN. , 3/8-in. dia form (white slug).
J4 - Coaxial panel jack, UG-22B/U, Amphenol L4 - 3 turns no. 14 enam., 5/8-in. long x
The output traps are built into a box thoroughly checked for mistakes. The high
Fig. 29 - Bottom view of fhe amplifier. A thin fabricated from copper-clad circuit board. voltage present can be lethal if nottreatf;d
sheet of Teflon has been placed between the The easiest approach to mounting it is by- carefully. Be sure no wires are touching
electrolytic capacitors and their aluminum means of a double-male UHF-type con- anything they shouldn't.
mounting strap. At upper right is the power Primary voltage should initially be
supply rectifier board. Details of the grid cir-
nector to the jack on the rear of the plate
cuit are also visible. compartment. applied through a variable line transfor-
An input attenuator may be required to mer. This allows the voltage to be brought
reduce transceiver output power down to up slowly so that if something is awry it
the 2-watt level necessary to drive the can be noticed before any damage can
and passes through the tube anode cooler. amplifier. Standard composition resistors occur. Once the primary has been safely
Tube-warmed air exits through the chim- can be used in attenuator construction. brought up to its normal voltage, voltage
ney and out the screened hole in the top. The attenuator is built in a separate box checks on the tube should be made. Screen
A small portion of the cooling air passes from the amplifier. voltage should be about 325 and bias
through the tube socket, cooling the tube should be about -130 V. AC filament vol-
base before flowing through the underside Tune-up and Operation tage should be about 6.
of the chassis. Upon complet~on, all wiring should be Shorting 12 places the amplifier in the
~
IW
04
1N4_
05
0.001
transmit mode. With no drive applied, and disconnect the temporary load resis-
adjust bias for an idling plate current of tors.
SO MA. This establishes class of operation Connect a source of drive to 1 I through
AB2. Adjustment of the warning-LED an· SWR indicator. A 50-ohm dummy
potentiometers requires the use of tem- load should be connected to 13. Applying
porary load resistors. Remove power a small amount of drive, adjust grid
Fi.g. 32 - The tube and plate line is in place,
from the amplifier. Temporarily connect a capacitor C for a dip in SWR on the WIth the top and side of the compartment
150-kO resistor from the grid terminal to indicator. The SWR may not be close to removed for clarity. The plate-tuning vane is at
ground. Turn on the power. Do not short I: L If not, readjust the position of L 1 with bottom center. A bracket is attached to the
12. Adjust R 13 until the grid-sensing LED respect to L2. Recheck the SWR. Continue side panel to support the rear of the Teflon'rod
supporting the tuning vane. The coil at the op-
just comes on. Turn the power off again. the process until the input SWR is close to pOSite end of the plate line is RFC1 connected
Connect a string of 10 1Ooo-ohm, I-watt 1: I. Since no voltage is applied to the am- between the high·voltage·bypass pl~te and the
resistors from the screen terminal of the plifier, adjustment is very simple. In the, top section of the plate-line sandwich. Items
tube socket to ground. Turn the power on amplifier shown, the Li adjustment was outside the tube enclosure include the filament'
transformer, blower motor, relays, and a power
again. Adjust R26 until the screen sensing optimum when placed approximately 1/8- supply to operate a VOX-controlled relay
. LED just comes on. Turn the power off inch (3-mm) from and parallel to L2 . system.
7-17 Chapter 7
one-turn link attached to the rf output
connector. Top and bottom covers are
~~fi~\\;'t .
Table 2
Pad Values for Input AUenuator
Attn (dB) R1 R2 R3 (ohms)
then secured. As with all cathode driven
amplifiers. excitation should never be
applied when the tube heater 'is activated
'm·'·····;,'· iii"'".....
~ "" ! +~ ~
j
f'lI ....
o none and plate voltage is removed. Next, turn on
6
10
18
'27
18
27
68
39 the tube heater and blower simulta-
neously, allowing 90 seconds for warm-up. r;.t.
~:;
20 43 43 11
A plate potential between 2400 and 3000
volts then may be applied and its presence ;.
verified on the multimeter. The power
supply should be able to deliver 800 mA •
•
or so. With the VOX relay actuated,
resting current should be indicated on the Fig. 34 - Front·panel layout of the 220·MHz
Table 3
cathode meter. A small amount of drive is, kilowatt.
Performance Oeta applied and the plate tank circuit tuned
Power input, walts 1000 1600
for an indication of maximum relative
Plate voltage 2600 2450 power output. The cathode circuit can
Plate current (single tone) 385 rnA 660 rnA now be resonated, tuning for minimum
Plate current (idling) 50 rnA 50 rnA reflected power on the reflectometer, and
Grid bias -10V -10V not for maximum drive power transfer.
Grid current (single tone) 35 rnA 54 rnA
Drive power, walts 1,8 41 Tuning and loading of the plate-tank
Efficiency (apparent) 59.5% 61.8% circuit follows the standard sequency for network. Medium values of Q were cho-
Power gain (apparent) 15.2 dB 13.9dB any cathode driven amplifier. Resonance sen to provide high efficiency. Both the
Power output, walts 595 1000 is accompanied by a moderate dip in cathode and the heater are operated at the
plate/cathode current, a rise in grid same rf potential; the heater is held above'
current and a considerable increase in rf ground by the impedance of the filament
relative power output. Plate-current dip is choke. The plate tank is a pair of quarter-
not absolutely coincident with maximum wavelength striplines placed symmetrically
Table 4 power output, but it is very close. Tuning about the tube. 2 This arrangement per-
Operating Conditions and output-loading adjustments should be mits a more uniform flow of current
432 MHz 432-MHz for maximum efficiency and output as through the anode, preventing "hot spots"
drive power output power indicated on the output meter. Final on the anode conducting surface. Addi-
watts watts adjustment for lowest VSWR at amplifier tionally, tube output capacitance is effec-
2 30 input should be done when the desired tively halved, as one-half the tube capaci-
4 50 tance (13 pF) is used to load each stripline.
5 80
plate input-power level'has been reached.
7 100
Striplines act as low-pass circuit elements
9 140
A 220-MHz High-Power Amplifier even with the high unloaded-Q conditions
Ep-l000V
Circuits for 220-MHz power amplifiers found at 220 MHz. Linear inductors also
Ip - 60 rnA, zero signal. have long been designed around the offer contr~l of odd-mode harmonics. No
Ip - 300 rnA, single tone (cw), 140 Woutput. external-anode tetrode. While these tubes spurious responses could be found in this
offer high gain, instability problems have amplifier up through the 900-MHz region.
caused many builders considerable con- A strip-line impedance can be varied by
sternation over the years. Multiple- changing its width and relation to its
tube amplifiers are often necessary to ground planes. Physical dimensions of the
obtain the high power levels many tube limit the position of the stripline
moonbouncers and weak-signal specialists above one ground plane. In order to
require. Push-pull amplifiers have been utilize cOll}mercially available chassis, the
lOoo-watt output point and worked tried with moderate success, and recently .stripline was placed 1-1/4 inch (32-mm)
backwards to calcur'!te apparent stage parallel-tube designs have found favor" above one side of an inverted 4-inch
gain and efficiency. Modern computer-aided tube designs (l02-mm) high chassis. This means that
Efficiency measurements also were have brought forth high-" triodes such as approximately 75 percent of the rf current
made employing the "tube air-stream the 3CXI500A7/8877, a 1500-watt dis- flows through the chassis, but only 25
heat-differential" method. Several runs sipation external-anode triode with maxi- percent flows through the top shield
were made at 885 watts static dc and mum ratings good through 250 MHz. The cover. The small percentage flowing
normal rf input. Apparent efficiencies of ceramic insulation allows a heavy flow of through the top reduces the effect of any
62 to 67 percent were noted. These values rf current through the tube, with no loss mechanical anomalies associated with a
were about five percent higher than th~ of stability in a properly designed circuit. . removable cover.
actual power output values given in Table Low heater requirements (5 V at to.5 A) Fbr quarter-wavel~ngth lines, the ratio
1. Both efficiency measurement schemes add to the appeal of the 8877. This ampli- of line impedance to reactance should be
serve to confirm that the amplifier is fier. employs the 3CX 1500A7/8877 in. a between 1.5 and 2.0 for the best band-
operating at the upper limit of the cathode-driven circuit. The grid is grounded width. Taking stray capacitance into ac-
theoretical 50 to 60 percent efficiency directly to the chassis, adding to the stabil- count, expected tuning capacitance and
range for typical Class AB2 amplifiers. ity. The amplifier (Figs. 34 to 38) is un- tube output capacitance gives a value of
To commence rou.tine operation, the conditionally stable - more so than some 55 ohms for Xc. Values of line impedance
variable capacitor in the input circuit amplifiers built for the hf region. versus line length for resonance at 222
should be set at the point where lowest MHz were computed on a programmable
input VSWR was obtained during the Circuit Details calculator for impedances between 30 and
"cold-tube" initial tune-up. The ability of The input circuit consists of a T
the plate tank to resonate at 144~145 MHz
'Barber, Rinaudo, Orr and Sutherland, "Modern
with the top cover in place should be IKnadle, "A Strip-Line Kilowatt Amplifier for 432 Circuit 'Design for VHF Transmitters," CQ,
verified with a grid-dip meter, via a MHz," QST, April and May 1972. November and December 1965. '
~
Jl: I
I
I. 500
I-- :,..--CATHOOE ENCLOSURE
I S.M.
I I
I I
I
(C~;+; ~
r--------------------------, I
+ PLATE
03 220
I
VOLTAGE I
I
21-----'
6801
3f-------' I
4 I
I
5f------.J 500,uF I
61---------t
'25V
~~~§=~~~~Tl ;
- - - - - - - - - - - - - - - - - - - - - - - - - -~
I
02
Fig. 35 - Schematic diagram of the 220·MHz amplifier. Unless otherwise specified. all capacitors are disc ceramic and resistors are 1/2-watt carbon
composition. Inches x 25.4 = mm.
C1 - Air variable. 15 pF. J3 - High-voltage connector (Millen). RFC2, RFC3 - 10 turns no. 18 enam. bifilar
C2. C3 - Button mica, 500·pF, 500-V rating. L1 - 3 turns No. 14, 1/4-inch 10, 3/4-inch long. wound on 3/4-inch Teflon rod close wound.
C4-C9, inclusive - Teflon capacitor (use 10-mil L2 - 1I4-inch wide, 2-3/8 inch long copper RFC4 - 5. turns no. 16 enam. wound on 1-MO,
Teflon sheet). flashing strap. 2-watt composition resistor.
C10 - Doorknob capacitor, 500 pF, 5-kV rating. L3 - Plate inductor (see Fig. 38}. T1 - Filament transformer, 5.0 V at 10.5 A.
01-04 - 1000 PRV, 3A. RFC1 - 8 turns no. 18 enam. 1/2-inch dia.,
J1, J2 - Coaxial receptacle, type N. 3/4-inch long.
lOO ohms. These were plotted on a graph. The plate blocking capacitor consists of along the same principles. A piece of
Final dimensions were determined using a sandwich of brass plate and the stripline, circuit board was once again sandwiched
this system, choosing dimensions that fell with Teflon sheet as the dielectric. This with Teflon sheet to the side wall of the
into the middle of the graph, thus forms a very low-loss, high-voltage capa- chassis. This technique is used effectively
allowing for any unpredicted effects. citor. The plate bypass capacitor is built throughout as an inexpensive bypass or
7·19 Chapter 7
1~---1'-3/~~~
Q-1!8" it
-r
2 /2
~L'
I
"1 _ _ _-.. ' 2-1/4" 1
16 1/4"
~H/2'~
L3
MAT~RIAL 01/16" BRASS
1'"2-1/2''1
--.
D
(2) BLOCKS TO SUPPORT
':J.4"
ENDS OF L3
MATER IAL .1/4" BRASS
II-I__~
_1/2"
~ ~I~T
IJ
,--1-1_311_6"
I• IO-3/4'~'-~-1
PLATE FOR C6
MATERIA L -1116" 8RASS
1 2 "1
r- 3 - 1/2 "--!
Fig. 36 - Bottom view of the amplifier. RFC2 and RFC3 can be seen above tube socket (bifilar
winding). Copper strap is L2 shown connected to C1. Small coil is L1 and larger coil is RFC1. The
grid of the tube should be grounded to the chassis with finger stock similar to that used in the
plate line. Component mounted on the heat sink at left is the Zener diode used for biasing
ltt'O
L
(2) PLATES FOR C8
ED
(2)PLATES FOR C7
MATERIAL' 1116" BRASS MATERIAL "/'6" BRASS
purposes.
~
HUTGRIO
10
Equipment for 432 MHz varies in style, 8
size, complexity and ancestry. Some 6.3V
-H~
BIAS
The design criteria for a desirable ADJ.
amplifier were simple - a table-top FT· FEEOTHROUGH
10k 1000
conduction-cooled (quiet) unit that would
deliver lOO-watts output at a drive level of
J5
T/R
02 ..1.QL
!lW 100 ~
~~"
03
less than 10 watts. The table-top configura-
tion would be more attractive 'to many 20V ",'-'"
5W
station owners than would the old reliable
+ 300V R2 ~ 10 9
rack-and-panel system of days gone by.
3/100k 01
OA2 OA2 ~
'Moretti, "A Heterodyne Exciter for 432 MHz,"
QST, November 1973 (also see Feedback, QST, 3
l'farch 1974, page 83). +2!10 TO 300V
..l.Q2.Q.
2!1W
7·21 Chapter 7
protuberances to worry about behind the
cabinet. A standard-size chassis is used to
COPPER
fill the gap between the panel and the
HEAT
SINK I
BLOCK
BeO THERMAL
liNK
amplifier proper, and incidentally to
provide mounting space for peripheral
\ '" '"
electronics. As long as the parts placement
~560A f.4-------~---------7-1I2"--- ----= ----l
..1.
within the amplifier grid and plate
~ (t90mm) compartments is not changed from the
"" ,-L2 design given here, it will not matter what
,C3 is done externally.
~
~
~
.. C4 EtP
1 (127 X 178 X 50-mm) aluminum chassis
I I
.
~ 0 0
/1-1/2"
1 ! with captive nuts in the bottom lip to
permit securing the bottom plate. For the
e-- /~~~~~T ECCENTRIC
plate compartment a 5 X \0 X 3-inch (127
r ~
X 254 X 76-mm) aluminum chassis was
I I C~ @.-Yt
modified to provide better mounting
II
surfaces for the heat sink and to allow the
LL9
'-V
Tube Placement
means of a capaci'tive ~robe to the grid probes,7 it ~as determined that the tube An Eimac SK-630 socket and SK-1920
line. A small tab' of copper is soldered. to would stay within maximum temperature' thermal link are used in mounting the
the grid line and forms one side of the ratings while dissipating 100 to 200 watts tube and conducting the heat away from
capacitor. A disk on the center conductor of dc. A liberal coating of thermal- the anode. The thermal link is made of
of a coaxial section is"the movable portion conducting grease was used to aid heat toxic beryllium oxide (BeO). The manu-
of the coupling. This coaxial section is transfer. More on this subject later. facturer's caution against abrasion, frac-
fabricated from pieces of brass tubing that Early tests with only dc applied, and tures or disposal should be heeded. Parts
will slide together, telescope fashion. A later ones with the full dc and rf voltages placement in the anode-block area is
BNC chassis-mount fitting with the present, confirmed that at the IOO-watt critical if efficient heat transfer and
threads filed down is soldered into the output level no forced-air cooling was minimum strain on the tube are to be
inner, movable piece of tubing to allow required. At higher output levels of 175 to obtained. The tube socket must have
ease of connection from the exciter. A 200 watts, the temperatures on the anode sufficient clearance in its mounting hole
piece of copper wire and a couple of and heat sink were still below the that some lateral movement toward or
Teflon discs extena the center conductor maximum allowed by the manufactuier, away from the heat sink is allowed. The
for attachment of the capacitor plate but high enough that it was felt prudent to socket is secured to the chassis with the
inside the grid compartment. Once ad- add a quiet "whisper" fan for safety. usual toe clamps supplied. Because of the
justed, the sliding portion is held in place Operational tests proved that the added . rim formed on the socket by the integral
by means of a small compression clamp. background noise was not distracting to screen-bypass capacitor, a spacer is
The plate line is the familiar haif- the operator. needed between the thermal link and the
wavelength variety, with capacitive tuning heat sink. A piece of copper, 1/4-inch
provided by movable vanes or "flappers." Construction (6.3-mm) thick and about 2-3/4 X 4-1/2-
In earlier versions using this tuning There are several configurations pos- inches (70 X 114-mm) serves as the spacer,
scheme, the flappers were moved by sible for the package, and the constructor as well as providing excellent heat transfer
means of string that was allowed to wind should feel free to mold them to fit his to the inner face of the heat sink. This
or unwind around a shaft, providing idea of how things should be assembled. copper spacer and the BeO thermal link
front-panel control. After a few instances An LMB cabinet (CO-I) was selected for are both held in place between the tube
of loss of control, caused by the nylon an enclosure because it matches many of anode and the heat sink by the pressure
fishing line melting or becoming untied, the "gray boxes" found in a lot of shacks. applied by the ceramic pillars. The anode
the writers decided that there had to be a Rather than mount the heat sink through end of the plate line is bent up to form a
better way. Accordingly, the cam-on- an unsightly hole in the rear panel of the surface that will permit screws to thread
a-rod method was tried and found satis- cabinet, it was decided to mount the into the insulators. In the early version of
factory. Both plate-tuning and output- amplifier parallel to the front panel. This the amplifier this shaped and bent piece of
coupling flappers are adjusted in this places the heat sink inside, but there is aluminum, was only long enough to
manner (Fig. 42). adequate ventilation through the box to provide some mounting surface to which
allow proper cooling. This mounting the plate line (double-sided pc board in
Cooling scheme also permits a fan to be mounted this instance) was fastenep by means of
Several tests were performed to check inside, so that there are no awkward five no. 6-32 screws and nuts with lock
the effectiveness of the thermal-link/heat- 'Temprobes"'Test Kit, by Tempil~ Hamilton Blvd. washers. The photograph shows this
sink cooling system. With the aid of Tem- South Plainfield, NJ 07080: particular scheme in the top view. A later
7·23 Chapter 7
B+
J3
R7 RS
\~~
IW
I
I
I
--------1
I
I
RELAY CONTACTS'
I
I NO - NORMALLY OPEN
'------ NC - NORMALLY CLOSED AMPLIFIER
RF SWITCHING
(CONTACTS
R3
NOT SHOWN)
=TI
21 V/50W
04
+----I'~~>--_-o-U
K~ ..--o"C
_. 52 T2
II
EXCEPT AS INDICATED, DECIMAL VALUES OF
CAPACITANCE ARE IN MICROFARADS (JlF) ;
OTHERS ARE IN PICOFARADS (pF OR ~~F I;
RESISTANCES ARE IN OHMS;
k-IOOO, 1'-1000 000.
£l
FI
3A
J4
B-
"7
NEUTRAL HOT
117 117
VAC YAC
functions as the plate-bypass capacitor. is switched in to supply near-cutoff bias With the ~ception of the multi meter
The half-wavelength cathode line is com- during standby periods. MI is used solely functions the metering and bias circuits
prised ofW2, L2 and C2. L1 and CI serve to monitor plate current in the high- are similar to those in the 220-MHz
to match the tube input impedance to the voltage supply negative-return lead. M2 is amplifier. 10
amplifier 50~ohm input. As the grid is switched to read grid current, high voltage
'McMuJl!n, "The Line Sampler," QST, April 1972.
grounded for dc as well as rf, DI is used to and relative output. The latter function is "Sutherland, "High Power Linear Amplifier for 220
develop operating bias at the cathode. R3 by means of an external line sampler. 9 MH1," Ham Radio, December 1971.
~ "'(I)
dJ
TUBE SOCKET
HOLE 1'1I4'"0IA
1-118"
-L§=1'1/:~1/4. 6-3/4"
8·4"
7·25 Chapter 7
1-II2-inch (38-mm) copper pipe. This underdrive an amplifier, of this type,
~z'-I pipe has .an outside diameter of 1-5/8 resulting in an efficiency of only 20 to 30
inches (41 mm) and should fit snugly into percent. If a wattmeter is not available,
T the hole. The Teflon chimney will be held tune for maximum forward power as in-
~/.~ 2' . firmly in place and no air should leak dicated on the line sampler. For one-
from the box without passing through the kilowatt input, drive power must be at
n~L
D~' '/<;~:::"'- __
'1
...J~2·
anode cooler.
Operation
' least 20 watts. Grid current will vary from
30 to 70 rnA as a result of tube dif-
ferences. Less grid current is a sign of in-
CATHODE LINE Adjust R9 to place maximum resistance sufficient drive and too much loading.
NO. 29 ---.l in series with the tube heater. Apply Decreasing the load capacitance should
0.136' h-1/4'1
heater power and .allow two minutes for result in reduced plate current and in-
m .. 'INCHES.ZIt.4 D~' the element to reach operating tempera- creased grid current. Slight readjustment
LOADING ture. Now energize K2 and adjust R9 to of the' plate-tuning capacitor should also·
place 5.7 volts at the socket pins. Apply be required. Table 5 contains a summary
Fig. 50 - Dimensions of the cathode line and plate voltage (about 2000 volts). Idling of amplifier operating conditions at the
the flappers used to tune and load the plate plate current should be approximately 30 one-kilowatt level. If 40 watts or more of
circuit are given here. Additional information is excitation are available, the cw plate effi-
contained in the text.
rnA. Apply drive and adjust its level to
bring the plate current up to ISO rnA. Ad- ciency can be boosted to about 59 percent
of 1-5/8-inch (41-mm) 00 copper pipe just C6 (plate tuning) for maximum out- by means of increasing the bias to 21
was used as a form to make the chimney. put. Input capacitors CI and C2 may then volts. A switch can be installed on the
:rhe Teflon is held together with RTV be coarsely adjusted for maximum plate front panel to select another diode in
(room-temperature vulcanizing) adhesive. current. Simultaneously increase drive place of 01 when operating cwo
Air in the plate compartment must now and adjust plate tuning and loading for
flow through the anode cooli'ng fins to maximum output until input power Conclusions
escape. The air outlet is built on a reaches one kilowatt or the desired level. From a cold start, the amplifier reaches
2-1/4-inch (57-mm) square copper plate. The input circuit may be adjusted for full output five seconds after dri ve is ap-
A I-SIS-inch (41-mm) diameter hole is minimum reflected power when the pro- plied.' After the first transmission, full
made in the plate and a piece of copper per drive level is established. output is obtained in one second, with no
window screening is soldered over it. On An accurate wattmeter is strongly furtht!r drift noticed. This amplifier has
the side opposite the screening is soldered recommended for use during initial ad- been operated for several months without
a 3/8-inch (9.5-mm) long piece of justments. It is very easy to overload and need for retuning.
Receiving Systems
How good should receiver performance being offered by some manufacturers, and . practical receiver operates. Design data
be? A suitable answer might be, "As good although performance is acceptable in and related philosophy are included for
as is pos~ible, consistent with the state of many instances, there is considerable those who are inspired toward developing
the amateur art and the money available room for improvement. At least there is a homemade receiver.
to the purchaser." That opens up a wide only one mixer to cause intermodulation
area for debate, but the statement is not distortion (lMD) and overloading prob- Sensitivity
meant to imply that a receiver has to be lems in a single-conversion superhetero- One of the least understood terms
costly or complex to provide good dyne receiver. A strong doubly bal- among amateurs is sensitivity. In a casual
performance: Some very basic, inex- anced mixer (DBM) and careful gain definition the word refers to the ability of
pensive homemade receivers offer out- distribution in such a receiver can yield a receiver to respond to incoming signals.
standing performance. superb performance if a proper design It is proper to conclude from this that the
'For many years the .evolution of effort is put forth. Of course, the local better the sensitivity, the more responsive
commercial amateur receivers seemed to oscillator should be stable and low in the receiver will be to weak signals. The
stagnate except for the window dressing noise components to further enhance popular misconception is that the greater
and frills added to the front panels. performance. Thus far, not many com- the receiver front-end gain, the higher the
Emphasis was placed on "sensitivity" . mercially built amateur receivers meet the sensitivity. An amateur who subscribes to
(whatever was really meant by that term) foregoing criteria. In terms of dynamic this concept can ruin the performance of a
in the advertising. Some amateurs. con- range, some manufactured receivers ex- good receiver by installing a high-gain
'cluded, as a result of the strong push for hibit an MDS (minimum discernible preamplifier ahead of it. Although this
sensitive receivers, that the mark of a good signal) of -145 dBm (referenced to the will cause the S meter to read much higher
unit was seen when atmospheric noise on noise floor); blocking of the desired signal on all signals, it can actually degrade the
the hf bands could push the S-meter does not occur (1 dB of compression) until receiver sensitivity if the preamplifier is of
needle up to an S2 or S3. Very little the adjacent test signal is some 116-dB inferior design (noisy).
thought, if any, was given· to the above the noise floor, and the two-tone A true measure of receiver sensitivity is
important parameters of a receiver - IMD is on the order of 85 dB. Greater obtained when the input signal is ref-
high dynamic range, fine readout reso- detail concerning this measurement tech- erenced to the noise generated within the
lu,tion and frequency stability. Instead, nique will be given later in the chapter. A receiver. Since the significant noise genera-
countless receivers were placed on the receiver with the approximate figures just ted inside a receiver of good design
market with 5- or lO-kHz dial increments given is considered to be an acceptable originates in the rf and mixer stages
and excessive amounts of front-end gain. one for use where fairly strong signals (sometimes in the post-mixer amplifier), a
The latter caused the mixer (or mixers) to prevail. However, it is possible to improve low-noise front end· is vital to high
collapse in the presence of moderate and those numbers considerably; it has been sensitivity. The necessary receiver gain
strong signals. Double-conversion super- done by amateurs who designed and built can be developed after the mixer - usually
heterodyne receivers were for a long time their own receivers. Examples are W7Z0I's in the i-f amplifier section. The internal
the choice of manufacturers and ama- "Competition Grade Receiver" (March noise is generated by the thermal agitation
teurs. The second i-f was often 100 or 50 and April 1974 QS1) and the WICER of electrons inside the tubes, transistors or
kHz, thereby enabling· the designer to get receiver described in June and July 1976 ICs. It is evident from the foregoing
fairly reasonable orders of selectivity by QST. discussion that a receiver of high sen-
means of high-Q i-f transformers. That The foregoing suggests strongly that sitivity could be one with relatively low
concept predated the availability of amateurs should consider designing and front-end gain. This thought should be
crystal-lattice and mechanical filters. The building their own receivers. Certainly, kept in mind as we enter the discussion of
low-frequency second i-f dictated the use such an endeavor is within the capability dynamic range and noise figure.
of a double-conversion circuit in order to of many experimenters. The satisfaction
minimize image responses. derived from such an effort can't be Noise Figure
Single-conversion receivers offer much measured. The following sections of this The lower the receiver noise figure
cleaner performance in terms of spurious chapter are written for those who wish to (NF), the more sensitive it is. Receiver
responses and dynamic range. They are acquire a better understanding of how a noise figures are established primarily in
8·1 Chapter 8
the rf amplifier and/or mixer stages.
Low-noise active devices (tubes or transis-
tors) should be used in the receiver front
end to help obtain a low noise figure or
signal-to-noise 'ratio (SNR). The un- t
wanted noise, in effect, masks the weaker '"o
::l
signals and makes them difficult to copy. I-
•
meters and above. spect to i-f selectivity which has been
established by means of high-Q filters
(LC, crystal, monolithic or mechanical).
Noise-Figure Measurements But in a broader sense, selectivity can be
Amateurs can use a thermal noise employed to reject unwanted signal
source for determining receiver noise energy in any part of a receiver - the
figure. The resistance 'of the noise- front end, i-f section, audio circuit or
generator output must match that of the local-oscillator chain. Selectivity is a
receiver input, 50 ohms to 50 ohms, for relative term, since the degree of band-
example. Fig. I shows a setup for making width can vary from a few hertz to more fO
these measurements. The first reading is than a megahertz, depending on the -0
taken with the noise generator turned off. design objectives. Therefore, it is not 5
The receiver audio. gain is adjusted for a uncommon to hear terms like "broadband
-10
convenient noise reading in dB, as filter" or "narrow-band filter."
observed on the audio power meter. The The degree of selectivity is determined 5
noise generator is turned on next, and its by the bandwidth of a filter network, The -20
outpui is increased until a convenient bandwidth is normally specified for the
power ratio, expressed by N2/Nl, is
observed. From these numbers the noise
minus. 3-dB points on the filter response
curve; the frequencies where the filter
..
CD
I
-2 5
-30
, ' figure can be calculated by output power is half the peak output -3 5
NF = (excess) dB - 10 log ITN2/Nl) -IJ power elsewhere in the passband. The
difference in frequency between a minus -40
where (excess) dB = the difference in noise 3-d8 point and the filter center fre- -4 5
power between the receiver noise quency is known asAf. The bandwidth of -50
and that from the noise generator. ' the filter then becomes 2 A f. Fig. 2 t~ 1.8 1.85 1.9 - 1.95 2.0 2.05
MHz
If a thermal diode such as a 5722 tube is illustrates this principle and shows how
used as the noise source, and if the circuit (B)
the unloaded Q of a tuned circuit or
is operated in the "temperature-limited resonator relates to the bandwidth charac-
mode" (portion of the tube curve where teristic. Fig. 3 - A tunable Cohn type of filter is shown at
saturation occurs, dependent upon cathode A. L5 and L6 are the bottom-coupling inductors
If a tuned circuit is used as a filter, the (1.45/LH). l1 and L4 are 70 /LH and L2, L3 are
temperature and plate voltage), the (ex- higher its loaded Q, the greater the 140 /LH. A response curve for the tunable filter is
cess) dB can be calculated by selectivity. To make the skirts of the given at B.
(excess)dB = 10 log (20Rd1d) response curve steeper, several high-Q
resonators can be used in cascade. This
where Rd = the noise source output re- aids the selectivity by providing greater
sistance rejection of signals close in frequency to
I d = the diode current in amperes
the desired one. The desirable effect of
Most manufacturers of amateur com- cascaded filter sections can be seen in Fig. curve is included to illustrate the se-
munications receivers rate the noise 3. The circuit is that of a tunable Cohn lectivity obtained.
characteristics with respect to signal type of three-pole filter for use in the front An ideal receiver with selectivity ap-
input. The common expression is S+ Noise end of a 160-meter receiver. The response plied ~o various significant parts of the
~
some 100 dB from l;l no-signal to a 2'
large-signal condition, a misinformed Diode Detectors R1 AF
OUTPUT.
individual might claim that the dynamic The simplest detector for a-m is the
range of the receiver is 100 dB. A receiver diode. A germanium or silicon crystal is AF GAIN
with a true dynamic range of 100 dB an imperfect form of diode (a small (Cl
,would be a very fine piece of equipment, current can usually pass in the reverse
indeed! direction), but the principle of detection in
Dynamic range relates.specifically to the a semiconductor diode is similar to that in Fig. 4 - Simplified and practical diode detector
amplitude levels of multiple signals that a vacuum-tube diode. circuits.- A. the elementary half-wave diodE!
can be accommodated during reception.' Circuits for both half-wave and full- detector; B. a practical circuit, with rf filtering and
audio output coupling; C. full-wave diode
This is expressed as a numeric ratio, wave diodes are given in Fig. 4. The detector, with output coupling indicated. The
generally in decibels. The present state of the simplified half-wave circuit at Fig. 4A circuit. L2C 1, is tuned to the signal frequency;
receiver art provides optimum dynamic includes the rf tuned circuit, L2C I, a typical values for <;2 and R 1 in A and Care 250
ranges of up to 100 dB. This is the coupling coil, LJ, from which the rf pF and 250 kO, respectively; in B, C2 and C3 are
100 pF each; R1, 50kO; and R2, 250 kO. C4 is
maximum dynamic range attainable when energy is fed to L2CI, and the diode, PI, 01 ILF.
the distortion products are at the sen- with its load resistance, RI, and bypass
sitivity limit of the I;,eceiver. Simply stated, capacitor, C2.
dynamic range is the dB difference (or The progress of the signal through the
ratio) between the largest tolerable re- detector or rectifier is shown in Fig. 5. A R I is smoothed out, as shown in C. C2
ceiver input signal (without causing audible typical modulated signal as it exists in the thus acts as a filter for the radio-frequency
distortion products) and the minimum tuned circuit is shown at A. When this component of the output of the rectifier,
. discernible signal (se~sitivity). signal is applied to the rectifier, current leaving a dc component that varies in the
Poor dynamic range can cause a host of will flow only during the part of the rf same way as the modulation on the
receiving problems when strong signals cycle when the anode is positive with 'original signal. When this varying dc
appear within the front-end passband. respect to cathode, so that the output of voltage is applied to a following amplifier
Notable among the maladies is cross the rectifier consists of half"cycles of rf. through a coupling capacitor (C4 in Fig.
modulation of the desired signal. Another These current pulses flow in the load 4), only the variations in voltage are
effect is desensitization of the receiver circuit comprised of R I and C2, the transferred, so that the final oqtput signal
from a strong unwanted signal. 'Spurious resistance of R I and the capacitance of C2 is ac, as shown in D.
signals may appear in the receiver being so proportioned that C2 charges to In the Circuit at 4B, RI and C2 have
tuning range when a strong signal is the peak value of the rectified voltage on been divided for the purpose of providing
'Hayward, "Defining and Measuring Receiver Dy- each pulse and retains enough charge a more effective filter for rf. It is
namic Range," QST. July, 1975. between pulses so that the voltage across important to prevent the appearance of
8-3 Chapter 8
DETECTOR RFC
MODULATED
SIGNAL APPLIED
~AiJDIO
~OUTPUT
TO DETECTOR
0.1
(A) r--~------,
1 :
SIGNAL
AFTER
RECTIFICATION
VARYING DC
AFTER RF IS 0
t&
o -~-
(8)
t ~/"\. ~
-~--
___ &
~--
~ INPlJT
;0
RF
.
I
I
I
I
L _________ --1
I
I
I
I
FILTERED OUT
(e)
(Al
AFTER PASSINGt ! ' r'\ r'\
THROUGH ~ I \ I \
~~~~II~~R ~ V \...JC RFC
~AUDIO
(0) ~OUTPUT
r----------, 0.1
~
I :
Fig. 5 - Illustrations of the detection process.
RF I I
INPUT I I
I . I +12V
any rf voltage in the output of· the I I
. L ________ -.l
detector, because it may cause over-
loading of a succeeding amplifier stage.
The audio-frequency variations can be (81
transferred to another circuit through a
coupling capacitor, C4. R2 is usually a
Fig. 6 - Plate-detection circuits. In each example the input circuit is tuned to the signal frequency.
"potentiometer" so that the audio volume Typical Rl values for the tube circuit at A are 1000 to 5600 ohms. For the FET circuit at B, Rl is on the
can be adjusted to a desired level. order of 100 to 3900 ohms.
Coupling from the potentiometer (volume
control) through a capacitor also avoids
any flow of dc through the moving
contact of control. The flow of dc through DETECTOR
a high-resistance volume control often 01
tends to make the control noisy (scratchy) 2N5459
'''1 ~-------- _ !
circuit only in that both halves of the rf RFJT:b
cycle are utilized. The full-wave circuit has
the advantage that rf filtering is easier
than in the half-wave circuit. As a result,
J ~~-----oOUTPUT
AUDIO
i
The product detectors shown in Fig. 8 PRODUCT DETECTOR
01 BFO
are called "passive." The term means that
the devices used do not require an T2pINJ.
operating voltage. Active devices (transis- • "''''.1
tors, ICs and tubes) do. require an
operating voltage. Passive mixers and
detectors exhibit a conversion loss, whereas
active detectors provide a conversion
gain. Passive detectors' usually require a
substantially greater level of BFO in- ...------0 AF
OUT
jection voltage than is the case with active
detectors. Therefore, the' primary draw- ;h0.1
backs to the use of diodes in these circuits (Cl
are the loss in gain apd the high injection
level required. A typical conversion loss
for a two-diode detector (Fig. 8A) is 5 dB. Fig. 8 - Examples of diode product detectors. Singly balanced types are sho'l"n at A and B. A doubly
The four-diode detectors have a loss of l1alanced version is illustrated at C.
approximately 8 dB. The BFO injection
level for each of the diode detectors shown
in Fig. 8 is + 13 dBm, or 20 mW. Since the though some designers use cores with less been added to improve the circuit balance
terminal impedance of the detector is initial permeability. An Amidon.FT-50-43 while presenting a more sYIUmetricalload
roughly 50 ohms, an rms BFO voltage of is entirely suitable as a transformer core to the BFO. The result is better isolation
1, or a pk-pk voltage of 2.8 is required. for the circuits shown. Fifteen trifilar between the BFO and i-f input ports.
The advantages to the use of diodes in a turns are ample for each transformer. Two broadband transformers are used
product detector are circuit simplicity, High-speed silicon switching diodes are to provide the doubly balanced' detector
low cost, broadband characteristics, low satisfactory for use in the circuits of Fig. of Fig. 8C. The advantage with this
noise figure and good port-to-port signal 8. They should be as Closely matched as configuration is that alt three ports are
isolation. This type of detector is excellent possible for forward and back resistance. isolated from one another effectively.
at the input of a direct-conversion receiver Closely matched diodes can be had by
(to be treated later in the chapter). using a diode-array IC, such as the RCA Simple Active Product Detectors
The transformers shown in the circuits CA3019 or 3039. Hot-carrier diodes are Fig. 9 contains two examples of
of Fig. 8 are broadband, toroidal-wound excellent for the circuits shown. Matched single-ended active detectors which em-
types. The black dots near the windings of IN914s are the choice of many amateur ploy FETs. They are quite acceptable for
Tl and T2 indicate the phasing required. . designers. use in simple receivers which do not
The core material is ferrite and the A singly balanced detector is seen at A require high performance characteristics.
windings are trifilar. Core permeability in Fig. 8. An improved singly balanced The circuit at A uses a JFET which has
can be 950 for most applications, al- detector is shown at B. Two diodes have BFO injection voltage supplied across the
8-5 Chapter 8
converted to. a-m by means of trans-
PRODUCT DETECTOR
former Tl. The voltage induced in the Tl
secondary is 90 degrees out of phase with
01 the current in the primary. The primary
MPFW2
signal is introduced through a center tap
SI G. " - .. -.Jt--e-~4rl-r-T-.-J on the secondary, coupled through a
IN v-----'l capacitor. The secondary voltages com-
bine on each side of the center tap so that
the voltage on one side leads the primary
signal while the other srde lags by the
same amount. When rectified, these two
voltages are equal and of opposite
polarity, resulting in zero-voltage output.
BFO A shift in input frequency causes a shift in
INJ. the phase of the voltage components that
(Al
result in an increase of output amplitude-
on one side of the secondary, and a
corresponding decrease on the other side.
The differences in the two changing
voltages, after rectification, constitute the
PRODUCT DETECTOR audio output.
01
RCA developed a circuit that has now
40673 RFC become standard in entertainment radios
and which eliminated the need for a
preceding limiter stage. Known as the
ratio detector, this circuit is based on the
idea of dividing a dc voltage into a ratio
which is equal to the ratio of the
1!500
amplitudes from either side of a dis-
criminator transformer secondary. With a
detector that responds only to ratios, the
input signal may vary in strength over a
wide range without causing a change in
the level of output voltage - fm can b<:
detected, but no a-m. In an actual ratio
detector, Fig. 12B, the dc voltage required
is developed across two Imid resistors,
Fig. 9 - Active product detectors. A JFET example is provided at A and a dual-gate MOSFET shunted by an electrolytic capacitor.
type is at B. Other differences include the two diodes,
which are wired in series-aiding rather
than series".opposing, as in the standard
source resistor. Because the source is not conversion gain it provides, it performs
discriminator circuit. The recovered audio
bypassed, instability can occur if the similarly to the singly balanced diode
is taken from a tertiary winding which is
circuit is used as a mixer which has ah i-f detector of Fig. 8B. Doubly balanced
tightly coupled to' the primary of the
that is close to the signal frequency. This active detectors are seen at Band C of Fig.
transformer. Diode-load resistor values
problem is not apt to become manifest 10. These ICs contain two sets of
are selected to be lower (5000 ohms or
when the output is at audio frequency. differential amplifiers each. The "diff
less) than for the discriminator.
Slightly more injection power is needed amps" are cross-connected in the exam-
The sensitivity of the ratio detector is
for circuit A than is necessary for the ples shown to obtain doubly balanced
one half that of the discriminator. In
detector at B. An rms voltage of roughly circuits. The virtues of these detectors are
general, however, the transformer design
0.8 is typical (6.5 mW). similar to the equivalent four-diode types,
values for Q, primary-secondary coupling,
The detector of Fig. 9B operates in a but they exhibit several dB of conversion
and load will vary greatly, so the actual
similar fashion to that of A, but the BFO gain. The MCI496G is made by Motorola
performance differences between these
is injected on control gate no. 2. and the CA3102E is an RCA device.
two types of fm detectors are usually not
Approximately I volt rms is needed (O~ I
FM Detectors significant. Either circuit cah provide
mW). FETs with proper injection levets
excellent results.
and moderate signal-input amounts have The first type of fm detector to gain
A crystal discriminator is shown in Fig.
excellent IMD characteristics. Generally, popularity was the frequency discrimi-
12C. This provides an adjustment-free
they are preferred to single-ended, bipolar- nator. The characteristic of such a
discriminator by virtue of the quartz
transistor detectors. The circuits at A and detector is shown in Fig. 11. When the fm
resonator. The components without as-
B contain rf chokes and bypass capacitors signal has nq modulation,. and the carrier
signed values are selected to give the
in the drain leads to minimize the transfer is at point zero, the detector has no
desired bandwidth. A unit of this type is
of BFO energy to the succeeding audio output. When audio input to the fm
sold by Spectrum International Co.
stage. The bypass capacitors are useful transmitter swings the signal higher in
also for rolling off the unwanted high- frequency, the rectified output increases in
frequency audio components. the negative direction. Over a range TRF Receivers
where the discrimination is linear (shown Tuned-radio-frequency receivers have
Active Balanced Product Detectors as the straight portion of the line), the little value in Amateur Radio today, but
Examples of active IC product de- conversion of fIp to a-m which is taking in the early days they were suitable for the
tectors are given in Fig. 10. A singly place will be linear. reception of spark and a-m signals. They
balanced version is shown at A. It uses an A practical discriminator circuit is consisted mainly of a couple of stages of
RCA differential-pair IC. Except for the shown in Fig. 12A. The fm signal is selective rf amplification, an a-m type of
~
, I I
, I I
I I I
l ! I
I I ~
I I
I I
I
+
+
FREQ.
o
1234567
overlap. These tests were performed in the
ARRL laboratory with the 10 signals
separated from one another by equal
14i312'1i 10 98
BOTTOM
amounts.
(el A major problem associated with the
use of regenerative and superregenerative
Fig. 10- Exam pies 01 IC product detectors. At A 'IS a singly balanced version. while those at Band C
receivers was oscillator (detector) re-
are doubly balanced. radiation. The isolation between the
detector and the antenna was extremely
poor, even when an rf amplifier was
employed ahead of the detector. In many
detector and an audio amplifier. Varia- could be copied, if they were loud, when instances the radiated energy could be
tions were developed as regenerative and the regeneration control was set for a heard for several miles, causing intense
superregenerative receivers. The straight non-oscillating condition. For weak-signal interference to other amateurs in the
regenerative detector was simply a self-os- a-m reception the regeneration control community.
cillating detector which provided in- was advanced to increase the detector
creased sensitivity (similar in function to a sensitivity and the signal was tuned in at Direct-Conversion Receivers
product detector) and a beat note for cw zero beat, thereby eliminating the het- A more satisfact<;>ry type of simple
reception. Amplitude-modulated signals erodyne from the carrier. Present-day receiver is called the direct-conversion or
8.7 Chapter 8
synchrodyne type. Although there is a
1N67A 6Bk distinct possibility of signal radiation, it is
FROM considerably lower in level than with
.I-F 0---.:..--.....--1
AMP. regenerative receivers. This results from
better isolation between the antenna and
0.001 the source of the oscillation. A modern
direct-conversion receiver uses a separate
100 oscillator and a balanced or doubly
balanced detector. Both features help to
IN67A
reduce unwanted radiation.
TI-MILLER 12-C45 The detection stage of this receiver is
1000 RFC
actually a product detector that operates
+B FM DISCRIMINATOR (Al at the desired signal frequency. The
product-detector circuits described eadier
FROMQ----, in the chapter are suitable in this kind of
I-F receiver. A tuned rf amplifier is useful
AMP.
ahead of the detector at 14 MHz and
+~ higher, but it is seldom necessary from 160
25V through 40 meters. This is because the
atmospheric and man-made noise from
1500 the resonant antenna usually exceeds that
of the detector below 14 MHz. When no rf
stage is used, it is desirable to include a
TI-MILLER 1606
tuned network ahead of the detector.
Fig. 12 shows typical front ends for
AF OUl
(Sl direct-conversion receivers. One circuit
(D) employs a passive detector. The other
(E) contains an active detector. The latter
r------------------j is desirable in the interest of increased
ZI I FM CRYSTAL DISCRIMINATOR I AF
gain.
: (Cl I OUT The circuit of Fig. 12D shows a singly
balanced passive detector. Frqnt-end
X'
OO
selectivity is provided by a tuned cireuit.
FROM I ':
TI is a broadband, trifilar-wound toroid
LIMITER I
CI I ~ transformer. It is tapped on the input
I I tuned circuit at the approximate 50-ohm
I C4, I point. An rf filter is used after Oland 02
I " ZI-SPECTRUM I to prevent LO energy from being passed
L __________________
rf-7 INT'L ASS'Y ~
,,_,,"n'·' ""'
~
O--=--+-T--' T RAN.
AF
TO
HIGH-GAIN
is used at the detector output. The
primary winding should have low dc
resistance to provide dc balance between
the collectors of the differential-amplifier
t;:
ANT. '--......- - . . . , pair in the Ie. Alternatively, a ccntcr-
. AUDIO AMP.
2
2000 tapped primary can be used. I f this is
done, pin S should be connected to one
TO.OI..-JVV\J--II--_ _- - - - . end of the winding and the B+ fed to the
rh 1200 center tap. The impedance between pins 6
+h
AC11VE DETECTOR
(E)
1
3.5 MHz
0 001
.
50;Frh
560
and 8 is approximately 8000 ohms.
In order to obtain ample headphone
volume during reception of weak signals it
is neceSl'ary to use an audio amplifier
LO +12V
which has between 80 and 100 dB of gain.
The first af amplifier should be a
Fig. 12 - Popular types 01 1m detector (A, Sand C) and typical detectors which can be used in the
Iront ends 01 direct·conversion receivers (0 and E). A passive detector is seen at O. The version at E low-noise type, such as a JFET. The
provides considerable conversion gain. audio-gain control should follow the first
Superheterodyne Receivers
Nearly all of the present-day communi- shunt audio noise limiters. An assortment purpose of making the second i-f lower
cations receivers are structured as super- of techniques is being used to improve the than the first. This helps to increase the
heterodyne types. Fig. 15 shows a simple overall selectivity of these receivers - overall selectivity in some designs, but it
block diagram of a single-conversion elaborate i-f filtering, RC active or LC often degrades the receiver dynamic range
superheterodyne circuit. This basic design passive audio filters. However, the basic through the addition of a'second mixer.
has been popular since the 1930s, and only circuit concept remains unchanged. The Multiconversion receivers are more prone
a few general circuit enhancements have advancement of greatest significance in re- to spurious responses than is the case with
been introduced in recent years. Sophisti- cent years is the changeover from vacuum single-conversion designs, owing to the
cated versions of this type of receiver use tubes to semiconductors. This has in- additional oscillator and mixing fre-
various alternatives to the circuits in- creased the life span of the of the equip- quencies involved. The "cleanest" per-
dicated in the block diagram. The local ment, improved overall efficiency, aided formance is obtained [rem properly
osci11ator, for example, might utilize a stability (reduced heating), and con- designed single-conversion receivers.
phase-locked loop or synthesizer type of tributed to greater ruggedness and'
LO chain rather than a straight VFO. miniaturization. ' Circuit Function
Digital readout is used in some models in Some manufacturers still produce dou- In the example of Fig. 15 it is assumed
place of the more traditional analog ble- or multiconversion superheterodyne that the receiver is adjusted to receive the
readout method. Rf types of noise receivers, but the circuits are similar to 20-meter band. Front-end selectivity is
blankers (often very complex) are chosen that of Fig. 15. Multiconversion receivers provided by the resonant networks before
by some designers in preference to simple have a second mixer and LO chain for the and after the rf amplifier stage. This part
ANT. {>--.?--t
AGC AGC
LINE LINE
to
AGC
AGC
RECT,
f
9,0'014
MHz ~
T
,---, 8,999
MHz
SINGLE-CONVERSION RECEIVER
Fig., 15 - Block diagram of a single-conversion superheterodyne receiver for 20 meters. The arrows indicate the direction of signal and voltage
components.
8·11 Chapter 8
3.5 MHz C1b OSC.
'-~__~____O+9V
Or-j.01 ~
REG.
ll-.L---+.-H r r - +.......-if-o g~~·PUT
S.M. 50
RFC
lmH
+9V
REG.
o.u
COLPITTS
(Al PIERCE
(Bl
TO 14-MHz
RCVR WHICH
....___.-~~~~~~~~~+9V
REG. HAS 3.3-MHz
I-F
. MAIN
TUNING
vxo VFO
lOOK
1-'¥+-....---0+9V
REG.
7 MHz
OUT
L-+-----J~~r_!f--o
5.0-5.5 MHz
>4------------....------------~--~~OUTPUT
22
+9V (Al S.M.
(81 REG.
8-13 Chapter 8
more of frequency change at 144 MHz. oscillator has a crystal for each amateur 25 percent of the total number of turns for
band accommodated by the receiver. The proper feedback.
VFO Circuits crystals and appropriate bandpass filters A comparison is drawn in Fig. 22
Variable-frequency oscillators are simi- are switched by means of a panel- between a bipolar transistor and JFET
lar in performance to the VXOs which mounted control. The band-pass filter version of a series-tuned Colpitts oscil-
were described in the foregoing text. The (Fig. 20) is desirable in the interest of lator (some times called a "series-tuned
essential difference is that greater fre- preventing 5- and 12.3-MHz energy from Clapp"). This type of oscillator can be
quency coverage is possible, and no reaching the receiver mixer. A doubly made very stable by using polystyrene
crystals are used. The practical upper balanced mixer is recommended if mini- capacitors in' the frequency-determining
frequency limits for good stability range mum unwanted energy is desired at the part of the circuit: Silver-mica capacitors
between 7 and 10 MHz. For operation at mixer output. are satisfactory if they have been graded
higher frequencies it is better to employ a Some typical VFOs are shown in Figs. out for temperature stability, but most
heterodyne type of VFO. This calls for a 21 and 22. A vacuum-tube ECO (electron- run-of-the-mill, silver-mica units have
VFO operating at , say,S MHz. The VFO coupled oscillator) is compared to a unpredictable characteristics in this re-
output is heterodyned in a mixer with similar one which utilizes a dual-gate gard. Some are very stable, while others
energy from a crystal-controlled oscillator MOSFET in Fig. 21. The capacitor shown from the same lot may exhibit positive or
to provide a resultant sum or difference in dashed lines (Cl) can be used in that negative drift characteristics. Polystyrene
frequency at the desired LO-chain output part of the circuit rather than at the low capacitors are considerably better in this
frequency. A block diagram is given at end of the tank coil if greater bandspread respect. These VFO circuits can be scaled
Fig. 20 to illustrate the concept. Most . is desired. Cl is the main-tuning capacitor to other operating frequencies by using
modern receivers employ this style of and C2 is the padder for calibrating the the values shown to determine the
local-oscillator circuit. The heterodyne oscillator. The coil tap is approximately reactances of the capacitors. This in-
formation will enable the designer to
select approximate values in pF for other
frequencies.
VFO Fig. 23 illustrates a composite VFO
which has a buffer stage that is followed
VFO by an amplifier. D 1 can be included to
5.0-5.5
MHz MPF102 provide the necessary frequency offset
when switching the receiver from upper to
( r __~~~~.-4i~~--~---o+9V
REG. lower ssb. This is necessary in order to
eliminate the need to readjust the receiver
5.0-5.5 calibration dial when changing sidebands.
MHz
Cl is adjusted for the desired offset
amount. The pi-network output from the
amplifier stage is designed to transform
10 500 ohms to 50 ohms. The low-impedance
P output is desirable in the interest of
0.01
minimum frequency pulling from load
P.POLYSTYRENE changes. A 3300-ohm swamping resistor is
used across the pi-network inductor to
(A) (8)
broadband the tuned circuit and to
prevent any tendency toward instability
when a high-impedance load is attached
Fig. 22- Colpitts VFOs. A bipolar transistor type is seen at A, while a FET version i~ given at B. to the circuit. Long-term drift measure-
+9V REG.
VFO
BUFF AMP.
~--1~I----:--i~50-0HM
~OUTPUT
0.01
~-n...--n+12V
Fig. 23 - SUggested circuit for a stable series-tuned Colpitts VFO. Buffering follows the oscillator to increase the output level and provide load isolation.
RF AmpU/Iers
ments with this type of circuit .at the front-end network may have a 3·dB band- It was implied earlier in the chapter that
frequency specified indicated a maximum width of 100 kHz for a given Q and load rf amplifiers are useful primarily. to
shift of 60 Hz over a three-hour period. factor. At 40 meters the same type of net- improve the receiver noise figure. When
Output was measured at, approximately 1 work would be 200-kHz wide at the 3-dB atmospheric and man-made noise levels
volt rms across 50 ohms. points of the response curve. It is for this exceed that of the' mixer it should be
reason that most receivers have a tunable possible to realize better dynamic range
Receiver Front Ends front-end section (preselector). If fixed- by not having an rf amplifier. The gain of
The designer has a number of options tuned filters were used, at least two such the rf stage, when one is used, should be
available when planning the input section filters would be necessary to cover from set for whatever level is needed to override
of a receiver. The band-pass characteristics 3.5 to 4 MHz or 1.8 to 2.0 MHz, this the mixer noise. Sometimes that is only a
of the input tuned circuits are of consider- would complicate the design and cost of few dB. A good low-noise active device
able significance if' strong out-of-band the equipment. should be employed as the rf amplifier in
signals are to be rejected - an ideal de- Fig. 24 shows the two concepts just such instances. For hf-band work 40673s,
sign criterion. Most of the commercial re- discussed. The circuit at A covers an of MPF102s and 2N4416s are good. Most
ceivers available to the amateur use tuned the 80-meter band, and if selective enough modern receiving tubes specified for use
circuits which can be adjusted from the offers some in-band rejection. A pair of as rf amplifiers are suitable also. For vhf
. front panel of the equipment. The greater Butterworth band-pass filters might be work the Siliconix E300 and U310 FETs
the network Q, the sharper the frequency used at FLl and FL2 of Fig. 24 to cover are excenent low-noise devices.
response, and hence, the better the adja- an of the 80-meter band. A lot of A wen-designed receiver should not
cent-frequency rejection. For a given net- additional components would be required, have agc applied to the rf amplifier. The
work design the bandwidth doubles for and the in-band rejection of unwanted best noise figure and rf-stage dynamic
each octave higher. That is, an 80-meter signals would be less than in the range win result when agc is not applied.
8-15 Chapter 8
This is because 'the. agc voltage changes
the' operating characteristics of the rf
amplifier from Class A to a less linear INPUT
mode. TUNED CIRCUIT
~~",--oTO RF
Fig. 26 shows some typical rf amplifiers AMP.
for use in amateur receivers. Tube-type
circuits have not been included because ANT.o----;~
they are not much in vogue, nor do they
offer any particular advantages over
solid-state amplifiers.
The circuit at A in Fig. 16 is likely to be
the least subject to self-oscillation of the (A)
four examples given. The common-base
hookup helps to ensure stability if the gate
lead is kept as short as is physically
possible. The gain (rom a common-gate .--_+_____OTO RF
amplifier of this type is lower than that of AMP.
a common-source amplifier. However,
gains up to 15 dB are entirely typical. The
drain of the FET need not be tapped
down on the drain coil, but if it is there
will be less loading on the tuned circuit,
thereby permitting somewhat greater
PAD-I PAD-2
tuned-circuit selectivity: The lower the
drain tap, the less the stage gain.
All of the FET amplifiers in Fig. 26 are (Bl
capable of providing low-noise operation
and good dynamic range. The common-
source circuits at illustrations Band C can
provide up to 25 dB of gain. However, Fig, 25 - Front-end attenuators. A simple typ.e is given at A and a step-attenuator version is seen at B.
+12V +12V
+12V
O.~
TO
FILTER
o--i0.01 1.8-30 MHz
(50n)
+12V
Fig. 26- Narrowband rf amplifiers are shown from A to C. A fed-back broadband,rf amplifier is seen at D.
8·17 Chapter 8
dynamic range.
An LC, crystal-lattice, or mechanical
type of band-pass filter is almost always
SINGLY BAL. MIX.
used after the mixer or the post-mixer
amplifier. This helps to establish the
overall selectivity of the receiver. It also
HP-2800 rejects unwanted mixer products that fall
~~]"
, OR 11-12
1N914 12-11
outside the passband of the filter.
'---+---1 L ,",1- F In the interest of optimum mixer'
r-'-'OUT
0.01 performance, the L'O energy supplied to it
should be reasonably clean with respect to
frequencies other than the desired LO
one. Many designers, for this reason, use a
+7dBm LO 12 0 11----.... band-pass filter between the LO output
0.01
and the mixer input. Excessive LO noise
will seriously degrade receiver performance.
LO noise should be 80 dB or more below
the peak level of the desired LO frequen-
(Al cy. Excessive noise will appear as noise
side-bands in the receiver output.
ir '"'
requires somewhat more LO power than
T2 l- F
,would be used if injection was done at the
gate. However, there is less occasion for
LO pulling when source injection is used,
and there is better isolation between the
LO and antenna than would be the case
with gate i!1jection.
The circuit at B in Fig. 27 is similar to
that of A with regard to general perform-
D1,-D4, INCL. ~LOf2
HP-2800 OR 1N914 ~+7dBm
ance. The major difference is that a dual-
0.01 gate MOSFET is used to permit injection
of the LO energy at gate 2. Since there is
considerable signal isolation between gates
1 and 2, LO pulling is minimized and.
(Bl antenna-LO isolation is good.
A singly balartced active mixer is
illustrated at Fig. 28. Two 40673 dual-gate
MOSFETS are connected in push-pull,
Fig. 29 - Singly and doubly balanced diode mixers.
but with the, LO frequency injected in
parallel at gate 2 of each device. A
potentiometer is used in the sources of the
DIPLEXER I-F AMP. transistors to permit circuit balance. This
mixer offers superior performance to
those shown in Fig. 27.
I-F OUT
SIG. 0----1 One of the 'least complicated or
IN
expensive mixers is the two-diode version
(singly balanced) seen in Fig. 29A. A
~T trifilar-wound broadband toroidal trans-
LO
rh ~7MH'l 1. XL,XC.50n
former is used at the mixer input. The
shortcoming of this mixer over the one
l
INPUT seen at B is that signal isolation between
all three mixer ports is not possible. A
51 better version is that at B in Fig. 29. In this
case all three mixer ports are well isolated
from one another. This greatly re-duces the
probability of spurious responses in the
receiver. Conversion loss with these
Fig. 30- Method far diplexing the mixer output to improve the IMD characteristics. mixers is approximately 8 dB. The"
impedance of the mixer ports is ap-
than loss. This means that the stages whereas an FET active mixer may exhibit proximately 50 ,ohms.
following .the mixer need not have as a conversion gain as great as 15 dB. The Improved IMD characteristics can' be
much gain as when diode mixers are used. cost of gain stages in this era is rdatively, had from a diode-ring mixer, by placing
A typical doubly balanced diode mixer small. This easily justifies the use of strong a diplexer after the mixer as seen in
will have a conversion loss of some 8 dB, passive mixers in the interest of high Fig. 30. The diplexer consists of a
7 "'-,,,,,,,--
HIGH-LEVEL",. .~1
•
INPUT ~ I - -........--,;~
MIXER
(8)
·Fig. 31 - CA3028A singly balanced mixer. The circuit for the IC is given at B.
OUTPUT
+12V +12V
;---------~~----o6
.----4-~-~e
1300
80-_--1
.1~ .1 HIGH-
100
f-;+., LEVEL
SIGNAL IN
~o
.01
HIGH- LEVEL I"\-....II-_~-t~!:./B
SIGNAL IN ....-----, 40-------1'
LOW-LEVEL
MC1496G SIGNAL IN
LOW-LEVEL ~1 9 10-------~----------_+--~
SIGNAL IN ....-----,I----+--......~
~~~::---"J
~o
BIAS
eo-_........--r
+12V
(AI
(8) 10
Fig. 32- MC1496G doubly balanced mixer and circuit of the IC.
high-pass network (Ll) and a low-pass mixer as that of Fig. 29. However, it has CA3028A and MCI590G ICs are the
one (L2). L2 is tu·ned to the i-f and good conversion gain and a fairly low most popular ones for amateur work
serves as a matching network between 50 noise figure. because they are easy to obtain' and are
ohms and R L, the FET gate resistor. L I relatively low in cost. With careful layout
and th.e associated series capacitors are I-F Amplifiers techniques either device will operate in a
tuned to three times the i-f and terminated The amount of i-f amplification used in stable manner. Bypassing should be done
in 50 ohms. This gives the mixer a proper a receiver will depend upon how much as near to the IC pins as possible. Input
resistive termination without degrading signal level is available at the input to the and output circuit elements must be
the 9-M Hz i-f. The high-pass network has i-f strip. Sufficient gain is needed to ensure separated to prevent mutual coupling
a loaded Q of I. ample audio output consistent with which c~n cause unstable operation. If IC
driving headphones or a speaker. Another sockets are used they should be the
IC Mixers consideration is the amount of age- low-profile variety with short socket
Although there' are numerous ICs initiated i-f gain range. The more i-f stages conductors ..
• available for use as mixers, only two are used (a maximum of two is typical) the Fig. 33 contains examples of bipolar
shown here. Fig. 31 shows a CA3028A greater the gain change caused by agc . transistor and FET i-f amplifiers. Typical
singly balanced active mixer. The diagram action. The range is on the order of 80 dB component values are given. A CA3028A
at Bshows the inner workings of the Ie. when two CA3028A ICs are used in the i-f IC, connected for differential-amplifier
The LO is injected at pin 2 of the Ie. strip. A pair of MC 1590G ICs will operation, is shown in Fig. 34 as an i-f
Conversion gain is on the order of 15 dB. provide up to 120 dB of gain variation amplifier. Up to 40 dB of gain is possible
F.,ig. 32A illustrates a doubly balanced with agc applied. with this circuit. The IC is useful up to 120
IC active mixer which employs an Nearly all modern receiver circuits MHz and has a low noise figure.
MC1496G. A schematic diagram of the utilize ICs as i-f amplifiers. Numerous A Motorola MCI590G IC will provide
IC is shown at B. The performance of this types of ICs are available to provide linear, up to 50 dB of stage gain when used as an
mixer is excellent, but it is not as strong a rf and i-f amplification at low cost. The i-t' amplifier. An example of the circuit is
8-19 Chapter 8
I-F AMP I-F AMP
AGC +12V
(MAX. GAIN 3-V
MIN. GAIN I-V)
Fig. 35 - Age is applied to an MC1590G IC.
(Al
I-F AMP
v-------~~+AGC
22k
(MAX. GAIN +4V
;:h.01 MIN. GAIN -2V)
Table 1
Min. 3 dB Min. 4 dB Max. 60 dB Max. 60 dB Max. RV . Max. RV Max.IL Max.IL Min. 60 dB S&L Res. Cap.
Part & Type BW@25°C BWOTR BW@25°C BWOTR @2SoC OTR @25° COTR SBR -5% +5%
Numbers (kHz) (kHz) (kHz) (kHz) (dB) (dB) (dB) (dB) (kHz) ohms· (pF)
526-9689-010 0.375 0.375 3.5 4.0 3.0- 4.0 10.0 12.0 445-F60L .2000 350
F455FD-04 F60H-465 350
526-9690-010 1.2 1.2 8.7 9.5 3.0 ~.O 10.0 12.0 445-F60L 2000 350
• F455FD-12 FSOH-465 350
• F526-9691-01 0 1.9 1.9 5.4 5.9 3.0 4.0 10.0 12.0 445-FSOL 2000 330
F455FD-19 F60H-465 330
526-9692-010 2.5 2.5 6.5 7.0 3.0 4.0 10.0 12:0 445-F60L 2000 510
F455FD-25 F60hi-465 510
526-9693-010 2.9 .2.9 7.0 8.0 3.0 4.0 10.0 12.0 . 445-F60L 2000 510
F455FD-29 F60H-465 510
526-9694-010 3.8 3.8 9,0 10.0 3.p 4.0 10.0 12.0 445-F60L 2000 1000
F455FD-38 F60H-465 1000
526-9695-010 5.8 5.8 14.0 15.0 3.0 4.0 10.0 12.0- 445-F60L 2000 • 1100
F455FD-58 F60H-465 1100
OTR = Operating Temperature Range, RV = RippleVoltage, IL ::: Insertion Loss. SBR = Stop Band Range. S & L = Source and Load
Courtesy of Collins Radio Co.
8-21 Chapter 8
receivers which use a second 'i-f of 455
kHz contain mechanical filters. Table I RS CURRENT FORCE FORCE CURRENT
lists various mechanical filters which are
INPUT DISK OUTPUT
manufactured by Collins Radio Co. TRANSDUCER RESONATORS TRANSDUCER
At intermediate frequenci~s above 500
kHz it is com.mon practice to use crystal
filters. These can be designed with just one
crystal (Fig. 36A), or with two or more
crystals. Fig. 36B illustrates a two-crystal,
half-lattice filter and a cascaded half- Fig. 37 - Block diagram of a mechanical filter (Collins Radio).
lattice filter is shown at C of Fig. 36.
The single-crystal example shown at A
of Fig. 36 is best stated for simple receivers
intended mainly for cw use. C I is adjusted
to provide the bandpass characteristic An illustration of how a mechanical amount of shunt capacitance will depend
shown adjacent to the circuit. When the filter operates is provided in Fig. 37. on the filter model used. The manu-
BFO frequency is placed on the part of Perhaps the most significant feature of a facturer's data sheet specifies the proper
the low-frequency slope (left) which gives mechanical filter is the high Q of the capacitor values.
the desired beat note respective to resonant metallic disks it contains. A Q . Most bipolar transistor i-f amplifiers
f ° (approximately 700 Hz), single-signal figure of 10,000 is the nominal value have an input impedance of 1000 ohms or
reception will result. To th~ right of obtained with this kind of resonator. If L less. There are situations where the output
f ° in Fig. 36A the response drops sharply and C constants were employed to acqui're impedance of the stage preceding the filter
to reduce output on the unwanted side of a bandwidth equivalent to that possible is similarly low. In circuits of this variety
zero beat, thereby making single-signal with a mechanical filter, the i-f would it is best to use series-resonating capacitors
reception possible. If no i-f filter was used, have to be below 50 kHz. in preference to parallel ones. Stray circuit
or if the BFO frequency fell at f 0, nelj,rly Mechanical ,filters have excellent fre- capacitance, including the input and output
equal response would exist either side of quency-stability characteristics. This makes capacitances of the stages before and after
zero beat (double-signal response) as is it possible to fabricate them for fractional the filter, should be subtracted from the
the case with direct-conversion receivers. bandwidths of a few hundred hertz. value specified by the manufacturer.
QRM on the unwanted-response side of Bandwidths down to 0.1 percent can be Collins mechanical filters are available
the i-f passband 'WOUld interfere, with obtained with these filters. This means with center frequencies from 64 to 500
reception. The single-crystal filter shown that a filter having ,a center frequency of kHz and in a variety of bandwidths.
is capable of at least 30 dB of rejection on 455 kHz could have a bandwidth as small Insertion loss ranges from 2 dB to as
the high-frequency side of zero beat. The as 45.5 Hz. By inserting a wire through the much as 12 dB, depending on the style of
filter termination R T, has a marked effect centers of several resonator disks, thereby filter used. Of greatest interest to ama-
on the response curve. It is necessary coupling them, the fractional bandwidth teurs are the 455-kHz mechanical filters
to experiment with the resistance value can be made as great as 10 percent of the specified as F455. They are available in
until the desired response is obtained. center frequency. The upper limit is bandwidths of 375 Hz, 1.2 kHz, 1.9 kHz,
Values can range from 1500 to 10,000 governed primarily by occurrence of 2.5 kHz, 2.9 kHz, 3.8 kHz and 5.8 kHz;
ohms. unwanted spurious filter responses adja- Maximum insertion loss is 10 dB, and the
A half-lattice filter is shown at B in Fig. cent to the desired passband. characteristic impedance is 2000 ohms.'
36. The response curve is symmetrical and Mechanical filters can be built for Different values of resonating capacitance
there is a slight dip at center frequency. center frequencies from 60 to 600 kHz. are required, for the various models,
The dip is minimized by proper selection The main limiting factor is disk size. At spreading from 350 to 1100 pF. Although,
of RT. YI and Y2 are separated in the low end of the range the disks become some mechanical filters are terminated
frequency by the amount needed to obtain prohibitively large, and at the high limit of internally, this series requires external
cw or ssb selectivity. The bandwidth at the the range the disks become too small to be source and load' termination of 2000
3-dB points will be approximately 1.5 practical. ohms. The F455 filters are the least
times the crystal-frequency spacing. For The principle of operation is seen in expensive of the Collins line.
upper or lower sideband reception Y I and Fig. 37. As the incoming i-f signal passes Most modern receivers have selectable
Y2 would be 1.5 kHz apart, yielding a through the input transducer it is con- i-f. filters to provide suitable bandwidths
3-dB bandwidth of approximately 2.25 verted to mechanical energy. This energy for ssb and cwo Most of the commercial
kHz. For cw work a crystal spacing 0[0.4 is passed through the disk resonators to receivers use a 500- or 600-Hz bandwidth
kHz would result in a bandwidth of filter out the undesired frequencies, then filter for cw and a 2.1- or 2.4-kHz
roughly 600 Hz.' The skirts of the curve through the output transducer where the bandwidth for ssb. The input and output
are fairly wide with a single half-lattice mechanical energy is converted back to ends of a filter should be well isolated
filter, which uses crystals in the hf region. the original electrical form. from one another if the filter characteristics
The skirts ca\'l be steepened by placing The transducers serve a second func- are to be realized. Leakage across a
two half-lattice filters in cascade, as tion: They reflect the source and load filter will negate the otherw.ise good
shown in Fig. 36C. R I and R T must be impedances into the mechanical portion performance of the unit. The problem
selected to provide minimum ripple at of the circuit, thereby providing a becomes worse as the filter frequency is
the center of the passband. The same rule termination for the filter. increased. Mechanical switches are not
for frequency spacing between the crystals Mechanical filters require external re- recommended above 455 kHz for filter'
applies. CI is adjusted for a symmetrical sonating capacitors which are used across selection because of leakage across the
response. the transducers. If the filters are not switch wafers and sections. Diode switch-
The' circuits of Fig. 36 can be built resonated there will be an increase in ing.is preferred by most designers. The
easily and inexpensively by amateurs. The insertion loss, plus a degradation of the switching diodes for the filter that is out of
transformers shown are tuned to center passband characteristics. Concerning the the circuit are usually back-biased to
frequency. They are wound bifilar or latter, there will be various unwanted dips ensure minimum leakthrough.
trifilar on ferrite of powdered-iron cores in tne nose response (ripple), which can In the interest of reducing wideband
of appropriate frequency characteristics. lead to undesirable effects. The exact noise from the i-f amplifier strip it is
\
make the agc take hold as quickly as
possible to avoid the ailments mentioned
in the foregoing text
TIME -CONSTANT For best receiver performance the i-f
NETWORK
filters should be contained witHin the agc
loop, which strongly suggests the use of
Fig. 39 - A system lor developing receiver agc voltage. rf-derived agc. Most commercial receivers
follow this rule. However, good results
can be obtained with audio-derived agc,
despite the tendency toward a clicky
response. If RC active audio filters are
used to obtain receiver selectivity, they
worthwhile to use a second filter which March and April 1974 QST. The second . should be contained within the audio-agc
has exactly the same center frequency as filter, FL2, has somewhat wider skirts loop if possible.
the first. The second filter is placed at the than the first, FLI. An RC active audio Fig. 39 illustrates the general concept of
end of the i-f strip, ahead of the product filter after the product detector has a an agc circuit. Rf energy is sampled from
detector. This is shown in Fig. 38. The similar effect, but the results are not quite the output of the last i-f by means of light
technique was described by W7ZOI in as spectacular as when two i-f filters are coupling. This minimizes loading on the
Table 2
CWor
Application SSB Tran. SSB Rec. Digital Data A-M A-M CW' FM
Filter type XF-9A XF-9B XF-9NB XF-9C XF-90 XF-9M XF-9E
No. 01 crystals 5 8 8 8 8 .4 8
6-dB bandwidth 2.5 kHz 2.4 kHz 0:5 kHz 3.75 kHz 5.0 kHz 0.5 kHz 12 kHz
Passband ripple < <1 dB <2dB <0.5 dB <2dB <2dB <1 dB <2dB
Insertion loss <3 dB <3.5 dB <6.5 dB <3.5 dB <3.5 dB <5 dB <3dB
Term. impedance 500{} 500{} 500{} 500{} 500{} 500n 1200B
Ripple capacitors 30 pF 30 pF 30 pF 30 pF 30 pF 30 pF 30 pF
Shape lactor 6:50 dB 6:60 dB 6:60 dB 6:60 dB 6:60 dB 6:60 dB 6:60 dB
1.7 1.8 2.2 1.8 1.8 4.4 1.8
Stop-band atten. >45 dB . >100 dB >90 dB >100 dB >100 dB >90dB >90dB
8·23 Chapter 8
tuned circuit of the i-f amplifier. The i-f
energy is amplified by theagc amplifier,
thC?n converted to de by means of an agc +2 TO +9V
DC AMP AGC FOR
rectifier. RI and Cl are selected to DC AMP. 2N4126
CA3028A.
provide a suitable decay time constant
AGC
(about I second for ~b and cw). QI and RECT.
Q2 function 'as dc amplifiers to develop
lN914 22k
the dc voltage needed for agc control of lN914
the i-f (and sometimes, rf) amplifier
stages. The developed agc voltage can be 01,
S5
Cf
AGC
ON
SOURCE
FOLL,
AGC
DIFF.
AMP.
AItC TO
U2.U3 (FIG.5)
QI1
5000
£/
CAPACITANCE ARE IN MICROFARADS I JlF I ;
'-----+......_'VI.I'v....-<l +f2V
formance is certain to result. Agc is OTHERS ARE IN PICOFARADS I pF OR JlJIFI;
I-F GAIN
RESISTANCES ARE IN OHMS;
mainly an operator convenience: It pre- kolOOO. Mol000 000.
C B E
vents loud signals from blasting out of
the speaker or headphones when the op-
erator tunes the band at a given af-gain Fig. 41 - An age syst!lm for CA3028A i-f amplifiers, An op-amp is used as difference amplifier to
setting. provide age voltage while operating an S meter, .
Beat-Frequency Oscillators
The circuits given for crystal-controlled
oscillators earlier in this chapter are
suitable for use in BFa circuits. A beat A BFa need not be crystal controlled. fier helps to boost the oscillator output
oscillator generates energy which is It can use a VFO type of circuit, or it can to satisfy the requirement.
supplied to a product detector for \>e tuned by means of a Varactor diode
reception of cw and ssb signals. The BFO (see chapter 4 for a discussion of S Meters
frequency is offset by the appropriate semiconductor theory). Elimination ofthe Signal-strength meters are useful when
amount with respect to the center crystals represents a cost savings to the there is a need to, make comparative
frequency of the i-f filter. For example, a builder, but fr~quency stability may not readings. Such might be the case when
BFa used during cw reception is usually be as good as when crystal control is another operator asks for a comparison
some 700 Hz above or below the i-f center employed. between two antennas he is testing.
frequency. During ssb reception the offset When the BFa is operated at frequen- Because S meters are relative-reading
is slightly more - approximately 1.5 kHz cies above 3· MHz it is helpful to use a instruments, signal reporting based on the
above or below the i-f center frequency, buffer stage after the oscillator to mini- amount of needle deflection is generally
depending upon the need for upper or mize the effects of pulling. Furthermore, without meaning. No two receivers render
lower sideband operation. Typically, the if a passive product detector is used in the the same reading for a given signal, unless
BFa is placed roughly 20 dB down on the receiver, a substantial amount of BFO by coincidence. This is because the gain
slope of the i-f passband curve for ssb output power will be required - ap- distribution within an amateur receiver
reception or transmission. proximately .+7 dBm. The buffer/ampli- varies from band to, band. Since most S
8·25 Chapter 8
LAST DET.
01 02 r--------..,
I
I-F TRANS.
I TO
I I >4---QAF
TO I-F I I 01 02 AMP.
AMP. I I
I I
I I Cl 1.5 MEG. C2
330 47k
LEVEL
L:T T3~0 ' ~9 ,.OFF
0.5;FrJ, rh rh
L~~io-......-"'--'\47f\0f\kr---~
AF SHUNT
Fig. 43- A simple audio limiter/clipper. R1 sets
the bias on the diodes forthe desired limiting (Al
level.
TO AGC RECT..
LAST I-F f----<OPROO. DEc. AND
AMP. A-M DET.
NOISE
AMPLIFIER SWITCH lTO DRAIN OF
AGC PULSE FIRST I-F
Q3 .
DETECTOR MPF102 . 0.001
G
2.2M
lN914
FROM INPUT
TO FIRSTo---j
I-F AMP. 47
o.o~
+12Vo-----~------~----------------......--------------------~
Fig 45 - Diagram of a noise blanker.'C 1 and L 1 are tuned to the receiver i-f,
RL
8-27 Chapter 8
RC ACTIVE FILTER I~UT
r------i_----------_---O + 12 V Rl
~-------.
R3 = 2Q
ClJ C
o
If Ho = 2,fo = SOO Hz, Q = 5
'e • 1000 Hz and C = .022 ~F
R\ = 22.6kO(use 22k)
R2 =9420 (use lQOO)
R3 =90.4kO (use 9lK, or lOOk)
Fig. 47A - Circuit example of one pole section of an RC active audio filter which uses discrete Fig. 478 - Equations for designing an RC ac·
active devices, Q1 an~ Q2. tlve audio filter.
3.5-W
DRIVER AF OUTPUT
470
08
2NII8801
!l700!
100,!!F
AF I!!V 1000 10k
PREAMP. r-P+ 1000
2000,!!F
2!!V J3
• ~.np~gNis
1000
OR
TO .h!E 1000 8-0HM
PROD. 111\ SPKR.
DET.
1..'- ~--.....+-{
e-.--4
XI
(FIG.!!)
09 .
2N!!881/S7002
~ 1.8M
POLY.• POLYSTYRENE
seOk
24k
"i%
FL5
RC ACTIVE +12V
750-Hz CW
FILTER
Fig. 48- Practical circuit for a two-pole cw RC active filter. showing how it can be switched into and out of the audio channel of a receiver.
~11~t}:
to provide cw selectivity, but reception of
cw and ssb signals is entirely adequate for
I this first effort at receiver building ..
The circuit of Fig. 50 is designed for
headphone output. The overall receiver
sensitivity is suitable for comfortable
reception of even the weaker signals when
a resonant antenna is used. The front-end
filter, LljL2jCljC2jC3, is designed to
work into a 50-ohm antenna. This network
is fixed tuned, so it does not have to be
adjusted across the tuning range once it 50pF
has been preset fJr midrange in the r-----------~----------
coverage of the main-tuning dial. If a I 7.0 -7.2' MHz VFO
50-ohm antenna is not available, the user I OR Q4
, 3.65-3.8 MHz MPF102
9.1 V
I C6 0
1ao
I
SHIELD a o x _ '
I
I
I
,
. MAIN I
TUNING ,
L-----------;h----------~ ,
Fig. 50 - Schematic diagram of the direct-conversion receiver. Capacitors are disk ceramic
unless noted otherwise. Polarized capacitors are electrolytic. Resistors are 114· or l/2·watt
C2, C3 - Mica compression trimmer, Arco 427 for 40·m eter version,
Fig. 49 - Front view of the beglnners·recelver. orequiv, D1 - Silicon switchino diode. 1N914 or equiv,
The panels are made from double·sided pc C5- Miniature air variable, Hammarlund HF·15 D2 - Zener diode, 9, l·Y, 400 mW or 1 watt,
board material. or similar for BO·meter version, HF·10 or similar D3 - Silicon rectifier diode, 50 PRY, 1 A suitable,
8~29 Chapter 8
Table 3
Compqnent Data for the Direct·Convers'on Receiver
Bafld LI, L2 (II-H) CI (pF) C2 (pF) C3(pF) C4 (pF) C5 (pF) C6.(pF) C7, C8 (pF)
40 M 2 .5 240 nom. 198 nom . 42 s.m. 10 max. 200 560
20 ts. no. Arco 427 Arco 427 poly. or poly. or
, 20 enam. on s.m. s.m.
T68-6 toroid
core. Use 2-
turn link on L1.
80M 8.7 7 184 nom. 116 nom. 68 S.m. 15 max. ' 300 950'
40 ts. no. Arco 427 Arco 427 poly. or poly. or
24 enam.on s.m. S.m.
T68-'6 toroid
core. Use 2-
turn link on L1.
S.M. = silver mica. Poly = polystyrene. L3 nom. inductance for 40 meters = 4.51'H (J. W. Miller 42A476CBI or equivalent). L3 for 80 meters has nom. in·
ductance of 9.61'H (J. W. Miller 42A105CBI or equivalent). T1 consists of 15 bifilar turns of no. 24 enam. wire on an Amidon FT 50·43 toroid core. Twist
wires 6 times per inch before winding. C2 and C3 are miniature mica compression trimmers.
can install a small antenna-matching method is to adjust the matching network through Tl, driving the source of Q2.
circuit between the receiver and the by means of a transmitter and SWR VFO injection is on t1K: detector gate. Q3
antenna system (a small Transmatch) to indicator. Then the receiver can be is the audio preamplifier. It is direct
provide the proper termination for FLl. connected in place of the transmitter and coupled to the drain of the product·
A rough approximation of the proper FLl adjusted as outlined later. detector. An additional 40 de of audio
settings for the tuner can be found by Q 1 of Fig. 50 operates as a fed-back gain is provided by U 1, an operational-
adjusting for maximum 'sensitivity while broadband amplifier. Output from Q1 is amplifier IC. A 330-ohm resistor is in
listening to a weak signal. A better routed to the product detector (Q2) series with the headphone line to permit
the use of 8- and 16-ohm phones. Without
,the resistor the IC tends to self-oscillate at
audio frequencies. The resistor is not
required whe.n the headphones have an
AF impedance of 600 ohms or greater:
AF AMP. A protection diode, D2, is included to
PREAMP. prevent accidental damage to the receiver
if the power supply is attached with the
+40d8
wrong polarity. Positive voltage will pass
03 2"F 820k through the diode, but negative voltage
+16~V
2~3906
RI . will be blocked.
B
Ok PHONES The VFO uses only one transistor.
50Q,uF
16V
r-
1000
AF
. 2)1t
16v f---o----v~
+2)1F
16V
J2
Voltage to the drain of the FET, Q4, is
regulated at 9,1 volts by means of a Zener
diode, D2. D 1 helps to stabilize the
L
GAIN 50)lF oscillator by limiting the positive swing of
16V
the sine wave during oscillation. Table 3
/~
J: 16V
contains the correct values for the VFO
parts, respective of the operating band,
which do not have specified values on the
J3
DI +13V schematic diagram of Fig. 50.
+12.3V
Construction
The receiver is built from pieces of
Q2,Q4 Q3 Q1 double-sided pc board, except for the
etched-circuit board which is single sided.
The overall layout is larger than need be,
18
2
3
4
TOP
U1
8
7
6
5
in order to make the project less difficult
for inexperienced builders to assemble.
The VFO is enclosed in a separate
compartment made from three pieces of
• • PHASING VIEW
EXCEPT AS INDICATED, DECIMAL
double-sided pc board. The side pieces
VALUES OF CAPACITANCE ARE measure 2-1/2 X 3 inches (64 X 76 mm).
IN MICROFARADS (JIF I ; OTHERS / The rear wall of the VFO box is 2 X 2-1/2
ARE IN PICOFARADS (pF OR JlJIFl;
RESISTANCES ARE IN OHMS; inches (51 X 64 mm). The rear corners of
k -1000. M'I 000000 the box are joined by flowing solder along
the inn~r seams. A 1-1/2 X 2-inch (38 X 51
mm) plate is soldered inside the front of .
composition. Numbered components which do not have assigned values are so deSignated for
the compartment, 1/2 inch (13 mm) from
layout purposes. Others appear in Table 3. FT indicates feedthrough type. the front panel. This plate is also made
J1 -:.. SO-239. R1 - 1O-kO carbon control, audio taper from double-sided pc board. It is the
J2 - Two-circuit phone jack. preferred. mounting plate for the main-tuning
J3- Single-hole mount phono jack. U1- 741 op amp, 8-pin dual in line. Mount capacitor. Three no. 6 spade bolts hold
02,04 - Motorola MPF1 02 JFET. directly on pc board or use 8-pin IC socket. the VFO box to the main circuit board.
Receiving Systems 8~30
Fig. 52- Outside view of the simple superhet
receiver. The front and rear panels are made
from d.ouble-sided circllit board.
TO Rl HIGH
TO J3 VIA D3 (+13V)
Fig,53- Scale pattern of the pc board showing parts placement. Shown from the co~ponent side of the board,
to front;end overloading and sensitivity. i-f selectivity element, FL2. This p~rt was
.
Miller 8814 transformer/filter is specified.
Five transistors, four diodes and one IC chosen because of the low cost. However, It contains a monolithic filter that
comprise the semiconductor count in the any of the Collins 455-kHz mechanical provides a 4-kHz ban9width at the 3-dB
design. A supply voltage of 11 to 14 is filters designed for ssb bandwidth can be points of the response curve. It is
suitable. The current drain is on the order used, provided the insertion less is low. suggested that the Miller 8814 be used in
of 50 rnA maximum. The only circuit changes necessary would place of a mechanical filter to ensure
Fig. 54 shows the schematic diagram of be modification of the terminating re- maximum overall receiver gain and
this receiver. A fixed-tuned Butterworth sistances, the filter resonating capacitors, reduced cost. An extra stage of i-f
front-end. filter permits coverage from and the BFO crystals (Y 1 and Y2). This amplification milY be required with some
3.75 to 4.0 MHz without the need for model calls for 2700 ohms at each end of mechanical filters in order to have
retuning the filter. A high-transconductance the filter, plus two 360 pF resonating I sufficient headphone output on weak
dual-gate MOSFET, Q 1, serves as the capacitors. The filter bandwidth is 2.2 signals. ' .
mixer. Conversion gain with this device kHz at the 3-dB points. The 60-dB Output from FL2 is routed to a
(3N211) is very high, owing to the gm of bandwidth is 5.5 kHz. A lower-cost single-stage i-f amplifier, Q2, another
30,000 micromhos. A Collins 'Radio CB- 455-kHz filter alternative is offf!red in the 3N211 FET. The gain of this amplifier,
type mechanical filter is shown as the modification diagram of Fig. 55. A l w. plus that of the audio-amplifier ICUI is
Receiving Systems 8-32
Fig. 54~ Schematic diagram cif the ssb receiver.
Fixed-value cS"pacit6rs are disk ceramic unless
MIXER
noted otherwise'. Polarized capacitors are'
electrolytic. Fixed-value resistors are 1/4-or 1/2-
Ql
walt composition, . 3N211
Cl, C2- Mic;a compression trimmer, 300 pF 3.75- 4.0 MHz
max. Arco 427 or equiv.
C3- Miniature 25-pF air variable, Hammarlund
RFC1
HF-25 or similar. IOmH'
C4- Circuit-board mount subminiature air .
variable or glass piston trimmer,·l 0 pF mix. ANT.
(50n)
NPO; miniature ceramic trimmer suitable as
second choice.
01,- LED, any colo(or size. Used only as 1.5-V
reference diode.
02, 03 - Silicon switching diode, 1N914 or
equiv. 50
rL°. 1
,controlled manually by means of a dual that take-off point. Some oscillator Audio gain is provided by UI, a40-dB
control, RIA/RIB. The bias on gate 2 of pulling (slight) will be observed when FL I op amp. The receiver output is adequate
Q2 is varied at RIA to set the i-f gain is aligned. However, once the front-end for weak.signals (l /.tV or greater) with the
level. In order to obtain a wide range of filter is tuned the effect will not be arrangement shown in Fig. 54. Addition
control it is necessary to place gate 2at a noticed. A VFO buffer/amplifier would of a 0.5 or I-watt audio IC would enable
volt or two minus with respect to gate I. resolve this condition, but is not necessary the builder to employ a speaker rather
This is achieved by "bootstrapping" the in practice if circuit simplicity is to be than phones, but the current drain of the
stage with DI, an LED which conducts at keynoted. D2 is used as a switching diode receiver would be considerably higher.
roughly 1.5 volts.' Therefore, when RIA to offset the VFO frequency when
has its arm at ground, gate 2 of Q2 is changing from upper to lower sideband. Construction Notes
effectively at -1.5 volts (minimum gain). This eliminates the need, to readjust the The' etched-circuit board is the single-
D I serves purely as a reference diode in main-tuning dial of the receiver. This part sided variety (copper on. only one side).
this instance. Alternatively, two IN914 of the circuit need not be included if dial Double-sided pc board material is used
diodes can be used in series from source to calibration is not essential when changing for the front· panel, rear panel and the
ground to provide a reference of roughly sidebands. . VFO enclosure. Constructional details
1.4 volts. A 3N211 is employed at 0'4 (gates tied and dimensions follow closely those given
A 3N2ll FET serves as the VFO. Gates together) as the product detector. Output for the direct-conversion beginner's re-
I and 2 are connected together to sim~late from the i-f amplifier is fed to the source ceiver treated earlier in this chapter.
a single-gate transistor. Injection for the of Q4. BFO energy is supplied to Q4 from Therefore, that information wiIl not be
mixer is taken from the gate of Q3 in another 3N211 (Q5). YI and Y2 are repeated here .. Pc boards, negatives and
order to realize a 3-volt pk-pk lev~1 at gate selectable to permit upper and lower parts kits for this receiver are available
~ of Ql. A pure sine wave is available at sideband reception. Since lower-sideband from Circuit Board Specialists, listed in
transmissions are the general rule on 75 the text for the foregoing beginner's
meters, the builder may elect to use only receiver. An adaptor pc board is available
'The LED should be installed so that it illuminates
the Isb crystal and eliminate SI and the from that supplier to permit using the
when R.1 is set for maximum receiver gain. VFO offset circuitry. modification shown in Fig. 55.
8~33 Chapt.r 8
I-F AMP. 453.5 kHz
02
3N211 r------..,
T1
~~----~--~I~~, I
: I
12k!. ~;
I I
~-- ----~ 1
TO
04
SOURCE
--------------------------------------------~~~---------4~~._~ +12V
(TOP)
B
7
6
5
receiver showing the locations of the various
components.'The VFO and its enclosure are at
the center of the etched-circuit board adjacent to
the front panel. The layout seen here differs in
some areas from the artwork in Fig. 57 due to late
(BOTTOM)
k -1000. M-l,OOO,OOO. circuit changes.The speaker on the rear panel is
not used.
AF AMP.
Fig, 57 - Scale template of the circuit board and the parts-placement guide,
regular 455-kHz Collins mechanical filters The main frame of this receiver is the spread of the VFO. A 10: 1 vernier drive is
in amateur receivers, FL2 was chosen for on'e described by WIFB in QST for June recommended for the VFO if that is done.
this project because the price is roughly and July of 1974. It tunes the loo-meter Analog readout over a. 500-kHz tuning
one-third that of the mechanical filters band and provides coverage of the 8P-, range is rather impractical, owing to the
used by most amate,ur builders. 40-,20-, and IS-meter bands by means of lack of dial resolution which will result. If
down-converters which are switch- the wider frequency coverage is desired,
A Medium-Grade Receiver selected from the front panel. The the bandwidth of the converter circuits
This circuit, represents a fair tradeoff loo-meter tunable i-f receiver covers only will not cover all of the 80- and 40-meter
between simplicity and good dynamic 200 kHz. Hence, any 200-kHz segment of bands. There will be some loss of front-
range. The packaging format (Fig. 58) is the higher bands can be accommodated end gain at the edges of those bands.
pretty much up to the builder. Etched cir- with a given down-converter. The builder
cuit boards and pa~ts kits can be obtained must decide if cw or ssb is his or her Main-Frame Circuit
from a supplier. 2 choice when ordering the converter and The tunable i-f receiver (maIn frame) is
BFO crystals. Those wishing to use a shown schematically in Fig. 59. Ll pro-
'Circuit Board Specialists, P. -0. Box 969, Pueblo, digital frequency readout can extent the vides a 50-ohm input to the tuned circuit
CO 81002. Tel. 303-542-5083. Negatives, pc tuning range of the main frame to 400 or (L2) from the down-converters or a
boards and complete parts kits available for this
project. 500 kHz by increasing. the frequency lOO-meter antenna. Tl is tapped across
8-35 Chapter 8
receiver. Output is sampled from the pro- ,
duct detector, then amplified by means of
U3. D3and D4 rectify the amplified audio
and drive a cascaded dc amplifier con-
siiting of Q6 and Q7. The agc time con-
stant is set by a I-j.lF capacitor in parallel
with a IM-O resistor. The time constant
\=an be varied to suit the operator by
changing the value of the charging'
capacitor in that "network. Values of less
than I j.lF will shorten the discharge time
and vice versa. D5 serves as a gating diode
to prevent loss of the agc qrive voltage
through the i-f gain control, R3.
The agc strip drives the S meter, Ml,
through Q7. Agc voltage for U1 and U2 is
obtained at the emitter of Q7. R2 should
be set so that maximum receiver input
signal provides + 9 volts at pin 4 or Ul
and U2.
Local Oscillator
Q13 operates as a highly stable series-
Fig. 58 - Photograph of the miniature 5-band superheterodyne receiver. This layout arrangement tuned Colpitts oscillator. Polystyrene
was developed by WA0UZO of Circuit Board Specialists. fixed-value capacitors are used to ensure
stability and offer drift compensation
which corrects for the positive drift of the
core material in Ll2. Silver-mica capaCi-
tors can be substituted with a possible
degradation in long-term stability. C2 is
part of the tuned circuit at 500 and is a + 9. Minimum gain is at the + 2-volt agc the main-tuning'capacitor. C4 is optional.
center-tapped broadband transformer. level. If it is used it should be panel-mounted to
permit dial calibration from band to
Mixer and Filter Product Detector and Audio Channel band. This will be necessary because the
Two 40673 or 3N211 MOSFETs are A singly balanced product detector is converter crystals may not, be precisely on
used as a singly balanced active mixer (QI fed from the low-Z secondary winding of the specified frequencies.
and Q2). Local-oscillator injection is ap- . T4. BFO injection is supplied by Q5, QI4 serves as a source-follower buffer
plied to the paralleled gates no. 2 of QI which is shown with a 454.3-kHz crystal stage. The signal level is built up by means
and Q2. Output from the mixer is by for use with the Collins cw filter specified of Q15, a Class A broadband amplifier.
means of T2,a broadband trifilar-wound for FLl. If a Collins ssb i-f filter is used at Sufficie'nt LO output is available at 500 to
pot-core transformer. A to-kO resistor is FLl, the BFO crystal will have to be of a swing the no. 2 gates of the balanced mix-
bridged across the mixer drains to limit frequency that falls 20 dB down on the er to 3 volts pk-pk. The added power
the signal swing - an aid to mixer IMD. filter curve, upper or lower sideband. The developed by Q IS is necessary to provide
FL2 filters the LO voltage to keep har- manufacturer specifies which frequencies the required injection voltage across the
monic energy to the mixer at a low level. are correct for the filter being used. For 50-0 10acJ presented by harmonic filter
The 68-pF capacitor between FL2 and the upper and lower sideband operation it will FL2. A 40673 can be used \n place of the
mixer gates can be changed in value to ob- be necessary to use two BFO crystals. 3N211 specified at Q13. Similarly, a
tain the 3-volt pk-pk LO voltage which is Thus, a selector switch will be needed be- 2N4416 will be satisfactory at Q14. '
specified. tween the gate of Q5 and the two crystals.
An MPF102 or 2N4416 JFET is used as RC filtering is used at the output of the Front-End Converters
a post-mixer amplifier (Q3). The drain produ"t detector to keep the BFO and i-f ' The same circuit-board pattern is used
load resistor is chosen to match the filter leakage from reaching the audio for the 40, 20 and IS meter converters. In
impedance. A 2-kO value is specified for preamplifier, Q4. Any JFET will suffice order to obtain a 200-kHz bandwidth on
use with the Collins F455FD-04 filter at at Q4~ such as a 2N4416. 80 meters, FL3 is used. Also, no rf amp-
FLI. If a different filter is used, such as Q17 functions as an audio muting lifier stage is necessary on 80 meters.
anssb type, the drain load resistor may switch during transmit. It is actuated via Therefore, the pc-board pattern is dif-
have to be changed to match the im- J2 by grounding the Q 17 b~se line through ferent from that for the other hf bands.
pedance of the filter. Also, the end an external set of relay contacts or The conv~rters of Fig. 60 are designed
resonating capacitors at FLl may need to bipolar-transistor dc switch. for a broadband i-foutput of 1.8 to 2.0
be a different value (see manufacturer's U4 amplifies the audio to speaker level. MHz. They are, selected for the desired
data sheets). LCfiltering is used at the output of U4 to operating band by means of S I. When this
suppress unwanted hf oscillations which switch is placed in the 16O-meter position
I-F Amplifier could interfere with overall receiver per- the' converters are bypassed to permit
A two-stage CA3028A i-f amplifier is formance (spurious responses). The routing the antenna directly to the mixer
used following the i-f filter, FLI. Series MC1306P IC is designed for low-voltage of the main frame for reception on 160
regulation is applied to the operating operation. Therefore, the 3-terminal meters.
voltage of this circuit in order to provide regulator, US, has been included in the Q8 performs as a low-gain, common-
9.1 volts. Automated gain control i~ circuit. gate rf amplifier. The source tuned circuit
developed by the agc strip and supplied to is peaked for the center of the band seg~
pin 4 of the CA3028As through Q7. Maxi- AGCSystem ment of interest. It is broad enough in
mum gain occurs when the agc voltage is Audio-derive~ agc is used in this response to require no additional tuning.
Receiving Systems 8-36
POST-MIXER
BAL.
MIXER AMP.
\
03
01 455 kHz 455kHZ
MPFI02,
lfl"~::O"' !~
40673
1,8-2.0 MHz
r----<'---,
Za40:1
•
,.to. 1
08
3,9V
56 1'W
100 +13V
0,1
2255-2455,ktlz
(50 OHMS)
CALIBRATOR
;J:'200 ;;::'200
+13V
ON
10 •
S' TTOARM
OF R5
(FIG,601
C7
45
1000
SOURCE BROADBAND
OSC.
FOLLOWER AMP,
013
3N211
180 +13V
2255-2465 kHz
0,1
MAIN
TUNING POLY.
Fig. 59 - Schematic diagram of the receiver mllin frame. Fixed·value capacitors are disc ceramic unless noted otherwise. Poly. signifies polystyrene
capacitor. Polarized capacitors are electrolytic or tantalum. Fixed·value resistors are 1/2·watt composition unless noted differently.
C1 - Miniature 100·pF variable (panel· J2 - Phono j a c k . ' L12 - 17 to 41 IlH high·Q slug·tuned inductor
mounted). J3 - Three-conductor closed-circult jack. (au = 175 at fo ). (J. W. Miller Co.
C2 - 50·pF air variable (vernier driven). Should L1 - 2 turns no. 26 enam. wire over the 43A335CBI or e~ulv.).
have double·bearing format and low torque. ground end of L2 winding. M1 - Panel meter, 0-1 mA dc.
C3 - Miniature 30·pF air trimmer. L2 - 70 turns no. 26 enam. wire on a T68-1 R1 _ Audio-taper, SO-kn composition control.
C4- Miniature 10·pF air variable (panel· toroid core (55 IlH). This core and others R2 - Pc-mount SO·kO control.
mounted).. avail. from Amidon Assoc., G. R. Whitehouse R3 - L1near·taper, 10·kO composition control.
01·06, incl. - Silicon switching diode. and Palomar Engineers, (see QST ads). R4 - Pc-mount, 100·0 comPosition control.
07,08 - Zener d i o d e . . L10, L11 - 3-IlH inductor. 26 turns no. 26 RFC1- Miniature 1Q-mH,choke (J. W. Miller
FL 1 - Collins mechanical filter (see text). enam. wl,re on t37·2 toroid core. 70F102A1 or equiv.). .
The same is true of the drain tuned circuit. cordingly. This will depend on the actual The principle of operation for the
A .broadband bifilar-wound transformer, gain of QIO and the activity of Y2. Sub- SO-meter converter (Fig. 61) is similar to
T6, couples the mixer output to L1 of the miniature coaxial cable, such as RG- that of the other unit~. FL3 is a fixed-
main receiver. LO injection to gate no. 2 174/U, should be used for all connections tuned bandpass filter. The 120-pF series
of Q9 should be on the order of 3 volts' pk~ to SI and other distant parts ofthe circuit. center capacitor could be replaced by a
pk. The 15-pF coupling capacitor between The shield braid must be grounded at each 200-pF miniature trimmer to permit pre-
QIO and Q9 may need to be chosen ac- end of each run of coax. cise peaking atmicj, frequency. This
1.2K
* HE'AT SINK
** PC MOUNT
S.M.' SILVER MICA
• • PHASING
TO 07 (AGC) POLY.' POLYSTYRENE
I~F
+~~F
J3 PHONES
~ +/~~~t-r---1
50k l-
5.6 K
BFO
f
AF
GAIN
SPKR
05
MPF102 AGC AMP
820 K
~470
RFCI
10mH
rl°. 1
100
+13V
.---------------~--~~----~+13V
,,+;+ 470~F
25V
RFC2 - Miniature 33·,..H choke (J. W. Miller T2 - Trifilar·wound bobbin with 50 turns of no. U3 - 741 op·amp IC (any brand).
70F335A1 or equiv.). 28 enam. wire on PC·2213·77 pot core U4 - Motorola audio IC.
RFC4, RFC5 - Miniature 1·mH choke (J. W. (L = 10 mH). Amidon Assoc. core. U5 - Three·terminal regulator IC, 8 V.
Miller 70F103A1 or equiv.). T3, T4 - Miniature 455·kHz i·f transformer Y1 - BFO crystal, HC-6/U holder. International
S2 - Spst toggle or slide switch. (J. W. Miller 2067 or equiv.). Crystal Mfg. Co. type GP, 30·pF load
T1 - Primary with 6 turns no. 26 enam. wire T5 - Broadband bifilar·wound transformer. 10 capacitance. Choose frequency for filter
over ct portion ·of secondary. Secondary has bifilar turns no. 24 enam. wire on FT50·43 used (see'text).
36 turns no. 26 enam. wire, center tapped, on toroid core U<; = 950),50 ,..H. Y4 - 1oo·kHz crystal.
FT50·43 toroid cor~ ijl; = 950). U1, U2 - RCA IC.
practice could be extended also to the two (R5) to prevent front-end overloading in 50-0 input impedance.
750-pF fixed-value capacitors. In that case the presence of extremely strong signals.
a pair of 680-pF silver-mica units could be The builder may wish to replace this with Construction
used. Each would be in parallel with a a T or pi type of step attenuator which The layout of this receiver is left to the
200-pF miniature trimmer. presents a constant 50-0 impedance to the whims of the builder, The major consi-
A simple rf attenuator is included ahead receiver. This is especially significant with deration is that the various oscillators be
of the converters and the main receiver respect to FL3, since it is designed for a shielded from one another and from other·
MIXER
7 MHz
Jl 7 MHz
ANT. (eon)
~0.01
t-~+~13~~--------+-----------~------~~--~.-o+13V
47>
o
o
100
5.M.
Fig. 60 - Schematic diagram of the 40·, 20· and 15·meter converters, shown for 40·meter operation. Fixed·value capacitors are disc ceramic unless
otherwise indicated. Resistors are 1/4· or 1/2·W composition. -
C5, C6 - Miniature ceramic or mica trimmer, wire (8 ,..H) On T50-6 torbld core. Tap 4 turns wafers to aid Isolation.
5 tb 25 pF. above ground. 15 meters: 25 turns no. 28 T6 - Bifllar broadband.transformer (70,..H).
Cfb - Feedback capacitor. 39 pF for 20 and anam. wire (4 ,..H) on T50-6 toroid core. Tap. Use 12 bifllar turns no. 24 enam. wire on
15 meters. 4 turns above ground. FTSO·43 toroid core (,..1 5 950).
L3 - 2 turns no. 28 enam. wire over ground L5 - 40, 20, 15: same as L4 but tap each coli Y2 - International Crystal Mfg. Co. type GP
end of L4 winding (40, 20 and 15 meters). at midpoint. (30-pF load) In FM·1 holder, fundamental
L4 - 40 meters: 50 turns no. 28 enam. wire (13 R5 - 500·() IInear·taper cqmposllion control. mode. 40 meters: 5.20 MHz. 20 meters:
,..H) on T50·2 toroid core. Tap 4 turns above 51 - 3·pole, 5·posltlon rotary wafer switch, 12.20 MHz. ~5 'meters: 19.2 MHz.
ground end. 20 meters: 44 turns no. 28 enam. phenolic or ceramic Insulation. Use three
8·39 Chapter 8
current amateur literature. Therefore, this oscillations should not ,be allowed to take . quartz element and the circuit to minimize
section of chapter 8 is devoted to design place in any part of a receiver. the generation of 1M products.
objectives, circuit techniques and some 5) Wide-range age. The agc circuit Careful attention must be given to cor-
practical examples. This will serve as the shou'ld engage at low signal levels and rect filter termination and input/output
basis for individual design'S which can be hold the receiver output at a constant resonance to ensure minimum passband
carried out. by the more experienced plateau over a wide range of input-signal ripple (unwanted dips in the nose of the
amateur. . levels. For example, the audio output response curve): Most filters require a
The interest in building homemade should remain constant in amplitude over specified external terminal capacitance to
receivers of the more complex variety has a range of input signal from less than a resonate the inpJ1t and output trans-
waned in a tragic fashion. during the past microvolt to better than 10,000 /AV, formers within the filter module;' Similar-
decade. This has been brought on by an depending, on the external noise level ly, each filter has a characteristic input
increased interest in operating and which reaches the receiver front end. The and output impedance which must be
through the availability .ofsophisticated agc attack time should be set so that matched to the source and load.
receivers and transceivers found on the "pumping" and "clicking" is not noted 9) Detector and audio channel. An
commercial market. For this reason it when strong signals are received. otherwise excellent receiver can be spoiled
seems prudent to devote this portion of 6) Local oscillator. Not' only must the by an inferior product detector or audio-
the Handbook to design approaches. The local oscillator be stable, it needs to have amplifier strip. The detector must be able
information given here is based on circuit low noise and good spectral purity at the to handle the highest output signal from
and performance. investigations in the output. Ideally, the LO noise floor should the i-f strip without saturating. Although
ARRL laboratory. It is slanted toward the be 8tl to 100 dB below the peak output active product detectors are sometimes
practical side of design and application in voltage. ~purs and harmonics in the out- used, they are the most prone to the fore-
order to be of greater use to amateurs who put should fall at least 50 dB' below peak going malady. The preference of most
have no formal background in electronics. output. ,LO output energy must be con- designers is a passive diode detector of the
fined to the mixer' by means of ap- singly or doubly balanced variety. This
Perfoi'mfJnce Objectives propriate shielding and filtering. type of de~ector can handle high signal
What should an amateur look for in 7) I-f amplifiers. An i-f amplifier strip levels with large amounts of BFO injec-
terms of high performance When building needs to have sufficient gain to drive the tion. Since the detector is the lowest-level
. a receiver? A subjective outlook would detector and provide ample excitation to part of the audio channel, hum and noise
call for a lot of "bells and whistles" with the audio channel. The design should in- shou.ld be minimal at. that point in the cir-
which to play, but a discerning operator is clude active devices which can ensure a cuit. Passive detectors do not need
interested in performance under all of the collective 80 dB or greater agc swing over operating voltages; hence one primary
adverse conditions one might ehcounter in the input-signal range mentioned in item source of hum is avoided .
. the course of operating an amateur sta- 5. A low-noise audio preamplifier should
tion. The following are representative of Wide-band noise is generated within follow the detector in a quality design. It
the major considerations in receiver per- most i-f amplifier chains. An improve- should be able to withstand the maximum
formance: ment in receiver "noise bandwidth" can output from the detector at peak receiver
1) High dynamic range. This is the be realized by adopting the W7ZOI filter input signal without operating in a non-
ability of the receiver to perform well in "tailending" scheme which calls for use linear manner.. The audio gain control can
the presence of strong signals within and of a second i-f filter immediately after the be used to the best advantage when it is
outside the amateur band of interest. last i-f amplifier. The second filter can located after the af prea,mplifier. If it is
Poor dynamic range results in cross- have slightly greater bandwidth than the used ahead of the preamplifier, the noise
modulation effects, receiver desensitiza- filter used ahead of the first i-f amplifier. figure of the audio channel may be de-
tion and spurious responses from the mix- The tailend filter will reduce the wide- graded at low settings of the gain control.
er which appear in the tuning range as band noise components. This technique is Audio shaping is normally applied to
additional signals (mixer IMD). discussed in greater detail in Solid State the af channel to provide a low-frequency
2) Good selectivity. This feature in- Design for the Radio Amateur, by The rolloff at some frequency well above 60 or
cludes the receiver front end (rf amplifier ARRL. 120 Hz. This greatly reduces the chance of
and mixer) along with the i-f and audio 8) I-f filters.· In order for a filter to power-supply ripple appearing in the
selectivity. The objective (ideal) is to have function as such, there must be some in- audio output. Also, there is little need for
the receiver pass only those frequencies to sertion loss (IL) if a passive network is low-frequency response below, say, 300
which it is tuned, while rejecting all being used. The IL is typically highest Hz in a communications receiver.' Simi-
others. This Utopian goal can not be when a mechanical filter is used. This fac- larly, the high-frequency response should
realized, but it can be approached closely tor must be taken into account when plan- be restricted so that rolloff starts around
enough to ensure good performance. ning the receiver gain distribution. Most 2000 Hz. A satisfactory tailoring of the.
3) Low noise figure. The noise figure mechanical filters have an IL of 8 dB or audio passband can often be done by
should be such that it is somewhat below greater, whereas a well designed crystal- proper selection of the Rand C com-
the level of the receiver antenna noise lattice filter has a characteristic IL of less ponents in that part of the circuit.
under typical "quiet" band conditions. than 5 dB. All of the audio stages must operate as
This means that the noise generated In some designs the i-f filter is the linearly as possible up to peak signal
within the receiver - notably the early limiting factor in achieving high perfor- levels. This will minimize distortion and
stages - should be kept to an absolute mance in mixer IMD. This is because of aid weak-signal reception greatly. It is
minimum so that it does not mask the movement of the mechano-electrical prudent, therefore, to use an output stage
(degrade) weak incoming signals. contacts within the filter, which generate which is capable of delivering greater un-
4) High order of stability. All of the. 1M products which are independent of distorted power output than will ever be
receiver oscillators, crystal-controlled or those in the mixer. Laboratory investiga- needed. Cross-over distortion is to be
LC types, need to be drift .free in an ideal tions indicated that mechanical filters avoided also. The effects of this are most
circuit. Since this is practically impossible, were somewhat worse than crystal filters apparent under weak-signal conditions.'
maximum drift (long term) should not ex- in this respect, limiting the receiver IMD The signal has a "fuzzy" sound when this
ceed 50 to 100 Hz in a good design. The profile to roughly 95 dB. The crystal type of distortion is present. Some of the
greater the i-f selectivity, the more impor- holders in lattice filters must be able to audio-power les have' significant croSs-
tant the oscillator stability. Self- provide positive electrical contact with the over distortion which can not be cor-
Receiving Systems 8·40
rected. This is because the biasing is done fashioned inexpensively from pieces of networks are not highly lossy or poorly
within the IC, and it can't be changed. double-sided pc board. Modular construc- matched to the source and load. This
For this reason it is helpful to use discrete tion permits the amateur to try new cir- rationale applies to frequencies up to ap-
devices in the audio channel. This enables cuits within the receiver without disrupt- proximately 14. MHz. At 20 meters and
the designer to bias the amplifiers for ing the remainder of the circuitry. higher an rf amplifier may be needed to
minimum distortion. 11) "Bells and whistles": This discus- ensure 'an acceptable receiver noise figure.
Tuned audio amplifiers can be used to sion does not include such themes as syn- As a general rule the designer should
provide steep skirts outside the desired thesizers, i-f passband tuning, noise use no more gain in the rf stage than is
passband. An example of a simple blankers, computer-programmed func- necessary to obtain /!.n acceptable noise
application of this, using a single pot-core tions and digital frequency readout. These figure. The higher the stage gain, the
inductor with variable Q, was described are primarily matters of whim and subjec- greater the sensitivity. But, more gain
by K1TX in April 1979 QST. Various tivity, however useful they mignt be. than is needed will degrade the receiver
types of passive LC filters can be used to For reasons of practicality the builder dynamic range markedly, by virtue of the
obtain cw or ssb selectivity at audio. llJust declde whether he will use analog or mixer being fed larger amounts of input
RC active audio filters with variableQ digital readout of the receiver frequency. signal than if no rf amplifier was used. So,
and adjustable peak frequency offer an There are two disadvantages attendant to even at the very early part of a receiver it is
excellent means for limiting the audio analog systems: .(1) Quality dial vital to pay attention to gain distribution.
bandwidth, minimizing wide-band noise mechanisms are scarce and highly expen- This fundamental rule applies from stage
and reducing QRM. Ideally, these filters sive. (2) Readout resolution is usually to stage throughout the receiver.
should be contained in the low-level part poor if more than 206 kHz of any band is There should be sufficient selectivity
of the audio channel rather than at the covered. The major advantages of analog ahead of the rf amplifier (and in most in-
receiver output. This will prevent frequency readout are reduced circuit stances between it and the mixer) to
overloading of the filter, which can impair complexity, lower cost (sometimes) and restrict passage of signals outside the
the performance and introduce intolerable less current drain from the receiver power amateur band of interest. This will greatly
amounts of distortion. supply. Heating is diminished also - a reduce the probability of unwanted im-
10) Structural considerations: There definite benefit to stability. ages in the tuning range. Furthermore; it
can be considerable latitude in, the A frequency counter and a digital will help prevent very strong out-of-band
mechanical approach one takes when lay- display, on the other hand, permit commercial signals from entering the
ing out a high-performance receiver. 500-kHz frequency spreads with good receiver front end and impairing perfor-
Aesthetics have no place in this discus- resolution. A shaft encoder is needed for mance. This form of selectivity is called
sion. We will address the matter of struc- synthesized LO systems to avoid thumb- "preselection." It can take the form of
Jure versus performance and leave the wheel frequency selection. But, it is easy LC circuits which are very narrow in
beauty of the front panel to the builder. to use parts of the synthesizer circuit for bandwidth, and tracked' manually from
The major points of concern are rigidity the frequency countl'!r, thereby making the front panel: Alternatively, fixed-tuned
of the overall assembly and shielding the two circuits compatible. In this type of LC filters can be used to provide selective
against incidental pickup and radiation. system, or in one which has a conven- circuits. A bandpass type of filter or tuned
The chassis and panels should be strong tional'LO and a counter; a 10:1 vernier circuit is the choice of most designers
enough to prevent undue stress on the pc drive without detectable backlash is because rejection is offered above and
boards during flexing or vibration. In a almost mandatory to keep the tuning rate below the frequency band of interest.
like manner the local oscillator should be within practical limits. The choice between small-signal FETs
relatively immune to any mechanical It should be stressed here that counters and bipolar transistors in an rf preampli-
stress which is imposed on the receiver. can create noise and spurious responses if fier is more than arbitrary. FETs exhibit
An excellent assembly technique is one they aren't designed and used correctly. low noise figures at hf and they consume
that uses a modular approach for the Careful shielding and filtering must be ap- less dc power than bipolars for an
various key circuit assemblies in the plied to prevent the counter from affect- equivalant output intercept. Generally
receiver. Each module is contained in its ing other parts of the receiver circuit. spealdng, FETs are less subject to block-
own shield box. All signal leads entering The same general considerations apply ing in the presence of strong input signals.
and leaving the, various modules are made to synthesizers. The design must be car- Bipolar transistors, on the other hand,
from RG-174/U or similar coaxial cable. ried out with care to minimize phase have rather well defined input and output
The shield braid is grounded at each end noise, which can degrade the mixer noise impedances and can be used more easily
of each cable. Leads which carry dc are figure and the ultimate i-f selectivity. with negative feedback than is true of
decoupled where they leave the module Three QST articles are offered as FETs. These features make them ideal for
shield. L-C or R-C decoupling networks references on high-performance receivers. ensuring a proper and constant filter ter-
are suitable in most instances. Feed- They contain information which will be of mination (filters must be terminated cor-
through capacitors can be mounted on the value to the amateur designer.' rectly in order to perfofm in a proper
box walls of each module to serve as ter- manner). A common-source FET which
minal connections for the dc voltages, RF Amplifiers operates in the hf spectrum can not meet
while functioning also as parts of the When it is deemed necessary to use an rf the foregoing condition. The use of feed-
decoupling networks. Since 50-0 amplifier ahead of the receiver mixer, back (negative and degenerative) in a
miniature cable is suggested for intercon- thought must be given to gain, linearity, bipolar-transistor preamplifier makes
necting leads, the points to which they signal-handling ability and noise figure . possible a low noise figure and a good in-
.t connect in the circuit/should be designed The choice between bipolar transistors put and output match.
for a like impedance. This form of and FETs is another consideration. An rf Bipolar transistors which are designed
modular construction and shielding great: amplifier should not be necessary in a for CATV and uhf oscillator work, such
ly reduces the chances for "birdies" by properly designed receiver, even if a as the 2N5179 (biased for about 20 rnA)
keeping rf energy where it belongs. It also passive mixer is used, provided the input and the 2N5109 (biased for roughly 50
prevents unwanted external signals from rnA) are excellent for use as rt amplifiers
'Wes Hayward, W7Z0I, "A Competition-Grade
being picked up by low-level parts of the Receiver," QST, March and April 1974. Doug ahead of a mixer. Fig. 66 shows a practical
circuit. Miniboxes or die-cast aluminum DeMaw, WIFB, "His Eminence, the Receiver," example of an ARRL circuit which uses a
QST, June and July 1976. Jay Rusgrove, WIVD,
boxes are excellent for use in modular "Human Engineering the Station Receiver," QST, 2N5109. A photograph of the assembled
work. But, homemade enclosures can be January 1979. module is at Fig: 63_
8-41 Chapter 8
RF AMP. \
, +12V
1.8-2.0 MHz
FL2
30/S.M. '
1500
T5.M. 220 C4 150 150
s:M. 250
f CO ·2.2MHz
Fig: 62 - Schematic diagram of a high·level rf amplifier for a receiver. L1 contains 41 turns of no. 24 enam. wire on a T68·2 toroid core.(10 "H). L2
and L3 have 26 turns of no. 24 enam. wire on a T68·2 toroid core (4 "H). L4, L5 and L6 consist of 43 tl,lrns of·no. 24 enam. wire on a T68·6 toroid core
(8.74 "H). T1 contains 16 bifilar turns of no. 26 enam. wire on an FT50·43 ferrite core.
R~J~fl)
l
• • ,T1
II
02
03
01
04
OBM
~---------------+~
OUT +-----4f--I°.°f-('-----09 MHz IF OUT
501'
(51fl)
~
e:'-PHASING
T ~ i1S QL-1
Fig. 65 - Practical circuit for a doubly balanced diode·ring mixer. The components are discussed
Stackpole no. 57·9130 ferrite balun core. in the text.
the component values shown. Noise figure band input and output transformers at version gain of 6 dB. Output intercept is
is 4 dB at 30 MHz. A I-dB saturation specific lower- levels. (A thorough discus- + 23 dBm and the input intercept checks
power output of 3.7 watts was observed, sion concerning diode mixers and their out at +
17 dBm. indications are that
indicating the suitability of this type of behavior is presented in the League's with further experimentation with VMOS .
circuit for high signal-handling applica- book, Solid State Design for the Radio devices, mixer biasing and LO injection
tions. Amateur, chapter 6.) Fig. 65 shows a power the input intercept could be im-
The basic circuits of Figs. 66 and 68 are practical circuit for a DBM. Itindudes a proved to at least .+ 25 dBm in an op-
suitable also as post-mixer amplifiers in diplexer at the i~f port to establish a 51-0 timized case. The circuit of Fig. 66 was
receivers. The VMP4 is a fairly expensive termination for the mixer. This offers an biased for· a total mixer current of 50 rnA
Siliconix transistor. It is likely that one of improvement to the IMD level by a few with LO power (+ 15 dBm) applied. The
the lower-priced pieces, such as the dB over a similar rriixer with no diplexer. use of'Siliconix VMP-4 transistors should
VN66AK, would provide good service at The diodes can be HP2800 hot-carrier lead to even better mixer performance. A
h.f in the circuit of Fig. 64. types. Carefully 'matched IN914s are photograph of the assembled experimen-
sometimes used as substitutes. T1 and T2 tal mixer with its post-mixer amplifier and
High-Performance Mixers are broadband toroidal transformers filter is shown in Fig. 67.
Doubly balanced diode-ring mixers (baluns). For wideband use in the hf spec- Fig; 66 shows that a broadband linear
(DBMs) of the type discussed early in this trum the cores should have a high post-mixer amplifier is used. It is followed
chapter are often used to obtain high permeability. A 0.37-inch (9.4-mm) by a 6-dB pad and diode clamps. The pad
dynamic range. Among the advantages diameter ferrite core (Amidon FT37-43) provides a constant load for the amplifier,
are low noise (diode mixers generate very with a mu of 950 will work nicely. Ten .Q3, And stabilizes the filter termination
little noise) and broadband charac- trifilai" turns of no. 30 enamel wire can be (FL3). The diodes prevent, damage to FL3
teristics. The mixer noise figure is used for the windings. Output intercept in the presence of very strong receiver in-
approximately the conversion loss of the for this circuit is typically + 13 dBm with put signals. FLl filters the LO output to
diode ring - typically 7 to 8 dB. The the LO injection at + 7 dBm. This pro- ensure a clean injection voltage to the
balanced mixer circuit provides port-to- vides an input intercept of 20 dBm (output mixer. The VMOS balanced mixer ex-
port isolation which is not possible with intercept plus the 7 dB conversion loss =
hibits a noise figure of roughly 8 dB at 14
single-ended or singly balanl;ed mixers. 20 dBm). Calculations for a high-level MHz.
This feature can aid the mixer IMD and diode mixer, assuming a + 17 dBm LO
help to minimize spurious responses level (recommended), the output intercept (-F Amplifiers
resulting from the LO energy entering will be + 23 dBm. Again, assuming a 7-dB The criteria for i-f amplifiers are pretty
other par~s of the receiver circuit. conversion loss the input intercept well defined in the philosophy section of
The main shortcomings of diode mixers becomes quite desirable - + 30 dBm. this part of the chapter. The choice of
are the high level of LO injection needed This is based on the respective perfor- active devices for i-f strips usually leads to
(approximately + 7 dBm for most) and mances of the commercially available ICs. The Motorola MCI590G of
the necessity of proper mixer termination, SRA-l and SRA-IH DBMs. It can be seen MC1350P are commonly found in high-
especially at the i-f output port. This type from the foregoing that better mixer per- performance receivers. These and the
of mixer is subject to harmonic mixing - formance can be realized at the higher RCA CA3028A IC offer good gain and
another trait which the designer must deal LO-injection levels. The actual LO power agc range with low noise figures. So, the
with. applied will depend upon the ability of the choice will depend in part on availability
Some high-level diode-ring mixers are diodes to handle the current. and whim. Normally, just two IC stages
available commercially. They require a are used in an i-f strip.
high amount of injection power (+ 17 Singly Balanced VMOS Mixer Dual-gate MOSFETs are used as i-f
dBm for acceptable' performance). The circuit of an experimental active amplifiers by some amateurs, but at least
Laboratory analysis suggests that high- mixer with highlevel capability is shown in four stages are needed to approach the
level mixers misbehave as a result of diode ~ig. 66. Two VMOS power FETs are gain of two ICs designed for .the applica-
imbalance at specific current levels. The employed in a singly balanced format. tion. Furthermore, FETs do not provide
effect is one of the IMD not dropping 3 Laboratory measurements ,of the circuit the agc range of ICs unless the control
dB when the input tones are lowered 1 dB between T1 and T2 (50-0 terminations) at gates can be made to swing slightly
in level. This phenomenon could be 14 MHz, with a LO frequency of 5 'MHz negative. The usual gate no. 2 agc range is
caused in part by saturation of the broad- and an i-f of 9 MHz, 'yielded a mixer con- from - 2 volts to + 4 volts for full agc
8-43 Chapter 8
+6dB
BAL. MIX. 9MHz
FL2
T2 O.Ot +(2dB
POST - MIX.
AMP.
O.t 390
• PHASING \
* USE HEAT SINK
FB • FERRITS BEAD
FLt
aSk
'co· &.7 MHz
r
930 20V
,L0.t 0.4W •
to
tW
-I- 24
+24V
-6dB - edB
39 9MHz
400
eon r-~---1r-~~~~--~----o TO 500-n
1-1' AMP.
t
30 •
-I-tedBm
eo-nLO
INPUT
(s.O-e.eMHzl-
EKCEPT AS INDICATED, DECIMAL VALUES OF
CAPACITANCE ARE IN MICROFARADS I J!F I ;
OTHERS ARE IN PICOFARADS ( pF OR JlJIF1;
RESISTANCES ARE IN OHMS;
k'IOOO, M'IOOO 000.
Fig. 66 - Circuit details are for a VMOS power FET singly balanced mixer. L1 and L2 have 13 turns of no. 24 enam. on a TSO-6,torold core. L4 con·
talns 21 turns of no. 26 enam. wire on a TSO-2 core. T1 and T2 have 12 trlfilar tuins of no. 26 enam. wire on an FTS0-61 toroid core. T4 contains 7
primary turns of no. 26 enam. and 21 secondary turns of no. 26 enam. on an FTSO-61 toroid core. FL3 Is a Spectrum International a-pole crystal lat·
tlce filter. Bandwidth Is 500 Hz. •
control.
Fig. 72 contains the circuit of an i-f strip
which uses two Collins mechanical filters
in the "tail-ending" scheme discussed
earlier. The ssb filter is at the front end of
the strip and the narrower (400 Hz) filter
is diode-switched into the circuit for low-
noise cw reception. For'phone-only recep-'
tion both filters can be of a 2.I-kHz band-
width, and both would remain in the cir-
cuit' at all times. At considerably greater
expense one could have a cw and a ssb
filter at the output end of the i-f strip.
They would be selected by means of diode
switching to permit a tail-end filter to be
in the circuit for either mode. The
photograph in Fig. 69 was taken before
some circuit changes were made, so it does
not conform exactly to the diagram in Fig.
68. It does, however, illustrate the recom-
mended layout for good filter isolation.
During s$b operation FL2 is shorted
across by means of 01 and 02. Q2 is used
to equalize the overall gain of the receiver
when the modes are changed. It compen- Fig. 67 - Photogr~ph of the experimental VMOS high·level mixer. Circuit boards for this and
sates for the 10 dB of insertion loss caused other modules shown photographically In this section of chapter 8 are available from Circuit
by FL2. During ssb operation the gain of Board Specialists, Box 969, Pueblo, CO 81002.
5.6k
0.001 O.OO~
F. T.
F.~ +12V
~ I AGC
(+3;0 +9V)
AGC
I. AMP. T2
455~Hz
0.1
CW SSB
~o-_SI_C___0~.0>01_/O~T_.________________________~
,.L0"
EXCEPT AS INDICATED, DECIMAL VALUES OF
CAPACITANCE 'ARE IN MICROFARADS (JAF »;
OTHERS ARE IN PICOFARADS (pF OR jlJlF);
RESISTANCES ARE IN OHMS;
'-1000. M-IOOO 000.
FB = FERRITE BEAD
F. T. = FEEOTHROUGH
Fig. 68 - Schematic diagram of a 455·kHz i·f strip which uses filter tail·ending. See the text.
Q2 is reduced by virtue of SIC and Rl. operation. Sampling ahead of Q2 would pedance of VI provides a suitable ter-
Agc amplifier QI is used to prevent cause a to-dB differential in the agc action mination for FLl - roughly 20000. T2
loading across T2. The i-f energy is sam- between the cw and ssb modes. FLl should have a 20: 1 impedance step-down
pled at the drain,of Q2 so that the agc will should have an input termination of ratio for going into a diode type of pro-
be relatively constant for both cw and ssb 20000. The characteristic input im- duct detector. The value is not critical.
VMOS FET Audio Amplifier
The importance of clean audio in a
receiver was stressed earlier in this section.
An output level of I watt or greater with
less than to percent distortion is recom-
mended for amateur receivers .
. Fig. 70 shows the circuit of a simple
VMOS power FET Class A audio
amplifier which was based on a Siliconix
design and adapted for amateur use in the
ARRL laboratory. Output power into a
4-0 load is 2.:S watts at less than lO percent
distortion.
QI need not be a dual-gate MOSFET.
The 40673 was used in this model because
a JFET, such as a 2N4416 or MPFlO2,
was not handy at the time of assembly.!
Any low-noise JFET can be sqj)stituted at
QI. A I-volt pk-pk input tone (at SOO) will
produce 2.S watts of output. QI draws 1.2
Fig. 69 - Photograph of the assembled i·f strip from Fig. 72. Double·sided pc·board material is rnA and Q2 idles at 300 rnA, rising to
used. roughly 325 rnA at peak signal periods.
8·45 Chapter II
1~·~--------------------GPZ30dB--------------------~~
AF PREAMP
CLASS A
AF OUTPUT
40673 1k
0.01
1rV;oKr;.PK o--1
(2.5mW)
L-------------------------~+~+24V
CI is necessary to provide a roll-off at recommended in the interest of preventing The shield compartment is maddrom sec-
high auoio frequencies. It serves also to vhf oscillations. The copper on the com- tions of double-sideQ pc board. A feed-
. prevent hf and vhf self-oscillations at Q2. ponent side of the board serves as a through capacitor serves as a terminal for
The peak voltages and currents which ground plane and should be made com- . the 12-volt line to the BFO and helps to
result from self-oscillation can destroy Q2 mon to the ground foils on the etched side prevent rf energy from leaving the BFO
immediately (gate-source and drain- of the board at several points. A V-shaped module via the )2-volt lead.
source breakdown). 01 and 02 are heat sink-(not shown in Fig. 71) should be
necessary across TI to damp transients used at Q2. It can be made from 16-gauge Local-OsciUator Structure
which can occur when the field in T1 col- aluminum stock. The vertical members of The local oscillator system shown in
lapses. Without the Zener diodes 'in the the sink should be at least 1-112 inches (38 Fig. 74 is an assembled version of the
circuit, drain-source breakdown poses a mm) high. practical VFO shown in chapter 6. The
significant threat. compartment is made from pc-board sec-
. Feedback is taken from an 8-0 tap on Universal BFO tions. A V-shaped press-fit aluminum
Tl and supplied to the source of QI to Fig. 72 provides a diagram of a BFO cover is used to enclose the top of the box.
keep distortion at a minimum (observe the which can be tailored for use at any of the This kind of shielding is important for the
polarity at Tl). An 8-0 speaker should not popular intermediate frequencies. The prevention of stray radia,tion from the
be attached to the 8-0 tap, as it will lower constants shown are for use at 455 kHz. VFO into other parts of the receiver cir-
the feedback and increase distortion. For higher operating frequencies it will be ~.uit. The enclosed module tends to pre-
Almost any large-signal VMOS tran- necessary only to modify the feedback vent rapid internal changes in ambient
sistor can be used at Q2. The Siliconix capacitors of the oscillators (CI-C5, in- temperature -- an aid to frequency stabili-
VN67 AF and N89AF devices (plastic clusive). The higher the operating fre- ty. Double-sided etched-circuit boards are
TO-202 format) would be good choices at quency the lower the capacitance value. not recommended for LC oscillators. This
Q2. The division ratio estaolished by C4 and is because the etched foils and the ground
Tl is a 24-0 to 4 and 8 ohms C5 should be maintained at all frequencies plane form numerous low-value
transformer. A Triad TY -29X is shown in on which this circuit is used. This will pre- capacitors, none of which are stable: The
the model of Fig. 71. A homemade vent the 50-0 input impedance of Q4 from pc-board insulating material represents a
transformer would be easier to realize and loading the oscillators and preventing poor dielectric material for' stable
it would be quite inexpensive. The core them from functioning. capacitors.
stack from a I-ampere or large filament This circuit was designed especially for
transformer would be suitable as a foun- use with a diode-ring product detector, A High-P~rformance AGC Circuit
dation for a homemade unit. The turns which requires a substantial amount of in- Fig. 75 contains the circuit of an i-f strip
ratio would be 1.73:1 for the primary/8-0 jection power. A 50-0 pad can be placed and agc chain which offers excellent per-
windings. For the primary/4-0 combina- in the output of Tl if lower injection formance. This circuit was designed by
tion it would be 2.45:1. No .. 26 enameled levels are needed. Tl should be terminated WtZOI for use in his Competition Grade
wire will be suitable for the current taken in approximately 500 for the best perfor- CW Receiver. The complete receiver cir-
by Q2. Vse the maximum number of turns mance. cuit is found in Solid State Design for the
possible, consistent with the turns ratios YI, Y2 and Y3 provide the proper BFO Radio Amateur,. chapter 9.
mentioned. This will ensure sufficient in- frequencies for upper sideband, lower This agc circuit is suitable for use with
ductance to prevent the loss of low- sideband and cw. JEETs can be substi~ the i-f strip of Fig. 72. It employs a full
frequency audio. tuted for the 40673s at QI, Q2 and Q3. "hang" action. The agc is defeated by
Double sided pc-board material is The assembled BFO is seen in Fig. 73. means of S1. The time constant is selec-
BUFFER AMP.
C7
56 0.1
Y2 ;;:.; ~0.1
456.4 kHz 0.1
nl-----~
02
40673 3300
USB
.......~----""'f\//',v-. BFO 453.6-456.4 kHz
15 50- .n
(-,--0 OUTPUT
CW 0.1 (10 V pk-pk)
nnn
56
Y3
::h 0 1
. v vv
453.6 kHz
!1~-----------'lSB
40673 BFO
G2@OG.10
DO 0 S
EO
B
0 C
04
BASE
DC
2_2
pk-pk
8
COlL. 11.6 20
.BOTTOM VIEWS EMITTER 2.5 2
Fig. 72 - A 3·channel BFO for universal use in receivers. T1 contains 25 bifilar turns of no. 30 enam. wire on an FTSQ-43 toroid core (950 mu).
8·47 Chapter 8
r--c)--o----<O+~2V
---,
+42V
5000;::J;
c;,~-
I-F AMP.
0.1
47 ~ I
lNPUT
II
~1"
L24
1N4152
DC AMPLIFIER
+12V
20k
METER
ZERO
FROM
AUDIO AMP. 0.1
lK
~~~!--+-i
~
SLOW
52 GROUND
0.1 TO
FAST MUTE
'--------' AGC
Fig. 75 - Circuit details of the W7Z01 i-f amplifier and high-performance agc system. This circuit provides full-hang agc characteristics.
able at S2. RI at U3 should be set for + 5 in Fig. 41 of this chapter. It does not in- A broadcast-band filter is a good exam-
volts at pin 6 of U3 with the agc off. With '. corporate the hang feature used in Fig. 75. ple. The circuit and construction informa-
Q5 and Q6 as part of the circuit, the tion for such a filter appear later in this
receiver is practically silent after a strong RECEIVER ACCESSORIES section.
signal disappears from the passband. But, A number of receiver accessories have
after a timing period associated with net- become popular either through necessity A Stable to-Meter Preamp
work. CI-R2, the receiver will return to or through the operating convenience that When receivers run out of gas in terms
full gain in roughly 50 milliseconds. This they provide. Notable among these items . of front-end gain and noise figure, it
is very advantageous when loud pulses of are noise blankers, audio filters, usually happens on 15 and 10 meters. A
noise enter the receiver. The effect is preamplifiers, frequency converters and typical symptom is the need to carry the
similar to that of a noise blanker. A de- interference filters. Although only audio gain wide open to copy a weak
tailed description of this type of agc cir- relatively few amateurs choose to build signal. This problem can often be resolved
cuit can be found in chapter 5 of Solid their own receivers from the' ground up, by adding a preamplifier between the
State Design for the Radio Amateur, by many more are inclined to try their hand receiver input and the antenna.
the ARRL. at a somewhat less complex station ac- It is not beneficial to add gain ahead of
A less complex agc circuit for use with cessory - especially if it is something that a receiver that has sufficient sensitivity.
RCA CA302S-A i-f amplifiers is provided is needed for .successful station operation. The end produt't may be excessive front-
Receiving Systems 8;48
Fig. 76 - Exterior view of the enclosed
preamplifier. A U-shaped, press-fit aluminum
cover Is used. Fig. 78 - Inside view of the preamplifier.
enqgain, leading to the demise of common-gate (grounded gate) type. regardless of the load connected to the
dynamic range. If the latter is degraded Stability is relatively easy to obtain with amplifier. Properly applied feedback en-
significantly, cross modulation, IMD and the latter, but a tradeoff in gain will ac- sures unconditional stability. Circuits of
desensitization will become manifest when company modification. A common-gate this kind are ideal for amateur builders
strong signals are present in .and near the JFET amplifier usually provides between who have limited practical experience in
band of interest_ In other words, don't use 10 and 14 dB of gain, whereas a common- the workshop.
a preamp if your receiver performs source version will yield up to 25 dB of Fig. 77 shows the preamplifier circuit.
satisfactorily now. All that will be gain in some circuits. TI is a toroidal input transformer which is
achieved is a higher S meter reading and a An alternative to using a common-gate tuned to the operating frequency by
higher ambient noise level when no signals configuration was highlighted in Solid means of Cl. Rl and C2 form a shunt
are present. State Design for the Radio Amateur, feedback-network which aids stability by
where Hayward and DeMaw gave design lowering the gain. R2 provides additional
Circuit Description data for fed-back bipolar-transistor rf stability by introducing degenerative feed-
A major problem experienced by some amplifiers. The same principles apply to back. R2 is not bypassed for rf.
builders who use common-source FETs in FET amplifiers, where gain is traded for The drain circuit contains a second
rf amplifiers is instability. Despite careful bandwidth and stability. tuned transformer, T2. R4 and C5 form a
layout and input-output isolation, in- A design objective with any amplifier decoupling network in the 12-volt supply
stability seems to occur_A quick solution should be unconditional stability - no line. This helps prevent unwanted signal
might be to change the circuit to a self-oscillations at any frequency, energy from entering the preamplifier via
the power-supply leads.
Construction Data
RF AMP. A printed-circuit layout is given to scale
RI C2 in Fig. 79. Pc-board material (double
sided) is 1.Jsed as a shield enclosure for the
2.7. 01 0.001 preamplifier. The etched-circuit board is
MPFI02 also double sided, with the copper on the
28 - 29.7 MHz
T2 component side serving as a ground plane.
J2
This aids stability. The outer dimensions
of the box are (HWD) 1-112 X 1-3/4 X
~OUTPUT
3-15/16 inches (38 X 44 X 100 mm). A 3
JI
~ rl-, 50.1\. X 4-112 inch (76 X 115 mm), pc-board
ANT.~
base is used as the bottom cover, and a
rl,
50Jl. ~ U-shaped, press-fit aluminum cover serves
C4
0.01 R4 C5 as a top shield for the assembly.
100 rt0.OI 01 The pc board is soldered in place inside
EXCEPT AS INDICATED, DECIMAL VALUES OF
CAPACITANCE ARE IN MICROFARADS (JlF) ;
OTHERS ARE IN PICOFARAD.S ( pF OR JlJlF);
RESISTANCES ARE IN OHMS;
k.l000. M'IOOO 000.
~
C6
0.001 q,.R the box (Fig. 64) after the circuit has been
built and tested. A single-hole mount
phono jack is used at each end of the box
to provide input and output connections
for the 50-ohm coaxial cables of the
+12V
system. A O.OOI-J.lF feedthrough capacitor
is mounted at one .end of the box to allow
Fig. 77 - Schematic diagram of the l()..meter preamplifier. Data for 15-meter operation can be found In routing of the 12-volt supply to the
the text. Fixed-value capacitors are disc ceramic. Resistors are 1/4-walt composition types. Numbered preamplifier .
components not appearing in the parts list are so identified for pc layout purposes only.
Cl, C3 - Mica compression trimmer, 100 pF Tl, T2 - 0.6 ",H with Hurn link. Wind 12 turns Performance
max. Elmenco 423 suitable. of no. 24 enameled wire on Amidon or
Jl, J2 - Coaxial connector of builder's choice. Palomar T50-6 powdered-iron toroid core.
A 3-dB bandwidth of 1 MHz is
al - Motorola JFET, MPF102 or equlv. (see Spread turns evenly around core and cement characteristic of this preamplifier, The
text). in place. measured gain is 15 dB. Stability is
8~49 Chapter 8
about 80 dB. This is sufficient to suppress
interference since the rf signal-to-noise
ratio very rarely exceeds 60 dB. Practical
experiments have shown this noise
To J blanker to be superior to other configua-
tions. The famous "woodpecker" com-
pletely disappears when it is switched in .
. The circuit layout should not be too
critical. However, some care is required to
build the switching gate without leakage.
Good balance is required. Slightly better
performance can be expected using HP
3081 PIN diodes, but this is achieved at
considerable higher expense.
Fig. 79 - Scale pattern and layout for the pc board. A Space-Perception CW Filter
The circuit described here was originally
I
conceived as an experiment to test the
excellent under all conditions, including Hahn-Meitner Institute, Berlin, West Ger- ability of the brain/ear subsystem to dif-
an open-loop situation (no termination at many.'" Martin's circuit is very involved ferentiate between tones of different
either end of the circuit). The noise figure and expensive. The simpler version pitch. An example of where this would occur
is under 2 dB at 30 MHz. A slightly better published here can be added to any hf is in a "pileup" situation where many sta-
noise figure and increased dynamic range receiver with a first i-f between 9 and 70 tions are transmitting at the same time on
might be possible with a Siliconix U310 MHz. It is assumed that the receiver has or about the same frequency. When these
JFET. However, at 10 and 15 meters no rf preamplifiers and that the amplifier signals are of the same relative amplitude,
either the MPF102 or 2N5484 are entirely following the mixer has a low enough it is difficult to separate them. It was
suitable. A 2N4416 FET would be an noise figure to make such a preamplifier thought that if the brain/ear system was
acceptable substitute. unnecessary. This circuit originally ap- given more information than just frequen-
Operation on 15 meters can be had by peared in an article by Rohde in the May cy difference between tones, that separa-
adding two turns of wire to teh main wind- 1980 issue of QST. tion of the individual signals would be
ings of TI and T2. No ohter changes are The noise blanker uses a Siemens enhanced. Initial tests with several users
necessary. The circuit constants given in TCA440 IC that incorporates all the of this filter indicate that this is likely the
Fig. 77 are for lO-meter operation. A elements of a single-conversion receiver. case.
parts kit, pc board or pc negative are The i-f chosen is about 2 MHz and the The filter detailed here is actually com-
available from Circuit Board Specialists, values of the input coils are selected for an prised of two filters - a high-pass and a
P.O. Box 969, Pueblo, CO 81002. This input frequency of 9 MHz. [Editor's low-pass. The' output from the low-pass
circuit is suitable for use on 20, 40 and 80 Note: This circuit is provided for tutorial filter is fed to the left ear and the infonna-
meters by merely changing the TI/Cl and purposes. Further design information on tion from the high-pass filter is applied to
T2/C3 circuits for the desired frequency. the TCA440 is contained in "Designed the right ear. Stereo headphones are re-
Examples of Scmiconductor Circuits," quired. The frequency response is tailored
A High-Dynamic-Range Noise Blanker Siemans Corp., Issue 1975/76. Copies of as shown in Fig. 82. As can be seen, the
The two most violent sources of pulse the material relevant to the TCA440 are roll-off is such that a,l 700 Hz (the center
interference are discharges during a available from Siem!.!ns Corp., IC Com- of the pass-band) each tone is down 6 dB.
lightning storm and noise generated by ponent Group, 186 Wood Ave. South, It is assumed that the receiver i-f band-
jamming stations and pulse radar stations. Iselin, NJ @8830.j width is limited to approximately 400 Hz
A nuisance called the "woodpecker" is a The 9-MHz signal is taken from the for the filter described.
several-megawatt over-the-horizon pulse mixer of the receiver and fed to a CP643 With this filter in the receive line the
radar system that apparently has its origin amplifier and a BF2;::16C source follower. user will notice that signals move from
in the U.S.S.R. This system produces The source follower drives a series tuned one ear to the other when tuning across
pulses up to several hundred microvolts at circuit. The signal is then applied to the the signals. Assume that the receiver pro-
the receiver input and interferes with com- TCA440 single-conversion receiver and duces a low-pitched note, increasing in
munications. LORAN is the bane of converted to the 2-MHz i-f. An external audio frequency as the receiver is tuned
amateurs using the 160-meter band. germanium diode provides fixed age through the signal. As the signal is first
Naturally occurring noise discharges such voltage to pin 9 of the TCA440. An audio heard (at a 500-Hz tone) it will appear
. as lightning add to the man-made noise test output is available to monitor the age predominantly in the left earpiece. As the
sources to make a noise blanker a necessi- action of this receiver section. The 2-MHz signal is tuned and the pitch rises, more of
ty in modern communications receivers. i-f output is taken from a BF246 source the signal will be heard in the right ear.
In general, the rise and decay times of follower and drives a BC 177 with an When the 700-Hz note is approached, the
man-made and naturally occurring noise adjustable-trigger threshold. The 74LS173 tones will be equal in each ear. If the
pulses are substantially faster than the ris~ IC has the proper rise and decay times to signal is tuned farther in the same direc-
and decay times of desired signals. This drive the four-diode switching gate via a tion, the higher-pitched tone will move
phenomenon can be used to differentiate 2N2219 driver. across to the- right ear.
between the two types of interference. It is It was determined that the intercept When two different tones are applied to
therefore desirable to build a pulse point of this arrangement is about 26 dBm the filter simultaneously (as when two
receiver that can become part of the ex~ and the switching gate has a depth of signals close in frequency are received),
isting receiver system without degrading the filter will split the signals, placing the
the overall receiver performance. lower-pitched tone more in the left ear
'Martin, "Moderner Stocraustaster mit hoher luter-
The noise blanker example of Fig. 80 is modulationsfestigkeit," cq-DL magazine, July and the higher-pitched tone in the right
based on a publication by M. Martin, 1978, p. 300. . ear. By doing this, the signals are
'Martin, "Grosssignalfester Stocraustaster fuer
Kurzwellen-und UKW-Empfaenger mit grossem separated not only in tone, but also in
'Circuit Board Specialists, Box 969, Pueblo CQ·81002. Dynamikbereich," UKW-Berishte, Feb. 1979, p.74, relative position. With a little practice,
.---------------------------~--------~------------------------------------------_.--O+12roI5V
Ul
'6 14 TCA440
470
:h 0 l
•
10k
tOk
470
,l0" 10k
THRESHOLD
r----------------- ----
I
---------1 2000
--------1-------------------
1
TEST AUDIO
I
I
I
2000 ~
4.7k
1000
tOO 2N2219
tk
50n
FROM
I
RECEIVER I
MIXER I
I
I
I
I
I
I
I
I
I I
L __________________________________ .1_ -,..r., - - - - - - - - - - -- - - -.
8·51 Chapter 8
,
components may be used. No doubt, the the receiver so that a 700-Hz note is ap- from adjacent commercial or amateur sta-
unit could be made smaller than the one plied to the filter. With the filter in the tions.· A simple cure for this problem is to
shown in the photographs. The rotary Ip/hp (low-pass/high pass) mode, adjust install between the antenna and receiver a
switch is mounted to the left of the board, controls RI and R2 so that the tone is of filter that will sufficiently attenuate the
and hookup wire is used to make connec- equal volume in each ear. It' should be out-of-band signals but pass those signals
tions throughout the unit. Other types of noted that the gains of the driver of interest with little or no attenuation. If
connectors may be used for the input, out- amplifiers should not be adjusted too the receiver is designed for reception of
put and power connections. The ones shown high, as there might be some hum pickup frequencies below and above the broad-
here were chosen for use in the writer's sta- in the low-pass section. Experimentation cast band, a 550-, to 1600 kHz band-stop
tion. with the settings of RI and R2 should pro- filter will be required. However, if recep-
Adjustment of the filter involves setting vide an optimum gain/noise settin~. tion is desired only below or above the
the level contro\s for each of the two broadcast band, then a less complex low
channels. This is accomplished by first in- A BC- Band Energy Rejection Filt.er or high-pass filter will suffice. Because a
serting a pair of stereo headphones at J2 Inadequate front-end selectivity, of majority of ham receivers are used for
and making a connection between the out- bipolar-transistor if amplifier and mixer reception above 1600 kHz, a high-pass
put of the station receiver and J I. Apply a stages which perform poorly, can result in filter will generally be preferable to the
regulated 10 to 15 volts dc at 13. Adjust unwanted cross-talk and overloading band-reject filter. For the same number of
Fig. 81 - Photograph of the space-perception cw filter. Press-on lettering was used to label the
control and jack functions.
9V +12
4.7k TO 15 V 583Hz 840 HZ
700 Hz
I
500 Hz I 900 Hz
-------, I
I I
I
I
I
----1.---
I
I
I
I
-6 dB
Fig. 82 - Relative frequency response of the two filter sections. The left ear receives information
from the low-pass filter and the right ear receives signals from the high-pass filter.
0.05
.}
1k
8
15k
Vee
;L
Vee LPF
10k
AUOIO
INPUT JI 10k
Vee
Ik
7
0.Q1 0.01
0.01
HPF
'I S.9k
Fig. 83-_Schematic diagram of the cw filter. All resistors are 114-watt, 5-percent types. All polarized capacitors are electrolytic. Others are miniature
ceramic types. J3 - Miniature power-supply connector.
J1 - Phono connector. Sl - Rotary, 4 pole, 4 position. U1, U2 - TL084C BiFET quad op-amp, Radio
J2 - Stereo headphone connector. T1, T2 - Primary 1000 ohm, secondary 8 ohm. Shack 276-1714 or equlv.
8·53 Chapter 8
tenuatioh rise, a filter oflO elements was ficulty in obtaining the desired response. 100 MHz, the insertion loss of the filter
considered necessary. Finally, to simplify If the attenuation peaks (f2, f4 and f6) at gradually increased to 2 dB. The
construction, only those designs permit- Q.677, 1.293 and 1.111 MHz are not ob- measured input impedance versus fre-
ting the use of standard-value capacitors tained, a slight squeezing or separating of quency was in good agreement with the
were considered. the toroidal-inductor winding should be calculated input impedance between 1.7
all that is required to tune the series- amd 4.2 MHz. (The frequency range
Building th~ Filter resonant' circuits. Note that series circuit above 4.2 MHz was not tested.) Over the
The filter layout, schematic diagram C6-L6 should resonate at f6 = 1.111 range tested, the input impedance of the
and response curve, the component values MHz, but from the response curve' it ac- filter remained within the 37 to 67.7 ohms
used, and the toroidal-inductor winding tually resonated at about 1.130 MHz. This input'-impedance window (equivalent to a
specifications are all shown in Fig. 88. The frequency error of about 2 percent is small maximum VSWR of 1.353).
design paramete(s and the calculated filter enough to ignore. The As value was Construction of the filter is relatively
components values and other calculated selected to be 58.3 dB, and examination of simple, as shown in, the photograph, and
parameters are shown in the upper right- the reponse curve shows the measured no difficulty should be experienced if the
hand corner of Fig. 88. The standard- filter response to be in good agreement. Mallory SXM polystyrene capacitors are
value capacitors used are listed under the The measured values of cutoff frequency used. These capacitors have a standard
filter schematic diagram. Note that all (at the attenuation level of 0.0988 dB) and tolerance of 2.5 percent and are available
standard values are within 2.8 percent of the measured value of fAs (the frequency through all Mallory distributors. The
the design values. Since the maximum where As is first reached) are also in good Micro-metals iron powder'T50-2 toroidal
deviation between the actual capacitance agreement with the calculated values. The cores are available through either Amidon
used and the design value will be only 5.3 measured pass-band loss was less than 0.8 or Palomar Engineers. This material
percent, there should be little or no dif- dB from 1.8 to 10 MHz. Between 10 and originally appeared in a QST article by
Wetherhold, W3NQN, in February 1978.,
J2
from a length of cable and connected in
parallel with the antenna at the receiver,
as shown in Fig. 89. It is desired that the
notch be at 4.24 MHz and a quarter-
wavelength section of cable open at the
far end will be used. The velocity factor of
the cable must be included in the calcula-
tions. As shown in the diagram, the cable
length is calculated from the standard for-
mula for wavelength.
One must be careful when using notch
filters such as these because there will be
additional notches 'across the frequency
spectrum. For an open quarter
wavelength of line there is an additional
notch at each odd multiple of that fre-
quency. The spectrum analyzer
photograph shown in Fig. 90 is an exam-
ple. The horizontal axis is lO-dB per divi-
sion. This display is of a 38-foot, 3-inch
(11.6 m) piece of RG-58A/U solid dielec-
tric coaxial cable open at the far end. As
can be seen, the first notch appears at
roughly 4.24 MHz and repeats at each odd
harmonic of that frequency. The notches
get smaller in depth as the frequency in-
creases. This is primarily because of the
RECEIVER/TRANSCEIVER _CABLETQ
ANTENNA
r ®
I~ ®'"
_"T"
Y
CONNECTOR
• ANTENNA
TERMINAL
>'<4 AT lNTERFERING
FREQUENCY
LINE LENGTH. 9B04 (VF) (;:IS'J"OF CABLE FOR
4t(HZ)
A VF OF 0.661
Fig. 86 - Full-scale etching pattern for the pc board used in the cw filter. Black areas indicate
VF. VELOCITY FACTOR OF CABLE
unetched copper.
OPEN AT THIS END ~•
increased loss of the cable at higher fre- quency has been shifted to twice the fre-
quencies. quency as that shown in Fig_ 90. Now, ad-
Fig. 89 - A quarter-wavelength of coaxial
The display in Fig. 91 is for the same ditional notches occur at every even multi- cable is used as a notch filter. The length of
piece of cable, only this time it is shorted ple of the quarter-wavelength frequency. line can be calculated using the formula that
at the far end. The fundamental notch fre- All notch frequencies for filters such as is given.
8·55 Chapter 8 -
fZ f4
70 ~~----.-----r--r--'--T-··--'-
I~
CD C5 = 1322 PF. L6 = 4.82 /lH
~ 40 - - - - \ + - . - - - - - t - - - - - - t -.. - . - - - - ... " - - - - 1
~. 88 - Filter·response curve, insertion loss, layout and schematic diagram. Terminal impedance is 50 ohms for this 1.7·MHz, high·pass filter.
CLOSED
LOOP
.~
___.)~----~A~G_.~"~A~/~U______~(~.) OPEN ~~~------RG---5-8-A-/U----~(-~~
MEASURED MEASURED SHORT
END HID
AdeqUate recelVlng capability is es- though simple to build and economical to come increasingly critical with respect
sential in vhf and uhf communications, use, is inherently lacking in selectivity. to signal-to-noise ratio as the operat-
whether the station is a transceiver or a With this general information in mind, ing frequency rises.
combination of separate transmitting and this section will cover vhf and uhf receiver
receiving units, and regardless of the "front end," stage by stage. Noise. Temperature, Noise Factor and
modulation system used. Transceive,rs and Noise Figure
fm receivers are treated separately in this RF Amplifiers Noise, in the context of this discussion,
Handbook, but their performance involves Signal-to-Noise Ratio (SjN): The limit- is produced by the movement of electrons
basic principles that apply to all receivers ing factor in the reception of signals is in any substance (such as resistors, tran-
for frequencies above 30 MHz. Important noise. Noise can be classified into two sistors and FETs) that has a temperatur.
attributes are good signal-to-noise ratio broad forms, random and nonrandom. above absolute zero ( - 273 0 C or 0 0 K).
(low noise figure), adequate gain, sta-' Nonrandom noise such as interfering Electrons move in a random fashion col-
bility, and freedom from overloading and signals are reduced or eliminated through liding with relatively immobile ions that
other spurious responses. techniques aimed at directly filtering or make up the bulk of the material. The
Except where a transceiver is used, the otherwise suppressing detection of the final effect is that, in most substances
vhf station often has an hf-band com- unwanted signals. This is only possible there is no net current in any particular
munications receiver for lower bands, because the nonrandom noises are dis- direction on a long-term average, but-
with a crystal-controlled converter for the crete in nature and are relatively pre- rather a series of random pulses. These
vhf band in question ahead of it. The dictable. pulses produce what is called thermal
receiver serves as a tunable i-f system, Random noise is generated by sources agitation noise, thermal noise or Johnson
complete with detector,. noise limiter, both internal and external to the receiver. noise.
BFO and audio aJ;Ilplifier. Unless one The external noise problem varies con- As the currents caused by electron
enjoys work with communications re- siderably with frequency of reception. movement increase with temperature, so
ceivers, there may be little point in build- Below about 25 MHz, man-made, atmos- does the noise power. Also, as the pulses
ing this part of the station. Thus our con- pheric, and galactic noise picked up by the are random, they spread out over a broad
cern here will be mainly with converter de- antenna is usually far· greater than the frequency spectrum. As it turns out, if we
sign and construction. noise generated within the receiver. In a examine the power contained in a given
Choice of a suitable communications majority of cases, noise output from a passband, the value of that power is in-
receiver for use with converters should not receiver tuned to the hf range drops dependent of the center frequency of the
be made lightly, however. Several degrees dramatically when the antenna is discon- passband. This is expressed as
of selectivity are desirable: 500 Hz or less nected from the receiver. When this is'
for cw, 2 to 3 kHz for ssb, 4 to 8 kHz for observed, it clearly demonstrates that p =kTB
a-m phone and 12 to 36 kHz for fm phone reception is limited by external noise. where p is the thermal noise power, k is
are useful. The special requirements of fm Since the ionosphere is less active at 50 Boltzmann's constant (1.374 X 10-23
phone are discussed in chapter 13. Good MHz, atmospheric noise is of less concern joule per °K),T is absolute temperature in
mechanical design and frequency stability than at hf. Even in a quiet location, OK and B is the bandwidth in hertz.
are important. Image rejection should be however, external noise usually overrides Notice that the power is directly propor-
high in the range tuned for the converter receiver noise in a well-designed system. tional to tempe'rature, and at 0 0 K the
output. This may rule out 28 MHz with Above 100 MHz, external noise other noise power is zero.
receivers of the single-conversion type than man-made is rarely a problem in Active devices normally exl;1ibit noise
having 455-kHz i-f systems. weak-signal work. The noise charac- temperatures different from their ambient
Broadband receiving gear of the sur- teristics of a receiving system become temperatures. The thermal noise pro-
plus variety is a poor investment at any important as they are the primary limita- duced by a semiconductor device will limit
price, u,nlessone is interested only in local tion in weak-signal work. Unfortunately, its ability to respond to input signals
work. The superregenerative receiver, circuit design and component choice be- below the level of the internally generated
9·1 Chapter 9
noise. Noise temperature,noise factor Brief analysis of this equation shows that
and noise figure' are all measures of this 3000
the first stage of a receiving system is the
device noise. The results are expressed in most important with regard to noise figure.
terms of temperature, ratios and decibels, 2000 1/ If the gain of this and succeeding stages is
respectively. •
V greater than unity, the denominator of
Consider a 50-ohm termination con-
V
'000 each successive term becomes greater. The
'00
nected to the input of a device with the 100
100
numerical value of terms beyond the
termination cooled to absolute zero. 800
/ second or third approaches zero and can
800
There would be no noise produced by this 400 NF (dB)_oK be ignored. .
source, and the noise output from the 300 1/ 0,1 - 6,75 It might seem that the more gain an rf
device would be that of the internally amplifier has, the better the signal-
generated noise. If the termination were 200 I 0,2
0,3
0.4 -
13,67
20,74
27,98 to-noise ratio and therefore the better the
0,5 35,39 reception. This is not necessarily true. The
now heated to a temperature that would
raise the output noise of the device by 3 l '00
90
/ 0,6
0,7
0.8
_
t=
42,96
50.72
58,66
primary function of an rf amplifier is to
dB (thermal agitation noise equal to the '"
IE 80 0,9 i-- 66,78 establish the noise figure of the system.
internally generated noise of the device) .
::>
I-
10
60 1,0
1,1
i-- 75,09 One good rf stage is usually adequate
and the temperature of the termination ..'"
IE 80
40
1,2
1,3
I-- 83,59
I-- 92,29
101,20
unless the mixer is a passive type with loss
measured, the effective input noise ''"" 1.4 I-- 110.31 instead of gain. Two rf stages are the usual
temperature (T E) of the device would be
this value. The noise temperature
.
I-
'"
(;
z
30
20
1.5
1.6
1,7
i-- 119.64
I-- 129,18
138,94
maximum requirement.
Once the system noise figure is es-
1,8 i-- 148,93
specification is independent of bandwidth 1,9 159.16 tablished, any further gain necessary to
2,0 169:62
and is directly proportional to noise '0 bring a signal to audible levels may be ob-
1234&678810
power. For example, if we were to halve tained from intermediate-frequency stages
NOISE FIQURE (dB)
the noise temperature we would double or in the 9udio channel. Use of the
the signal-to-noise ratio. minimum gam necessary to set the overall
In order to convert a noise tempe~ature Teceiver noise figure is desirable in order
Fig. 1 - Relationship between noise figure
measurement to noise figure an in- and noise temperature. to avoid overloading and spurious signals
termediate calculation is required - noise in subsequent stages.
factor (f). Noise factor is by definition the Further examination of the equation
ratio of the total output noise power to temperature, the equation can be re- points out the desirability of mounting the
the input noise power when the termina~ arranged as follows: first stage of the receiver system at the
tion is at the standard temperature of antenna. The transmission line from the
290° K (17° C). The noise power caused TE = 290 [antilog(NF110) - I J° K antenna to the receiver can be considered
only by the input noise of the termination
is simply the noise power of the source
multiplied by the gain of the device.
Mathematically
~
Typical Circuits INPUT
'''''it] ~
RF Selectivity
Ever-increasing occupancy of the radio
spectrum brings with it a parade of
receiver overload and spurious responses.
Overloading problems can be minimized
by the use of high dynamic range receiving
techniques, but spurious responses such as
the receive image must be filtered out
before mixing occurs. Conventional tuned
circuits cannot provide the selectivity
necessary to eliminate the plethora of
signals found in most ham neighborhoods.
Other filtering techniques must be used.
Fig. 4 - Cascode amplifier circuit combines grounded·source and grounded·gate stages, for high Although some amateurs use quarter-
gain and low noise figure. Though JFETs are shown, the cascode principle is useful with wavelength coaxial cavities on 50, 144 and
MOSFETs as well. 220 MHz, the helical resonators. shown in
Fig. 6 are usually a better choice as they
are smaller and easier to build. In the
frequency range from 30 to 100 MHz,
where it's difficult to build high-Q
achieve stability and best noise figure. An common-gate amplifier, shown in Fig. 3. inductors, and because coaxial cavities are
example of this approach is shown in Fig. The feedback reduces the stage gain and very large, the helical resonator is an
2A. Neutralization may be required; Fig. lowers the input impedance, increasing excellent choice. At 50 MHz, for example,
2C shows capacitive neufralization ap- the bandwidth,of the stage. An additional a capacitance-tuned, quarter-wavelength
plied to dual-gate MOSFET amplifier. benefit of common-gate amplifiers is coaxial cavity with an unloaded Q of 3000
Common-source JFET amplifiers usually reduced susceptibility to overload as com- would be about 4 inches (100 mm) in
require neutralization to attain satisfac- pared to common-source amplifiers. diameter and nearly 5 feet (1.5 m)long. On
tory operation. Inductive neutralization , The cascode circuit of Fig. 4 combines the other hand, a helical resonator with
as shown in Fig. 2B is commonly used. the common-source and the common-gate the same unloaded Q is about 8.5 inches
Using the gate as the common stage ele- amplifiers, securing some of the advan- (216 mm) in diameter and 11.3 inches (287
ment introduces negative feedback and tages of each. Increased gain over a single mm) long. Even at 432 MHz, where
eliminates the need for neutralization in a stage fs its greatest asset. coaxial cavities are common, the use of a
(A)
Fig. 6 - Round and square helical resonators,
showing principal dimensions. Diameter, D (or
side, S) is determined by the desired unloaded
Q. Other dimensions are expressed in terms of
OUTPUT
D (or S) as described in the text.
'0
00,
QO,
,0 "
'0 40 20
'0 '0 20 '000
20 '0 0, 900
40 100 '0 '0
'0 ,0 '00
l!)oO '0 40 '0
02
0,
,0
" ·'00
.0
.0
"
20
'0 '0
20
'0
.00
20 20
'0 ,0 ,0 '0
2000 '0
'0 "
20
40
'0
'0
'0
• '00
400
40 10
" '0 '00
20 40 '00
'0
.0 '0 '00
'000 70 200 - Z!)O
'0 '0 100 250
40 '00 200
4000 '0
.0 "0 U~:~TR_ 3
"0
.0 200
'000 "
'00 260
.,00
100
Fig. 7 - Design nomograph for round helical resonators. After selecting unloaded au, required shield diameter is Indicated by index line from au
scale to frequency scale (dashed index line shown here indicates a shield of about 3.8 inches (97 mm) for an unloaded a of 500 at 7 MHz). Number
of turns, N, winding pitch, P, and characteristic impedance, Zo, are determined by index line from the frequency scale through previously determined
shield diameter on rlght·hand side of the chart (Index line indicates P = 0.047 InCh, 1 mm, N = 70 turns, and Zo = 3600 ohms).
soldered carefully from one end to the Most helical resonators are tuned by selectivity of the circuit. However, to keep
other. means of a brass tuning 'screw or high dissipation loss to 0.5 dB or less (as is the
Best results are obtained when little or quality air-variable capacitor across the case for low-noise vhf receivers), the
no dielectric is used inside the shield. This open end of the helix. Piston capacitors unloaded Q must be at least 18 times the
is usually no problem at vhf and uhf , also work well, but the Q of the tuning loaded Q. Although this may be difficult
because the conductors are large enough capacitor should ideally be several times to achieve in practice, it points up the
that a supporting coil form is not the unloaded Q of the resonator. Varactor necessity of considering both selectivity
required. The lower end of the helix diodes have sometimes been used where and insertion loss before choosing the
should be soldered to the inside of the remote tuning is required, but varactors unloaded Q of any resonant tuned circuit.
shield at a point directly opposite from the can generate unwanted harmonics and
bottom of the coil. other spurious signals if they are excited CoupUng
Although the external field is mini- by strong, nearby signals. Signals may be coupled into and out of
mized by the' use of top and bottom When a helical resonator is to be tuned helical resonators with inductive loops at
covers, the top and bottom of the shield by a variable capacitor, the shield size is the bottQm of the helix, capacitive probes
may be left open with negligible effect on based on the chosen unloaded Q at the at the top of the helix, dii:ecl taps on the
frequency or unloaded Q. If covers are operating frequency. Then the number of coil, or any combination of these. Al-
provided, however, they should make turns, N and the winding pitch, P, are though the correct tap point can be cal-
good electrical contact with the shield. In based on resonance at 1.5f o. Tune the culated easily, coupling by loops and
those resonators where the helix is resonator to the desired operating fre- probes must be determined experimentally.
connected to the bottom cover, that cover quency, f o. When only one resonator is used, the
must be soldered solidly to the shield to input and output coupling is often
minimize losses. Insertion Loss provided by prob~s. For maximum isola-
The insertion loss (dissipation loss), IL, tion the probes are positioned on opposite
Tuning in dB, of all tuned resonant circuits is sides of the resonator.
A helical resonator designed from the given by When coupling loops are used, the
nomograph of Fig. 7, if carefully built, plane of the loop should be perpendicular
will resonate very close to the design fre"
quency. Resonance can be adjusted over a IL 20 log ( 1 _ = ~d/Qu
) dB
to the axis of the helix and separated a
small distance from the bottom of the coil.
small range by slightly compressing or ex- For resonators with only a few turns, the
panding the helix. If the helix is made where Qd = loaded 0 .plane of the loop can be tilted slightly so it
slightly longer than that called for in Fig. Q u = unloaded Q. is parallel with the slope of the adjacent
8, the resonator can be tuned by pruning conductor. Helical resonators with in-
the open end of the coil. However, neither This is plotted in Fig. 10. For the most ductive 'coupling (loops) will exhibit more
of these methods is recommended for practical cases (Qd > 5) this can be attenuation to signals above the resonant
wide frequency excursions because any closely approximated by IL == 9.0 (Q d/Q u) frequency (as compared to attenuation
major deviation in. helix length will dB. The selection of a loaded Q for a tuned below resonance) whereas resonators with
degrate the unloaded Q of the resonator. circuit is dictated primarily by the required capacitive coupling (probes) exhibit more
~
while at the same time maintaining a
reasonable noise figure. MOSFET over-
0.005 load characteristics are suitable for the
vast majority of appljcations. Local-
,L oscillator energy can be applied at one of
osc. the MOSFET gates, effectively isolating
INJECTION
+12 V the local oscillator from the other signals.
The gate impedance is high, so relatively
little injection is needed for maximum
conversion gain. A typical example is
(C) shown in Fig. 12B.·
JFETs are close to the MOSFET in
Fig. 12 - Examples of single-ended mixers. The diode mixer, A, is usable through the microwave mixer performance but are more difficult
region. FET mixers, Band C, offer conversion gain and low noise figure. to apply in practical hardware. As with
the MOSFET, input impedance to a JFET
mixer is high, and substantial conversion
gain is available. JFET bias for mixer
passband_ Proper mixer operating con- these frequencies. For recelVlng ap- service is critical and must be adjusted for
ditions will alleviate IMD proble~s, and plications, amateurs typically want to best results. The output impedance of a
also reduce gain-compression problems. detect only one of the mixing products, JFET is lower than a dual-gate MOSFET;
A third type of distortion is cross usually the first order mixing product. typically around 10 kO.Although other
modulaiion. This is most readily observed Filtering must be applied to separate the possibilities exist, local-oscillator injection
on a-m signals. When the carrier is on, desired signal from the rest. Post-mixer should be made at the JFET source for
cross modulation is evidenced by modula- filtering is not adequate, as input images best results. The source is a low-
tion characteristics of another signaf can be mixed to the same intermediate impedance point, so considerably more
being superimposed on the received car- frequency as the desired signal. Input local-oscillator power is required than if a
rier. Techniques to improve IMD charac- filtering discriminates against these images dual-gate MOSFET were used as mixer.
teristics also improve cross modulation and prevents unwanted out-of-band sig- Noise figures as low as 4 dB are possible
performance. nals from possibly overloading the mixer. with circuits like that shown in Fig. 12C.
A problem inherent to all mixing The injection level of the local oscillator
systems is image generation. Whenever Single-ended Mixers affects mixer performance. Raising LO
two signals are mixed, components are level increases conversion gain in an FET
produced at the sum and difference of the Most mixers are single-ended. The mixer. The local-oscillator signal should
two signal frequencies, and at multiples of simplest type of mixer is the diode mixer. be as large as possible without pushing the
VHF and UHF Receiving Technique. 9-8
FET into its pinchoff region. The gate
RFC
junction of the FET should never conduct
in mixer' applications. Increased IMD
products result from either of the afore-
mentioned conditions and should be OSC.
carefully avoided. The local-oscillator INJECTION 1152 111Hz
energy should be as pure as possible.
Distorted injection energy not only in-
creases IMD production but also in- 3-dB
creases stage noise figure. HYBRID 1---0
Proper termination of the output of an COUPLER IF
RF OUTPUT
FET mixer optimizes overload per- INPUT
formance. If the impedance seen at the
drain of an FET mixer is too high at any of ' 1296 MHz
the mixer product frequencies, large
voltage excursions can occur on the FET
drain. If thevoltage excursion on the drain
is large enough, output distortion will be 01,02 HP5082-2717
evident. Often these high-voltage ex-
cursions occur at frequencies outside the
desired passband, causing distortion from
signals not even detectable by the receiver. 5000
A resistor within the output matching TO
network is often used to limit the 15k
broadband impedance to a suitably low
level. ~v BIAS-+:
Balanced Mixers ,
(Al
Use of more than one device in either a
singly or doubly balanced mixer offers
many advantages over a single-ended
01-04 HP 5082-2835
mixer. The balance prevents energy
injected into a mixer port from re- RF
INPUT •
appearing at another port. The implications osc.
1
of this are significant when minimum
r
mixer distortion is sought. The port-
to-port isolation inl).ibits any signals other
than the mixing products from reaching
any other stages further along in the
system where they might be mixed,
causing undesirable signals. The usually
large local-oscillator signal is kept away (B)
from the rf amplifier stages where it might
cause gain compression because of its
magnitUde. Any amplitude-modulated noise
foupd on the local-oscillator signal is
suppressed from the mixer output, where QIA U430
it might be later detected. In a singly RF
balanced mixer only one port, usually the INPUT
r
other two. A 'doubly balanced mixer
isolates all three ports from each other.
The most common balanced mixer uses
diodes. The disadvantages presented ear-
lier with respect to single-ended diode
mixers apply here also. A singly balanced
diode mixer is shown itlFig. 13A.
Hot-carrier diodes are normally used for
Dl and D2, as they can handle high
currents, have a low noise figure, and are
available for use up through several
gigahertz. ' ,
The doubly balanced mixer is more
(C)
common today. Commercial modules, Rl
very reasonably priced, are often used
instead of homemade circuits. Large-scale
manufacturing can usually offer electrical
balance 'not so easily attained with the
Fig. 13 - Balanced mixers for vhf and uhf. The singly balanced mixer, A, provides isolation of the
homemade units. Isolation of 35 to 40 dB local oscillator from the output. The doubly balanced diode mixer, B, has all ports isolated from
is typical at vhf, with only 6 to 7 dB of each other, and is broadband throughout vhf. A special dual JFET is used in C to give high
conversion loss. High local-oscillator dynamic range with low noise figure. ,
9·9 Chapter 9
injection is needed to reach optimum
performance with these mixers .. Proper +
broadband termination of all the mixer
ports is necessary to prevent unwanted
signals from being reflected back into the
mixer "rat 'race," only to emerge at
another port The i-f port (shown ill Fig.
13B) is the most critical with respect to
termination and should be terminated at
50 + jO ohms. Transmission line trans-
formers provide the necessary phase shift,
as half the bridge is fed 180 degrees out of
phase with respect to the other half. These ' - - - - - 0 "&-MHz OUTPUT
can be wound on ferrite toroid forms to
". ",±
effect a broadband response. Careful 10 k
l~o Ir-r--+---..-I---t
improves port-to-port isolation.
Active devices can be used very
effectively in balanced mixers. Both FETs
and bipolars can be used successfully.
Active balanced mixers offer all the
benefits of balanced diode mixers plus the 47
that band. Cl I
In Fig. 16 the voltage dropping resistor, INPUT I EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPACITANCE ARE
R4 and the Zener diode, Dl, may be of IN MICROFARADS I JlF I ; OTHERS
the value necessary to obtain 9 to 12 Y dc ARE IN PICOFARADS I pF OR JlJlF);
for operation of the unit. By increasing RESISTANCES ARE IN OHMS;
the resistance and dissipation rating of R4 k 01000, M-IOOOOOO
and DI, the preamplifier may be operated
from the 150- to 200-Y supply found in
many tube-type receivers.
The layout of the board is so simple Fig. 16 - Schematic diagram for the preamplifier. Part designations not listed below are for pc
that it is hardly worth the effort of making board placement purposes. Alternative input circuit for use with microwave diOde mixer is shown
a negative for the photo-etch process. A at B.
Cl, C4 - See Table 1. L 1, L2 - See Table 1.
Kepro resist-marking pen was used with C2, C3, C5, C6, C7, C9 - Disc ceramic .. R4 - 3 t~rns no. 28 enam. on ferrite bead. A
success on several boards. Another ap- C8 - 0.001 feedthrough capacitor. 220-ohm, 1/2-watt resistor may be sub-
proach ...... and one that is highly Jl, J2 - Coaxial connectors. Phono-type, BNC stituted.
recommen~d - is to cover the copper or SO-239 acceptable. ' RFC2 - 33 IIH, iron-core inductor. Millen
~300-33 or J. W. Miller 70F335Al.
with masking tape, transfer the pattern
with carbon paper, then cut away the tape
to expose the part to be etched. On small, T.ble1
simple boards the masking-tape method is 28 MHz 50 MHz 144 MHz
hard to beat. L1 17 turns no. 28 enam. on 12 turris no. 26 enam. on 5 turns no. 20 tinned
The pc board may be mounted in Amidon T-50-6 core. Tap at Amidon T-37-10 core. Tap at 1/2-lnch 10 x 1/2-inch
almost any small enclosure. Construction 6 turns from ground end. 5 turns from ground end. long. Tap at 2 turns from
ground end.
is nof tricky or difficult. It should take
L2 Same as L1, without tap. Same as L 1, without tap. 4 turns no. 20 tinned like
only a few minutes to complete the unit L1, without tap.
after the board is prepared. The board is C1, C4 15 to 6O-pF ceramic trimmer. 1.8- to 16.7-pF air variable. 1.5- to H.6-pF air variable.
fastened in the enclosure by means of one Erie 538-002F. E. F. Johnson 189·506-005. E. F. Johnson 189-504·005.
metal standoff post and a no. 4 screw and
'*11 Chapter'
'drain, at 12 to 15 volts d~. Touch the
TO +V SUPPL'V
THROUGH C8
neutralizing coil, L2. If there is any
change in current, tile stage is oscillating.
Keeping contact with L2 (to prevent
oscillation), readjust RI for 5 mA~ Using a
strong 220-MHz test signal, adjust C4 for
maximum signal indication. Set CI to
minimum capacitance, and peak C2.
Increase CI slowly until signal no longer
rises, then back off one turn and readjust
C2 and C4 for maximum signal.
Now reverse the preamplifier, con-
necting J I to the receiver input, and (Al
feeding the signal to J2. With the dc still
applied, tune L2 to minimum signal
Fig. 17 - Full·scale layout and parts place· feed-through. If L2 has an ungrounded
ment guide for the pc board. Foil side shown. brass slug, the amplifier attenuation
should be about 50 dB. Drain current
should remain at 5 mAo
Connect the amplifier normally, and
The dc voltage for the preamplifier repeat the process outlined above, until
is fed through one arm of a coaxial the tuning of C4 remains nearly constant.
T fitting at the receiver input. This Finally, adjust CI for best signal-to-noise
assumes use of some sort of blocking ratio (lowest noise figure) and readjust
capacitor in the receiver input, to prevent C2. This should yield a noise figure of 1.5 (8)
grounding the dc through a coupling loop to 2 dB, and gain of 12 to 18 dB,
or tap on a grounded tuned circuit. The depending on the transistor used. Often Fig. 19 - At A, 432·MHz GaAs FET preampli·
fier built by K2UYH. The transistor is mounted
rf choke in the preamplifier circuit, the lower-gain condition will also give the at the central shield by soldering the source
RFC I, and the one used at the receiver best noise figure. lead directly to the copper foil. The drain lead
input (to isolate the dc from the rf) are not of the transistor passes through a hole in the
critical. Any reasonably good vhf choke Low-Noise GaAs FET Preamplifiers for shield. At B, a 1296·MHz GaAs FET
432 and 1296 MHz preamplifier built by WA211F. In this model,
should do. If you're still willing to take the the transistor is connected to striplines etched
losses involved in the line, and you want Gallium-arsenide field-effect transistors on glass·epoxy board. SMA·type coaxial con·
to use the preamp at the receiver input, (GaAs FETs) have recently come into use nectors are Shown although type N or BNC
leave RFC lout of the circuit, and connect as low-noise microwave amplifiers. Ama- connectors may be used.
the dc as shown in Fig. 18. teur experimentation has shown that they
can provide excellent performance on the
Adjustment uhf and lower microwave amateur bands. ticularly the ones characterized as C-band
First set R I for about 5 mA current These devices are rather expensive, par- and X-band (4-12 GHz) microwave
low-noise amplifiers. However, other
GaAs FETs, characterized as power
amplifiers for low and medium-power (up
to 1/4 watt) microwave applications will
EXCEPT AS INDICATED, DECIMAL provide almost the same noise figure at
VALUES OF CAPACITANCE ARE
IN MICROFARADS ()IF) ; OTHERS
uhf and are being made available to
ARE IN PICOFARADS (pF OR ~)lF); amateurs. The power devices also have
RESISTANCES ARE I N OHMS;
wide dynamic range, providing less inter-
k -1000. M'I 000 000
modulation distortion and lower sus-
~~_~~~rr~--<l17---<r__--1C~EtT~~ER ceptibility to. burnout. The receiver pre-
amplifiers to be described are relatively
~
simple to construct and have sufficient
INPUT
\\L3 : tuning range for almost any GaAs FET
~RFC1
available. They were first described by
K2UYH and WA2ZZF in June 1978
QST.
~~-----------<~-O+
Construction
12V
These preamps for 432 MHz (Fig. 19A)
and 1296 MHz (Fig. 19B) use power
GaAs FETs made by Microwave Semi-
conductor Corp.; however, devices made
Fig. 18 - Circuit and parts information for the WB6NMT 220·MH~ preamplifier. by NEC (Nippon Electric Co.) perform at
C1 - 0.8J)r 1 to 10 pF glass trimmer, least as well, and many similar devices will
Johanson 2960 of JFD VAM or MVM series. 1. turn from top, subject to adjustment for also certainly work. Construction details
C2 - Like C1, or Corning Direct Traverse lowest nf. Air·wound coils also usable, but are shown in the photographs and sche-
CGW. 0.8 to 10 pF. toroids preferred.
C3 - 390·pF silver mica. L2 - 9 turns no. 28 enam. on 114·in. (6.3 mm) matic diagrams. The 432-MHz prea~p is
C4 - Like C1, C2 or less·expensive type with slug·tuned form (Miller 4500, brass slug). Do built in a 2-1/4 X 1-1/2 X I-inch (57 X 38
1 to 10·pF range. not ground the slug. X 25-mm) box made of double-sided
C5 - Experiment with values 1 to 5 pF, for L3 - Like L1, but no tap. printed-circuit board. A cover 'plate is
maximum gain in system as it will be used. Q1 - 2N5245, 2N5486, MPF·107, TIS·88.
J1, J2 - SMA or N·type connector. R1 - 200· or 250-ohm control. recommended but does not significantlx
L1 - 4 turns no. 22 enam. or Micrometals RFC1 - Vhf rf choke, 0.8 to 3"H. Use only affect tuning. The GaAs FET source is
T·37·0 toroidal core (Amidon Associates). Tap when preamp is antenna mounted (see text). soldered to the central shield board with
+flTWO
41-mm) Minibox (BUD CU-30ooA or u-v
equivalent). The GaAs FET is bolted. BATTERIES
~6
between two pieces of 1/16-inch (1.6-mm)
printed-circuit board, using O-SO screws
(available at many hobby shops). The lead
height is just right to sit on top of the
+1 ---- -vgI
9-13 Chapter 9
Table 2
Manufacturer Relcom Anzac MCL- MCL MCL MCL
Model M6F MD·108 SRA·1 SRA·1H RAY·1 MA·1
Frequency Range (MHz)
lO 2·500 5·500 5·500 5·500 5·500 1·2500
rf 2·500 5·500 51·500 5·500 5·500 1·2500
i·f DC·500 DC·500 DC·500 DC·500 DC·500 HOOO
Conversion loss 9 dB max. 7.5 dB max. 6.5 dB typo 6.5 dB typo 7.5 dB typo 8.0 dB typo
Mid·range
Isolation, lO·RF 34·40 dB min. 40 dB min. 45 dB typo 45 dB typo 40 dB typo 40 dB typo
Mid·range lO i·f· 25·35 dB min. 35 dB min. 40 dB typo 40 dB typo .40 dB typo 40 dB typo
Total input power
l{) power requirement:
50mW
+7 dBm (5 mW)
400mW
+7 dBm (5 mW)
500mW
+7 dBm (5 mW)
. 500mW
+ 17 dBm (50 mW)
1W
+ 23 dBm (200 mW)
50mW
+ 10 dBm (10 mW)
Signal2·dB
compression level Not spec. Not spec. + 1 dBm + 10 dBm + 15 dBm +7dBm
Impedance, all ports 50 ohms 50 ohms 50 ohms 50 ohms 50 ohms 50 ohms
Relcom, Division of Watkins·Johnson, 3333 Hillview Ave., Palo Alto, CA 94304.
All specifications apply only at stated lp power level. Anzac Electronics, 39 Green Street, Waltham, MA 02154.
MCl - Mini·Circuits laboratory, 2625 East 14th St., Brooklyn, NY 11235.
, DRAIN
0.13713.4mm)
(A)
mixer-device current is generally main-
tained at a low level.' This can reduce
dynamic range, increasing overload po-
tential, as defined in the terminology I
I
I
-I
I
I
I
I
L..
I _________ _ I
appendix. Gain contributions of rf ampli-
fiers (used to establish a low system noise
0.I2e TYP.
~ATE
figure) further complicate the overload
_ _tL---I_3._'m_m_)_ L..J problem. LO-noise leakage to the rf and
10.e~·~~~-1 f-- i-f ports adversely affects system per- Fig. 23 - The i·f port of a double·balanced
formance. Mixer dynamic range can be . mixer is matched at flO - frf and reactive at
limited by conversion of this noise to i-f, flO ± frf. In this configuration conversion
0.420
r~~l
placing a lower limit on mixer system loss, rf compression and desensitization levels
sensitivity. Generally 20 dB of mixer .can vary ± 3dB while harmonic modulation and
third·order IMD products can vary ± 20 dB.
midband interport isolation is required,
and most passive DBM can offer greater
than 40 dB. requirement, as will be demonstrated
A commercially manufactured doubly later. Additional LO power is fairly easy
balanced diode mixer offers performance to generate, filter, and measure. If we
(B) predictability, circuit simplicity and flexi- accept the fact that more LO power is
bility. Closely matched Schottky-barrier necessary for the DBM than is used in
hot-carrier diodes, commonly used in conventional single-device active mixing
most inexpensive mixers of this type, circqits, we leave only two real obstacles
0.I2e TVP. provide outstanding strong-signal mixer to be overcome in the DBM, those of
13.Imm)
performance (up to about 0 dBm at the rf conversion loss and i-f output int~rfacing.
, tl input port) and add little (0.5 dB or so) to To minimize conversion loss in a DBM,
0.023.-.1 L the mixer noise figure. Essentially, diode the diodes are driven by the LO beyond
10.GOmm)l I conversion loss from rf to i-f, listed in their square-law region, producing an
Table 2 represents most of the mixer con- output spectrum which in general includes
Fig. 22 - Dimensional information for the
GaAs FET packages supplied by MSC. At A, tribution to system noise figure. Midband the terms 2 :
case style 98, top view, and at B, top view of isola.tion between the LO port and the rf 1) Fundamental frequenciesjLO andJrf
case style 97. Drain and source leads are and'i-f ports of a DBM is typically> 35 dB 2) All of their harmonics
spaced 0.065 in. (1.65 mm) above the bottom of - far greater than that achievable with 3) The desired i-f output, JLO ± ftf
the case. MSC designation for these case
styles is Flipac.
conventional single device active-mixing 4) All higher order products of njLO ±
schemes. This isolation is particularly ad- mJrf, where nand m are integers.
vantageous in dealing with low-level local- The DBM, by virtue of its symmetry
types? What does it offer, and what are its oscillator harmonic and noise content. Of and internal transformer balance, sup-
disadvantages? To answer these questions, course, selection of LO devices with low presses a large number of the harmonic
a look at more conventional "active" audio noise figures, and proper rf filtering modulation products. In the system
(voltages applied) mixing techniques and in the LO putput, will reduce problems described here, JLO is on the low side of
some of their problems is in order. The from this source. Jrf, therefore, numerically, the desired i-f
reader is referred to an article in QST ~ Often-listed disadvantages of a diode output isJrf - JLO. Nonetheless, the term
dealing with mixers. Briefly reiterated, DBM are (a) conversion loss, (b) LO JLO ± Jrf appears at the i-f-output port
common single-device active mixers with power requirements, and (c) i-f-interface equal in amplitude to the desired i-f
gain at vhf and uhf are beset with problems. The first two points are closely signal, and this unused energy must be
problems of noise, desensitization and interrelated. Conversion loss necessitates effectively terminated to obtain no more
small local-oscillator (LO) isolation from some low-noise r-f amplification to es- than the specified mixer-conversion loss.
'DeMaw and McCoy; 'Learning to Work With Semi- tablish a useful weak-signal system noise
conductors," Part IV. QST, July 1974. figure. Active mixers also have this 'See appendix.
VHF and UHF Receiving Technique•. 9-14
EKCEPT AS INDICATED, DECIMAL YAWES OF
CAPACITANCE ARE IN MICROFARADS I JlF I ;
OTHERS ARE IN PICOFARADS· I pF OR JlJIFI;
RESISTANCES ARE IN OHMS;
kalOOO, MalOOO 000.
I
AMP.
14 OR 28 MHz r-h! I
I
14 OR 28 MHz
I
C2 C3 I
I
rL X 0.001
)FT
0.001 :
)"Ff"1
i-~---------~- --~- -1
I
I
I
1000 001
f1-l
0,01
I
I
I
1·001
I +15V
r-hl
Fig. 24 - A schematic diagram for the double·balanced mixer and i·f post amplifier. The i·f can be either 14 or 28 MHz. Parts values are given in
Table 3.
This is not the image frequency,jLO -fi of, Also greatly affected are third-order directly at the mixer i-f terminal. This
which will be discussed later. inter-modulation-product ratio and the approach is easiest for the amateur to
In any mixer design, all rf port signal suppression of spurious signals, both of implement and duplicate, so a form of it
components must be bypassed effectively which may vary ± 10 dB or more. It is was tried - with success. In our circuit,
for best conversion efficiency (minimum ironic that the i-f port is the most sensitive Cl serves a dual purpose. Its reactance at
loss). Energy not "converted" by mixing to a reactive termination, as this is a ftf + JLO is small enough to provide a
action will reduce conversion gain in receiving system point where sharp- low-impedance "short-circuit" condition
active systems, and increase conversion skirted filters are often desired. to this term for proper mixer operation.
loss in passive systems such as the diode Briefly, here is what happens with a Additionally, it is part of the input
DBM. Rf bypassing also prevents spur- reactive i-f port termination. Fig. 23 reactance of the mixer i-f-amplifier inter-
ious resonances and other undesired shows a DBM with "high side" LO face. Fortunately the networkimpedance-
phenomena from affecting mixer per- injection and an i-f termination matched transformation ratio is large enough, and
formance. In this system, rf bypassing at atjLO - ftfbut reactive tojLO +frf. The in the proper direction, to permit a fairly
the i-f-output port will be provided by the latter term re-enters the mixer, again large amount of capacitance (low react-
input capacitance of the i-f interface. The combines with the LO and produces terms ance) at the mixer i-f-output port. The
DBM is not a panacea for mixing ills, and that exit at the rf port, namely 2jLO +frf, capacitor, in its dual role, must be of good
its effectiveness can be reduced drastically a dc term, and JLO + ftf - JLO (the . quality at vhf/uhf (specifically frf + JLO),
if aU ports are not properly terminated. original rf-port input frequency). This with short leads, to be effective. The mixer
condition affects conversion loss, as condition (frf + jLO)/frf - JLO) > 10 is
DBM Port TermltUltlons mentioned earlier, in addition to rf-port inet at 432 and 220 MHz with a
Most DBM-perforniance inconsisten- VSWR, depending on the phase of the 404/192-MHz LO (2g-MHz i-t) and on 14
cies occur because system source and load 'reflected signal. The term 2jLO + frf also MHz with a l3O-MHz LO (14-MHz i-t)o
impedances presented to the mixer are not affects the harmonic spectrum resulting in At 50 MHz, with a 36 MHz LO, we are
matched at all frequencies encountered in spurious responses. slightly shy of the requirement, but no
normal operation. The terminations (at- One solution to the i-f-interface prob- problems were encountered in an operat-
tenuator pads) used in conjunction with lem is the use of a broadband 50-ohm ing unit. The pi-type interface circuit
test equipment by manufacturers to resistive termination, like a pad, to assures a decreasing impedance as i-f
measure published performance char- minimize reflections. In deference to operation departs from midband, thereby
acteristics are indeed "broadband" match- increased post-conversion system noise lessening IMD problems.
ed. Reactive mixer terminations can cause figure, it seemed impractical to place such
system problems, and mUltiple reactive a termination at the mixer i-C output port. TheLO Port
terminations can usually compound these While a complimentary filter or diplexer The primary effect of a reactive LO
problems to the point where performance (high-pass/low-pass filters appropriately source is an increase in harmonic modula-
is very difficult to predict. Let's see how terminated) can be used to terminate both tion and third-order IMD products. If
we can deal with reactive terminations. frf + JLO and frf - JLO', a simpler the drive level is adequate, no effect is
method can be used if ftf + JLO is less noted on conversion loss, rf compression
The I-F Port than I GHz andfrf + (fLO)/(frf - JLO) > and desensitization levels. A reactive LO
The i-f port is very sensitive to 10. Place a short-circuit termination to ftf source can be mitigated by simply
mismatch . conditions: Reflections from + JLO, like a simple lumped capacitance, .padding the LO port with a 3· or 6·dB pad
the mixer/i-f amplifier interface (the pi and increasing the LO drive a like
'Presentation and calculation format of these terms is
network in Fig. 24) can cause the based on "low-side" LO injection. See the appen- amount. If excess LO power is not
conversion loss to vary as much as 6 dB. dix for explanation. available, matching the LO source to the .
mixer will iMprove performance. This the modular~onstruction approach per-.
method is a~ble for 'Single~frequeJK;y Table 3 mits good signal isolation and enables the'
LO applications, when appropriate test DIM l·F Ampllfler"'arts List mixer-amplifier/i:f system to be used at a
equipment is availabl~ to evaluate match- ,variety of rf and LO-input frequencies, as
ing results. For simplicity, a 3·db pad 14 MHz i-f output 28 MHz i-f output mentioned earlier.
Cl 470 pF JFD 471J 300 pF JFD 301J Most commonly available, inexpensive
was incorporated at the LO-input port as or equal. or equal.
an interface in both versions of the mixer. C2 390 pF SM not used DBM are not constructed to take ad-
Thus the LO port is presented with a C3 180 pF SM 51 pF SM vantage of LO powers much above + 10
reasonably broadband termination, and is C4 39 pF SM 18 pF SM dBm (10 mW), To do so requires
relatively insensitive to applied frequency, C5 56 pF SM 27 pF SM additional circuitry which could degrade
C6 300 pF SM 150 pF SM
as long as it is below about 500 MHz:"This other mixer characteristics, specifically
implies that freq\lencies other than ama- L1 9 turns no. 18 enam., 9 turns no. 24 enam., conversion loss and interport isolation.
teur assignments'may be covered - and close wound on a 3/8- close wound on a 1/4- The advantage of higher LO power is
inch (9.5 mm) diameter inch (6.3 mm)diameter
such is indeed the case when appropriate primarily one of improved strong-signal-
LO frequencies and rf amplifiers are used.
red-slug coil form.
L2 18turns no. 26enam.,
green-slug coil form.
12 turns no. 26 enam.,
. handling performance, At lea~t one
Remotely located LOs, when adjusted for manufacturer advertises a moderately
close wound on a 3/8- close wound on a 1/4-
a 50-ohm load, can be connected to the inch diameter red-slug inch diameter green-
priced "high-level" receiving DBM that
mixer without severe SWR and reflective- coil form. slug coil form. can use up to +23 dBm (200 mW) LO
loss problems in the transmission line. power, and still retain excellent con-
Tap down 7 turns from No tap used.
Broadband mixers exhibit different top for 3N 140 drain
version loss and isolation characteristics,
characteristics at different frequencies, connection. See text. shown in Table 1. The usefulness of mixers
due to circuit resonances and changes in with LO power requirements above the
diode impedances resulting from LO L3 Same as L2 but no tap. Same as L2, spaced 1 commonly available +7 dBm (5 mW)
spaced 1 1/8-i nch inch (25 mm) center-to-
power-level changes. Input impedances of (29 mm)center-to- center with L2. level in amateur receiving applications
the various ports are load dependent,.even center with L2. may be a bit moot, as succeeding stages in
though they are isolated from each other AI, 300 ohm 1/4W, carbon. 430 ohm, 1/4W, most amateur receivers will likely over-
physically, and by at least 35 dB R3 carbon. load before the DBM. Excessive over-
electrically. At higher frequencies, this A2 160hm, 1/4 W, carbon.' 11 ohm, 1/4 W, carbon. design is not necessary.
effect is more noticeable, since isolation In general, mixer selection is based on
tends to drop as frequency increases. For Ferrite beads can be replaced by a 10-ohm, the lowest practical LO level requirement
1/4 W carbon resistor at one end of the choke, that will meet the application, as it is more
this reason, it is important to maintain the ifd(lsired.
LO power at its appropriate level, once economical and results in the least LO
SM = Silver Mica.
other ports are matched. leakage within the system, As a first-order
approximation, LO power should be 10
The RF Port dB greater than the highest anticipated
A reactive rf source is not too image-response potential in the 84 to input-signa! level at the rf port. Mixers
with LO requirements of +7 dBm are
detrimental to system performance. This 88-MHz range. TV channel 6 wideband-
quite adequate for amateur receiving
is good, since the output impedance of fm audio will indeed appear at the
most amateur preamplifiers is seldom 50 i-f-output port near 28 MHz unless applications.
ohms resistive. A 3-dB pad is used at 'the rf appropriate rf-input filtering is used to
port in the 50- and 144-MHz mixer to 14 eliminate it. While octave-bandwidth Application Design Gulclelines
MHz. and.a 2-dB pad is used in the vhf/uhf "imageless mixer" techniques can While the material just presented only
220/432-MHz to 28 MHz, although they improve system noise performance by scratches the surface in terms of DBM
add directly to mixer noise figure. Rf about 3 dB (image noise reduction), and theory and utilization in amateur vhf/uhf
inputs between about 80 and 200 MHz are image signal rejection by 20 dB - and receiving systems, some practical solutions
practical in the 14-MHz i-f-output model, much greater with the use of a simple to the non-ideal mixer-port-termination
while the 28-MHz-output unit is most gating scheme - such a system is a bit problem have been offered. To achieve
useful from 175 to 500 MHz. Mixer -esoteric for our application. Double or best performance from most commercially
contribution to system noise figure will be multiple-conversion techniques can be manufactured broadband DBM ill ama-
almost completely overcome by a low- .' used, to advantage, but they further teur receiver service, the following guide-
noise rf amplifier with sufficient gain and complicate an otherwise simple system. lines are suggested:
adequate image rejection. Image noise and signal rejection ~ill 1) Choose i-f and LO frequencies, which
depend on the effectiveness of the filtering. will provide maximum freedom from
Image Response provided in the rf-amplifier chain. interference problems. Don't "guessti.
Any broadband mlXlng scheme will mate"; go through the numbers!
have a potential image-response problem. Mixer Selection 2) Provide a proper i-f-output termina-
In most amateur vhf/uhf receiver systems The mixer used in this system is a tion (most critical).
(as in these units) single-conversion. Relcom M6F, with specifications given in 3) Increase the LO-input power to
techniques are employed, with the LO Table 2. Suitable substitute units are also rf-input power ratio to a value that will
placed below the desired rf channel for presented. The M6F is designed for provide the required suppression of any
non-inverting down-conversion to i-f. printed-circuit applications (as are the in-band interfering products. The speci-
Conversion is related to both i-f and LO recommended substitutes), and the lead fied LO power (+ 7 dBm) will generally
frequencies and, because of the broad- pins are rather short. While mixers are accomplish this.
band nature of the DBM, input signals at available With connectors attached, they 4) Provide as good an LO match as
the rf image frequency (numerically flO are more expensive. The simple package is possible,
- fl-f in our case) will legitimately appear suggested as, aside from less expense, 5) Include adequate pre-mixer rf-image
inverted at the i-f-output port, unless improved interface between mixer and i-f filtering at the rf port.
proper filtering is used to reduce them at' amplifier is possible because of the short When the mixer ports are terminated
the mixer rf-input port. For example, a leads. The combining of mixer and i-f properly, performance usually in.excess of
144-MHz converter with a 28-MHz i-f amplifier in one converter package was published specifications will be achieved
output (l16-MHz LO) will have rf done for that reason. Along these lines, - and this is more than adequate for
VHF and UHF Receiving Techniques 9.18
Fig. 26 - The bottom view of the DBM/i-f
amplifier shows component and shielding Fig. 27 - This is a side view showing con-
layout. L1, the mixer-amplifier interface induc- struction details for the double-tuned i-f output
tance and associated components are in- circuit. The 3N140 drain lead passes through
dicated. C1, with its wide silver-strap leads, is the shield wall via a small Teflon press-fit
connected directly between the mixer i-f output bushing and is connected directly to L2. A dc-
pin and the copper-clad ground plane with input illolation compartment along with device
essentially zero lead length. Connection be- gate 2 biasing components (bias configuration
tween the mixer output pin and other com- modified slightly after photograph was tak~n),
ponents (L 1, C2 and the rf choke for doc return) can be seen to the left of the i-f output com-
is made by using excess lead from C1. The ponents. L2 and L3 are spaced 1-1/B-inch
43-ohm, 1/4-W resistor in the 3N140 gate 1 lead (29-mm) center-to-center in the 14-MHz model
is connected between the high-impedance end shown, and 1 inch (25 mm) apart in the 2B-MHz
of L1 and a space terminal on the coil form. unit.
The device gate no. 1 lead and resistor are
joined at this point. It is important that in-
put/output isolation of the 3N140 be maintain-
ed as it is operating at high gain. Mixer
Fig. ~ - This top view of the DBM/i·f packages other than the M6F may have dif-
ments were made on the DBM/i-f ampli-
amplifier shows the plastic mixer package plus ferent pin connections and require slightly dif- fier, using the test setup shown in Fig. 28,
rf/LO inputs and i·f output jacks clearly marked ferent input-circuit layout and shielding. from both tones of a two-equal-tone rf-
for cabling. The unit is mounted on the open Double-sided copper-clad board was used input test signal consisting of -10 dBm
face of a standard 6 x 4 x 2·inch (152 x 102 throughout.
x 51·mm) aluminum chassis. This shielding is each tone_ The tones were closely spaced
necessary to prevent the 3N140 from picking in the 144-MHz range, and converted to
up external signals in the 14-MHz region. 14 MHz LO, Close spacing was necessary
sociated inductance. Both amplifiers were to ensure third-order products would ap-
tuned independently of their respective pear essentially unattenuated within the
most amateur vhf/uhf receiver mixing mixers, and checked for noise figure as relatively narrow i-f-output passband. In
applications. well as gain. With each i-f amplifier operation, as simulated by these test
pretufied and connected to its mixer, conditions, equivalent output signal levels
The Comb/nedDBM/I-F Ampllfter signals were applied to the LO and at 13 would be strong enough to severely
A low-noise i-f amplifier (2 dB or less) rf-input ports. The pi-network inductance overload most amateur receivers. Perhaps
following the DBM helps ensure an in the i-f interface was adjusted carefully the early Collins 7SA series R390A and
acceptable system noise figure when the to see if performance had been altered. No those systems described by Sabin 4 and
mixer is preceeded by a low-noise rf change was noted. I-f gain is controlled by Hayward' would still be functioning well.
amplifier. A pi-network matching system the externally accessable potentiometer. A high-performance, small-signal, vhf/
used between the mixer i-f-output port Passband tuning adjustments in the drain uhf receiving amplifier optimized for
and gate I of the 3N140 transforms the circuit are best made with a sweep IMD reduction and useful noise figure is
nominal SO-ohm mixer-output impedance generator, but single-signal tuning tech- only as good as any succeeding receiving-
to a ISoo-ohm gate-input impedance (at niques will be adequate. While there system stage, in terms of overload. The
28 MHz) specifically. for best noise should be no difficulty with ,. the non- DBM/i-f-amplifier combination present-
performance. The network forms a narrow- gate-protected 3N 140, a 40673 may be ed significantly reduces common first-
band mixer /i-f-output circuit which serves substituted directly if desired.
two' other important functions: It helps
achieve the necessary isolation between DBM/I-F Ampllfter 'IMD Evaluation 'Sabin, "The Solid-State Receiver," QST, July 1970.
'Hayward, "A Competition-Grade CW Receiver"
rf-and i-f signal components, and serves as Classical laboratory IMD measure- QST, March and April 1974. '
a 3-pole filter, resulting in a monotonic
decrease in match impedances as the
operating i-f departs from mid-band. This
action aids in suppression of harmonic-
distortion products.
The combined DBM/i-f amplifier is
shown schematically in Fig. 24 and
pictorially in the photographs, In the
14-MHz model, the 3N140 drain is tapped
down on its associated inductance to
provide a lower impetiance for better HP 8518/85518
strong-signal-handling ability. The 3N140
produces about 19 dB gain across a
LPRECISION 1.dS
STEP ATTENUATORS
...J SPECTRUM
ANALYZER
9·17 Chapter 9
+110 level. of two input tones by 10 dB will
cause the desired output to increase by 10
+40 dB, but the third-order output will
increase by 30 dB. If the mixer exhibited
+30 no compression, there would be a point at
3RD-ORDER which the level of the desired output
+20
INTERCEPT .POIN~ would be equal to that of the third-order
product. This is called the third-order
MIXER intercept point and is the point where the
+ 10 COMPRESSION
REGION
desired-output slopes and third-order
o slopes intersect (Fig. 29).
Noise Figure
- 10
Noise figure is a relative measurement
-20
based on excess noise power available
from a termination (input resistor) at a
- 30
particular temperature (290 0 k). When
measuring the NF of a double balanced
... - 40
mixer with an automatic system, such as
::) the HP-342A, a correction may be
...
a..
::)
-110
necessary to make the meter reading
o consistent with the accepted definition of
....
I receiver noise figure .
-80
In a broadband DBM, the actual noise
-70
bandwidth consists of two i-f passbands,
. one on each side of the local-oscillator
frequency (fLO +fi-f and}LO - fi-f). This
- 70 - 60 -110 ~ 40 .;.. 30 - 20 -10 0 + 10 + 20 + lSO + 40 + 110 double sideband (dsb) i-f response in-
cludes the rf channel and its image. In
RF INPUT POWER (dam) general, only the rf channel is desired for
further amplification. The image con-
Fig. 29 - A third-order intercept point is determined biextrapolatlng the desired product curve tributes nothing but receiver and back-
beyond the mixer compression point and Intersecting with the third-order 1M-product curve. In this ground noise.
case LO power is '+ 7 dBm, conversion loss is 5 dB. When making an automatic noise-
figure measurement using a wideband
noise source, the excess noise is applied
mixer overload problems, leaving the signal that causes the small-signal con- through both sidebands in a broadband
station receiver as the potentially weak version loss to increase by 1 dB, i.e., DBM. Thus the instrument meter in-
link in the system. When properly reducing a weak received signal by I dB. dicates NF as based on both sidebands.
understood and employed, the broadband /ntermodulation Products This means that the noise in the rf and
DBM followed by a selective low-noise i-f image sidebands is combined in the mixer
Distortion products caused by multiple
amplifier can be a useful tool for the i-f-output port to give a double con-
rf signals and their harmonics mixing with
amateur vhf/uhf receiver experimenter. tribution (3 dB greater than under ssb
each other and the LO, producing new
Appendix output frequencies. conditions). For equal rf-sideband re-
sponses, which is a reasonable assumption,
Mixer Terminology and in the absence of preselectors, filters,
Mixer /ntermodulatlon Intercept Point
Irf - rf input frequency or other image rejection elements, the
}LO - local-oscillator input frequency Because mixers are nonlinear devices, automatic NF meter readings are 3 dB
fi-f ~ i-f output frequency all signals applied will genel"ate others. lower than the actual NF for DBM
When two signals (or tones), Fl and F2, measurements.
By convention, mixing signals and their are applied simultaneously to the rf-input
products are referred to the LO frequency The noise figure for receivers (and most
port, additional signals are generated and DBM) is generally specified with only one
for calculations. In the mixer system appear in the output as}LO ± (nF I + sideband for the useful signal. As men-
presented, Irf is always above }LO, so we mF2). These signals are most troublesome
will refer our signals to Irf, with the tioned in the text, most DBM diodes add
when n ± m is a low odd number, as the no more than 0.5 dB (in the form of NF)
exception of Fig. 23 which uses the }LO resulting product will lie close to the de-
reference. to conversion loss, which is generally
sired output. For n - 1 (or 2) and m - 2
measured under single-signal rf-input
Overload (or 1), the result is three (3), and is called the (ssb) conditons. Assuming DBM con-
A generic term covering most undesired two-tone/third-order intermodulation prod- version efficiency (or loss) to be within
operating phenomena associated with ucts. When FI and F2 are separated by I specifications, there is an excellent proba-
device non-linearity. MHz, the third-order products will lie I bility that t~e ssb NF is also satisfactory.
MHz above and below the desired Noise figure calculations in the text
Harmonic Modulation Products outputs. Intermodulation is generally were made using a graphical solution of
Output responses caused by harmonics specified under anticipated operating the well known noise-figure formula:
of}LO and Irf and their mixing products. conditions since performance varies over
RF Compression Level the broad mixer-frequency ranges. Inter- IT = fl +f2 -1
g]
The absolute single-signal rf input~ modulation products may be specified at
power level that causes conversion loss to levels required (i.e., 50 dB below the converted to dB.
increase by I ·dB. desired outputs for two O-dBm input
signals) or by the intercept point. Improved Wide &md I-F Responses
RF Desensitization Level The intercept point is a fictitious point The following information was de-
The rf input power of an interfering determined by the fact that an increase of veloped in achieving broad-band per-
~O~_5~____~__Jlc)~l
]:C1 ,LC2
+15 vn--~·A,"""""
Fig. 30 - Suggested changes in the mixer-to 3N140 pi-network interface circuit, producing lower Fig. 31 - Completed 6- and 2·meter converters
QL and better performance. (left and center) with power supply.
formance in the mixer-to-amplifier cir- cross-modulation as a prime system In at both choices of i-f. Best inter-
cuitry. In cases where only a small portion consideration, this 3: 1 change (reduction) modulation figure for the 3N201 was
of a band is of interest the original circuit in source resistance implies a 3: 1 im- obtained with a small forward bias on
values are adequate. For those who need provement in cross-modulation and total gate 1, and the bias-circuit modification
to receive over a considerable portion of a harmonic distortion. shown may be tried, if desirt:d.
band, say one to two MHz, a change of Tests on the 3N201 dual-gate MOSFET
some component will provide improved have shown device noise performance to Converters for SO, 144 and 220 MHz
performance over a broad fange while be excellent for source impedances in the The converters here were originally
maintaining an acceptable noise figure. 1000- to 2000-ohm region. For optimum designed by the Rochester VHF Group
The term "nominal 50-ohm impedance" noise and good cross-modulation per- and appeared in August 1973 QST. The
applied to diode DBM ports is truly a formance, the nominal 50-ohm mixer i-f design was the basis for a club project so
misnomer, as their reflective impedance is output impedance is stepped up to about the same board is used for all three
rarely 50 ohms +;jO and a VSWR of 1 is 1500 ohms for i-f amplifier gate 1, using converters, with only slight modifications
almost never achieved. Mixer perform- the familiar low-pass pi network. This is a for each band. Design features include
ance specified by the manufacturer is mismatched condition for gate 1, as the 1) Low noise figure.
measured in a 50-ohm broadband system, device input impedance for best gain in 2) State-of-the-art freedom from cross
and it is up to the designer to provide an the hf region is on the order of lOkO. modulation.
equivalent termination to ensure that the Network loaded-Q values in the article are 3) Sufficient gain to override the
unit will meet specifications. Appropriate a bit higher than necessary, and a design front-end-noise of most receivers.
matching techniques at the rf and LO ports for lower Q L is preferred. Suggested 4) Double-tuned bandpass interstage
will reduce conversion loss and low-power modified component values are listed in and output circuits to achieve a flat
requirements. Complex filter synthesis can Fig. 30. High-frequency attenuation is response over a 2-MHz portion of either
improve the i-f output match. However, if reduced somewhat, but satisfactory noise band.
one does not have the necessary equip- and bandwidth performance is more 5) Filtering of the local oscillator chain
ment to evaluate his efforts, they may be easily obtained. Coil-form s'ize is the in the 144- and 220-MHz models to
wasted. Simple, effective, easily reproduced same, so no layout changes are required reduce sp~rious responses.'
reduced circuitry was desired as long as for the modification. Components in the 6) Small size and low power con-
the trade-offs were acceptable, and measure- interface must be of high Q and few in sumption.
ments indicate this to be the case. number to limit their noise contribution 7) Freedom from accidental mistuning
The most critical circuit in the com- through losses. The 28-MHz values during the life of the converter.
bined unit is the interface between mixer provide satisfactory interface network Other points considered were such things
and i-f amplifier. It must be low-pass in performance over a 2-MHz bandwidth. A as freedom from the necessity of neu-
nature to satisfy vhf signal component higher Q L in the 28-MHz interface can be tralization and the use of moderately
bypassing requirements at the mixer i-f useful if one narrows the output network priced transistors.
port. For best mixer IMD characteristics and covers only a few hundred kilohertz Several breadboard models were con-
and low conversion loss, it must present to bandwidth, as is commonly done in structed and tested as the design evolved.
the i-f port a nominal 50-ohm impedance 432-MHz weak-signal work. Fig. 31 shows two completed converters
at the desired frequency, and this im- Device biasing and gain control methods and a power supply.
pedance value must not be allowed to were chosen for simplicity and adequate
increase as i-f operation departs from performance. Some sort of gain adjustment Circuit Design
midband. The impedance at the i-f is desirable for drain-circuit overload pro- Schematic diagrams ~or the three ver-
amplifier end of the interface network tection. It is also a handy way to "set" the sions are shown in Figs. 32, 33 and 34.
must be in the optimum region for receiver S meter. A good method for gain The configuration of the rf and mixer
minimum cross-modulation and low noise. adjustment is reduction of the gate-2 bias portions of the circuit are virtually
A dual-gate device offers two important voltage from its initial optimum-gain bias identical for all three with the values of
advantages over most bi-polars. Very point (greater than +4 V dc), producing a the frequency-determining components
little, if any, power gain is sacrificed in remote-cutoff characteristic (a gradual re- being scaled appropriately. The major
achieving best noise Tigure, and both duction in drain current with decreasing gate difference between the converters is a
parameters (gain and NF) are relatively bias). The initial gain-reduction rate is higher change in the local oscillator chain. A
independent of source resistance in the with a slight forward bias on gate I, than fm minor change in the method .of interstage
optimum region. As a result, the designer Vg IS = O. Input and output circuit coupling was necessary to prevent stray-
has a great deal of flexibility in choosing a detuning resulting from gain reduction capacitance effects from making the
source impedance. In, general, a 3: 1 (Miller effect) is inconsequential as the alignment critical on the 50-MHz con-
change in source resistance results in only gate-l and drain susceptances change very verter.
a I-dB change in NF. With minimum little over a wide range of Vg 2S and All inductors in the 50-MHz model and
9-19 Chapter 9
28 MHz
R1
50 MHz 270 50 MHz
AMP.
L1 QI
C3 I-F
(
II 22
Is
300
L5 OUTPUT
INPUT
C4 R2
0.01 330
osc.
-1-
Q3
MPF102
--- -- - -----.- - ------J,--.
EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPACITANCE ARE
IN MICROFARADS (pF I ; OTHERS
ARE IN PICOFARADS' ~F OR JOpF);
RESISTANCES ARE IN OHMS;
k -1000. M-I 000 000
R9
330
L.-_ _ _ _ _~N'~-------- .....--o.12 V
Fig. 32 - Schematic diagram of the 50·MHz converter. All reSistors are 1I4·watt composition. C2, C8, C10 and C15 are 0.001 I'F disc ceramic. C4 is
0.01·I'F disc ceramic. All other capacitors are dipped mica.
L1·L6, incl. - All no. 28 enam. wire wound on L3, 12 turns; L4, 18 turns; L5, 18 turns tap· tapped at 6 turns from hot end.
Amidon T·25·6 cores as follows: L1, 14 turns ped at 4 turns from cold end; L6, 26 turns Y1 - 22·MHz crystal. International Crystal
tapped at 4 turns and 6 turns; L2, 13 turns; Mfg. Co. type EX.
R1 MIXER
270
AMP.
28 MHz
INPUT
I·F
OUTPUT
L5
C7 C12
47
100
I i
I I
----------------~-----------~
osc. ----rL---
EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPAC ITANCE ARE
IN MICROFARADS (pF I ; OTHERS
ARE IN PICOFARADS (pF OR JOpF);
RESISTANCES ARE IN OHMS;
k -1000.101-1000000
38.6 MHz
Ylt:::J
RIO
330
~----------------~------------~---~+12~
Fig. 33 - Schematic diagram of the 144·MHzconverter. All resistors are 1/4·watt composition. C8, C10, C15 and C18 are O.OOI·I'F disc ceramic. All
other capacitors are dipped mica units.
L1, L2, L3, L7, L8 - All no. 20 enam. wire from hot end. cold end.
formed by using the threads of a 1/4·20 bolt L4 - 18 turns no. 28 enam. wound on Amidon L6 - 16 turns no. 28 enam. wound on Amidon
as a guide. L1, 5 turns tapped at 1·3/4 turns T·25·6 core. T·25·10 cote.
and 314 turn from cold end; L2, 5 turns; L3, L5 - 18 turns like L4, tapped at 4 turns from Yl - 38.666·MHz crystal. International Crystal
4 turns; L7 and L8, 5 turns tapped at 2 turns Mfg. Co. type EX.
C12
47
----rL---
EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPAC ITANCE ARE
IN MICROFARADS (JiF I ; OTHERS
ARE IN PICOFARADS (pF OR JlJIFI;
RESISTANCES ARE IN OHMS;
k -1000, M-I 000000
48.0 MHz
Y1CJ
Rl0
330
+12V
Fig. 34 - Schematic diagram of the 220·MHz converter. All resistQrs are 1/4·watt composition. C8, C10, C15 and C18 are 0.001·,..F disc ceramic. All
other capacitors are dipped mica units.
L1, L2, L3, L7, L8 - All no. 20 enam. wire cold end, and L8, 4 turns tapped at 2·1/2 cold end.
formed by using the threads of a 1/4·20 bolt turns from hot end. L6 - 18 turns no. 28 enam. wound on Amidon
as a guide. L1, 4 turns tapped at 1'1/2 turns L4 - 18 turns no. 28 enam. wound on Amidon T·25·10 core.
and 3/4 turn from cold end; L2, 4 turns; L3, T·25·6 core. Y1 - 48.000 MHz crystal. International Crystal
2 turns; L7, 3 turns tapped at 1·1/2 turns from L5 - 18 turns like L4, tapped at 4 turns from Mfg. Co. type EX.
the 28-MHz output circuit are wound on winding wire around the threads of a feedback in this case. The drain tank is
Amidon toroid cores. The tuned circuits. 1/4-20 bolt. The turns of Ll are spread to modified to provide output at the third
are aligned by spreading or compressing permit adding taps prior to mounting on harmonic, thus eliminating the need for a
the turns around the toroid core. After the board. The degree of interstage separate'tripler stage. Q4 is used as an
alignment the coils are glued in place with coupling in the two-meter model is isolation amplifier running at very low
Silastic compound (sold as bathtub controlled by the positions of L2 and L3. current level (as controlled by R9) to
caulk). Since they are mounted at right angles, the provide attenuation of the adjacent
The rf amplifier; Ql, is used in a coupling is very light. By changing the harmonics. This stage is not needed for
grounded-gate configuration. The input angle between these two coils, the amplification of the oscillator signal but
circuit is tapped to provide a proper passband. may be optimized. without the additional filtering, severe
match between the antenna and source of In the 144- and 220-MHz converter oscil- "birdies" may result from nearby fm or
the FET while maintaining a reasonable lator stages, Q3 is changed to an oscil- TV stations. In all three versions, a
Q. The 50-MHz inter stage coupling lator/tripler by replacing the source bias, number of printed-circuit pads will be left
network consists of C3, C5, L2 and L3. resistor with L6. Replace bypass capa- over when construction is completed.
Band-pass coupling .is controlled by the citor, C13, with a suitable value to These are the result of providing both
capacitive T network of C3 and C5 in resonate L6 near the crystal frequency. bands on a common pc layout. For
ratio with C6. A 40673 dual-gate MOS- Source-to-gate capacitance provides the example, the isolation amplifier following
FET is used in the mixer circuit (Q2).
Gate 1 receives the signal, while gate 2 has
the local-oscillator injection voltage ap-
plied to it through C7. A slight amount of
positive bias is applied to gate 2 through
R2. A top-coupled configuration, using
toroid inductors, serves as the 28~MHz
output circuit of both converters.
The oscillator circuit in the 50-MHz
model is straighforward, relying on the
drain-to-gate capacitance of the FET for
feedback. A tap at four 'turns from the hot
end of the toroid winding provides the
injection to the mixer through capacitor
C7. In the 144- and 220-MHz converter
the rf stage is identical to the 50-MHz
version except for the tuning networks.
Ll, L2, and L3 are air wound, self- Fig. 35 - Scale·size layout for the pc board. The same pattern is used for each band. Foil side
supporting, and are formed initially by shown here.
9·21 Chapter 9
'=
F
o
LI
.£L.
"til
'IIR ...u...;u.,I Re C8 1CIO .
I
Lt, co I
I
s't
L.4
II ICt 0
using a sweep generator, diode probe, and
oscilloscope is a necessity to assure the
flat response over \ the tuning range.
Commercial attenuators can be used to
calibrate each converter by the sub-
RF I "L.3 01 0
0
\I-F stitution method.
OUT
o Tuning of the air-wound rf circuit for
I
0
SHIEL.D YI 0 0 RII L __ 02
_CI2 I-F
0 144 and 220 MHz is accomplished by
spreading or compressing the turns of the
o ~ R "c131 " ~ 0 ~7 0 0
OND
coils. After aligrt'ment, the windings are
Ie: CI~ 0
87' 0 0 +12V
O~ICI4
secured by a bead of Silastic compound
" 0 0 " 0
0 along the coil to hold t4e turns in place.
" Q3
G S D
0" 0
The transistors used in the if stage are
also subject to some variation in noise
(A) figure. When this occurs, an rf FET
should be carefully traded with an
'jl Jell
i:
oscillator FET, since performance of the
~ ~ ~ FET as. an oscillator usually is satisfac-
fl
• I'R2 e. I
~
o CO C8 Ic 10 0
tory.
LI etI IJl
I La, LS I • 't
, 1M L' L' Small ceramic trimmers can be used in
r
R3 a 0
lell
1-,
the tuned circuits of these conyerters. The
- - - __ ~L ___ - -~
....
i ",
01 R8 cia 1-" midrange of the trimmer should be
SHIILD VI
o ~
L
1" 1!8
~
--
CI7 C7 RIO OND
0 approximately the value of the mica
capacitors replaced. This procedure may
• "oGIJ,~ ILl---
,COl simplify the tuning process of the con-
0
• 0' ' .
03 0
GJnI S D -
'.1
~
0 S...
0
-
+It,
0
verters where a sweep generator setup is
not available. A little careful tweaking
should give a reasonably flat response.
(B) If trimmers are used, the rf input circuit
should be tuned to the center of the
desired response, 50.5 MHz as an
Fig. 36 - Parts·placement guide for the 50·MHz converter, A, and the 144· and 220·MHz con·
verters, B. View is from the foil side of the board. Dashed lines show the location of shields that
example. This circuit tunes broadly and is
are soldered to short' pieces of wire which project thro.ugh holes In the pc board. The shields may not too critical. The rf interstage circuits
be fabricated from sheet brass or copper, or scraps of copper·clad board material. should be stagger tuned, one at 50.0 MHz
~
rON T1
lead clippings are soldered into these hqles
to provide a mounting for the shield
partitions, which are soldered to the wires
where they extend through the board. Fig.
36 shows the parts layout for the three
converters. Notice that one lead of C3 SWITCH AND TRANSFORMER I,
!'lOT 01'1 pc BOARD
must reach past the ground hole and I
connect to the foil. R3 is not used on the (A)
50-MHz converter.
A/~gnment and Test
Perhaps the most difficult task in the
project was the test and tune-up of the
finished converter. A single test setup
RI R2
FROM
CO!'lVERTER R3 TO I-F
I-F OUT RECEIVER
28 MHz
r 39
rrrn
L5 J,
1
(84MHzI
51
~XCEPT AS INDICATED, DECIMAL S.M. =SILVER MICA
VALUES OF CAPAC ITANCE ARE
IN MICROFARADS (Jjf I ; OTHERS
ARE IN PICOFARADS (pF OR JjJjF);.
RESISTANCES ARE IN/OHMS; 100
k -1000. M.I 000 000
+12V
Fig. 41 - Schematic diagram of the passive mixer, diplexer and post-mixer amplifier. Fixed-value capacitors are disc ceramic except those shown
as feedthrough types. Resistors are 1/2-watt composition.
C6-C8, inci. - 35-pF ceramic or Mylar trimmer. L5 - 5 turns no. 24 enam. wire on T-50-6 03 - Siliconix E300 JFET or 2N4416.
Solder 47- and 33-pF silver mica capacitors toroid core (0.09 "H). U1 - Diode-quad doubly balanced mixer. MCl
across pc foils for C7 and C8, respectively. l6 - 21 turns no. 24 enam. wire on T-50-6 SRA-1 or SBL1, or Cimarron CM-1 suitable.
toroid core (1.8 "H).
1
33
ISO
9.1V
40QmW
01
+12V
h E·aol FLS
'rJ,FT 116 MHz
DOUBLER C12
47
1N914
560 TO
....+---0 U1
MIXER
C9 can be adjusted to shift the oscillator LO chain as viewed on an HP analyzer. It disrupting the complete converter. Most
frequency by a small amount. Greater can be seen that all spurious responses are of the circuit boards are much larger than
range for netting the crystal can be 72 dB or greater below the desired necessary. This was done with a view
obtained by inserting a small amount of II6-MHz energy level. Imagine that the toward possible revisions to the circuits of
inductive reactance between C6 and Y I. carrier' amplitude is full scale when the first model. Those who are skilled at
A tuned toroidal transformer, TI, is comparing the levels of the spurs. The layout work may want to compress the
adjusted for resonance at arproximately II6-MHz carrier has been suppressed by circuits somewhat.
58 MHz. ClO is set for reliable starting of means of a trap to prevent front-end Small Teflon press-fit feedthrough bush-
the oscillator (consistent with high out- overloading of the analyzer. This has no ings are used as input and output
put) when the operating voltage is turned effect on the accuracy of the spur-level terminals for the various modules. They
on. Zener-diode regulation (DI) is used to readings. The full-scale line at the far were purchased at a flea market, so the
aid oscillator stability. left in Fig. 43 is the zero-reference blip original source is not known. Any
Output from Q4 is routed to a fed-back, from the analyzer. It should be ignored. low-loss miniature bushing should be
broadband, Class A amplifier, Q5. The Q4, Q5 and the related circuitry are suitable as a substitute. Alternatively, the
feedback provides a 50-ohm input charac- contained in a module made from RG-I74 miniature coaxial cables can be,
teristic and contributes to excellent stability single-sided pc board. The copper foil is routed directly into and out of the
of the 58-MHz amplifier. Negative feed- on the inside of the box. D2, D3, Q6 and modules for connection to the circuits.
back and emitter degeneration are used at related components are in a separate For attachment to the + 12-volt supply,
Q5 and Q6 for this purpose. T2 is a broad- container. Doubl~-sided pc board is used. each module has O.OOI-/!,F feedthrough
band 4: I toroidal transformer. It trans- capacitors. These components are mount-
forms thy collector impedance ofQ5 to 50 Construction Data ed on the box walls. Small Teflon
ohms. Output is taken at this point and Modular construction is used in this feedthrough bushings are used at Q I and
fed to a separate module which contains design so that various portions of the Q2 to permit circuit connections for the
T3 through FL5. circuit can be isolated from one another in FET leads inside the module. The FETs
The output from Q5 is used to drive a an effective manner. Another benefit to are installed on the outer wall of the
two-diode balanced doubler, D2 and D3. this style of construction is that the front-end module.
A trifilar-wound broadband toroidal trans- builder can experiment with other circuits All 'of the modules are affixed to the
former (T3) couples the energy to the (substitutes) and install them without main chassis 7 X II X 2 inches (180 X 280
doubler. L9 and Cll comprise a II6-MHz
resonator which serves as an impedance
transformer between the diodes and Q6. 05
This resonator also suppresses energy at
other than 116 MHz. The diodes and the
base of Q6 are tapped close to ground on 117
L9 to effect an impedance match. The VAC
?1-......-~o---o+12V
stage gain from L9 to the outpJ.lt of FL5 REG.
1000
was measured as 16 dB. The gain is 5W
needed to ensure a +7-dBm injection level
at Ul, the mixer. 51 ON
~
former, T4, is used to effect an impedance EXCEPT AS INDICATED, DECIMAL VALUES OF ~3(Qndl,
CAPACITANCE ARE IN MICROFARADS I JlF 1 ;
U2
match. It transforms' the collector im- OTHERS ARE IN PICOFARADS I pF DR JlJIF1;
TOP
RESISTANCES ARE IN OHMS;
pedance of Q6 to the 50-ohm input k·1 000, M"OOO 000. 1 32
impedance of band-pass filter' FL5. The
latter greatly suppresses the 58-MHz'
Fig. 45 - Schematic diagram of 'the regulated 12·volt power supply used with the 2·meter con·
energy passing through Q6. It also rejects verter. CapaCitors are disc ceramic except the one with polarity marked, which is electrolytic.
the harmonics of the I 16-MHz LO chain. 04·07, inc!. are 1·,!., 50 PRV silicon diodes. S1 is a spst toggle switch. T5 delivers 24V ac at 1 A,
Fig. 43 shows the spectral output of the and U2 is a National Semiconductor three·terminal regulator.
9-25 Chapter 9
INPUT
Fig. 46 - Scale layouts and parts·placement guides for the, three circuit boards, Views are shown from the bolt om (etched sides) of the boards.
x 50 mm) by means of no. 6 spade bolts. amplifier assembly are 5-3/4 X 1-7/8 terialls used for all of the etched circuits.
The latter are bolted to the side walls of inches (155, X 50 mm), length and width. A ground-bus copper strip is retained
the modules. The rf amplifier assembly The box height is 1-1/2 inches (3.8 cm). around the entire perimeter of each board.
uses the chassis surface as part of the box This container and the one for', the This permits the builder to solder the
shield. An aluminum plate is attached to low-level stages of the LO chain do not boards into the shield boxes.
the side wall of the module by means of have copper on the outside. Therefore it is The low-level section of the LO chain is
spade bolts. This type of construction was necessary to provide a grounding contact housed in a box which is 5-1/4 inches (135
used to ease assembly of the front end for the press-fit U-shaped aluminum mm) long, 2-1/4 inches (58 mm) wide and
. filter and amplifiers. Spring-brass fingers covers. Shim-brass or flashing-copper 2 inches (51 mm) high. The last half of the
are soldered to the compartment dividers strips are located opposite one another on LO circuit is contained in an enclosure
to assure solid contact to the alumin'um the upper lips of the side walls. The strips which is 4-1/8 inches (105 mm) long,
side plate (see Fig. 44). The aperture size are approximately 1-1/2 inches long (38 1-1/2 inches (38 mm) wide and 1-1/4
between L2 and L3 of FL2 is 1 x 5/8 inch mm) and 3/4 inch (19 mm) wide. They are inches (32 mm) high. The end com-
(25 X 16 mm). Each resonator compart- soldered to the inner walls of the box, then partment houses the bandpass filter. It is
ment measures 1-1/4 X \-1/4 X 2 inches bent over the edges and down the,outside 1-1/2 X 1-1/2 inches (38 X 38 mm) square.
(32 X 32 X 50 mm). The' coils are of the box walls This provides a ground The depth of the compartment is 1 inch
centered in the compartments. contact for the box covers (25 mm). '{he lower surface of the filter
Dimensions for the mixer Ipost-mixer Single-sided glass-epoxy pc-board ma- compartment is part of the main pc board,
II A II
1 I
i
--F "1-
IC
I
I
I
I
I
'l':'::"':l
E! --rr
r--
f T i1)( /MET~~:~ATE 1
l
I I
ternatively, angle stock can be made of I I
flashing copper and soldered across each
corner of the box. The V-shaped lids will
! li)(
I
I
I
I
rr--01
IB
I
I
X
L>/,""ro'l
I : Y OR MYLAR
: Ie)! I
·I~
I J
AUgnment [2 ___ .QI
i "
:J
I
I 01 ~
Choose a 200-kHz segment of the I Cl
I ~
2-meter band that suits your purposes. t.
~
.25" (6.32"")
T
Place a 2-meter signal at the midpoint of
that tuning range (144.1 MHz for
coverage from 144.0 to 144.2 MHz). ~ PLASTIC
SHOULDER
f. ~75"
(;9~;;;j
Connect the converter to a receiver tuned ~ASHER
and peaked at 28.1 MHz. Apply power to SOFT COPPER TUBING
SOLDERED TO WALL 4-40 SCREW
the converter and locate the 2-meter signal AND NUT
in the receiver tuning range. It should fall
close to 28.1 MHz. Observe the receiver S ENCLOSURE DIMENSIONS
meter and adjust all of the converter tuned FREQ. x Y z
circuits for maximum meter reading. This '296 MHz
2.00"
(50.8",,,,1
2.00" 2.25'
(50.1 ... 111) (57. 1m",)
procedure should be repeated two or three , -, ~ r~~ .9375" 1.00" 1.25"
+-~ ~r+ .375 2304 MHz {IS.'mm, (25.4m'" (31,7111111'
times again, using the weakest 2-meter \ _/ '-(., (9.25,"") L-_ _~=::::::'~::":':::::.L=='-'
Power Supply
A well-filtered 12-volt regulated dc
supply is recommended for use with this
converter. It should be capable of
accommodating a load current of 100 rnA
or greater. The unit shown here contains a
12-volt regulated supply under the main
chassis. A red LED is used as the on-off
indicator. The circuit is shown in Fig. 45.
A bottom plate can be added to the
c;hassis. Adhesive-backed plastic feet (four)
should be affixed to the bottom plate to
prevent scratching the operating desk.
Pc-board layouts and parts-placement
guides are given in Fig. 46. The drawings
show the etched sides of the boards. Each Fig. 48 - The converter for 1296 MHz. This unit was built by R.E. Fisher, W2CQH. While the mixer
board is illustrated to scale. assembly (top center) in this model has solid brass walls, it can be made from. lighter material, as
explained in the text and shown in Fig. 47. The i-f amplifier is near the center, just above the
Interdigital Converter for 1296 or 2304 mixer-current-monitoring jack, J1. A BNC connector at the lower left is for 28-MHzoutput. The
MHz local oscilll;ltor and multiplier circuits are to the lower right. Note that L6 is very close to the
chassis, just above the crystal. The variable capaCitor near the crystal is an optional trimmer to
In a world where rf spectrum pollution adjust the oscillator to the correct frequency.
9·27 Chapter 9
INPUT
Fig. 46 - Scale layouts and parts·placement guides for the, three circuit boards. Views are shown from the bottom (etched sides) of the boards.
X 50 mm) by means of no. 6 spade bolts. amplifier assembly are 5-3/4 X 1-7/8 terialls used for all of the etched circuits.
The latter are bolted to the side walls of inches (155. X 50 mm), length and width. A ground-bus copper strip is retained
the modules. The rf amplifier assembly The box height is 1-1/2 inches (3.8 cm). around the entire perimeter of each board.
uses the chassis surface as part of the box This container and the one for .the This permits the builder to solder the
shield. An aluminum plate is attached to low-level stages of the LO chain do not boards into the shield boxes.
the side wall of the module by means of have copper on the outside. Therefore it is The low-level section of the LO chain is
spade bolts. This type of construction was necessary to provide a grounding contact housed in a box which is 5-1/4 inches (135
used to ease assembly of the front end for the press-fit U-shaped aluminum mm) long, 2-1/4 inches (58 mm) wide and
. filter and amplifiers. Spring-brass fingers covers. Shim-brass or flashing-copper 2 inches (51 mm) high. The last half of the
are soldered to the compartment dividers strips are located opposite one another on LO circuit is contained in an enclosure
to assure solid contact to the aluminum the upper lips of the side walls. The strips which is 4-1/8 inches ( 105 mm) long,
side plate (see Fig. 44). The aperture size are approximately 1-1/2 inches long (38 1-1/2 inches (38 mm) wide and 1-1/4
between L2 and L3 of FL2 is 1 X 5/8 inch mm) and 3/4 inch (19 mm) wide. They are inches (32 mm) high. The end com-
(25 X 16 mm). Each resonator compart- soldered to the inner walls of the box, then partment houses the bandpass filter. It is
ment measures 1-1/4 X 1-1/4 X 2 inches bent over the edges and down the outside 1-1./2 X 1-1/2 inches (38 X 38 mm) square.
(32 X 32 X 50 mm). The' coils are of the box walls This provides a ground The depth of the compartment is 1 inch
centered in the compartments. contact for the box covers (25 mm). 1'he lower surface of the filter
Dimensions for the mixer Ipost-mixer Single-sided glass-epoxy pc-board ma- compartment is part of the main pc board,
t'
VHF and UHF Receiving Technique. 9-26
the end of which has not been stripped of
copper. 4.8"
TO BOTTOM
GROUNDPLANE \ '
)( RI
I' _
through the box walls at four points per ----F=---F ~-~-r=-=r---J: ~- co .f--, _" c--.,
i
-i
r=~T.--TI.-~~~~-ff,-~~--~~hr~
wall, then soldering the wires in place on : A :
each surface (inner and outer). Al- I I
ternatively, angle stock ca"'n be made of
flashing copper and soldered across each i i x
l
I I I I
corner of the box. The U -shaped lids wiII
:: : I :
complete the electrical contacts when I I I I Y I I y
1 ~F
I TEFLON
installed. lor-'o I I I I I~ OR MYLAR
Alignment
: IC) i ::,:: 1:1 :1
1 t //INSULATOR
-
Choose a 200-kHz segment of the '12:-=.Q, ': : OIl CI :! 02
J
2-meter band that suits your purposes. t
~
.25" (S.32mm)
Place a 2-meter signal at the midpoint of
that tuning range (144.1 MHz for
coverage from 144.0 to 144.2 MHz).
Connect the converter to a receiver tuned
and peaked at 28.1 MHz. Apply power to •
J
SOFT COPPER TUBING
!"'--SHOULDER
~ASHER
PLASTIC
4-40 SCREW
t
r 75 "
(;9m;;.i
•
SOLDERED TO WALL
the converter and locate the 2-meter signal AND NUT
in the receiver tuning range. It should fall
close to 28.1 MHz. Observe the receiver S ENCLOSURE DIMENSIONS
Power Supply
A well-filtered 12-volt regulated dc
supply is recommended for use with this
converter. It should be capable of
accommodating a load current of 100 rnA
or greater. The unit shown here contains a
12-volt regulated supply under the main
chassis. A red LED is used as the on-off
indicator. The circuit is shown in Fig. 45.
A bottom plate can be added to the
chassis. Adhesive-backed plastic feet (four)
should be affixed to the bottom plate to
prevent scratching the operating desk.
Pc-board layouts and parts-placement
guides are given in Fig. '46. The drawings
show the etched sides of the boards. Each Fig. 48 - The converter for 1296 MHz. This unit was built by R.E. Fisher, W2CQH. While the mixer
boarc;l is illustrated to scale. assembly (top center) in this model has solid brass walls, it can be made from, lighter material, as
explained in the text and shown in Fi9. 47. The i·f amplifier is near the center, just above the
Interdigital Converter for 1296 or 2304 mixer·current·monitoring jack, J1. A BNC connector at the lower left is for 28·MHzoutput. The
MHz local oscillator and multiplier circuits are to the lower right. Note that L6 is very close to the
chaSSis, just above the crystal. The variable capaCitor near the crystal is an optional trimmer to
In a world where rf spectrum pollution adjust the oscillator to the correct frequency.
9-27 Chapter 9
EXCEPT'AS INDICATED, DECIMAL
VALUES OF CAPACITANCE: AR£
IN MICROFARADS ( .IIFl ; OTHERS
ARE IN PICOFARADS (pF OR .II.11FI;
R£SlST ANCES ARE I N OHMS; ,
'k -1000. M.I 000000
01
129S-MHz
INPUT J4
'Fig, 49 - Schematic diagram of the 1296-MHz converter with OSCillator and multiplier sections included, Dimensions for the filter and mixer assembly are
given in Fig. 47,
C1, C2 - 30·pF homemade capacitor. See text 01 - Hewlett Packard 5082·2577 or 5082-2835. nominal).
and Fig, 2, 02 - Hewlett Packard 5082·2811 or 5082·2835. L3 - 10 turns like L 1 (0.5 "H).
C3, C4 - 0,8- to 10·pF glass trimmer, Johanson J1 - Closed·circuit jack. L4, L5 - 6 turns like L 1 (0.2 "H).
2945 or equiv, J2 - Coaxial connector, type BNC acceptable. L6 - Copper strip, 1/2-inch wide x 2·1/2 inches
C5 - 0,001-"F button mica. L1, L2 - 18 turns no. 24 en am. on 1/4·inch (127 x 635 mm) long. See text and photo·
C6 - 2· to 20·pF air variable, E.F. Johnson 189- (6.3-mm) 00 slug·tuned form (1.5 "H graphs.
507-004 or equiv. RFC1 - 33 "H, J.W. Miller 74F33SA1 or equiv.
is becoming more serious, even into the copper-clad epoxy fiberglass. One ground and serves as the coupling section to the
microwave region, it is almost as im- plane is made larger than the microwave filter input. Rod B is the high-Q resonator
portant to keep unwanted signals out of a assembly and thus provides a convenient and is tuned by a 10-32 machine screw.
receiver as it is to prevent radiation of mounting plate for the remainder of the Rod C provides the filter output-coupling
spurious energy. An interdigital filter was converter components. section to the mixer diode, D I.
described some years ago, featuring low The sidewalls are bent from 0.032-inch The mixer diode is a Hewlett-Packard
. insertion loss, simplicity of construction, thick sheet brass or they can be made from 5082-2577 Schottkey-barrier type which is
and reasonable rejection to out-of-band 1/4 X 3/4-inch (6 X 19 mm) brass rod. available from distributors for about $4.
signals. It could be used in either One edge of each sidewall is soldered to the The cheaper 5082-2835, selling for 90
transmitters or receivers. larger ground plane. Thl? other edge is cents, can be used instead, but this
This twice-useful principle has now fastened to the smaller ground plane by substitution will increase the 2304-MHz
been put to work again - as a mixer. 4-40 machine or self-tapping screws, each mixer noise figure by approximately 3 dB.
Again, the ease of construction and located over the centerline of a rod. The One pigtail lead of the mixer diode is'
adaptation leads many to wonder that it sidewall edges should be sanded flat, tack-soldered to a copper disk on the
had not been thought of before. It was before the ground plane is attached, t'0 ungrounded end of rod C. Care should be
first described by W2CQH in January assure continuous electrical contact. N ott> taken to keep the pigtail lead as short as
1974 QST. that no end walls are required since there possible. If rod C is machined from solid
are no electric fields in these regions. brass stock, then it is feasible to clamp one
A Filter and Mixer Electrically, rods A, B, and C comprise of the mixer-diode leads to the rod end
A layout of the microwave portions of a one-stage, high-loaded-Q (Q L = 100), with a small setscrew. This alternative
both converters is shown in Fig. 47. The interdigital filter6 which is tuned to the method facilitates diode substitution and
structure consists of five interdigitated incoming signal frequency near 1296 or was u~ed in the mixer models shown in the
round rods, made of 3/8-inch (9.5 mm) 2304 MHz. The ungrounded end.ofrod A photographs.
OD brass or copper tubing. They are is connected to a BNC coaxial connector Fig. 47 also shows that the other end of
soldered to two sidewalls and centrally D 1 is connected to a homemade 30-pF
located between two ground-planes made 'Fisher, "lnterdigital Bandpass Filters for Amateur bypass capacitor, CI, which consists of a
of 1/16-inch (1.6 mm) sheet brass or VHF/UHF Applications," QST, March 1968. 1/2-inch-square copper or brass plate
VHF and UHF Receiving Techniques 9-28
EXCEPT AS INDICATED, DECIMAL VALUES OF
CAPACITANCE ARE IN MICROFARADS (pF I ;
OTHERS ARE IN PICOFARADS (pF OR ppFI;
. RESISTANCES ARE IN OHMS;
b 1000, M'IOOO 000.
I-F AMP:
220k
.6V
,.L0.OOI
Jl
Fig. 50 - Schematic diagram of the 2304·MHz version of the converter, with the i·f amplifier. The oscillator and multiplier circuits are constructed
separately.
C1, C23 - 30-pF homemade capacitor (see 02 - Hewlett Packard 5082·2811 or 5082·2835. slug-tuned form (0.25 "H). .
text). J1 - Closed·circuit jack.' L3 - Copper strip 1/2 x 2-11/16 inches (13 x
C3, C4, C5 - 0.8- to 10-pF glass trimmer, J2, J3, J4 - Coaxial connector, type BNC. 69 mm). See text and photographs.
Johanson 2945 or equiv. L1 - 5 turns no. 20 enam., 1/4·inch 10 x 1/2- RFC1 - Ohmite Z-144 or equiv.
01 - Hewlett Packard 5082·2577 or 5082-2835.' inch long. (6.35 x 12.7 mm). RFC2 - Ohmite Z·460 or equiv .
. L2 - 6 turns no. 24 enam_, on 1/4-inch 00
24V
.15 VC
(TO FIG.49)
1 0.001
L!.!....
\ ~ (SEE TEXT)
R1
Fig. 51 - Schematic diagram of the oscillator and multiplier for the 2304 MHz converter. As explained in the text, a fixed-value resistor may be
substituted for R1 after the valUE( that provides proper performance has been found.
C1, C2, C3 - 0.8· to 10·pF glass trimmer, L4 - Copper strip 1/2 x 1-1/2 in. (13 x 38
Johanson 2945 or equiv. L 1 - 10 turns no. 24 enam. on 1/4-in. 00 slug· mm). Space 1/8 inch (3.2 mm) from chassis.
C4 - 0.001-"F button mica. tuned form. RFC1 - 10 turns no. 24 enam. 1/8-in. 10, close
J1 - Coaxial connector, type BNC or equiv. L2, L3 - 3 turns like L1. wound.
9-29 Chapter 9
clamped to the sidewall_ with a 4-40 The multiplier circuits are tuned to
machine screw. The dielectric material is a Table 4 resonance in the usual manner by holding
small sheet of 0.004-inch (0.1 mm) thick Converter Specifications a wavemeter near ea~h inductor being
Teflon or mylar. A 4-40 screw passes tuned. Resonance in the Q3 collector
through an oversize hole and is insulated 1296 MHz 2304 MHz circuit is found by touching a VTVM
from the other side of the wall by a small Noise figure 5.5 dB 6.5 dB probe (a resistpr must be in the probe) to
Conversion gain 20 dB 14 dB
plastic shoulder washer. 3-dB bandwidth 2 MHz 7 MHz
C2 and adjusting the Johanson capacitors
In the first converter models con- Image rejection 18dB 30 dB until about -1.5 volts of bias is obtained.
structed by the author and shown in the I·f output 28 MHz 144 MHz The 317- to 1268-MHz multiplier cavity is
photographs, CI was a 30-pF button mica then resonated by adjusting the 10-32
unit soldered to the flange of a 3/8-inch machine screw until maximum mixer
diameter threaded panel bearing (H. H. current is measured at J I. When resonance
Smith No. 119). The bearing was then shielded housing. is found, RI should be adjusted so
screwed into a threaded hole in the that about 2 mA of mixer current is
sidewall. This provision made it con- Oscilhllor and Multipliers obtained. As an alternative to mounting a
venient to measure the insertion loss and The non microwave portion of the potentiometer in the converter, once a
bandwidth of the interdigital filters since converter is rather conventional. QI, a value of resistance has been found that
the capacitor assembly could be removed dual-gate MOSFET, was chosen as the provides correct performance it can be
and replaced with a BNC connector. 28-MHz i-r amplifier since it can provide measured and the nearest standard fixed-
Rods C, D, and E comprise another 25 dB of gain with a 1.5 dB noise figure. value resistor substituted. Some means of
high loaded-Q (Q L = 100) interdigital The mixer diode is coupled to the first adjusting the collector voltage on the
filter tuned to the. local oscillator (LO) gate of Q I by a pi-network matching multiplier stage must be provided initially
frequency. This filter passes only the section. It is most important that the to allow for the nonuniformity of transis-
fourth harmonic (1268 or 2160 MHz) proper impedance match be achieved tors.
from the multiplier diode, D2. The two between the mixer and i-f amplifier if a
filters have a common output-coupling low noise figure is to be obtained. In this A 2304-MHz Version
section (rod C) and their loaded Qs are case, the approximately 30-ohm output Fig. 50 and 51 show the schematic
high enough to prevent much unwanted impedance of the mixer must be stepped diagrams of the 2304-MHz converter and
coupling of signal power from the antenna up to about 1500 if Q I is to yield its rated multiplier. The mixer and i-f preamplifier
to the multiplier diode and LO power noise figure of 1.5 dB. It is for this reason was built on a separate chassis since, at
back out to the antenna. that a remote i-f amplifier was' not the time of their construction, a mUltiplier
The multiplier diode is connected to the employed, as is the case with many chain from another project was available.
driver circuitry through C2, a 30-pF contemporary uhf converters. An·i-f of 144 MHz was chosen, althou.e;h
bypass capacitor identical to C1. D2 is a Q2 functions in an oscillator-tripler cir- 50 MHz would work as well. An i-f output
Hewlett-Packard 5082-2811 although the cuit which delivers about 10 milliwatts of 28 MHz, or lower, should not be used
5082-2835 works nearly as well. Fifty of 158.5-MHz drive to the base ofQ3. The since this would result in undesirable
milliwatts drive at one-quarter of the LO emitter coil, L3, serves mainly as a choke interaction between the mixer and multi-
frequency is sufficient to produce 2 mA of to prevent the crystal from oscillating at plier interdigital filters.
mixer diode current, which represents its fundamental frequency. Coils L4 and - The 2304-MHz mixer and i-f amplifier
about 1 milliwatt of the local-oscillator L5, which are identical, should be spaced section, shown in Fig. 50, is very similar to
injection. A Schottky-barrier was chosen closely such that their windings almost its 1296-MHz counterpart. QI, the dual-
over the more familiar varactor diode for touch. gate MOSFET, operates at 144 MHz and
the multiplier because it is cheaper, more Q3 doubles the frequency to 317 MHz, thus has a noise figure about I-dB higher
stable, and requires no idler circuit. providing about 50 milliwatts drive to the than that obtainable at 28 MHz.
Fig. 49.shows the schematic diagram of mUltiplier diode. It is important that the The multiplier chain, Fig. 51 , has a
the 1296 to 28 MHz converter. All emitter lead of Q3 be kept extremely separate oscillator for improved drive to
components are mounted on a 7 X 9-inch short; 1/4-inch (6.3 mm) is probably too the 2N3866 output stage. Otherwise the
(178 X 229 mm) sheet of brass or long. L6, the strip-line inductor in the circuitry is similar to the 1296-MHz
copper-clad epoxy-fiberglass board. As collector circuit of Q3, consists of a 112 X version.
mentioned earlier, this mounting plate 2-l/2-inch (127 X 635-mm) piece of
also serves as one ground plane for the flashing copper spaced 1/8-inch (3-mm) References
microwave mixer. When completed, the above the ground plane. The cold end of Fisk. "Double-Balanced Mixers." Ham Radio. March
1968.
mounting plate is fastened to an inverted L6 is bypassed to ground by C5, a Ress. "Broadband Double-Balanced Modulator."
aluminum chassis which provides a O.OOI-~F button mica capacitor. Ham Radio, March 1970.
•
book described complete shielding active development in Europe. The group electrical' system having the lowest im-
methods for inductive-discharge ignitions. studying the situation there is CISPR pedance. If the regulator is adjustable, set
Late-model automobiles employ (Comite International Special des Pertur- the voltage no higher than is necessary to
sophisticated high-energy electronic igni- bations Radioelectrique). The SAE is ensure complete battery charging. Radio
tion systems in an attempt to reduce ex- cooperating with CISPR to promote inter- equipment manufacturers combat voltage
haust pollution and increase fuel mileage. national standardization. FCC docket variations by internally regulating, critical
With increased sophistication comes 20654 is a Notice of Inquiry concerning circuits wherever possible. . C
greater sensitivity to modification - solu- the automotive RFI issue. The manufac- Both' conducted and radiated noise can
tions to RFI caused by older Kettering turers and users of radios in the land be suppressed by filtering the alternator
systems cannot be uniformly applied to mobile service Oust above the 2-meter leads. Coaxial capacitors (about 0.5 /AF)
the modern electronic ignitions. band), represented by the Electronics In- are suitable, but don't connect a capacitor
Such fixes may be ineffective at best, dus~ries Association, responded to the to the field. The field lead can be shielded
and at worst may impair the engin'e per- NOI by stating that even vehicles comply- or loaded with ferrite beads if necessary.
formance. One should thoroughly under- ing with J551 can seriously degrade A parallel-tuned LC trap in this lead may
stand an ignition system before at- mobile vhf communications. SAE J551 be effective against radiated noise. Such a
tempting to modify it. One of the signifi- was never intended to protect vehicular trap in the output lead must be made of
cant features of capacitive discharge receivers - only fixed receivers located no. 10 wire or larger, as some alternators
systems, for example, is extremely rapid near roadways. conduct up to 100 amperes. The altef-
voltage rise, which combats misfire caused' On the whole, modern automobiles are nator slip rings should be kept clean to
by fouled spark plugs. Rapid voltage rise cleaner from an RFI standpoint than prevent excess arcing. An increase in
depends on a low RC time constant those of 20 years ago. The interference "hash" noise may indicate that the
presented to the output transformer. For problem, however, at least at vhf and uhf, brushes need to be replaced.
this reason, high-voltage suppression persists because present-day receivers are
cable designed for capacitive-discharge about 10 dB more sensitive than their , Instrument Noi~
systems is wound with monel wire. It ex- predecessors. Some automotive instruments are
hibits a distributed resistance of only Useful tips for solving ignition noise capable of creating noise. Among these
about 600 ohms per foot, as contrasted problems can be found in Giving Two- gauges and senders are the heat- and fuel-
with 10 kilohms per foot for carbon- way Radio Its Voice, published by the level indicators. Ordinarily, the addition
center cable used with inductive-discharge Champion Spark Plug Company. This of a 0.5 /AF coaxial capacitor at the sender
systems. Increasing the RC product by well-prepared publication covers noise- element will cure the problem.
shielding or installing improper spark- suppression fundamentals, preliminary Other noise-generating accessories are
plug cable could seriously compromise the procedures, methods for pinpointing in- turn signals, window-opener motors,
capacitive-discharge circuit operation. terference, and techniques for solving heating-fan motors and electric·
Ferrite beads represent a possible means noise problems in automotive and marine windshield-wiper motors. The installation
for RFI reduction in newer vehicles. Both environments. A copy of this booklet can of a 0.25-/AF capacitor will usually
primary and secondary ignition leads are be obtained for $1 from Champion Spark eliminate their interference noise.
Chapter 10 10·2
Faraday shield between LI and L2 ensures
SIGNAL
ANTENN-A that .the coupling is purely magnetic. The
coupling between L2 and L3 is purely elec- .
trostatic. Adjusting the coil coupling
causes the noise to null. The block
diagram of Fig. 3B illustrates a more
modern broadband approach to noise
'----.......--....\.~I RECEIVER cancellation. A short wire near the igni-
INPUT
tion coil couples impulse energy Into the
active impedance transformer, which is
simply an FETsource follower stage. The
amplitude and phase of the noise are con-
trolled by the attenuator and delay line,
respectively. The signal combiner can be a/
hybrid ferrite transformer at hf or a
·transmission line multicoupler at vhf.
NOISE VOLTAGE
PROBE BATTERY POWER
(The following material was assembled
by Dave Geiser, WA2ANU.) The
RECEIVER availability of solid-state equipment
INPUT
makes practical the use of battery power
under portable or emergency conditions.
Hand-held transceivers and instruments
(8) are obvious applications, but even fairly
powerful transceivers (100 W or so out-
put) maybe practical users of battery
power. Solid-state kilowatt mobile
Fig. 3 - Automotive noise-cancelling systems. At A, the circuit used in the BC-342 hI receiver. At
amplifiers exist, but these are intended for
B, a suggested broadband noise·cancelling scheme. . operating from an auxiliary battery that is
constantly charged. The lower-power
equipment can be powered from either of
two types of batteries, the "primary" bat-
Corona-Discharge Noise The Soviet "woodpecker" over-the- tery intended for one-time use, and the
Some mobile antennas are prone to cor- ~ horizon radar has inspired some serious storage (or "secondary") battery that
ona build-up and discharge. Whip anten- development work on noise blankers that . may be recharged many times.
nas which come to a sharp point will don't degrade receiver dynamic range. A battery is a group of chemical cells,
sometimes create this kind of noise. This Receivers for vhf fm service are generally usually series connected to give some
is why most mobile whips have steel or designed for optimum noise figure at the desired multiple of the cell voltage. Each
plastic balls at their tips. But, regardless expense 'of resistance to overload. Recent assortment of chemicals used in the cell
of the structure of the mobile antenna, advances in rf amplifier design have pro- gives a particular nominal voltage, and
corona buildup will frequently occur dur- ven that low noise figure and high this must be taken into account to make
ing or just before a severe electrical storm. dynamic range are not mutually exclusive .. up a particular battery voltage.
The symptoms are a high-pitched A high-performance noise blanker is Primary Batteries
"screaming" noise in the mobile receiver, • useless if the front end of the receiver
which comes in cycles of one or two overloads on the noise pUlses. A helical The most common primary cell is the
minutes duration, then changes pitch and resonator at the receiver input affords carbon-zinc flashlight type, in which
dies down as it discharges through the some protection against noise overload chemical oxidation converts the zinc into
front end of the receiver. The condition because it restricts the total noise energy salts and electricity. When there is no cur-
will repeat itself as soon as the antenna delivered to the front end. rent flow, the oxidation stops until the
system charges up again. There is no cure Some fm receivers suffer from impulse next time current is required. Some
for this condition, but it is described here noise because of inadequate a-m rejec- chemical action does continue, so even-
to show that it is not of origin within the tion. The cure for this ailment is to ensure tually stored batteries will degrade or d'ry
electrical system of the automobile. hard limiting in the i-f stages and to use a out to the point where the battery will no
detector that is inherently insensitive to longer supply the desired current. If this
amplitude variations. has happened without battery use, the
Electronic Noise Reduction Particularly troublesome vehicular im- time taken for the degradation is called
When all electrical noise generated pulse noise can sometimes be cancelled at shelf life.
within a vehicle has beef\ eliminated, the the receiver input. The technique involves The carbon-zinc battery has a nominal
mobile operator can be annoyed by RFI sampling the noise voltage from a separate voltage of 1.5 volts, as does its "heavy
from passing vehicles. Some measures can "noise antenna" and adjusting its phase duty" or "industrial" brother. These lat-
be taken in the receiver to reduce or reject and amplitude to cancel the noise ter types are capable (for a given size) of
impulse noise. (Noise limiters and noise delivered by the "signal antenna." For producing more milliampere hours and
blankers are discussed in the hf receiving this sytstem to be effective, the signal less voltage drop than a carbon-zinc bat-
chapter.) The placement of a noise antenna must be positioned to provide the tery of the same size, and also have longer
blanker in the receiver is important. The best possible signal-to-noise ratio, and the shelf life. Alkaline primary blltteries have
blanking circuit must be placed ahead of noise antenna located close to the noise even better characteristics and will retain
the sharp selectivity, otherwise the i-f source and effectively shielded from the more capacity at low temperatures.
filter will stretch the noise pulses, and they desired signal. Fig. 3A shows the noise Nominal voltage is 1.5 volts.
cannot be blanked without destroying a. cancellation circuit used in some models Lithium primary batteries have a
major portion of the received intelligence. of the BC-342, a WW II receiver. The nominal voltage of about 3 volts per cell
types should not be discharged more than stance, should not be less than 10.5 volts.
1.2 V below nominal battery voltage. (It is also good to keep a running record .00
The most widely used storage battery is of hydrometer readings, but the conven-
the lead-acid type. In automotive servic~ tional readings of 1.265 charged and 1.100
the battery is usually expected to discharged apply only to a long, low-rate 90 SIZE 21
discharge partially at a very high rate, and discharge. Heavy loads may discharge the
then to be recharged promptly while the battery with little reduction in the
alternator is also carrying the electrical hydrometer reading.)
load. If the conventional auto battery is Batteries that become cold have less of
~
allowed to discharge fully from its their charge availaele, and some attempt !70
SIZE 24
nominal 2 V per cell to 1.75 V per cell, to keep a battery warm before use is
only about 50 cycles of charge and worthwhile. The battery may lose 700-;0 or
60
discharge may be expected, with reduced more of its capacity at cold extremes but
storage capacity. recover with warmth.
The most attractive battery for extend- All batteries have some tendency to
ed high-power electronic application is the freeze, but those with full charges are less
AMPERES
so-called "deep-cycle" battery (intended susceptible. A fully-charged lead-acid bat-
for such use as powering electrical fishing tery is safe to _30°· F (_26° C) or colder.
motors and the accessories in recreational Storage batteries may be warmed
Fig. 4 - Output capacity as a function of
vehicles). The size 24 and 27 batteries fur- somewhat by charging. Blowtorches or discharge rate for two sizes of lead·acid
nish a nominal 12 volts and are about the other flame should never be used to heat batteries.
Chapter 10 10·4
traffic and DX-chasing situations the time
spent listening should be much greater
than that spent transmitting.
Caring for Storage Batteries
In addition to the precautions given
above, the following are recommended.
(Your manufacturer's advice will prob-'
ably be more applicable.)
Gas escaping from storage batteries
may be explosive. Keep flame away.
Dry-charged storage batteries should be
given electrolyte and allowed to soak for
at least half an hour. They then should be '
charged at perhaps a 15 A rate for 15
minutes or so. The capacity of the battery
will build up slightly for the first few
cycles of charge and discharge, and then
have fairly constant capacity for 'many
cycles. Slow capacity decrease may then
be noticed.
No battery should be subjected to un-
necessary heat, vibration or physical
shock. The battery should be kept clean.
Frequent inspection for leaks is a good
idea. Leaking or spraying electrolyte
should be cleaned from the battery and
surroundings. The electrolyte is chemical-
ly active and electrically conductive, and
may ruin electrical equipment. Acid may
be neutralized with sodium bicarbonate
(baking soda), and alkalies may be
neutralized with a weak acid such as
vinegar. Both neutralizers will dissolve in Fig, 5 - TheWA7ARK mobile NiCad charger, The power cable has an in-line fuse and ister-
water, and themselves be quickly washed minatedby a cigarette:lighteraccessory plug,
off. Do not let any of the neutralizer enter
the battery.
Keep a record of the battery usage, and
include the last output voltage and (for
lead-acid storage batteries) the
hydrometer reading. This allows predic- + 12 -14 V
tion of useful charge I'emaining, and the
recharging or procuring of extra batteries,
thus minimizing failure of battery power
during an excursion or emergency.
Charging Storage Batteries
The rated full charge of a storage bat-
tery, C, is exp'ressed in amper~-hours. 10k
-
current depends on the supply voltage.
This varies from about 12.5 V in a parked
vehicle to about 14.8 V when the engine is
running, causing the charging current to
vary from about 140 rnA to about 185
rnA. This will charge a 500 mAh battery
pack in five or four hours, respectively. If
-$-, WA7ARK HBK81
your battery requires a lower charging
current, you can scale RI (doubling RI
halves the charging rate, and so on.)
Ed Kalin, KIRT, designed the pc art- Fig, 7 - Etching pattern forthe mobile NiCad charger pc board, The foil side is shown; black
work and built the unit pictured in Fig. 5. represents ,copper.
Chapter 10 10-8
Table 1
Some Solar Battery Manufacturers and
DIstrIbutors
Solar Power Corporation
c/o Lindberg Company
4163 Montgomery, NE
Albuquerque, NM 87109
Tel. 505-881-1006
Solarex Corporation
1335 Piccard Dr.
Rockville, MD 20850
Tel. 301-948-0202
Applied Solar Energy
15251 E. Don Julian Rd.
City of Industry, CA 91746
Tel. 213-968-6581
Solec International
12533 Chadron Ave.
Hawthorne, CA 90250
Tel. 213-325-6215
AUTOMOTIVE
GROUND
Fig. 8 - Parts-placement guide forthe mobile NiCad charger pc board. The component side is
shown with an x-ray view of the foil.
+ 14V DC
+
~ _ (REGULATED)
BTl
~
BT2
~ * HEAT SINK
I
BT3
0-25
',' I
I~+ +
-I ' O-'~ -1",
1+
Fig. 12 - Photograph of a commercial dc·to·ac
'inverter that operates from 6 to 12 volts dc and
delivers 117 volts ac (square wave) at 175
SI 8T4 BT5 L watts.
VOLT
REG. "'-_________- ________--'+-{
I TEST
VOLT
TEST
R4
REFERENCE
470k
R5
4700
+ +
12V TO + + COMPARATOR
FROM SOLAR
RIGS
"';;'BT1 Cl
PANEL
1000
35V
+
C2 VOLTAGE ADJUST
0.33
35V
CASE IS
e'
03 COLLECTOR
BOTTOM \
FRONT TOP VIEW
..
",,1 0 ~
B VIEW
1 8
EOC BOE
4 5
IN ~ OUT 741 2N2907 2N3055
GNO
Fig'. 11 -Schematic diagram of the electronic switch. Resistances are in ohms; k = 1000; capacitance values are in microfarads (jAF).
BT1 - Automotive storage battery, lead'acid rent rating sufficient to pass full output of 9 in. 2 (5800 mm 2) or more.
type.' the solar panel. R7 - 10 kO, 1/2 watt, carbon control,
C1 ' - 1000-"F, 35-V electrolytic. 06 - Light·emitting diode, any type. linear taper, pc mount.
C2 - 0.33·"F, 35·V. 02 - Low·frequency power transistor; 2N3055, U1 - 3 terminal, 5·volt regulator.
03 - Silicon diode, PIV of 50 or more, cur· HEP S7000, or equivalent. Use heat sink of U2 - Op amp, any of the 741 family usable.
Chapter 10 10·8
Some manufacturers recommend a high-
detergent oil that comes in various service
grades such as MS, SO, and similar types.
Examine the top or side of the cans in
which the oil is sold and see if the letters
correspond to those recommended by the
engine manufacturer. ,
3) The carburetor mixes gasoline with
air, which is then burned in the engine.
·Before entering the carburetor, the air
must be filtered so that it is free of dust
and other foreign matter that might other-
wise be drawn into cylinder(s). Particles
that do get by the air filter are picked up
by the oil. That should be changed more
often if the alternator is operated in a
dusty location. Also, it is important to
clean the air filter frequently. It contains a
foam-like substance which can be cleaned
in kerosene and then soaked in fresh
Fig. 14 - The numbers indicate the primary
Fig. 13 - Large gasoline generators of the kW motor: oil. Squeeze excess oil from the maintenance points of a large power generator
and higher class are excellent for powering filter before replacing. Also, consult the (see text for details).
several amateur stations from a complex field instruction manual for further recommen-
site. Mainte,nance, as discussed in the text, is dations.
a vital matter to ensure reliable operation.
Here, WB9QPI has just finished a maintenance 4) Once the gas/air mixture' enters the
check of the group's Field Day power plant. cylinder, it is compressed by the piston in- Two very important safety precautions
to a very small volume and ignited by·the should be observed with regard to the ex-
spark plug. During the rapid burning that haust system. Never operate an alternator
then occurs, the expansion caused by the in closed surroundings such as a bUilding.
more complicated maintenance pro- resulting heat forces the piston down and Dangerous gases are emitted from the ex-
cedures (such as carburetor overhaul) are delivers the mechanical power to the alter- haust which are highly toxic. Secondly,
not required. nator. never refuel an engine while it is running
2) Another important factor often As might be expected, proper operation or if the exhaust system is still very hot.
neglected in maintenance of alternator of the ignition system is an important fac- Unfortunately, this last precaution is
engines is oil. While lubrication is one job tor in engine performance. Power for the disregarded by many, which is extremely
oil has to perform, there are other con~ spark is supplied by a device called a foolish. (Experienced service station
siderations as well. The engine oil in the magneto that is normally installed on the operators will refuse to refuel an
cran~case also collects a large amount of front of the engine. The magneto seldom automobile with the motor running,
solid combustion products, bits of metal requires servicing and such work should . which is often prohibited by law.) Don't
worn away by the moving parts, and any only be done by those qualified to do so. become an unnecessary statistic,
dust or other foreign matter that enters (This is one reason why the magneto is 6) Most alternators are air-cooled as op-
the carburetor intake. For instance, it is often located under a flywheel that is dif- posed to the water-cooled radiator system
especially important to observe the ficult to remove by the inexperienced.) of the automobile. A fan on the front of
manufacturer's recommendations con- On the 'other hand, faulty spark plugs the engine forces air over the cylinder and
cerning the length of time the engine may ate the usual cause of ignition problems. an unobstructed entrance for this air flow
be operated before an oil change is re- Special equipment is required to test a is necessary. Avoid operating the alter-
quired during the break-in period. If you spark plug properly, but an easier solution nator in areas where obstruction to this
ever have the opportunity to examine the is to have a new one handy. In fact, keep flow might result (such as in tall grass).
oil from a new engine, you will note a two spare plugs on hand. Spark plug life Alternators should be operated such that
metallic sheen to it. This is from the ex- can be notoriously short on occasion. a sufficient amount of air circulation is
cessive 'amount of metal that is worn However, repeated plug failure is also ab- present for cooling, caburetion and ex-
away. After the break-in period, much normal and other causes such as a poor haust.
less metal is abraded and the oil doesn't gas/air mixture might be the culprit.
have to be replaced as often. Replace the spark plug with a type Storage
The oil level should be checked fre- similar t\) the one that came with the alter- Proper maintenance of an alternator
quently during engine operation. Each nator or a substitute recommended by the when it is ,not being used is just as impor-
time fuel has to be added the oil should be manufacturer. Some models have resistor- tant as during the time it is in operation.
checked also. When storing an alternator, type plugs which are desirable for The usual procedure is to run the engine
it is also wise to drain the oil and replace it ignition-noise suppression. Resistor plugs dry of gasoline, drain the crankcase and
with fresh stock. This is because one of are usually indicated by an R prefix. For fill it with fresh oil, and remove the spark
the combustion products is sulfur which instance the resistor version of a plug. Then pour a few tablespoons of oil
forms sulfuric acid with water dispersed in Champion CJ-S would be an RC-JS. into the cylinder and turn the engine over
the oil. The acid then attacks the special 5) Little maintenance is required in a few times with the starter and replace
metal in the bearing surfaces causing pit- regard to the exhaust system. In some the plug. But never crank the engine with
.ting and premature replacement. forested areas, a spark-arrester type of the plug removed and the ignition or start
Also note the grade and weight of oil muffler is 'required, so be sure that your switch in the on or run po~i,tion. The
recommended by the manufacturer. unit is so equipped before contemplating. resulting no-load high voltage might cause
Unlike their larger counterparts in the operation in such a- location. "Quiet damage to the magneto. It is also a good
automobile, mo~t small engines do hot hours" may also be imposed in some idea to ground the spark-plug wire to the
have oil filters, which is another reason places during the nighttime hours if engine frame with a clip lead in case the
why required changes are more frequent. generator exhaust noise is too loud. switch is accidentally activated.
~rrw'"
iron product - such as a generator, in base-ioading coil requires fewer turns of
storage. The coating of oil helps retard wire than one for center loading, and this
rust formation here which might actually is an electrical advantage because of
weld the two surfaces when the engine is reduced coil losses. A base-loaded anten-
restarted, resulting in premature wear. BLACK WIRE GREEN WIRE na is more stable during wind loading and
Consequendy it is important to store the sway. If a homemade antenna system is
alternator in'an area of low humidity. contemplated, either system will provide
Although the maintenance procedures BLACK WHITE good results, but the base-loaded antenna
outlined may seem like a chore, the long- GREEN may be preferred for its mechanical ad-
r---
term benefits include low repair costs and I vantages.
like-new performance. Engines for alter- I
nator combinations must be able to han- I Loading Coils
I CIRCUIT )
dle a variety of loads while maintaining a I BREAKIlR There are many commercially built
constant speed in order to keep the output I antenna systems available for mobile
frequency constant. A mechanical gover- I
I operation, and some manufacturers sell
nor performs this latter function by ~---- the coils as separate units. Air-wound
metering the fuel supplied to the engine coils of large wire diameter are excellent
under different load conditions. BLACK GREEN WHITE for use as loading inductors. Large
However, the system cannot function pro- Miniductor coils can be installed on a
perly with an engine in poor mechanical solid phenolic rod and used as loading
condition because of lack of proper coils. Miniductors, becaus.e of their turns
maintenance. Fig. 15 - A simple accessory that provides ' spacing, are 'easy to adjust when
overload protection for genEirators that do not resonating the mobile antenna, and pro-
Grounds have such provisions built in. vide excellent Q. Phenolic-impregnated
Newer generators are supplied with a paper or fabric tubing of large diameter is
three-wire outlet and the ground connec- antennas are favored by some, it is better suitable for making homemade loading
tion should go to the plug as shown in Fig. to place the antenna mount on the rear coils. It should be coated, with liquid
15. On older types, the ground would deck of the vehicle, near the rear window. fiberglass, inside and out, to make it
have .to be connected separately to the This locates the antenna high and in the weatherproof. Brass insert plugs can be
generator frame and then to the common clear, assuring less detuning of the system installed in each end, their centers drilled
terminal in the junction box. A pipe or when the antenna moves to and from the and tapped for a standard 3/8 x 24 (9.5
rod can then be driven into the ground car body. Never use a base-loaded anten- mm x 24) thread to accommodate the
and a wire connection made to either a na on a bumper mount. Many operators mobile antenna sections. After the coil
clamp supplied with the rod or by means avoid cutting holes in the car body for winding is pruned to resonance it should
of a C-cJamp for larger sizes of pipe. fear of devaluation when selling the be coated with a high-quality, low-loss
From an ignition-noise-suppression stand- automobile. Such holes are easily filled, compound to hold the turns securely in
point, the ground is desirable along with and few car dealers lower the trade-in place, and to protect the coil from the
safety considerations when power tools price because of the holes. weather. Liquid polystyrene is excellent
are, being used. The choice of base or center loading a fol' this. Hobby stores commonly stock
The ground connection goes to the mobile antenna has been a matter of< this material for use as a protective film
green wire in commercially made three- controversy for many years. In theory, the for wall plaques and other artwork.
wire conduit. Conduit purchased from an center-loaded whip presents a slightly Details for making a home-built loading
, electrical store comes with a color-coded higher base impedance than does the base- coil are given in Fig. 16.
insulation and the colored wires should be loaded antenna. However, with proper
connected as shown in Fig. 15. Consult impedance-matching techniques employed Impedance Matching
the owner's manual for the generator for there is no discernible difference in per- Fig. 17 illustrates the shunt-feed
further details on power hookup that
might apply to your particular model.
Table 2
HF MOBILE ANTENNAS
Approximate Values for 8·foot Mobile Whip
The antenna is perhaps the most impor-
tant item ,in the successful operation of a Base Loading
mobile installation. Mobile antennas, Loading RC(Q50) RC(Q300) RR Feed R' Matching
whether designed for single or multiband f(kHz) L(,..H) Ohms Ohms Ohms Ohms L(,..H)
use, should be securely mounted to the 1800 345 77 13 0.1 23 3
automobile, as far from the engine com- 3800 77 37 6.1 0,35 16 1,2
partment as possible (for reducing noise 7200 20 18 3 1,35 15 0.6
pickup), and should be carefully matched 14,200 < 4.5 7.7 1.3 5.7 12 0.28
21,250 1,25 3.4 0.5 14,8 16 0,28
to the ooaxial feed line connecting them to 29,000 36 0,23
the transmitter and receiver. All antenna
connections should be tight and weather- Canter Loading
proof. Mobile loading coils should be pro- 1800 700 158 23 0.2 34 3,7
tected from dirt, rain and snow if they are 3800 150 72 12 0.8 22 1.4
to maintain their Q and resonant frequen-
cy. The greater the Q of the loading coil,
7200
14,200
21,250
40
8.6
2.5
.
36
15
6,6
6
2,5
1.1
3
11
27
19
19
29
0.7
0.35
0.29
the better the efficiency, but the narrower RC = Loading·coil resistance; RR = Radiation resistance.
will be the bandwidth of the antenna 'Assuming loading coil Q =
300, and including estimated ground·loss resistance.
system. Suggested coil dimensions for the required loading inductance are shown in a following table,
Though bumper-mounted mobile
Chapter 10 10-10
Table 3 ;----..
Suggested Loadlng·Coll Dimensions
~ BR~AS;
PL~~W -~-
Req'd No. Wire 018. Length
-Ii)
L(,..H) Turns Size In. In.
= ,ADJUSTABLE
(
1I~::::"J..~'I1 FIELD - STRENGTH
TOP SECTION
700 190 22 3 10 NO.6 3/8" x 24 METER, SEVERAL
BRASS ~ THREADED FEET AWAY FROM
345 135 18 3 10 SCREW \ HOLE - S E T SCREW
MOBILE ANT.
150 100 16 2-112 10
/
SOLDER LUG (4) 6-32
2-1/2 AND WASHER THREADED
77 75 14 10
~HDLE
77
40
29
28
12
16
5
2-112 2
4-1/4
\
-~
PHENOLIC TUBING
40 34 12 2-1/2 4-1/4 --0
11
lIt6" OR 1/8 WALL
THICKNESS LOADING
20· 17 16 2-1/2 1-1/4 COIL
20 22 12 2-1/2 2-3/4 LI
(AI
8.6 16 14 2 2
8.6 15 12 2-1/2 3 1/4" D R A I N e
HOLE ' ~ _PLUG TO MAKE
4.5 10 14 2 1-1/4 - SNUS FIT IN TUBING
4.5 12 12 2-1/2 4
2.5 8 12 2 2 BRASS INSERT
_____ PLUG
2.5 8 6 2-3/8 4-1/2
MOISTURE
1.25 6 12 1-3/4 2 SEAL 1ffi;~~'" ___
1.25 6 6 2-3/8 4-1/2
NO.6 SCREW
To obtain dimensions in millimeters, multiply
inches by 25.4.
~..L.~I'111 ......__ WINDING
1i'R'",,------"II'IIlICOAT WITH COIL DOPE)
'"
resonance.
SECTION
/
TELESCOPING Top-Loading Capacitance
Because the coil resistance varies with
NO. 10 OR LARGER
WIRE
CAPACITY
HAT
Chapter 10 10-12
TL'~'.66 T I
T19"
/
1/16" BRASS
ROO
(A)
(8)
SEAL
WITH
EPOXY
CEMENT
ANT.---"
50-OHM
(FEET)
(e) "4
1
~
(D)
I' LINE
J A
T 1
~V
JAMES MILLEN
37001
CONNECTOR
50-OHM LINE
®
~. CAR
(ANY LENGTH)
- @,_.,----=-,ROOF
"
TO TRANS.
I
TO
75-0HM COAX
,- - - . • ---=:: TRANS,
(ANY LENGTH)
~~___ ==- -3~~--
t 36"
made from lightweight TV twin lead. The
characteristic impedance of this type of
dipole is near 300 ohms, but it can be
t~ansformed to a 50-ohm source or load
by means of a simple matching stub.
1
118 II BRAZING ROO
Fig. 26 illustrates the construction
!INCHES II. 25.4 = mm\ method and irpportant dimensions for the
6-32 THREAD, twin-lead dipole, A silver-mica capacitor
til LONG--'" is shown for the reactive element, but an Fig. 24 - Photograph Of the 220-MHz
NUT~~ 5/S·wavelength mobile antenna. The bottom
6-32
, <!?,:-NO.6 SOLDER LUG open-end stub of twin lead can serve as end of the coil is soldered to the coaxial
• I~ well, provided it is dressed at right angles connector.
NO. 6 ~ TO SOLDER LUG
LOCK WASHER ,-- to the transmission line for some distance.
The stub method has the advantage of
6- 32 THREAD.
easy adjustment of the system resonant.
T
I" DEEP PLEXIGLAS FORM
3/4" JI, 3-t/2" frequency.
NO.IO OR 12 The dimensions and capacitor values
WIRE 10-112 TURNS
COIL TAP TO F. T.
BUSHING
for twin-lead dipoles for the hf bands are
GROUND COIL END
given in Table 4. These dimensions were
'Ml
10-32 THREAD. TO COAX CABLE 28.5"
I" DEEP calculated by Jerry Hall, KITD, and Jay
CORK OR RUBBER ~O-32 It 3/4" SCREW
Rusgrove, WI VD. To preserve the
GASKET
/ (COAT WITH EPOXY GLUE balance of the feeder, a I: I balun must be
AT ASSEMBLY TIME)
NO.IO LOCK WASHER used at the end of the feed line. In most
ON BOTTOM OF BRACKET
applications the balance is not important,
(Al and the twin lead can be connected direct-
ly to a coaxial output jack, one lead to the
center contact, and one lead to the shell.
Because of its higher impedance, a folded
dipole exhibits a wider bandwidth than a
single-conductor type. The antennas
described here are not as broad as a stan-
dard folded dipole because the impedance
6-112 TURNS NO.12
transformation mechanism is frequency ENAMEL COVERED WIRE.
selective. However, the bandwidth should TAP AT 1- 3/4 TURNS
FROM BOTTOM.
be adequate. An antenna cut for 14.175
MHz, for example, will present a VSWR
of less than 2: lover the entire 20-meter Fig. 25 - Diagram of the 220·MHz mobile
band. antenna.
WHIP
50-11.
LINE
QRP Transmatch Idea' provide an inductance range of 1.9 to
TO RIG
Perhaps the most common type of cir- roughly 20IlH. The rod is a 4-inch
cuit in use by amateurs for Transmatches (102-mm) piece of 125-mu stock which
(antenna tuners) is the T -match variety of has an OD of 112 inch (13 mm). The rods
(el Fig. 27A. Under some conditions it offers are available from Amidon Associates.
good suppression of harmonics by func- (see chapter 17.)
Fig. 23 - Structural details for the 2-meter tioning as a bandpass tuned circuit. But, it If the method of Fig. 27B is adopted, a
antenna are provided at A. The mounting can also operate as a high-pass network, panel bushing should be used at the front
bracket is shown at B and the equivalent cir- in which case there is no suppression of . panel to provide support for the tuning
cuit is given at C. harmonic energy. Just how it functions shaft. This will keep the ferrite rod
will depend upon the impedance it looks centered in Ll throughout its travel. The
into. Despite this limitation it is a simple end of the drive shaft nearest the rod can
and easily adjusted network which is be epoxy-cemented to the rod. A I12-inch
SIMPLE ANTENNAS FOR HF capable of matching 50 ohms to practical- (13-mm) diameter washer can be soldered
PORTABLE OPERATION ly any impedance an antenna might pre- to the end of the 1/4-inch (6.3-mm) drive
sent. shaft before the cementing is done. This
The typical portable hf antenna is a Small components can be used for QRP will ensure adequate mating surface at the
random-length wire flung over a tree and equipment because the developed rf point where the cement is applied. The
end-fed through a Transmatch. QRP voltage and current are 10w.CI and C2 end of the rod should be . scored
Transmatches can be made quite com- can be miniature bc-band variable (checkered) by means of a metal file to aid
pact, but each additional piece of equip- capacitors for power amounts up to 25 the bond between the rod and the drive
ment necessary makes portable operation watts. A small-diameter coil can be used shaft.
less attractive. The station can be at LI. If additional inductance is needed to en- .
simplified by using resonant impedance- An experimental model of the coil ar- sure a wide range of matching conditions,
matched antennas for the bands of in- rangement of Fig. 27B was tried in the simply insert a small amount of fixed-
terest. Perhaps the simplest antenna of ARRL laboratory: This method proved to value fnductance between the ground end
this type is the half-wave dipole, center- be unique in eliminating the need for coil of Ll (Fig. 27 A) and chassis ground. An
fed with 50- ·or 75-ohm coax. Unfor- taps and a switch. The ferrite-rod slug is spst switch can be utilized to short out the
tunately, RG-58, RG-59 or RG-8 cable is pushed into Ll from the front panel to added coil when it is not needed. Similar-
Chapter 10 10·14
(DC) receivers are so simple they .can't
~I·-----------A----------~~I possibly compete with a good superhet,
right? Wrong! DC receivers have only one
significant disadvantage when compared
to superhets: the presence of an audio im-
age which doubles the amount of noise
::t:
300-0HM and interference heard.
t-
(!)
TWIN LEAD The only other inherent disadvantage is
Z
the inability to generate other than audio-
'"
...J
derived agc. The same careful attention to
>-
Z
<f detail and potential problems is required
in designing the DC receiver as is required
C ~ SILVER- MICA
s for a top-quality superhet, if comparable
CAPACITOR ACROSS, performance is to be realized. This last
FEED LINE OR STUB.
point is frequently overlooked', and that
may be one reason why the DC receiver is
often' looked upon as a mediocre per-
former.
All other problems can be overcome
with careful design, and even the two in-
50.1'\. herent disadvantages can be overcome to
some extent. On cw, narrow af filters may
Fig. 26 - A twin·lead folded dipole makes an excellent portable antenna that is easily matched be used; reducing the image bandwidth·
to 50·ohm stations. See text and Table 4 for details. along with the desired signal bandwidth.
RIT helps also: When an image signal pro-
duces the same beat note as the desired
signal, adjusting the RIT will move one up
Table 4 in pitch and the other down, thus
Twin·Lead Dipole Dimensions and Capacitor Values separating them.
Frequency Length A Length B Cs Stub Length As for agc, this rig does perfectly well
3.75 MHz 124'·9·1/2" 13'·3·1/2" 289 pF 38'·2·1/2" without. This receiver is on a par with all
7.15 MHz 65'·5·112" 6'·11·1/2" 152 pF 20'·1/2" but the best superhet for any type of
14.175 MHz 33' • 3'·6" 76 pF 10'·1" operation, except perhaps during contests
21.225 MHz 22'·1/2" 2'·4" 51 pF 6'·9"
28.5 MHz 16'·5" 1'·9" 38 pF 5'·1/2"
in conjunction with a. high-power
transmitter, but at a fraction of the com-
meters = It x 0.305
mm = in. x 25.4
plexity. Note also that to use a superhet in
a transceiver, an additional oscillator and
mixer must be added to the transmitter to
convert the VFO to the transmit frequen-
Iy, the minimum inductance of L I can be . than 20 mA; reasonable transmitter effi- cy. In a transceiver using a DC receiver the
halved by placing another 1.9-~H induc- ciency; . high-performance direct- required shift is only a few hundred hertz,
tor in parallel with L I. This can be ac- conversion receiver; and small size (\-112 and can easily be accomplished by pulling
complished by means of a switch also. X 2-112 X 3-112 in. [40 X 70 X 90 the VFO.
. In order to realize the full inductance mm}). Since the rig was designed for portable
range of the ferrite slug and L I, the rod This material has been excerpted from use, current'drain was a major considera-
travel should permit total withdrawal an August 1980 QST article by Roy tion. Experience indicates that many solid
from the coil to full entry into it. An SWR Lewallen, W7EL. Fig. 28 shows the ex- contacts may be had using simple anten-
indicator should be used with the terior of the transceiver. nas and operating during the night with 2
Transmatch to permit proper adjustment. watts on 40 meters. This power level is
Some Underlying Philo/iophy also more than adequate for short-range
AN OPTIMIZED QRP "High-performance direct-conversion daytime operation. Power drain is low-
TRANSCEIVER receiver" may seem to be self- enough that the rig will run for about a
contradictory. After all, direct-conversion week of evening operation from one
Many QRP rigs have been described in
Amateur Radio publications over the
years. The distinguishing characteristic of
this transceiver is·that it was designed and 3.5-29 MHz
WIRE ANT.
then optimized for high performance. It's 50-I\.
INPUT J2
relatively easy to build something that J1 C2
works, but it usually requires a great deal
more effort to produce something that 365
10k
-L
.----.~~~-----_o ()-----{) +VREG
ZERO
V REG
T2
II~
ANTENNA
rhO. 1
0.1
L3
--- -
L4
__
50n
• - PHASING
04~
J:, ;;: 51
*- NPO CERAMIC 05
KEYED V TO
SIDETONE OSCILLATOR
~
T4 •
VFO FROM
BUFFER •
47
011 270
KEYED V FROM Q6
390k 4.7k
390k
WOk
013
;f;O.06~
0.033
~ 0.022~
SIDETONE OSCILLATOR
further by limiting the swing of either or tically allows dense packing. It helps a
both Q 10 and U I, or biasing Uland U2A great deal to mount components on both
Fig. 29 - The circuit of the W7EL 40·meter outputs closer to the postive supply sides of the board, and to ignore the con-
ORP transceiver. Resistors are 114 or 1/8·watt, voltage. vention of placing parts in neat rows. The
5%. All ferrite cores are available from Amidon
Associates. When winding the inductors that Oscillator frequency shift is obtained by use of 1/8-watt resistors saves a surprising
use BLN·43·2402 cores, the wire should be changing the bias on the Zener diode amount of space compared to 114-watt •
passed once through both holes of the core for (used as a voltage-variable capacitor) in units. Another gr,eat space-saver is the use
each "turn" specified. See the .illustration. the VFO circuit when going from transmit of tantalum, rather than aluminum, elec-
C1 - 1,..F, 3·V non polarized ceramic.
02, 03, 05·013, incl. - Silicon general·pur· to receive or vice versa. The timing is pro- trolytic capacitors. Small parts are nearly
pose/switching dioc!e; 1N914, 1N4152 or vided by Q8, which comes on fast when always more expensive and less available
equiv. the key is closed, but goes off some five than their larger counterparts, so each
04 -.Zener, 33·V, 400·mW, 1N973 or equiv. milliseconds after the transmitter output builder must decide if the trade-off is a .
014 - Zener,10·V, 400·mW,1N961 or equiv.
L1 - Approx. 3,..H; 26 turns on a T·44-6 drops to zero following the release of the good one.
core. Tap at seven turns from ground end. key. Shaped transmitter keying is provid-
L2 - Approx. 10 ,..H, 43 turns on a T·5O·2 ed by Q7 and associated components. Adjustment and Operation
core. Simultaneous keying of the base and col-
L3, L4 - 1 ,..H; 19 turns on a T·37·6 core.
The only adjustments required are the.
L5 - 9.4 ,..H; 58 turns on a T·37-6 core. lector circuits. of the driver stage was re- VFO trimmer, used to set the VFO fre-
01,011 - Silicon n·channel JFET, 300 mW, quired to give the desired rise and fall quency at the lower band edge; the drive
2N4416. times of a few ms at the transmitter out- level pot, used to set power input at· 2
02, 03, 010 - General purpose, silicon npn, put. The sidetone oscillator is keyed from
310·mW, 2N3904. watts (although no major problem will
04 - General purpose, silicon npn, 1.8 W, the same line. arise if driven at higher or lower levels, ef-
2N2222. ficiency may drop slightly); and the
05 - Rf power, silicon npn, 7 W, 2N3553 Construction transmitter rf-amplifier tank circuit,
or 2N5859. .
06, 07 - General purpose, silcon pnp, 310
Fig. 30 shows the interior of the which is peaked at the center of the fre-
mW,2N3906. transceiver. It would be very difficult to quency range. None should require re-
08 - General purpose, npn, 310 mW, duplicate the unit (!xactly, because malJY adjustment once set.
2N4124 or 2N3565. hard-to-find miniature components are Operation is, by design, simple. The
09 - General purpose, Silicon npn, 310. mW,
2N3565.
used. Lewallen suggests starting with a only point worth noting is that, as with
RFC1 - 100·,..H subminiature choke, wound on box and building the rig into it, rather any direct-conversion receiver, signals
a 1/4·watt·resistor·sized ferrite form. Oc reo than the other way around. Get a good must be tuned on the correct side of zero
sistance is approx. 8 O. idea of the placement of controls, con- beat so that the transmitter will be on the
T1 - Primary 15 turns, secondary 3 turns.
Wound on a BLN·43 2402 core.
nectors and large components before you same frequency as the received signals.
T2 - Primary ~9 turns (approx. 6.7 ,..H), begin. Since it's difficult to troubleshoot Guest operators have picked this up 'in a
secondary 5 turns. Wound on a T·44-6 core. or modify such a rig once built, ideally a few minutes, so the SPOT button is seldom
T3, T4 - Five trifilar turns on a BLN·43·2402 larger breadboard version should first be used. When the rig is new, however, it's
core.
U1 - op amp, LM301. constructed, perfected and operated. nice to have the assurance of knowing just
U2, U3 - Oual opamp, LM358N (one section Building the circuitry on small pieces of where the transmitter will be when the key
of U3 unused). perfboard and ~ounting the boards ver- is pressed.
Chapter 10 10·18
Chapter 11
Code Transmission
RadiotelegraPh; is a popular medium To illustrate· the key-click problem we problem is a function of power supply
of amateur communications because it is used a square-wave model. In real life, the load regulation (see chapter 5). Note that
highly effective and relatively un- situation may be worse because the keying in a class C amplifier the change in power
complicated. The process by which a envelope often has a sharp spike on the output related to plate or collector voltage
radio signal is interrupted to generate a leading edge and decays exponentially regulation is proportIonal to the square of
coded message of dits and dahs is called throughout the duration of the pulse. This the voltage change (see chapter 6). The
on-off l<eying or make-break keying. On-' type of wave is even richer in harmonics power supply for a solid-state transmitter
off keying may also be used in than a square wave. The severity of the may be regulated easily and inexpensively
Radioteletype transmission~ although this
method is seldom used today. In the early
years of Amateur Radio, keying a
transmitter consisted of inserting a
telegraph key in any convenient power
lead - even the ac input! Such a
haphazard approach today is an invitation
to trouble, (rom the standpoints of signal
quality and safety. Our cw sub bands may
shrink in the future. Successful operation
under crowded conditions and adherence
to FCC regulations concerning stability
and purity of emissions requires clean
signals. In this chapter we define good
keying, analyze keying defects and show
how radiotelegraphy can be implemented (A) (8)
11-1 Chapter 11
+5-SV by electronic means, but such a scheme
applied to a high-power tube type of
transmitter is costly and dissipates con-
siderable power. Fortunately, the plate-
lOOK
voltage waveform can be corrected with a
passive circuit (see Dome, May 1977
QS1). Note also that the power source
4.7K must be nearly pure dc to ensure that the
transmitter output signal is not broadened
by hum modulation.
TV Once the power supply voltage has I;>een
HORIZONTAL
DEFLECTION brought under control, it is a simple mat-
TRANSISTOR
ter to shape the keying envelope with an
~OGIC INPUT O--+I--J\.I\I\r1o-,-+-i RC network. The figures in this section il-
1N914
lustrate the application of time-constant
circuits to various keying methods.
When a circuit carrying current is
opened or closed mechanically, a spark is
Fig. 2 - Cathode keying. Envelope shaping is accomplished by means of the RC network. 01 generated. This spark causes the circuit to
must be able to withstand the plate voltage of the keyed stage. Some suitable types are: DTS·423, radiate energy throughout the elec-
2N6457 (400V), SDT 13305 (500V), DTS·801 (800V), MJ12010 (950V), 2SC1.308K, ECG 238 (1500V). tromagnetic spectrum. When a transmit-
These are high-energy devices and are capable of switching any value of plate current the tube is
likely to draw. For plate voltages below about 350, the 2N3439 is adequate (and much less ter is keyed manually or through a relay,
expensive). the spark at the contacts can cause local
BCI, but this spark has no effect on the rf
output signal. A simple filter (O.OI-/AF
capacitor in series with 10 ohms) across
the key or relay contacts will usually
reduce the local clicks to a tolerable level.
Solid-state switching methods significant-
ly reduce the current and voltage that
must be switched mechanically, thereby
1N914
reducing local clicks and enhancing
operator safety. Modern transistorized'
transmitters incorporate this type of key-
1M
ing. With proper device selection, solid-
state keying may be implemented in older
2M
tube types of designs as well.
Amplifier tubes may be keyed in the
1M,
cathode (filament transformer center-tap
22K for directly heated types), grid-bias supply
•
BLO<;:KING
VOL TAGE or screen. Transistors should be keyed in
+ one leg of the collector supply. The low
LOGI C
INPUT ~ impedance of rf power transistor circuits
usually requires the emitter to be ground-
Fig. 3 - Blocked-grid keying. The rise time of the keying pulse is determined by C1 and its ed as directly as possible; therefore, no
aSSOCiated network. The decay time is governed by the R1C2 product. Rg is the existing grid leak. solid-state analog of cathode keying ex-
Typical values for R1 and C2 are 220 kO and 0.022 IiF. Some transistors suitable for Q1 are: ists. Similarly, blocked-grid keying has no
2N5415 (200V), MM4003 (250V), MJE350 (300V), 2N5416, RCS882 (350V), 2N6213 (375V), 2N6214 transistor equivah:;nt, because a reverse
(425V).
bias sufficient to cut the stage off in the
r----------------r~---------------__,' presence of heavy excitation would cause
+~/ KEYED STAGE breakdown of the base-emitter junction.
Mechanical contacts frequently bounce
several times before stabilizing in the
closed state. The beginnings of keying
pulses formed by bouncing contacts are
Rg poorly defined. This defect can degrade
the readability of a code signal under
adverse conditions. Relays and
semiautomatic keys are especially prone to
2 6 this malady. The circuit of Fig. 9 will help
2N3439
2N2222
clean up the pulses generated by
2N4401 mechanical contacts.
2N3904
100pF A satisfactory code signal can be
r'"
22K
LOGIC
INPUT
amplified by means of a linear amplifier
lN914 without affecting the keying characteris-
300K
-300V
tics. If, however, the signal is amplified by .
one or more nonlinear stages (e.g., a class
rh rh rh rh rh
C multiplier or amplifier), the signal'
Fig. 4 - If a suitable high-voltage pnp transistor cannot be obtained, an npn unit can be used
envelope will be modified, possibly in-
with an optical isolator. The rise time of the keying envelope is controlled by the "integrating" troducing significant key clicks. It is possi-
capacitor connected to the baseot the phototransistor. ble to compensate for this effect by using
.Code Transmission 11-2
longer-than normal rise and decay times in
+5V
the exciter and letting the amplifier
modify the signal to an acceptable one.
Any clicks generated by a linear amplifier 10K
are likely to be the result of low-frequency
parasitic oscillations.
lK
A change in frequency at the beginning
of a. keying pulse is called a chirp. If the
2N3904
oscillator isn't keyed, chirp is the result of 2N2222 3.9V NEGATIVE
changing dc operating potentials or 2N4401
L--1e--O KEY LINE
changing rf load conditions on the
oscillator. The voltage to the oscillator LOG I c o--~",,-../IN'---+-t
INPUT POSITIVE
can be regulated easily, as most transmit-
, . . - - - - - 0 KEY LINE
ters use fairly low power oscillators. If the
oscillator frequency is pulled by the 1K
loading effect of subsequent keyed stages,
. better load isolation is indicated. Chapter
6 gives a thorough treatment of buffering
techniques. __
If break-in operation is desired (see
below), it may be necessary to key the
transmitter's oscillatOr. Oscillators may Fig. 5 - Circuit to interface digital logiC with positive or negative key lines. 01 and 02 must be
be keyed by the same methods used for able to withstand the expected negative and positive keying voltages and currents.
amplifiers, but greater care is required to
obtain good results. In general, the goals
of clickless and chirpless oscillator keying
are mutually exclusive. This is because a
key-click filter will cause the operating +
voltage to be applied slowly, thereby
creating a chirp. Crystal oscillators may RF IN V - - - - I ~-......,RF OUT Vee
SUPPLY
be keyed satisfactorily if active crystals are
used. A keyed oscillator may exhibit a lOW
11-3 Chapter 11
If the transmitter oscillators run con-
RFOUTPUT tinuously, they may be audible as a
back wave between keying pulses. A
strong backwave may indicate the need
for neutralizing one or mOle transmitter .
stages .. In general, if the backwave con-
forms to the -40 dB spurious signal rule, it
won't be objectionable.
AUOI°0--:-11--.------------'
INPUT The figures in this section illustrate
+1Z-15V methods by which various solid-state and
lOOK thermionic devices may be keyed. In these
circuits, the armature of the hand key is at
ground potential and the voltages across
the key are imperceptible. The current
through the key is generally less than one
milliampere. A neon bulb with a proper
series resistor across the key will alert the
operator to junction breakdown of the
~~;~~ u---*--AJV\.,---......--t--!
high-voltage transistors. As long as safety
is given due consideration, the key-at-
ground convention need not be followed,
but this standardization is useful for
equipment interconnections. Digital con-
Fig. 8 - Keying a doubly balanced modulator in a cw-ssb transmitter. trol is shown in all of the examples. This
feature simplifies the simultaneous keying
+ VODI5-15VI of transmitter stages, T -R switches, side-
tone oscillators and muting systems. The
+VDD Ies used to perform the control functions
1M
rJ,01 are very plentiful and inexpensive. These
systems use a logic "one" to indicate a
key-down condition.
F 10'----QDEBOUNCED
OUTPUT
Break-In
C04011
Break-in (QSK) is a system of
radiotelegraph transmission in which the
2
station receiver is sensitive to other signals
fD
11
0.01 between the transmitted keying pulses.
Rl 1M 13 This capability is very important to traffic
handlers, but can be used to great advan-
tage in ragchewing as well. Break-in gives
1M
cw communication the dimension of more
natural conversation.
Fig. 9 - Debouncing circlJit for hand keys an.d relay contacts. The minimum dit length is deter· Most commercially manufactured
mined by the R1C1 product. transceivers feature a "semi break-in"
mode in which the first key closure ac-
tuates the VOX relay. The VOX controls
+Voo
AMPLIFIER
are usually adjusted to hold the relay
closed between letters. With proper VOX
adjustment, it is possible for the other
1" VDO operator to break your transmission be-
(5-15V)
tween words, but this system is a poor
SIDE TONE
MONITOR
substitute for true break-in.
Separate Antennas
~01 The simplest way to implement break-
in is to use a separate antenna for receiv-
1M ing. If the transmitter power is low (below
50 watts or so) and the isolation between
OSCILLATOR
transmitting and receiving antennas is
good, this method can be satisfactory.
51 K
Best isolation is obtained by mounting the
antennas as far apart as possible and at
51K
RECEIVER right angles. Smooth break-in involves
MUTING
1N914 protecting the receiver from permanent
12 II damage by the transmitter power and·
U1' CD4001
U2: CD4011 F
"
13
U2d
Fig. 10 - Differential (sequential) keying system for fast break·in with oscillator·multiplier
assuring that the receiver will "recover';
fast enough to be sensitive between keying
pulses. If the receiver recovers fast enough
but the transmitter clicks are bothersome
(they may be caused by receiver overload
transmitters. . and so exist only in the recfiver) their ef-
Code Transmission 11-4
feet on the operator can be minimized r----<~---~.-------<~-_o+18V
through the use of an output limiter. The
separate antenna method is most useful
on the 160-, 80- and 40-meter bands, S.2M 2,2M
where the directional effects of the anten-
nas aren't pronounced.
0.001
G2 o 40673
MPF121
*
Switching a Common Antenna Gl
When powers above about 50 watts are
used, where two antennas are not laOK 1M REC
1M
available, or when it is desired to use the
_1'"
lN914
O'mOl
1000
same antenna for transmitting and receiv- 1N914 lN914
ing (a "must" when directional antennas
are used), special treatment is required for
quiet break-in operation on the transmit-
ter frequency.
Vacuum relays or reed switches may be
used to switch the antenna between the
transmitter and receiver in step with the
keying. This method is satisfactory for
power levels up to the legal limit, but the
relays are expensive and the system timing
is critical.
Perhaps the most modern and elegant Fig. 11 - A T·R switch can be connected to the input side of the transmitter pi network. For
approach is the use of PIN diodes to powers up to about 100 W, Cl can be a 5·10 pF, loo0·V mica unit For high·power operation a
smaller "gimmick" capacitor made from a short length of coaxial cable should be used.
switch the antenna. These devices are
available in power ratings up to about 100
watts, but are quite expensive at present.
There are no keying-speed constraints
when PIN diodes are used, and if the pro- son ANT
per devices are selected, the spectral purity aXMTR~'0PF
(SHORT
. LEAD) lKV
of the output signal won't be affected.
The important electrical parameter in tkis lN9t4
regard is carrier lifetime.
An easy and economical way to imple-
ment break-in with a single antenna is to 62K
use an electronic T-R switch. With such a
device the antenna is connected to the 22K
O·5W
,.L0m
transmitter at all times. Itl the most com- LOGIC ()-<~IV'--I*-~'__..-ir-l
INPUT lN914
mon type of electronic T-R switch, a tube
is used to couple the antenna to the 22K
receiver. When the transmitter is keyed,'
the rf output causes the tube to draw grid
current through a high-resistance grid t300V
leak. The high negative bias thus
developed cuts off the plate current, -300V
limiting the signal delivered to the (MAX)
receiver. I
Unfortunately, when the grid circuit is Fig. 12 - An external T·R switch. The primary of Tl is 50 turns of no. 30 enameled wire on an
driven into rectification, harmonics are FT37·43 toroid core. The secondary is 15 turns of no. 30 wound over the primary.
generated. A commercially manufactured
low-pass filter after the T-R switch can
help to eliminate TVI caused by the har-
monics, but the lower-order harmonics prevents any grid current flow at power received signal. It is commonly stated that
may cause interference to other com- levels up to 800 wattS. This power figure electronic T-R switches are usable only
munications. Another common shortcom- assumes a 50-ohm system with a unity with transmitters having Class C output
ing of T-R switches is that the transmitter VSWR. The circuit can withstand peak rf stages because the "diode noise"
output circuit may "suck out" the re- voltages up to 300. The power capability generated by the resting current of a linear
,ceived signal. In a transmitter having a must be derated if the impedance at the amplifier will mask weak signals. Actual-
high-impedance tuned eutput tank circuit, point of connection is higher than 50 ly, the class-of-service designation is not
both of these problems can be cir- ohms. Although the signal path has a related to key-up conditions, so there is no
cumvented by connecting the T-R switch diode, it is effectively "linearized" by the reason that a linear amplifier can't be
to the input side of the tank circuit. With high-value series resistor, and should not biased off during key-up periods.
this configuration, the "grid rectification significantly degrade the spectral purity of
harmonics are suppressed and the received the transmitter output. Reduction of Receiver Gain
signal peaked' by means of the tuned cir- External T-R switches should be well During Transmission
cuit. Fig. 11 shows a MOSFET T-R switch shielded and the power leads carefully For absolutely smooth break-in opera-
that works on this principle. filtered. In general, the coaxial cable to tion with no clicks or thumps, means must
A T-R switch for use external to any the transmitter should be as short as possi- be provided for momentarily reducing the
transmitter is shown in Fig. 12. The tube is ble, but some experimenting may be gain through the receiver. A muting
grid-block keyed, and the fixed-value bias necessary to eliminate "suckout" of the function completely disables the receiver
11-5 Chapter 11
FROM
T-R SWITCH
OR .
RECEIVfNG
ANTENNA
( 50tl.)
r"
0
+1SV-
IN914
1 - - -.....- - - -.....----~--l4--..---i~ECEIVER
rle
IN914
2.2K
0.01
INPUT CIRCUIT
( 50 n )
10K
15K
IN914
OUTPUT
°1
LOGIC INPUT
6
rL°1
Fig. 15 - A 555 universal timer used as a
0.1
sidetone generator. Pin 4 is taken to ground to
interrupt the tone. The frequency of oscillation
-15 V
rL is about 500 Hz with the constants shown.
Fig. 1,3 - A diode attenuator for receiver gain reduction during keying. The logic threshold is If the audio output· isn't muted, the
determined by R1 + R2 (+ VCC). For + 1o-volt CMOS logic, R1 = R2 = 1M. For + 5-volt receiver can be used to monitor one's key-
R2
TIL operation, R1 = 130 kO, 'R2 = 15k. • ing, provided both stations are on the
same frequency. Some DX operators
transmit and listen on separate frequen-
cies. When using your receiver as a
rL°1 monitor, you should be careful about
drawing any conclusions concerning the
quality of your signal. The signal reaching
510K
the receiver must be free of any line
voltage effects induced by the transmitter.
. LOGIC o--+-+---\. 270K To be certain of your signal quality you
INPUT
TO.O I 0.01 should listen to your station from a
'..-_-..,~ TO NEXT distance. Trading stations with a nearby
750 K ,.).,
O~
STAGE
amateur is a good way to make signal
checks.
3N211
GI Keying Speeds
0.01
lit radiotelegraphy the basic code ele-
270K 270K ~0.01
ment is the dit, or unit pulse. The time
SIGNAL
INPUT
duration of a dit and a space together is
that of two unit pulses. A dah is three unit
pulses long. The space between letters is
GATE PROTECTED
three unit pulses; the space between words
or groups is seven unit pulses. A speed of
Fig. 14 - Gain-reduction circuit for recei~rs using a fixed-bias dual-gate MOSFET in the first one baud is one pulse per second.
stage. As much as 40 dB of attenuation is possible with this method. The logic threshold is
calculated as in Fig. 13. Assuming that a speed key is adjusted
to give the proper dit, space and dah
values mentioned above, the code speed
can be found from
audio during key-down periods. Assum-
ing the transmitter signal at the receiver is
monitor one's sending. A 555 timer con-
nected as an astable multivibrator is com-
Speed (wpm ) = dits/min
2.5
=
held below the damage level, muting the monly used for this purpose. This device 2.4 X dits/sec.
audio output can be an effective means of delivers rectangular output pulses, and the For example,a properly adjusted elec-
achieving smooth break-in, provided no resulting signal often sounds quite tronic key gives a string of dits which are
age is used. Age systems suitable for cw raucous. A variation of the standard 555 counted as 10 dits per second. Speed =
operation are characterized by long circuit appears in Fig.' 15. The di()des 2.4 X 10 = 24 wpm.
"hang" times. Unless the transmitter maintain the symmetry' of the waveform Many modern electronic keyers use a
signal in the receiver is at a level similar to independently of the pitch and the RC clock or pulse-generator circuit which
that of the other station, the age system filter removes many of the objectional feeds a flip-flop dit generator. For these
will seriously desense the receiver, render- harmonics. A keying monitor can be keyers the code speed may be determined
ing the break-in system useless. A diode powered by the rf output of the transmit- directly from the clock frequency
attenuator suitable for use with T-R ter. Such a circuit is shown in Fig. 16.
switches or separate antennas is shown in Keying monitors often have built-in Speed (wpm) =1.2 X clock'frequency (Hz)
Fig. 13. If the receiver uses a dual-gate loudspeakers, but it is less expensive and
MOSFET with no age in the first stage, more convenient to inject the monitor For a quick and simple means of deter-
the-method of Fig. 14 may be used. signal into the audio output stage of the mining the code speed, send a continuous
receiver. With this system one always string of dahs and count the number of
Monitoring . hears his sidetone from the same source dahs which occur in a five-second period.
If the receiver output is muted, an (speaker or headphones) as the other sta- This number, to a close approximation, is
audio sidetone oscillator must be used to tion's signal. the code speed in words per minute. A
Code Transmilssion 11-6
method for checking the speed of a Morse Speed (wpm) = (zeros/min) x 0.44.· generation. A single- or dual-lever. paddle
keyboard is to send a continuous string of may be used. If operated with a dual-lever
zeros, with proper interletter spacing. A Single Ie Keyer paddle, the 8044 is capable of iambic
Most keyboards will automatically insert Fig. 17 shows a circuit for an electronic operation (an alternating series of dits and
the correct space if the key is released and keyer built around a single 8044 CMOS in- dahs is produced when both contacts are
reactivated before the end of the tegrated circuit. Features designed into closed). A dit memory is included to pre-
character. If zeros are sent for one this IC include contact debouncing, rf im- vent loss. of dits caused by the operator
minute, the speed is approximately munity and self-completing character leading the keyer. The circuit shown in-
corporates a weight control to alter the
dit-space ratio. This feature allows partial
compensation for delays and slow rise
RF OUTPUT 1501\) times in the transmitter.
The 8044 has a built-in sidetone
generator with adjustable pitch. At the
keyer output, a high logic level represents
the key-down state. This output can
source enough· current to turn on most
npn transistors. With the timing com-
ponents shown, the speed range is about
8-50 wpm. The quiescent current drain of
this keyer is on the order of 50
microamperes, so no on.off switch is re-
quired. This keyer is ideal for mounting
inside a transceiver for station compact-
ness. The 8044 IC is marketed by Curtis
Electro Devices, Inc., Mountain View,
CA. The price range is $15.
SIDETONE
OUTPUT
An Inexpensive Buffered Morse Keyboard
270K
Until recently, the price of commercial-
ly manufactured keyboards and the cost
of many home-built units have made the
keyboard keyer an expensive station com-
ponent. But, with the explosive growth of
home computers, large supplies of surplus
Fig. 16 - An rf·powered keying monitor suitable for power outputs from about 20 to 100 W. This keyboards and digital-logic material have
circuit should be installed inside the transmitter or a shielded enclosure to minimize RFI.
been made available at reasonable prices.
The Morse keyboard described in earlier
05
Handbooks was a cost-effective keyer that
worked well but lacked some of the deluxe
features found on commercial units. This
revised version is an extension of the
previous design, with a 32-character buf-
fer memory added. An investment of
about $50 (1980 prices) can secure the
16 parts for an alert shopper. AI Helfrick,
IS
K2BLA, did the electrical ·design work.
The pc-board development, mechanical
lK
CURTIS 14
r-~~~~--~----------~~)KEYED
design and construction were performed
8044 OUTPUT in the ARRL laboratory.
U1 13
5 12 Circuit Description
R8 A 64-kHz scan oscillator drives a binary
counter. The Q outputs of the counter are
used to program the multiplexers. Each
SIDE
TONE combination of scanning lines is selected
in sequence until a key-switch closure is
detected. When the feedback loop is
closed the scan oscillator is inhibited. At
.this point the binary number represented
by the counter-output sUites is loaded into
~: .
the first-in/first-out (FIFO) registers. This
parallel data ripples down the first set of
+ -
"'" IIIIII~BTI FIFOs directly into the second set. If all
l
i<lDDLE 3 TO 9 V rh four FIFOs are full they inhibit further
data entry. The parallel data from the
EXCEPT AS INDICATED, DECIMAl,. VALUES OF
CAPACITANCE ARE IN MICROFARADS (JIF I ;
RESISTANCES ARE IN OHMS; FIFOs is then converted to serial form by
k -I 000, M'IOOO 000.
. OTHERS ARE IN PICOFARADS (pF OR JlJIFI; the two shift registers.
The seven-bit (six data plus one carry
Fig. 17 - Circuit diagram for the keyer. 01·06, incl., 1N~70 or equiv. All potentiometers are linear out) data format has 128 possible com-
taper and reSistors are 112· or 1/4·watt. binations. By connecting the key switches
11-7 Chapter 11
t'o the proper scanning lines (given in ,zero is added as an "end bit," and the ones. An optional circuit "examines" the
Table I), the combinations corresponding characters are sent fr9m right to left. All contents of the FIFOs and displays the
to the valid Morse characters are selected. of the characters must b~ represented by ,buffer statu~ on three LEQs. The com-
In the keyer, binary zero represents dit, seven-bit binary numbers where aU of the plete circuit is given in Figs. 18, 19 and 20.
and binary one represents dah. An extra unused bits to the left of the end bit are This circuit descrip'tion is necessarily
+ v 0----;---{>
4
.t" ., 6
-f'>
=1
U4 5 v
~A >
~
10. V+ 270'
16
voo 4528
RXe. DUAL ONE SHOTRST
,-TIll
p..-o·v+
i 'l
2
°i V+C III
C(1) Vss - fR tTR RST
3
INIOUT rh
1'4 'V 11 12 131
VDD o. B1
~
CO~
lID
81 ~ lID C
9 3
07 ~
B2 ~ lID 131
1::·
10
83
/1 1
lID
Ul
• 4 00
us "ST 10
.2
4051 4024
~
11 5 BINARY
B4 110
DEMUX A
•• C.oUNTER 2 4U5~~ r!!-
"ST
~ ~
9 03,
,
v•• ~' ONE SHOT VD ~
~ +TR~RX(2)
B5 110 INH
B6
t=
110 11
VEE t!-- Q2
' l'4
B7 110
VDD Vss, ~
12
. ••
ClK rL-
± -"
10'
v.
-{) V+
·~-~r
6
r; ;L'DO
V+ ~O.,
;{>
1,·
~
.,
I/O
9 6 10
AI
~ I/O
C
1.0 S
D2
10 7
0' -t>
• D' D3 Q2 11
A2 <J--~ 110
U2 A
11 4
D'
U7 .2 11 6 D2 UB QI 12
A3 ~ 110 4051 .' 40105 40105
A4 <f----!. 110
DEMUX INH
6
3
7
D3 FIFO
REGISTER .'
••
12
13
5 DI FIFD
REGISTER
Of 13
I.
INIOUT
4
D'
16
A5 <}-----J 110 COM V+ VDD
3
VEE j2- d- DO" st
A6
~ 110
°l~
15 2 16
A7
~ 110
V.. j-L- CTRl.
Vs • RIT
"3
SO
1
DI"
V..
CTRL ..
RST
Voo
fO--~r
r:1--J ~B 9" 8 9 15
V+ r;1--J
,L0"
9 3 9 15
RIT RST so
<]---11 "
O.
..
AB 110 9 4 10
C 03
A9
~ 110
• 10 5
01 '
1.0 7
d. Q2
11
AID ~ I{O U3
4.051
A 11 6
D2
U9
401.05 02 11 6
D2
UIO
401.05 01
12
All
~ 110 DEMUX
IN/OUT 3 FIFO
01 12 5
01
FIFO
o. 13
REGISTER REGISTER
AI2 <J---L 110
CON
INH
6 7
D. O. 13 4
O.
A13 <}-----2 110
L- I DOR
14 3
VEE St
<}---2
+
A14 110
16 2 16
V.. ..!........ V+rc 10
15
DI, ..nV+
A15 ~ 110
, CTRL. VSS CTAL V,S
0,1 0"1
'm~"~
10.
1-
ri
, 2 RX U6A
POWER . """ . + '
V+ -AA/I" Cx 4528 .TR
= 6-15V
10.0,.+:;, • 9NE SHOT cx
rh
Fig. 18 - Keyswitch encoder and buffer circuits for the K2BLA keyer.
U1, U2, U3 - CD4051B decoder/multiplexer. U5 - CD4024B binary counter,
U4, U6 - MC14528 dual monostable multi-vibrator. U7, U8, U9, U10 - CD40105 FIFO register.
<J
A
~D
V+
01 1
.
1~
5
VDD
U(3
4013
"D' F/!'
ClK
3
'J
C DI QI~
RaT SET VSS SET D. RST elK
41 61 71 81 91
10
1 111
-,-j,
V+ . ~. 4 UI7B 6 10 U17C 9 B
KEYIN~::;J
2N2222
0.1;[
OPTIONAL <J- ;O~~
SIDETONE
i6~ KEYING
~1\
Voo TIC rl7
~
DI
Ji!l
D. UI1 K 4 4001
'J 4035
2 I
a,
~
SHIFT I ,
D, REGISTERS
~ D4 0. fl?--
~
Qs~
r
k
RST
Vss Q 13 lUI8D)
4
Pill elK UI6D 12 UIBA 2
2hl
7 6
tOk
,
~
-
_5
U!8C 9 ~UI8B 4 71 8
~
t q 6}--
8 RST SET VSS
DUMP
° .l.-
~
U15A
VDD
%
+v V+
4027
F/F
3
UK
<J-{>c 7 6
2 J ..!..-
'~-
.
A 10 PIS elK
D. elK
'""
<1----!! D Q.
15
6 12 10
3
'" UI2
4035
U20
5 1460 -
14
~
,<]---2
16
D, SHIFT
REGISTERS
0.
13- 4- Ell rh 19 12 15
~i:: -
°4 SET RST Q
15 13 elK
1 9 °
°1
L. UI4B UI5 B
L
Vss TIC K J DI 4027 4027
9 F/F F/F
21 3 41
r
11
K
6v+
J elK SET RST J K
91
'b
16 5 I. 12 10
%=l ~
10
V+v ~ Voo VSS K
v+
O.~ UI4A 6 DOT/DASH rh
4027 J
F/F
1 2
.
° Q
---~
'i~"
9-
_5 47k 1
SPACE BAR 4- . / UI98 3 U19A
"\.. 6 4011 2
<J--o-L I~~
TOA8
D~lI~TOP 0.1
-
~,
TO 87 IN$14
Fig. 19 - Decoder and output circuits for the keyboard. The open-collector output is suitable for transmitters having positive key lines up to 40 V.
Arrangements for other key lines are given earlier in the chapter. .
U11, U12 -'CD4035B parallel shift register '
U13 - CD4013B dual D flip-flop U19 - CD4011B quad 2-input NAND gate
U14, U15 - CD4927B dual JK flip-flop 'U16, U17, U18 - CD4001B quad 2-input NOR gate U20 - MC14501 8-input NANDiAND gate
11-9 Chapter 11
Table 1 +v
Keyswitch Connections
Connect
Character From To
A 87 A10
'8 86 A1
C 86 A5
~
D 87 A1
E 87 A12
F 87 A4· 0.01 I
G 87 A1 I
86 Alb +v rh
*__
H HALF
I 87 A8 ...---------lc - - -.- - - T - - --,.
J 86 A14 . ~::- : 1
470
K 87 A5 I ,~
'L 86 A2 I I
A11 470 I I
M 87 L--"uI\A..------IH ____ - ...L _____ •
N 87 A9 Uill'IN4 FULL IN-USE
0 87 A7
P 86 A6
U7 PINI4 14f--------' +v~
Q 86 A11
R 87 A2
S 87 Alb U7 PIN 2
T 87 A13
U 87 A4
V 86 A8
W 87 A6 Fig. 20 - Optional sidetone generator and buffer status indicator.
X 86 A9 U21 - CD40018 quad 2-input NOR gate. U22 _ 88C30 dual differential line driver.
Y 86 A13
Z 86 A3
1 J 85 A14
2 85 - A12 dc voltage from 5to 15; a regulated supply recovery speed of the receiver. If the
3 85 A8 receiver agc time constant is fairly fast, it
85 Alb isn't necessary. Penlight cells will work
4
5 84 Alb fine, but if the optional circuitry is in- should be possible to hear signals between
6 84 A1 cluded, the larger C or 0 cells should be characters at keying speeds of up to 25
7 84 A3 used. wpm. If the receiver agc is turned off, or
8 84 A7 set for very fast recovery, signals can be
84 A15 Figs. 27 and 28 reveal some of the con-
9
SK 82 A8 struction details. An Apollo cabinet heard between characters at speeds of up
AR 84 A10 houses the assembly, but a home-made to 50 wpm. If you prefer, the receiver can
KN 84 A13 enclosure will work as well. The dimen- be muted during characters and two out-
AS 84 A2 sions and mounting arrangements will be puts, the + mute and - mute, are pro-
8T 85 A1
AA 86 A5 dictated by the keyboard the builder ob- vided for this purpose. Several transmit-
83 A3 tains. ters were tried with this system and it was
82 A10 possible to use them without modifica-
Operating the Keyboard tion, so long as the final amplifier was
Sending MorSe code with this keyer is . biased off under key-up conditions. No
very much like typing. Character spacing background hash was noticable.
is automatic (provided one types in step The PIN diode T-R system is not
brief. Helfrick's original QST article with or ahead of the output), and weird .plagued by problems commonly asso-
(January 1978) contains a more com- - spaces are made with the space bar. With ciated with some other systems. First,
prehensive treatment of the keyboard cir- .the buffer the operator may type up to 32 "suckout" (receiver desensing) has been
cuitry. characters ahead of the output. The unit eliminated, as has the problem of critical
has a "dump" switch, so if the operator interconnecting line lengths: Also, since
Hardware Assembly must abort a transmission, pushing the the saturated diode technique has been
The keyer circuits are assembled on two button will instantly terminate the code abandoned there should be no chance for
6 X 4-II2-inch (152 X 114-mm) double- output. Using the buffer takes a little TVI. No high-priced vacuum relays are
sided pc boards. One board holds the key- getting used to. Most operators like to used. No amplifiers are placed ahead of
switch encoder and buffer, while the other copy exactly what they send, and with the the receiver that could affect receiver per-
holds the shift registers and output de- buffer you can't do that. You must type formance. In short, the system described
coding circuitry. Figs. 21 through 24 con- and ignore the output (or disable the side- here provides excellent performance and
tain the ~tching patterns and component- tone), or you may become quite confused. suffers none of the ills of earlier systems.
placement guides for these boards. The
two boards are designed to be stacked one A PIN Diode T·R Switch The Circuit
over the other. A small third circuit board The T-R switch system described here is A schematic diagram of the circuit is
holds the optional sidetone oscillator and usable with most any 100-W (output) shown in Fig. 30. The diagram is divided
buffer indicator. This is a single-sided power level transmitter/receiver or trans- into two sections, as is the actual circuitry.
board and its etching pattern and parts ,ceiver /receiver combination. This system That portion of the circuit to the left of
layout appear in Figs. 25 and 26. The buf- is designed for flexible operation and in- the dotted line is intended to be mounted
fer indicator draws 100 times the current terconnection to various pieces of com- at the operating position for easy access.
of the main circuit; therefore a separate mercial or home-made equipment. For the Circuitry to the right of the line can be
, power switch should be used to disable the .- simplest setup, no. modifications to the mounted remotely, perhaps behind the
indicator. All of the ICs are CMOS units. transmitter or receiver are necessary. All station equipment.
The oscillator circuits using gates require that's req'uired is to plug the station equip- The transmitter is connected to the
B-series devices, but any series can be used ment into the system. The only limitation antenna through a quarter-wavelength,
elsewhere. The keyboard will work on any associated with this simple setup is the lumped-constant circuit. SI selects the ap-
- J = WIRE- JUMPER .
x = THROUGH CONNECTION
Fig. 22 - Component-side etching pattern and component-placement guide for the encoder/buffer board.
11-11 Chapter 11
o
o
o
Fig. 23 - Botto~-side etching pattern for the decoder/output board.
KEYING
OUTPUT
Fig. 24 ~ Component-side etchi~g pattern and component-placement guide for the decoder/output board.
Fig. 25 - Etching pattern for the buffer indicatorlsidetone generator board, shown from the foil
side. '
propriate circuit for the frequency in use: D2 connected in series and D3 in shunt. The station keyer (or straight key, bug
Quarter-wave circuits are required to pre- . Combination switches provide superior or keyboard) is connected to either 16 or
vent "suckout" of the received signal. attenuation performance to either the 17 depending on the output keying poten-
"Suckout" occur~ with tube-type trans- series or shunt elements alone. Approxi- tial. Q 1 and Q2 provide a suitable signal
mitters when the high-impedance end of mately 50 dB of isolation from the anten- for driving Q3through Q9. Q3 controls
the transmitter pi network becomes na to the receiver connection is provided D3 and turns the diode on during trans-
unloaded; during receive periods, for ex- throughout the rf range. These results mission and off during receive. Q4 and Q5
ample. As the pi network is one type of should be reproducible if the same type of control Dl and D2, biasing the diodes on
impedance-inverting network, the high PIN diodes are used and the board layout for receive and off for transmit. At first
resistance . presented by the non- shown is followed closely. glance it might appear that some
operational tube causes the low-
impedance end of the network to ap-
proach 0 ohms. The quarter-wavelength
lumped-constant sections provided in the
T-R system serve to. step the nearly zero
impedance level of the transmitter output
up to an almost' infinitely high impedance
that will not reduce the received signal
level. As shown in the schematic diagram,
the antenna is connected directly to! 13
which feeds the PIN diode switch section
of the T-R circuit.
The components located between 13
and J4 comprise the switch that protects
the receiver from the transmitted signal. A Fig. 29 - Exterior view of the PIN Diode T·R Switch. The box at the right is mounted at the
"T" configuration is ,used, with Dl and operating position. The box at the left can be mounted remotely.
11-13 Chapter 11
+12V
. .
1't
t80
Ct5
R2
t80
Ct6
O.t,utL O.,,u~
RFC'II RFC 211
:.HrJ-,
J' ANTENNA J3
®-£:F'
0' 02
RECEIVER
Lt
r-h' ~OCV RFC 3 3~
~T.c,
,.L Ct8
O.I)JF
S.M'r-h RFC 4
--
C20
rL°.,,uF
80
TRANS- 03 2N2222A R7
MITTER
R6
J2
StB
20~ +,2V tOk
to
'5.(.:=~". r L4
RB
,Ok
t5~TC7 t50.'T
CB
S.M~ S.M:-+,
L5
+KEY
Rt4 Rtl Q7
t5k
RFC 7
04
TI -KEY
II
RFC B
+MUTE
Rt2
C3
T '
rh O.Ot,uF tOk
---;;;v
----,
• I EXCEPT AS INDICATED, DECIMAL
TO C27 VALUES OF CAPACITANCE ARE
OW 1ft cn I IN '''CROFARADS' pF I ; OTHERS
~tOOO
rh' POWER ~ F.'T.""
I ARE IN PICOFARADS·, pF OR ),pFl; Rt3 09
REstSTANCES ARE IN OHMS~
rh I
'I---C:N
52. / . Ct2 k -1000. M.. ' 000 000 1.5k
0----0----0--
T ~
F.T.
TO C28 I
rh Ct3 I
L------____O---O- tOOO
I
r+.
TO C29
F.T. -MUTE
Fig. 30 - Schematic diagram of the T·R switch system: All resistors are 1/4-watt composition types. All capacitors are miniature ceramic, 50·volt
types unless polarization is indicated. Polarized types are aluminum electrolytic or tantalum. Component designations listed in the schematic, but
not called out in the parts list, are for text or layout reference only. '
C1, C2 - Mica, 820 pF, 500 V. . T-80-6 core.
C3, C4 - Mica, 470 pF, 500 V. choice). RFC1, RFC2 - Toroidcnoke, 20 turns no. 26
C5, C6 - Mica, 220 pF, 500 V. J6, J7 - Phone, 1/4 inch or builder's choice. enam. wire on a FT-37-75 core.
C7, C8 - Mica, 150 pF, 500 V. J8-Jll, incl. - Phono·or builder's choice. RFC3-RFC9, incl. - Toroid' choke, 26 turns
C9, Cl0 - Mica, 110 pF, 500 V. Ll - Toroid, 20 turns no. 18 enam. wire on a no. 30 enam. wire on an FT-23-75 core.
C32 - Electrolytic, 1000 JJF, 35 V. T-80-2 core. Sl - Rotary wafer, 2 sections, 5 positions,
C33 - Tantalum; 1 JJF, 35 V. L2 - Toroid, 15 turns no. 18 enam. wire on a ceramic.
01-03, incl. - PIN diode, Unitrode lN5763 or T-80-2 core. S2 - Toggle,spst.
equiv. L3 - Toroid, 11 tl,Jrns no. 18 enam. wire on a T1 - Miniature power, primary 117 V, sec·
04-09, incl. - Power, 100 PRV, 1 A. T-80-6.core. ondary 12 V at 300 mAo Radio Shack 273-
010 - Light-emitting diode. L4 - Toroid, 9 turns no. 18 enam. wire on a 1385 or equiv.
Fl - Fuse, 1/2 A. T-80-6 core. Ul - Three·terminal regulator, 12N output.
Jl-J5, incl. - Rf connector, female ~builder's L5 - Toroid, 8 turns no. 18 enam. wire on a Radio Shack RS-7812 or equiv.
Fig. 32 - Il)side view of the remotely mounted portion of the system. Short lengths of wire are
used to attach the connectors to the appropriate circuit board foils. All power supply components.
simplification of the switching diode and are mounted on the circuit board.
control circuitry might be possible. Since
it was desired to power the system from a
12-V dc source (for portable operation),
and high negative voltages could not be
used to reverse bias the diodes during
transmit, an unusual arrangement was
devised. Hence, the more complicated cir- parts-placement patterns are shown in component leads are grounded. Many ot
cuit. J8 and J9 are provided for keying the Figs. 33 and 34. An interior :view of this the component ground connections are
transmitter; one of the two outputs should unit is shown in Fig. 31. The rotary-switch not made on the pattern (bottom) side of
be suifible for almost any transmitter. JIO wafers are positioned to line up closely the board. These components must be
and Jll are for muting the station receiv.er with the appropriate circuit board connec- soldered on the top side to' complete the
during transmission. Again, dual-polarity tion points. An extra ceramic spacer is in- ,gound connection.
outputs are provided. Choose the one ap- serted in each of the switch section sup- Garden-variety components are used,
plicable to the equipment in use. port rods to provide the needed separa- with the exception of the PIN diodes. The
The power supply is depicted at the bot- tion. Short lengths of no. 18 tinned wire diodes are Unitrode IN5763 types, which
tom center of the schematic diagram. are used' for the connections. from the can be obtained from many supply
Power to the system is routed through an board to the switclr contacts. RG-58A/U houses. All of the rf chokes are hand
on/off switch that is mounted at the con- cable is used to make the connection from wound on small ferrite cores. Since encap-
trol head. An LED indicator is included as the antenna coaxial connector to the. front sulated chokes are relatively expensive and
a reminder that power is switched on. wafer. The cable braid is soldered to cores are not, the time spent winding the
. Connection from the control head to the ground lugs at each end. chokes can result in reduced cost.
remote unit is made via feedthrough The second enclosure used for the T-R
capacitors at each box. These capacitors switch system is constructed from sheet Circuit Checkout
ensure that each cabinet remains rf-tight. aluminum and measures 2-1/8 X 4-1/8 Interconnection of the two modules re- ,
The power supply is of the usual variety in X 7 inches (54 X 105 X 178 mm). An quires four lengths of hookup wire, each
all respects. A step-down transformer, ideal commercial enclosure would be the long enough to reach between the two
full-wav.e rectifier, filter and three- Bud CU-247 die-ca:st aluminum box. units when installed in their operating
terminal regulator provide the necessary These boxes are, extremely rugged and rf positions. The wires are twisted, cable-
voltage. A fuse is included in one leg of tight. The power supply, PIN-diode tied or laced together. Wires of different
the ac line. switch and control circuitry are mounted colors will help distinguish the connec-
on a circuit board. Circuit board etching tions and prevent possible suq>rises the
Construction and parts-layout information are shown in first time power is applied! Connections
As mentioned earlier, the T-R switch Figs. 33 and 34. Double-sided circuit- are as follows: Cll to C27, C12 to C28
system is constructed in two enclosures. board material was used, with the top side and C13 to C29. The fourth wire, g;ound,
The circuitry intended for mo'unting at the of the board left substantially unetched to connects the two boxes. A coaxial cable is
operating position is built into a Bud act as a ground plane. Copper must be used to connect 13 and J4.
Minibox that measures 6 X 4 X 5 inches removed from around circuit-hole loca- Connection to the station equipment is
(76 X 102 X 127 mm). The part number tions for components that are not con- a simple matter. The keyer is plugged into
of this box is CU-3005A. This enclosure nected to ground. This can be accom- either J6 or 17. If the keyer provides a
houses the circuit-board mounted quarter- plished in the etching process with the aid positive voltage when keyed, use 17. If the
wave-length sections, rotary switch, of the top-side etching pattern. Alter- keyer provides a ground, use J6. Connect
on/off power switch, LED indicator, natively, the copper can be removed from the antenna to J 1 and the station receiver
coaxial connectors and feedthrough around holes with a large drill. Do not . to J5. Do not connect the transmitter at
capacitors. The circuit board etching and remove copper from around holes where this time. A check of the system operation
11-15 Chapter 11
;::: i5 iii
5-
,~
'3 m iii
3 §:
w w 3 3 a: a:
l-
w W
::>
I-
::> >-
w >-
w >-
w >-
w
RECEIVE LINE :;: I :;: ::.:: ::.:: ::.:: ::.::
RCVR (J5) (J4)
I I + + +
0.01
TOAC
PLUG TO FEEDTHROUGH
CAPACITORS
C27·C29
Fig. 33 - Parts·layout patterns of the two printed·circuit boards. Each board is shown from the component side.
can now be made. If all is in order at this step attenuator· and received. signals. At- receiver noise when the keyer is activated.
point, signals should be heard in the tenuation should be on the order of 50 dB. Connect the transmitter output to J2 and
receiver. Actuating the keyer should cause If no measurement equipment is available, install a cable between the transmitter key
the signals to become inaudible. The exact a received signal and the receiver S-meter jack and J8 or J9. If muting of the
amount of attenuation can be measured may be used. A strong signal should receiver is desired, make the appropriate
using a calibrated signal generator or a become almost completely buried in the connection at J 10 or } I l.
\,
0 0
000 0
0 •0
0
0
0
0
00
00
• • II II 0 0
0 0
00
o 0 0 0 0 0 0 0 0 o 0 00 00
o 0
0 0 0 o 0 0 00
0
00 0 0 o 000 o 0 000 o 0 o 0
0 0 0 0 0 0 0
o 0 0
0 0 0 0 0 o 0 0
0 0 0 0 0
0 0 0
o 000 0 o 000
0 0 0 0
8
• • •••• • •
0
0
Fig. 34 - Etching patterns for the two printed-circuit boards. THe smaller board is single-sided while the large one is double-sided. Patterns are pro-
vided for both sides of the board. Patterns are to scale. Black areas represent unetched copper.
11·17 Chapter 11
Chapter 12
Single-Sideband Transmission,
the high-frequency amateur All of the intelligence is contained in the cent a-m signals. Perhaps the most impor-
bands, sing1e-sidebanq is the most widely sidebands, but two-thirds of the rf power tant advantage of eliminating the carrier is
used radiotelephony mode. Since ssb is a is in the carrier. The carrier serves only to that the overall efficiency of the transmit-
sophisticated (or simplified, depending on demodulate the signal in the receiver. If ter is increased. The power consumed by
one's point of view) form of amplitude this can~er is suppressed in the transmitter the carri~r can be put to better use in the
modulation, it is worthwhile to take a and reinserted in the proper phase in the sidebands. The power in the carrier is con-
brief look at some ,-m fundamentals. receiver, several significant communica- tinuous and an a-m transmitter requires a
Modulation is a mixing process. When rf tions advantages accrue. If the reinserted heavy-duty power supply. A dsb (double
, and af signals are combined in a standard carrier is strong compared to the incoming sideband) transmitter having the same
a-m transmitter (such as one used for double"sideband signal, exalted carrier power output as an a-m transmitter can
commercial broadcasting) four output sig- reception is achieved in which distortion Use a much lighter power supply because"
nals are generated: the original rf signal, caused by frequency-selective fading is the duty cycle of voice operation is low.
called the carrier, the original af signal, greatly reduced. A refinement of this
and two sidebands, whose frequencies are technique, called synchronous detection Balanced Modulators
the sum and difference of the original rf ,uses a phase-locked loop to enhance the The carrier can be suppressed or nearly
and af signals, and whose amplitudes are rejection of interference. Also, the lack of eliminated by using a balanced modulator
proportional to that of the original af a transmitted carrier eliminates the or an extremely sharp filter. In ssb trans-
signal. The sum component is called the heterodyne interference common to adja- mitters it is common practice to use both
upper sideband. It is erect, in that in-
creasing the frequency of the modulating
audio signal causes a corresponding in-
crease in the frequency of the rf output
•
signal. The difference component is called
the lower sideband, and is inverted, mean-
ing an increase in the modulating frequen-
cy results ina decrease in the output fre-
quency. The amplitude and frequency of
the carrier are unchanged by the modula-
tion process, and the original af signal is
rejected by the rf output network. The rf
envelope as viewed on an oscilloscope has
the shape of the modulating waveform.
Fig. IB shows the envelope of an rf
signal that is modulated 20 percent by an
af sine wave. The envelope varies in am- j
pIitude because it is the vector sum of the
carrier and the sidebands. A spectrum
analyzer or selective receiver will show the
carrier to be constant. The spectral photo-
graph also shows that the bandwidth of an '
a-m signal is twice the highest frequency
component of the modulating wave.
An amplitude-modulated signal cannot
be frequency multiplied without special'
processing because the phase/frequency
relationship of the components of the
modulating waveform would be ~everely
distorted. For this reason, once an a-m
signal has. been generated at a fixed fre-
quency, it can be moved in frequen'cy only Fig. 1 - Elecitronlc displays of a·m signals in the frequency and time domains. (A), Unmodulated
by heterodyning. ' carrier or single·tone ssb signal. (8) Full-carrier a-m signal with single-tone sinusoidal modulation.
i:'~
AUDIO AMP. AUDIO AMP.
+12V +12V
ML
WMOD. BAL. MOD.
10k 455 kHz
I-F TRANSFORMER
MILLER 2042
270k r-------,
AUDIO
I r----:I--o
INPUT
I I
rI
O.OlV
I I TO FIRST
TO I I MIXER
SIDEBAND
FILTER
O-j I i
0.1 I I
L---;h--J
(Al (Bl
BAL. MOD.
01
2N2925
820k
+9V
BAL. MOD.
220
5.55 MHz
O.Ol
01
5.5 MHz
CARRIER
INPUT
i 0.0'
§ o
5.55 MHz
FILTER
,
Tl
9MHz 1000
150
PHASE
D1
9.1V
DSB OUTPUT
(0.2V)
(el
+12V
9MHz. USB
•
45
TO 9,MHz
AMI'! OR
II '----e---<J
FILTER
Fig. 2 - Typical circuits of balanced modulators. Representative parts values are given and should serve as a basis for design.ing one's own
equipment.
SINGLE·SIDEBAND EMISSION Fig. 4 - Additional balanced·modulator circuits in which integrated circuits are used.
A further improvement in communica-
tions effectiveness can be obtained by
transmitting only one of the sidebands.
When the proper receiver bandwidth is BALANCED MODULATOR
(4 EA) HPA-5082- 5826
used, a single-sideband signal will show an
FROM
effective gain of up to 9 dB over an a-m CARRIER T 2 ( ' OUTPUT
DSS
signal of the same peak power. Because GENERATOR
the redundant information is eliminated,
the required bandwidth of an ssb signal is
half that of a comparable a-m or dsb emis-
sion. Unlike dsb, the phase of the local
carrier generated in' the receiver is unim-
portant.
~.01
Generating the SSB Signal: Filter Method
If the dsb signal from the balanced
modulator is applied.to a bandpass filter, Fig. 5 - Balanced·modulator design using hot·carrier diodes.
one of the sidebands can be greatly at-
tenuated. Because a filter cannot have in-
: finitely steep skirts, to obtain adequate
suppression of the unwanted sideband the Table 3
response of the filter must begin to roll off Guidelines for Amateur,SSB Signal Quality
, within about 300 Hz of the phantom car- Parameter Suggested Standard
rier. This effect limits the ability to
Carrier suppressjon At least 40-dB below PEP.
transmit bass frequencies, but as will be Opposite·sideband suppression At least 40·dB below PEP.
. shown in the section on speech prbcessing, Hum and noise At least 40·dB below PEP.
these frequencies have little communica- Third·order intermodulation distortion At least 30·dB below PEP.
tions value. The filter rolloff can be used Higher·order intermodulation distortion. At least 35-dB below PEP.
Long·term frequency stability At most 100·Hz drift per hour.
to obtain an additional 20 dB of carrier Short·term frequency stabil,ity At most 10·Hz pk·pk deviation in
suppression. The bandwidth of an ssb fil- a 2·kHz bandwidth.
ter is selected for the specific application.
For voice communications, typical values
are 1800 to 4000 Hz. image rejection. This consideration re- tions. High quality components and
Fig. 6 illustrates two variations of the quires the heterodyne-oscillator frequency careful adjustment are required for good
filter. method of ssb generation. The to be above the fixed ssb frequency on results with this type of filter. An alternate
heterodyne oscillator is represented as a some. bands and below it on others. To possibility is a "synthesized" filter com-
simple VFO, but may be a premixing sys- reduce circuit complexity, early amateur prised of high-performance operational
tem or synthesizer. The scheme at B is per- filter types of ssb transmitters did not in- amplifiers used as gyrators or "active in-
haps less expensive tlian that of A, but the clude a sideband selection switch. The ductors." A further drawback of ssb
heterodyne oscillator frequency must be result was that the output was Isbon 160, generation in this frequency range is that
shifted when changing sidebands if the 75 and 40 meters, and usb on the higher multiple conversion is necessary to reach
dial calibration is to be. maintained. The bands. This convention persists despite the desired output frequency with ade-
ultimate sense' (erect or inverted) of the the flexibility of most modern amateur ssb quate suppression of spurious mixing pro-
final output signal is influenced as much equipment. ducts.
by the relationship of the heterodyne~ Mechanical filters are an excellent
oscillator frequency to the fixed ssb fre- Filter Types choice for ssb generation in the 400- to
quency as by the filter or carrier frequency For' carrier frequencies in the 50- to 500-kHz region.' These filters are de-
selection. The .heterodyne-oscillator fre- l00-kHz region, a satisfactory filter can scribed in some detail in the receiving
quency must be chosen to allow the, best be made up of lumped-constant LC Sec- chapter. For wide dynamic range receiving
12·4 Chapter 12
applicatons, the more modern types using
piezoelectric transducers are preferred for
BALANCED
lowest interlllodulation distortion_ In
transmitters, where the 'signal levels can be
closely predicted, the types using
magnetostrictive transducers are entirely
FILTER suitable.
Quartz crystal filters are commonly
used in $ystems in which the ssb signal is
CARRIER generated in the' high-frequency range.
OSC. Some successful amateur designs have
also employed crystals at 455 kHz.
Generally, four or more crystal elements
(A)
are required to obtain adequate selectivity
for ssb transmission. Crystal-filter design
BALANCED
is a sophisticated subject, and the more
esoteric aspects are beyond the scope of
this Handbook. The discussion of
piezoelectric crystal theory in the Elec-
trical Laws and Circuits chapter is suffi-
cient background material for the general
understanding of the concepts outlined in
CARRIER this section.
OSC. A fundamental crystal filter section is
the half-lattice, shown in Fig. 7. The pass-
band of this type of filter is slightly wider
than the frequency spacing between the
crystals. The antiresonant (parallel
(B)
resonant) frequency or pole of the low-
frequency crystal must be equal to the
series-resonant frequency or zero of the
I high-frequency crystal. Such a filter is
,Fig. 6 - The filter method of ssb generation. Two sideband selection schemes are commonly
used. useful for casual receiving purposes, but
the ultimate stopband attenuation is poor,
and numerous spurious responses will
A SERIES PARALLEL exist just outside t~ passband. Cascading
o f----e_--o
0
I
I
CRYSTAL A
CRYSTAL B
~
(\T+ *:)
:
/ / FREQUENC;Y
FREQUENC.Y
two of these sections back-to-back, as in
Fig. 8, will greatly suppress these parasitic
_L resonances and steepen the skirts without
I I I
'r I I I materially affecting the passband. An im-
I
portant factor ih the design of this type of
filter is the coefficient of coupling be-
tween the two halves of the transformer.
The coupling must approach unity for
proper operation. A twisted-pair or bifilar
winding on a high-permeability ferrite
core most nearly approximates this ideal.
Fig. 7 - The half·lattice crystal filter. Crystals A and B should be chosen so that the parallel· • Some crystal filters have tuned input and
resonant frequency of one Is.the same as the series·resonant frequency of the other. Very tight
coupling between the two halves of the secondary of T1, Is required for optimum results. The
output transformers. The flatness of the
theoretical attenuation·vs.·frequency curve of a half-lattice filter shows a flat passband between passband is heavily dependent on the ter-
the lower series-resonant frequency and higher parallel-resonant frequency of the pair of crystals. minating resistances. Lattice filters exhibit
fairly symmetrical response curves and
can be used for Isb or usb selection by
A B' means of placing the carrier frequency on
the upper or lower skirt.
An asymmetrical filter is shown in Fig ...
9. Good unwanted sideband suppression
can be obtained with only two crystals
using this approach. The crystals are
ground for the same frequency. The
potential bandwidth here is only half that
obtained with a half-lattice design. The
maximum bandwidth of almost any
B A' crystal filter can be 'increased by using
o I-+---+-.----i 01-----' plated crystals intended for overtone
operation. ,
The home c()nstruction of crystal filters
can be very time-consuming, if not expen-
Fig. 8 - Half-lattice filters cascaded In a back-to-back arrangement. The theoretical curve of such sive. The reason for this is that one must
a filter has increased skirt selectivity and fewer spurious responses, as compared with a simple experiment with a large number of crystals
half'lattlce, but the same passband as the simple circuit. to produce a filter with satisfactory per-
INPUT
FREQUENCY
70
80
9673 74
.r 75 76 77 78 79 9680 81 82 83
FREQUENCY (kHz)
Fig, 9 - An a$ymmetrical filter and theoretical Fig, 10 - Some ladder filters based on CB crystals, with the response that can be expectetl from
attenuation curve, the 6·pole unit .
formance. Crystal grinding and etching very little effect on the 3-dB bandwidth.
can be a fascinating and highly educa- Ladder filters having six or more elements reble 4
tional activity, but most'home builders are suitable for ssb· transmitting and CB Frequencies
would prefer to spend their time on other receiving' service. In' general, the band- Frequency Frequency
aspects of equipment design. High-quality width is inversely proportional to the Channel (MHz) Channel (MHz)
filters are available from several manufac- values of the shunt capacitors and directly 1 26.965 21 27.215
turers in the $50 to $100 price class. Most proportional to the term:inal impedances. 2 26.975 22 27.225
amateurs who build their own ssb equip- Table 4 lists the frequencies of the CB 3 26.985 23 27.255
thent adopt a "systems engineering" ap- channels. Overtone crystals for CB service 4 27.005 24 27.235
5 27.015 25 27.245
proach and design their circuits 'around have fundamental resonance at approxi- 6 27.025 26 27.265
filters of known performance. Some filter mately one third of the listed fr~quency. 7 27.035 27 27.275
suppliers are listed in the Construction 8 27.055 28 27.285
and Data Tables chapter. It is still worth- Filter Applications 9 27.065 29 27.2!!5
10 27.075 30 27.305
while to have an appreciation for the basic The important consi'derations in circuits 11 27.085 31 27.315
design ideas, ,however, for many of the using bandpass filters' are impedance 12 27.105 32 27.325
less expensive filters can be improved matching and input/output isolation. The 13 27.115 33 27.335
markedly by the addition of a couple of requirements for the latter parameter are 14 27.125 34 27.345
15 27.135 35. 27.355
'\lcrystals external to the package. The less severe in transmitting applications 16 27.155 36 .27.365
technique is to steepen the skirts by group- than they are for receiving, but with prop- 17 27.165 37 27.375
ing sharp notches on either side of the er layout and grounding, the, opposite 18 27.175 38 27.385
passband. sideband suppression should be deter- 19 27.185 39 27.395
20 27.205 40 27.405
An important exception to the above mined by the shape factor rather than
commentary is the ladder filter. Although signal leakage~ The filter must be ter-
this type of filter is treated in textbooks, it minated with the proper impedances to
has received attention in the amateur ensure a smooth bandpass response.
literature only recently .. The significant Fig. llA shows a typical ssb generator transformer of the modulator. The tank
feature of ladder filters is that all of the using a KVG (see'QST ads) crystal filter. circuit is broadly resonant at 9 MHz and
crystals are ground for the same fre~ The grounded-gate JFET presents a rejects any spurious signals generated in
quency., Low-cost CB crystals are ideally broadband 50-ohm termination to the the modulator that might be propagated
suited to this application. Representative balanced modulator and transforms the throu,gh the filter. Crystal filters should be
designs by F6BQP and G3JIR are given in impedance to the 500 ohms required by isolated from any dc voltages present in
Fig. 10. Filter sections of this type can be the filter. The dc return for the source of the circuit.
cascaded for improved shape factor with the' JFET is through the output A circuit using a Collins mechanical
12·6 Chapter 12 6·
r---------4~_o +'2V
0.0'
DOUBLY BALANCEO
~
MODULATOR 2N4857 9MHz
0.01
0.01
(-----<:) Z.'::¥5 O..n..
AUDIO
INPUT
+12V
(A)
+12V
+12V
4.7 k ,.+,0.1
'100
...----4~_t ~O '" 3 k
1llL-t-rI~TAGE~1
6 F400FD - 25
r-+-1 L..----.oo'
TRANSFORMER
4!>5 kHz
(8)
CARRIER
GENERATOR
Fig. 11 - Connecting a packaged filter into an ssb generator. (A) 9-MHz crystal filter. (B) 455-kHz mechanical filter.
filter is illustrated in Fig. liB. The i-f The phasing method was used in many
pre-I960 amateur ssb exciters, but became Tlble 5
transformer prevents spurious responses
and removes dc bias. The output ter- less populat after the introduction of uriwlntedSldeblnd Suppresllon
minating network does double duty as the relatively inexpensive high-performance II I Function of. Phlse Error
bias network for the transistor amplifier bandpass filters. The phase shift and Phase Error (degrees) Suppression (dB)
stage. The filter output transformer is the amplitude balance of the two channels 0.125 59.25
dc return for the base circuit, This tech- must be very accurate if the unwanted 0.25 53.24
sideband is to be adequately attenuated. 0.5 41.16
nique is legitimate so long as the current is 1.0 41.11
limited to 2 rnA. Table 5 shows the required phase accuracy 2.0. 35.01
of one channel (af or rt) for various levels 3.0 31.42
SSB Generation: The Phasing Method of opposite sideband suppression. The 4.0 28.85
Fig. 12 shows another method ·for ob- numbers given assume perfect amplitude 5.0 26.85
10.0 20.50
taining a single-sideband signal. The balance and phase accuracy in the other 15.0 16.69
audio and carrier signals are each split in- channel. It can be seen from the table that 20.0 13.93
0
to components separated 90 in phase and a phase accuracy of ± 10 must be main- 30.0 9.98
applied to balanced modulators. When tained if the signlll quality is to satisfy the 45.0 6.0
the dsb outputs of the modulators are criteria tabulated at the beginning of this
combined, one sideband is reinforced and chapter. It is difficult to achieve this level
the other is cancelled. The figure shows of overall accuracy over the entire speech 0
± 2 if the peak deviations can be made
sideband selection by means of transpos- band. Note, however, that speech has a to occur within the spectral gap.
ing the audio leads,but the same result complex spectrum with a large gap in the The major advantage of the phasing
can be had by means of switching the car- octave from 700 to 1400 Hz. The phase- system is that the ssb signal can be
rier leads. accuracy tolerance can be loosened to generated at the operating frequency
BALANCED MODULATOR
204
/'
AF INPUT
300- ~
II ._----0 + 12 V
3000 Hz o------J
Tt 6
SEC:600.n.
Fig. 13 - A circuit using the B&W 2Q4 audio phase shift network.
lOOk
>--<~-o TO BAL.
MOD NO.t
LOW- Z
INPUT
tOOk
tOOk
>-<~--{) TO BAL.
tOOk
MOD N~. 2
C1 : O.044)JF (2 x O.022)JF)
C2 : O.033}JF
C3: 0.02jJF Ul, U2 : HIGH PERFORMANCE
C4: O.OljJF OPERATIONAL AMPLIFIER
C5: 5600pF
C6: 4700pF
Fig. 14 - A high-performance audio phase shifter made from ordinary loose-tolerance components.
12·8 Chapter 12
CARRI£R f -ro 8Al
fNPUT MOD NO 1
Ie 0----_
RVC - : ' TOBALMOD
NO.1
C .. R Xe=RATfc
TO BAl:
' - - - - - - - - - 0 ~g ~AL MOD
MOD NO.2
~D6IY
tolerance components, this phase shifter
provides approximately 60 dB of opposite
sideband attenuation over the range of
300 to 3000 Hz. DELAY
LINE
Numerous circuits have been developed (S) (Al
to synthesize the required 90° phase shift
electronically. Active-filter techniques are
used in most of these systems, but preci-
Fig. 17 '7 (A) A digital rf phase shift network. The bandwidth of this system is limited only by the
sion components are needed for good capabilities of the ICs. The RC network connected to the set lines of the flip·flops ensures a con·
results. An interesting phasing system sistent phase relationship every time the circuit is energized. (B) The required frequency ,
-described in Electronics for April 13, multiplication in (A) can be obtained from this exclusive OR gate circuit. Additional gates can be
1978, makes use of a tapped analog delay used for the delay lines, The more conventional push·push multiplier circuits can also be used.
line. These "bucket brigade'" devices are
becoming available at reasonable prices quency, problems similar to those in the small, such as in the 144.1- to 145-MHz
on the surplus market. audio networks must be overcome. range, the rf phase shift can be obtained
A differential rf phase shifter is shown conveniently by means of transmission-
RF Phasing Networks in Fig. 15. The amplitudes of the quadra- line methods. If one balanced modulator
If the ssb signal is to be generated at a ture signals won't be equal over an entire fee,d line is made an electrical quarter
fixed frequency, the rf phasing problem is phone band, but this is of little conse- wavelength longer t,han the ,other, the two
trivial; any method that produces the quence as long as the signals are strong signals will be 90"0 out of p'hase. It is im-
proper phase shift can be used. If the . enough to saturate the modulators. portant that the cables be properly ter-
signal is produced at the operating fre-, Where percentage bandwidths are minated.
(A)
AUDIO COMBINER
INFORMATION e
INPUT
8+90,°
.--+---'\
INFORMATION
INPUT
(B)
Fig. 18 - Independent-sideband generators. (A) Filter system. (8) Phasing system. The block marked "rf combiner" can be a hybrid combiner or a
summing amplifier. '
One method for obtaining a 90° phase microphone input, F5 emission will result. Many cominercially manufactured ssb
shift over a wide bandwidth is to generate Two alternating tones from an afsk RTTY transceivers have provisions for control-
the quadrature signals at a fixed frequency keyer will cause the transmitter to produce ling the transmit or rec~ive frequency with
and heterodyne them individually to any an FI signal. A keyed audio tone will be an external VFO or receiver. With slight
desired operating frequency. Q!ladrature translated into an Al signal. This tech- modification it should be possible to slave
hybrids having multioctave bandwidths nique is a perfectly legitimate way to two transceivers to a single VFO for isb
are manufactured commercially, but they operate cw·with an ssb transceiver, and is operation. The oscillators in the trans-
cost hundreds of dollars. Another prac- simpler than the more traditional method ceivers must be aligned precisely.
tical approach is to use two VFOsin a of upsetting the balanced modulator for The most obvious amateur application .
master"slave phase-locked loop system. carrier leakage. One can vary the for independent sideband is the transmis-
Many phase detectors lock the two signals transmitting frequency independently of sion of slow-scan television with
in phase quadrature. A doubly balanced the receiving frequency by means of simultaneous audio commentary. On the
mixer has this property. One usually changing the audio tone. The strength of vhf bands, other combinations. are possi-
thinks of a phase-locked loop as having a ' the tone determines the transmitter power ble, such as voice and code or SSTV and
, VCO locked to a reference signal, but a output. Good engineering practice re- RTTY.
phase differential can be controlled in- quires that the tone be frequency-stable
dependently of the oscillator. The circuit and that the total harmonic distortion be The Speech Amplifier
in Fig. i6illustrates this principle. A less than one percent. Also, the carrier The purpose of a speech amplifier is to
digital phase shifter is sketched in Fig. 17. and opposite sideband must be suppressed raise the level of audio output from a
If ECL ICs are used, this system can work at least 40 dB. Of course the rise and microphone to that required by the modu-
over the entire hf spectrum. decay times of the audio envelope must be lator of a transmitter. In ssb and fm trans-
controlled to avoid key clicks. This sub- mitters the modulation process takes place
Other SSB Modes ject is treated in detail in Chapter 11. at low levels, so only a few volts of audio
An., ssb transmitter is simply a frequency are necessary: One or two simple voltage-
translator. Any frequency- or amplitude- Independent Sideband Emission amplifier stages will suffice. A-m
varying signal (within .the bandwidth If two ssb exciters, one 'usb and the transmitters often employ high-level plate
capabflities of the transmitter) applied to other Isb, share a common .carrier modulation requiring considerable audio
the input will be translated intact oscillator, tWo channels of information power, compared to ssb and fm. The
(although frequency inversion takes place can be transmitted from one antenna. microphone-input and audio voltage-
in Isb) to the chosen radio frequency. If Methods for isb generation in filter and amplifier circuits are similar in all three
, amplitude-limited tones corresponding to phasing transmitters are shown in Fig. 18. types of phone transmitters, however:
the video information of a slow-scan May 1977 QST carried an article on con- When designing speech equipment it is
television picture are fed- into the verting the popular Drake TR4-C to isb. necessary to know (1) the amount of
12·10 Chapter 12
peaks (sharp rises in the reproduction' is suspended in a magnetic circuit. When'
MPSA10
curve) limit the swing' or modulation to sound impinges on the diaphragm, it
the maximum drive voltage, whereas the moves the coil through the magnetic field,
usable energy is contained in the flat part generating an alternating voltage.
of the curve. A microphone must be ter-
4700 minated in its specified load impedance if Electret Microphones
the designed frequency response is to be The electret microphone has recently
realized. . appeared as a feasible alternative to the
Microphones are generally omnidirec- carbon, piezoelectric or dynamic micro-
.....JVV'v~_-----,-+-<:J+12V
tional, arid respond· to sound from all phone. An electret is an insulator which
directions, or unidrectional, picking up has a quasi-permanent static' electric
CARBON
sound from one direction. If a miCro- charge trapped in or upon it. The electret,
MPF102
phone is to be used close to the operator's operates in a condenser fashion which
mouth, an omnidirectional microphone is uses a set of biased 'plates whose motion;
ideal. If, however, speech is generated a caused by air pressure variations, creates a
foot (0.31 m) or more from the changing capacitance and an' accompany-
microphone, a unidirectional microphone ing change in voltage. The electret acts as
will reduce reverberation by a factor of the plates would, and being charged, tt re-
1. 7: 1. Some types of unidirectional quires no bias voltage. A low voltage pro-
CRYSTAL,CERAMIC, OR HI-Z DYNAMIC
microphones have a proximity effect in vided by a battery used for an FET im-
that low frequencies are accentuated when pedance converter is the only power re-
the microphone is too close to the mouth. quired to produce an audio signaL
OUTPUT
Electrets traditionally have been suscep-
~'---+-i+ f-o Carbon Microphones tible to damage from high temperatures
~F The carbon microphone consists of a
met~l diaphragm placed against a cup of
and high humidity. New materials and dif-
ferent charging techniques have lowered
3900 loosely packed carbon granules. As the the chances of damage, however. Only in
diaphragm is actuated by the sound extreme conditions (such as 120°F or
pressure, it alternately compresses and 49°C at 90 percent humidity) are prob- .
~----_>--O+12V
decompresses the granules. When current lems present. The output level of a typical
is flowing through the button, a variable electret is higher than that of a standard
dc will correspond to the movement of the dynamic microphone. .
Fig. 19 .2 Speech circuits for use with diaphragm. This fluctuating dc can be
standard·type microphones. Typical parts used to provide voltage corresponding to Microphone Amplifiers
values are given. the sound pressure. The output of a car- The circuit immediately following the
bon microphone is extremely high, but audio input establishes the signal-to-noise
nonlinear distortion and instability has ratio of the transmitter. General-purpose
reduced its use. lCs such as the 709 and 741 op amps are
widely used in speech amplifiers, but they
audio power the modulation system must Piezoelectric Microphones are fairly noisy, so it is best to precede
furnish, and (2) the output voltage Piezoelectric microphones make use of them with a lower-noise discrete device
developed by the microphone when it is the phenomenon by which certain (FET or bipolar transistor). The circuits in
spoken into from normal distance (a few materials produce a voltage by mechanical Fig. 19 fulfill this requirement.
inches) with ordinary loudness. It then stress or distortion of the materiaL A
becomes possible to choose the number diaphragm is coupled to a small bar of Voltage Amplifiers
and type of amplifier stages needed to material such as Rochelle salt or ceramic The important characteristics of a
generate the required audio power made of barium titanate or lead zirconium voltage amplifier are its voltage gain,
without overloading or undue distortion titanate. The diaphragm motion is thus maximum undistorted output voltage,
anywhere in the system. translated into electrical energy. Rochelle- and its frequency response. The voltage
salt crystals are. susceptible to high gain is the voltage-amplification ratio of
Microphones temperatures, excessive moisture, or ex- the stage. The output voltage is the maxi-
The level of a microphone is its elec- treme dryness. Although the output level mum af voltage that can be secured from
trical output for a given sound intensity. is higher, their use is declining· because of the· stage without distortion. The
The level varies somewhat with the type. their fragility. amplifier frequency response should be
It depends to a large extent on the distance Ceramic microphones are impervious to adequate- for voice reproduction; this re-
from the sound source and the intensity of temperature and humidity. The output quirement is easily satisfied.
the speaker's voice. Most commercial level is adequate for most modern ampli- . The voltage gain and maximum un-
transmitters are designed for the median fiers. They are capacitive devices and the distorted output voltage depend on the
level: If a high-level mic is used, care output impedance is high. The load im- . operating conditions of the amplifier. The
should be taken not to overload the input pedance will affect the low frequencies. output voltage is in terms of peak voltage
amplifier stage. Conversely; a microphone To provide attenuation, it is desirable to rather than rms; this makes the rating in-
of too low a level must be boosted by a reduce the load to 0.25 megohm or even dependent of the waveform. Exceeding
preamplifier. lower, to maximize performance when the peak value causes the amplifier to
The frequency response (fidelity) of a operating ssb, thus eliminating much of distort, so it is more useful to consider
microphone is its ability to convert sound the unwanted low-frequency response. only peak values in working with
uniformly into alternating current. For amplifiers.
high articulation it is desirable to Dynamic MIcrophones A circuit suitable for use as a
reproduce a frequency range of 200-3.500 The dynamic microphone somewhat microphone preamplifier or the major
Hz. When all frequencies are reproduced resembles a dynamic loudspeaker. A light- gain block of a speech system is shown in
equally, the microphone is considered weight coil, usually made of aluminum Fig. 20. The response rolls off below 200
"flat." Flat response is highly desirable as wire, is attached to a diaphragm. This coil Hz to reduce hum pickup. Ordinary 741
~~.
~~~.-.t
INPUT 1).1f
15V
t
180'
IK t-__+:-;1 0
-+12 V (Al
0.1
-------f-
(Al 10K
--r I
10K
,, ' J , AV. PEAl(
ENV,
INPUT
0---)
0.1 10 K 2
3
1 uf
-=--=--=-_-__ 1_
1
0.1
1
180·
Fig. 24- (A) Typical ssb voice'modulated
signal might have an envelope of the general
nature shown, where the rf amplitude (current
or voltage) is plotted as a function of time,
which increases to the right horizontally. (8)
1 uf
~----4..---l t - - - -....O
1 Envelope pattern after speech processing to in·
crease the average level of power output.
J,wv
varying. The power input peak-to-average 470pF
ratio, as the power-output peak-to- 0.002
average ratio, depends on the voice 4.7pF
characteristics. Determination of the in- MIC 15V
put ratio is further complicated by the fact 3 OUTPUT
~-------------1~~~---+--~n BOrnV
that there is a resting value of dc plate in-
put even when there is no rf output. No
AGe
exaet figures are possible. However, ex-
,.l,0.47 GENERATOR
perience has shown that for many types of 4
voices and for ordinary tube operating
conditions where a moderate value of U2 5
reSting current is used, the ratio of PEP SL620C
220
input to average input (during a modula-
tion peak) will be in the neighborhood of +
47pF
2: 1. That is why many amplifiers are rated
for a PEP input of' 2 kilowatts even ,.L 15V
'I
01 02
• MPSAlO MPSA10
~_-+--+-0ClIPPED AF
;.:i:,' OUTPUT
(Al
Q1 02
2N3391A 2N2925
Fil TER
~_ _-_-f-'--;'..--c:)CLI P P ED A F
~' OUTPUT
+12V . (81
Fig, 26 - This drawing illustrates use of JFETs or silicon diodes to clip positive and negative voice peaks.
tors and four diodes are used in U2. this purpose should have relatively little
The compressor will hold the output attenuation below about 2500 Hz, but
level constant within 2 dB over a 4O-dB high attenuation for all frequencies above o
range of input signal. The nominal output 3000 Hz. ~ ~ 8t-------1---+---+-?<----I
level is 80 mY; the microphone used The values of Land C should be chosen ~o:
should develop at least 3 mV at the gate of to form a low-pass filter section having a ~~ 6't-------1---+-7"---+-----I
Ql. ' cutoff frequency of about 2500 Hz, using ~o
Q.T 4
Fig. 25B shows an IC audio compressor the value of the terminating resistor load ~o
I-
circuit using the National Semiconductor resistance. For this cutoff frequency the "'I
~~ 2f----~~~~~~~~~~__I
LM-370. This IC has two gain-control formulas are '
points, pins 3 and 4; one is used for the in- ~ 0 ~~:d~~=:i~~RF~C~O~M~PR~ES~S~IO~N
put gain adjustment while the other Ll, = 7:S0 Cl = C2 = 6i,6 <J) 0 5 10 15
dB OF PEAK COMPRESSION
20
FINAL AMP.
6146 (2 EA.)
C3
CONTROL
----<p-*,
VO LTAGE O--.....
OUTPUT
0.01
LEVEL
(A) SET R1
1M
(8)
CONTROL
+ 40V VOLTAGE
OUTPUT
+200V
ALC
.....-H---<P-l4----....---·-QOUTPUT
22"
;ho5
(e)
ALC
0-50 dB GAIN REDUCTION
FOR 0-20V ALC APPLIED
AMP.
I NPU To-""Ho::-<Ib--._-'-f"o....
30
>'!..-_--~~""OOUTPUT
(E)
Fig. 29 - (A) Control voltage obtained by sampling the rf output voltage of the final amplifier. The diode back bias, 40 volts or so maximum, may be
taken from any convenient positive voltage source in the. transmitter. R may be a linear control having· a maximum resistance of the order of 50 kn.
01 may be a 1N34A or similar germanium diode.
(B) Control voltage obtained from grid circuit of a Class AB1 tetrode amplifier. T1 is an Interstage audio transformer having a turns ratio, secon·
dary to primary, of 2 or 3 to 1. An inexpensive transformer may be used, since the primary and secondary currents are negligible. 01 may be a 1N34A
or similar; time constant R2C3 is discussed in the text. -
(C) Control voltage is obtained from the grid of a Class AB1 tetrode amplifier and amplified by a triode audio stage.
(0) Ale system used in the Collins 328·3 transmitter.
(E) Applying control voltage to the tub~ or (F) linear IC·controlied amplifier.
12·16 Chapter 12
ISOLATION AMP. I-F AMP.
TO
r-------------_---lt---oTRANS.
50 MIXER
,- ----------,
I
3.395 MHz
I
I TO
I RCVR.
I SECOND I-F
I
I I
L ______ ~~--.J
22k
~D.DD5
1W +300V
100k
TO
TO AVC
ALC
MECHANICAL
I-F FILTER
I-F AMP.
,---------,
TRANS FOR MER
:TI' IL ____ _
+10V (REC.)
MIXER
r-----,
1 :
'r
1 1
]I ] 1
if I I
2N2672 2N2672
I .1
I 1
I I 0.01
I I
L___ -1
GROUND
( TRANS,)
+1DV (REC.)
+10V (el
t
Fig. 30 - Transceiver circuits where a section is made to operate on both transmit and receive. See text for details.
speech, but such a compressor added little ping. The effect of such clipping on a two- dB is typical. An ale circuit with shorter
to the intelligibilty' threshold at the tone test pattern is shown in Fig. 28. time constants will function as an rf
receiver, only about 1-2 dB. Automatic level control, although a syllabic compressor, producing up to 6 db
Evaluation of rf clipping from the form of rf speech processing, has found improvement in the intelligibility thresh-
receive side with constant4evel speech, its primary application in maintaining the old at a distant receiver. The Collins
and filtering to restore the original band- peak rf output of an ssb transmitter at a Radio Company uses an ale system with
width, resulted in an improved in- relatively constant level, hopefully below dual time constants (Fig. 29D) in their
telligibility threshold of 4.5 dB with 10 dB the point at which the final amplifier is S/Line transmitters, and this has proven
of clipping. Raising the clipping level to 18 overdriven when the audio input varies to be quite effective.
dB gave an additio'nal 4-dB improvement over a considerable'range. These typical Heat is an extremely important con-
at the receiver, or 8.5-dB total increase. ale systems, shown in Fig. 29, by the sideration in the use of any speech pro-
The improvement of the intelligibiiity of a nature of their design time constants offer cessor which increases the average-to-
weak ssb signal at a distant receiver can a limited increase in transmitted average- peak power ratio. Many transmitters, in
thus be substantially improved by rf c1ip- to-PEP ratio~ A value in the region of 2-5 particular those using television sweep
Single·Sideband Transmission 12·17.
tubes, simply are not built to stand the ef- from the filter is amplified and dent of signal waveform. A thermocouple
fects of increased average input, either in capacitance-coupled to the transmitter ammeter connected in series with the load
the final-amplifier tube or tubes or in the mixer. The relay contacts also apply ale would be a typical example of such a
. power supply. If heating in the final tube voltage to the first i-f stage and remove system. The output power would be equal
is the limiting factor, adding a cooling fan the screen voltage from the. second i-f to PR, where I is the current in the am-
may be a satisfactory answer. amplifier, when transmitting. meter and R is the load resistance (usually
Bilateral amplifier and mixer stages, 50 ohms). In order to find the PEP output
SINGLE·SIDEBAND TRANSCEIVERS
first used by Sideband Engineers in their with the latter m.ethod (using a two-tone
A transceiver combines the functions of SBE-33, also have found application in test input signal), the power output is
transmitter and receiver in a single other transceiver designs. The Circuits multiplied by two.
package. In contrast to a packaged shown in Fig. 30B and C are made to A spectrum analyzer is capable· of
"transmitter-receiver," it utilizes many of work in either direction QY grounding the giving the most information (of the three
the active and passive elements for both bias divider of the input transistor, com- methods considered here), but it is also
transmitting and receiving. Ssb transceiver pleting the bias network. The application the most costly method and the one with
operation enjoys widespread popularity of these designs to an.amateur transceiver the greatest chance of misinterpretation.
for several justifiable reasons. In most for the 80- through to-meter bands is Basically, a spectrum analyzer is a receiver
designs the transmissions are on the same given in the Fifth Edition of Single Side- with a readout which provides a plot of
(suppressed-carrier) frequency as the band for the Radio Amateur. signal amplitude vs. frequency. The
receiver is tuned to. The only practical The complexity of a multiband ssb readout could be in the form of a paper
way to carryon a rapid multiple-station transceiver is such that most amateurs buy chart but usually it is presented as a trace
"roundtable" or net operation is for all them fully built and tested. There are, on a CRT. For a spectrum analyzer to
stations to transmit on the same fre- however, some excellent designs available provide accurate information about a
. quency. Transceivers are ideal for this, in the kit field, and any amateur able to signal, that signal must be well within the
since once the receiver is properly set the handle a soldering Iron and follow instruc- linear dynamic range of the analyzer. For
transmitter is also. Transceivers are by tions can save himself considerable money a thorough explanation of the function
nature more compact than separate trans- by assembling an ssb transceiver kit. and application of this instrument see
mitter and receiver setups and thus lend Some transceivers include a feature that Rusgrove, "Spectrum Analysis - One
themselves well to mobile and portable permits the receiver to be tuned a few kHz Picture's Worth a . . . ," August 1979
use. either side of the transmitter frequency. QST.
Although the many designs available on· This consists of a voltage-sensitive
the market differ in detail, there are of capacitor, which is tuned by varying the Two-Tone Tests and Scope Patterns
necessity many points of similarity. All of applied dc voltage. This can be a, useful A very practical method for amateur
them use the filter type of sideband device when one or more of the stations in applications is to use a two-tone test signal
generation, and the filter unit furnishes a net drift slightly. The control for this (usually audio) and sample the transmitter
the receiver i-f selectivity as well. The car- function is usually labeled RIT for output. The waveform of the latter is then
rier oscillator doubles as the receiver receiver incremental tuning. Other applied directly to the vertical-deflection
(fixed) BFa. One or more mixer or i-f transceivers include provision for a plates in an oscilloscope. An alternative
stage or stages will be used for both crystal-controlled transmitter frequency method is to use an rf probe and detector
transmitting and receiving. The receiver S plus full use of the receiver tuning. This is to sample the waveform and apply. the
meter may become the transmitter plate- useful for "DXpeditions" where net resulting audio signal to the vertical-
current or output voltage indicator. The operation (on the same frequency) may deflection amplifier input.
VFO that sets the receiver frequency also not be desirable. If there are no appreciable non lin-
determines the transmitter frequency. The earities in the amplifier, the resulting
same signal-frequency tuned circuits may Testing a Sideband' Transmitter envelope' wiII approach a perfect sine-
be used for both transmission and recep- There are three commonly used Wave pattern (see Fig. 31A). As a com-
tion, including the transmitter pi-network methods for testing an ssb transmitter. parison, a spectrum-analyzer display for
output circuit. These include the wattmeter, oscilloscope, the same transmitter and under the same
Usually the circuits are switched by a and spectrum-analyzer techniques. In conditions is shown in Fig. 31B. In this
multiple-contact relay, which transfers the each case, a two-tone test signal is fed into case, spurious products can be seen which
antenna if necessary and also shifts the the mic input to simulate a speech signal. are approximately 30 dB below the
biases on several stages. Most commercial From the measurements, information amplitude of each of the tones.
designs offer VOX (voice-controlled concerning such quantities as PEP and As the distortion increases, so does the
operation) and MOX (manual operation). intermodulation-distortion-product (IMD) level of the spurious products and the
Which is preferable is a controversial sub- levels can be obtained. Depending upon resulting waveform departs from a true
ject; some operators like VOX and others the technique used, other aspects of sine-wave function. This can be seen in
prefer MaX. . transmitter operation (such as hum prob- Fig. 31 C. One of the disadvantages of the
lems and carrier balance)' can also be scope and two-tone test method is that a
Circuits checked. relatively high level of IMD-product
The use of a filter-amplifier combina- As might be expected, each technique voltage is required before the waveform
tion common to both the transmitter and has both advantages and disadvantages seems distorted to the eye, For instance,
receiver is shown in Fig. 30A. This circuit and the suitability of a particular method the waveform in Fig. 31C doesn't seem
is used by the Heath Company in several will depenq upon the desired application. too much different from the one in Fig.
of their transceiver kits. When receiving, The wattmeter method is perhaps the 31 A but the IMD level is only 17 dB below
'the output of the hf mixer is coupled to simplest;one but it also provides the least the level of the desired signal (see analyzer
the crystal filter, which, in turn, feeds the amount of information. Rf wattmeters display,in Fig, 3lD). A 17- to 20-dB level
first i-f amplifier. The output of this stage suitable for single-tone or cw operation corresponds to approximately to-percent
is transformer-coupled to the second i-f may not be accurate with a two-tone test distortion in the voltage waveform. Con-
amplifier. During transmit, Kl is closed, signal. A suitable wattmete( for the latter sequently, a "good" waveform means the
turning on the isolation amplifier that case must have a reading that is propor- IMD products are at least 30 db below the
links the balanced modulator to the band- tional to the actual power consumed by desired tones. Any noticeable departure
pass filter .. The single-sideband output the load. The reading must be indepen- from the waveform in Fig. 31A should be
12·18 Chapter 12
CROSS-OVER
DISTORTION
\
\
.... - _./
-.~ ,
'.'
,
Fig. 36 - A voice signal can be represented as an a·m waveform, which results from multiplication of a relatively slowly varying envelqpe (8) with a
carrier (C). Note the carrier peak amplitude is constant. The speech processor separates components 8 and C, and filters out 8, leaving the carrier
portion only.
+15V
RFC1
J2
OUTPUT
TO
XMTR
100k
3300
3300
+15V
12·20 Chapter12
MIC
HEADPHONES
However, a definite mathematical rela- intermodulation-distortion products. The quency and can't be eliminated by filter-
tion does exist between the desired com- "order" of such products is equal to the ing. As pointed out previously, such terms
ponents in an ssb signal and the "distor- sum of the multipliers in front of eaCh fre- do not normally result from fundamental
tion signals.'~ Whenever nonlinearities quency component. For instance, a term components beating with harmonics. An
exist, products between the individual such as (3fl - 2f2) would be called a exception would be when the fundamental
components whifh make up the desired fifth-order ~erm since 3 + 2 is equal to 5. signal along with its harmonics is applied
signal will occur. The' mathematical result In general, the third, fifth, seventh, and to another nonlinear stage such as a mix-
. of such mUltiplication is to generate other similar "odd-order" terms are the most er; Components at identical frequencies as
signals of the form (2fl- f2), (3fl), important ones since some of these fall the IMD products will result.
(5f2 - fl) and so on. Hence the term near the desired transmitter output fre- When two equal tones are applied to an
amplifier and the result is displayed on a
spectrum analyzer, the IMD products ap-
pear as "pips" off to the side of the' main
signal components (Fig. 31). The
amplitudes associated with each tone and
10k
the IMD products are merely the dB dif-
ference between the particular product
and one tone. However, each desired tone
is 3 dB down from the average power out-
put and 6 dB down from the PEP output.
Since the PEP represents the most im-
EXCEPT AS INDICATED, DECIMAL portant quantity as far as IMD is concern-
VALUES OF CAPAC ITANCE ARE ed, relating IMD-product levels to PEP is
IN MICROFARADS (JIF I; OTHERS
ARE IN PICOFARADS (pF OR JlJIF);
one logical way of speci(ying the
RESIST,ANCES 'ARE IN OHMS; "quality" of a transmitter or amplifier in
k 01000. M'l 000 000 '
regard to low distortion. For instance,
+15V IMD levels are referenced to PEP in
"Product Review" specifications of com-
mercially made gear in QST. PEP output
can be found by multiplying the PEP in-
put by the efficiency of the amplifier. The
input PEP for a two-tone test signal is
given by
10-
0.57 -1-
p
where
Ep =: the plate volfage
Ip = the average plate current
+i'5V FROM ' RFC 2 TO ALL
10 = the idling current
POWER~+15V IN Fig. 38 - Circuit diagram fOF the speech processor.
SUPPLY 1mH.1. __ , CIRCUIT Generally speaking, most actual voice
01,02 - Pair of matched silicon diodes (see
~0.01 (text).,
patterns will loo'k alike (in the presence of
03, 04 - Same as 01, 02. distortion) except in the case where severe
R1, R4, R7 - Use panel·mount type pots. flattopping occurs.' This condition is' not
-15V FROM RFC3 TO ALL R5, R6 - Circuit·board type pots. too common since most rigs have an ale
POWER ~-15V IN Sl - Opdt toggle switch.
SUPPLY , 1mH .1.. _. CIRCUIT T1 - 1000-ohm to 8-ohm audio transformer, system which prevents overdriving the
amplifiers. However, the voice pattern in
.rL°. 01 250 milliwatts.
U5, U6, U8, U9, UH, U12, U13 - Type 741
op amp, 8·pin DIP package.
a properly adjusted transmitter usually
has' a "Christmas tree" shape when
U1, U2, U3, U4, U7, U10 - National Semicon·
ductor LM301 op amp, 8'pin DIP package.
observed on a scope; an example is shown
10TE: U3 - U13 POWER CONNECTIONS ARE NOT
SHOWN ABOVE, CONNECT +15V TO PIN 7 U14 - National Semiconductor LM1595 (or in Fig. 33.
AND -15V TO PIN 4 Of EACH, Motorola MC1595). 14·pin DIP package.
An Audio Speech Processor
Over the years, different speech pro-'
12·22 Chapter 12
.'
tries to implement the preceding scheme. at the processor input. Output from VI is mended for rf shielding.
MIC PREAMP. AUDIO AMP. COt.1PA~ATOR
R2
1M
R7
47k
R6
AUDIO AMP. 41k
DELAY AND RELAY· DRIVER
10M _--"'''',_~
RFC2 4.1M DELAY ADJ,
*
9
Fig. 39 - Schematic diagram of the VOX unit. Unless otherwise specified, resistors are 1I4-watt composition. Capacitors with polarity marked are
plastic-encapsulated tantalum; others are disc oeramic. . ,
C1 - For text reference. similar. A4, A5, AS - Miniature control (see text).
01-04, incl. - Germanium diode, 1N34A, 1NS7 K1 - Miniature type, 12-volt coil (see text). AFC1-AFC3, incl - Ferrite bead.
or equivalent. 01 - 2N5139 silicon pnp. U1 - National Semiconductor LM3900.
05 - Silicon diode, 50 PAVor more, 1N4001 o~ A1-A3, incl., AS, A7 - For text reference. U2 - Type 555 timer.
Procurement of parts should present no wave. Disconnect the generator, recon- readings may be used. With the processor
particular problems. As of the time of this nect the, microphone, and plug head- switched "out" speak into the micro-
writing, the 741 and LM30l operational phones into J3. Advancing volume con- phone and increase the transmitter's
amplifiers used in the circuit can be pur- trol R8, one should now be able to hear microphone gain control until the ale
chased from mail order houses fol' about himself talking, although background meter starts to deflect. Note the peak
30 cents apiece. The LMI595 integrated noise and ac hum will probably be very readings. Switch the processor "in" and
circuit, probably the single most expensive high. Adjust noise blanking control R4 adjust R7 to obtain the same peak
item in the processor, was bought for for the desired degree of background reading.
under $2. Matched diodes for 01,02,03 noise suppression.
and 04 should cost less than a dollar. Those who do not have access to test A Modem Solid-State VOX
The circuit is powered by a dual dc equipment may do the following:' Set R4 Voice-operated T-R control is a great
power supply that provides plus ahd to the center of its range. Connect a convenience in ssb operation. The unit·
minus 15 volts, as is typically used with microphone to 11 and headphones to H. described here is compact and uses inex~
most operational amplifiers. Current con- Speaking into tire microphone, advance pensive components. It is ideal for inclu-
sumption is approltimately 50 rnA from input gain control RI and monitor volume sion ipa home-built transceiver or exciter,
each side of the supply. control R8 to the poi-nt where the .speech or retrofitting to commercial gear that
becomes audible in the headphones. Ad- does not have VOX. The performance of
Initial Adjustments just offset controls R5 and, R6 for mini- this unit is improved over previously
If an oscilloscope and audio sine wave mum distortion as monitored in the head- published versions because of a modifica-
generator are available, the following phones. The final setting of RI is not tion suggested by W7KGZ.
alignment procedure should be followed: critical. It should be high enough so that
Set R4 to minimum resistance. Connect a the circuit functions properly (is set too Circuit Description
microphone to 11 and the oscilloscope low, the audio output will sound broken The schematic diagram of the VOX
probe to pin 6 of U I. Adjust R I, the input up and "grainy") but not so high that the device is shown in Fig. 39. Three of the
gain control" while speaking into the speech amplifier itself distorts the signal LM3900 sections have been configured as
microphone so that the-voice peaks view- by clipping. Adjust R4 to suppress back- high-gain audio amplifiers. UIA and UIB
ed on the oscilloscope are slightly oelow ground noise as desired. amplify the signal from the microphone.
the output clipping level of U I (approxi- Finally, connect the processor output at UIC functions as .an amplifier for audio
mately 14 volts peak). Remove the micro- J2 to the transmitter's 'microphone jack. sampled at the station speaker. Coupling
phone and connect the signal generator to Switch the processor out of the line by capacitors in the audio stages have ,been
J I. Set the generator frequency to about means of S1. If a Monitorscope is avail- chosen ,to reduce response below 300 Hz.
1000 Hz and adjust its output level to pro- able to view rf output, speak into the This will minimize hum problems.
duce about 10 volts peak at pin 6 of U 1. microphone and note the level of the voice • Outputs from the microphone and
Place the oscilloscope probe on pin 6 of peaks. Switch the unit "in" and adjust speaker amplifiers are capacitively cou-
U12. Adjust offset controls R5 and R6 for output level control R7 for the same peak pled to rectifier stages which convert the
the best-looking sine wave. It should be voice output level. If a monitorscope is audio signals to varying dc voltages. Ger-
possible to produce a nearly perfect sine not available, the transmitter's ale meter manium dioe/es, because of their lower
Single.sldeband Transmission
. 12·23 ,
\' ,
~,"
DRIVER'
04
2N2102
LOCAL
OSC.
VI
o I---------t
0.1
TO~
'''''''''i ~ L RF AMP.
MIXER
I
I
I
/
Ll I
/
/
i
I
I
LI
_ _ _ _ _,
____ ___ ________ I
~
~
~
,~0.01
Fig. 42 - Schematic diagram far.the transl(erter. Resistors are 1/2-watt composition and capacitors are disc ceramic, unless otherwise noted.
C1 - Dual·section air variable, 140-pF per 02 - Zener diode, 6.8-volt, 1-watt (1N4736). L2, L4 - 19.5 to 24.3-"H variable. inductor
section, or two 150 pF air variable units. J1 - Phono type, chessls mount. (Miller 46A225CPC).
C2 - Air variable, Millen 19280. J2 - Coaxial receptacle, chassis mount. L3 - 22 turns of no. 28 enafTI. wire wound on
C3 - Dual-section broadcast variable, 365 pF K1, K2 - 12-V dc, 2·A contacts, dpdt relay L4 coliform.
per section, both sections connected In (Radio Shack 275-206). L5 - 18.8 to 41.0-,,101 variable inductor, Miller
parallel. L1 - 11 turns of no. 28 enam. wire wound 42A335CPC.
01 -·SlIicon,50 PRV, 100 mAo over L2.
12·24 Chapter 12
stops. Ql allows the SSS to'be retrigger~d
.
qUirements. This writer's unit uses pc- . type ().f relay will depend on how tbe VOX
continuously. On~ of the major difficul- mount controls which are aligned on the device will be used. Any 12-volt. relay
ties of earlier VOX circuits was that the 'board so that they may. be accessed' which requires less than 200 mA of cur-
capacitor discharge circuits were usect through small holes in the rear pimel of rent canbe employed. When the VOX re-
where' the capacitor would not always be the transceiver. If panel-mount controls lay must drive a second relay, such as the
fully charged, so the time delay produced are desired, Mallory MLC units may be antenna relay in a transceiver, the fast
would vary .. used for R4, RS and R8. operating time of a reed relay is needed to
The 555 has a' current-switching capa- The VOX device is small enough so it prevent clipping of the first syllable
bility of 200 mA, sufficient to directly can be mounted inside most rigs. If a spoken. The total close time of all relays'
drive either a relay or a solid-state switch- separate VOX unit is needed, a small util- connected in tandem should be 10 milli-
ing arrangc:ment. OS is induded to protect ity box or Minibox will make an appro- seconds or less. If the VOX relay will per- .
the IC from trallsients generated when priate housing. Rf interference can cause form all switching functions directly, a
switching an inductive load such as a relay trouble, so the unit should be shielded in miniature control relay such as the Potter
coil. any application where rf fields may be' & Brumfield RIO series is appropriate.
present. The bypass capacitors for the These relays are available in 2-, 4- and
Compon.nts and Construction audio inputs are located on the circuit 6-pole versions, part numbers RIO-EI-Y2-
The VOX unit is constructed on a 2-3/8 board. If the leads from the audio connec- 185, RIO-EI-Y4-VI85 and RIO-EI-Y6-
x 2-3/4-inch (60 X 70-mm) etched cir- tors are more than a few inches long, the V90, respectively.
cuit board. The photo indicates that one- bypass capacitors and their associated The circuit board is designed for
third of the board real estate i~ unused, so ferrite-bead chokes should be mounted at 1I4-watt resistors which are mounted flat.
a smaller version is possible. The type of the connectors. If II2-watt units are used, they must be
controls and. relay el1Jployed will be deter- No provision has been made for mount- positioned vertically. Care must be
mined .. by the builder's individual re- ing the relay on the circuit board, as the employed· when mounting and soldering
KIC
F
ANT..
1.8 MHz
tttt
I
I
I
I
LS, L8 - 35 to 43.0·"H variable Inductor, Miller L 10 - 42 turns, no. lS enam. wire equally 102-572.
4SA395CPC. spaced on a T-200 Amidon core. RFC2 - 56-"H rf choke, Millen J-302-5S.
L7 - 13.2 to lS.5·"H variable Inductor. Miller Ml - 500 mA, panel mount, Simpson 17443 or Yl - 19.3-MHz crystal is used for a 21-MHz
4SA155CPC. similar. i-f, 2S.5-MHz crystal used for a 28 MHz i-f.
L9 - 10.8 to 18.0-"H adjustable coil, Miller 05, OS'- RCA40673 MOSFET. Zl, Z2 - 2 turns, no. 18. enam. wound over
21A155RBI. RFCl - l-mH, 5OO-mA rf choke, Johnson 47-ohm, '2-watt composition resistor.
Fig. 43 - Top view of transverter with cover removed. Final amplifier Fig. 44 - The bottom view of the chassis. The sockets for the 6146B
circuit is at the left. The rear apron has an accessory socket for an tubes are at the lower center. The etched·circuit board-is above the
external power supply (transceiver), rf and remote·keylng connectors. final amplifier tube sockets and the T·R relays at the upper right. The
The plate meter is at toe lower left. different supply voltages are obtained from tlJe I!ssociated transceiver.
12-28 Chapter 12
\.
+I2V
SPEECH AMP.
tOOk
BAL. MOD.
o.Ot
~
TO' STBY SWITCH I+I2V)
cw 52
I2V
r+ '
SSB
USB
CW DRIVE
4321
SUI
27 B
NPo
A
GSD
E@C GI@2
00
0 00 5
• = PHASING Q5 5678
o. DC VOLTAGE
Q3, Q4, Q6
BOTTOM VIEWS
QI. Q2
+I2Vo---------------------~--------------------~
o . PK-PK VOLTAGE
Fig. 45 - Schematic diagram of a practical !I·MHz ssb generator. Fixed-value capacitors are disc ceramic unless otherwise noted. Polarized
capacitors are aluminum or tantalum. Fixed-value resistors are 1I2-watt composition.
C1, C2 - Miniature 30-pF trimmer. NPO R2 - 1000-0 pc-board-mount control. wire on a T50·2 Iron core. I'e 10, dla 0.5 = =
ceramic ·preferred. R3 - 25·kO linear·taper control, panel ' inch. Link has 10 turns of no. 30 insulated
C3, C4 '"'7 Miniature 6O·pF trimmer. Mica mounted. wire over 01 end of primary.
compression type suitable. S1, S2 - Single·pole, double·throw miniature T3 - 10·/lH. 44 turns of no. 26 enam. wire on
01 - 9.1·V, 400·mA zener diode. switch, panel mounted. a T50-2 iron core. Link has 22 turns of no.
FL 1 - Spectrum International 9·MHz T1 - 15 trifilar tl.ll'ns of no. 26 enam. wire 30 Insulated wire over cold end of primary.
crystal·lattice filter. Type XF·9A. (twist 10 times per inch) on an FT·50-61 Y1, Y2 - Crystals to match FL1. Obtain from
(see QST ads). toroid core II"e = =
125, dia 0.5 Inch/13 mm). filter manufacturer or International Crystal
R1 - 10·kO audio·taper control, panel T2 - 10'/lH primary. 44 turns no. 26 enam. Mfg. Co.
mounted. .
of the transceiver. Then connect an an- 21.1-MHz cw drive power at J1. Tune L6, amplifier in Fig. 45. Output from Ul is
tenna to J2 and listen for signals. Peak the L7, L8 and L9 for maximum meter supplied to the gates of Q 1. Q 1 and Q2 are
incoming signals with the preselector con- reading. While monitbring the plate cur- used as a balanced modulator. MOSFETs
trol. The slugs of L2 and L4 should be ad- rent, tune C2 for a dip. C3 is the I'A are used to prevent changes in gate-source
justed for the highest S-meter reading on loading control. When the P A capacitors capacitance when the audio level is in-·
the transceiver. L5 should be set for maxi- are properly adjusted, the p~ate current creased by means of R1. JFETs will not
mum output at 21 or 28 MHz. If the re- will be about 220 rnA. work properly at Ql and Q2 because the
ceiving converter is functioning properly. junction capacitance changes with in-
it will be possible to copy a 0.1 IAV signal A Sideband-Generator Module creased audio drive, thereby unbalancing -
without difficulty in areas where at- A practical circuit for a simple ssb the modulator.
mospheric and man-made noise are at a 'generator is given in Fig. 45. Output is at 9 Tl is a broadband trifilar-wound
minimum. If no signals can be heard, MHz. This circuit can be followed by ap- toroidal transformer. It provides the
check Ql to make certain that it is work- propriate stages of the designer's choice, necessary 180 0 phase difference for the
ing properly. A wavemeter or general- thereby making it possible to heterodyne drains of Ql and Q2 while coupling the
coverage receiver can be employed to see the 9-MHz energy to a desired amateur balanced modulator output to the i-f
if the crystal oscillator is operating. band. A block diagram for a typical cir- preamplifier, Q3.
Attach a 50-ohm load to J2 before test- cuit lineup is provided in Fig. 46. Circuit- . R2 compensates for differences m.
ing the transmitter section. Set Rl (or an design information for the additional operating level between Ql and Q2. Th~s
indicated restillg plate current of 50 rnA stages is found elsewhere in this book and balancing control makes it unnecessary 'to
on M1. This adjustment should be made in Solid State Design for the Radio use matched transistors in the balanced
without drive applied but with Kl and K2 Amateur. modulator. R2 is adjusted for maximum
energized. Next, apply about one watt of A 741 op amp functions as the audio 'carrier suppression, which will be on the
12·28 Chapter 12
Chapter, 13
Frequency Modulation
and Repeaters
j
WW\N~MhfflME
taneous amplitude of the modulating
Frequency and Phase Modulation signal is small, and is greatest when
It is possible to convey intelligence by the modulating signal reaches its peak,
modulating any property of a carrier, either positive or negative.
including its frequency and phase. When As shown in the drawing, the amplitude (el
the frequency of the carrier is varied in of the signal does not change duripg
accordance with the variations in a modulation.
modulating, signal, the result is frequency Fig. 2 - Graphical representation of frequency
modulation (fm). Similarly, varying the Phase Modulation modulation. In the unmodulated carrier ~t A,
phase of the carrier current is called phase If the phase of the current in a circuit is each rf cycle occupies the sa'me amount of
time. When the modulating signal, B, is. ap·
modulation (pm). , changed there is an instantaneous fre- plied, the radio frequency is increased 'and
Frequency and phase modulation are quency change during the time the phase is decreased aocording to the amplftude and
not independent, since the frequency being shifted. The amo.unt of frequency polarity of the modulating signal.
13-2 Chapter 13
/ .
C2 50 osc. ~ANK
r-----------H---'--_~H::_-+_---......- T O osc.
+12V 0.001
C1
' - - - - - -......--O+12V
I
~-
narrow-band fm transmitter modulated by a
1-kHz tone.
MIC. ANT.
(Al
13-4 Chapter 13
.TS-175, will provide sufficient' accuracy.
Frequency counters that will work di-
,----~----.,p_--.._-----:::-p_-~----o+ 9-1~V
rectly up to 500 MHz and higher are
available, but their cost is high. The less
T1 expensive low-frequency counters can be
II~
220k employed using a prescaler. a device
TO
47k which divides an input frequency by a
DISCRIMINATOR preset ratio, usually 10 or 100. Many
4-,'100k 0.1
1N4~7
prescalers may be used at 148 MHz or
higher, using a counter with a 2-MHz (or
10)JF +
more) upper frequency limit. If the
15V counting system does not have a sufficient
upper frequency limit to measure the
output of an fm transmitter directly, one
of the frequency-multiplier stages can be
sampled to provide a signal in the range of
'the measurement device. Alternatively, a
PEAK DEVIAT,ION METER
crystal-controlled converter feeding an hf
/ receiver which has accurate frequency
Audio Deviation Produced readout can be employed, if a secondary
Frequency 1st Null 2nd Null 3rd Null -standard is available to calibrate the
905.8 Hz ±2.18 kHz ± 5.00 kHz ± 7.84 kHz receiving system.
1000.0 Hz ±2.40 kHz ± 5.52 kHz ± 8.65 kHz
1500.0 Hz ±3.61 kHz ± 8.'28 kHz ±12.98 kHz Deviation and Deviation Linearity
1811.0 Hz ±4.35 kHz ±10.00 kHz ±15.67 kHz
2000.0 Hz ±4.81 kHz ±11.04 kHz ±17.31 kHz A simple deviation meter can be
2079.2 Hz ±5.00 kHz ±11.48 kHz ±17.99 kHz assembled following the diagram of Fig.
2805.0 Hz ±S.75 kHz ±15.48 kHz ±24.27 kHz SA. This circuit was designed by K6VKZ.
The output of a wide-band receiver
Fig. 8 - (A) Schematic diagram of the deviation meter. Resistors are 1/2,watt composition and discriminator (before any deemphasis) is
capacitors are ceramic, except those with polarity marked, which are electrolytic. 01·03, incl., are
high·speed silicon switching diodes. R1 is a IInear·taper composition control, and Sl, S2 are spst fed to two amplifier transistors. The
toggle switches. T1 is a miniature audio transformer with 10·kll primary and 20·kll center·tapped output of the amplifier section is trans-
secondary (Triad A31X). (8) Chart of audio frequencies which will produce a carrier null when the former coupled to a pair ofr.ectifier diodes
deviation of an fm transmitter is set for the values given. to develop a dc voltage for the meter, M I.
There will be an indication on the meter
with no signal input because of detected
added at the transmitter to proportionally indirect method (B) often produces su- noise, so the accuracy of the instrument
attenuate the lower audio frequencies, perior results. will be poor on weak signals.
giving an even spread to the energy in the To calibrate the unit, signals of known
audio band. This results in an fm signal of Testing an FM Transmitter deviation will be required. If the meter is
nearly constant energy distribution. The Accurate checking of the operation of to be set to read 0-15 kHz, then a 7.5-kHz
reverse process, cal1ed deemphasis, is an fm or pm transmitter requires different deviation test signal should be employed.
accomplished at the receiver to restore the methods than the corresponding checks R I is then adjusted until M I reads half
audio to its original relative proportions. on an a-m or ssb set. This is because the scale, 50IlA. To check the peak deviation
Sample circuits are shown in Fig. 6. common forms of measuring devices of an incoming signal, close both S I and
either indicate amplitude variations only S2. Then, read the meter. Opening first
FM Exciters (a milliammeter, for example), or because one switch and then the other will indicate
FM exciters and transmitters take two their indications are most easily inter- the amount of positi've and negative
general forms. One, shown at Fig. 7A, preted in terms of amplitude. deviation of the signal, a check of
consists of a reactance modulator which The quantitH!s to be checked in an fm 'deviation linearity.
shi'rts the frequency of an oscillator to transmitter are the linearity and frequency
generate an,rm signal directly. Successive deviation and the output frequency, if the Measurement of Deviation Using Bessel
multiplier stages provide output on the unit uses crystal control. The methods of Functions
desired frequency, which is amplified by a checking differ in detail. Using a mathematical relationship known
PA stage. This system has a disadvantage as the Bessel Function it is possible
in that, if the oscillator is free running, it is Frequency <7hecking to predict the points at which, with certain
difficult to achieve sufficient stability for Crystal-controlled, channelized fm oper- aUdio-input frequencies and predeter-
vhf use. If a crystal-controlled oscillator is ation requires that a transmitter be held mined deviation settings, the carrier
employed, because the amount that the within a few hundred hertz of the desired output of an fm transmitter will disappear
crystal frequency is changed is kept small, channel even in a wide-band system. completely. Thus, by monitoring the
it is difficult to achieve equal amounts of Having the transmitter on the proper carrier frequency with a receiver, it will be
frequency swing. frequency is particularly important when possible to identify by ear the deviation at
The indirect method of generating fm operating through a repeater. The rigors which the carrier is nulled. A heterodyne
shown in Fig. 7B is currently popular. of mobile and portable operation make a signal at either the input or receiver i-f is
Shaped audio is applied to a phase frequency check of a channelized trans- required so that the carrier will produce a
modulator to generate fm. Since the ceiver a good idea. at three-month beat note which can easily be identified ..
amount of deviation produced is very intervals. . Other tones will be produced in the
smal1, a large number of multiplier stages Frequency. meters generally fall into modulation process, so some concen-
is needed to achieve wide-band deviation two categories, the heterodyne type and the tration is required by the operator when
at the operating frequency. In general, the digital counter. For amateur use, the making the test. With an audio tone
system shown at, A will require a less vhfluhf counterparts of the popular selected from the chart (Fig. 8B), advance
complex drcuit than that at B, but the BC-221 frequency meter, the TS-174l!nd the deviation control slowly until the first ~
o +
FREQUENCY
FM Filters
Center Nonimal Ultimate Impedance (r) Insertion Crystal
Manufacturer Model Frequency Bandwidth Rejection In Out Loss . Discriminator
KVG (1) XF-9E 9.0 MHz 12 kHz 90 dB 1200 1200 3 dB XD9-02
KVG (1) XF-107A 10.7 MHz 12 kHz 90dB - 820 820 3.5dB XD107-01
KVG (1) XF-107B 10.7 MHz 15 kHz 90 dB 910 910 3.5 dB XD107-01
,
KVG (1) XF-107C 10,7 MHz 30 kHz 90dB 2000 2000 4.5dB XD107-0l
Heath Dynamics (2)- 21.5 MHz 15 kHz 90 dB 550 550 3 dB -
Heath Dynamics (2)· 21.5 MHz 30 kHz 90 dB 1100 1100 2dB -
Clevlte (3) TCF4-12D3CA 445 kHz 12 kHz 60 dB 40k 2200 6 dB -
Clevlte (3) TCF4-18G45A 455 kHz 18 kHz 50dB 40k 2200 6 dB -
Clevite(3) TCF6-30D55A 455 kHz 30 kHz 60 dB 20k 1000 5 dB -
Fig. 11 - A list of 1m-bandwidth filters that are available to amateurs. Manufacturer's addresses are as follows: (1) Spectrum Internati6nal, P. O. Box
1084, Concord, MA 01742; (2) Heath DynamiCS, Inc., 6050 N. 52nd Ave., Glendale, AZ. 85301, tel. 602-934-5234; (3) Semiconductor Specialists, Inc.,
P. O. Box 66125, O'Hare International Airport, Chicago, IL 60666.
null is heatd. If a higher-order null is side of the selectivity curve. When the Oth~rwise the functions, and often the
desired, continue advancing the control frequency of the signal varies with circuits, of the rf, oscillator, mixer and
further until the second, and then the modulation it swings as in4icated in Fig. audio stages will be the same in either
third, null IS heard. Using a carrier null 9, resulting in an a-m output varying receiver.
beyond the third is generally not practical. between X and Y. This is then rectified as In operation, the noticeable difference
For example, if a 905.8-Hz tone is used, an a-m signal. between the two receivers is the effect of
the transmitter will be set for 5-kHz With receivers having steep-sided I se- noise and interference on an incoming
deviation when the second null is reached. lectivity curves,' the method is not very
The second null achieved with a 2805-Hz satisfactory because the distortion is quite
audio input will set the transmitter severe unless the frequency deviation is PLATE
deviation at 15.48 -kHz. The Bessel- small, since the frequency deviation and CURRENT
function approach can be used to calibrate output amplitude is linear over only a LIMITER OUTPUT
13-8 Chapter 13
signal. From the time of the first spark
transmitters, ",rotten QRN" has been a
major problem for amateurs. The limiter
and discriminator stages in an fm set can
eliminate a good deal of impulse noise,
except noise which manages to acquire a .
frequency-modulation characteristic. Ac-
curate alignment of the receiver i-f system
and phase tuning of the detector are
required to achieve good noise sup-
pression. Fm receivers perform in an
Fig. 13 - (A) Input wave to a limiter stage shows a·m and noise. (8) The same Signal, after
passing through two limiter stages, is devoid of a·m components.
unusual manner when QRM is present,
+250V
INPUT OF
DISC. TRANS.
r------
FROM 470 I
1- F
AMP.
0---1 f-..---+---+- I
I
I
6BH6 (2)
!
I
L ______ _
~4
6.3V AC ~,).,
68K
+250V
(Al
FIRST Llt.\ITER SECOND LIMITER
PART OF
DISCRIMINATOR TRANSFORMER
r-----------
FROM
I-F
AMP.
E
EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPAC ITANCE ARE
IN MICROFARADS (Jlf) ; OTHERS
ARE IN PICOfARADS (pf OR JlJlfl;
+cr--~------1~--~--~--------------r----~
12V RESISTANCES ARE IN OHMS;
k -1000.101-1000000
4.7):IF
,£h 15V
(8)
PART OF
DISC. TRANS.
r--~--
LIMITER
I
I
.---~----._--.---._--~+9V
I
I,
I
I
I
I
L __ _
FROM
I-f
AMP.
I
;h.l
L ___ ,_________ .JI +IIV
(e) (0)
Fig. 14 - Typical limiter circuits using (A) tubes, (8) transistors, (C) a differential IC, (D) a high-gain linear IC.
FROM
I-Fo--.;--..._--I
AMP, 6Bk 0.1
0.001
100
lN34A
+B
13-8 Chapter 13
FROM I-F
AMP. PLATE
t>LL ,DETECTOR
(A) •
68
.--.....: : - : - - - - _ - o + 5 V
cz
;+.,0.1 AUDIO
1000 OUTPUT
AUDIO
+e ·OUTPUT
Fig. 17 - A ratio detec!lor of the type often used in entertainment radio and TV sets. T1 is a ratio-
detector transformer such as th.e Miller 1606.
-5V
(8)
(-~-
bands need additional receiver gain and L ___ _ 1_ ... _
0.001 \0,001
noise-figure improvement for weak-signal ·F.T. "'F'T
work. Too much gain can seriously de- ,GAIN~ HI- 20dB ' - - - E K ' I - - - - - - -......- ....."V'I/'v--Q+f2V
NF.5 dB . 100
grade the receiver dynamic range, so care O.O~
must be exercised when adding a pre- -e---. FERRITE BEAD (...,i·125)
amplifier ahead of an existing receiver
front end. The temptation of some inex-
perienced amateurs is to usc a preamp Fig. 21 - Schematic diagram of the 440·MHz preamp.' The 10·pF capacitors are silver mica.
FT indicates feedthrough capaCitor. '
which has a gain of 25 dB or greater. As a C1·C3, incl. - 1.4 to 9.2·pF miniature air vari- approx. 114 inch (6.3 mm) below C2 and C3
result: strong local signals can overload able', Johnson 189-0563·001 or equiv. ends of line:01, 02 source taps on L1 and
the receiver and cause severe mixer IMD: J1, J2 - BNC·type connector.soldered to case L2 are approx. 314 inch (19 mm) up from
The two preamplifiers described here are outer wall. ground.
L1-L3, incl. - 2·518 x 114-inch (67 x 6.3·mm) 01,02 - Siliconix JFET.
tailored for useful but not excessive gain brass strip. Input and output taps on L1 and RFC1, RFC2 - 420·MHz choke J. W. Miller
amounts. They should enhance the' per- L3 are approx. 112 inch (13 mm) up from 4584 or equiv. Ferrite beads assoc. with
, formance of receivers or converters that ground (see text). Attach 01102 drain taps these chokes are Amidon miniature. type.
are marginal in terms of overall gain and
,noise figure. They should not have a
serious effect on the receiver dynamic
range.
The 2-meter version shown in Fig. zo- CI of Fig. 20 is adjusted for lowest A strip-line preamplifier for use at 440
utilizes a single Siliconix U310 JFET in a noise figure. The Clcoil tap can be ad- MHz is shown in Figs. 21 and 22. The
common-gate circuit. This helps to ensure juSted 'also if further improvement is maximum attainable gain is roughly 20 dB
stability and provide a gain of over 10 dB. needed. C2 and C3 should be high-Q trim- with the circuit shown. Noise figure
The U310 is well known for its low noise mers for best performance. Miniature should be better than 5 dB when the taps
up to 450 MHz (about 3 dB at 450 MHz ceramic trimmers should be suitable for onto LI are optimized. Sqme experiment-
and 1.5 dB at 144 MHz). This transistor usc at CI, C2 and C3. Ideally, Teflon ing will be necessary. The 'loaded Q of the
also has excellent dynamic-range charac- trimmers or small air variables would be rthree resonators can be increased by
teristics (in excess of 100 p8). A less costly used at those circuit points, moving the Q 1IQ2 taps closer to the
substitute is the Siliconix E300, which If LI 'and L2 are at right angles to one ground ends of each line. The tradeoff is
comes in a plastic case. The performance another and spaced well apart, it may not. in preamplifier gain. The increased Q may
traits are approximately the same, but be necessary to use a shield divider across be important when gain requirements
stability may be harder to realize because QI as shown in Fig. 20. However, a small aren't too gr~t (as in a repeater installa-
the E300' has no metal case which can be piece of cOPRer, brass or double-sided pc' tion), but when rejection of out~of-band
grounded automatically when the gate is board should be easy to add to the circuit commercial signals are vital to good per-
grounded. board. ' formance. L1, L2 and L3 are silver plated
I
I
I I
I I
I I
Jl I
INPUT I
I
I
I
I
I 0.001
0.001
F.T.
I INNER SHIELD
~~-~IVv_----~----~Ar----~
I 220
I
I
L _ _ _ _ _ _ _ -,- _ _ , I
--l
+
12VDC
SHI:DT-E~;,:-A;-I~;;~E;.:;ALV~~;--------1....---o
Ql,Q/1? rh
0:":
CAPACITANCE ARE IN MICROFARADS ( JlF I ;
OTHERS' ARE IN PICOFARADS ( pF OR JlpFl;
G S D RESISTANCES ARE IN OHMS;
k·I000. M o l,OOO.OOO. .
Fig. 23 - Circuit diagram of the preamplifier. Heavy lines indicate the pc-board shield box and dividers. The outer shield box is shown in daslled
lines. Fixed-value capacitors'are disc ceramiC unless otherwise noted. S.M. indicates silver mica. Resistors are 112 watt carbon.
C1-C4, inc I. - 11-pF subminiature air variable, Q1, 02 - Vhf or uhf JFET (see text). Keep
E. F. Johnson 189-584. Piston trimmers or gate lead as short as pos~ible, 118 inch
Johnson 160-0104-001 suitable also. L1, L3 - 3-112 turns no. 14 tinned bus wire, (3 mm) or less.
C5-C7, incl. - Feedthrough capaCitor. 112-inch 10 x 314-lnch (13 x 19 mm) long. RI=C1 - 144-MHz rf choke, approximately 2.7
01,02- High-speed silicon switching diode, Tap source at 1-314 turns from trimmer end. I'H. James Millen 34300-2.7 or equivalent.
1N914 or equivalent. Tap L1 also at 112 turn from ground. Alternatively, wind 20 inch (508 mm) of.no.
03 - 15-volt, 1-watt Zener diode. L2, L4 - 3-113 turns no. 14 tinned bus wire, 30 enam. wire on the body of a 2700-ohm
J1, J2 - Coaxial connector of builder's choice. 112-inch ID x 314-inch (13 x 19 mm) long. 1-watt carbon resistor. Use pigtails as
(Type BNC used in this modeL) Tap L4 112 turn above C6. anchor points for ends of winding.
e
Ul,
2,3,4
TO PINS 4 AND 6 OF U 4 0 _ - - - - - - . . . . , . 4
PIN 4 OF Ul,2,3 8> FROM
VCO
2 1 2
ALL DIODES
SILICON SWITCHING 1 2 1
TYPE lNS14 OR , 'S2B S2C
EQUIV. S2A 4
rL- 4
..!L ~ ~
C C
\C
~~ ~ ~~ ~.
,. ~ A~ ~~
. 270
- r;h
>270
?
14
~
A
~270
T. 11
14
8
270
14
270
11
~
~270
~
Ul U4
Mc~gl6P
6 I U2 6
1 6 1
MC4016P MC4016P MC4016P
.;oN .;oN .;oN 7N
f- ~ 5 5 \
,r
2 2 2 2
'1" f ~
,r ,
T
,, ,r
270
vV'v rL.A,j\ty
270
,
, , ,r
1"
.,
- ~ ~ ~ ~
I 2 4
l C
lc
J'
Ql
2N363e
Q2
2N51BS
I
t ~ ~
R1
+5V
125~~ ~"
100'0 v
. .J\" "
WOOV
7
FROM J1,PINe
(FIG.4-40)
R3
625rv FROM
DIVIDE-BY-N C4
CIRCUIT 0.22 AMP
(FIG. 4-38) DC AMP
. 500/n R5
lOOk
10 R4
8200
C3
+
r+;5V
22
100
R6
1000
:h 0l
~22
'100 l5V
DC TO REF. OSC.
CIRCUIT
REGULATOR
;h.01
+12V
~
+12V
'0
CAPACITANCE ARE IN MICROFARADS (jlF ) ; U6
OTHERS ARE IN PICOFARADS (pF OR jljIF); 3-CASE
RESISTANCES ARE IN OHMS;
ko 1 000, M.l000 000.
1234~67
TOP VIEW
<fC5
.,
.2
BOTTOM VIEW
.
Fig. 27 - Circuit of the VCO and PLL portion of the synthesizer. Resistors can be either 1/4 or 1/2 watt composition.
C1 - g. to 60·pF compression trimmer, Arco enam. closewound on high·value 2·watt reo P1 - 9-pin chassis-mount connector, male.
• 404 or equiv. sistor. Miller 4506 slug tuned inductor also U5 - Phase-frequency detector IC, Motorola
01 - 15- to 60-pF voltage-variable capacitance suitat:>le. . MC4044P or equiv.
diode. Motorola MV2205 or MV2209 suitable. L2, L3 - 8.9 H miniature audio choke. Aladdin U6 - 5-volt regulator IC. 1A, National Semi-
L1 - 15 I'H nominal. Approx. 40 turns no. 28 86-101 or equiv. conductor LM309K or equiv.
13-14 Chapter 13
influences 'the lockup time' of the VCO, i~ The 5.5- t06.5-MHz output from the call for. extremely rigid ritechanical as-
ad<,fttion to reducing the amplitude of any VCO buffer,,'Q7, is divided by four pro- sembly, and very good filtering and
625-Hzenergy present Values for Ll and grammable stages, called divide-by-N shielding techniques: With a VFO that
D t are selected so as to cause Q4 to (Fig. 25). The divisor, N,. is ~e\ected by will be used with several stages of
oscillate - on the frequency needed for setting the thumbwheel switches to the frequency multiplication for an output on
transmitting, in this case (; MHz. proper positions. These switches have the 144 MHz, the importance of·the afore-
A trimmer capacitor, Ct, is added to necessary binary-coded output connec- mentioned practices cannot be over-
the circuit to adjust the oscillator to the tions to program the divided ci~cuits. Two stressed. The coil, voltage-variable capacitor,
frequency 'nee"ded for receiving. In the separate sets of thumbwheel switches are and the trimmer capacitors associated
receive mode, a positive voltage is applied connectt;d to the divider chain through with the frequenty-determining parts of
to D2, causing it to conduct and isolating diodes. A positive voltage, ap- the circuit should be mounted securely to
effectively place "C 1 in parallel with D l. plied to Q2, causes the diodes to conduct, . eliminate microphonic effects. The leads
This lowers the frequency of oscillation to thereby connecting the receive thumb- carrying dc supply voltages should be well
S.S MHz, which is multiplied by external wheel to the chaln. When there is no filtered and shielded.· Since it takes only
circuits to the frequency needed for voltage applied to Q2 and the tr'ansmit- millivolts of change to cause a wide
receiver mixing. At the same time that a receive bus is grounded, QI will conduct, excursion of frequency, the leads carrying
voltage is applied to D2, the divide-by-N ca\lsing the transmit .channel selector the correction voltage to the VCO should
counter is reprogrammed to provide a switches to be connected to the divider be well shielded against hum and noise.
62S-Hz signal for comparison, with the chain. The 625-Hz output from the In the model described here, the VCO
reference, thus keeping the VCO sta- divide-by-N circuit is connected to the re- coil, L I, is wound on a 2-watt resistor of
bilized. QS, Q6, and Q7 are buffers to maining input of the phase detector. high ohmic value, thus .obtaining a
isolate the oscillator from any load and to mechanically rigid mount. Dipped silver-
provide two output connections. Output Phase Detector mica capacitors are used in the oscillator
from Q7 is connected to the divider chain. A circuit to compare the phase of the circuits for temperature stability. All dc
Q6 provides output to the frequency two 62S-Hz signals is provided in a single and correction voltages are connected to
mUltipliers in the transmitter or receiver. IC, in this case a Motorola MC4044P, is the pc boards via feed through capacitors
shown in Fig. 27, the signals are applied to of the solder-in variety. A blank copper-
Frequency Dividers clad board, to provide shielding, is
the inputs of the phase detector, U5. Any
A stable 625-Hz reference signal is derived phase difference is detected and a dc mounted between the frequency divider
from the 5-MHz oscillator and' a fre- voltage that is proportional to this dif- board and the one containing the VCO
quency divider chai.. as shown in Fig. 28. ference i~ developed. A dc amplifier is and phase detector. The boards are
,A hex inverter, SN74H04, is used as a built into the IC to bring the voltage up to fastened securely to each other and to the
crystal 'controlled oscillator. CS is a trim- a useful level. Q3, with R3, R4, RS and C4 chassis by threaded metal posts at the
merc.apacitor to provide a means of ad- forms an active filter that helps to corners.
justing the frequt;l)cy to zero beat with eliminate any tendency to lock up on har-
WWV. The oscillator is followed by three Adjustment and Operation
monics of the input frequencies. The cor-
. divide-by-IO stages, then a divide-by-4 rection voltage is filtered additionally by As complex as the circuitry for the
and a divide-by-2 stage for a total division L'2 and L3 in the VCO circuit. synthesizer is, it requires few adjustments
of 8000. The 625-Hz output from this cir- during operation. The S~MHz oscillator
cuit is applied to one input of the phase Construction can be set to zero beat with WWV by
detector. Construction practices with any VFO adjusting CS. If the reference-oscillator
OSCILLATOR
+5V
AND BUFFER
9'r---------~~--------------------~------------------------------
625 Hz
r
~..I::;~L~:....-!.f...L.:.:-' 12OUTPUT
C5 14
2
14 14 14 14 TO
U7 . PHASE
3 U6 U8 U9 Ul0 Ul1 e DET.
HEX DECADE DECADE eECADE DUAL FF (FIG.4-40)
INVERTER COUNTER CoUNTER COUNTER FF 7
4
6 ';-10 12 ';-10 12 ';-10 12 ';-4 6 14
5
68pF
S.M. 7
150
TOP VIEW
o
U6 THROUGH Ul1 Ul1A NOT USED
S.M .• SILVER MICA
1234567
13-18 Chapter 13
the collector should have extremely low de
15 resistance. Even a drop of 1 volt can signi-
1W ficantly reduce the power output of the
EXCEPT. AS INDICATED, DECIMAL amplifier. A good goal is less than 0.5-volt
VALUES OF CAPAC ITANCE ARE drop fr()m the car battery to the transistor
IN MICROFARADS (pF) ; OTHERS collector. With operating currents of
~"",-,---o+13.6V
ARE'IN PICOFARADS (pF OR J/pF);
RESISTANCES ARE ·IN OHMS;
several amperes, the total dc resistance
CIO
k -1000. M·I 000 000 500 should be only a fraction of an ohm. A
F.T:
standard commercially made heat sink is
used for the 50-~att amplifier, and it is
L2
30nH 146MHz
adequate for amateur communications.
L3 Forced-air cooling across the heat sink
should be used for any application requir-
ing longterm key-down operation at 40
watts or more of output.
Tune-Up Procedure
Generally, the best way to tune a tran-
sistor power amplifier is for maximum rf
power output. If this approach results in
exceeding the power ratings of the tran-
sistor, then the power output should be
reduced by reducing the drive level, not be
detuning the final. In the' case of an out-
board P A stage, such as decribed here,
both the input and output networks can
be tuned for maximum rf outpltt, if the
driving source has an output impedance
of approximately 50 ohms. However, a
better procedure consists of tuning the
output tank circuit for maximum rf out-
'put and tuning the input circuit for
minimum SWR as measured between the
exciter and the final amplifier. This tune-
up procedure has the added advantage of
assuring that the amplifier presents a
RELAY TOP VIEW RELAy SIDE VIEW 50-ohm load to the exciter. A dc ammeter
to check collector current is a useful tune-
up aid. Since tuning is for peak output, 'a
SHORTING
Monimatch-type SWR bridge is adequate
BARS REMOVE for the job. The best tuning procedure is
ADDED
CONNECTING
WIRES
to monitor simultaneously both output
(C)
power (absolute or relative) and the SWR
between the exciter and amplifier.
First, apply dc voltage with no rf drive.
No collector current should flow. Then
FIL TER
Ll
apply a low level of rf drive - perhaps 25
INPUT~OUTPUT percent or less of the rated 10 watts maxi-
mum drive - and tune the input network
;hC' ;hC2 for maximum indicated collector current.
(0) The networks may not tune to resonance
at this low drive level, but you should at
least get an indication of proper operation
Fig. 30 - (A) Diagram of the amplifier which provides 40' to 50 watts output. Capacitors are mica by smooth tuning and lack of any erratic
unless otherwise noted. The heat sink is a Thermalloy 6169B, Allied Electronics no. 957·2B90. (B) behavior in the collector-current reading.
COR circuit. Capacitors are disc ceramic. (C) The COR relay is modified by removing the connect· Gradually inc~ase the drive until full
ing wires from all four wiper arms and adding two shorting bars, as shown: Only'the stationary·
contact connections are used. (D) Pi·section output filter, C1 and C2 are 39·pF mica capacitors,
rated output is reached.
Elemenco 6ED3900J03 or equiv, and L1 consists of 2 turns of no. 1B tinned wire, 1/4 inch ID, 0.2
inch (61 x 5 mm) long (approximately 44 nH). K1 - 4pdt open·frame relay, 12·V contacts A Solid-State PA for 440 MHz
C1, C7 - 5· to BO·pF compression trimmer, (Comar CRD·1603·4S35 or equiv., Sigma Whether the application is by a person
Arco 462 or equiv. 67R4·12D also suitable), modified as de·
C2, C4, C5, C6, CB - Mica button, Underwood scribed above. using a hand-held transceiver as a mobile
J·101,· L1 - 12 nH, no. 10 tinned wire, 1·1/4·inch rig or· by an' experimenter building a
C3, C9 - 9· to 1BO·pF compression trimmer, (32'mm) long straight conductor. repeater, a medium-power, amplifier
Arco 463 or equiv. L2 - 30 nH, 1·3/4 turns, no, 10 tinned wire, would come in quite handy for increasing
C10 - Feedthrough type. 3lB·inch ID, 3/4 inch (10 x 19 mm) long,
C11 - Tantalum. L3 - 15 nH, no. 14 tunned wire, 3/4·inch (19· the effective range of his station. This arti-
C12 - Ceramic disc. mm) long straight conductor. , cle describes a compact, inexpensive,
D1 - 1oo·PRV or more, 500·mA or more silicon L4 - 2 turns oi no, 18 tinned wire 1/4·inch lO-dB gain power amplifier which is sim-
diode (Motorola 1N4001 or equiv.). ID, 0.2·inch (6 x
5 mm) long (approximately ple enough for nearly any experimenter to
D2, D3 - High,speed, low·capacitance 100· 44 nH). . ,
PRV silicon diode (Motorola MSD7000 dual 01 - Motorola silicon power transistor. build. This circuit originally .appeared in
package used here). 02 - Npn silicon Darlington transistor, HFE February 1977 QSTin an article by Olsen,
J1, J2 -Coaxial connector, panel mount. of 5000 or more, Motorola MPS·A13 or equiv. WA7CNP.
. f
Frequency Modulation and Repeate,. 13-17
RG-58A/U laying out the hole_ When' the hole has
been formed, the next thing is to ensure
that a good rf path is continuous from the
ground plane on one side to that on the
other. Connections can be made with cop-
/'
OUTPUT per or brass eyelets crimped and soldered
. CONNECTOR
MOUNTED
to both sides of the board. If no eyelets
IN VERTICAL
BULKHEAD'
are available, these connections may be
made by drilling a nO. 50 hole through the
board, inserting a piece of no. 18 wire
through the hole, and soldering and trim-
/'
INPUT
ming both sides flush with the board. Be
CONNECTOR 'sure that there is one such connection
MOUNTED
IN VERTICAL
made under each Unelco capacitor and
BULKHEAD alongside' the microstrip line and dc feed
point.
Next, the Unelco capacitors are
mounted as closely to the transistor
package as possible, and at the same time
THESE FOUR AREAS, REQUIRE AN ISOLATED CONNECTION' POINT.
they double as mounting surfaces for the
THIS CAN BE FORMED BY ETCHING A MOAT IN PC BOARD,OR BY transistor emitter leads. Connections to
ATTACHING A SMALL SQUARE OF PC BOARD ON TOP OF MAIN PC BOARD. the input and output lines may be made
with i- to 5-mil copper strap or foiL
Fig. 31 - Parts-layout diagram for the 50-watt amplifier (not to scale). A 4 x 6-inch (102 x The transistor can be mounted at this
152-mm) pc boc!rd is used as the base. time. The holes in the transistor flange are
Circuit Analysis
RFC 1
The circuit employed is essentially a
basic narrow-band amplifier capable of
being tuned over a broad range of fre-
quencies - 430 to 450 MHz_ Input-match
and collector-load transformations are ac-
complished by using multiple L sections
comprised of 50-ohm microstrip-line and
mica-compression variable capacitors.
The ac~ive device is the Motorola MRF618 !SO-OHM
OUTPUT
- an internally matched, 12.5-volt,
controlled-Q transistor designed for ap-
plication froJl1 420 to 512 MHz_ J
Construction
Fig. 33 - Schematic diagram of the 15-watt amplifier.
The amplifier is built on double-sided Cl, C7 - 0.9- to 7-pF mica compression Cl0 - 680-pF feedthrough capacitor.
G \0 glass-epoxy board _ Fig_ 34 is a I: I trimmer, Arco 400. Cll - l-"F, l5-V tantalum.
drawing of the board layout. Care should C2, C5 - Unelco l5-pF mica. Lt, L4 - 50-ohm microstrip line, 2.~inches
be taken in etching to maintain the line C3, C4 - Unelco 25-pF mica. long, 0.110-inch wide.
C6 - 3- to 35-pF mica compression trimmer, RFCl - Ferrite bead on cold lead of L2.
width of the microstrip at 0.1 \0 inch for a Arco 4 0 3 . ' RFC2 - 8 turns no. 22 enam., 1I8-inch 10,
Zo of 50 ohms. C8 - 0.D18 "F chip capacitor. ATC or equiv. close wound.
After the board has been etched, the (a 250-pF Unelco mica or a O.OOl-"F Erie RFC3 - 4 turns no. 22 enam., 1I4-inch 10,
first step is to cut the hole in the board for Redcap may work as a substitute). close wound.
C9 - O.l-"F disc ceramic.
the transistor heat sink (flange). The tran-
sistor flange can be used as a stencil for
Fig. 32 - Photograph of the completed Fig. 34 - A parts-placement guide for the amplifier board. Be sure to provide rf-connecting paths
430-450 MHz amplifier. between the top and bottom ground surfaces, as explained in the text.
13-18 Chapter 13
. and -output, and tune until the des'ired
12.0
operating conditions are achieved. Fig, 35
shows examples of typical data taken in
11.5
VCC<13.5V the lab. You will find that operating fre-'
11.0 PoUT·15W ~
V;
25
20
./'
V - r--
PsAT coverage that is superior to that obtained
by most stations. A major improvement is
usually found when a repeater is used be-
..
f-
f-
~
15
I
Fig. 36 - This typical 144-MHz amateur
repeater uses GE Progress·Line transmitter
tween vhf mobile stations, which normally
are severely limited by their low antenna
>- 10
V VCC-13.5V
10-450 MHz-
and receiver decks. Power supplies and meter· heights and resulting short communica-
ing circuits have been added. The receiver tions range. This is especially true where
"
0
I located on the middle deck is a 440·MHz con·
0-
5 I rough terrain exists.
I I trol receiver, also a surplus GE unit. A
I preamplifier, similar to that shown in Fig. 30, The simplest repeater consists of a re-
2345676 has been added to the 2-meter receiver to im· ceiver with its audio output direciIy con-
ORIVE POWER (WATTS) prove the sensitivity so that a 0.2-flV input
AMPLI FI ER TUNED AT PSAr nected to the audio input of an associated
(Sl signal will produce 20 dB of quieting.
transmitter tuned toa second frequency.
But, certain additional features are re-
Fig. 35 - These gain and output·power graphs quired to produce a workable repeater.
show what performance can be expected from These are shown in Fig. 37A. The "COW'
an amplifier utilizing the MRF618.
or carrier-operated relay is a device con-
nected to the receiver squelch circuit
which provides a relay contact closure to
key the transmitter when an .input signal
of adequate strength is present. As all
made to clear no. 4-40 screws. Drill and amateur transmissions require a licensed
tap two no; 42 holes in the heat sink, using operator to control the emissions; a "con-
the flange as a drill guide. Next, clean the trol" switch is provided in the keying path
heat sink and bottom of the transistor so that the operator can exercise his
flange so that foreign matter will not pre- duties. This repeater, as shown, is suitable
yent the transistor from seating properly for installation where an operator is pre-
on the heat sink. Apply a very sm'all sent, such as the home of a local amateur
amount of thermal compound to the with a superior location, and would re7
flange and bolt it firmly to the heat sink. quire no special licensing under existing
The transistor leads may then be soldered rules.
to the circuit. In the case of a repeater located where
The base-return choke and dc collector- no licensed operator is available, provi-
feed circuit may now be put into pla~e. sionsAnust be made to control the equip-
C 10 may easily be installed by drilling a ment over a telephone line or a radio cir-
/
0.192-inch hole in a small copper strap OR/ cuit on 220 MHz or higher. Fig. 37B
/
(0.02-inch thick) and then putting a tight- / sh,ows the simplest system of this type:
angle bend in the strap so that the capaci- -----~ The control decoder may be variously
TELEPHONE LINE (8)
tor can be mounted upright above the designed to respond to simple audio
board. Be sure to make the base return tonesl dial pulsed tones, or even "Touch-
and dc-feed connections as close to the Fig. 37 - Simple repeaters. The system at A is Tone" signals. If a leased telephone line
transistor package as possible. for local control; remote control is shown at B. with de continuity is used, control
Now comes the remaining turning ele- voltages may_be sent directly, requiring no
ments and output dc-isolation capacitor. decoder. A three-minute timer to disable
Fig. 34 shows the mounting position. Care the repeater transmitter is provided for
should be taken to solder the ground taps fail-safe operation. This timer resets dur-
of the trimmer to the ground plane to ing pauses between transmissions and
reduce the amount of lead inductance in" may be used, ranging from a piece of coax does not interfere with normal communi-
herent in these capacitors, at uhf. C8 to RCA phono plugs. cations. The system just outlined is
should be mounted after cutting a small suitable where all operation is to be
break in the output line near the end. This Tune-Up through the repeater and where the fre-
capacitor need not be a 0.018-J.lF chip, but Tuning is simple: Apply low power quencies to be used have no other activity.
care should be taken in choosing a substi- (about 3/4 watt) to the input and tune the
tute that will not grossly affect the output input capacitor until a small amount of Remote Base Stations
load characteristic of the circuit (see parts collector current begins to flow. Then The remote base, like the repeater,
list). Depending upon your application, tune the output capacitors for peak out- utilizes a superior location for transmis-
any 50-ohm outside-world connection put. Switch back and forth between input sion and reception, but is basically a
CONTROL
STATION
across the U.S. and Canada. Some
lO-meter repeaters are operational, and
most use loo-kHz separation. A popular
arrangement on 6 meters uses 52.525
MHz as either the input or output, with
several choices for the other half. Many
(B) stations, however, are moving toward
either a 6oo-kHz or I-MHz offset. On 2
meters the standard is 600 kHz. The
Fig. 38 - A remote base is' shown at A. A repeater with remote·base operating capability is 220-MHz band uses 1.6-MHz separation.
shown at B. Control and keying circuits are not shown. Telephone·line control may be substituted On the 450-MHz band it is 5 MHz. The
for the radio-control channels shown. - ,
choice and usage is a matter for local.
agreement.
In some cases where there is overlap-
simplex device. That is, it transmits and The audio interface between the re- ping geographical coverage ef repeaters
receives on a single frequency in order to peater receivers and transmitters can, with using the same frequencies, special meth-
communicate with other staticlns also some equipment, consist of a direct ods for selecting the desired repeater have
operating on that frequency. The operator connection bridging the transmitter mic- been employed. One technique requires
of the remote base listens to his hilltop rophone inputs across the receiver speaker the user to transmit automatically a
receiver and keys his hilltop transmitter outputs. This is not recommended, how- 0.5-second burst of a specific audio tone
over his 220-MHz or higher control chan- ever, because 0f Jhe degradation of the at the start of each transmission. Different
nels (or telephone line). Fig. 38A shows audio quality 1 in the receiver-output
tones are used to select different repeaters.
such a system. Control and keying fea- stages. A cathode follower connected to Standard tone frequencies are 1800, 1950,
tures have_been omitted for clarity. In each receiver's first squelch-controlled 2100, 2450 and 2400 Hz.
some areas of high activity, repeaters have audio amplifier stage provides the best Because of growing congestion among
all but disappeared in favor of remote results. A repeater should maintain a flat same- al}d adjacent-channel repeaters, an
bases because of the interference to response across its audio pa~sband to increasingly popular access method is a
simplex activity caused by repeaters maintain the repeater intelligjbility at the continuous sub audible tone. Popularly
unable to monitor their outplit frequency same level as direct transmissions. There known as PL (Motorola trademark for
from the transmitter location. should be no noticeable difference be- Private Line), it must be applied to a
tween repeated and direct transmissions. transmitter after the clipperlfilter stage
A Complete System The intelligibility of some repeaters has shaped the voice audio. Table I lists
Fig. 38B shows a repeater that com- suffers because of improper level settings the EIA-standard frequencies.
bines the best features of the simple which cause excessive clipping distortion.
repeater and the remote base. Again, The clipper in the repeater transmitter Practical Repeater Circuits
necessary control and keying features should be set for the maximum system Because of their proven reliability,
have not been shown in order to simplify deviation, 5 kHz, usually. Then the commercially made transmitter and re-
the drawing, and make it easier to follow. receiver level driving the transmitter ceiver decks are generally used in repeater
This repeater is compatible with simplex should be set by applying an input signal installations. Units 'designed for repeater
operation on the output frequency be- of known deviation below the maximum, or duplex service' are preferred because
cause the operator in control monitors the and adjusting the receiver audio gain to they have the extra shielding and filtering
output frequency from a receiver at the produce the same deviation at the repeater necessary to hold mutual interference to a
repeater site between transmissions. The output. Signals will then be repeated minimum when both the receiver and
control operator may also operate the linearly up to the maximum desired transmitter are operated simultaneously_
system as a remote base. This type of deviation. The only incoming signal that Wide-band noise produced by the
system is almpst mandatbry for operation should be clipped in a properly adjusted transmitter is a major factor in the design
on one of the national calling frequencies, repeater is an overdeviated signal. of any repeater. The use of high-Q tuned
such as146.52 MHz, becaus~ it minimizes The choice of repeater input and output circuits between each stage Qf the trans-
interference to simplex operation-and per- frequencies must be carefully made. In mitter, plus shielding and filtering through-
mits simplex communications through the general, check with the appropriate out the repeater installation, will hold the
system with passing mobiles who may not volunteer frequency coordinator, who is wideband noise to approximately 80 dB
have facilities for the repeater-input fre- listed in the ARRL Repeater Directory, below the output carrier. However, this is
quency. since about 4000 repeaters are operating not sufficient to prevent desensitization -
13·20 Chapter 13
DELAY SWITCH
H ......-<>-12V
+12V
Z
f-
f-
c(
Rl
5M
TIME
SET
20 40 60 80
VERTICAL SPACING (FEET)
(A) INPUT
(Al \ (81
Fig. 40 - (A) COR circuit for repeater use. R2 sets the length of time that K1 will stay closed
Z after the input voltage disappears. K1 may be any relay with a 12·volt coil, although the long·life
f-
f-
reed type is preferred. 01 is a silicon diode. (B) Timer circuit using a 5ignetics NE555. R1, C1 set
c( the timer range. C1 should be a low·leakage type capacitor. 51, 52 could have their contacts
paralleled by the receiver COR for automatic START and RESET controlled by an incoming signal.
Rl (--0TO HIGH
50k 047 IMPEDANCE
"'---+,r--o . TO
CARBON
"'+r-oINPUT
5)JF
15V~INPUT
MIKE
'r---"--~ ~~~~~T~
(e) (D)
Fig. 42 - Typical connections for the encoders manufactured by Western- Electric' (A) and
Automatic Electric (B). If low·impedance output is needed to drive a carbon·microphone input; the
circuit at C can be employed for either encoder. Likewise, the circuit at 0 will provide a high·
impedance output. Rl can be any miniature composition control; the types made for mounting on
circuit boards are ideal.
\
Fig. 41 - This Western Electric Touch·Tone
encoder has been mounted in a 4 x 4 x
2·inch {102 x 102 x 51·mm) utility box (Bud
AU·1983). Encoders are sold by most telephcme
supply houses, including Telephone Equipment
Co., P. O. Box 596, Leesburg, FL 32748. Tel.
904·728·2730.
Q----rp----rp BLUE
I I 820
I I ~
I • I I
12-button encoder pad consisting of digits
zero through nine and symbols pound #
and star '" are employed. A typical en-
coder is shown in Fig. 41, and the connec-
&---¢----$ .
I
I
I
I
I
I
tions for pads manufactured by Western
Electric and Automatic Electric are shown G----m---~ 02
in Fig. 42. The telephone pads will work
GREEN.
~RN.
with as little as nine volts or as much as 24
volts dc applied. Either high- or low- ~
--<>---<>
impedance output may be employed, as WHITE/BLUE
1J~
shown in Figs. 42C and D.
A circuit diagram of a typical tele-
phone-company pad is given in Fig. 43.
Individual models will vary slightly, but WHITE
BLACK
the basic circuit used in all models is the
same. A single transistor produces two
tones. Two LC circuits are used, one for
the high tone group and one for the low
tones. Some people are bothered by the
use of it single transistor to generate two
audio frequencies, so the lower tone can
be considered the frequency of oscillation
while the high tone is called a parasitic Fig. 43 - Diagram of a typical Western Electric Touch·Tone generator. T1 and T2 are special
m~lti·winding·transformers manufactured by Sangamo Erectric and others. 01·04, incl. are silicon
oscillation, for purposes of explanation. varistors.
A Homemade Touch-Tone Encoder
To be compatible with all repeaters and
telephone systems, a T~uch-Tone signal from the rig ,it is used with; there is no necessary to hang onto the push-to-talk
.must be accurate and stable in frequency, need to depend on separate batteries for switch while fumbling with the Touch--
and have a nearly sin~soidal waveform. power. When a tone pair is selected by Tone pad, and there are no squelch tails
Simpler encoders than 'the one described pressing the keyboard switch, the trans- between digits. It has a low~impedance
here can be built, but they will not provide mitter is automatically keyed. When the audio output which is electronically
its high performance. key switch is released, a delay timer keeps disconnected from the transmitter audio
This encoder features internal voltage the transmitter on long enough for the system when no keyboard switches are
regulation, allowing power to be taken next tone pair to be selected. It's no longer pressed. The encoder may be connected to
\
13-22 Chapter 13
the mic input of transceivers having either
15M
high- or 'low-impedance mic inputs -
with negligible loading of the transmitter
audio circuitry. The audio frequencies are
crystal-controlled, meaning there is no 9 10
drift. This circuit was originally described RI 3
10k
'0·
before reaching emitter follower Ql. QI TOP VIEW
o o
disadvantage of the scanning decoder is
Fig. 46 - Circuit board etching pattern for the Touch·Tone encoder. The board is single sided, the slow response time resulting from the
shown at actual size from the foil size, with black representing copper. need to wait for each decoder to find tbe
received tone. Also, a delay is built in
which requires both decoders to halt for at
least one full clock period before a digit is
registered. The operation is thus relatively
impedance dc voltmeter, and a scope if The mechanical details of the installation immune to spurious responses from voice
'one is available. are left to the discretion of the reader. signals, yet takes one-half second or less .
First ensure that 01 is regulating the en- Since the encoder will not load the to respond properly to any digit.
coder supply voltage at +5.1 V dc ± 10 audio system, it should not be necessary The type 567 tone decoder is not
percent. UI may be damaged if more than to change the setting of any transmitter satisfactory for use in this. circuit because
6 volts is applied to pin 16. mic-level controls. Adjust only Rl in the neither side of the frequency determining
The two operating states for the en~ encoder for proper tone deviation. The RC network is grounded. A Motorola
coder are (1) no keyboard buttons prototype unit has provided excellent per- MC13IOP was tried because\one had been
depressed and (2) one or more buttons formance on 'both a Tempo VHF/One used previously for tone decoding and was
depressed. Connect the 12-V dc supply 2-meter rig and a Kenwood TR-8300 uhf found to work well in this frequency
and measure the voltage at the test points rig. range. Its intended use is as a phase-
shown on the schematic diagram. locked-loop fm stereo decoder. In this
Voltages measured should be hi accord- A Scanning Touch-Tone Digit and Word application it locks onto the 19-kHz pilot
ance with those shown. Decoder tone which is present, along with the
If any voltages are incorrect, look for The Touch-Tone encoding system, used audio signal, and turns on an open-
wiring errors. If the collector voltage of extensively in autopatch operations on fm collector output to light a stereo indicator
, Q2 is not at least 4 volts with no buttons repeaters across the country, offers a lamp. Its internal oscillator runs at 76
pressed, the problem may be a leaky tran- ready-made source for dual-tone codes, kHz, and an internal frequency divider
sistor at Q 1 or Q2. If a scope is available it and advances in microcircuitry design gives the 19 kHz for the pilot tone'
may be used to inspect the audio output. have produced a single device that can be detection. A 19-kHz monitor output is
Pressing anyone button should produce a used to decode these dual-tone codes for a provided. In the present circuit the
signal, while depressing any two buttons varie5Y of remotely controlled functions. oscillator is run at four times the
simultaneously should produce a single However, one device is required to decode Touch-Tone frequencies, and the stereo
tone. decode~ function (except for the indicator
each tone. This section describes how a
scanning decoder evolved as an attempt to lamp output) is ignored .. Since the
Installation avoid using seven of these decoder ICs, frequency determining resistor runs to
The electrical portion of the installation and how a simple counter circuit can ground, it is programmed easily for
simply involves running four wires from recognize specific four-digit word se- scanning operation by using four re-
the encoder to the transceiver: + 12 V, quences to provide a unique approach to a sistance values and four, open-collector,
ground, push-to-talk (PTT) and audio . remote-control decoder. 15-volt NAND I gates (7426).
output. Shielded audio cable is recom- There are 16 tone pairs possible -
mended for the audio output; which is selecting one from the low group, .697, Digit Decoding
connected to the transceiver mic input. 770, 852 and 941 Hz, and one from the The digit decoder is shown in Fig. 47. A
The PIT lead is connected to the hot side high group, 1209, 1336, 1477 and 1633 Hz. 12-volt power supply is required for the
of the mic PTT switch. The + 12-V and Two phase-locked-loop types of tone MC13IOP circuits. The: voltage on the
ground leads are self-explanatory. The en- decoders should therefore be sufficient if ·programming resistance network is some-
coder PTT circuit is designed for rigs with each one sequentially scans the four tones what high for the usual open-collector
an antenna relay coil which is connected of one group. In this way two decoders NAND gates, so a 7426 is used. The
to the + 12-V ,bus and the PTT .switch. with some added scanning circuits take O.OI-and 0.01 5~~F capacitors should be
The latter grounds the cold side of the the place of eight. Parts of the scanning mylar or silver mica for temperature
relay coil during transmit. Assure that circuit such as the clock oscillator and stability, and some experimentation with
your rig has this type of PTT circuit and digit decoder would be required in any the resistance values to achieve the correct
.' that the relay coil draws less than 300 rnA. case for word decoding, and the parts' frequencies may be necessary. A fixed
I believe all the popular vhf and uhf fm cost of the present system using primarily resistance of 10 kO or 12 kO was used and
ham rigs have this type of PTT circuit. low cost TTL logic is reasonable. One a jumper-wire or selected value of fixed
13·24 Chapter 13
resistance was inserted between all of the Each prefix decoder consists of a dual J-K and ease of use becomes an important
l000-(>hm potentiometers. The circuit flip-flop (UlS) two AND gates (U20A and factor in choosing codes. Often a single
time constants resulting from the use of B) and two NAND gates (U21A and B). digit is better to turn something off
1-,.,.F ceramic capacitors appear to be the The prefix of first threedegits of the four- because it's quicker and less likely to be
correct value for the present system. The letter word being decoded are selected by forgotten. In any case, 'all system codes
SZO-ohm pull-up resistor from the + 5- connecting inputs digit I (DG I f, digit 2 must be mutually compatible.
volt supply makes the output TTL com- (DG2), digit 3 (DG3) and digit 4 (DG4) to
patible; the monitor output is correct for the desired outputs from U12 and U13. Construction
driving TTL devices. Likewise, the other word-decoder inputs The unit was constructed on double-
The 555 timer U9 and .inverting gate - DHI, DH2, DH3 and DH4 are connected sided, copper-clad pc board. Thetone and
U7D provide a positive clock pulse for all to four outputs from U12 and U13. The digit decoder circuit of Fig. 4S was con-
the 7473 flip-flops which toggle on the first four-letter word such as the sequence structed on a 6 X 6-inch (152 X 152-mm)
trailing edge Qf the pulse. When no tone 4639 would turn U22 on, and the second pc board. All of the components are
inputs are received, U3 and U4 count such as I * S # would turn it off. mounted on the top side of the board and
through four states each and cause the This sequence detection is achieved by are soldered on both sides of the pc board.
open collector gates U5 and U6 to the gating used on the J-K inputs of UIS The value of R I through R6 in each tone-
conduct in sequence A through D, thereby and U19. Each is a two-stage counter selection line should be selected so that the
sweeping the frequencies of Uland U2 _ which will advance in the state sequence potentiometer will'tune the circuit to the
upw(lrd through the low and high tone o - I - 3 - 2 - 0 only if the correct digit proper tone in the middle of its resistance
groups respectively. When either tunes to input is on in proper sequence. That is, in range. The values shown in the circuit
an incoming tone, it becomes phase order to advance from 0 - 1 DG 1 must be diagram of Fig. 4S are typical and can be
locked to it, its output at the test point on, to advance from 1 - 3 DG2 must be used as a starting point for selecting the
goes low, tne counter is stopped because on, and to advance from 3 - 2 DG 3 must final value. The word decoders are also
its J and K inputs are low and the detector be on. If any are off when they should be constructed on pc board w.ith a double-
remains locked on the incoming fre- on, the state goes directly tQ zero. When sided layout: The decoder board is 4 X 4
quency. Also, the monitor output can pass state 2 is reached, UISA is off and UISB inches (102 X 102 mm) with all of the
through gate U7B or U7C and can be is on, and two of the three AND J inputs components installed on the top side.
used for exact measurement of incoming _ of U22 are on. At this point the three-digit Several of these decoder boards can be
tone frequencies. When both tones are so prefix has been received successfully. If stacked and will make the addition of con-
detected, a logiC-one condition appears at the fourth digit received corresponds to trol functions a simple task.
the output of NOR gate U7A and counter DG4, the clock pulse (CLK) will also turn
US is permitted ·to advance from its on U22 since all its J inputs will then be Alignment
cleared condition. on. Similarly, the second four-digit-word Alignment is a matter of setting the
The J-K flip-flops in US are wired to sequence will turn off U22. Further scanned frequencies to the correct values
advance in count through states 0, 1, 3, 2, 2 interfacing between the TTL output of using a frequency counter connected to a ,
and become stopped in state 2 (USA off, U22 and a controlled system will depend monitor point. The associated test point is
USB on) until reset When one or the other upon its nature. A simple relay driver grounded, and with no input to the
tone detector drops out. Its purpose is to using two parallel-connected I5-volt decoder one of the frequencies can -be
provide a two-clock-period double check open-collector buffer'inverters (1/3 of a adjusted, depending on which of the four
on the decoder operation and yield a 7416) and a I2-vol~, 150-ohm relay is tones in any group happened to be on
single clock pulse (CLK) just before the shown. A small silicon diode connected as when it was stopped manually. It is best to
end of state 3 if the tone signal is so shown helps to avoid transient problems. stop the highest tone first and adjust the
validated. Also, during state 3 (USA and simplified by using the same prefix de- first potentiometer, along with R9 if
USB on) the decoder U 11 is enabled, and coder for both turn on and turn -off; only necessary, to get 941 Hz. Then select the
one of the digit outputs from U 12 or U 13 the fourth digit need be different. One second potentiometer and value of R 1
comes on. must be sure the word decoder is reset until S52 Hz can be tuned. In a like
The decoder makes use of the counter before ft will respond properly to a four- manner, adjust all-eight frequencies to the
states of U3 and U4 when they are digit word. This is just a matter of being correct values. After several months of
stopped by an incoming two-tone signal. sure that any code such as 4639 is not operation the response became sluggish
Since the tones are scanned from low to preceded by a 4, a 46, or a 463. If in doubt and finally the unit stopped working, but
high and the low group (top to bottom about what the last digit may have been in original performance was restored by
rows on a standard pad) is wired to the the system, an extra random digit other retuning R 1 through R6. Satisfactory
two least-significant-digit inputs on UIl, than 4 is generally sufficient. Alternative- operation is obtained with input levels
the output states of U 11 - would cor- ly, the reset inputs of U IS and U 19 can be between 0.1 and 1 volt ac.
respond to the tone button assignments of wired to some completely independent
a standard Touch-Tone pad. In order that source of a reset such as the carrier-input A Universal Touch-Tone Decoder
the digit outputs are correct for the actual detector. The control unit described here is ex-
assignment on the tone pad, the U II A few words about the choice of codes. tremely reliable, flexible and immune to
outputs are reordered and the 12 cor- For most Amateur Radio applications, the false signals. Any number of control func-
responding to the commonly used 12- four-digit word provides adequate se- tions can be built into this modular unit.
button pad are inverted to the positive curity. If a great deal of phone patch Starting with a simple"single-digit, on/off
logic form by U 12 and U 13. The acticity is present on the channel, the control, it'may be expanded to provide up
four gates U14, UI5, UI6 and UI7 are characters * or fI should be used in the to 45 different contra) functions, in-
optional and are used to obtain the binary code, as these do not appear in phone cluding a three-digit on/off command.
equivalent of the standard digit as- numbers. The four additional characters The application of the decoder system
signment of the Touch-Tone pad. generated only by a 16 button pad can be described here is not limited to repeater
used if four additional inverters are added use. With a little ingenuity one might
Word Decoding to the group in UI2 and U13. When a adapt the simpler systems to turn on
The word-decoding circuitry shown in number of four-digit codes are used to house lights or open garage doors.3 And
Fig. 49 consists of two', three-digit prefix operate a remote system of some sort, it there is also a voltage-to-frequency con-
decoqers and an output flip-flop U22. becomes difficult to remember them all, verter. More about that later.
13-2) Chapter 13
that pin 8 will go low when the tones are decoder without the latch. Connect the An Expanded System
received' and remain low until the base of transistors (Q 1, Q2) as shown in Fig. 52.
Ql is ungrounded or taken high. A relay is shown at the output although Next, proceed to construct a complete
.. To become even more familiar with the either a relay or transistor may be em- system offering up to 45 different control
circuitry, a s'imple '" (on) and # (off) ployed. When the'" is received, the relay functions and using a three-digit entry.
decoder will be constructed. Build the '" will close and remain closed (latched) until Use of either the'" or # sign as the first en-
decoder with' the latch function and the # a # signal is received. try for. a control function is recommend-
!.I. K
3
JI2
7473
71
~5
elK
~~
~A
~.,,,
10
U70
9
3 U9
555
. 7
39k
+5V
EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPAC ITANCE ARE
IN . MICROFARADS ( pF) ; OTHE RS
ARE IN PICOFARADS (pF OR
RESISTANCES ARE IN OHMS'
k -1000. Mo l,OOO,OOO
,elK U8A
Q JQI IJQ U88
4
r
+5V e e +5V-- 1
6 820
7402
Ii
4
13 14 8
1 6,h 4 1~ 7440 '~:F 9
~) vee 13
12
~
U7A
GNO
~
vee
~NO:/
'-6
1m12
13
U108
8
ClK
,h
18 19 UI38
21 8 3 _4
e 7 -0 o
5~.
20 0
1 6
0 -0
U12E
5 11 ._10
4 -0 2
8
9 ~4
12-0
U12B
2 3 ._4 ~
1 ~ 4
U120
5
.El
UI3F
9~ 5
+~
24
9
10 13 12
~ 6
~04
r+:.
3 1 2 -0
2 7
Ul1 UI3C
7 5 6
6 ~ 8
U13E
10
Ii I~-<l 9
U12F
4 13 12
3 ~ *(10)
II
13
~ -0
9 8 #(11)
14
12
+5V
h47~F
23 A 15
13
22 B
16
14
0p50V
17
15
+r \ 6 3 4 5 2 1 1112 3 4 5 1 2 6 1112f 11 I 212 3 4 5 6f 5 11 612 4 3 2 II
I
vee
U3
GND
4
I
vee
4
GND
U12,U13 47m47+"47+"~+" 8
vee
7430
8
BeDB,
vee
8
vee
BCOO
8
vee "
BCOA BCOC
DG1 o----+-----'-irv:r:r;-""
U2BB
.-------'3'-1 4'
+5V
DG 2 Q - - -.....--lf--"'-I
n.c.
DG30-------------------~--------------~
+5V
DH1 o----+------'-={
7 J Q 9
U19B
+5\1
DH20---~--+_~ ____~
DH3O--------------------------------+--~
+5Vu-~"",",-
820
Fig. 49 - Schematic diagram of the, relay-control pc board. Be sure to select relays that have contact-current capability for use in the desired ap-
plication,
ed, especially if the repeater is equipped "window" will remain open for only two To combine the three digits and
with an autopatch. In this manner, seconds. The ... is also used to reset all generate one output, a COMBINER
numbers alone cannot initi,ate a control other decoders. This was done so that if board is used. Each board contains iliree
function. Personal preference is to use the any of the decoders have been accidently SN74LSIO ICs. Fig. 55 shows the make-
... symbol to initiate the command and the activated, no command function will be up of a single 74LSIO which incorporates
# sign as an "all clear"; this also permits carried out. three triple-input NAND gates. Each board
system reset. It will also be necessary to invert all the furnishes nine functions; five boards are
To carry the logic 'required for the lows supplied by the decoders. This opera- used in 'a 45-function decoding system.
larger system, other circuit boards will be tion is performed by SN74LSOO quad, The last board required is (ap-
needed. All are of the 28-pin plug-in vari- two-input NAND gates. The input pins of propriately) the FINAL board, which
ety. A timer is also needed; an NE555 the gates are tied together, thereby contains the latches for the desired func-
serves nicely. The timer, activated by the ... creating inverters. The timer and inverter tions and the transistor drivers. Four
decoder, (constructed without a latch) will diagram is shown in Fig. 54. Only' one SN74L 73 dual J-K flip-flops and either
open a "window" through which the TIMER-INVERTER board is required 2N 1711 transistors are mounted on each
other two digits must be passed. This for any system of up to 45 functions. board. Five boards are used in the
13-28 Chapter 13
FREO. ADJ. EXCEPT AS Ili0ICATEO, DECIMAL OIGIT
RX RI VAL.UES OF CAPACITANCE ARt!:' FREQUENCIES VALUE OF RX
+~V IN MICROFARADS (JIF I ; OTHERS +~V
Fl 1209 1336 1477 F2 FREQ. VALUE
5k ARE IN PICOFARADS CpF-OR JlJIFI;,
(SEE
T,CI
rf-?0"
TEXT)
TPI RESISTANCES ARE IN OHMS;
k -1000. M. I 000 000
D 0 0 697 1209
1336
6.8k
5.6'
AL.L. RESISTORS 1/4W
CI- e4 ARE MYLAR
0 0 0 770 1477
697
5.tk
13k
GJ [!J 0 852 770
852
12k
10k
,..__o---"I/\J'I.,.........-i-i. 0 GJ 0 941
941 9.1k
~~~~~~~--------------~----------lr----~----~OUTPUT
2 +' 02
,l, C6
2.2}1F/50V
R5
15k
IN914 *
04
+ IN270**
C7
,l4.7}1F/16V
* PARTS FOR LATCH FUNCTiON
* * PARTS FOR UNLATCH FUNCTION
Fig. 50 - The circuit diagram of the Universal TOUCh-Tone Decoder. A pair of ICs is used to pro·
vide better reliability and immunity to "falsing." If desired, the capacitor at pin 1 of each IC may
be increased to 100 "F to provide a two·second decoding delay.-
Fig. 53 - With a little ingenuity, this simple
version of the decoder can find many uses .
S .r---- TO DESIRED
cr---I
L'RCU'T
r-----~.---~+5V
N.C. I 47k
I +12 TO +15V
I
~l OPT!ONAL
T lN914
I
W\r-+-+--..... - - - - - - .J
I.Sk
lk
non. 4
!~
OUTPUT I N PUT FROM D----'\f\/I..r----'
5 U2B COMBINER BOARD
+ . * CAUTION - 00 NOT USE 74LS73
47pF
T
DECODER . toV
INPUTS ~ Fig. 56 - A diagram of one section of the FINAL board. FF1 and FF2 are both part of the same
IC, a dual J·K flip·flop. Four of these dual flip·flops and.elght of their accompanying output-tran·
sistors are mounted on each board. See Fig. 580 and the text concerning the installation of the
optional diod\! shown.in dotted lines.
01-04
4,IN4003 ,----0+ 19V TO RELAYS
~ +5V TO ALL
.~ _., ~-- ~-
INVERTER - 74 LS00;L+ 47pF
-
CIRCUIT
SOARDS
IOV
(B)
(
I case. This allows everything to be discon- decoder and the third input by the "2"
INPUTS
FROM
nected easily for servicing. LEDs are" decoder. This forces the output of U3A
INVERTERS +-t-~---:~~ UIB _mounted on the front panel of the low. This low toggles -the J-K flip-flop
enclosure to provide an indication of the U4A (note that the clear or C input ef
\ 4
status of all of the functions. A Touch-
Tone pad was also installed on the front
panel. This pad may be switched into the
U4A is held high through RI). The Q out-
put of U4A will go high llfld remain high.
This causes transistor QI (the output tran-
. 3 ute system for local checks of the unit. sistor) to conduct. The corresponding
\
~ Refer to Fig. 50A. Note that the output LED will glow, indicating that the func-
of the'" decoder is used to start the timer tion has been carried out. If desired, a
*OUTPUTS TO FLIP-FLOPS
and reset all the decoders, The output of relay could be' used at the output of Q1.
U2 is applied to one input of the three- With this simple system, one must use the
Fig. 55 - there are three 74LS10 ICs on each input gate, U3A. The second input of the same codes ("', I ,2) to turn off the func-
COMBINER board although only one is shown. gate is satisfied by the output of the" I " tion. Tbis is not a sound idea since one
13·30 Chapter 13
+t2V
+t2V
*.1,2 ON *.1,2 OFF WITH #OFF
OUTPUT
U48
74L73
l FF2
_
B
=- t~ U38 8
N.c,
(8)
11
+t2V
+42 V
(A)
74L 73
Fig. 58 - The four circuits diagrammed here
FFI
are discussed in the text. The most simple cir·
74LS10
cuit (A) does have a weak point, wtlile that at 0
is the most reliable.
OUTPUT
the # key will ensure the function is in its *.1,2 ON #' OFF'
Appendix
XIO XIOO 'Ferroxcube components can be purchased from Elna
CAL.
CAL. Ferrite Laboratories, Inc., 9 Pine Grove St., Wood---
stock, NY 12498.
'Underwood mica capacitors are available from Alpha
.Electronic Laboratories, 2302 Oakland Gravel Rd.,
Xl0 Xl00 Columbia, MO 65201.
CIRCUIT CIRCUIT 'If such operation is intended, one should pay par-
(8)
ticular attention to sections 19.34(b), 19.35(c),
(C)
97.61,97.89 and 97.99 of the Communications Act
of 1934. See The Radio Amateur's License Manual
ARRL, $4.
Fig. 59 - The voltage-to-frequency converter is shown at A. Both dc and ac voltage amplitudes 'FM and Repeaters for the Radio Amateur, first
may be read out on a frequency counter at the output of the IC. The circuits at Band C may be edition, p. 119.
used as x 10 and x 100 multipliers at the input to the converter. 'Circuit boards and parts kits are available from
Circuit Board Specialists, P. O. Box 969, Pueblo,
CO 81002. Printed circuit board etching patterns
are available from the ARRL for 50 cents and
an s.a.s.e.
AC OUT
DC IN
ALL TRANSISTORS
2Nt711 (10)
Fig. 60 - Half-size component placement guides for the pc boards. Clockwise from upper
COMBINER. See note 5 of the appendix.
13·32 Chapter 13
DECODER' BOARD
DECODER BOARD CIRCUIT BOARD SPECIALISTS PO BOX 969 PUEBLO COLO 81002
N
o
o
;;;
.,;
u
o
~
...
'.,."
.,.'"
)(
o
GO
o
...
'">-
'"
~
«
...~
'"c
'"«
o
GO
=
"u
'"
u
, /
13·34 Chapter 13
Chapter 14
Specialized Communicati.ons
Systems
•
nonradio devices such as microcomputers
and telephone lines. .
Use your imagination; see what you can
do. In addition to the surveys presented
here, some good resources are the ARRL
Operating Manual, QST and AMSAT's
Orbit Magazine.
SATELLITE COMMUNICATIONS
Large antenna arrays and comparative~
lysophisticated equipment may rule out
some forms of space communications for
many amateurs, but the amateur satellite
program puts the excitement of vhf DX
within the reach of every amateur: The
possibilities grow with every new launch!
It began in 1961 with the successful
launch of OSCAR I, the world's first
nongovernmental satellite. OSCAR rode
"piggyback" on a regular launch, with
the tiny ham satellite serving as ballast.
The small breadbox-size satellite, built Fig. 1 - After several years of work, the first high·altitude Phase III spacecraft was prepared for
launch on May 23, 1980. As thousands of amateurs listened to the launch net, the Aria'ne launch
from out-of-pocket expenses, of $64, vehicle fell into the Atlantic soon after liftoff. Work began almost immediately on a Phase III B
transmitted the Morse code letters "HI" satellite. (photo courtesy W4PUJj
batteries were finally exhausted. The near- the effectiveness of satellites in pinpointing :!
:; 800 I
ly identical OSCAR 2 transmitted signals emergency locator transmitters ,similar to
for 18 days after its launch in 1962. those that are carried aboard downed air-
UJ
r
:::> 700 /
The first active communications craft, and in relaying complex medical data ~ 1/
VI 600
~
satellite in the OSCAR series was OSCAR such as electrocardiograms' from coast to
g500
3, launched in March, 1965. More than coast, simulating in-transit work from a
100 stations in 16 countries helped make disaster site. Finally, when OSCARs 6 and
UJ
g400 /
satellite history with OSCAR 3, the first 7 and 8 have been in close proximity, r
/~
free-access satellite. OSCAR 4 followed satellite to satellite linkups have been ~300
later that year and achieved another com- achieved: another milestone for free ac-
UJ
t: 200 / MAP RANGE·
;!~ COSo, R:H'-
munications first despite a bad orbit - cess satellites. In the future, with the
.J
.J
UJ
/
.R ·EARTH RAD.(3960STAT Mt)
!;;100
the first U.S.-to-U.S.S.R. satellite con- greatly extended access time and insignifi- 1/ I
H'YTEL1LlTE ALTJ(STAT MI.)
tact. OSCAR 5, constructed by a team at cant Doppler shift near AMSAT Phase
VI
o I I
i
Melbourne University, was launched in Ill's apogee, and with the flixibility that o I
500
1000
1500
I2000
2500
3000 I
1970, the first launch coordinated by a results from its sophisticated on-board M'AP RANGE IN STATUTE MILES
newly formed group of amateurs' in the computer, many more experiments are
Washington, DC area: AMSAT. The planned in areas of remote-store-and- Fig. 2 - Satellite altitude above earth versus
spacecraft internal batteries powered its forward data transmission, computer ground station map range (statute miles).
2-meter and lO-meter beacons for more software exchange and emergency com-
than six weeks. munication.
AMSAT-OSCAR 7 joined the series late In recognition of the experimental and
YOUR CONTACT COMPLETED
in 1974, providing another major step for- educational potential of these free-access I OR NOT!
ward for the amateur satellite program: it is satellites, NASA has been most generous
scheduled to remain on continuously, alter- in providing "secondary payload" launch
nating between two different modes of opportunities to AM SAT to get its
operation. Though it has recently suffered OSCARs in orbit. NASA has a particular-
from periods of "falsing," occasIonal, un~ ly strong interest in the OSCAR Educa-
predictable mode switching, OSCAR 7 is tion Program, designed to bring satellite
still providing long range vhf communica- and space technology' into classrooms
CONTACT MAY
tions in its sixth year of service. A throughout the world. Using readily START HERE
cooperative international effort, OSCAR available commercial equipment, students
7 was constructed in module form by are experiencing firsthand the unique
amateurs in West Germany, Canada and aspects of space communications: Dop- Fig. 3 - Satellite pallses through the range of
Australia as well as in the U.S. .pier shift, orbital mechanics, Faraday two stations, enabling contact.
The latest in the Phase II (low orbit, rotation, telemetry decoding and much
long life) OSCAR series of amateur more. (If you are interested in par-
satellites, AMSA T -OSCAR 8, was ticipating in this program locally, contact
launched in March of 1978, joining its OSCAR Education at ARRL head-
predecessors, though at a slightly lower quarters.)
altitude. A joint effort by radio amateurs distance vhf communication for long
in Canada, Japan, the United States and Altitude, Time and Range periods of time.
West Germany, this satellite is intended to' The determining factor in the maximum The maximum range of two-way com-
continue support for the OSCAR Educa- theoretical range of satellite communica- munication through the satellite is il-
tion Program 'while providing ariother tions is the height of the satellite. Fig. 2 lustrated in Fig. 3. The greater the range
spacecraft for experimentation and can be used to determine, the range for the circle overlap for the two stations, the
routine communication. Several weeks' low orbital characteristics of early longer the time that these stations can re-
after launch, the ARRL assumed opera- satellites in which the altitude is assumed main in contact. With low-orbit satellites,
tions responsibility to free AMSAT per- to be constant throughout each orbit. To communication at maximum range may
sonnel for intensive work on the AMSAT determine when you can hear the satellite, last less than one minute between stations
Phase III-A project. The OSCAR 8 draw a circle with a radius. equal to the whose range circles Qverlap only a small
transponders and other on-board systems map range from Fig. 2. For OSCAR 7 this amount; the ,effective range, therefore, is
appear to be functioning flawlessly; con- is about 2450 miles (4000 km) for its slightly 'less than twice the radius of your
tinued good health and useful operation is 91O-mile (1490-km) high orbit and for range circle. The key here is in mutual ac-
anticipated for years to corne .. OSCAR 8, orbiting at an altitude of 560 'cess of the satellite with the other stations.
Late in 1978 satellite users throughout the miles (910 km), the map range is about As you can see, at times with Phase III
world were pleased to welcome two addi- 2000 miles (3265 km). When a given satellites, you'll be able to communicate
tional low circular orbit satellite entries by satellite passes through your range circle, with stations halfway around the world.
the Soviet Union: Radio 1 and Radio 2. In- it is within range of your location and you' The length of time the'satellite is within
tended for educational and communication should be able to hear it. range of your statiqn is determined, as is
use with very low power ground stations, RS With the Phase III elliptical-orbit the range, by the height of the orbit. As
1 and RS 2 are no longer routinely available. satellites, communication range is, of many satellite users can tell you, the
Those who have' communicated through course, still determined by altitude, but higher the orbit, the slower the satellite
them are grateful for the work of our fellow the altitude continuously varies between a moves, and the longer the sa(ellite will be
radio amateurs in die Soviet Union and perigee of915 miles (1500 km) and apogee in range. With the Phase II low, nearly
hope these satellites ml:\rk the beginning of of about 22,300 miles (35,900 km). Near circular orbit satellites, the altitude is
continued active involvement. apogee, the satellite illuminates, fully, assumed to be constant. With tile Phase
More than a challenging means of com- that half of the earth's surface directly III high elliptical orbit, however, as the
munication for harris, the amateur satellites below. This enables reliable, very long altitude varies, so does the .speed of the
14·2 Chapter 14
Table 1
Amateur Radio Spacecraft Orbital Parameters
Satellite Inclination (Deg.) Apogee (km) Perigee (km) Period (min) Increment (Deg. W)
OSCAR 7 101.4467 1471.11 1448.11 114.9422 28.7374
OSCAR 8 98.8668 940.30 905.91 103.2026 25.8023
These parameters are based on Project OSCAR predictions dated August 1980.
14·4 Chapter 14
Glossary of Satellite Terminology storage by ground stations for later retrans· longitude (0·360').
AMSAi -' The Radio Amateur Satellite Cor· mission. erp - Effective radiated power - system
poration, a nonprofit organization located in COSMAC1802 - A CMOS, 8 bit micro· power output after transmission·line
Washington, DC; has overseen the OSCAR processor made by RCA. Its low power losses and antenna gain are considered.
program since the launch of OSCAR 5. consumption, high noise immunity, wide ESA - European Space Agency - Agency
(AMSAT, P. O. Box 27, Washington, DC temperature tolerance and flexible Ito the responsible for theAMSAT Phase III·A launch.
20044.) programmer) architecture, make it well Experiment Day - Routinely scheduled days
AOS - ' Atquillition of sigrial - The time you suited to use in orbital hardware. The 1802 during which the satellite is closed to
can first hear satellites, usually just after it was the heart of the on·board Phase III·A casual use and reserved for scientific ano
rises' above the horizon. computer. educational experimentation; a secondary
Apogee - That point in a satellite orbit where Descending Node - The peint where the purpose is to allo~ for battery recharge.
it is farthest above the earth. satellite crosses the equator traveling Falsing - Unscheduled, unintentional and
Area Coordinators - Volunteers in the AMSAT from north to the south. undesirable mode switching initiated by
organization who coordinate satellite activity Desense - A problem characteristic of Mode conditions in the spacecraft.
in their regions. Most states have at least J operation in which the strong 2·m uplink Geostationary orbit ':""An orbit at such an alti·
one; many countries are also represented. signal overloads a low·noise 70·cm tude (22,300 miles) and in such a direction
Ascending Node - The pOint where the preamplifier or converter. rN to E) over the equator that the satellite
satellite crosses the equator travelling Dopp/~r Effect - An apparent shift in frequen· appears to be fixed above a given point.
from the south to the north. cy caused by satellite movement toward or Groundtrack - The imaginary line traced on
ASSC - Amateur Satellite Service Council - away from your location. the surface of the earth by a satellite
A coordinating body comprising equal Down·link - The frequency of signals subsatellite point.
representation from Project OSCAR, AMSAT transmitted from the satellite to earth. IHU - Integrated housekeeping unit -
and ARRl. Earth sensor - A device to be used on the Phase Ill's on·board computer that will
Az-el mount - Antenna mount that allows AMSAT Phase III satellite that will manage many of the routine in·flight
antenna positioning in both the azimuth enable the computer to determine the tasks automatically.
and elevation planes. spacecraft orientation to the earth. Inclination - The angle at which the satellite
Azimuth - Dfrection (side·to·side in the Eccentricity - That orbital parameter used to crosses the equator at its ascending node;
horizontal plane) from a given point on describe how much an elliptical orbit also the highest latitude reached in an orbit.
earth; usually specified in degrees (N =0', deviates from a circle; eccentricity values An orbit crossing directly over the North
E = 90', S = 180', W = 270'). vary between 0 and 1: e = 0 for a circle. Pole would have an Inclination of 90', east
Circular Polarization - A special case in Elliptical Orbit - Those orbits in which the of the pole less than 90' and west of the
which the electric field component of a satellite path traces an ellipse with the pole greater than 90'. ,
transmitted radio wave is equal in both the earth at one focus. Increment - The number of degrees longitude
vertical and horizontal planes and effectively Elevation - Direction .(up·and·down in the ver· that the satellite appears to move westward
rotates. The sense of polarization, whether tical plane) from a given pOint on Elarth at the equator with each orbit, caused by the
right·hand circular or left·hand circular, is usually specified in degrees (0' = plane of earth's rotation under the satellite during
determined from behind the antenna, the earth's surface at your location; each orbit. (The earth rotates 360' in a
looking out alond its axis of propagation. and 90' = straight up, perpendicular to 24·hour period.)
Codes tore - A special system that allows the plane of the earth, overhead). Kick Motor - A one·shot motor on·board the
digital information IMorse Code, e.g.) to EQX - Equator crossing, usually specified in Phase III satellite that will be fired at
be placed in on·board electronic memory time (UTC) of crOSSing, and in degrees west perigee a few weeks after launch to adjust
amplifiers in the lOO-watt class are now on polarization is the special case where the cularly polarized antennas are a require-
the market or you can build a suitable vertical and horizontal components are ment - in fact it would be a better idea to
amplifier yourself. The narrow beam- equal; the field effectively rotates. begin with simple antennas and modify
width of high-gain antennas will be much The sense of polarization, whether them to suit your needs after you've had a
less a problem with the Phase III satellite right-hand circularly polarized (clockwise) little experience.
elliptical orbit as the spacecraft moves or left-hand circularly polarized (counter- The lO-meter antennas on OSCARs 7
across the sky at a comparatively slow clockwise), is determined, by convention, and 8 are dipoles that were deployed after
speed (± 3 hours at apogee). as though viewed from behind the anten- orbital insertion. With a linearly polarized
na, looking out along the axis of propaga- lO~meter, ground-station antenna, you'll
Satellite Hints and Kinks tion. The important factor for our pur- notice periodic fading as the satellite spins
Satellite communications confronts the poses, however, is that one achieves max- on its axis: maximum signal. strength oc-
newcomer with many unfamiliar imum down-link signal strength when the curs when the antennas are parallel,
phenomena that will challenge his or her circularly polarized antenna has the same minimum when perpendicular. With a
operating techniques. Here are a few ex- sense as the incoming wave. Cross full-wave horizontal loop or a circularly-
planations and "tricks of the trade" to polarization, on the other hand, can result - polarized turnstile the 'effect of this fading
help make your satellite work as effective in 30 dB or more attenuation of the signal. will be minimized.
and enjoyable as possible. How does this affect OSCAR opera-
tion'? The 2-meter antennas on OSCARs Full Duplex
Circular Polarization 7, 8 and Phase III satellites are circularly In satellite communications you can
When considel-ing a travelling polarized. Theoretically, to maximize sta- continuously monitor your transmitted
wavefront of rf energy, we can describe it tion performance, you'll want antennas signal through the satellite, as the down-
in terms of polarization. Radio waves are that are circularly polarized in the same link signal from the satellite is on a dif-
made up of both electric and magnetic sense as those on the spacecraft. A better ferent band from the up-link. You can
fields which are perpendicular to each choice would be an antenna that is switch- thus evaluate both the strength and quali-
other andto the direction of propagation. able between right-hand and left-hand cir- ty of your own signal throughout con-
The classification of polarization is deter- cular polarization. This is because as the tacts.' It is through this full-duplex
mined by the plane in which the electric downlink signal from the satellite passes capability that one is able to adjust for
component lies: If the plane of the electric through the ionosphere the polarization Doppler shift (see Doppler Shift below).
component is vertical, the polarization is sense may switch. This effect is commonly To locate yourself in the satellite pass-
vertical; if horizontal, the polarization is referred to as Faraday rotation. An exam- band, select an appropriate up-link fre-
horizontal. In the case where components ple of a switchable right-hand and left- quency, calculate the approximate down-
exist in both the vertical and horizontal hand polarized antenna is given later in .link frequency, transmit a string of dits
planes, polarization elliptical, and circular this chapter. This do~s not mean that, cir- and tune a few kHz either side of the
Specialized Communications Systems 14·5
the orbit to the desired final perigee and Pass An orbit of the satellite. so, the apogee will occur near the ~qitator.
inclination. Passband ..... The range of frequencies handled Project OSCAR - Califomla-based group,
LHCP - Left-hand circular polarization - by a satellite transponder. among the first to recognize the potential
counterclockwise. Perigee - That point in a satellite orbit where of space for Amateur Radio; responsible
LOS - Loss of signal - The time when the it passes ,closest to earth. for OSCARs 1 through 4. I
satellite passes out of range. Period - The time it takes for a complete QRP test - Special orbits set aside for
Mode A - Transponders wJth 2-meter uplink orbit, usually measured from one EQX to the operating through the satellites while
and 10-meter downlink. next. The higher the altitude, the longer using a maximum of 10 watts erp; output
Mode A/J - Simultaneous operation of the the period. powers of less than 1 watt have proven
Mode A and Mode J transponders on Phase I - The term given to the earliest, short- effective in some cases.
AMSAT-OSCAR 8. lived OSCAR satellites that were not equipp- Reference orbit - The orbit beginning with the
Mode B - Transponders with 70-cm uplink and ed with solar cells. When their batteries first ascending node during a given day UTC.
. 2-meter downlink. were depleted, they ceased operating . RHCP - Right hand circular polarization -
Mode C - Equivalent to Mode B with less Phase II - The term given to low altitude, clockwise.
power output; of no discernable difference' long-lived satellites. Equipped with solar RS1 and RS2 - The first two Soviet Amateur
than Mode B to the user. ' panels that powered the spacecraft systems Radio satellites.
Mode 0 - Battery-recharge mode; and recharged its batteries, these satellites Secondary payload - Usually smaller
transponders off. have been shown to be capable of lasting up packages that in essenc;e share the launch
Mode J - Transponders with 2-meter uplink to five years. (OSCARs 6, 7 and 8). with the primary payload which is the main
and 70-cm downlink. Phase III - Extended-range, high-orbit purpose for the launch. OSCAR 8, for exam-
NASA - National Aeronautics and Space satellites, typically in either elliptical orbit, ple, was a secondary payload "hitchhiker" to
Administration - U,S. Government agency as AMSAT Phase III-A, or in geostationary the LANDSAT C Earth Resources
that has provided "piggyback" launch oppor- orbit. Technology Satellite primary payload;
tunities for AMSAT OSCARs 5, 6, 7 and 8 in Power budget - A determination of how much secondary payloads, in effect, substitute
recognition of the OSCAR program's power is actually available to operate the for ballast weight.
contributions. on-board satellite systems, taking into Spin modulation - Periodic amplitude-fade
OSCAR - Orbiting Satellite Carrying Amateur account such things as solar cell surface and-peak resulting from Phase III SO-rpm
Radio; there have been eight Amateur area, solar cell efficiency and angle toward spin; the effect is a 3-Hz "modulation" of the
Radio satellites named OSCAR at the the sun. A positive power budget means passband.
end of 1980 and two Soviet Amateur Radio that ample power will be available to power SSC - Special service channels - Frequen-
satellites designated RS1 and RS2. the desired systems; a negative power cies in the downlink passband of AMSAT
OSCAR "Education Program - A special pro- budget means that periods <if shutdown and Phase III that are designated for
gram that brings live demonstrations of recharge must be periodically scheduled. authorized, scheduled use in such areas as
the OSCAR satellites to classrooms, helping Precession - An effect that will be education, data exchange, scientific experi-
teach students physics, space science, characteri$tic of the AMSAT Phase III mentation, bulletins and official traffiC.
,astronomy and related subjects. Teachers orbit; the satellite apogee after the SSB - Subsatellite point - That point directly
use ARRL curriculum materials to structure firing of the perigee kick motor will occur beneath a satellite on the surface of the
their courses,around the OSCAR satellites. at about 37" N latitude, but'will gradually earth at a given instant; usually defined i'n
OSCARLOCA TOR - A satellite tracking device rise higher to 57° N latitude within a few terms of latitude and longitude. '
consisting of a ranging oval and ground- years. Then, gradually, it will move lower Sun sensor - A device to be used on AMSAT
tracks superimposed on a polar projection in latitude until after five years or Phase III to determine the s'pacecraft
map. orientation to the sun.
calculated down-link frequency_ Once Mode-A Down Link will overload or "desense" most 435-MHz
you've located your signal in the pass- low-noise preamplifiers l!nd converters. If
band, you'll know approximately how When first, listening to die OSCAR your transmit and receive antennas are
much to change the transmit frequency to satellites, one is likely to tune to the located close to each other your 2-meter
"hit" a given receive frequency_ Simply Mode-A transponder down·link between up-link signal may obliterate the 435-MHz
change your transmit frequency by the 29.4 and 29.5 MHz_ This may prove down-link while you are transmitting!
proper amount - with the transmitter disappointing in that the signal may h~ve The problem is common and the solu-,
off! Swishing the passband With the key a warbling quality, reminiscent of weak hf tion is, fortunately, simple_ For about $5
down is inconsiderate of the other DX. If this is the case, check the re- you can build an effective 70-cm, cavity
users_ mainder of the lO-meter band. At this filter from materials readily available at
point in the sunspot cycle with high solar your neighborhood hardware or plumbing
Doppler Sh~t activity, you may find the band "wide
Doppler shift is caused by the relative open." This means that the ionosphere is
motion between you and the satellite_ As bending terrestrially originated lO-meter
the satellite is moving toward you, the fre- signals back down to earth and long
quency of down-link signals will increase distance communication at 28 MHz is suc-
by a small amount. As the satellite passes cessful. The OSCAR satellites orbit hun-
overhead and begins to move away from dreds of miles above the ionosphere and
you, there will be a sudden drop in fre- so 29-MHz down-link signals are also bent
quency of a few kilohertz, much the same by the ionosphere - but away from 'the
way as the tone of a car horn or train earth! You'll usually find that the down-
whistle drops as the vehicle moves past' link signal degradation is worse at low
you. This Doppler effect will be different elevations where the angle of incidence
for stations located at different distances with the ionosphere is shallow; with the
from the satellite. The result is that signals satellites at high elevations, the angles are
Oh,,,+'rr~rn'+TTrrhn~
passing through the satellite move slowly much steeper and the probability of a
around the calculated down-link frequen- signal penetrating the ionosphere is much o 5 10 i5 20 25
MAXIMUM OOPPLER SHIFT (kHz)
cy. Locating ,your own signal is a little greater.
more difficult than simply computing the
relation between input and output fre- Mode-l Desensing Fig. 5 - Satellite transmitter 'frequency versus
Doppler shift for satellite in 200- or
quency. as the rather hard-to-predict ef- As most members of the Mode J Club 1000-statute-mile orbits. For a translator, use
fects of Doppler must be taken into ac- (details in Fig. 6) will tell you, the strong the difference between uplink and downlink
count. 2-meter up-link signal from your station frequencies as the "frequency."
14·6 Chapter 14
codestore, synthesized-speech telemetry,
1'l\HORIZONAL LOOP
-magnetometer, . particle/radiation detec- . (PARALLEL TO GROUND
SURFACE)
tors, and two experimental beacons on 1.3 1/S·l\OR HIGHER
and lOA GHz. This launch is scheduled ABOVE GROUND
Mechanical Details
The TR-44 rotator is mounted inside
the tripod by means of a rotator plate of
the type commonly used with a top section
of Rohn 25 tower. V-bolts around the
tripod legs secure the plate to the tripod. iRG-58A/U
A length of I-inch (25-mm) galvanized MATCHING
water pipe (used as the mast) extends from SECTION
,,~,\
I _ _ _ _ ..,... _ _ _ _ 1I
'-
145-MHz uplink signal. This "Mode J
(
50-OHM
filter" should eliminate the problem. The
filter is inserted between the 435-MHz
antenna and the 435-MHz preamplifier or
1/4-1\, 95-0HM 50-OHM LINE converter. The insertion loss is roughly 0.4 "-----+---../COAXIAL. "T'I
PHASING LINE TO STATION
to 0.5 dB and the "nose" of the filter is COAXIAL liT" CONNECTOR
approximately 7 MHz wide. One impor- CONNECTOR
Fig. 13 - A drawing of the switchable tant feature of this filter is the excellent
polarization antenna system complete with rolI:off characteristic lower in frequency
cable specifications. When calculating the than the passband. This allows a 50-dB at-
length of individual cable be sure to include tenuation (or more) of the 145-MHz
the velocity factor of the cables. Fig. 14 - This is a drawing of the basic anten·
• uplink signal frequency.
na system for sWitchable right· or left-hand cir-
Most small plumbing businesses can cular polarization. The quarter-wavelength lines
supply you with the materials needed for between the antennas and th'e relay step up
construction. A local plumber cut several the antenna 50-ohm' impedance to 100 ohms.
pieces of the 3/4- and 3-inch copper pipe The phasing line is made from 95-ohm coaxial
a quarter-wavelength transformer bet- cable so as to provide a good match to the
. ween the antenna and the relay. This and charged the writer only a few dollars . 1oo-ohm system. See text for a detailed
quarter wavelength of 75-ohm line steps Circuit-board material can be used for the description of the system.
up the 50-ohm impedance of each antenna top and bottom plates. Silver plating all of
to roughly 100 ohms. As shown in. the the parts is a good idea to protect against
drawing, each fixed contact of the relay is poor contacts. However, this is not man-
also connected to the quarter-wavelength datory. The dimensions for the filter are
(90-degree) section of cable that acts as given in Fig. 15.
the phasing line. The phasing line was EME thus presents amateurs with the
constructed from RG-133/U cable, which EARTH·MOON·EARTH ultimate challenge in strengthening radio
has a characteristic impedance of 95 Popularly known as moonbounce, systems. Before amateur involvement the
ohms. This provides a very dose match to EME is the second-most popular method only other known mo~m relay circuit was
the l00-ohm impedance of the system. If of space communication after OSCAR. operated by the U.S. Navy between
RG-133/U proves difficult to locate, RG- The concept is straightforward: Stations Washington, DC and Hawaii. Their 400
63/U (125-ohm impedance) may be used which can simultaneously see the moon megawatts of effective radiated power
with a slightly higher mismatch. As can be communicate by reflecting vhf and uhf carried four multiplexed RTTY channels.
seen in the drawing the phasing line is signals off the lunar surface. Unlike The first two-way amateur link took place,
always in series with the system feed point OSCAR, though, the two stations have a between the Eimac Radio Club, W6HB,
and one of the antennas. As shown, the relatively stable target and maybe and the Rhodendron Swamp VHF
antenna on the left receives energy 90 separated by virtually 180 degrees of arc Society, WIBU, on 1296 MHz in July
degrees ahead of the one on' the right. If on the earth's surface, which translates to 1960. Only a few amateurs heard anything
the relay were switched, just the opposite more than /11,000 miles (17,600 km). more than their own echoes during the
would be true. There is a trade off, though; Since the next few years. Hams at government and
In reality it is not necessary to use single moon's mean distance from earth is private institutions began conducting tests
quarter wavelengths of line. For example, 239,000 miles (385,000 km), path losses with other hams by using very large arrays
the 75-cihm impedance transforming lines are huge when compared to "local" vhf such as the 150-foot (45.7-m) steerable
between each antenna and the relay eouid work. Thus, each station on an EME cir- dish at WA6LET (Stanford University) or
be any odd mUltiple of one quarter cuit demands the most out of the the l000-foot (305-m) paraboliC surface at
wavelength, such as 3/4, 5/4, 7/4 transmitter, antenna, receiver and KP4BPZ (Arecibo). Amateur-to-amateur
wavelength, etc. The same is true for the operator skills. Even with all those factors contacts did not become established until
95- or 125-ohm phasing line. One must in an optimum state, the signal in the the early 70s, a notable effort being be-
keep track while using different lengths headphones may be barely perceptible tween VE7BBG and WA6HXW. Activity
for the phasing line: This is especially true above the noise. Nevertheless, for any spread to all continents - except South
when figuring out which position of the· type of amateur communication over a America. In July 1976, the Mt. Airy VHF
relay will yield right-' or left-hand distance of 500 miles or more at 432 MHz, Club of Philadelphia (Packrats) staged an
polarization. The builder is apt to find for example, moon bounce comes out the expedition to Barraquilla, Colombia,
that it will be necessary to. use one of the winner over terrestrial methods when which allowed K2UYH to become the first
odd multiples of a quarter wavelength various ·factors are figured on a balance amateur to work all continents on 432
since a single qijarter wavelength of line, sheet. MHz.
SpeCialized Communications Systems 14·9
universal operating procedure will evolve.
Parts List
Piece No. EME Scheduling
1 Pipe, Copper Cut ends square. The best days to schedule are usually
3" dla. 5" long Drill or punch for when the moon is at perigee (closest to the
connectors 3·3/4" earth) since the path loss is typically 2 dB
from bottom.
2 Pipe, Copper
less than when the moon is at apogee (far-
Solder to cer.ter
3/4" dia. 4" long of 10 thest from the earth). The moon's perigee
3 Disc, Copper Drill thru center. and apogee dates are listed monthly in Sky
3/4" dla. Solder solid hook and Telescope. If the distances are located
1/16" - 1/8" thick up wire between on the EME Path Loss Nomograph, Fig.
disc and connector
to space disc 16, the nominal EJ\.fE path loss can be
3/16" Jrom pc 2. determined readily for the most popular
4 Disc, Copper Drill thru center. frequencies. The moon follows a sine-
3/4" dia. Solder solid hook wave orbit pattern. Hence, the day-to-day
1/16" - 1/8" thick up wire between path changes at apogee and perigee are
disc and connector
to space'CIi8c minor. The greatest changes take place at
3/16" from pc 2. the time when the moon is traversing be-
5 Connector, Coax BNC, SMA or N tween apogee and perigee.
type. Solder to Also the SD (semi-diameter in minutes
prevent turning. of arc), listed in Tables 4 through 7, can
For large
connector use be translated into apogee and perigee by
chassis punch use of the EME Path Loss Nomograph.
6 Connector, Coax BNC, SMA or N An SD of 16.53 is an approximate earth-
type. Solder to to-moon distance of 225,000 miles (typical
prevent turning.
For large
perigee), and an SD of 14.7 is an approxi-
connector use 5) The receiving system should have a mate distance of 252,500 miles (typical
chassis punch. very low noise figure and sharp filters. apogee). However, there are several other
7 Nut, Brass Don't let these requirements scare you! factors that must be considered for op-
1/4"-20 hex Most EME-ers started out as listeners, timum scheduling besides the path loss.
8 Nut, Brass and in the' first, second and third ARRL If perigee occurs near the time of a new
1/4"-20 hex
International EME Competition opera- moon, one to two days will be unusable
9 PIC Board, Drill hole in center
double sided. to clear 1/4·20 bolt
tors with nothing more than a single Yagi, since proximity of the moon to the sun's
Top 4" x 4". Solder 7 and 8 preamplifier and multimode transceiver orbit will cause increased sun-noise.
each side of hole. were hearing the stronger stations. For pickup. Therefore, schedules should be
(Use bolt 11 to those who are seriously interested in avoided when the moon is within 10° of
hold nuts in place
whe'n soldering.) assembling a complete station, the Eimac the sun (and .farther if your antenna has a
10 PIC Board, Solder 2 in center. Division of Varian has assembled a com- wide beam or strong side lobes). Lately,
double sided. prehensive package, on . the technical perigee usually occurs ne~r new moon in
Bottom 4" x 4". details. Write to Eimac, 301 Industrial the summer and near· full moon in the
11 . Bolt, Brass Insert thru 12 Way, San Carlos, CA 94070. winter. These references to seasons apply
1/4·20 x 3" then thru 7 and 8. A short section about operating tech- only to the northern hemisphere. Hence,
12 Locking Nut, To hold piece 11 niques is offered as a guide to the begin-
Brass after resonance
it is usually easier to schedule in the
1/4·20 Hex adjustment. ner. It should be noted, however, that the daytime in the summer and during the
details differ from one band to another to nighttime in the winter. The fall and
some degree. Such differences are slight, spring seasons are a toss-up. The moon's
Fig. 15 - Details of the "Mode J desense filter."
and should cause no great concern. perigee does shift during its 18- to 19-year
Perhaps as the ranks grow an accepted cycle so this rule will not always apply and
.i
1296 - 272.5 dB 5.5
design considerations must be taken
because it is weak-signal work.
1) Transmissions must be made on cw !
240
V / 2304-278.0dB ::::
.7
15.6
L'
.. 220-255.0dB 8.5
~ 225t--.+/-----.t'-.NOMINAL PERIGEE 432-261.0dB-- .-.--- - ... -- ----,6.8
.rotators are needed for both azimuth and
o 1296-270.5dB
elevation. Since the half-power beam- ~ . I 2304·276.0d9 ::::
width of a high-gain antenna is quite 220 -0.4 0 +0.4 to.8 ....-tz-·+.'.6 -"+2:O-Tz...-- +2.8 -
sharp, the rotators must have an ap- AODITIONAL PATH LOSS IN DECIBELS
propriate accuracy.
4) Transmission-line losses should be Fig. 16 - Variations in EME path loss can be determined from this graph. SO refers to the semi·
held to a minimum. diameter of the moon, which is indicated at the bottom of each page in,'the Nautical Almanac.
14·10 Chapter 14
Table 4
Greenwich hour angle (GHA) and declination (decl.) of the'moon at 0000 UTe for each day of the month for January through June 1981. The 2nd and 4th
columns indicate hourly increments for GHA and decl., respectively. The 5th column indicates the semi-diameter (SOrof the moon in minutes of arc. See
text for full explanation.
1 241 20 14.555 -9 90 -0 t4} 14.8 1 227.00 14 480 -19.49 -0.034 15.3 1 246. s8 14.478 -20. 17 -0.007 15. J
n0 ~2 14.53:5 -13 ?4 -0 122 14 9 2 214.53 14.463 -20 31 0.010 15.4 2 233. 85 14. 46] -20. 35 f.\. €I37 15.5
;>19 7:7 14 51] -16 28 -0 094 15. ~ 3 201.65 14.453 -20.97 o 056 15. ~ 3 220.96 14.455 -19.46 0.084 15.7
4 ?07 f,7 14.490 -1,8 53 -0. €I59 15.1 4 188.53 14.452 -lS 72 0.102 15.8 4 207 88 14.453 -17.45 0.128 16.0
., :~9~. 45 14.47} -19.95 -0. '018 15.1 5 175.38 14 458 -16.27 0.143 16.0 5 194.75 14.456 -14.37 0.167 16.2
6 182 79 :14 46;;:: -20.37 0.027 15.4 6 162.38 14.467 -12 83 0.176 16.1 6 181.69 14.460 . -10.36 0.196 16.3
169.87 14 46n -19 72 01373 1S6 7 14958 14.475 -8.61 0.197 16.2 7 168.73 14.463 -567 0.212 16.4
8 156.91 14 465 -17 98 o li5 1!) 7 8 137 €Ie 14.480 -3.87 1'1.2137 16 2 8 155." 14, 462 -0.58 B. 214 16.5
9 144 08 14.475 -1521 {1 151 15.8 9·124.52 14.4813 1. 09 0.204 16.2 9 '142.92 14.457 4.56 0.2132 16.4
10 131 49 14 485 -1159 o 179 15:9 113 112. €I3 14 473 '599 13.189 16.2 10 129.913 14.449 9.413 13.176 16~ 4
11 119.1:1 14.491 -7.313 0.1% 1613 11 99.39 14.462 113.53 0.163 16.1 11 116.68 14.441 13.64 0.1413 16.2
12 106 92 14.492 -2.59 • 13 2133 16.1 12 86 48 14.449 14.45 0127 16.0 12 1133 26 14.435 16.99 0.1395 16.1
13 9473 14.485 2 29 13.199 16. :t 13 73 26 14.439 17.513 13.1383 16.13 13 89.713 14.435 19.27 13.046 15.9
14 82 37 14 472 7137 13.184 162 14 5981 14. 436 19 50 0.1335 15.9 14 76.14 14.444 29.37 ,0.004 15.7
15 69 713 14.454 11.48 0157 16.2 15 46 26 14.441 213 34 -13.015 15.7 15 62.79 14.460 213.28 '-0.1351 15.6
16 56 60 14.437 15 25 e.l'.9 161 16 32.86 14 457 19.98 -13.061 15.6 16 49.83 14.482 19.136 -0. 092 15.4
17 '43.138 14.424 ' 18.11 e. en 16 1 17 19.82 14.479 18.51 -0. 11'12 15.5 17 37 39 14.5136 16.86 -0.125 15.3
18 29 25 14.421 19.85 0.1321 16.13 .18 7.30 14.504 16.137 -13 133 15 :;; 18 25.52 14.528 13.86 -13.149 15.2
19 15.37 14.431 20.36 -0.039 15.9 19 355 39 14.528 12 B8 -0.156 15.2 19 14.20 14.548 10 28 -0.166 15. "
20 1.72 14.452 19.63 -0.077 15 7 20 344.05 14.548 9.14 -0.169 15.1 20 3.34 14.562 6.29 -0. 175 14.9
21 348.57 14.480 17.79 -0.115 15.5 21 333 20 14.563 5.137 -0.175 15. Ii! 21 352 82 14.5713 2.10 -0.176 14.8
22 336.09 . 14.509 15.04 -0 143 15 4 22 322.70 14.571 0.87 -0.174 14.9 22 342.50 14.572 -2.13 -0.171 14.8
23 324.31 14.535 11 61 -0.161 15.2 23 312.413 14.572 -3.31 -0.167 14.8 23 33~. 23 14.568 -6.23 -0.160 14.7
24 313.16 14.556 7 74 -0.171 15.0 24 "3:e2.1s 14.567 -::7 52 -0.. 154 14.8 24 321.87 14.560 -113.08 -9.143 14.7
25 30250 14.569 .3.63 -0.174 14.9 25 291.73 14.555 -11. 03 -0. B6 14.8 25 311.30 14.546 -13 52 ':'0121 14.7
26 292.16 14.575 -13.55 -0.171 14.8 26 281.06 14.539 -14.39 -0.113 14.8 2~ 3013.41 14.530, -16.42 -0.093 14.8
27 281 95 14.572 -466 -0.163 14 8 27 269 99' 14.519 -17 91 -0.1384 14.9 27 289. 12 14.512 -18.65 ~9. ~60 14.9
28 271.69 14.563 -8.56 -0.149 14.8 28 258. 44 14.497 -19.01 -0.1348 15.1 28 277. 41 14.495 -213.08 -0.022 15.1
29 26120 14. 547 -12.13 -0.130 14.8 29 265.29 14.481 -20.60 0.021 15.3
30 250.32 14.526 -15.24 -0.194 14.9 30 252.83 14.471 -213.11 ·0.965 15.5
31 23894 14.503 -17.75 -0.973 15.1 31 240.13 14.466 -18. 54' 0.109 15.7
1 227 30 14.463 -15.92 0.150 16.0 t 221. 18 14.468 -5.213 9.216 16.4 1 201.12 14.4113 1286 o 163: 16 7
2 214.43 14.462 -12.31 9.185 16.2 2 208.413 14 454 -0 02 0.221 16.6 2 186.96 14.3913 16.78 13 116 16 6
3' 2131.52 14.460 -7.87 0.209 16.4 } 195.31 14.436 5.29 13.211 16.7 } 172.32 14.383 1955 e. A5~: 16.5
4 188.56 14.455 -2.85 0.220 16.6 4 181.78 14.417 10 H 0.183 16 7 4 157.51. 14.393 213.94 -0002 16 3
'5 175.47 14.445 2.44 9.216 16.7 5 16778 14.4130 14.74, 0.141 1.6.7 5 142.93 14.418 20.89 -9057 16 1
6 162.16 14.433 7.63 13.196 16.6 6 153 39 14.,393 1812 0.1388 16.5 6 12896 14. 453 19.51 -0.103 15.8
7 148. 56 14.422 1232 0.161 16. '5 7 138. 82 14.399 2023 0.0313 16.3 7 115.82 14 489 17 04 -0.137 15.6
8 134.69 14 415 16.18 0.115 16.4 B 124.40 14.419 20.94 -0.026 16. "' 8 103.57 14.5n 13 76 -0.160 15.3
9 120.64 14.416 18. 93 13.062 16.2 9 110.46 14.449 20.32 -0.974 15.8 9 92.11 14.549 9.92 -0.173 151 ,
113 106.62 14.427 20.42 0.009 16.0 10 97.23 14. 483 18. 53 -13.113 Ij 5 10 81. 28 14.566 5.76 -13.179 15.0
11 92.86 14.447 29.63 -0.041 15.7 11 84.81 14.514 15.82 -0.142 1 .. 3 it 70.87 14.575 1.45 -0.179 14.9
12 79.59 14.474 19.64 -0.084 15.5 12 ?S16 14.541 1242 -0.161 151 12 69.68 14.576 -2 85 -0173 14.8
13 66.96 14.501 17.62 -0.119 15.3 13 62.14 14.561 8. 55 -0.173 15.0 13 50.51 14.5713 -7. 00 -0161 147
14 54.99 14.527 14.77 -0.145 15.2 14 51.60 14.573 4 413 -0.178 14.9 14 40.19 14.558 -113.88 -0.144 14.7
15 43.65 14.548 11.30 -0.163 '15.9 15 41.35 14.577 0.14 -0.177 14.8 15 29.57 14.541 -14. 34 -9.121 14.8
16 32.81 14.564 7.39 -13.171 14.9 16 31.21 14.575 -4.10 -9.169 14.7 ·16 18. 56 14.523 -17.23 -13.091 148
17 22. :13 14.572 3.23 -0.177 14.8 17 21.1313 14 566 -8.17 -9.156 14.7 17 7.11 14.506 -19 42 -0.056 14.9
18 12.97 14575 -1.02 -0.174 ' 14.8 18 10.59 14.553 -11.92 -0.137 14.7 18 355.25 14.493 -20.75 -0.015 15. til
19 1.86 14.571 -5.20 -0.165 14.7 19 359.87 14.537 -15.21 -0.112 147 19 343.08 14.486 -2112 e. e27 15.1
29 351.57 14.563 -9.17 -0.150 14.7 20 348. 76 14.521 -17.89 -0.080 14.8 213 330.75 14.486 -20.47 0.070 15.2
21 341.07 14.550 -1277 -13.129 14.7 21 357.25 14.506 -19.82 -0.044 14.9 21 318.42 14.491 -18. 79 0.1113 15,4
22 330.26 14.535 -15.87 -0.1132 14.7 22 325.40 14.495 -20.88 -0.004 15.9 22 306.21 14.498 -16.14. 0146 15.5
23 319.09 14.519 -18.33 . -e. 0713 14.8 23 313. 29 14.490 -20.97 0.038 15.1 23 294.17 14.504 -1263 0.175 15.7
24 307.54 14.504 -20.131 -0.1333 14.9 24 301,05 14.490 -20.05 0.081l 15.3 24 282.25 14.505 -8. 43 0.196 15.9
25 295.64 14.493 -20.80 0.007 15.1 25 288.80 14.492 -18. 12 0.120 15.4 25 270. J6 14.499 -3.72 0.208 16.0
26 283 47 14.486 -20.62 0.1350 15.2 26 276.62 14.495 -15.24 0.155 15.6 26 258. 34 14.485 1. 28 0.210 16.2
27 271.13 14.482 -19.42 0.093 15.4 27 264.50 14.495 -11.52 0.184 15.9 27 245.98 14464 6.32 0.199 16.3
28 258.70 14.481 -17.19 0.133 15.7 28 25239 14.4913 -7.10 13.205 16.1 28 233. 11 14.438 11.09 0.174 16.4
29 246.24 14.480 -14.013 0.169 15.9 29 240.15 14.478 -2.17 9.216 16.3 29 219.61 14.411 15.28 0.135 16.5
30 233, 75 14.476 -9.94 0.198 16.2 30 22764 14.4613 3.01 0.214 16.5 30 205.48 14.392 18.52 0.084 16.5
31 214.67 14.435 8.'14 0.197 16.6
must be modified accordingly. sities. Generally, low elevation angles in- been used very effectively 'at 144 MHz,
Low moon declinations and low aiming crease antenna~noise pickup and increase but has been more elusive above 420
elevation generally produce poor results tropospheric absorption, especially above MHz. It is hoped that current tests on 144
and should be avoided if possible. Con- 420 MHz, where the galactic noise is very and 432 MHz, using this mode of propa-
versely, high moon declinations and high low. This situation cannot be avoided gation, will yield more predictable resutts.
elevation angles should yield best results. when one station is unable to elevate the . Usually, signals are stronger in the fall
Good results ar~ usually obtained when antenna above the horizon or when there and winter months and weaker in the sum-
both stations are using similar elevation is a great terrestrial distance between sta- mer. Also, signals are generally better at
angles, since then both stations are look- tions. Ground gain (gain obtained when night than during the day. This may be a.t-
ing through comparable electron den- the antenna is aimed at the horizon) has tributable to decreased ionization or less
Specialized Communications Systems 14-11
Table 5
GHA and dec!. of the moon at 0000 UTe for each day of the month for July through December 1981. Hourly increments for·GHA and dec!. and the moon's
SO are also indicated.
1 190 89 .14, 187 20 53: A A::'~ th.t 167. 19 14 480 16. 54 -0. 147 1:'i 7 151 96 14 560 L 62 -13186 15 1
" 176 1.8 14. 400 ;::1 13 --;-C:l 0]4 1to 2 154.71 14 51J n.132 -0.17B 15 5 141 39 14 565 -2. 85 -A. 178 14 9
1: 161. 78 14 427 ?0?-2 -A 0S,; 16 0 142 q7 1,4 ~37 8 94 -0. 182 15 J 3 130 96 14 564 -7 n -f) 164 '148
4 • 148. ~3 14 462 1::: ;:::.~ -C'1 1;'7 15.::: 4 HI. 86 14 ~56 4 56 -0 186 15 1 4 120 49 14 556 -11 07 -13 145 14 8
5 155 13 14 498 15 21 -~ :1'i" 15. F. 5 t2.1 213 14 566 0. 11 -e. 182 15. e 5 1139 83 14. 543 -14 54 -0 1,:0 14 8
6 1.23 08 14 5:<9 1.1. 47 -0 17} 15 4 6 11~ 73 14 51S9 -4 26 -0 172 14 '3 6 98. 88 14. 521 -17. 41 -0 1390 14 8
7 111. 78 14 ')'ic 7 11 -A 182 15. 2 le~ 44 14 564 -8.4" -I) 157 14 I> '7 87:<9 1.4.517 -19 56 -0.055 14.9
8 1(\1 A4 14 51>7 :;-~. 96 -FJ 18? 15 Fl 8 89 98 .14 553 -.12, 17 -0 137 14 8 8 75 79 14. 495 -213 87 -0. 015 15.13
9 99. 65 14 "iT; -1 42 -~ 177 14 9 9 79 24 14 ~B( -15 45 -0. 111 14. 8 9 63. 67 14 483 -21. 23. f) 1329 15.2
HI 80 41 14 ')71 -~ IS:=: -A 167 14 8 10 68 1? t4 '5.18 -18. 113 -1'1. 079 14 9 10 51. 27 14. 477 -20. 54 0 071 15.4
11 713 12 14 ')62 -9 68 -0 150 14 8 11 'iF. 54 14 499 -2". 0fl -0. 042 15. 13 11 38 71 14 476 -18. 79 0. 117 156
t2 59 61 14 547 -F: 29 -13 '129 14 8 12 44~: 14 48~ -;:-1. £10 (I flBe 15. 1 12 26 13 14 478 -15. 98 0. 156 15.8
l~ 48 74 14 ~?9 -\F. 18 -~. lAl 14 'f: t1 12 ltS ·14 47F=.: -21 0A 0 fJ4~ 1~,.--::~ 13 13 613 14 481 -12 25 0. 188 16. e
14 17 42 14 5A9 -18 Al -0 ~68 14 9 14 19 58 1.4 474 -19 93 ,,08915 4 14 1. 14 14. 482 -7 75 0. 2139 16.2
to 7564 14 49, -2f1 4::: -C'1 1?128 -1:-i. ft 1!) h SI~ 14 478 -17 79 e 13:A 15 6 1.5 348. ('2 14. 488 -2. 71 0 218 16.3
16 1::: 47 144ft? -::'1. 1J ~ Aj,5 15 1 16 154 42 14 485 -14 66 0.165 15.8 16 336 23 14. 472 2 51 0, 214 16.3
17 . 1 ~4 14 478 -:;:'A 76 f1 A!l9 15 7, 17 342. 01; 14 492 -10 69 e. 191 15 9 17 323 55 14 459 7 64 0. 196 16.3
18 <:4S 02 14 4:?? -l9 15 0.1At 154 18 ,'29 86 14 495 -6 f\q 0. 207 J.6. e 18 310 56 14 443 12. 33 0 164 16 3
1:4 ?;:':<f=; AS 14 49,9 -I'; 92 (l139 ie, ') 19 317 n 14.492 -1 12 "21.1 16 1 19 297. 19 14 429 16. 27 e. 122 '16.2
2F1 127(:=:::::: 14 498 -1'3: 5::: 0 17A .15 (' ?0 305 '55 t4 483 "95 13 2fl? 1.6 1 20 28~ 49 14. 420 19 19 0 071 16.1
~'1 1t1 78 1.4 504 -q ~e n 192 15 8 21 ?9" 14 14 468 8 81 . 0 HiS 16 2 21 269 57/ 14. 42:1 20.913 e.017 16.13
?:~' ;::'9g :::f, j" 504 -4. S8 n 2f1~ 1~ 9 22 :?8(1 38 14 45A 13 23 A t5~ 16 2 22 255 67 14.432 21. 30 -13.1337 15.9
?3: :=?87 9':: 14 497 0. 135 A 2fl8 16 a 23 267. 17 14 432 16 88 0 110 16. .t 23 242. A5 14 45] 20. 42 -0. 085 15.7
24 .::-75 89 14 48]: 'i 134 0. :t99 '16. 1 24 253 53 14 419 19 51 '0 059 1f; 1 24 228 ~ 14 479 18. 39 -0. 125 156
,,:.• J :=?h~: 47 14 461 q 8'1 A '17:=: jF, 2 ::?5 2j:g 60 14. 417 20. 94 B. Are 16. A 25 216. 41 14 505 15 40 -0 155 15.5
?6 ;~5A:-i4 14 4:;7 14 08 13 145 lt, <:. 26 225 61 14 428 21. 05 -0. 049 15 9 26 71'14 51 14 528 11. 69 -e. 175 153
27 217 04 14 4J.5 17 56 iii 1138 16:;: 27 2:t1. 88 14 449 19 89 -j;I 091; 15 8 27 1,,, 21 14 547 7. 48 -a. 186 15.2
28 ;'2J~:1 1.4. 40] t9 95 fl"4~ 162 28 198.,;5 14.475 1.7 58 -fl 1:1.5 15 7 28 182 13 14. 559 3 92 -0. 188 15 t
29 ?e8 67 14 4'~4 21, "6 -0 011 16 2 29 186 06 14 503 14 14 -13 16:;: "15. 'i 29 1.71. 74 14. 564 -1. SA -e. 183 14 9
:SA 194 17 1.4 420 20.79 -0.1366 loS. 1 30 174. 14 14 528 10. 43 -13 180 15. 4 313 1.<;J 29 14 ;;64 -5 913 -~ 171 149
'::! 1810 4'i 14 44~: J.9.2? -0.112 J.5 9 ]1 tl;2 8.: 14 547 6. 12 -0 .t87 15.;:
1 158 82 14. 558 -113 130 -0. :t53 14 8 D6 go 14 513 -213 64 -e. 0413 14. 8 1 t31 29 :14 51'15 -21 16 0. 071 1~ 1'\
2 1413 21 14 548 -H 67 ~0 129 14 7 2 125 22 14. 5135 -21 613 -0 01'le :t4 8 2 119 4fl J4 5139 -19 46 0. lUI 15 1
J.29 15 14. 534 -16. 78 -0. 1013 14. 7 U3:n l4 5fle -n 6A A A42 '15 13 1137 62 1.4 ';:14 -16. 82 0 145 15]
4 U8. le. 14.' 520 -19. 18 -B.B66 14.8 4 101 14 'l4 499 -213 to" 0. 1383 15.:t q" 96 t4. 517 -H. 33 0 175 15. 5
'.' 106. 66 14. 507 -~0. 77 -0. B28 14 9 5 89. 32 14. 5.11<1 -18 60 13. 123 15.:> 5 84 Sh 14!iJ 4 -9 12 13 199 1'5 8
F, 94. 82 14. 496 -21. 45 B. en 15. e 6 77 12 14. ~Al -1~ 64 e .. 159 15 6 6 7? 69 14 503 -414 13215160
7 82. 72 14. 489 -21. 13 13 056 15.2 7 6'" 34 14 498 -11. 82 13 190 15. 8 7 6fl 76 14 484 ft. 8~' 0 220 16 ~
8 70. 46 14 486 -19. 78 0.099 15.4 8 ~3, ~A 14 49"1 -7 26 e· 213: 16. 1 48 s7 14 456 6 tf1 QI ?12 t~ 5
9 58 12 14 485 -17 59 0.149 15." 9 41 08 .1.4 4,7 -2. 16 9. 224 16. 4 9 ~5 11 14 4~'4 .1 j 18 13 187 16':;
113 45. 76 14 485 -14. 13:;: 0.176 15.9 10 28 52 14. 456 ':3 23 e. 222 16. 6 1e n. 49 14 s91 1~ 68 e. :146 16 7
11 :;:3: 413 14 482 -9 81 B.294 16.1 11 15 46 14 430 8 57 0. 204 16. 7 11 6 89 14 169 19 18 13089 Jh 7
12 20. 97 14 476 -4 92 e. 221 16. 3 12 1. 79 14 4134 13 47 a. 168 16. 7 12 :'51 ,4 14 1 .. 4 21.:~ ,,024 16 6
E 8 39 14. 464 e. 37 B 224 16.5 13 347. 48 14.31'14 17.51 0.117 16. 7 13: '116. 48 14 7;:81 21 92 -13 1340 '16 ';
14 355. 54 14 448 5 75 0.212 16." 14 332 70 14 378 20 53 0. 057 16. 6 H 3?1 62 :t4.4J4 20 91; -0 096 16 2
15 342 29 14 4;'9 10. 84 B 184 16 . ., 15 317 77 14. 389 21. 69 -0. 006 16. 4 15 307. 56 14 455 18. 6f, -e. 13S 16. (:)
16 328. 59 14 412 15. 26 13.142 16.5 16 303. 11 14. 416 21 55 -0. 063 16. 1 16 294 47 14 494 15 3:1. -[1 11;7 15 7
17 :;:14 48 14. 403 18. 67 0. Ba9 16.4 17 289. 09 14. 451 213. 04 -0. 110 15. 9 17 2:::2 31 1.4 528 11 ";2 -0 184 15!i
18 390. 15 14. 405 29 82 0.032 16 2 18 275 92 14 488 17. 40 -0. 145 15. 6 18 2713 99 14. 552 6 '313 -0 191 15 2
19 285. 86 14 419 21 58 -0. 025 16.0 19 263 63 14. 520 13 93 -0 169 15. 4 19 ?fi0 24 14. 566 2.3·2 -I-?i 1q1 15 A
213 271 92 14. 444 20. 98 -0.076 15.8 20 252. 12 14. '546 9 88 -B. 183 15. 2 20 249 8] 14. '572 -2 26 -fl 184 14 9
21 258. 58 14 474 19. 16 -13.117 15.6 21 241 22. 14. 563 5. 49 -0. 189 15. 1 21 239 ~€, 14 570 -6. 67 -(1 171 14. 8
n 245. 97 14. 5134 16. 35 -B.149 15 4 22 230. 72 14 571 0. 97 -0. 187 14. 9 22 229 n 14 561 -113. 77 -0 15:t 14 7
23 234. 137 14. 5313 12. 78 -0.171 15. 3 23 220. 43 14. 572 -3. 53 -0. 1813 14. 8 21 218. 69 14 547 -14 43 -A 128 14 7
24 222. 78 14. 550 8. 68 -0.184 15.1 :?4 210 16 14. 566 -7. 85 -0 166 14. 8,~ 24 2137. 81 14 531 -17.51 -0 098 14 7
25 211. 97 14. 562 4. 27 -0. 188 15.0 25 199. 75 14 555 -11. 84' -0. 146 14. 7 25 196. 55 14. '516 -13. 87 -fi 17163 14 7
26 201. 46 14. 568 -0. 25 -0. 186 14 9 26 189. 138 14 542 -15 35 -0. 120 14. 7 26 184. 93: 1.4 ~05 -21:;9 -e. 1123 14. 8
"27 191. 139 14 567 -4. 71 -0.176. 14 8 27 '178. 138 14. 527 -18. 23 -B. 088 14. 7 27 173 134 14 499 -21 95 O. "1.9 14 9
28 180. 71 14 561 -8. 95 -0.160 14.8 28 166. 73 14. 514 -20. 35 -0. 1352 14. 7 28 161. 02 14. 500 -21 5~ n 061 14 9
29 170 17 14 551 -12. 80 -0.158 14.7 29 155. 07 14 506 -21. 60 -13 1312 14. 8 29 149 03 14. 507 -20 .•,4 e. 1131 15. 1
:;:1'1 159. 39 14. 538 -16. 12 -B.111 14.7 30 143. 22 14 503 -21. 88 . 0. 03B 14. 9 sa 137 19 14 515 -17 62 0.136 15.2
31 148. 31 14. 525 -18 77 -0.078 14 7 31 125. 56 14. 523 -14.14 e. 166 15 3
Faraday rotation. Scorpius and Sagittarius at southern of the week must be considered since most
Whenever the moon crosses the galactic declinations. Positions of the moon with of us have to work for a living and cannot
plane (twice a month for three to five days respect to these constellations cail be always be available for schedules.
each occurrence), ,the sky temperature will checked with Sky and Telescope magazine Naturally, weekends and evenings are
be higher. Hence, some degradation (1 to or the Nautical Almanac. The galactic preferred, especially when perigee occurs
2 dB) may be observed; especially above plane is biased toward southern declina- on a weekend.
420 MHz where the normal background tions, which will cause southerly declin~
sky temperature is lower. Areas of the sky tions to be less desirable (with respect to General Considerations
to avoid are the constellations Orion and noise) than are northern declinations. It helps to know your own EME win-
Gemini at northern declinations and Finally, the time of the day and the day dow as accurately as possible. This can be
14·12 Chapter 14
Table 6
GHA and decl. of the sun at 0000 UTe for each day of the month for January through June 1981. Hourly increments for GHA and decl. are also indicated.
An eclipse of the sun occurs on February 4 and ~. '
'1 179 15 14. 995 -21~:l a alB 1 176 61 14. 999 -17 18 o ('1.2 t 176. 88 15.13132 -7 70 Iil 016
2 179 03 14 995 -22 94 1;\ "~4 2 176 57 14. 999 -16 913 o 012 2 176 93 1~. 1302 -7 3:2 0.016
3 178 91 14 995 -22 85 1'\ 004 3: 176 54 14.999 -16 61 0.01;;: 1 176 98 15. 131'12 -6.94 13 016'
4 178 80 1'4. 995 -22 75 0.004 4 176. 52 14 999 -16 31 e.012 4 177. 03 15. 0132 -6.5"0 a 1316
5 178 69 14 995 -22 64 o ~e5 5 176 513 14. 999 -16. I'll o. on 5 177 09 15.002 -6.17 a. ell',
6 178. 57 14 995 -22. 53 o el'l5 6 176. 48 14 999 -15 71 o 013 6 177 15 15.002 -5.78 0.016
7 178. 46 14 995 -22 41_ o 01'\~ 7 176. 46 14 999 -15 40 e.en 7 177. 20 15 1'1133 -5.39 o 016
8 178 36 14 996 -22 2>: 01306 8 176. 45 15 1300 -15 138 0. en 8 177. 26 15. 13133 -~. 01 0.1316
9 178. 25 14 996 -2:>. 14 01306 9 176.44 15 000 -14. 77 013'13 9 177 D 15. 013? -4.62 11 016
'10 178. 15 14 996 -22. 01'l 0.0136 113 176. 43 15 131313 -14. 45 a ~14 1.1'1 177 39 15. aln -4.22 1'1.1'116
11 178. 135 14 996 -21 85 0. el'l7 11 176 43: 15. El00 ~14 12 o 014 11. 177. 45 15 01'1:> -3 83 0.016
12 177 '35 14 99f: -21 Fo9 1'\ lle7 12 176. 43 15.0130 -13 79 0.1'114 12 177 52 15. 003 -3.44 0.016
1'1 177 85 14. 996 -21. 51 13 007 13 176 41 15. 13130 -15 46 o 014 13 177. 59 1~. 130:> -3.05 0.016
14 :177 76 14 9q6 -21.?F. ;; l'Iil8 14 176.44 15.131313 -13 12 131'114 14 177 65 15.13133 -2.65 o 016
15 177 67 '14 996 -2'1 18 ~ A08 15 176. 45 15 001 -12 78 0.014 15 177. 72 15. 003 -226 0.1316
1.6 "177 58 14 997 -20. 99 a 13138 16 176. 46 15. eel -12 43 13.1314 16 t 77 8e 15. €le3 -1.86 13.1316
17 177 ~VI 14 q'~7 -?fl.8~ " 1'11'18 1.7 176 48 15. 1301 -12 "9 13 015 1.7 177 87 15. 1303 -1.47 o 1317
18 "177 4:' 14 997 -2" 6fJ "l >lPl9 18 176 50 15 0el -11 74 13 1315 18 177. 94 15. 003 -1. 07 1'1.016
19 177}4 14 997 -2~ 4"l 1'1 009 19 176 52 15. 1301 -11 18 ~ 015 19 178 131 15 1303 -1'1.68 13.016
:~A 177 2F. -14 997 -?f1 18 "l 009 213 176.54 15.0131 -11. 03 0. 1315 20 178. 1'19 15. 003 -0.28 0. e16'
?1 177:l q 14 997 -1~ 97 >1 0Aq ?1 176 57 15 eel -10 67 a "l15 21 1.78 16 15. 0133 13.12 0.016
22 J 77 1? .t4 997 -19 74 A 1'\10 22 176 613 15. 1301 -10 30 13 1315 22 178 24 15.003 e. 51 0.016
2J 177 05 14 997 -19 51. 1'1.011'1 n 176 64 15 1301 -9 94 1'1.1315 23 178. 31 15.0133 13.90 o 1316
24 176 99 14 998 -19 28 '" 1310 24 176. 67 15. 002 -9 57 1'1.1315 24 178 3:9 15.003 L 30 0.016
25 176 93 14 998 -:\9 03 " 1311'1 2'5 176 71 15. 0132 -9. 213 0.1316 25 1.78. 46 15. 1'103 L 69 0.1316
26 176 88 14 998 -18 79 " fl1.1 26 176 7'5 15 13132 -8 83 13 1316 26 178. 54 15. 0133 2.08 0.016
27 176.82 14 998 -'\8 5, a ill1 2? 176 79 15 002 -8. 4", 0. e16 27 178 62 15 1'103 2.48 0.1316
28 176 77 14. 998 -18.27 1'1 13'11 28 176 84 15 13132 -8. 138 13016 28 178 69 15 0133 2 87 0.1316
29 176 7} 14 998 -18 01 1'1 fl'11 29 178. 77 15. 003 3 26 o 016
}a 176 68 14. 998 -17 74 o Ai;::' 30 178 84 15 e133 3.65 13.016
?,1 176 64 14. 998 -17 4f. a 0:l2 31 178 92 15. 003 4 114 0. ell',
1 1.78. 99 15 01')3 4 42 0.016 1 1813. 72 15. 01'l1 14.98 0. an 1813 59 14 998 22 131 1'1 0136'
2 179.07 15 005 4 81 13.016 2 180. 75 15 1301 15 28 o 012 2 IB0. 55 14. ~98 22 15 0.005
3' 179 14 15. 1'103. 5 19 13.1316 3 1:?.0. 78 15 ')101 15. "08 13 01? :> 180. 51 14 998 22.27 €I a.10~
4 179. 22 15. alB 5.58 1'1.1316 4 180. 81 15 013'\ 15, 88 o 012 4 1813. 47 14. 998 ;;:2.40 0 005
5 179 29 15. 1303' 5.96 1'11316 5 1813 8?- 15. 0131 16 16 0.''112 5 1813. 42 14 998 22 51 0. «135
6 '179 36 15. 13133 6.34 13.016 6 ,18'" 85 15 001 1645 o 012 6 180. 38 t4 998 22 62 e ~e4
7 179. 43 15 003 6.71 13 016 7 180 87 15. eel 16 n 13 012 7 181l 3} 14 998 2? 72 ~' 01.34
8 179. 50 15. 1303 7. 139 o 1316 8 1813 88 15. 131'11 17 00 001.1 1813 29 1.4. 998 22 8;) 0.1304
9 179 57 15 0133 7.46 13.015 9 180.913 15 0013 t7 27 o 011 9 180 24 14. 998 22- 91 0 en3]:
Hl 179 64 15. 1303 7.83 13.1'115 1@ 180. 91 15. ee0 17 54 01311 113 181'1 19 14. 998 22 99 0. ~03
11 179. 71 15. 003 8.2e e. El15 11 180 92 15. 000 17 813 1'1.1311 11 180 14 14 998 23 06 1'1 01;\"
12 179. 77 15. 01'13 8.57 0.015 12 :18e 92 15. 13013 18 1'16 0010 12 180 139 14. 998 23: 11 ~3 0fG
13 179. 84 15. 0133 8.93 1'1.1315 13 180 93 15 000 1:3 11 n 1310 13 180 134 14 998 2?- 19 0.1302
14 179 913 15. 00:5 9.313 13 015 14 180. 9:; 15 01313 18 5~ 11 Ole 14 179 99 14 998 21 25 I{I 002
15 179 96 15 0133 9 65 0. 1'115 1") :1813 91 15 13('113 18.7g 013113 1.5 179 :33: 14 998 2?- 1~ 111.0""
16 180. 132 15. 1'1133 1001 0.1315 16 180 93 15 0130 1.9 02 0. 13113 16 179. 88 14 998 2?-.34 111 0H1.
17 180. 138 15. 131'12 10.37 0.015 17 lR0. 92 15 130" 19.2£ n 1309 17 179.83 14 998 2]: 17 "0.0'11
18 1813. 14 15 002 113.72' 13 014 18 1813 n 15 ')00 19 48 13.009 18 179. 77 t4 998 23.41) 0.0131
19 Ib~ 20 15. 002 1107 13.1'114 19 :t80 91 15 0130 1971'1 13 009 19 :t79. 72 t4 99A 23.42 1'1 001
21'1 180. 25 15. 002 1141 13.1314 20 1813. 910 14 999 19 91 o 009 20 179 67 14 998 234} 13 0013
21 1813. 30 15. 002 11 75 13.014 21 180 88 14 999 20 12 0.1308 21. 179 61 t4 998 n 44 -0.0f\f'
22 181'1. 36 15 1302 12139 13.1314 22 181l 87 14. 999 13.1308 22 179.56 14.998 :?J 44 -0 0fJ.,
23 1813. 40 1:1. 002 12.43 13.1'114' 23 180 85 14 999 .20 52 0.0108 2J 179 5f\ 14 9.98 . 23 43 -0 01'11
24 180 45 15. 1302 12.76 0.1314 24 180. 83 14. 999 20.71 0.,13138 24 179 45 14 998 23 42 -13 0131
25 180. 513 15. 1'11'12 13139 13.014 25 1813 80 14. 999 21'1 89 e 007 25 179 40 14. 998 23. 40 -0. 00t
26 180 54 15. 01'l2 B.41 13.013 26 180 78 14. 999 21 07 00137 26 179 34 14 998 ~?; 3:7 -0 O~l
27 1813. 58 15. 1'102 1371 13.013 27 1813. 75 14. 999 21.24 o 1307 27 179 29 14.998 2, 34 -,~ 002
28 181'1. 62 15. 002 14135 e. on 28 180 72 14 999 n.41 f1 007 28 179 24 14. 998 21. s0 -0 0a.12
29 180. 66 15. eel 14.37 13.013 29 180 69 14. 999 21 57 13 006 29 179 19 14 998 23 25 -0 00~
31'1 18e 69 15. 13131 14.68 13013 :10 180. 66 14 999 21 72 0. 1'106 30 179 14 14. 998 27,. 1~ -A 00]
31 18(1 62 14. 998 n 87 0. ~j6
determined best with the help of the charts EME stations are limited in some way by systems exhibit some pattern skewing
and tables discussed in a later section of local obstructions, antenna-mounting which must be accounted for: A simple
this chapter. Most EME operators deter- constraints, geographical considerations, calibration method is to peak your anten-
mine their local window and translate it and the like. Therefore, the accuracy of na· on received sun noise and then align
into GHA (Greenwich hour angle) and each station's EME window is very impor- the bore sight tube on the sun. The
declination. This information is a con- tant for locating common windows and boresight of the antenna is now calibrated
stant, so once it is determined it is usable setting schedule times. and can be used to aim the antenna at the
by other stations just as one would use A bore sight of some type is practically moon. Readers are cautioned against us-
UTe Likewise, jt helps to know the win- mandatory in order to align your antenna ing a telescope or other device employing
dow of the station to be scheduled. Most accurately with the moon. Most antenna lenses as a boresight device! Even the best
Specialized Communications Systems 14·13
•
GHA and decl. of the sun at 0000 UTe for each day of the month for July through December 1981. Hourly increments for GHA and decl. are also indicated.
An eclipse of the sun occurs on July 31.
1 179 ~9 14.998 25.13 -a 093 1 178 43 15 e91 18.19 -111.011 . 1 179.97 15.903 8.40 -0. 015
2 179 04 14.998 2J 06 -0. e03 2 178 4~ 15.1391 17.85 -9.911 2 180. 05 15.003 8. a3 -0.1315
3 178 99 14. 998 22.99 -€I 003 3 178. 46 15.001 17 59 -0.011 3 1813.13 15.003 7.67 -9.015
4 178 94 14.998 22 91 -0 004 4 178. 48 15.001 17. :n -€I 011 4 180.21 15.13133 7.3:0 -13.015
5 17890 14 998 22 82 -0.004 5 17851 15 001 17.06 -0.011 5 180.29 15.003 6.93 -0.1315
6 178 86 14 998 22.72 -0.1304 6 178.53 15. B01 16.79 -0.011 6 180.38 15.004 6. 56 -0.016
7 178. 81 14.998 22.62 -13.1305 7 178. 56 15.001 16.51 -0.012 7 180.46 15.004 6.19 -0.016
8 178.77 14.998 22.51 -0.005 8 178.59 15.001 16.23 -0.012 8 180.55 15.0B4 5.81 -0.016
9 17874 14.998 22.40 -0.005 9 178. 62 15.001 15.95 -0.012 9 189.63 15.904 5.44 -9.016
10 178. 70 14.999 22 28 -e. 985 10 178.66 15.0e2 1566 -0012 10 180.72 15.0134 5.06 -e. 016
11 178. 66 14.999 22.15 -0.006 11 178 70 15 002 15.36 -0.012 . 11 180.81 15.004 4.68 -0.016
12 178.63 14.999 22.92 -0.1306 12 17874 15.002 15.137 -13.1313 12 180.90 15.004 4.30 -0. 1316
13 178. 613 14.. 999 21.88 -13.0136 13 178 78 15.13132 14 77 -0 en n 180.98 15.1304 3.92 -9.016
14 178. 57 14.999 21.7J -0.006 14 178 83 15.002 14 46 -0.913 14 181.137 15.004 3.53 -0.1316
15 178. 54 14.999 21.58 -913137 15 17887 15.002 14 15 -9.913 15 181.16 15.1304 3.15 -13.1316
16 178. 52 14.999 2142 -e. 007 '16 178.92 15002 1384 -0913 16 181.25 15.13134 2.77 -0.916
17 178.59 14.999 21 25 -9.007 17 178 97 15.1302 B.52 -0.913 17 181.34 15.004 2.38 -0.016
18 178. 48 14.999 21.08 -9.13137 18 179. es 15.092 13 29 -0.1313 18 181.43 15.004 1.99 -0.016
19 1';>8.46 14.999 20.91 -0 008 19 179 08 15.992 12 88 -13.914 19 181.52 15.004 1.61 ~e. 016
29 17844 14.999 20 72 -13.ee8 29 179.14 15.9133 12 55 -9914 20 181.61 15.1394 1.22 -0.016
21 178 43 15.900 29.5$ -13.1308 21 179. 20 15.903 12 22 -13.014 21. 181.70 15.0134 0.83 -13.1316
22 178.42 15.13013 20.34 -0.1308 22 179 26 15 13133 11 89 -13 1314 22 181.78 15.004 0.44 -0.016
23 178.41 15.1390 20.14 -0.999 23 179 1]: 1~ 001 115e", -0 914 23 181.87 15.0134 0.95 -0.016
24 178. 40 15.000 19 93. -0 >l09 24 179 3:9 1~ 9133 . 11 21 -13 1314 24 181. 96 15.004 -0.34 -0.016
25 1.78 40 15 0013 19 72 -13 9139 25 179 46 15 903 19.87 -9.1314 25 182.04 15.004 -0. 73 -0.016
26 178 39 15.000 19 51 -0.13139 26 179 ~1 1.509] 113 52 -13.915 26 182. n 15.0134 -1.12 -9.016
27 178.'39 15..13913 19 29 -13 1309 27 17'3.6'1 15 903 113 18 -0 ~15 27 182.22 15.003 -1.51 -9.016
28 178 40 15. 0~0 19 e6 -0 01e 28 179 67 15.1303 9 83 -0. 915 28 182. J9 15.003 -1.99 -0.1316
29 178 41"; 15 e130 18 83 -0.9U} 29 17974 15 003 9 47 -a 015 29 18238 15.00J -2.29 -13.916
10 178 41 15.0130 18 59 -9.0113 30 179 82 15 9133 9.12 -'J 915 313 182.47 15.003 -2.68 -e. 016
11 178.4i :15 13013 18 3!5 -90113 31 179 89 15 093 8 76 -r, 015
1 182.55 15.003 -3.07 -0.016 184.09 15.1300 -14.33 -0.013 1 182.78 14.996 -21.76 -0.006
2 182.63 15.003 -3.45 -0.016 2 184.10 15000 -14.65 -0. el3 2 182.68 14.996 -21.91 -0.006
3 182.70 15.003 -3.84 -0.016 3 184.113 15.000 -14.97 -0.013 3 182.59 14.996 -22.06 -13.006
4 182.78 15.003 -4.23 -0.1316 4 184.19 15.090 -15.28 -0.013 4 182.49 14.996 -22.2111 -0. 006
5 182.86 15.003 -4.61 -0.016 5 184 10 15.000 -15.59 -0. en 5 182.39 14.996 -22.33 -0. 005
6 182.9J 15.003 -5.00 -0.016 6 184. e9 15.0e0 -15.89 -0.012 6 182.28 14.996 -22.46 -0.005
7 183.01 15.003 -5.38 -0.016 7 184.08 14.999 -16.19 -0.012 7 . 182.18 14.995 -22.57 -0.005
8 183.08 15.003 -5 76 -0.016 S 184.06 14.999 -16 48 -0.012 ' 8 182.07 14,99!:, -22.69 -00134
9 183.15 15.093 -6.15 -0.016 9 184.05 14.999, -16.77 -0.012 9 181.96 14.995 -22.79 -13.0134
10 183 22 15.1303 -6.53 -0.016 10 184.02 14.999 -17 06 .-0 012 10 181 85 14.995 -22 89 -13 1304
11 183.28 15.003 -6.90 -0916 11 184.013 14.999 -1.7. 34 -0.011 11 181.73 14.995 -22.97 ~0. 003
12 183.35 15.003 -7 28 -0. e16 12 i8:?; 97 14.999 -1761 -0.011 12 181.62 14.995 -23.136 -0.0133
13 183.41 15.003 -7 66 -0 016 13 183.94 14 999 -17.88 -0.011 13 181.. 50 14.995 -2313 -0 e03
14 183.47 15. 002 -8. 03 -0. €l15 14 183 90 14 998 -1815 -0.1311 14 181.38 14.995 -23 19 -0.1302
15 183.53 15.002 -840 -0.915 15 183.87 14 998' -18 41 -0.1311 15 181.27 14.995 -23.25 -0.0132 .
16 lB3.5B 15.002 -8. 77 - -0. 015 16 183 82 14.998 -1866 -13.010 16 181.15 14995 -23.30 .20 002
17 183.64 15.002 -9.14 -0.015 17 183 78 14.998 -18. 91 -0. e10 17 181.92 14.995 -2!" 3:5 -0 001
18 183.69 15.002 -9 50 -0.015 18 183.73 14.998 -19.15 -0.010 18 180.90 14.995 -;'3:.38 -0.eill
19 183.73 15.002 -9 87 -0.915 19 183 67 14 998 -19.39 -0. elB 19 18a 78 14 995 -23 41 -0 0~1.
20 183.78 15.002 -10.23 -0.015 20 18362 14.998 -19 62 -0 009 20 18065 14 99~ -23 43 -0. 000
21 183.82 15.002 -10.59 -0.015 21 18156 14.997 -19.85 -0.009 21 18053 14.995 -23.44 -13.000
22 '183.86 15.002 -10.94 -0.015 22 183 49 14 997 -20. e7 -0.009 22 180. 4.1 14.995 -23.44 0.13013
23 183.90 15.1301 -11.30 -0.1315 23 183 43 14 997 -20.28 -0. 1309 23 ISil 28 14. 995 -23.44 013131
24 183 93 15.001 -11.65 -0.014 24 183 36 14 997 -20.49 -e. 008 24 180 16 14.995 -23.42 o 001
25 183.96 15.001 -11.99 -0.014 25 183 28 14.997 -20.69 -0.1308 25 1.80133 14 995 -23.411\ " 091
26 183.99 15.001 -12.34 -13.014 26 183.21 14.997 -.20 88 -0 'l08 26 17q 9:1 14 995 -2~. 38 o 1302
27 184.02 15.001 -12 68 -9.014 27 183.13 14.997 -21 07 -0. Be8 27 179 78 14 995 -23 }4 0. 0~?
28 184.04 15.001 -13.02 -0.914 28 183 04 14.996 -21.25 -0.0137 28 17q 66 14 995 -23 30 9.002
29 184.06 15.001 -13 35 -0.014 29 182.96 14.996 -21.43 -0.1307 29 1('9 ~4 14.99" -23 25 B.002
;;:0 184.07 15.001 -13 68 -13.014 39 182.87 14.996 -2159 -111.007 3:0 179. 42 14 995 -23.19 O. 130~
31 184 138 15.000 -14.01 -13.013 }1 17~ 29 14. q9~ -23 12 fl.llm
of optical filters will not eliminate the ience. Accuracies of ± 2 0 are usually Locating the Moon
hazard from solar radiation' when viewed necessary and can be attained with syn-
directly. A simple piece of tubing of small cros. A remote readout is particularly im- The moon orbits the earth once in ap-
diameter and two or three feet long can portant for scheduling 'when the moon is proximately 28 days, a lunar month.
serve the purpose in this instance. A sym- within 45 0 of the sun or when the sky is Because the plane of the moon's orbit is
metrical spot of light cast upon a piece of overcast. Very few of us are not bothered tilted from the earth's equatorial plane,
paper near the back end of the' tube .will by occasional fog, rain, snow or overcast. the moon swings in a sine-wave pattern
indicate alignment. Aiming the antenna blindly seldom pays both north and south of the equator. The
A remote readout (such as a syncro or off. angle of departure of the moon's position
selsyn) is a highly recommended conven- at a given time from the equatorial plane
14·14 Chapter 14
is termed declination. Declination angles error, but such error will likely be negligi- the data from Tables 4 through 7 into
of the moon; which are continually chang- . ble for Amateur Radio purposes. Worst- ' useful information. The GHA and decl.
ing (a few degrees a day), indicate the case conditions exist when apogee or information may be converted into
latitude on the earth's surface where the 'perigee occurs near mid-day on the date in azimuth and elevation angles with the
moon' will be at zenith.'For this presenta- question. Under such conditions the total mathematical equations and procedure'
tion, positive declination angles are used angular error in the position of the moon given later in this section. A less tedious
when the moon is north of the equator, as determined by this procedure may be as (but also perhaps less accurate) method is
and negative angles when south. much as a sixth of a degree. Because it to use the appropriate chart (or charts)
The longitude on the earth's surface takes a full year for the earth to orbit the from Figs. 17 through 24.
where the moon will be at zenith is related sun, the similar error for determining the Each chart of Figs. 17 through 24 is
to the moon's Greenwich Hour Angle, ab- position of the sun will be no more than a' prepared for a fixed-observer latitude on
breviated G.H.A. or GHA. "Hour few hundredths of a degree. the earth's surface. Use Fig. 17 if your
angl~" is defined'as the angle in degrees to Assume that we wish to know the GHA position is on the. equator, 0 degrees
the west of the meridian. If the GHA of and dec!. of the \1100n at 1115 UTC on latitude. Use Fig. 18 if your latitude is 10
the moon were zero degrees, it would be June 7, 1981. From Table 4 we see that the' degrees, either north or south' of the
directly over the Greenwich meridian. If GHA' on that date is 115.82. In the col- equator. Use Fig. 19 if your latitude is 20°
the moon's GHA were 15 degrees, the umn to the right we see that the moon's N. or S. The charts progress in this man-
moon would be directly over the meri- GHA is changing at an average rate of ner so thai you should use Fig. 24 if your
dian, which is designated as 15° W. 14.489 degrees per hour on that date. At latitude is 70° N. or S. If your latitude is
longitude on a globe. As one can readily 1115 UTC on June 7, 11.25 hours will between those for which the charts are
understand, the GHA of the moon is con- have elapsed since the clock time was 0000 prepared, you will obtain more accurate
tinually changing, too, because of the or- UTC. Multiplying the hourly increment, aiming data by interpolating between,
bital velocity of the moon and because of 14.489, times 11.25 tells us that the GHA values obtained from two charts, as ex-
the earth's rotation inside the moon's or- of the moon will have increased 163.00 plained later.
bit. The moon's GHA changes at the rate degrees in that time. Adding this 163.00 to To use the charts, it is first necessary to
'of approximately 347° per day. the GHA at ()()()() urc, 115.82, yields determine the local hour angle (LHA) and
GHA and declination are terms that 278.82 degrees as the GHA of the moon at dec!. of the celestial body, using data
may be applied to any celestial body. The 1115 UTC on June 7. From the same line from Tables 4 through 7. LHA equals
Nautical Almanac lists the GHA and dec!. in the table we see that the dec!. of the GHA plus or minus your longitude (plus
of the sun and moon (as 'well as for other, moon at 0000 UTC is 17.04 degrees, and if east long., minus if west). Let's assume
celestial bodies that may be used for the hourly increment is - 0.137. Using the that your latitude' is 40° N., and you wish
navigation) for every hour of the year. same procedure, we multiply - 0.13.1 to position, your antenna system in
Tables 4 through 7 are based on informa- times 11.25, yielding - 1.54. This tells us azimuth and elevation for EME contacts.
tion contained in The Nautical Almanac that the declination at 1115 UTe is 1.54 Let's further assume that we've deter-
for the same year as this edition of the degrees less than it was at 0000 UTC; mined the moon's LHA to be 320°, and
Handbook. This information may be used 17.04 - 1.54 = 15.50 degrees declination its declination to be 10 0. From Fig. 21,
to point an antenna with precision, rather at 1115 UTC on June 7. The moon's semi- which is prepared for 40° latitude, we may
than merely looking up in the sky and diameter on June 7 is 15.6 minutes of arc. read azimuth and elevation information
pointing one's antenna by eye. Tables 6 The procedure for using Tables 6 and 7 directly from the polar scale. The infor-
and 7 are for the sun. That information for determining the sun's position is iden- mation is read at the intersection of the
may be useful for boresighting an antenna tica!. 320° LHA line and the 10° dec!. line. The
array, as explained earlier in this chapter. 2 In using the tables, remember that azimuth, indicated by the radial coor-
These tables indicate the position of the negative declination angles denote that the dinates" is seen to be 119.1 0. T14e eleva-
sun and moon at ()()()() UTC for each day sun or moon is south of the equator. As a tion, indicated by the concentric citcles, is
of the year (GHA and DECL). Also matter of information, negative hourly in- seen to be 42.9° degrees. (The outer edge
shown are hourly increments (HR. crement figures indicate the sun or moon of the charts corresponds to 0° e1eyation,
INCR.) for both GHA and decl., which is moving south, while positive figures in- or the horizon, while 'the center cor-
may be used to interpolate the position of dicate northward movement. In inter- responds to 90° elevation, at zenith.) In-
the sun or moon for any time on a. given polating for flositions at times other than termediate angles of LHA and dec!. may
date. Tables 4 and 5 further indicate the ()()()() UTC, however, it is necessary only to be interpolated on the chart to obtain the
semi-diameter (SD) of the moon in observe algebraic signs listed in the ta~les az-el information.
and to add the values algebraically.
minutes of arc for eacllday of the year. If your latit~de is 40 degrees south of
This data may be used as an aid in deter- If a polar mount 3 is used, information the equator, Fig. 21 still applies, but all
from Tables 4 through 7 may be used
mining path losses, as previously dis- the data is "inverted." To enter the chart,
cussed. directly to point the antenna array. The first multiply the decl. angle by - 1, and
local hour angle (LHA) is simply the
subtract the LHA from 360. If the LHA is
Using the Tables GHA plus or minus the observer's 320° and the dec!. is 10°, we would find
The hourly increment (HR. INCR.) in- longitude (plus if east long., minus if the az-el information at the intersection of
formation from Tables 4 through 7 may west). The· LHA is the angle west of the the 10 X (- 1) or - 10° declination line
observer's meridian at which the celestial
be used to make linear interpolations for and the 360 - 320 or 40° LHA line. The
body is located. elevation angle may be read, directly,
the positiens of the sun or moon for any
time on a given date. Because the orbital Azimuth and Elevation for the Moon 26.9° in this case. The azimuth angle is
velocity of the moon is not constant, An antenna system which is positioned 180° from that indicated in Fig. 21. From
linear interpolations will yield some small in azimuth (compass direction) and eleva- the chart we read 225.7 0; the actual
tion (angle above the horizon) is called an azimuth is 225.7 .,-'180 = 45.7°.
'The Nilutical Almanllc for the yea,..···, where •••• az-el system. For such a system, some ad- An observer in the southern latitudes
indicates the calendar year for the data, This annual may find it more convenient to turn the
publication is printed by the U ,S. Government ditional work will be necessary to convert
Printing Office, Washington, DC. It is available appropriate chart upside down and relabel
from the Superintendent of Documents and from the LHA and the declo lines in reverse
many dealers of marine products.
'Also see Bray and Kerchner, "Al1tenna Patterns from 'See Michael, ,"Tracking the Moon _ In Simple order from the way they are printed. The
the Sun," QST, July 1960. English," QST, July 1960. azimuth scale should also be relabeled so
,. ~\."
'" '
Fig. 17 - Chart for determining moon az-el information for an observer Fig. 18 - Moon az-el chart for an observer at 10· latitude.
at O· latitude. See text.
Fig. 19 - Moon az-el cnart for an observer at 20· latitude. Fig. 20 - Moon az-el chart for an observer at 30· latitude.
O· is now at the "top," where 180· is Fig. 21 nor Fig. 22 will provide accurate 40.3· elevation). At any latitude in the
printed. az-el information for your location, but' range of these charts, when the celestial
An observer at any latitude may find it data from the two may be interpolated. In body is well above the horizon, linear in-
more convenient to relabel the LHA lines the previous example, where the moon's terpolation should yield az-el information
with actual GHA values_ For example, .if GHA was 320· and the dec!. 10·, we with a total angular error of no more than
your longitude is 100· W., the printed O· learned that an observer at 40· N. lat. 0_6 degree, assuming the' charts are read
LHA line for your location corresponds would aim his antenna at 119.10 azimuth, with 0.2-degree accuracy. '
to a GHA of 100·. Values for lines.to the 42.9· elevation. From Fig. 22 we see that
an observer ai 50· N. lat., for the same Azimuth and Elevation jor the Sun
right will correspond to Greenwich Hour
Angles of 90·, 80·, 70·, and so on, while position of the moon, would position his Visualize two observers on opposite
those to the left will correspond to 110·, antenna at 126.4· azimuth, 37.4· eleva- sides of the earth who are pointing their
120·, etc. Relabelihg the chart in this tion. A linear interpolation of these values antennas at the moon. Imaginary lines
manner will eliminate the need to convert for 45· N. lat. yields 122.8· azimuth and representing the boresights of the two
GHA to LHA information for each az-el 40.2· elevation. This result is reasonably antennas will converge at the moon at an
determination. close to the values obtained by angle of approximately 2·. Now assume
Suppose your latitude is 45· N_ Neither mathematical means (123_0.· azimuth and these observers aim their antennas at some
14·16 Chapter 14
Fig. 21 - Moon az-el chart for an observer at 40' latitude. I Fig. 22 - Moon az-el chart for an observer at 50' latitude.
Fig. 23 - Moon az-el chart for an observer at 60' latitude. Fig. 24 - Moon az-el chart for an observer at 70' latitude.
distant star_The bo.resight lines now may Suppo.se o.ur latitude is 40°, the LHA o.f mathematical means fo.llo.ws a pro.cedure
be co.nsidered to. be parallel, each o.bserver the sun is 320°, and the sun's declinatio.n· similar to. calculating great-circle bearing
having raised his antenna in elevatio.n by is 10°. Fig 21 indicates an antenna and distance fo.r two. po.ints o.n the earth's
approximately 1 ° _ The reaso.n fo.r the azimuth o.f 119.1 ° and an elevatio.n o.f surface. Equatio.ns fo.r az-el calculatio.ns
necessary change in elevatio.n is that the
earth's diameter in co.mpariso.n to. its
42.9°, if the mo.o.n were at this ·LHA and
declo The azimuth angle, 119.1°, is also.
are:
sin E
.
= (sin lat.) (sin dec\.)
dIstance from the mo.o.n is significant. The co.rrect fo.r the sun. To. o.btain the co.r-
rected elevatio.n angle for the sun we enter
+ (co.s lat) (co.s dec\.)
same is no.t true fo.r distant stars, o.r fo.r (co.s [lo.ng. -' GHA])
the sun. Table 8 with the indicated elevatio.n, (Eq.l)
Figs. 17 thro.ugh 24, altho.ugh prepared 42.9°, and determine the co.rrected eleva-
fo.r the mo.o.n, may be used fo.r the sun if a tio.n angle, 43.6". So.me interpo.latio.n may tan F = (sin E) - K (Eq.2)
small co.rrectio.n in elevatio.n angle is be necessary when using Table 8. Co.s E
made. Table 8 lists co.rrected elevatio.n
angles fo.r po.sitio.ning an antenna system Az-El Data by Mathematical Means C sin dec\. - sin lat. sin E
o.n the sun. Determining az-el info.rmatio.n by Co.s = co.s lat. co.s E (Eq. 3)
14-18 Chapter 14
:;.;
tt:
l!::l
,,,
1+ ::
Fig. 25 - Chart recording of moon echoes received at W2NFA on July 26, 1973, at 1630 lITC. Antenna gain 44 dB; transmitting power 400 watts and
system temperature 400 K.
The reflected wave shown in Fig. 26B the moon from moment to moment. Since (not depth of fade) is coincident with
consists of many scattered contributions the lunar surface is very irregular, the minimum total Iibration:. Calculation of
from the rough moon surface. It is reflected wave will be equally irregular, minimum total libration is at best tedious
perhaps easier to visualize the process as if changing in amplitude and phase from and can only be done su'tcessfully by
the scatttering were from many small in- moment to moment. The resultant con- means of a digital computer. It is a prob-
dividual flat mirrors on the moon which tinuous summation of the varying lem in extrapolation of rates of change in
reflect small portions (amplitudes) of the multi path signals at, your antenna feed coordinate motion and in small dif-
incident wave energy in different direc- point produces the effect called Iibration ferences of large numbers.
tions (paths) and with different path fading of the moon-reflected signal.
lengths (phase). Those paths directed The term lib ration is used to describe EME Operating Techniques
toward the moon arrive at your antenna small perturbations in the movement of Most EME signals tend to be near the
and appear as a collection of small wave celestial bodies. Earth Iibration consists threshold of readability, a condition
fronts (field vectors) of various mainly of its diurnal rotation; moon Iibra- caused bya combination of path loss,
amplitudes and phases. The vector sum- tion consists mainly of its 28-day rotation Faraday rotation and libration fading.
mation of all these coherent (same fre- which appears as a very slight rocking mo- This weakness and unpredictability of the
quency) returned waves (and there is a tion with respect to an observer on earth. signal has led to the development of
near-infinite array of them) takes place at This rocking motion can be visualized as techniques for exchange of EME informa-
the feed point of your antenna (the col- follows: Place a marker on the surface of tion that differs from those used for nor-
lecting point in the antenna system). The the moon at the center of the 'moon disc, mal vhf work - the usual RST reporting
level of the final summation as measured which is the point closest to the observer, would be jumbled and meaningless for
by a receiver can, of course, have any as shown in Fig. 27. Then over a period of 'many EME contacts. Dahs are often
value from zero to so'me maximum. time we will observe that this marker chopped into pieces, a string of dits would
Remember now that we assumed the earth wanders around within a small area. All be incomplete, and complicated words
and moon were stationary, which means this means is that the surface of the moon would make no sense at all.
that the final summation of these as seen from the earth is not quite fixed Unfortunately, there is no universal
multi path signal returns from the moon but changes slightly as different areas of agreement as to procedures for all the
will be one fixed value. The condition of the periphery are exposed pecause of this bands, although there is similarity. Two-
relative motion between earth and moon rocking motion. Moon Iibration is very meter operators generally use the
being zero is a rare event which will be slow (on the order of 10 - 7 radians per "T M 0 R" system, while those on 432
discussed later in this section. second) and can be determined with some MHz use a similar system but applied at
Consider now that the earth and moon, difficulty from published moon ephemeris somewhat different levels of readability.
are moving relative to each other (as they tables. The meanings, and typical use, of each
are in nature), so that the incident radio Although the lib ration motions are very part of the, sequence are given in Tables 9
wave "sees" a slightly different surface of small and slow, the large surface area of through 12.
the moon, having nearly an infinite
number of scattering points (small area),
means that even these slight geometric
~ ~ROUGH
movements can alter the total summation
INCOMING ft MOON
{ROUGH MOON of .the returned multipath echo by a
PLANE f_ SURFACE - suRFACE
significant amount. Since the Iibrations of
WAVE t \ REFLECTED
t WAVE the earth and moon are calculable, it is
(A) (B) only logical to ask if there ever occurs a
time when the total Iibration is zero or
Fig. 26 - How the rough surface of the moon
near zero? The answer is yes, and it has Fig. 27 - The moon apppears to "wander" in
its orbit about the earth. Thus, a fixed marker
reflects a plane wave as one having many field been observed and experimentally verified on the moon's surface will appear to move
vectors. on radar echoes that minimum fading rate about in a circular area.
Specialized Communications Systems 14·19
Table 9 "even" sequence. Note that odd or even .ing. High.speed cw is hard to copy at
Signal Reports Used on 144-MKz EME refers to the sequence number, not the marginal signal levels for most amateurs,
T - Signal just detectable minutes designated within that sequence. and the fading that is typical of an EME
M - Portions of calls copied On 432 MHz, there are 12 ~equence path can make it well nigh impossible to
o - Complete call set has been received periods to the half hour. The eastern-most decipher the content.
A - Both "0" report and call sets have been station calls first, and since two
received Other Modes
SK - End of contact 2-112-minute periods fill a 5-minute
space, it works out conveniently that the Only a few stations have the capability
eastern (or first) station will call starting of sending (and receiving) signals of a
with every five minute mark, and start . strength sufficient to allow experimenta-
Table 10 listening 2-112 minutes later. Thus a tion with other than cw. There have been
Signal Reports Used on 432·MHz EME schedule starting at 0030 would be an some ssb contacts and echoes of RTTY
"odd" period, although operators on 432 and fm signals have been heard, but no
T - Portions of calls copied
M - Complete calls copied MHz seldom label them as such. It is con- two-way communication by these latter
o- Good signal - solid copy (possibly venient for the operators to simply start modes has been accomplished to date. In
enough for ssb work) with the eastern-most station calling on general, only the stations with large
A - Calls and reports copied the hour or half hour, unless arranged parabolic reflectors try these more dif-
SK - End of contact
otherwise. ' ficult means of EME work. Such installa-
Of course there is much room for tionsare often "borrowed" from some
change in these arrangements, but they do research program for the amateur
Table 11 serve as vital guidelines for schedules. As endeavors.
144·MHz Procedure - 2·Mlnute Sequence signals become stronger, the rules can be
relaxed to a degree, and after many con- Frequencies
Period 1-112 minutes 30 seconds
1 Calls (W6XXX de tacts, stations can often ignore them com- Most amateur EME work is conducted
W1XXX) . pletely, if the signals are strong enough. within a few ,kHz of some convenient spot
2 W1XXX de W6XXX I TTTT Calls are often extremely difficult to frequency. On 144 MHz there is some
3 W6XXX de W1XXX, 0000 hear in their entirety. A vital dit or dah room to move about, but most operation
4 AO AO AO AO . de W1XXX k
5 AAAAA de W6XXX k can be missing, which can render a com- is very near the low edge, consistent with
6 QAZ? EME de W1XXX k plex call unreadable. To copy both calls the ability of the station to stay within the
completely requires much patience and a band. Thesituation is further confused by
good ear. Both calls must be copied, the requirement that Technician licensees
because even though most work is by must stay above 145 MHz. There are some
Table 12 timed schedules, there can be last-minute EME enthusiasts among the Technician
432·MHz Procedure - 2·112·Mlnute Sequence substitutions because of equipment trou- class licensees, and those amateurs with
ble at one station, unexpected travel, phm large enough antennas can work both por-
Period 2 minutes 30 seconds
1 VE7BBG de K2UYH
changes or the like, which make it im- tions of the band. The antenna problem is
2 K2UYH de VE7BBG possible for the scheduled station to ap- mentioned because most high-gain anten-
3 VE7BBG de K2UYH TTT pear. Thus, rather than have one station nas will not maintain their performance
4 K2UYH de VE7BBG MMM spend the entire period listening, only to over such a wide portion of the band.
5 AM AM AM AM de K2UYH k Operation on 432 ,MHz is generally
6 AAAAA de VE7BBG SK
find that no one was there, a system of
standby stations is becoming more within I or 2 kilohertz of that frequency,
popular. This is good, because nothing with a few stations going as far· afield as
will demoralize a newcomer faster than 431.997 or 432.003 for general schedules.
At the moment, there is no widespread several one-sided schedules. It is not unknown for a pair of stations to
system in use for bands other than 144 An exchange of signal reports is a useful move up 10 or 15 kHz for a contact, while
and 432 MHz. There are so few par- and required bit of information: useful others are on the nominal ".000" spot.
ticipants on 50 and 220 MHz that because it helps in evaluating your station There has been a movement afoot recently,
presumably they will have no difficulty in performance and the prevailing condi- to reach a gentleman's agreement to avoid
arranging techniques by correspondence tions at the time, required because it is a any short-range, local, or non-DX opera-
prior to scheduies for EME tests. The "non-prearranged" eXGhange, thereby re- tion between 432.0 and 432.025 MHz - a
amount of operation on 1296 MHz is low, quiring that you copy what was sent as concept that most EME enthusiasts
but on the increase. Perhaps an operating part of the contact. Obviously there are heartily approve of.
technique can evolve that will be accept- other things that could be incIiided in an For operation on 1296 MHz, most sta-
able to those on any band. The important "exchange of unknown information," tions are within a very few kilohertz of
consideration is that an. exchange of infor- and when conditions permit stronger that spot frequency. Many devices, tubes
mation takes place. This information signals, many operators do include names, and transistors, would work much better
should include three basic parts: calls of elaborate on the signal reports, arrange at the low end of the 1215-MHz band, but
both stations, signal reports, and confir- next schedule times, and so on. Unfor- the 1296.0 spot became popular because it
mation that previous information was tunately, such exchanges are rare. was convenient to triple from an existing
received. Confitmation is essential for comple- 432-MHz exciter.
, In the schedule sequence for both 144 tion of the exchange. There is no way that Of course, it is obvious that as the
and 432 MHz, the initial period starts. on you can be sure that the other operator number of stations on EME increases, the
. the hour, but because of the difference in copied what you sent until you hear him frequ'ency spread must become greater .
sequence lengths for 'the two bands, say so. That final R or "roger" means that Since the moon is in convenient locations
schedules starting on the half hour will he has copied your information, and your only a few days out of the month, and
_ not be the same. On two meters, there are two-way contact is complete. only a certain number of stations can be
15 sequence periods to the half hour, a
Sending speed is usually in the 5- to scheduled for EME during given even-
which would make the period 0030 to lO-wpm range, although it can be ad- ing, the answer will be in the use of
0032 an "even" sequence. This could justed according to conditions a{ld simultaneous schedules, spaced a few
make a difference, depending on which operator skill. Characters sent too slowly kilohertz apart., The time may not be too
\ operator was assigned an "odd" or tend to become chopped up and confus- far away - QRM has already been
14·20 Chapter 14
experienced on each of our three most A moderately sized Vagi array has the' use 160-element or larger systems. As with
active EME frequencies. advantage that it is relatively easy to con- Vagi and quagi antennas, the' collinear
struct and can be positioned in azimuth cannot be easily adjusted for polarity
EME Net In/ormation and elevation with commercially available c\langes. From a constructional stand-
An EME net meets on weekends at 1600 equipment. Matching and phasing lines point there may be little difference in
and 1700 UTC for the purpose of arrang- present no particular problems. The main complexity and material costs between the
ing schedules and exchanging pertinent in- disadvantage of a Vagi array is that the collinear and Vagi arrays.
formation. The frequency of operation is polarization plane of the antenna cannot The parabolic dish is another antenna
14.345 MHz. . . be conveniently changed. One way around that is used extensively for EME work.
this would be to use cross-polarized. Yagis Unlike the other antennas described, the
Antenna Requirements and a relay switching system to select the major problems associated with dish
The tremendous path loss incurred over desired polarization. This represents a antennas are mechanical ones. Dishes 20
an EME circuit places stringent re- considerable increase in system cost and feet in diameter are required for successful
quirements on the station performance. complexity_ Polarization shift at 144 MHz EME operation on 432 MHz. Structures
Low-noise receiving equipment, maxi- is fairly slow and the added comph;:xity of of this size and wind/ice loading place a
mum legal power and large antenna arrays the cross-polarized antenna system may severe strain on the mounting/positioning
are required for successful EME opera- not be worth the effort. At 432 MHz, systems ..Extremely rugged mounts are re-
tion. Although it may be possible to copy where the shift is at a somewhat faster quired for large dish antennas, especially
some of the better-equipped stations while rate, an adjustable polarization system of- when used in windy locations. Several
using a single high-gain Yagiantenna, it is fers a definite advantage over a fixed one. aspects of the parabolic dish antennas
doubtful whether such an antenna could A photograph of the Vagi antenna make the extra mechanical problems
provide reliable two-way communication. system used at KIZZ is shown in Fig. 28. worth the effort. For example, the dish
Antenna gain of at least 20 dB is required The system consists of four, 2-meter antenna is inherently broadband and may
for reasonable success. Generally speak- CushcraftBoomer antennas mounted on be used on several different bands by
ing, more antenna gain will yield the most a 70-foot, Rohn 25 tower. A CDE Ham- simply changing ,.the feed. The graph at
noticeable improvement in station perfor- III rotator is used for positioning the Fig. 2~ relates antenna gain, frequency
mance, as the increased gain will aid both antenna in azimuth and a TET KR-500 and size. As can be seen, an antenna that
the received and transmitted signals. rotator is used for elevation control. The is suitable for 432 MHz work is also
Several types oJ antennas have become gain of this array is approximately 20 dB, usable for each of the higher amateur
popular among EME enthusiasts. Perhaps taking into account phasing line losses. bands. Additional gain is available as the
the most popular antenna for 144-MHz Quagiantennas (made from both quad frequency of operation is increased.
w.ork is an array of either four or eight and Vagi elements) are also popular for Another advantage of this antenna is in
long-boom (14- to 15-dB gain) Yagis. The EME work. Slightly more gain per unit the feed system. The polarization of the
four-Vagi array would provide approxi- boom length is possible as compared to feed, and therefore the polarization of the
mately 20-dB gain, and the eight-antenna the conventional Vagi. Additional infor- antennfl, can be adjusted with little dif-
system would show an approximate 3 dB mation on the quagi is presented in the ficulty. It should be a relatively easy mat-
increase over the four-antenna array. At VHF and UHF Antennas chapter of this ter to devise a system whereby the feed
43.2 MHz, eight or 16 long-boom Yagis are book. could be rotated remotely from the shack.
used. Vagi antennas are available com- The collinear is another popular type of Changes in polarization of the signal
mercially or can be constructed from . antenna for EME work. A 40-element col- could be compensated for at the operating
readily available materials. Information linear array has approximately the same position! As polarization changes can ac-
on maximum-gain Vagi antennas is frontal area as an array of four Yagis. The count for as much as 30 dB of signal at-
presented in the VHF and UHF Antennas collinear array would produce approxi- tenuation, the rotatable feed could make
chapter of this volume. The dimensions mately 1 to 2 dB less gain. Of course the the difference between working a station
presented are based on figures developed depth dimension of the collinear array is and not. A photograph of the parabolic
by the National Bureau of Standards for considerably less than for the long-boom dish antenna used at K2UYH is shown in
Vagi design. At least one manufacturer Yagis. An .sO-element collinear would be Fig. 30. More information on parabolic
has used the NBS design information for marginal for EME communications, pro- dish antennas is available in the ARRL
their latest series of high-performance viding approximately 19-dB gain. Many Antenna Book.
antennas. operators choosing this type of antenna Antennas suitable for EME work are by
..
..
the signals to be copied on EME are bare- mode, Kl connects the antenna to the ANT
ly, and not always, out ofthe noise, a low- transmit line. K2 switches the preamplifier
noise-figure receiver is a must. The mark into the 50-ohm termination. Hence, two
to shoot for at 144 MHz is something relays provide the isolation between the
under 2 dB, as the cosmic noise will then transmitter connection and the preamplifier.
be the limiting factor in the system.'Noise If independent control of K2 is provided
figures of this level are relatively ,easy to for, the preamplifier can be switched be-
achieve, even with inexpensive devices tween the 50"ohm termination and the
that are available. antenna during receive. This feature is
As Iowa noise figure as can be attained especially useful when making sun-noise
will be usable at 432 MHz. Noise figures measurements to check system perfor-
on the order of 0.5 dB are possible with mance. For this measurement the. antenna is
GaAS FETs. As most GaAS FETs are directed toward the sun and the preamplifier
,currently still quite expensive and is alternately switched between the 50-0hm
(B)
somewhat fragile, many builders choose load and the antenna: The dB difference can
the more rugged bipolar, which offers a be recorded and used as a reference when
noise figure just under 1 dB. . checking system improvements. The com- Fig. 31 - Two systems for switching a
Since the loss in the transmission line plete circuit for this relay' system is preamplifier in and out of the receive line. At
that connects the antenna to the presented later in this chapter. A, a single length of cable is us.ed for both the
transmit and receive line. At B is a slightly
preamplifier adds directly to the system As the preamplifier is mounted ahead more sophisticated system that uses two
noise figure, most serious EME operators of the transmissiotl. line to the receiver, a separated transmission lines. See text for
mount a preamplifier at the top of the cable of mediocre performance can be details.
tower or directly at the antenna. If an ex- used. The loss of the cable, as long as it is
ceptionally good grade of transmission within reason, will npt add appreciably to
line is available,it is possible to obtain the system noise figure. Information con- operators may wish to consider similar
almost as good results with the tained in the VHF IUHF Receiving systems for terrestrial work, as a tower-
preamplifier located in the shack. Two chapter of this book explains how to mounted preamplifier usually means a
relay Ipreamplifier switching systems calculate system noise figures. Foam-type noticeable reduction in system noise
suitable for remote mounting are outlined RG-8 cable is acceptable for runs up to figure. The thoughts behind this system
in Fig. 31. The system at A makes use of 100 feet at 144 MHz. \ are outlined in the previous section, en-
two relays and a single transmission line It is important to get as much transmit- titled Receiver Requirements.
for both transmit and receive. The ter power as possible to the antenna. For
preamplifier is simply switched "in" for this reason Hardline, Heliax or similar The Circuit
receive and "out" for transmit. low-loss cable is specified for the transmit The relay switching system is separated
line. into two parts. One section is mounted at
the tower, and the other, the control cir-
Transmitter Requirements cuitry and power supply, is mounted at
In many EME installations the antenna the station. A length of four-conductor,
gain is not much above the minimum re- TV -rotator cable can be used to connect
quired for communications. It is highly the two units. The package that is
likely that the maximum legal limit of mounted at the tower consists of two rf
power will be required for successful EME coaxial relays and a preamplifier. The
work on up through 432 MHz. Since schematic diagram is Shown in Fig. 33. As
many contacts may require long, slow can be seen, Kl is used to switch the
sending, the transmitter lamplifier should
have adequate cooling. Also, an amplifier
with some power to spare rather than an
amplifier running "flat out" is desirable.
This is especially important should ssb
communication be attempted. An
amplifier run all out on ssb will likely pro-
duce large amounts of odd-order IMD
products that fall within the band. While
the splatter produced will not affect your
communications, it will certainly affect
that of ,others close in frequency!
Remote Preamplifier Switching System Fig. 32 - The preamplifier relay switching
The preamplifier-switching system system suitable for EME work. The control box
is mounted at the antenna and the control box
Fig, 30 - A newcomer to EME stands in awe described here is intended primarily for is located in the statiQn. A length of four·
of the K2UYH 28-foot dish. EME applications. Serious vhf and uhf conductor wire connects the two units.
14·22 Chapter 14
r -----KI~---- - --~ r - -' - - - - - - - - - - - ,
K2A I
I I
I
I I
I I
I I
I I
ANT I
I I
I I'
I I·
I I
I
+~
I.
I
TX LINE I
I
I RI~
I" I
'- _____________ J I _______________ JI
~
•
RX LINE
VOLTAGE
KIB TO PRE-
AMPLIFIER
~ _0.01
'kV
rh
TOWER TB3
STATION 2 t
,...-<TB~~
SI
r-_---<CY 0-----<.----.. . TI r-_ _ _ _ _ _ _ _ _ _ _ _ _
ON S2 4 CONDUCTOR /
CONNECTING
~ CABLE
TERM rhANT. ~~
SHORT FOR
r<
~
TRANSMIT
TBI
r0
+12 VOLTS o,....----<.~--,---------------'----'
TO PREAMPLIFIER
Fig. 33 - Schematic diagram of the preamplifier switching system. The diagram is divided into two parts; the top portion is for the circuitry at the
antenna and the bottom is that for use in the station.
DS1 r Neon indicator light with built-in
dropping resisto"r. Resistor built' into a PL-259 connector. 115-Vac secondary, 15 VA. Stancor P-6411
K1, K2 - Rf-coaxial relays suitable for the fre- S1 - Toggle, spst. or equiv.
quency range to be used. 'S2 - Toggle, spdt. TB1-TB3, inci. - Terminal block, screw
R1 - Termination, 50-ohms, non inductive. T1 - Isolation transformer, 115-Vac primary, connection, four terminals.
antenna between the transmit line and a relay coils. S2 controls the action of K2, watertight enclosure. The item shown at
line which connects with K2. K2 switches which is either connected in parallel with the bottom center of the enclosure is a
the preamplifier to either the antenna or Kl in 'the ANTENNA position,or activated commercial preamplifier. Power for the
to a 50-ohm termination. Two relays pro- separately for the TERMINATION position. preamplifier (12 V dc) is fed through the
vide more than adequate isolation be- A pilot light, fuse and ON/OFF switch are fourth wire of the four-wire cable that
tween the preamplifier and the transmit provided in this design. connects the two modules (tower and sta-
line_ Additionally, K2 can be switched tion) together. When all components are
between the antenna and the termination Construction properly mounted, the chassis is sealed
independently of Kl. This allows for sun- The iteqts to be mounted at the tower with silicone rubber (RTV). A terminal
noise measurements ~hen the system is in are enclosed in an ordinary chassis and block provides for connection to the four-
the receive mode. bottom plate assembly_ A photograph, conductor cable. Two O.OI-I"F capacitors
The portion of the system mounted at shown in Fig. 34, indicates the general are mounted across the relay coils at ter-
the station is essentially a power supply layout. Short lead lengths are used minal block TB3.
and control circuitry. A line isolation throughout. Bulkhead uhf feedthrough The station circuitry is mounted in a
transformer is used to power the 117-volt connectors are used to ensure an rf- and small aluminum cabinet. A neon in-
Specialized Communlcatlo~s Systems 14·23
sources on condition that they will be used and tape perforator are .required for this'
purely for amateur purposes and will not process. A reper!orator is a device that
. be resold for commercial use. may be connected to the conventional
Some dealers and amateurs around the teletypewriter for punching tape when the
country make it known by advertising that machine is operated in the regular way. It
they handle parts or may be a source for may thus be used either for an original
machines and accessory equipment. message or for "taping" an incoming
QST's Ham Ads and other publications message for later retransmission.
often show good buys in equipment as
amateurs move about, obtain newer Start, Data, Stop Pulses
equipment,. or change interests. In a teleprinter machine, the normal
Periodic publications are available "rest" condition of the selector-magnet
which are devoted exclusively to amateur solenoids is with loop current on. Inter·
RTTY. They carry timely technical ar- ruption of the loop current releases the
Fig. 34 - Interior view of the package that is ticles and operating information, as well selector magnet, allowing rotation of a
mounted at the antenna. The object at the bot·
tom center of the chassis is a commercial
as classified ads. Over the years QST has cam in the machine. Transmission of a
preamplifier. carried a number of articles on all aspects TTY character begins with a space pulse
of RTTY. For a list of surplus equipment (current off), called the start pulse. The.
dealers, information on publishers of start pulse signals to the machine that
RTTY periodicals, and a bibliography of reception of a character has begun. im-
dicator, . onloff power and termination all articles on RTTY which have appeared mediately after the start pulse, a series of
switch are mounted to the front panel. in QST, write to ARRL, 225 Main Street, data pulses is transmitted with mark or
The fuse and interconnection terminal Newington, CT 06111. U.S. residents space condition as indicated by the en-
blocks TBI and TB2 are mounted on the should enclose a stamped business-size coding for the desired character.. The
rear apron. Component layout is not at all:' envelope bearing a return address with number of data pulses used to represent
critical. their request. the letters, numbers and symbols varies
with the TTY code being used; Baudot
· Additional Thoughts Types of Machines code uses five data pulses, ASCII uses
Although the circuit described here per- There are two general types of eight. Immediately after the last data
formed flawlessly for many months, one machines, the page printer and the tape pulse, a stop pulse is included which is
· change might be considered by the pro- printer. The former prints on a paper roll always a mark pulse. The stop pulse,
spective builder. The change would in- about the same width. as a business let- therefore, always occurs in a fixed time
volve rewiring the relays so that the terhead. The latter prints on paper tape, after the start pulse (after five data pulses
preamplifier would be automatically usually gummed on the reverse side so it in Baudot and eight in ASCII). The stop
switched into the termination when the may be cut to letter-size width and pasted pulse gives the machine a "rest'time" to
system is de-energized. This would' protect on a sheet of paper in a series of lines. The prepare for the beginning of the next
the preamplifier from static or nearby page printer is the more common type in character, mai.ntaining receive machine
lightning strikes. As the circuit is presently the equipment available to amateurs. synchronization with the transmitted
shown, the preamplifier will remain con- The operating speed of most machines signal. The time length of the start and
nected to the antenna when the power is is such that characters are sent at .the rate each data pulse are the same and are often
switched off. Although no damage has oc- of either 60, 67, 75 or 100 wpm depending called the unit-pulse or select-pulse time.
curred to' the preamplifier used by the on the gearing ratio ofa particular The stop-pulse length varies from code to
author, some GaAS FET amplifiers may machine. Current FCC regulations allow code and even with speeds within a code,
not be able to tolerate the voltage levels amateurs the use of any of these' four as will be explained later. In general, the
produced by nearby lightning storms. speeds. Interchangeable gears permit most minimum stop-pulse length can be one or
machines to operate at. these speeds. Or- two times as long as the unit-pulse time;
RADIOTELETYPE dinary teletypeWriters are of the start"stop stop pulses may be as long as desired since
Radioteletype (abbreviated RTTY) is a variety, in which the pulse-forming the machine is "at rest" until the next
form of telegraphic communication mechanism (motor driven) is at rest until a start pulse is received. This type of TTY
employing typewriterlike machines for (I) typewriter key is depressed. At this time it code that uses start, data and stop pulses
generating a coded set of electrical im- begins operating, forms the proper pulse in the construction of each character is
pulses when a typewriter key correspond- sequence, and then comes to rest again called an asynchronous or start-stop serial
mg to the desired letter or symbol is before the next key is depressed to form code. Other codes also in commercial use
pressed, and (2) converting a received set of the succeeding character. The receiving include synchronous serial codes, in which
such impulses into the corresponding mechanism operates in similar fashion, start and stop pulses are not attached to
printed character. The message to be sent being set into operation by the first pulse the data pulses for each character, and
is typed out in much the same way that it of the sequence from the transmitter. parallel data codes, in which each data
would be written on a typewriter, but the Thus, although the actual transmission pulse is assigned a separate wire to and
printing is dorre at the distant receiving speed cannot exceed about 60 wpm (or from the terminal device. Such codes are
point. The teletypewriter at the sending whatever maximum speed the machine is found in common use with computer and
point may also print the same material. geared for), it can be considerably slower, line-printer devices. FCC regulations cur-
,The teleprinter machines .used for depending on the typing speed of the rently authorize amateurs to use either the
RTTY are far too complex mechanically operator. Baudot or the ASCII serial asynchronous
for home construction; and if purchased It is also possible to transmit by using TTY codes.
new would be highly expensive. However, perforated tape. This has the advantage
· used teletypewriters in good mechanical that the complete message may be typed The Baudot TTY Code
I •
condition are available at quite reasonable out in advance of actual transmission, at One of the first data codes used with
prices. These are machines retired from any convenient speed; when transmitted, mechanical printing machines uses a total
commercial service but capable of entirely however, it is sent at the machine's nor- of five data pulses to represent the
satisfactory operation in amateur work. mal maximum speed. A special tape alphabet, numerals and symbols. This
They may be obtained from several reader, called a transmitter-distributor, code is commonly called the Baudot or
14-24 Chapter 14
MARK
"''''"1
CONDITION
(CURRENT ON)
~
CHARACTER
I
I
,...--
;., -r '""'' ' '" ~
~---
Table 13
The Baudot Data Code
Bit Number
54321
ooodo
Letters
U,S.
Figures
BLANK BLANK BLANK
CCITT No.2
Figures
I 00001 E 3 3
I 00010 LF LF LF
SPACE
(CURRENT OFF)-- ' - / ,
:
I ,.,
~
2
(s)
,I
y
4
(m)
T
(5)
)
/
1 1
-
(m)
, '"""2
( 5)
00011
00100
00101
00110
00111
01000
A
SPACJ,:
S
I
U
CA
SPACE
BELL
8
7
CA
SPACE
8
7
CA
START STOP START 01001 0 $ WAU
PULSE TIME ~ 'PULSE PULSE 01010 A 4 4
01011 J BELL
01100 N ,
Fig, 35 - Time sequence of typical Baudot character, the letter 0, 01101 F .!
01110 C
01111 K
Murray telegraph code, after the work Keyboards on Baudot machines such as 10000 T 5 5
done by these two pioneers. Although the Teletype Corp. models 15 and 28 dif- 10001 Z +
10010 L ) )
commonly called the Baudot code in the fer from standard typewriter keyboards, 10011 W 2 2
United States, a similar code is usually having only three rows of keys with the 10100 H # .£
called the Murray code in other parts of related letter and number/symbol on each 10101 Y 6 6
the world and is formally defined as the keytop - Q and. I, K and (, and so on). 10110 P 0 0
10111 Q 1 1
International Telegraphic Alphabet No.2 The typist soon discovers this difference! 11000 0 9 9
Baudot Code in part 97.(1) of the FCC Newer electronic terminals such as the 11001 B ? ?
Rules and Regulations. This standard HAL DS2000 and D~3100 have standard 11010 G & &
11011 FIGS FIGS FIGS
defines the codes for letters, numerals and keyboard arrangements and automatically 11100 M
the slant or fraction bar but allo'oVs varia- insert L TRS or FIGS characters as they are 11101 X
tions in the choice of code combinations needed. The Baudot code itself is 11110 V
for punctuation. U.S. amateurs have restricted to upper-case letters only since 11111 LTAS LTAS LTAS
generally adopted a version of the so- insufficient codes are available to repre- Note: FIGS·H (10~0) may also be used for
called "Military Standard" code, arrange- sent lower-case letters. MOTOA STOP function, "1" = mark = hole in
punched tape
ment for punctuation, largely because of The Baudot code has seen extensive
the ready availability of military surplus commercial use throughout the world and
machines in the post-1945 years. is still actively utilized for international
Amateurs in other countries (particul,arly wire, press and weather communications. U.S. and CCITT No.2 alphabet. Notice
in Europe) have standardized on the Inter- Because of the ready availability of that the waveform drawing of Fig. 35
national Consultive Committee for Baudot mechanical equipment, this code shows the current waveform, with mark
Telephone and Telegraph (CCITT) No.2 will continue to be quite popular among represented by the upper deflection. Also,
code arrangement, which is similar to the radio amateurs. Nevertheless, the lack of the bits in Fig. 35 are arranged in a left-to-
U.S. standard but has minor symbol and code space for control, extended punctua- right order, as would be observed on an
code-arrangement differences. tion or lower-case letters is a severe limita- oscilloscope. The bits in Table 13,
Since each of the five data pulses can be tion of the five-unit Baudot code. These however, are arranged in descending
in either a mark or space condition (two limitations are particularly inconvenient order (b5 to bl), conforming to the stan-
possible states per pulse), a total of in computer-terminal applications, even dard binary representation. Thus the let-
2X2X2X2X2 =2s = 32 different though Vilrious serial and parallei data- ter 0 shown in Fig. 35 would be written as
code combinations are possible. Since it is coding schemes have been used with com- the binary character 01001.
necessary to provide transmission of all 26 'puters. Fig. 35 shows a time diagram of
letters, 10 numerals and punctuation, the • typical Baudot characters, and Table 13 ASCII
32 <;ode combinations are not sufficient. shows the Baudot data code for both the In 1968, the American National Stan-
This problem is solved by using the codes
f';"
twice; once in the letters (LTRS) case and
again in the figures (FIGS) case. Two
special characters, LTRS and FIGS, are
used to indicate to the printer whether the
following characters will be of the letters
or figures case. The printer has a latching
CONDITION
"''''"j '"' ' ' ' ".,
MARK
mechanism that "remembers" or stores (CURRENT ON) I ;-- ;-- ---
the last received LTRS or FIGS character so I I
I I
that it remains in the last received case un-
11 '
I
I
til changed. Control operations such as
LTRS, FIGS, carriage return (CR), line feed
SPACE __ I
2 3
I
4 5 '--s 7 8
-
"0"'" "" (5) I
=
II
(m) (m) ( 5) ( 5) (m) ( 5) (m)
(LF), space bar (SP) and blank (BLNK no
print or carriage movement) are assigned ,l )
y
to both the LTRS and FIGS case so that they START
DATA PULSES FOR "5
STOP
11
START
can be sent in either case. The remaining PULSE
~
PULSE PULSE
TIME
26 code combinations have different letter
or numeral/symbol meanings, depending Fig, 36 - Time sequence of typical ASCII character; the letter s, The eighth or parity bit may be
upon whether preceded by a L TRS or FIGS set for any of four conditions: (1) always mark, (2) always space, (3) odd parity or (4) even parity,
character. All four choices are in common usage, '
choices of gear ratios and motor-shaft rate. A limited amount of amateur use of nected, text typed on one keyboard is
speeds have resulted in the use of 74 baud ("100 speed") has been noted on reproduced on both printers. Connection
non integer wpm rates. Common usage, the high-frequen.cy bands'. Most commer- of the keyboard directly to its associated
however, has rounded the exact speeds to cial RTTY transmissions on high frequen- printer is called a local loop and results in
easily remembered numbers. Thus, "60 cies use 50, 57 and 74 baud, with little what is called half duplex (HDX), giving
speed" Baudot is actually sent at 61.33 45-baud activity. local copy of transmitted text, termed
wpm and "75 speed" is really 76.67 wpm. ASCII data rates are. commonly local echo. ' '
A major problem occurs with the use of specified as a baud rate, although a Selector magnets have been designed
words per minute as a TTY speed character-per-second (cps) or words-per- for mark loop currents of 60 or 20 'rnA dc,
specification because of the varying length minute (wpm) rate may also be given. The with 60 rnA being by far the most com-
of stop pulses in use. For example, "60 lowest standard ASCII data rate in com- mon for older machines such as the
speed" Baudot TTY has 22-ms-long start mon usage is 110 baud. ASCII characters Teletype Corp. model 15 or 28. Newer
and data pulses and a 31-ms stop pulse; sent at 110 baud are usually sent with a Baudot machines and most ASCII
the Western Union "65 speed" also has 2-unit-wide stop pulse, although the machines and terminals use electronic in-
22-ms start and data pulses, but the stop I-unit stop pulse may also be found in terface circuits that accept a wide range of
pulse is also 22-ms long; electronic ter- some applications. Above 110 baud, it is loop currents (10 to 120 rnA for the HAL
minals commonly use 22-ms start and data common to make the stop pulse one unit DS3100, for example); a 20-mA loop cur-
pulses and 33-ms stop pulses (1.5 times the pulse in length. The standard ASCII data rent is quite commonly used with ASCII
data-pulse width). All of these three codes rates commonly used with asynchronous terminals.
are compatible and may be received on the serial transmission are shown in Table 13. Since the dc resistance of the machine
same printer or terminal since the stop- The ASCII data rates up to 300 baud selector magnets is rather low '(100 to 300
pulse length is a minimum time. The com- are authorized for U.S. amateur use on ohms, typically), it would at first seem
mon factor between these codes is the frequencies between 3.500 and 21.250 that a .low-voltage loop supply could be
22-ms length of the data, or unit pulse. MHz. Data rates up to 1200 baud are per- used. However, the inductance of the
Therefore, a new data-rate specification mitted between 28 and 225 MHz; up to magnet is usually quite high (on the order
has been adopted~ the baud rate, which is 19,600 baud maybe used above 420 MHz. of 4 henrys for a model 15), causing a
the reciprocal of the data- or unit- or The llO-baud rate is by far the mpst prac- delay in the current rise time. This, in
select-pulse width: . tical for 3.5 to 21.5 MHz use, again turn, delays the selector magnet response
because of the ready availability of equip- to a mark pulse, distorting the signal. This
Baud rate = I/t, where t = length of ment as well as the increased susceptibility distortion can be severe enough to cause
unit pulse. misprinting of received text, particularly if
of the higher data rates to noise, static, in-
Using this definition, all three of the terference and so forth. Vhf fm amateur other forms of distortion are present (such
above codes have a data rate of 45.45 activity finds 110 and 300 baud useful for as caused by variations in the radio
baud, commonly abbreviated to "45 terminal-to-terminal communications, signal). The effect of this inductive distor-
baud." and 300 and 1200 baud for computer- tion is reduced considerably if the L/R
As noted above, the length of the stop related activities such as exchanging pro- ratio (L is solenoid inductance and R is.
pulse varies between codes, being from grams and the like. total loop resistance) is reduced by in-
1.0 to 2.0 times as long as the unit (or creasing R. Increasing R requires that the
data) pulse; mUltipliers of t.O, 1.42, and Loop Circuits dc voltage be increased to maintain 'the re-
1.5 are commonly used with the Baudot As discussed earlier, the printing quired 60-mA loop current. In general,
codes. Standard Baudot data rates and mechanisms use solenoids or selector the higher the loop voltage and loop
speeds are shown in Table 12. magnets to sense the presence (mark) or resistance used, the lower the distortion.
U.S. amateurs are authorized to use all absence (space) of the loop current. The In practice, loop power-supply voltages
of the Baudot data rates shown in Table letters typed on the sending keyboard are between 100 and 300 Vdc are common;
12, with the exception of 100 baud. This encoded with proper mark and space 130- and 260-voIt supplies were often used
rate has seen -limited commercial use in pulses by mechanically driven keyboard with model 15, 19 and 28 Teletype
Europe. The 45-baud data rate is by far contacts. Since the keyboards and selector machines. Modern TTY systems use a
the most popular worldwide amateur data magnets of both machines are series con- 150- to 200-volt loop power supply and a
14-28 Chapter 14
camera equipment with color film is thusi.ast. It fits nicely inside a TV set and
popular in this work because it affords presents no problem if one wishes to
'on-the-spot processing. Color reproduc- mount it on the antenna.
tion by this technique can be quite good. The TVC-I, sensitive and selective, per-
More advanced experiments in color forms exceptionally well, yet it is an exam-
SSTV reception are being contemplated ple of simplicity. Performance is en-
by storing frame data in digital memory hanced by a commercially manufactured,
and by applying picture information to a double-balanced mixer and a voltage-
standard color TV set, along with encoded controlled oscillator that is tuned by a
red-green-blue data, via an rf modulator. lO-kO potentiometer. The latter. feature
Fast-scan TV (ATV) is permitted on the enables the converter to be tuned remotely
bands above 420 MHz where wider band- by a similar potentiometer. Such remote-
Fig. 39 - Top view of the ATV iransceiver. The widths needed for this mode can be ac- control operation \Vould be useful if the
power supply is at the left. A coaxial relay and commodated. This article is concerned converter is to be mounted on the antenna
Bird power sensor are near the center of the with the latter method. or some location apart from the operating
back apron. The TVC-1 tunable converter has
been placed atop a Janel converter box for size
position.
comparison. A 12-V regulator chip has been
A Converter for Reception Physically, the TVC-I components are
provided for the converter because the main Understandably a station that transmits mounted on a very small pc board. There
supply furnishes 15 volts to enable the power. well but is deficient in receiving for want is a choice of a 50- or 75-ohrrj input. A
amplifier to reach full output.
of a good receiver leaves much to be coaxial-cable connector is provided for
desired. Efforts by some amateurs to the output on channel 3. The output
modify the home-TV uhf tuner have signal may be fed to an ordinary television
sending of three separate frames of the proved rather disappointing. Others have . set for viewing the picture and hearing the
same picture with a red, a green, and a geared up homemade devices that failed sound. No modification of the TV set is
blue filter successively placed in front of to live up to expectations. What is the required. Therefore the TV receiver may
the camera lens' for each or" the three alternative? be used in the normal manner for home
frames. At the receiving end of the circuit, P. C. Electronics, which produces the entertainment without any incon-
corresponding filters are used and each units one needs for an ATV station, has veniences.
frame is photographed .on color film. resolved the matter of providing a means As manufactured, the converter module.
After a tricolor exposure is made, the for obtaining excellent reception. The is supplit;d pretuned and ready to connect
photograph is developed and printed in TVC-I converter, available in ready-made to the ATV system. The TVC-I should be
the normal manner. The use of "instant" form, is tailor-made for the 439-MHz en- installed in a shielded chassis or enclosure.
REMOTE
TUNE
DO NOT EXCEED
+1!lVDC
RF
OUT
M~IC PHONE
114"
PH JACK
KIA
0:L RELAV
COIL
+14-I!lV
CAMERA
4!1MHz
OUT
Fig. 40 - Block diagram of the ATV transceiver. Principal units of the circuit are the VHF Engineering TX-432 transmitter. Motorola MHW·710
amplifier, P. C. Electronics TVC-1 converter, VM·3 video modulator, FM·A5 audio subcarrier unit and video i·d generator. Output of the converter is
connected to a standard TV set. The IC is a type 7812 12·V dc regulator.
GROUND
~;,"O
station. After the station has been found
by tuning with t,he lO-kO potentiometer, :TOB+
fine adjustments are made with Cl. When
the latter is properly set, tuning will be .~~4W----------'----------.... ~~~T
good for ± 10 MHz. For installations C44 T •
~~ ~~ ~
where the converter is mounted on the - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - .... INPUT
antenna, the lO-kO frequency control on
the VCO may be replaced by the remote Fig. 41 - Two minor changes are all that need be made to modify the TX·432 transmitter for ATV.
tuning circuit shown in Fig. 43. Use of this C28 and the jumper connecting it with RFC3 are removed. The line input at the lower right of the
arrangement allows the tuning ad- diagram is not used. .
.....--oTO L7
VIDEO IN
CONNECT TO
TX-432
---------,
I
~D 4.5-MHz I
,..;.:;'" INPUT
I
C2:J:
I
I
PART OF
I
TX-432 I
I
I
AMIDON
BEAD
AMIDON
BEAD
2N3055 * I
_ _ _ _ _ _ _ _ _ _ ...1I
SHORT LEADS
+100pF
T·25V
.2L
o~rh
RG58
1I2W
1N4OO1
FROM
TX4238 MDA80i
OUT
RG58
+12V DC
TO VM-3
AND FM-A5
~
RFAMP. veo
01 02
, !lO;1 Cl O.OIH_ _....,Mt'PS_-~H8
....3,...-_ _ _ _~---, MPS-H81
8 47
SBL-1
DOUBLE
BALANCED
MIXER 7
2
4 33 3
' - - -...........- - -.....--()+12V
4,7#H
O.22 .. H
I
I _
E
rt-----"
Fig. 43 - The circuit of the TVC-1 converter consists of a low-noise, high-gain, rf amplifier, a doubly balanced mixer and a Varicap-tuned VCO.
Remote tuning of the conve'rter permits it to be mounted on the antenna if desired. For remote operation, theremote tuning circuit replaces the
10-kll potentiometer shown below the oscillator transistor, Q2. C2 and C3 are Arco 400, 1-10 pF, capacitors. L2 consists of 1-112 turns no. 22wire,
1I4-inch diameter, tapped 3/8 inch from the lower end. L3 is a hairpin loop, 1/2 inch across the bottom and 5/8 inch high. It is made with no. 22 bus
wira '
+12V
+12V
100k 2200
----.---r
TO
470 2~00 VM-3 OR FM-A~
~
4.5MHZFM
OUTPUT
±25 kHz DEV
470 LEVEL
1000 1N914 ADJ.
100k
1000 MODULATOR
FM'AS
Fig, 44 - This ATV sound·subcarrier generator module.permits both voice and video to be transmitted. The trap shown in the diagram is nec~ssary
to isolate the capacitance of the long coaxial cable from the FM·A5 output. The capacitance would act as a bypass to the 4.5·MHz signals without
the trap. U1 is either a Motorola MC1458CP1 or Raytheon RC45580N operational amplifier.
,14-32 Chapter 14
/
COHERENTCW
While spectrum management has
received much attention in the recent
Amateur Radio literature, the problems
and possibilities of "more QSOs per
kilohertz" were first recognized more
than half a century ago. Frederick
Emmons Terman, 6FT, presented his vi-
sion of narrow-band communications in
"Some Possibilities of Intelligence
Transmission When Using a Limited '(B)
Band of Frequencies," published in Pro-
ceedings 0/ the Institute 0/ Radio Fig. 46 - Block diagram of the basic frequency compandor scheme. The circuit at A is used to
Engineers, January 1930. compress the audio bandwidth on transmit. The circuit at B expands this signal for receive.
An important part of Terman's paper
deals with the theoretical bandwidth re-
quired for radiotelegraphy. As early as As discussed in Chapter II of this The receiver output is sampled at twice the
1927, the Bell Telephone Company had Handbook, the bandwidth required for keying frequency. A block diagram of a
reported successful experiments with transmitting a radiotelegraph signal is ccw communications link is given in Fig.
200-wpm Baudot TTY communication in directly proportional to the keying rate. 47. Increased frequency stability and ac-
a 50-Hz bandwidth over undersea cables. For a speed of 12 wpm the unit pulse curacy can be achieved through phase-
The bandwidth· reduction resulted from length is 0.1 second. Since a dot and a locking both reference generators to a
synchronization of the 'transmitter and space each require 0.1 second, a string of standard frequency broadcast station. A
receiver. In proposing the application of dots at 12 wpm is a square wave having a good signal for this purpose is broadcast
synchronized telegraphy to radio, Terman fundamental frequency of 5 Hz. To on 60 kHz from WWVB. The 3.58-MHz
addressed the frequency stability problem preserve the square-wave characteristic of color burst signal generated by the teleyi-
attendant to any reduced-bandwidth the emission, an ssb transmission band- sion networks is another excellent
system by suggesting that transmitter and width of at least 15 Hz is required. A base- reference source.
receiver oscillators be locked to one of the . band (or dc wire telegraph) receiver needs Fig. 48 shows the elements of the audio.
standard frequencies broadcast by NBS. a similar bandwidth for conventional in- output filter in more detail. A combina-
He noted that given a reference frequen- formation recovery. Terman reported that tion of digital anq analog techniques pro-
cy, any other frequency can be derived by with synchronization techniques, the duces· a 3-dB bandwidth of 9 Hz, which is
means of multiplication, division or receiver bandwidth could be reduced to within the range predicted by Terman.
heterodyning .. Only the' word "syn- 1.5 to -2:0 times the keying rate. In con- When the receiver is properly tuned, the
thesizer" was missing! ventional (Morse) radiotelegraphy, the in- filter input signal frequency is I kHz.
Technology made giant leaps in the next telligence is ultimately received as an Since this frequency is "zero beat" with
45 years. In September 1975 QST, Ray- audio tone. Even a 15-Hz bandwidth filter the reference (LO) signal, the mixer out-
mond Petit, W7GHrvi, described the ex- centered on, say, 500 Hz, would require a put is Ii dc voltage proportional to the
periments of some radio amateurs with a Q of 33, causing intolerable ringing. cosine of the phase angle between the in-
mode he called "coherent cw." Petit did The ringing problem can be overcome put and reference signals. When the
not acknowledge Terman's paper, so we with time-domain processing at both ends signals are phase-cointident the mixer
must conclude that he -rediscovered the of the communications path. The produces a maximum positive voltage.
wheel. In any case, ccw is an idea whose transmitter is stabilized to within I Hz of The mixer output voltage swings negative
time has come. Adrian Weiss, W0RSP, the proper frequency by phase-locking to when the input and reference. signals are
disclosed some of the technical details of a reference standard. Precisely timed key- 180 0 out of phase. A 90 0 relationship
the ccw system in June and July 1977 CQ. ing pulses are derived from the same results in zero output voltage. The actual
The presentation contains some errors, reference standard. A similar reference hardware uses square waves for mixer LO
but the astute reader will be able to standard stabilizes the receiver frequency injection, so the phase detection transfer
recognize the significant principles. and synchronizes the audio output filter. characteristic is linear rather than
14-31 Chapter 14
n
r---- - - WWVB 60KHz - - - - - - - - - - ..,
I I
I
I I
I
FREQUENCY FREQUENCY
STANDARD
\11 STANDARD
FREQUENCY FREQUENCY
STABILIZER STABILIZER
'~
KEYING
SYNC
_r-
-, ~
KEYER
------'
TRANSMI'I'"TER ~ RECEIVER ~---l'" CCW FILTER
SIGNAL INPUT
1kHz
r--l00MS --1
..., i--1MS I
SA~
--l rIMS
tkHz
REFERENCE GENERATOR
RESET
SAMPLE
RE~
~ r- 1MS
I-o---:-l00MS-----i
10Hz
PHASE
~_ _--,,10:..:0.:.;H:::.Z_ _ _--t
<D®@®®®<v®®@
OECADE
COUNTER/DECODER
Fig. 48 - Block diagram of a ccw receiving filter. The operation is outlined in the text.
sinusoidal, but the mmtmum and maxi- domly with respect to the reference, even locked to the same standard frequency
mum voltages occur at the same points. though the reference generators at each transmission. This variation results from
.The phase of the input signal varies ran- end of the communications link may be changing propagation conditions. Phase
Specialized Communications Systems 14-35
variations of the input signal have little ef- RESPONSE, db
fect on the timing of the sampling window
because the sampling rate is only one one-
hundredth of the signal frequency. The
sampling window position is adjustable,
in any case. To prevent loss of output
voltage when the input/reference phase
relationship swings through 90 0, two
signal channels are driven in phase'
quadrature by the reference generator.
Thus, if the input signal is shifted 90 0
from the reference signal applied to one J _ L ___ -1 ____ 1 __ .l-_ t
mixer, that mixer output will be zero, but -60 ~50 -40 -30 -20 -\0 o 10 20 30 40 50 60
the other mixer will see a 00 or 180 0 rela- FREQUENCY FROM CENTER, Hz
tionship. The two channels are summed at
the filter output, so the output amplitude Fig. 49 - Amplitude vs. frequency response of the receiving filter.
is independent of the input phase, pro-
vided the frequency is zero-beat.
The voltage from the mixer is integrated
over a O.I-second period. Near the end of signal from the input mixer. If the input sized tranceivers can be outfitted for ccw
this interval, the timing logic causes the signal is 10 Hz away from the reference, more easily - replacing the internal
sample-and hold circuit to acquire the in- the mixer output will be a IO-Hz ac signal. reference oscillator with an external stan-
tegrated output voltage. One millisecond At the end of the O.I-second integrating dard is all that's required. To send ccw,
later, the logic resets the integrator and period, the mixer output waveform will the paddle-actuated clock in the keyer
the cycle repeats. The sample-and-hold have completed one cycle. Assuming the must be repli!ced by a continuous pulse
voltage controls the amplitude of the cycle started at zero volts, the sample-and- train from the frequency standard. Coor-
reference signal passed by the output hold will acquire the integrated mixer dinating one's paddle movements with the
modulator. Fig. 48 shows a separate voltage at the zero crossing and instruct "metronome" requires a different keying
phase-shift network at the output the output modulator to pass zero technique. A rbuffered keyboard (con-
modulator for clarity, but the input net- reference signal. (The above assumption . trolled by the standard) is the ideal ccw
work can serve both circuits. The square- isn't necessarily valid for a single channel, sending instrument.
wave outputs are in phase quadrature. but it holds for the resultant of the When more stations have ccw capabili-
The combined output waveform is a stair- quadrature channels.) Any whole number ty, the mode may prove highly useful for
case that can be filtered into a sine wave of beat-frequency cycles will cause the emergency communications. Another
with relative ease. sampled voltage to be zero. Since the possibility for ccw is in EME work. The
The timing signals are derived from the sampling interval is 0.1 second, the signal-to-noise ratio advantage should
I-kHz reference, which is synthesized response nulls occur every 10 Hz away reduce the station gain requirements.
from the master frequency standard used from the peak. If the input and reference Charles Woodson, W6NEY, played a
to stabilize the receiver LO. A decade signals differ by a multiple of 5 Hz, the stereophonic demonstration tape at
counter with a decimal decoder produces mixer voltage is sampled at the peak of a ARRL headquarters of a 20-meter ccw
10 outputs, each having the duration of half cycle, causing an alias. The aliases (Oontact with JRIZZR using a power of
one input pulse and a frequency of one- diminish 6 dB every time the beat frequen- 100 mW. While this type of contact isn't
tenth of the input clock. Output one goes cy doubles because the integrator is a first- too unusual in itself when conditions are
high during the first clock pulse, output order low-pass filter having a 6-dB-per- good and interference is absent, the chan-
two goes high during the second clock octave roll off. nel recorded without the ccw filter was
pulse, and so on. Every second pulse of a Noise bursts and strong adjacent- unreadable, while the ccw channel was
sequence of 10 commands the sample- channel signals result in an occasional ex- perfect copy. Woodson publishes a ccw
and-hold circuit to sample, and every tra dit or an elongated dah, but are other- newsletter whenever there is new informa-
fourth pulse resets the integrator. A sec- wise unnoticed. At the 12-wpm keying tion to report. In addition, Petit Logic
ond decade counter/decoder is cascaded speed used by ccw experimenters, a signal- Systems, P.O. Box 51, Oak Harbor, WA
with the timing control, and its phase- to-noise ratio improvement of about 20 98277 markets kits and accessories for ccw
adjustable output gates the timing signals dB can be realized over' the bandwidths operation.
to select a O.I-second integrating window ty'pically used for cw. Faster speeds are
that is synchronized with the incoming possible, but the bandwidth must be in-
keying pulses. Because the signal is creased at the expense of signal-to-nQise
sampled at the end of the integration in- ratio.
terval, the filter output is delayed 0.1 sec- To establish ccw contact, one station
ond with respect to the input. sends a prealJlble Of dits to allow the
Fig. 49 shows the approximate receiving operator to synchronize his
amplitude-versus-freq~ency response of filter. Experience thus far indicates that
the filter. Note the symmetry of the skirts. once the filter has been synchronized, it
Unlike analog filters composed of linear usually won't need adjustment for several
circuit elements, the sampling filter does. hours. Fig. 50 depicts a typical ccw sta-
not exhibit arithmetic -selectivity. The tion. The early experimenters built their
spurious responses on either side of the stations around simple QRP equipment to
main passband resemble the infinite rejec- dramatize the communications advan- Fig. 50 - This is the first complet~ amateur
tion notches characteristic of an elliptic tages offered by the mode and to em- station to be built for coherent cw operation.
filter and are called aliases. The frequency phasize the accessibility of the necessary Assembled by Andy McCaskey, WA7ZVC, it
consists of a modified Ten·Tec PM-2
response is quasi-periodic because signals technology. The simple gear requires transceiver and homemade modules which pro-
that are not zero-beat with the reference some add-on circuitry to allow' oscillator vide for the control and processing of signals
frequency produce a difference frequency stabilization. The more modern synthe- as required for coherent-cw operation.
14-36 Chapter 14
Chapter"15
I.nterference
with Other Services
Radio Frequency Interference (RFI) It is clear, therefore, that almost all RFI , matic diagrams showing the recommended
has probably been with us since the first problems experienced with home-enter- placement of bypass capacitors and other
amateur stations came on the air some 70 tainment devices result from basic design components to reduce rf susceptibility.
years ago'. Fed by the technology that deficiencies in this equipment. The few One large American manufacturer of hi-fi
developed during and following WW II, small components or filters which would equipment has in some cases supplied the
the problem has become iln increasing prevent RFI are often left out of otherwise necessary components free of charge,
source of irritation between radio opera- well-designed products as manufacturers although no consistent policy has been
tors and their neighbors. Home-entertain- attempt to reduce costs, and hence, to evident and the consumer must still pay to
ment electronics devices now abound, reduce the prices of their products. have a serviceman install the components.
with most families owning at least one While these are encouraging develop-
television receiver, an a-m or fm radio, The Solution - Consumer Protection ments, it appears likely that meaningful
and anyone of several audio devices (such Given the present unacceptable situa- and widespread corrective action by
as a phonograph, an intercom, an tion, what can we as" amateurs do to help equipment designers will require both
electronic guitar, or an electronic organ). the consumer resolve the RFI problem? pressure from consumers and establish-
Given the innate perversity of these One step which should certainly be taken ment of suitable government standards.
objects to intercept radio signals, it should is to advise our friends and neighbors to Voluntary after-the-fact measijres on
surprise no one to learn that RFI is one of inquire, before they make a purchase of the part of manufacturers simply are not
the most difficult problems amateurs face an electronic device, whether the product enough. It is a foregone conclusion that as
in their day-to-day operations. has been certified for operation in the pre- long as the inclusion of additional
sence of a radio transmitter. Manufac- components for susceptibility reduction
How Serious is the RFI Problem? turers must be made to recognize that RFI increases a manufacturer's cost, however
In one year alone, the FCC received protection oT their home-entertainment slightly, there will be reluctance to take
150,000 RFI complaints, up more than 200 equipment has become essential. Further, steps to improve equipment designs by the
percent from the number of complaints where interference is being experienced, manufacturers themselves. What appears
received in 1970. Of these, the great the consumer should be encouraged to to be nec~ssary, therefore, is federal
majority involved interference to home- contact the manufacturer of his equipment legislation giving the FCC the authority to
entertainment equipment. Most important, and to request that the manufacturer fur- regulate the manufacture of home-enter-
•
nearly all of these would never have come nish the components or services necessary tainment devices and thus protect the con-
to the Commission's attention if the manu- to eliminate RFI. sumer.
facturers had corrected design deficiencies
in their home-entertainment products at What Are Manufacturers Doing Today? It's Up to Us
the time of manufacture. It is of interest to Many responsible manufacturers have a If requests to manufacturers of home-
note that more than 60 percent of the policy of supplying filters for eliminating entertainment equipment for those com-
interference cases reported were related to television interference when such cases are ponents and installation services neces-
television interference (TVI). ·brought to their attention. A list of those sary to relieve RFI problems are to be suc-
In the case of television interference, manufacturers, and a more thorough cessful, each of us, when faced with an
FCC experience shows that 90 percent of treatment of the RFI problem, can be RFI problem, must make known out
the problems experienced can only be obtained by writing the ARRL. If a given position to. the manufacturers involved.
cured at the television receiver. Further, manufacturer is not listed, it is still While a respectful request for assistance
when it comes to audio equipment, the possible that he can be persuaded to will bring more cooperation than a blunt
only cure for RFI is to treat the audio supply a filter; this can be determined by demand, do not hesitate to let the
device experiencing the interference. There writing either directly to him or to the manufacturers know that they have a
is nothing an.amateur can do to his trans- Electronic Industries Association (EIA).l responsibility to the consumer for cor-
mitter which will stop a neighbor's phono- With respect to audio devices, some recting the . design deficiencies that are
graph from acting as a short-wave re· manufacturers will supply modified sche- causing the problem. Before casting the
eeiver. It should be emphasized that first stone, however, make sure you're not
phonographs and hi-fi units are not sitting in a glass house. Certainly, if your
designed to be receivers, but simply audio 'Electronic Industries Association, 2001 Eye Street, own television receiver experiences no
N.W., Washington, DC 20006. Attention: Director
devices. of Consumer Affairs. interference while you are on the airy it is
sistanc~ wherever older equipme~t designs increase listener cooperation, not destroy
are in use. it. .
88
first step obviously is to make sure that transmitting on' frequencies below 30
, the transmitter has no radiations outside MHz, the TV band of principal interest is
the bands assigned for amateur use. The the low vhf band between 54 and 88 MHz.
best check on this is your own a-m or TV If harmonic radiation can h,e reduced to
receiver. It is always convincing if you can the point where no interference is caused
demonstrate that you do not interfere to channels 2 to 6, inclusive, it is almost
with reception in your own home. certain that any harmonic troubles with
channels above 174 MHz will disappear
Do!"t Hide Your Identity also.
Whenever you make equipment changes The relationship between the vhf tele-
- or shift to a hitherto unused band or vision channels and harmonics of amateur
type of emission - that might be expected bands from 14 through 28 MHz is shown
to change the interference situation, check in Fig. 1. Harmonics of the 7~ and 3.5-MHz
with your neighbors. If no one is bands are not shown because they fall in
,experiencing interference, so much the every television channel. However, the
better; it does no harm to keep the harrIlOnics above 54 MHz from these
, neighborhood aware of the fact that you bands are of such high order that they are
are operating without bothering anyone. usually rather low in amplitude, although
Should you change location, make your they may be strong enough to interfere if
presence known and conduct occasional the television receiver is quite close to the
Fig, 1 - Relationship of amateur-band harmonics to
tests on the air, requesting anyone whose amateur transmitter. Low-order harmonics vhf TV channels. Harmonic interference from
reception is being spoiled to let you know - up to about the sixth - are usually the transmitters operating below 30 MHz is likely to be
about it so steps may be taken to eliminate most difficult to eliminate. serious in the low-channel group (54 to 88 MHz),
the trouble. Of the amateur vhf bands, only 50 MHz
will have harmonics falling in a vhf
Act Promptly television channel (channels 11, 12 and "fringe" areas where the received picture
The average person will tolerate a 13). However, a transmitter for any is visibly degraded by the appearance of
limited amount of interference, but the .amateur vhf band may cause interference set noise or "snow" on the -screen, it may
'sooner you take steps to eliminate it, the if it has multiplier stages either operating be necessary to go to extreme measures.
more agreeable the listener will be; the in or having harmonics in one or more of In either case the intensity of the
longer he has to wait for you, the less the vhf TV channels. The rf energy on interference depends very greatly on the
willing he will be to cooperate. such frequencies can be radiated directly exact frequency of the interfering signal.
from the transmitting circuits or coupled Fig. 2 shows the placement of the picture
Present Your Story Tactjully by stray means to the transmitting and sound carriers in the standard TV
Whenever a device intercepts your antenna. channel. In channel 2, for example, the
signals, it is natural for the complainant to picture carrier frequency is 54 + 1.25 =
assume that your transmitter is at' fault. If Frequmey Efj«tl 55.25 MHz and the sound carrier fre-
you are certain that the trouble is not in The degree to which transmitter har- quency is 60 - 0.25 = 59.75 MHz. The
your transmitter, explain to the listener monics or other undesired' radiation second harmonic of 28,010 kHz (56,020
that the reason lies in the receiver design, actually in the TV channel must be kHz or 56.02 MHz) falls 56.02 - 54 = 2.02
and that some modifications may have to suppressed depends principally on two MHz above the low edge' of the channel
be made in the receiver if he is to expect factors: The strength of the TV signal on andis in the region marked "severe" in
interference-free reception. the channel or channels affected, and the Fig. 2. On the other hand, the second
relationship between the frequency of the harmonic of 29,500 kHz (59,000 kHz or
Arrange jor Tests spurious radiation and the frequencies of 59 MHz) is 59 - 54 = 5 MHz from the low
Most listeners are not very competent the TV picture and sound carriers within edge of the channel and falls in.the region
observers of the various aspects of the channel. If the TV signal is very marked "mild." Interference at this
iriterference. If at all possible, enlist the strong, interference can be eliminated by frequency has to be about 100 times as
help of another amateur and have him comparatively simple methods. However, strong as the 5.6,020 kHz to cause effects
operate your transmifter while you see for if the TV signal is very weak, as in of equal intensity. Thus an operating
15·2 Chapter 15
o 2 3 5
MEGAHERTZ FROM LOW EDGE OF TV CHANNEL
Fig. 2 - Location of picture and sound carriers in a monochro/Tle television channel, showing the
relative intensity of interference as the location of the interfering signal within the channel is varied
without changing its strength. The ttiree regions are not actually sharply defined as shown in this
drawing, but merge into one another gradually. Fig. 3- "CrOSS-hatching," caused by the beat
between the picture carrier and an interfering
Signal inside the TV channel.
frequency that puts a harmonic near the look about as shown in Fig. 4. They result
picture carrier requires about 40 dB more from the variations in the intensity of the
harmonic suppression in order to avoid interfering signal when modulated. Under
interference,as compared with an opera- most circumstances modulation bars will
ting frequency that puts the harmonic not occur if the amateur transmitter is fre-
near the upper edge of the channel. quency- or phase-modulated. With these
For a region of 100 kHz or so either types of modulation the cross-hatching
side of the sound carrier there is another will "wiggle" from side to side with the
"severe" region where a spurious radia- modulation.
tion will interfere with reception of the Except. in the more severe cases, there is
sound program and this region also seldom any. effect on the sound reception
should be avoided. In general, a signal of when interference shows in the picture, r
intensity equal to that of the picture unless the frequency is quite close to the
carrier will not cause noticeable inter- sound carrier. In th& latter everit the
ference if its frequency is in the "mild" sound may be interfered with even though
region shown in Fig. 2, but the same the picture is clean.
intensity in the "severe" region will utterly Reference to Fig. 1 will show whether Fig. 4 ~ "Sound bars" or "modulation bars"
destroy the picture. or not harmonics of the frequency in use accompanying amplitude modulation of an
will fall in any television channels that can interfering signal. In this case the interfering carrier
Interference Pattems be received in the locality. It should be is strong enough to destroy the picture, but in mild
cases the picture is viSible through the horizontal
The visible effects of interference vary kept in mind that not onlY,harmonics of bars. Sound bars may accompany modulation even
with the type and intensity of the the final frequency may interfere, but also though the unmodulated carrier givelj no visible
interference. il(:omplete "blackout," where harmonics of any frequencies that may be cross-hatching.
the picture and sound disappear com- present .in buffer or frequency-multiplier
pletely, leaving the screen dark, occurs· stages. In the case of 144-MHz transmit-
only when the transmitter and receiver are ters, frequency-multiplying combinations radiation from the antenna, so measures
quite close together. Strong interference that require a doubler or tripler stage to taken to prevent harmonics from reaching
ordinarily causes the picture to be broken operate on a frequency actually in a the antenna will not reduce TVI if the
up, leaving a jumble of light and dark low-band vhf channel in use in the locality transmitter itself is radiating harmonics.
lines, or turns the picture "negative" - should be avoided. But once it ·has been found that· the
the normally white parts of the picture transmitter itself is free from harmonic
turn black and the normally black parts Harmonic Suppression radiation, devices for preventing har-
turn white. "Cross-hatching" - diagonal Effective harmonic suppression has monics from reaching the antenna can be
bars or lines in the picture - accompanies three separate phases: expected to p'roduce results.
the latter, usually, and also represents the 1) Reducing the amplitude of har-
most common type 'of less severe inter- monics generated in tht; transmitter. This Reducing Harmonic Generation
ference. The bars are the result of the beat is a matter of circuit design and operating Since reasonably efficient operation of
between the harmonic frequency and the conditions. rf power amplifiers always is accompanied
picture carrier frequency. They are broad 2) Preventing stray radiation from the by harmonic generation, good judgment.
and relatively few in number if the beat transmitter and associated wiring. This calls for operating all frequency-multiplier
frequency is comparatively low - near requires adequate shielding and filtering stages at a very low power level. When the
the picture carrier - and are numerous of all circuits and leads from which final output frequency is reached, it is
and very fine if the beat frequency is very radiation can take place. desirable to use as few stages as possible in
high - toward the upper end of the 3) Preventing harmonics from being fed building up to the final output power level
channel. Typical cross-hatching is shown into the antenna. and to use tubes that require a minimum
in Fig. 3. If the frequency falls in the mild It is impossible to build a transmitter of driving power.
region in Fig. 2, the cross-hatching may be that will not generate some harmonics, but
so fine as to be visible only on close it is obviously advantageous to reduce Circuit Design and Layout
inspection of the picture, in which case it their strength, through circuit design and Harmonic currents of considerable
may simply cause the apparent brightness choice of operating conditions, by as large amplitude flow in both the grid and plate
of the screen to' change when the a factor as possible before attempting to circuits of rf power amplifiers, but they
transmitter carrier is thrown on and off. prevent them from being radiated. Har- will do relatively little harm if they can be
Whether or not cross-hatching is. visible, monic radiation from the transmitter itself effectively bypassed to the tube cathode.
an amplitude-modulated transmitter may or from its' associated wiring obviously Fig. 5 shows the paths followed by
cause "sound bars" in the picture. These will cause interference jus~ as readily as harmonic currents in an amplifier circuit;
because of the high reactance of the tank erately introduced; t~ harmoni~ at that conditions, there is little or no difference
coil there is little 'harmonic current in it, so frequency will be tremendously increased between single-ended and push-pull am-
the haqnonic currents simply flow throu&h in amplitude. I plifiers in respect to harmonic generation.
the tank capacitor; the plate (or grid) Such resonances are unavoidable, but Push-pull amplifiers are frequently trouble-
blocking capacitor, and the tube capaci- by keeping the path from plate to cathode makers on even-order harmonics because
tances. The lengths of the leads forming and from grid to cathode as short as is with such amplifiers the even-harmonic
these paths is of great importance, since physically possible, the resonant fre- voltages are in phase at the ends of the
the inductance in this circuit will resonate quency usually can be raised above 100 tank circuit and hence appear with equal
with the tube capacitance at some MHz in amplifiers of medium power. This amplitUde across the whole tank coil if the
frequency in the vhf range (the tank apd puts it between the two groups of center of the coil is not grounded. Under
blocking capacitances usually are so large television channels. such circumstances the even harmonics
compared with the tube capacitance that It is easier to place grid-circuit vhf can be coupled to the output circuit
they have little effect on the resonant resonances· where they will do no harm through stray capacitance between the
frequency). If such a resonance happens when the amplifier is link-coupled to the tank and coupling coils. This does not
to occur at or near the same frequency as driver stage, since this generally permits occur in a single-ended amplifier having
one of the transmitter harmonics, the shorter leads and more favorable con- an inductively coupled tank if the
effect is just the same as though a ditions for bypassing the harmonics than coupling coil is placed at the cold end or
harmonic tank circuit had been delib- is the case with capacitive coupling. Link with a pi-network tank.
coupling also reduces the. coupling be-·
tween the driver and amplifier at har- Some TVI Tests
monic frequencies, thus preventing driver One of the difficulties in solving TVI
harmonics from being amplified. problems, particularly in fringe areas, is
The inductance of leads from the tube the number of possible causes and their
to the tank capacitor can be reduced not elusive nature. A "cure" seems to be
only by shortening but by using flat strip found, only to have the problem return
C2 instead of wire conductors. It is also better with renewed severity the next day.
to use the chassis as the return from the Consequently, some tests were performed
Cl
blocking capacitor or tuned· circuit to by the ARRL in order to isolate thO!
C3 C4 cathode, since a chassis path will have less causes, if possible. Although the results
inductance than almost any other form of weren't encouraging in regard to certain
connection. aspects, one bright spot was some "fall-
The vhf resonance points in amplifier out" in the way of additional suppression
tank circuits can be found by coupling a measures previously neglected.
Fig. 5-Avhf resonant circuit is formed by the tube
capacitance and the lead inductances through the grid-dip meter covering the 50-250 MHz
tank and blocking capacitors. Regular tank coils are range to the grid and plate leads. If a Test Procedures
not shown, since they have little effect on such resonance is found in or near a TV A ham ex;periencing some TVI in a
. resonances. C1 is the grid tuning capacitor and C2 is channel, methods such as those described fringe area (on his own set) generously
the plate tuning capacitor. C3 and C4 are the grid
and plate blocking or bypass capacitors,
above should be used to move it well out agreed to be the "patient" in the tests. A
respectively. " of the TV range. The grid-dip meter also large screened enclosure was transported
should be used to check for vhf reso- to his location. It could contain the TV set
nances in the tank coils, because coils along with a smaller version of the screen
made for 14 MHz and below usually will enclosure (Fig. 6) for the transmitter.
show such resonances. In making the Other equipment included a gasoline-
check, disconnect the coil entirely from powered generator that cpuld power
000 0 the transmitter and move the grid-dip either the TV set or transmitter. There is .
000 0
meter coil along it while exploring for a always the possibility that feedback of rf
0000
dip in the 54- to 88-MHz band. If a energy through the power line (or
resonance falls in a TV channel that is in "conducted interference") is a factor in a
LINE UHF
FILTER BULKHEAD use in the locality, changing the number of TVI problem. By running the equipment
FEEDTHRU
o 0 0 turns will move it to a less-troublesome on separate power systems, some idea of
KEY MIC
frequency. the importance of this type of coupling
would be ascertained.
117 V AC Operating Conditions Various low-pass filters, high-pass fil-
Grid bias and grid current have an ters, and power harnesses made up the
important effect on the harmonic content rest of the equipment list. Checks on
Fig. 6 - Construction details of the rf enclosure.
of the rf currents in both the grid and various TV channels indicated the most
Forthe model shown, thin aluminum sheet metal serio,us problem resulted from third-
was used to form a box 12 X 28 X 20 inches plate circuits. In general, harmonic output
increases as the grid bias and grid current harmonic energy on channel 3 during
(HWD) (304 X 711 X 508 mm). Small holes were
drilled for ventilation and a fan might be 15-meter operation. Tests were performed
are increased, but this is not necessarily
advisable if temperature rise is considered a
true of a particular harmonic. The third with the rig inside and out of its shielded
problem. Fe,ed-through connectors can be of the enclosure, the TV set inside and out of the
builder's choice but ac co'nduits and control and higher harmonics, especially, will go
larger enclosure, and with either the TV
leads should be filtered. for key and mic leads, through fluctuations in amplitude as the
bypass with 0.001 "uF disk-ceram ic capacitors grid current is increased, and sometimes a set or transmitter on independent power.
and install a small ferrite bead (if available). A rather high value of grid current will
commercially manufactured line filter was used, T~st Results
although a homemade one of the "brute force" minimize one harmonic as compared with
type would also be suitable. a low value. This characteristic can be Previous checks in the lab revealed that
Although dimensions and material are not used to advantage where a particular almost all currently manufactured ama-
critical, the cabinet should be deep enough fo harmonic is causing interference, remem- teur transmitters and transceivers emit
form as much of an overlap as possible between
the front of the equipment panel and the front of bering that the operating conditions that harmonics in the. form of "chassis
the cabinet. It is important that all leads be run minimize one harmonic may greatly radiation" to varying degrees. It should be
through the r~ar of the cabinet. increase another. For equ'al operating pointed out that no outstanding "saints"
15·4 Chapter 15
were found in this area but mostly only test that indiCated appreciflble reduc- measures will be equally ineffective under
"sinners." Once this energy escapeS from the tion ofTVl was the one with the transmitter all circumstances. However, the claims of
transmitter cabinet, it can be conducted to placed inside of its shielded enclosure. In all some manufacturers are open to question.
the antenna or back through the power other tests, there was no improvement or Items such as power supplies that
line via a single-conductor waveguide type the change was so slight as to be inconclu- eliminate TVI, and similar nostrums have
of mode. This mode is very similar to the sive. Some residual interference still re- come to our attention. Consequently,
propagation of rf energy over a two- mained with the rig enclosed in the shield. common sense is in order in judging
dimensional conducting surface in the. This was likely caused by rectification in whether or not a particular device will
form of a ground wave. But the important the external environment. One unexpected prove effective in eliminating interference
thing to keep in mind is that devices such result was that no noticeable difference or how it should be employed.
as filters, traps and grounds are ineffective was observed with the door of the enclo-
since the rf energy flows around the sure opened or closed. In fact, with the Suppression Practices
suppression network. The only effective door partially closed and touching the As the test results reported in the
measure is adequate shielding. shield at only a few points the TVI became previous section reveal, complete elimi-
As might be expected, the field tests worse! nation of TVI is often not a simple pro-
verified the importance of this aspect. The cess. It seldom happens' that a single mea-
Conclusions sure such as installing a high-pass filter at
The transmitter power for these tests the TV set will cure the problem. Rather,
was approximately 180 watts. Cohsider- a number of methods must be applied
able work has yet to be performed to simultaneously. The principal factor in
determine the important factors at higher any TVI situation is the ratio of TV signal
levels. However, there is hope for the ham strength to interference level. Th~s in-
experiencing TVI because of chassis cludes interference of all types such as
radiation. With the door open, it is , ignition noise, random or thermal noise ,
believed that the enclosure acts as a (which isn't really interference but sets the
waveguide below cutoff and still offers minimum signal that permits "snow-free"
some measure of suppression while per- reception), and unwanted signals·that fall
mitting access to the controls. In discuss- within the TV channel. A signal-to-
ing the tests with other amateurs experi- interference ratio greater than approxi-
encing TVI, reports from the field were mately 35 to 40 dB is required for good
Fig. 7 - Lab simulation of TVI tests discussed in favorable when similar measures were picture quality. .
the text. A dummy load placed next to the "rabbit
ears" served as the transmitting antenna. tried. In this regard, an area frequently
Some further experimentation along overlooked in TVI difficulties is the
these lines is in order. For instance, TV -set antenna. A poor antenna with
former shielding theory advocated the use little gain in the direction of the TV
of high-conductivity materials. Newer station, old and corroded wire and
methods often rely on the dissipation of connections (which can cause the har-
unwantecl rf energy in lower conductivity monic generation by rectification' of a
materials such as steel. Although rf energy "clean" signal generated in a nearby
can penetrate deeper into low-conductivity amateur transmitter), may result in a TVI
metals, and greater thicknesses are re- situation that is impossible to solve. For
quired to provide the same isolation (as instance, the "simulated" lab tests il-
that of copper, for example), other prob- lustrated in the photographs were per-
lems are simplified. Unwanted rf energy formed with a dummy load next to a set of
must be dissipated somewhere and when a "rabbit ears," which comprised the
good conductor is used for a shielding TV -set antenna. With a good outdoor
enclosure there is a greater tendency for system, the TVI would not have been
this energy to leak out through doors, present even though there was a leakage
Fig. 8 - Severe interference occurred with the conduits and other points of entry. By from the transmitter cabinet. Generally
setup shown above with the transmitter out of the
enclosure. Interference was about the same with dissipating energy internally on the shield speaking, if the picture quality on the JV
the leads running from the front of the cabinet walls, there is less chance for it to leak out. set experiencing the interference is poor to
instead of through the rear connectors. (However, it the unexpected attenuation begin with, even sophisticated suppression
with the door open was actually caused by measures are likely to prove futile.
a waveguide-below-cutoff effect as specu-
lated, high-conductivity material near the Grounds
door opening would be advisable.)
Grounding of equipment has long been
considered to be a first step in eliminating
Other Results interference. While the method is very
With the TV set in its shielded enclosure effective in the mf range and below, for all
and with power fed through a commercially practical purposes it is useless in sup-
manufactured line fIlter, there was no differ- pressing vhf energy. This is because eve~
ence in TVI with the rig or TV set on inde- short lengths of wire have considerable
pendent power. Rf signal energy from the reactance at vhf. For instance, suppose a
TV antenna was fed into the shield enclo- length of wire by itself has an inductance
sure through a commercially manufactured of 1 uH. At 550 kHz, the reactance would
'high-pass filter. Little change was noted be about 3.46 ohms. On the other hand,
when these measures were eliminated and the same wire would have a reactance of
Fig. 9 - The interference was either reduced the set operated on the same power service over 300 ohms at 56 MHz, which is the
considerably (as shown in this photo) or
eliminated completely, depending upon TV
as the transmitter, without a high-pass fil- frequency range of TV channel 2. (Ac-
signal strength, with ali the leads exiting from the ter and outside of the shield. tually, the impedance of a wire becomes a
rear of the cabinet. It should not be concluded that such more complicated entity to define at vhf.
I' 70
1---- ' l - ,-+- t-
t-
..-r-... "r-
I 60
I
I ..
I _.
I
I CHASSIS .. .
,.,.
III 50
."
.. .
en'" .. Z
f-
::::> ~
0-
f-
::::> 40
Fig. 14 - Additional lead filtering for har· 0
monies or other spurious frequencies in the w
::::>
high vhf TV band (174·216 MHz). 0 SPURIOUS ATTENUATION <30MHz
1< 30
C1 - 0.001·"F disc ceramic ::::>
0-
x ; POWER (WATTS)
C2 - 500· or 1000'pF feed·through bypass w
(Centralab FT·1000. Above 500 volts, u. POWER LEVEL ATTENUATION
0
substitute Centralab 8585·500.) z 20 X< 5 WATTS 30dB
RFC - 14 inches,135.S·mm) no. 26 enamel Q
f- 5 WATTS ~X~ 500 WATTS 40dB
close·wound on 3/16·inch (4.76·mm) dia. form «
::::> X> 500WATTS t3+10 LOG,a(X)dB
or composition resistor body. z
w to
f-
f- I-
« --.- .
-:-:T
+--
0
3 4 6789 3 4 6789 2 3 4!S 6789 1 2 3 456789
capacitors can be paralleled to reduce the 2 5 10 20 50 tOO tOOO 10,000
effect of lead inductance. OUTPUT POWER (WATTS)
A very desirable capacitor (C2) from an
RFI point of view is shown in Fig. 14.
Instead of having two or more plates
Fig. 15 - The FCC specifies that spurious signals generated by transmitting equipment must be
arranged '. in a parallel fashion, the reduced well below the level of the fundamental. This graph illustrates exactly how far the
conductors are coaxial and are separated spurious components must be reduced. This applies to amateur transmitters operating below 30
by the dielectric. Such feedthro'ugh capa- MHz.
50
- r--
40
-I--
SPURIOUS ATTENUATION 30-235 MHz
30
.-_. X = PQWE R (WATTS)
z
0
>= I· . POWER LEVE,L (OUTPUT) ATTENUATION
"z
::> 20
x< 0.\ WATT 50+10 LOG,o (X)dB
'"
l-
I-
-"-
0.\ WATT <: X 5. 0.25 WATT
0.25 WATT <. X"::: 25 WATTS'
40dB
46+10 LOG10{X)dB
<!
10
X >25 WATTS 60dB
r------- - - --
o i::?-" 1--
2 2 3 4 56789 2 :5" 5 6 789 2 3 " 5 6 789 4567891 23456789] 23456789
0.00001 0.000\ O,QOl 0.0\ 0.\ 10 100 1000
OUTPUT POWER (WATTS)
Fig. 16 - Th.is graph illustrates to what level spurious-output energy must be reduced for equipment designed to operate in the 30- to 235-MHz range.
interference to the reception of another ragio model was built and the response is shown
station, the licensee may be required to take in- Fig. 13. As can be seen, the filter came O.2646pH 0 .402996pH O.t8986~H
steps to' eliminate' the interference in ac- quite close to the design goals. Unfortu-
cordance with good engineering practice.
nately, as with most of the designs in this
NOTE: For the purposes of this section, a
spurious emission or radiation means any emis- section, alignment of the more compli-
. sion or radiation from a transmitter, cated filters requires some sort of sweep- O.06229tpH O.20096pH
transceiver, or external radio frequency power generator setup. This is the only practical
amplifier which is outside of the authorized way of "tweaking" a· filter to the desired 12-59.7 MHz ft- 4t.2 MHz
Amateur Radio Service frequency band being response. While building a sweep setup is ~tt4.tPF ~74.tt8PF
used. not beyond the talents of an advanced ex-
The numerical limits cited in 97.73 are perimenter, the lack of one is an obstacie
interpreted graphically in Figs_ 15 and 16. in the home construction of filters.
Note, however, that paragraphs (c) and Fig. 17 - Schematic diagram showing
An Absorptive Filter
(d) go beyond absolute limits in defining component values of an experimental elliptic
the amateur's obligation. The filter shown in Fig. 19 not only pro- function filter.
vides rejection by means of a low-pass
Filters and Interference section, it also includes circuitry that
The judicious use of filters, along with absorbs harmonic energy. A high-pass
other suppression measures such as section consisting of Ll,L2, Cl and C2 is
shielding, has provided solutions to terminated in a 50-ohm "idler load" and'
interference problems in widely varying this combination performs the latter
applications. As a consequence, con- function. The advantages of this tech-
siderable attention has been given to the pique are that degradation of filter
subject over the 'years that has resulted in rejection caused by antenna mismatch at'
some very esoteric designs. Perhaps the the harmonic frequency are not as severe
most modern approach is the optimiza- (with a filter of this type) and the
tion and/or realization for a particular transmitter is terminated in a resistive
application of a filter by means of a digital load at the harmonic.
computer. However, there are a number /
of other types with component values Construction and Test Techniques
Fig. 18 - Response curve of the filter shown
cataloged in tabular form. Of these, the If good performance above 100 MHz is in Fig. 17. Vertical scale represents 10 dB/div.
most important ones are the so-called not a necessity, this filter can be built and horizontal scale is 10 MHz/div.
Chebyshev and elliptic-function filters. using conventional fixed capacitors. Cop-
(Butterworth filters are often considered per-clad Teflon board may not be readily
a special case of Chebyshev types with a available in small quantities from many accurate grid-dip osciUator, an SWR
ripple factor of zero.) supply houses. Regular fiberglass-insulated bridge, a reactance chart or the ARRL
Elliptic-function filters might be con- board is satisfactory for low power. One Lightning Calculator (for L, C and f), a
sidered optimum in the sense that they such filter has been used with an SB-100 50-ohm dummy load, and a transmitter_
provide the sharpest rolloff between the transceiver running 100 watts. Although Once the value of a given capacitor has
passband and stopband. Computed values the Q of the fiberglass capacitors will be . been calculated,' the next. step is to
for a low-pass filter with a O.I-dB ripple in lower than -that of Teflon-dielectric capa- determine the capacitance per square inch
the passband and a cutoff frequency of citors, this should not greatly affect the of the double-clad circuit board you have.
30.6 MHz are shown in Fig. 12. The filter type of filter described here. This is done by connecting ORe end of a
is supposed to provide an attenuation of Test equipment needed to build this coil of known inductance to one side of
35 dB above 40 MHz. An experimental filter at home includes a reasonably the circuit board, and the other coil lead
15·8 Chapter 15
be maintained' for arc-over protect~on.
ct C2
Capacitors with Teflon dielectric have Table 1
been used in filters passing up to 2 kW
JJl
TV Receiving Antenna Gain, dB, vs Angle
501\ PEP. One further word of caution: No to Transmitter
.
Ll L2 IDLER
LOAD low-pass filter will be fully effective until
the transmitter with which it is used is 315-45°45·135° 135·225° 225-315°
properly shielded and all leads filtered. Fringe·Area + 10 <-10 <0 <-10
The terminating loads for the high-pass Antenna
Lower·Grade + 3 <0 <0 <0
50fl section of the filter can be made from Antenna
2-watt, lO-percent tolerance composition
L5
I re$istors. Almost any dissipation rating
can be obtained by suitable series:paralle1
combinations. For' example, a 16-watt,
Table 2
50-ohm load could be built as shown in
Fig. 19. This load should handle the Path Loss, dB, Transmitting Antenna
harmonic energy of a signal with peak to Receiving Antenna
Fig. 19 - Schematic diagram qf the absorptive fundamental power of 2 kilowatts. With Distance in Meters 10 20 30 40 50 100
filter. The pc·board used is MIL·P·13949D,
FL·GT·.062 in, C·2/2·11017, Class 1, Grade A.
this load, the harmonic energy will see an 44·MHz I·F 25 3.1 34 37 39 45
Polychem Bud Division. Capacitance between SWR under 2: 1 up to 400 MHz. For low TV -2 27 33 37 39 41 47
copper surfaces is 10·pF per square inch. power (under 300 watts PEP), a pair of TV 3 28 34 38 40 42 48
TV 4 29 35 39 41 43 49
Values are as follows for a dei?ign cutoff fre· 2-watt, lOO-ohm resistors is adequate. TV 5 36 40 42 44 50
'quency of 40 MHz and rejection peak in chan· 30
In the model shown the high-pass filter TV 6 31 37 41 43 45 51
nel2:
C1 - 52 pF C4 - 21.6 pF L3 - 0.3}'H
series capacitor~ are bonded and mounted
C2 - 73 pF L1 -0.125"H L4 - 0.212 f'H on Teflon standoff insulators.
C3 - 126 pF L2 - 0.52"H L5 - 0.24"H
How Much Harmonic Suppression Table 3
is Needed? TV Received·Signal Power
to the other side of the circuit board. Use TV Channel Class 8 Class A
the grid-dip oscillator, coupled lightly to While it's a fairly simple matter to in-
the coil, to determine the resonant stall a low-pass filter and hope for good 2 -55 dBm -34 dBm
3 -56 dBm -35 dBm
frequency of the coil and the circuit-board results, approaching the problem scien- 4 - 57 dBm -35 dBm
capacitor. When the frequency is known, tifically can reduce unnecessary expense 5 -58dBm -37 dBm
the total capacitance can be determined and aggravation. Rasmussen and Oerue 6 -59 dBm -38 dBm
by working the Lightning Calculator or presented an orderly method for deter-'
by looking the capacitance up on a mining the required harmonic attenuation
reactance chart. The total capacitance in "Harmonic TVI - A New Look at an
divided by the number of square inches on Old Problem," QST, September 1975. Add 3 dB for each doubling of the
one side of the circuit board gives the The following factors affect the power (1000 watts is 60 dBm).
capacitance per square inch. Once this tolerable harmonic levels: T A - Transmitter harmonic attenuation,
figure is determined, capacitors of almost T p - Transmitter power output in dBm. can be estimated from specifications
any value can be laid out with a ruler! TA - Transmitter harmonic attenuation. given for commercial transmitters.
High voltages can be developed across OT - Gain of the transmitting antenna OT - Conservative estimate () dB. .
capacitors in a series-tuned circuit, so the at the harmonic frequency, GR - Estimate relative bearing from the
copper material should be trimmed back G R - Gain of the.TV receiving antenna at TV antenna to the amateur antenna,
at least 1/8 inch (3-mm) from all edges of the interfering frequency, and use Table I.
a board, except those that will be soldered SA - Attenuation (path loss) from the SA - Estimate distance between the an-
to ground, to prevent arcing. This should transmitting antenna' to the receiving tennas, and use Table 2 for path loss
not be accomplished by filing, since the antenna, for the TV channel concerned.
copper filings would become imbedded in Ss - TV signal strength at the TV re- Ss - TV signal level from Table 3 for
the board material and just compound the ceiver input, TV channel and distance to TV station.
problem. The capacitor surfaces should be SR - Signal-to-interference ratio at the For under 50 miles, use Class A; for
kept smooth and sharp corners should be TV receiver needed to preclude visual or over 50 miles, use Class B.
avoided. . audio interference. SR - Signal/interference ratio required
If the filter box is made of double-clad The relationship between these factors, can be as high as 40 dB, with an
fiberglass board, 'both sides should be when expressed in dB, is operating frequency whose harmonic is
. bonded together with copper stripped
from another piece of board. Stripped
HR =
iT p + T A + GT + G R + SA) within I MHz of the picture carrier
frequency, or as low as 20 dB, if the
copper foil may be cleaned with a razor - Ss + SR
harmonic is more than 2 MHz from
blade before soldering. To remove copper where HR is the further transmitter har- both picture and sound carriers. 35
foil from a board, use a straight edge and monic reduction required (in dB). dB is a suggested reasonable value.
a sharp scribe to score the thin copper foil. The first step in determining HR is to . Let us assume a station operating on 15
When the copper foil has been cut, use a find the harmonic relationship between meters' with l-kW PEP output. The third
razor blade to lift a corner. Careful the bands you use and the frequencies of harmonic (in channel 3) is down 40 dB.
heating with a soldering iron will reduce the weaker TV signals in your area ..The The antenna is a triband beam pointed at
the effort required to separate the copper harmonics of the hf amateur bands that the TV antenna, with a separation of 40
from the board. This technique of fall in the lower vhf TV channels are given meters (130 feet). Fringe-area reception
bonding two pieces of board or two sides in Fig. 1. Then, for each combination of applies. The numbers are:
of a piece of board can also be used to amateur band and TV channel concerned,
interconnect two capacitors when con- determine H R, using the following infor- Tp +60=
struction in one plane would require too mation: TA -50=
much area. Stray inductance must be T p - Peak transmitter power output in G T 0 =
minimized and sufficient clearance must dB above I mW. 25 watts is 44 dBm. G R =:= -10
300_0~M
when interference occurs in a TV channel ~JL ...JL _H-'"
~
that is not harmonically related to the ~\ ~ 1\ ~ ,ov
amateur transmitting frequency, the pos- LINE TO EHASSJS TERMINALS
ANT. ~ L ANT. TERMINALS
15-12 Chapter 15
be applied successfully iii the case of the receiver chassis, but if this is not
50-MHz trlfnsmissions, because they do possible the shield should be grounded to
not have sufficiently sharp cutoff charac- the chassis right at the antenna terminals.
teristics to give both good attenuation at The use of shielded transmission line
50-54 MHz and no attenuation above 54 300- OHM. I
I I
I 300 -OHM
for t~e receiver also >will be helpful in
MHz. A more elaborate design capable of LINE TO TV ... - - - - - - - - ~- -+ LINE TO re~ucmg ~esponse to harmonics actually /
giving the r~quired sharp cutoff has been SET : ANT. bemg radiated from the transmitter- or
described (Ladd, "50-MHz TVI - Its ~ransmi~ting antenna. In most receiving
Causes and Cures," QST, June and July, I I mstallatlOns the transmission line is very
I I
1954). This article also contains other in- Io... ___ 3-30pF JI
much longer than the antenna itself, and is
_____
formation useful in coping with the TVI consequently far more exposed to the
~
~
hi-fi unit. If the volume control has no shown in Fig. 30A.
effect on the level of interference or very In this system, two rf chokes are
SPEAKER ~ TO OUTPUT
slight effect, the audio rectification of the installed in series with the speaker leads TERMINALS 'TRANSFORMER
C2
amateur signal is taking place past the from the output transformers, or ampli-
volume control, or on the output end of fier output, to the speakers. These chokes ~
the amplifier. This is by far the most are simple to make and help keep rf out of (A)
common type. It usually means that the the amplifier. In particularly stubborn
amateur signal iS,being picked up on the cases, use shielded wire for the speaker RFC3
15-14 Chapter 15
PREAMPLIFIER
filter,. the same type used for TVI. The
RFC5 filter should be installed as close as
possible to the antenna inpu, of the tuner.
The high-pass filter will attenuate the
amateur fundamental signal, thus pre-
venting overloading of the front end.
Shielding
Lack of shielding on the various
INPUT
components in a hi-fi installation can
permit rf to get into the equipment. Many
units have no bottom plates, or are
iristalled in plastic cases. One easy method
of providing shielding is to use aluminum
foil. Make sure the foil doesn't short
circuit the components, and connect it to
chassis ground.
15-16. Chapter 15
Chapter 16
Test Equipment
and Measurements
desired extent. The meter itself has an for Rl should be approximately twice the I I
inherent resistance - its intt:rnal resis- expected internal resistance of the meter. I I
tance- which determines the full-scale For highly sensitive meters (loo~ A and
current through it when its rated voltage is less) 1000 ohms should be adequate. For
applied. (This rated voltage is of the order less sensitive meters 100 ohms should
of a few millivolts.) By connecting an suffice.
external resistance in parallel with the The value for minimum resistance at R2
internal resistance, as in Fig. 1, the current can be calculated using Ohm's Law. For Fig. 1 - Use of a shunt to extend the calibration
will divide between the two, with the example, if the meter is a 0-1 rnA type and range of a current-reading instrument.
meter responding only to that part of the the supply is a 1.5-volt battery, the
current which flows through the internal ,minimum resistance required atR2 will be
resistance of its movement. Thus it reads
only part of the total current; the effect is R 1.5
to make more total current necessary for a 2 = 0.001
full-scale meter reading. The added R2(min) = 15000hms
resistance is called a shunt.
It is necessary to know the meter's In practice a 2000- or 2500-ohm
internal resistance before the required potentiometer would be used.
value for a shunt can be calculated. It may ,}JA,mA,
vary from a few ohms to ~ few hundred, Making Shunts ,......._ _---J
+ A
with the higher resistance values as- Homemade shunts can be constructed
sociated with higher sensitivity. When from any of various special kinds of +
known, it can be used in the formula resistance wire, or from ordinary copper
SUPPLY
below to. determine the required shunt for wire if no resistance wire is available. The
a given current multiplication: copper wire table in this handbook gives Fig. 2 - A safe method for determining the in-
ternal resistance of a meter.
Rm the resistance per 1000 feet (305 m) for
R= n-l various sizes of copper wire. After
computing the resistance required, deter-
where:
mine the smallest wire size that will
R = the shunt +~---------~
carry the full-scale current (250 circular
Rm = internal resistance I mils per ampere is a satisfactory figure for
n = the factor by which the original this purpose). Measure off enough wire to
MULTIPLIER
18-2 Chapter 11
voltmeter, the more accurate the meas- meter alone. Strictly, the internal re- The value of shunt resistance must be
urements, especially' in high-resistance sistance of the meter should be subtracted calculated from the known or estimated
circuits. Current flowing through d1e from the calculated value but this is maximum current expected in the circuit
voltmeter will cause a change in the seldom necessary (except perhaps for very (allowing a safe margin) and the voltage
voltage between the points where the low ranges), since the meter resistance will required for full-scale deflection of the
meter is. connected, compared with the be negligibly small compared with the meter with its multiplier.
voltage with the meter absent. This is multiplier resistance. An exception is
illustrated in Fig. 4. when the instrument is already a volt- Power
meter and is provided with an internal Power in direct-current circuits is
Multipliers multiplier, in which case the multiplier determined by measuring the current and
The required mUltiplier resistance is resistance required to extend the range is voltage. When these are known, the power
found by dividi~g the desired full-scale R=Rm(n-l) is equal to the voltage in volts multiplied
voltage by the' current, in amperes, where by the current in amperes. If the current is
required for full-scale deflection of the Rm = total resistance of the measured with a milliammeter, the read-
ing of the instrument must be divided by
instrument itself
n = factor by which the scale is to 1000 to convert it to amperes.
be multiplied The setup for measuring power is
shown in Fig. 6, where R is any de
For example, if a lO00-ohms-per-volt volt- "load," not necessarily an actual resistor.
250V meter having a calibrated range of 0-10
volts is to be extended to 1000 volts, Rm Resistance
1000n/V METER READS APF! 81V
is 1000 X 10 = 10,000 ohms, n is 1000/ Obviously, if both voltage and current
20kO/V METER READS APP. 98V
11 MEG METER RE ADS APP. 99V
10 = 100, and R = 10,000 (100 - 1) are measured in a circuit such as that in
= 990,000 ohms. Fig. 6 the value of resistance R (in case it
When extending the range of a volt- is unknown) can be calculated from
meter or converting a low-range meter Ohm's Law. For accurate results the
Fig, 4 - Effect of voltmeter resistance on into a voltmeter, the rated accuracy of the internal resistance of the ammeter or
accuracy of readings. It is assumed that the dc il}strument is retained only when the milliammeter, mA, should be very low
resistance of the screen circuit is constant at 100 multiplier resistance is precise. Precision
kilohms, The actual current and voltage without compared with the resistance, R, being
the voltmeter connected are 1 rnA and 100 volts,
wire-wound resistors are used in the measured, since the voltage read by the
The voltmeter readings will differ because the multipliers of high-quality instruments. voltmeter, V, is the voltage across MA
different type~ of meters draw different amounts These are relatively expensive, but the and R in series. The instruments and the
of current through the 150-kilohm resistor, home constructdr can do quite well with dc voltage should be chosen so that the
I-percent tolerance composition resistors. readings are in the upper half of the scale,
They should be "derated" when used for if possible, since the percentage error is
this purpose - that is, the actual power less in this region.
dissipated in the resistor should not be
more than 1/4 to 1/2 the rated dissipation The Ohmmeter
- and care should be used to avoid Although Fig. 6 suffices for occasional
overheating the body of the resistor when resistance measurements, it is incon-
soldering to the leads. These precautions venient when frequent measurements over
SHUNT will help prevent permanent change in the a wide range of resistance are to be made. '
resistance of the unit. The device generally used for this purpose
Ordinary composition resistors are is the ohmmeter. This consists fund-
generally furnished in 10- or 5-percent amentally of a voltmeter (or milliam-
tolerance ratings. If possible errors of this meter, depending on the circuit used)
order can be accepted, resistors of this and a small battery, the meter being
type may be used as multipliers. They calibrated so the value of an unknown
should be operated below the rated power resistance can be read directly from the
Fig, 5 - Voltmeter method of measuring current. dissipation figure, in the interests of scale. Typical ohmmeter circuits are
This method permits using relatively large v!jlues long-time stability.
of resistance in the shunt, standard values of shown in Fig. 7. In the simplest type,
fixed resistors frequently b'eing usable. If the DC Measurement Circuits: the Voltmeter shown in Fig. 7 A, the meter and battery
multiplier resistance is 20 (or more) times the are connected in serie.s with the unknown
shunt resistance, the error in assuming that all A current-measuring instrument should resistance. If a given deflection is obtained
the current flows through the shunt will not be of have very low resistance compared with
consequence in most practi,cal applications, with terminals A-B shorted, inserting the
the resistance of the circuit being measured; resistance to be measured will cau~e the
otherwise, inserting the instrument will meter reading to decrease. When the
cause the current to differ from its resistance of the voltmeter is known, the
value with the instrument out of the follpwing formula can be applied:
circuit. The resistance of many circuits in
radio equipment is high and the circuit R =eRm - - Rm
DC operation is affected little, if at all, by E
v
VOLTAGE addin.g as much as a few hundred ohms in where
series. In such cases the voltmeter method R = resistance to be found
of measuring clirrent, shown in Fig. 5, is e = voltage applied (A-B shorted)
frequently convenient. A voltmeter (or E = voltmeter reading with R connec-
low-range milliammeter provided with a ted, and
multiplier and operating as a voltmeter) Rm = resistance of the voltmeter.
Fig, 6 - Measurement of power requires both having a full-scale voltage range of a few The circuit of Fig. 7A is not suited to
current and voltage measurements; once these
values are known the power is equal to the
volts is used to measure the voltage drop measuring low values of resistance (below
product P = EI. The same circuit can be used across a suitable value of resistance acting a hundred ohms or so) with a high-
for measurement of an unknown resistance. /as a shunt. resistance voltmeter. For such measure-
16-4 Chapter 16
Rl A
VOC
~
01
R13
T1 FUSE
+
Qh
C4
R14
r
VAC
1
Fig. 9 - Vacuum-tube voltmeter circuit.
C 1, C3 - 0.002- to 0.005-uF mica. R6, R7- 2to 3MO. R15-10MO.
C2 - 0.01 uF, 1000 to 2000 volts, paper or mica. R8- 10-kO, variable (calibrate). R16-3MO.
C4 - 16 uF, 150 V electrolytiC. R9, R1 0 - 2000 to .3000 ohnk R17 - 1O-M 0 variable.
D1 - 400 PRY rectifier. R11 - 5000- to 1O,OOO-ohm control (zero set). T1 - 120-V, 15-mA transformer (only secondary
M - 0-200 microammeter. R12- 10 to 50 kO. shown).
R1-1 mO, 1/2waU. R13, R14- Approx 25 kO. A 50-kO slider-type V1- Dual triode, 12AU7A.
R2-R5, inc I. - To give desired voltage ranges, wire·wound can be used. V2 - Dual diode, 6AL5.
totaling 10 mO.
device actually measures a small vo.ltage vo.ltage divider having a resistance o.f 10 by R 7 co.nnected to. the grid o.f the seco.nd
acro.ss a po.rtio.n o.f a high-resistance megohms, tapped as required, in series tube sectio.n.
vo.ltage divider co.nnected to. the circuit with a I-mego.hm resisto.r inco.rpo.rated in Values to. be used in the circuit depend,
being measured, Vario.us vo.ltage ranges a probe that makes the actual co.ntact with co.nsiderably o.n the supply vo.ltage and
are o.btained by appro.priate taps' o.n the the "ho.t" side o.f the circuit under the sensitivity o.f the meter, M. R12 and
vo.ltage divider. measurement. The to.tal vo.ltmeter re- R 13-RI4, sho.uld be adjusted by trial so.
In the design o.f electro.nic vo.ltmeters it sistance, including pro.be, is therefo.re II that the vo.ltmeter circuit can be bro.ught
has beco.me practically standard to. use a mego.hms. The pro.be resisto.r serves to. to. balance, and to. give full-scale de-
iso.late the vo.ltmeter circuit fro.m the flectio.n o.n M with abo.ut 3 Vo.lts applied to.
"active" circuit. the left-hand grid (the vo.ltage cho.sen fo.r
this determines the Io.west vo.ltage range o.f
A Vacuum-Tube Voltmeter the instrument). The meter co.nnectio.ns
. A typical vacuum-tube vo.ltmeter can be reversed to. read vo.ltages that are
(VTVM) circuit is given in Fig. 9. A dual . negative with respect to. ground. The small
trio.de, VI, is arranged so. that, with no. circuit asso.ciated with V2 is fo.r ac meas-
vo.ltage applied to. the left-hand grid, urements.
equal currents flo.W thro.ugh bo.th sectio.ns. As co.mpared with co.nventio.nal. dc
Under this co.nditio.n the two. catho.des are instruments, the VTVM has the dis-
at the same Po.tential and no. current flo.WS advantages o.f requiring a so.urce o.f Po.wer
thro.ugh M. The currents can be adjusted fo.r its o.peratio.n, and generally must have
to. balance by Po.tentio.meter, Rll, which its "co.ld" terminal gro.unded in o.rder to.
takes care o.f variatio.ns in the tube sec- o.perate reliably. It is also. so.mewhat
tio.ns and in the values o.f catho.de resisto.rs susceptible to. erratic readings from rf
R9 and RIO. When a Po.sitive dc vo.ltage pickup when used in the vicinity o.f a
current thro.ugh that tube sectio.n increases; transmitter,. and in such cases may require
so. the current balance is upset and the shielding. Ho.wever, its advantages o.ut-
meter indicates. The sensitivity o.f the weigh these disadvantages in many ap-
meter is regulated by R8, which serves to. plicatio.ns. . .
adjust the calibratio.n. R12, co.mmo.n to.
the catho.des o.f bo.th tube sectio.ns, is a An FET Volt-ohmmeter
feedback resisto.r that stabilizes the sys- The circuit sho.wn in Fig. 11 makes use
tem and makes the readings linear. R6 of two. field-effect transisto.rs in a bal-
and C I fo.rm a filter fo.r any ac co.mpo.nent anced circuit. Since no. two. active devices
Fig. 10- Exterior view olthe FET Volt-ohmmeter. that may be present,' and R6 is balanced have exactly the same characteristics,
DCo.-<>-----'
CALIBRATION o AC
OHMS
DC S2,B
4.7M
OHMS 0.005
1SM
CALIBRATION
RS R9
~OOk 2.~M
1M
OHMS J4
P6~'>-----~------------------------------~
P3 01 6.2 V
400mW
Fig. 11 -.Schematic diagram of the FET YOM. All resistors are 1/4-watt carbon types except for the potentiometers. Numbered components not appearing
in the parts list are for text callout only. All controls except R,11 are for calibration. _ .
8T1 - Battery, 9-volt rectangular. P1-P3, incl. - Standard tip plugs. S 1 - Spst toggle.
01 - Zener diode, S.2-V. P4 - Standard test probe. Mount 1-M!} 1/4-watt S2 - 2-pole, 3-position rotary.
J1-J4, incl. - Standard tip jacks. resistor inside probe tip. S3- 2-pole, 5-position rotary:
M 1 - Panel meter, 0-50 uA dc; Radio Shack 22- P5, PS - Standard test probe.
051 or equiv. 01,02 - Motorola MPF102; do not substitute.
some ineans must be incorporated to system. Conventional VTVMs and VOMs double-sided, glass-epoxy circuit board
balance the circuit under static conditions. . use scales that are cramped on the high material with the overall dimensions
The ZERO potentiometer does just that end and expanded on the low resistance measuring 4 X 6 X 2-1/4 inches (100 X
since the meter will read exactly "0" end. This logarithmic system is im- 150 X 60 mm). All seams are soldered
when the circuit is balanced. Any im- practical for a home-constructed in- together along their entire length to
balance causes the meter to deflect, the strument since special meter faces are not ensure a rigid construction. The battery is
amount of deflection proportional to the generally available. Linear readout of held between two pieces of circuit board
degree of imbalance. resistance allows the user to read the value material 'soldered to the sides at several
Voltage scales for both ac and de are of resistance directly from a standard locations.
0-0.5, 0-5, 0-50 and '0-500. A series of meter face. All components other than the meter
dividers (R2 through R5) feed a portion of Potentiometers R6 thr.ough RIO are • switches, battery and zero potentiometer
the voltage being measured to the bridge used in place of precision, nonstandard- are mounted on a circuit board that
circuitry. A I-MO resistor is used in the tip value resistors. Each potentiometer con- measures 2-7/8 X 3-1/8 inches (73 X 79
of the de probe bringing the total input trols the voltage division for its associated mm). A suitable foil pattern with parts
impedance to approximately 7 megohms. range. layout is shown in Fig. 12.
The use of potentiometers in the divider Under normal circuit conditions with The schematic shows a number of
alleviates the need for precision, special- the instrument placed in the ohms . connections to ground. In this particular
value resistors, thereby reducing the cost position, the meter will rest gently against circuit, ground is not the cabinet of the
of the unit. . the peg, off scale at the high end. When instrument but rather a "floating" ground.
. Measurements o(ac voltage are fac- the ohmmeter leads are connected to- By not connecting any of the circuitry to
ilitated by rectifying the ac and reading gether, the zero potentiometer is adjusted the cabinet there is no 'chance of having
the resulting de directly. Two IN4007 so that the meter indicates zero resistance. dangerous voltages on the case. This
diodes, a 22-MO resistm, and a 0.05-IlF Separating the leads causes the pointer to means that the circuit-board ground foil
capacitor form the rectifier circuit. R 1 is return to its position resting against the should not be allowed to contact the
used to calibrate the instrument for ac high-end peg. Dl is used to limit the caoinet.
measurements. voltage fed to the bridge so that the Although the unit shnwn in the
Resistance measurements are made in pointer does not slam against the peg. photographs was left natural (tarnished
ohms using five ranges: 0-50, 0-500, copper) with a clear acrylic coating, there
0-5000, 0-50k and 0-500k. This circuit Construction is no reason why the builder should not
makes use of a linear ohms-readout The enclosure is made from pieces of paint the finished VOM. Treat the copper
18-8 Chapter 18
Fig. 13.,- The inside of the FET VOM. Leads are
dressed with wire-tying twine to provide a neat
appearance.
Fig. 12 - Circuit-board layout with parts overlay for the FET VOM. Shaded areas represent unetched
copper areas of circuit board. This view is from the foil side of the board.
done in a similar manner. A resistor that
will allow the meter to read approximately
midscale for each range will be required.
like any other metal surface when on the calibrated meter. Do the same for If the resistors are ofthe precision variety,
painting. Any type of labeling that suits each of the other ranges ~sing a voltage a calibrated ohmmeter will not be
the builder's fancy may be used. Dry that will allow the meter to read near required. However, if the resistors used
transfer-type labels were used on the unit midscale. Should 250 volts not be for calibration are of five-percent tole-
shown. available for the high-range calibration, rance or greater, it would be wise to use a
50 volts could be used, yielding only a calibrated meter. For example, a 27.ohm
Calibration small difference in accuracy. Care should precision resistor could be used for the
Adjustment of the completed FET be taken to touch only the plastic insulation lowest frequency range. R6 would be
YOM is simple. However, it does require on the potentiometers since potentially adjusted for a reading of exactly 27 ohms
the use of a calibrated meter and a source dangerous voltages are present in the on the FET YOM. A 10- or 20-percent
of . variable-voltage dc. The dc ranges circuit. tolerance resistor could be used provided
should be adjusted first. Connect the Ac calibration is somewhat simpler a calibrated meter is available. In that case
calibrated voltmeter in parallel with the since the basic voltage dividers have the FET YOM reading should be made
FET YOM and attach these connection already been calibrated. The ac line the same as the calibrated meter. Simply
points to the variable-voltage dc supply. voltage should be used for calibration, do the same for the remainder of the
Start with the lowest range (0-0.5) and set again conforming the reading on the FET resistance ranges. That completes the
the supply voltage for a midscale reading YOM with the calibrated meter. Rl is calibration of the instrument. It is now
. (0.25 volt). Adjust R2 so that the FET provided for this adjustment. ready for use in those many applications
YOM reading conforms with the reading Calibration of the ohmmeter circuitry is around your shack.
(A)
}-~
- (A)
I
-~---
---PEAK
1{-~---
--PEAK
o.e pos.
05 POS.
+' -----------AVG.(0.31S) + - - --- --- --'AVERAGE
t 0 --- ----
t 0 ---- (8)- -- (8)
05
l{~-fIfJJ-- ---.:...
---PEAK
NEG.
0.5
+t
i
0
- - - -
NEG
-
-~-J).~rs~-~0=CS--AVERAGE
- - -
_=.::J:~tt ___'C::..::.J= ___'ct::.\ __
PEAK
{~
--
~.e ____S~~.H
1{--H-J5..-J5..-n-- _______ --AVERAGE
ro ~Ij~j~j~I~:;~' PEAK
(0.636)
t 0 - - - - --
(D)
--
18-8 Chapter 16
..J
C1
Ri 01
AC
01
0---1 +
02 FORWARD mA
IN,PUT
AC 02 R2 DCV
INPUT OUT
_REVERSE VOLTS FORWARD VOLTS-
• HALF-WAVE
SERIES
(\REAKDOWN
0--\\ C1 R1 A
V + I REVERSEPf
AC AC
~R2 OCV
INPUT INPUT 01 OUT
Series and shunt peak-reading circuits and since the capacitive reactance de- The transition from high to low resistance
are shown in Fig. 17B. Capacitor Cl creases with increasing frequency, the occurs at considerably less than 1 volt, but
isolates the rectifier from dc' voltage on impedance also decreases with frequency. is in the range of voltage required by the
the circuit under measurement. In the The resistance is subject to some variation associated dc meter. With an average-
series circuit (which is seldom used) the with voltage level, particularly at very low reading circuit the current tends to be
time constant of the C2-RI-R2 com- voltages (of the order of 10 volts or less) proportional to the square of the applied
bination must be very large comparea depending upon the sensitivity of the voltage. This crowds the calibration
with the period of the lowest ac frequency meter movement and the kind of rectifier points at the low end of the meter scale.
to be measured; similarly with CI-RI-R2 used. For most measurement purposes, how-
in the shunt curcuit. The reason is that the The ac load resistance represented by a ever, it is far more desirable for the output
capacitor is charged to the peak value of diode rectifier is approximately ~qual to to be "linear;" that.is, for the reading to
voltage when the ac wave reaches its maxi- one-half its dc load resistance. In Fig. be directly proportional to the applied
mum, and then must hold the charge (so 17A the dc load is essentially the meter voltage.
it can register on a dc meter) until the next resistance, which is generally quite low To achieve linearity it is necessary to
maximum of the same polarity. If the time compared with the multiplier resistance use a relatively large load resistance for
constant is 20 times the ac period the Rl, so the total resistance will be about the diode - large enough so that this
charge will have decreased by about five. the same as the multiplier resistance. The resistance, rather than the diode's own
percent by the time the next charge capacitance depends on the components resistance, will govern the current flow. A
occurs. The average drop will be smaller, and construction, test lead length and linear or equally spaced scale is thus
so the error is appreciably less. The error disposition, and o~her such' factors. In gained at the expense of sensitivity. The
will decrease rapidly with increasing general, it has little or no effect at amount of resistance needed depends on
frequency, assuming no change in the power-line and low audio frequencies, but the type of diode; 5000 to 50,000 ohms
circuit values, but will increase at lower the ordinary VOM loses accuracy at the usually suffices for a germanium rectifier,
frequencies. higher audio frequencies and is of little depending on the dc meter sensitivity, but
In Fig. 17B R I and R2 form a voltage use at rf. For radio frequencies it is several times as much may be needed for
divider which reduces the peak dc voltage necessary to use a rectifier having very low silicon. The higher the resistance, the
to 71 percent of its actual value. This inherent capacitance. greater the meter sensitivity required; i.e.,
converts the peak reading to rms on Similar limitations apply to thc peak- the basic meter must be a microammeter
sine-wave ac. Since the peak-reading reading circuits. In the parallel circuit the rather than a: low-range milliammeter.
circuits are incapable of delivering ap- resistive component of the impedance is
preciable current without considerable smaller than in the series circuit, since the Reverse Current
error, R2 is usually the ll-megohm input dc load resistance, Rl/R2, is directly When voltage is applied in the reverse
resistance of an electronic voltmeter. Rl is across the circuit being measured, and is direction there is a small leakage current
therefore approximately 4.7 megohms, therefore iIi parallel with the diode ac, load in semiconductor diodes. This is equiva-
making the total resistance approach 16 resistance. In both peak-reading circuits lent to a resistance, connected across the
megohms. A capacitance of 0.05 /.L F is the effective capacitance may range from rectifier, allowing current to flow during
sufficient for low audio frequencies under lor 2 to a few hundred pF. Values of the the half cycle which shQuld be completely
these conditions. Much smaller values of order of 100 pF are to be expected in nonconducting, and causing an error in
capacitance suffice for radio frequencies, electronic voltmeters of customary design the dc meter reading. This "back re-
obviously. and construction. sistance" is so high as to be practically
unimportant with silicon, but may be less
Voltmeter Impedance Linearity than. 100 kilohms with germanium.
The impedance of the voltmeter at the Fig. 18, a typical current/voltage The practical effect of back resistance is
frequency being measured may have an characteristic of a small semiconductor to limit the amount of resistance that can
effect on the accuracy similar to the error rectifier, indicates that the forward dy- be used in the dc load resistance. This in
caused by the resistance of a dc voltmeter, namic resistance of the diode is not turn affects the linearity of the meter
as discussed earlier. The ac meter acts like constant, but rapidly decreases as the scale. .
a resistance in parallel with a capacitance, forward voltage is increased from zero. The back resistance of vacuum-tube
Test Equipment and Measurements 16-9
diodes is infinite, for practical purposes. Insert a half-inch grommet ~t the top of
the tube shield, and slide the shield over
RF Voltage the cable and flexible braid down onto the
Special precautions must be taken to tube socket. The spring should make good
minimize the capacitive component of the contact with the tube shield to insure that
voltmeter impedance at radio frequencies. the tube shield (probe case) is grounded.
If possible, the rectifier circuit should be Solder an alligator clip to the other end of
installed permanently at the point where the flexible braid and mount a phone plug
the rf voltage to be measured exists, using on the free end of the shielded wire.
the shortest possiblerf connections. The Mount components close to the term-
dc meter can be remotely located, inal strip, to keep lead lengths as short
however. . as possible and minimize stray capaci-
For general rfmeasurements an rfprabe tance. Use spaghetti over all wires to
is used in conjunction with an electronic prevent accidental shorts. Fig. 19 - Af probe for use with an electronic
voltmeter. The case of the probe is constructed
voltmeter, substituted for the dc probe The phone plug on the probe cable from a seven-pin ceramic tube socket and a
mentioned earlier. The circuit of Fig. 20, plugs into the dc input jack of the 2-1/4·-inch (57-mm) tube shield. A half-inch (13-
essentially the peak-reading shunt circuit electronic voltmeter and rms voltages- are mm) grommet at the top of the tube shield
of Fig. 17B, is generally used. The· series read on the voltmeter's negative dc scale. prevents the output lead from chafing. A flexible
copper-braid grounding lead and alligator clip
resistor, installed in the probe close to the The accuracy of the probe is within· provide a low-inductance return path from the
rectifier, prevents rf from being fed ± 10 percent from 50 kHz to 250 MHz. test circuit
through the probe cable to the electronic The approximate input impedance is 6000
voltmeter, being helped in this by the ohms shunted by 1.75 pF (at 200 MHz).
cable capacitance. This· resistor, in con-
junction with the IO-MO divider resistance RF Power
of the electronic voltmeter, also reduces Power at radio frequencies can be
the peak rectified voltage to a dc value measured . by means of an accurately C1·
equivalent to the rms of the rf signal, to calibrated rf voltmeter connected across' 0---1------~"'''~
make the rf readings consistent with the the load in which the power is being RF
regular ac calibration. dissipated. If the load is a known pure
INPUT
18-10 Chapter 18
Qne voltage (designated "incident" or . terminals will result in a finite voltmeter
"forward") and the power reflected from reading. When used in a transmission line
the load is represented by the other. this reading is proportional to the
Because the relative amplitudes and phase reflected voltage. To measure the incident
AC relationships are definitely established by voltage the secondary terminals ofTI can
SOURCE
(Al the line's characteristic impedance, its be reversed. To function as described, the
length and the load impedance in which it secondary leakage reactance of Tl must
is terminated, a bridge circuit can separate be very large compared to the resistance
the incident and reflected voltages for of R l . .
measurement. This is sufficient for de- Instruments of this type are usually
'termining the SWR. Bridges designed for designed for convenient switching be-
AC
this purpose are frequently called rej1ec- tween forward and reflected, and are often
SOURCE tometers. calibrated to read power in the speCified
(9l
Referring to Fig. 20A, if RI and R2 are characteristic impedance. The net power
made equal, the bridge will be balanced transmission is equal to the incident
when Rx = Rs. This is true whether power minus the reflected power.
p
Rx is an actual resistor or the input
··resistance of a perfectly mitched trans- Sensitivity vs. Frequency
W
1
mission line, provided Rs is chosen to
1
T1 In all of the circuits in Fig. 13 the
11
equal the characteristic impedance of the sensitivity is independent of the applied
line. Even if the line is not properly frequency, within practical limits. Stray
RFIN RF OUT
matched, the bridge will still be balanced capacitances and couplings generally limit
C2
for power traveling outward on the line, the performance of all three at the
(el since outward-going power sees only the high-frequency end of the useful range.
Zo of the line until it reaches the load. Fig. 23A will work right down to dc, but
However, power reflected back from the the low-frequency performance of Fig.
load does not "see" a bridge circuit, and 23B is degraded when the capacitive
Fig. 23 - (A) Generalized form of bridge circuit
for either ac or dc. (9) One form of ac bridge the reflected voltage registers on the reactances become so large that voltmeter
frequently used for rf measurements. (e) SWR voltmeter. From the known relationship impedance becomes low in comparison (in
bridge for use in transmission lines. This circuit between the incident and reflected vol- all these bridge circuits, it is assumed that
is often calibrated in power rather than voltage. tages the SWR is easily calculated: the voltmeter impedance is high compared
with the impedance ofthe bridge arms). In
SWR= Vo +V, Fig. 23C the performance is limited at
ments on tr~nsmission lines. The Vo -V, low frequencies by the fact that the
circuits at Band C, Fig. 23, are favorites transformer reactance decreases with
for thi.s purpose. The "ReOected Power Meter" frequency, so that eventually the reac-
Fig. 23B is useful for measuring both Fig. 23C makes use of mutual in- tance is not very high in comparison with
transmission lines and "lumped constant" ductance between the primary and sec- the resistance of Rl.
components. Combinations of resistance ondary of Tl to establish a balancing
and capacitance are often used in one or circuit. CI and C2 form a voltage divider The "Monimatch"
more arms; this may be required for in which the voltage across C2 is in the A type of bridge which is quite simple
eliminating the effects of stray capaci- same phase as the voltage at that point on to make, but in which the sensitivity rises
tance. the transmission line. The relative phase ' directly with frequency, is the Monimatch
Fig. 23C is used only on transmission of the voltage across Rl is determined by and its various offspring. The circuit
lines, and only on those lines having the the phase of the current in the line. If a cannot be described in terms of lumped
characteristic impedance for which the pu're resistance equal to the design constants, as it makes use of the
bridge is designed. i!p.pedance of the bridge is connected to distributed mutual inductance and ca-
the "RF Out" terminals, the voltages pacitance between the center conductor of
SWR Measurement - the ReOectometer across R I and C2 will be out of phase and a transmission line and a wire placed
In measuring standing-wave ratio ad- the voltmeter reading will be minimum; if parallel to it. The wire is terminated in a
vantage is taken of the fact that the voltage the amplitudes of the two voltages are also resistance approximating the character-
on transmission line consists of two equal (they are made so by bridge istic impedance of the transmission line at
components traveling in opposite di- adjustment) the voltmeter will read zero. one end and feeds a diode rectifier at the
rections. The power going from the Any other value of resistance or im- other. A practical example is shown later
transmitter to the load is represented by pedance connected to the "RF Out" in this chapter.
Frequency Measurements
The regulations governing amateQr individual amateur to see that he stays form is a high-stability oscillator generat-
operation require that the transmitted safely "inside." ing a series of signals which, when
signal be maintained inside the limits of This is not difficult to do, but requires detected in the receiver, mark the exact
certain bands of frequencies.! The exact some simple apparatus and the exercise of . edges of the amateur assignments. It does
frequency need not be known, so long as it some care. The apparatus commonly used this by oscillating at a low frequency that
is not outside the limits. On this last point is the frequency-marker generator, and the has harmonics falling on the desired
there are no tolerances: It is up to the method involves use of the. station frequencies.
receiver, as in Fig. 24. U.S. amateur band limits are exact
'These limits depend on the type of emission and multiples of 25 kHz, whether at the
class of license held, as well as on international The Frequency Marker extremes of a band or at points marking
agreements. See the latest edition of The Radio
Amateur's License Manual for current status. The marker generator in its simplest the subdivisions between types of emis-
Ll
10mH
Fig. 27 - Schematic diagram of the marker generator. Resistors are 1/4·watt composition types.
Diodes are lN914 or similar switching types. •
Fig. 26 - The marker generator is housed in a Cl - 20-p~ trimmer, Johnson 189-508-5 or SI - Rotary, 1 pole, 4 position.
4·112 x 3x 2·1/2 inch (114 x 76 x 64 mm) equiv. Ul - LF·353N.
aluminum cabinet. A 9·volt battery makes the Jl - Coaxial connector, builder's choice. U2 - 4001.
instrument completely self·contained. J2 - Miniature binding post. U3 - 4013.
18·12 Chapter 18
•
the nearest marker frequencies above and markers which bracket the signal and Frequency-Marker Circuits
below the transmitter signal are tuned in dividing 25 by the number of dial
and identified. The transmitter frequency divisions between them. This will give the The basic frequency-determining ele-
is obviously between these two known number of kHz per dial division. . ment in most amateur frequency markers
frequencies. is a lOO-kHz crystal. Although the marker
If the marker frequencies are accurate, Transceivers generator should produce harmonics at
this is all that needs to be known - except The method described above is ap- 25-kHz and 50-kHz intervals, crystals (or
that the transmitter frequency must not be . plicable when the receiver and transmitter other high-stability devices) for frequen-
so close to .a band (or subband) edge that are separate pieces of equipment. When a cies lower than 100 kHz are expensive
sideband frequencies, especially in phone transceiver is used and the transmitting and difficult to obtain. However, there is
transmission, will extend over the edge. frequency is automatically the same as really no need for them, since it is easy to
If the transmitter signal is "inside" a that to which the receiver is tuned, setting divide the basic frequency down to any
marker at the edge <if an assignment, to the tuning dial to a spot between two figure one desires; 50 and 25 kHz require
the extent that there is an audible beat known marker frequencies is all' that is only two successive divisions, each by
note with the receiver's BFO turned off, required. two. In the division process, the harmonic
normal cw sidebands are safely inside the The proper dial settings for the markers output of the generator is greatly en-
edge. (This statement does not take into are those at which, with the BFO on, the hanced, making the generator useful at
account abnormal sidebands such as are signal is tuned to zero beat - the spot frequencies well into the vhf range.
caused by clicks and chirps.) For phone where the beat disappears as the tuning
the "safety" allowance is usually taken to makes the beat tone progressively lower. Simple Crystal Oscillators
\
be about 3 kHz, the nominal width of one Exact zero beat can be determined by a Fig. 25 illustrates a few of the simpler
sideband. A frequl;ncy difference of this very slow rise and fall of background circuits. Fig. 25A is a long-time favorite
order can be estimated by noting the noise', caused by a beat of a cycle or less where vacuum tubes are used and is often
receiver dial settings for the two 25-kHz per second; incorporated in receivers. Cl in this and
OUTPUT
Fig. 29 - Parts·placement guide. The tomponent side of the board is shown. Shaded areas represent an ,x-ray view of the copper foil.
DIRECT
IN
0.1 1000
QI
T 1M
6
PRESCALE
PRESCALE
1~0.01 51
INPUT
~J2
~---l4
GATE
EXCEPT AS INDICATED, DECIMAL VAUJES ·OF
CAPACITANCE ARE IN MICROFARADS « pF ) : 10Hz
OTHERS ARE IN PICOFARADS (pF OR jlJlFJ:
S2
RESISTANCES ARE IN OHMS; 100Hz
k .. I ooq. M"IOOO 00"" + Vee 0
1Hz
220
DECIMAL' POINT
LOCATOR
1KHz
20pF/SM
Fig. 32 - Schematic diagram of the counter. Ail resistors are 1/4-watt, 5-percent types. Nonpolarized capaCitors are miniature ceramiC unless noted
otherwise. Polarized capaCitors are tantalum types .
.J1, J2 - Coaxial connector, BNC type. U1, U5 - 74500.
51 - Toggle, spdt. U2, U15 - 74510. U4 - 11C90.
52 - Toggle, spdt, center off. U3 - 74L5196. U6, U7, US, U11 ,U13 - 74L590.
18.14 Chapter 18
the other circuits is used for exact similar station. is given principally as an example of II.
adjustment of the oscillating frequency to Fig. 25B is a field-effect transistor simple transistor arrangement. A much
100kHz, which is done by using the analog of the vacuum-tube circuit. How- better oscillator is shown at C. This is a
receiver for comparing one of the oscil- ever, it requires a 100mH coil to operate cross-connected pair of transistors form-
lator's harmonics with a standard fre- well, and since the harmonic output is not ing a ~ultivibrator of the "free-running"
quency transmitted by WWV, WWVH or a strong at the higher frequencies the circuit or "asynchronous" type, locked at 100
I
I
CLEAR I
I
L _______'-t--_-_-_-_-_-_-_-_-_-_-_-+-_-_-_-_-_-+-_--_-_-_-_+--+-_-_-_-_-_--.
UIOB
BLANK
6 LATCH
BLANK
CARRY
OUT
+Vcc
+Vcc
OVERFLOW
DETECTOR
LEADING ZERO
SUPPRESSION
U10D
BLANK ENABLE 8
FL ASH STROBE
Electronic division is accomplished by a comparator to provide a waveform circuit itself; it has to be extracted from
"bistable" Jlip"Jlop or cross-coupled cir- suitable for driving the frequency divider the circuit being measured; hence the
cuit which produces one output change stages. A CMOS dual D flip-flop IC pro- name absorption frequency meter. ,It will
for every two impulses applied to its input vides two divide-by-two stages for the 50- be observed that what is actually meas-
circuit, thus dividing the applied frequen- and 25-kHz outputs. Signal routing and ured is the frequency of the rf energy, not
cy by two. All division therefore must divisor selection is handled by a quad the frequency to which the circuit in which
be in terms of some power of two. In NOR gate and a diode matrix. A single- the energy is present may be tuned.
practice this is no handicap since with wafer rotary switch and a trimmer The measurement accuracy of such an
modern integrated-circuit flip-flops, cir- capacitor are the only controls. The complete instrument is low, compared with the ac-
cuit arrangements can be worked out for schematic diagram is drawn in Fig. 27. curacy of a marker generator, because the
division by any desired number. ' Q of a practicable LC circuit is not high
As flip-flops and gates in integrated Construction enough to make precise reading of the dial
circuits come in compatible series - A component-placement guide and et- possible. Also, any two circuits coupled
meaning that they work at the same ching pattern are given in Figs. 29 and 30. together react on each others' tuning.
supply voltage and can be directly The layout is not critical, but the rotor (This can be minimized by using the
coimected together - a combination of a and lor adjustment screw of the trimmer loosest coupling that will give an adequate
dual-gate version of Fig. 25C and a dual capacitor should be grounded so that con- indication.)
flip-flop make an attractively simple tact with a screwdriver won't affect the The absorption frequency meter has one
combination for the marker generator. frequency. Similarly, the crystal should be useful advantage over the marker generator
There are several different basic types positioned so that a screwdriver won't - it will respond only to the frequency to
of flip-flops, the variations having to do come too near it during frequency adjust- which it is tuned, or to a band of frequencies
with methods of driving (dc or pulse ment. Fig. 28 shows a successful arrange- very close to it. Thus there is no harmonic
operation) and control of the counting ment. Give due respect to the CMOS ICs ambiguity, as there sometimes is when using a
function. Information on the operating during assembly - they can be damaged marker generator.
principles and ratings of a specific type by static charges. Keep them in their pro-
usually can be obtained from the manu- tective material and don't insert them un- Absorption Circuit
facturer. The counting-control functions til the other components have been install- A typical absorption frequency-meter
are not needed in using the flip-flop in a ed. The output coupling capacitor isn't circuit is shown in Fig. 30A~ In addition to
simple marker generator, although they critical - the value shown allows fairly the adjustable tuned circuit, Lt-Cl, it
come into play when dividing by some constant-amplitude harmonics up to 30 includes a pickup coil, L2, wound over
number other than a.power of two. MHz into a 50-ohm load. If the unit is in- Lt, a high-frequency semiconductor diode,
\ stalled in a receiver, a small twisted-wire Dl, and a microammeter or low-range
Marker Generator for 100, 50 and 25 kHz "gimmick" capacitor should be used. (usually not more than 0-1 mA) milliam-
The signal source in the accompanying This capacitor should be adjusted for meter. A phone jack is included so
illustrations will deliver usable calibration minimal loading of the input circuit, the device can be used for listening to the
markers throughout the hf spectrum. consistent with adequate marker strength. signal.
When built into an enclosure and powered The sensitivity of the frequency meter
from a battery or regulated dc supply, the Adjustment and Operation depends on the sensitivity of the dc meter
unit is a test instrument that is very helpful The unit can be set to precisely 100 kHz movement and the size of L2 in relation to
in aligning receivers. Alternatively, the pc by zero-beating a harmonic against wwv L 1. There is an optimum size for this coil
assembly can be incorporated into a com- or a known broadcast station. A small which has to be found by experiment. An
munications receiver; in this service it fixed-value capacitor can be shunted alternative is to make the rectifier
becomes a "crystal calibrator." across the trimmer if necessary. Use a connection to an adjustable tap on Lt, in
Three integrated circuits are used in the slIver mica or NPO ceramic component which case there is an optimum tap point.
generator. A l00-kHz crystal oscillator is for this purpose. The dissipation in the In general, the rectifier coupling should be
designed around a dual lFET-input crystal and op amp is minimal, so the a little below (that is, less tight) the point
operational amplifier. Crystals in this fre- generator is stable with time. As a test in- that gives maximum response, since this
quency range show considerable variation strument, the marker generator should be will make the indications sharper.
in their characteristics as compared to hf set before beginning an alignment job. As
ones. A characteristic they all share, a crystal calibrator, it should be located Calibration
however, is their high expense - for this away from the' heat-generating com- The absorption frequency meter must
reason builders often purchase "bargain~' ponents of the receiver. After the receiver be calibrated by taking a series of readings
crystals, sometimes with disappointing has reached operating temperature, the on various frequencies from circuits
results. When the crystal characteristics marker frequency should be checked carrying rf power, the frequency of the rf
(resonance mode, load capacitance and against WWV from time to time. The energy first being determined by some
equivalent series resistance) are known, generator can be powered from any 9- to other means such as a marker generator
it's a fairly simple matter to design an ap- 12-volt dc source. The frequency varies and receiver. The setting qf the dial that
propriate oscillator circuit. Instruments of with the applied voltage, so a Zener diode gives the highest mefer indication is the
this type usually have oscillators made regulator should be used where the calibration point for that frequency. 1'his
18·18 Chapter 18
point should be determined by tuning V27
through it with loose coupling to the
53
r---,---- -1
I
circuit being measured. I
I U28
18-18 Chapter 18
CD f... Z on ;:! Z I'- ~ ~ CD
Z Z
IO Z Z :x:Q Z Z Z Z LU
(!li= (!ll- ii: ii:
ii: ii: :::>U ii: ii: :::>U ii: ii: ...J
1-1-
OLU OLU CD «I-
'"'":::> a:Z
IZ
I'-
:;
I'-
:; a: Z
IZ
CD
:;
CD
:; '"
:; :; iii U:::>
(1)11.
U:::>
LUll.
1-0 1-0 0 LUZ ~Z
0 0 0 0 0 0-
0
I- U l- I- U ~ l- I- l- I- a:-
11.
Fig. 36 - Parts-placement diagram for the pc board as seen from the component side of the board.
Fig. 37 -Etching pattern for the pc board. The top pattern is for tile top side of the board and the iower pattern is for the bottom of the board. Both
are to scale and black areas indicate unetched copper.
18-20 Chapter 18
Table 1
Rl R2 R3
ColI·Wlndlng In'ormatlon
27k 50k Freq. Range Cl C2 L1
MHz pF pF Turns
C5
2.3-4 15 15 71-1/2
0.01
3.4-5.1 33 10 39-1/2
4.B-B 10 33 25-112
7.9-13 10 33 14-1/2
12.8-21.2 10 33 6-112
o -DC vOLTAGE WITH BOTH
POTENTIOMETERS SET
21-34
34-60
10
10
33
33
4-1/2
2-1/2
AT MID ROTATION 60-110 10 33
90-200 not not
* . SEE TEXT
used used
RS S.M.- SILVER MICA
1000
"denotes a 1-1/2-turn coil of no. 1B enam. wire
EXCEPT AS INDICATED, DECIMAL wound on a 1/2-inch (13-mm) form spaced 1/B
VALUES OF CAPACITANCE ARE inch (3 mm) between turns. It shOuld be placed
IN MICROFARADS l jlF) ; OTHE~S so that the coil is near the top of the coilform.
ARE IN PICOFARADS l pF OR jljIF); ""denotes a hairpin loop made from flashing copper.
RESISTANCES ARE IN OHMS; 3/B-inch (9.5-mm) wide X 1-7/B-inch (89 mm)
k .,000. M'1,000,000 total length.
All other coils are wound with no. 24 enam. wire.
Fig. 39 - Schematic diagram of the dual-gate MOSFET dip metl[!r. All resistors are 1/2·watt
composition type. Capacitors are disc ceramic unless noted otherwise.
C1, C2 - See Table 1. M1 - Edgewise panel meter, 0·1 mA,
C3 - Variable capacitor, 35 pF, Millen 20035 Calectro 01-905 or equiv.
or equiv. 01 - Dual-gate MOSFET, RCA 40673.
01 - 1N34A or equiv. 02 - Npn transistor, 2N2222A.
J1 - Socket, Amphenol type S4. R2 - Potentiometer, 50 kG.
_____ L1 - See Table 1 for values. All coils R7 - Potentiometer, 5000 ohms.
wound on Millen 45005 coit forms. S1 - Spst on-off switch mounted on R7.
700 70 7 0.7 1\
~
0.6
\
600 60 6
~ \
\ \
500 50 5 0.5 '" ~
(' '"
400 40 4 0.4 ~ ~
C'
1\
1\(
"~
"'",
'So'
\I'
~
1(
.
<ll.
~
300 30 3 0.3 'So
~ ~
~-s.
("''V..
~ Il(,
\I' '\I'
1(
250 25 2.5 0.25
C'
1( \I' "1(
"'.
C'.
C'
1(", "', <ll.
'", C'.
200 20 2, 0.2
0.,.
. 1\ 1\
150 15 1.5 "0.1 5
1\
I'. 1\
\ 1\
1\ \ 1\
100 10 O.1
1\
1.5 2 2.5 3 4 5 6 7 8 9 10 15 20 25 30 40 50
D C 8 A FREQUENCY IN MEGAHERTZ
Fig. 42 - Chart for determining unknown values of Land C in the range of 0.1 to 100 "H and 2 to 1000 pF, using standards of 100 pF and 5 "H.
circuits, since nonresonant circuits or series of harmonic responses, based on the turns can be substituted. Thli inductance
components will not absorb energy at a lowest resonant frequency, which may will be 5 j.lH within amply close tolerances
specific frequency. The circuit may be lead to false conclusions respecting the if the specifications in Fig. 43 are followed
either lumped or linear (a transmission- behavior of the system. closely. In any case, the inductance can
line type circuit) provided only that it has Measurements with the dip meter are easily be ~djusted to the proper value; it
enough Q to give sufficient coupling to the essentially frequency measurements, and should resonate with the lOO-pF capacitor
dip-meter coil for detectable absorption of for best accuracy the coupling between the at 7100 kHz.
rf energy. Generally the coupling is meter and circuit under checking must be The setup for measuring an unknown is'
principally inductive, although at times as loose as will allow a perceptible dip. In shown in Fig. 44. Inductance is measured
there may be sufficient capacitive coup- this respect the dip meter is similar to the with the unknown connected to the stan-
ling between the meter and a circuit point absorption wavemeter. dard capacitance. Couple the dip meter to
that is at relatively high potential with the coil and adjust the meter for the dip,
respect to ground to permit a reading. For Measuring Inductance and Capacitance using the loosest possible coupling that
inductive coupling, maximum energy with the Dip Meter
absorption will occur when the meter is . With a carefully calibrated dip meter,
coupled to a coil (the same coupling rules properly operated inductance and
which' apply to any two coils are operative capacitance in the values ordinarily used
here) in the tuned circuit being checked, for the 1.5-50 MHz range can be
or to a high-current point in a linear measured with ample accuracy for prac-
circuit. tical work. The method requires two ac-
Because of distributed capacitance (and cessories: an inductance "standard" of
sometimes inductance) most circuits re- known value, and a capacitance standard
sonant at the lower amateur frequencies also known with reasonable accuracy.
will show quasi-linear-type resonances at Values of 100 pF for the capacitance and 5
or close to the vhf region. A vhf dip meter j.lH for the inductance are convenieht. The
will uncover these, often with beneficial chart of Fig. 42 is based on these values.
results since such "parasitic" resonances The Land C standards can be quite or-
can cause unwanted responses at har- dinary components. A small silver-mica
monics of the intended frequency, or be capacitor is satisfactory for the Fig, 43 - A convenient mounting, using
responsible for parasitic oscillations in capacitance, since the customary bindingpost plates, for Land C standards
amplifiers. Caution must be used in tolerance is ::t: 5 percent. The inductance made from commercially available parts. The
checking transmission lines or antennas standard can be cut from commercial capacitor is a 100-pF silver mica unit, mounted
so the lead length is as nearly zero as possi-
- and, especially, combinations ofanten- machine-wound coil stock; if none is ble. The inductance standard, 5 "H, is 17 turns
na and line - on this account, because available, a homemade equivalent in of coil stock, 1·inch (25.mm) diameter, 16 turns
these linear circuits have well-defined diameter, turn spacing, and number of per inch.
18·22 Chlpte, 18
Vee
220
(A)
1000
Rl R2
C4
t----+-l~T
;hCl
~'
(8)
R2
Fig. 44 - Setups for measuring inductance 20'
and capacitance with the dip meter. FREQ.
1
Fig. 45 - Twin-T audio oscillator circuit. S1A 4
-----ll~
Representative values for R1·R2 and C1 range
from 18 kG and 0.05,.F for 750 Hz to 15 kG and
0.02 ,.F for 1800 Hz. For the same frequency
will give a usable indication. Similar pro- range, R3 and C2·C3 vary from 1800 ohms and
0.02,.F to 1500 ohms and 0.01 ,.F. R4 Should
'----f_o-,s
cedure is followed for capacitance
measurement, except that the unknown is
be approximately 3300 ohms. C4, the output :h0'os
coupling capaCitor, can be 0.05 ,.F for high- 0,005
connected to the standard inductance. impedance loads. ~------~------~
Values are read off the chart for the fre-
quency indicated by the dip meter.
Coemcient of Coupling and for simple trouble shooting in af
The same equipment can be used for amplifiers, an oscillator generating one or
measurement of the coefficient of coup- two frequencies with good wave form is EXCEPT AS INDICATED, DECIMAL VAUlES OF
CAPACITANCE ARE IN MICROFARADS (JIF I ;
ling between two coils. This simply adequate. A "two-tone" (dual) oscillator, aTHERS ARE IN PItOFARADS ( pF OR JlJIFI;
requires two measurements of inductance is particularly useful for testing sideband RESISTANCES ARE IN' OHMS;
k.IOOO, Y.IOOO 000.
(of one of the coils) with the coupled coil transmitters, and adjusting them for on-
first open-circuited and then short-cir- the-air use.
cuited. Connect the lOO-pF standard The circuit of a simple RC oscillator Fig. 46- A simple audio oscillator that pro-
capacitor to one coil and measure the useful for general test purposes is given in vides a selectable frequency range, R2 and R3
inductance with the terminals of the Fig. 45. This "Twin-T" arrangement control the frequency and R1 varies the output
level.
second coil open. Then short the terminals gives a waveform that is satisfactory for
of the second coil and again measure the most purposes, and by choice of circuit
inductance of the first. The coefficient of constants the oscillator can be operated at cuit draws 4 rnA. The frequency range is
coupling is given by any frequency in the usual audio range. selectable from IS Hz to ISO kHz,
RI, R2 and CI form a low-pass type net- although a 1.5- to IS-Hz range can be in-
work, while C2C3R3 is high-pass. As the cluded with the addition of two 5-IlF non-
phase shifts are opposite, there is only one polarized capacitors and an extra switch
frequency at which the total phase shift position. Distortion is approximately one
from collector to base is 180 degrees, and percent. The output level under a light
where oscillation will occur at this frequency. load (10 k ohms) is 4 to 5 volts. This can
k = coefficient of coupling Optimum operation results when CI is ap- be increased by using higher battery
L 1 = inductance of first coil with proximately twice the capacitance of C2 voltages, up to a maximum of plus and
terminals of second coil open or C3, and R3 has a resistance about 0.1 minus 18 volts, with a corresponding ad-
L2 = inductance of first coil with that of RI or R2 (C2 = C3 and RI = justment of Rf .
terminals of second coil shortened R2). Output is taken across CI, where the Pin connections shown are {or the TO-5
harmonic distortion is least. A relatively case. If another package configuration is
Audio-Frequency Oscillators high-impedance load should be used- used, the pin connections may be dif-
Tests requiring an audio-frequency 0.1 megohm or more. ferent. Rf (220 ohms) is trimmed for an
signal generally call for one that is a A small-signal af transistor is suitable output level about five percent below
reasonably good sine wave, and best for Q l. Eith~ npn or pnp types can be clipping. This should be done for the
oscillator circuits for this are RC-coupled, used, with due regard for supply polarity. temperature at which the oscillator will
operating as nearly Class A amplifiers as R4, the collector load resistor, must be normally operate, as the lamp is sensitive
possible. Variable frequency covering the large enough for normal amplification, to ambient temperature. Note that the
entire audio range is needed for determin- and may be varied somewhat to adjust the output of this oscillator is direct coupled.
ing frequency response of audio ampli- operating conditions for best waveform. If you are connecting this unit into circuits
fiers, but this is a relatively unimportant where dc voltage is present, use a coupling
type of test in amateur equipment. The A Wide-Range Audio Oscillator capacitor. As with any solid-state equip-
variable-frequency af signal generator is A wide-range audio oscillator that will ment, be cautious around plate circuits of
best purchased complete; kits are readily provide a moderate output level can be tube-type equipment, as the voltage spike
available at prices that compare very built from a single 741 operational caused by charging' a coupling capacitor
favorably with the cost of parts. amplifier (Fig. 46). Power is supplied by may destroy the IC. This unit was
For most phone-transmitter testing, two 9-volt batteries, from which the cir- originally described by Schultz in
Te~t Equipment and Measurements 18~23
November 1974 QST. to a kilowatt).
For receiver and minipower transmitter
o
, Resistors at Radio Frequencies testing an excellent dummy antenna can
be made by installing a 51- or 75-ohm
Measuring equipment, in some part of composition resistor in a PL-259 fitting as
its .circuit, often requires essentially pure
resistance - that is, resistance exhibiting
shown in Fig.' 47. Sizes from one half to /
two watfs are satisfactory. The disc at the
only negligible reactive effects on the end helps reduce lead inductance and
frequencies at which measurement is completes the shielding. Dummy antennas Fig. 47 - Dummy antenna made by mounting
intended. Of the resistors available to made in this way have good characteris- a composition resistor.in a PL-259 coaxial
amateurs, this requirement is met only by tics· through the vhf bands as well as plug. Only the inner portion of the plug is
small composition (carbon) resistors. The shown; tl)e cap screws on after the assembly
at all lower frequencies. is completed.
inductance of wire-wound resistors makes
them useless for amateur frequencies. Increasing Power Ratings
The reactances to be considered arise More power can be handled by using a
from the inherent inductance of the number of 2-watt resistors in parallel, or
resistor itself and its leads, and from small series-parallel, but at the expense of
stray capacitances from one part of the introducing some reactance. Nevertheless,
resistor to another and to surrounding if some departure from the ideal im-
conductor,s. Although both the induc- pedance characteristics can be tolerated
tance and capacitance are small, their this is a practical method for getting
reactances become increasingly important increased dissipations. The principal prob-
as the frequency is raised. Small com- lem is stray inductance which can be
position resistors, properly. mounted, minimized by mounting the resistors on
show negligible capacitive reactance up to flat copper strips or sheets, as suggested in
100 MHz or so in resistance values up to a Fig. 48.
few hundred ohms; similarly, the in- The power rating on resistors is a
ductive reactance is negligible in values continuous rating in free air. In practice, SIDE VIEW
higher than a few hundred ohms. The the maximum power dissipated can be
optimum resistance region in this respect increased in proportion to the reduction Fig. 48 ~ Using resistors in series'parallel to
is in· the 50 to 200-ohm range, ap- increase the power rating of a small dummy
in duty cycle. Thus with keying, which has antenna. Mounted in this way on pieces of flat
proximately. a duty cycle of about one half, the rating copper, inductance is reduced to a minimum.
Proper mounting includes reducing lead can be doubled. With sideband the duty Eight l00-ohm 2-watt composition resistors in
length as much as possible, and keeping cycle is usually not over about one-third. two groups, each four resistors in parallel', can
the resistor separated from other resistors be connected in series to form a 50-ohm
The best way of judging is to feel the dummy. The open construction shown permits
and conductors. Care must also be taken resistors occasionally (with power off); if free air circulation.
in some applications to ensure that the too hot to touch, they may be dissipating
resistor, with its associated components, more power than they are rated for .
.does not form a closed loop into which a of its travel snaps it back very quickly to
voltage could be induced magnetically. the starting point - time progresses to the
So installed, the resistance is essentially The Oscilloscope right.
pure. In composition· resistors the skin The electrostatically deflected cathode- Most cathode-ray tubes for oscilloscope
effect is very small, and the rf resistance ray tube, with appropriate associated work require a deflection amplitude of
up to vhf is very closely the same as the dc equipment, is capable of displaying both about 50 volts per inch. For displaying
resistance. low- and radio-frequency signals on its small signals, therefore, considerable amp-
fluorescent screen, in a form which lends lification is needed. Also, special circuits
Dummy Antennas itself to ready interpretation. (In contrast, have to be! used for linear deflection. The
A dummy antenna is simply a resistor the magnetically deflected television pic- design of amplifiers and linear deflection
that, in impedance characteristics, can be ture tube is not at all suitable for circuits is complicated, and extensive texts
substituted for an antenna or transmission measurement purposes.) In the usual are available. For checking modulation of
line for test purposes. It permits leisurely display presentation, the fluorescent spot transmitters, a principal amateur use of
transmitter testing without radiating a moves across the screen horizontally at the scope, quite simple circuits suffice. A
signal. (The amateur regulations strictly some known rate (horizontal deflection or 6O-Hz voltage from the power line makes
limit the amount of "on-the-air" testing horizontal sweep) and simultaneously is a satisfactory horizontal sweep, and the
that maybe done). It is also useful in moved vertically by the signal voltage voltage required for vertical deflection can
testing receivers, in that electrically it being examined (vertical deflection). Be- easily be obtained from transmitter rf
resembles an antenna, but does not pick cause of the retentivity of the screen and circuits without amplification.
up external noise and signals, a desirable the eye, a rapidly deflected spot appears as For general measurement purposes
feature in some tests. a continuous line. Thus a varying signal amplifiers and linear deflection circuits
For transmitter tests the dummy anten- voltage causes a pattern to appear on the are needed. The most economical and
na must be capable of dissipating safely screen. satisfactory way to obtain a scope having
the entire power output of the transmitter. Conventionally, oscilloscope circuits these features is to ,assemble one of the
Since for most testing it is desirable that are designed so that in vertical deflection many kits available.·
the dummy simulate a perfectly matched the spot moves upward as the signal
transmission line, it should be a pure voltage becomes more positive with Simple Oscilloscope Circuit
resistance, usually of approximately 52 or respect to ground, and vice versa (there Fig. 49 is an oscilloscope circuit that has
73 ohms. This is a severe limitation in are exceptions, however). Also, the all the esSentials f'br modulation
home construction, because nonreactive horizontal deflection is such that with an monitoring: controls for cente~ing, focus-
resistors of more than a few watts rated ac sweep voltage - the simplest form - ing, and adjusting the brightness of the
safe dissipation are very difficult to positive is to the right; with a linear sweep fluorescent spot; voltage dividers to
obtain. (There are, however, dummy - one which moves the spot at a uniform supply proper electrode potentials to the
antenna kits available that can handle up rate across the screen and then at the end cathode-ray tube; and means for coupling
18-24 Chapter 18
monitoring modulation arC( given in
earlier chapters.
CI Quasi-Linear Sweep 117~T11
E For wave-envelope patterns that require
a fairly linear horizontal sweep, Fig: 50
shows a method of using the substantially
60H~
0
omit A8 and connect the bottom end of A7 to on the relative amplitudes, frequencies
the top end of A9.
C1·C5, incl. - 1000-volt disc ceramic.
A1, A2, A9, A11 - Volume-control type, linear
taper. A9 and A11 must be well insulated
and phases of the two voltages. If the
ratio between the two frequencies is con-
stant and can be expressed in integers, a
1:1
D<J 2:1
rjJJ 00
A3, A4, A5, A6, A10 - 1/2 watt. The stationary patterns obtained in this
A7, A8 - 1 watt. way are called Lissajous figures. Ex- 3:1 3:2
V1 - Electrostatic-deflection cathode-ray tube.
2- to 5-inch (51 to 127-mm). Base connec- amples of some of the simpler Lissajous
tions and heaterratings vary with type figures are given in Fig. 51. The frequency
chosen. ratio is found by counting the number of
~
loops along two adjacent edges. Thus in
the vertical and horizontal signals to the the second figure on the left there are 4:3
deflection plates_ three loops along a horizontal edge and
The circuit can be used with electro- only one along the vertical, so the ratio of
static-deflection tubes from two to five the vertical frequency to the horizontal
inches in face diameter, with voltages up frequency is 3: 1. Similarly, in the bottom Fig. 51 - Lissajous figures and corresponding
to 2500. Either set of deflecting electrodes figure there are four loops along the frequency ratios for a 9O-degree phase relation-
;>hip between the voltages applied to the two
(Dl-D2, or D3-D4) may be used for either horizontal edge and three along the ver- sets of deflecting plates.
horizontal or vertical deflection, de- tical edge, giving a ratio of 4:3. Assuming
pending on how the tube is mounted. that the known frequency is applied to the
In Fig. 49, the centering controls are horizontal plates, the unknown frequency
not too high above electrical ground, so is
they do not need special insulation. thus cover the audio range useful for voice
However, the focusing and intensity con- communication.
trols are at a high voltage above ground
and therefore should be carefully in-
f2 =:~ n An oscilloscope having both horizontal
and vertical amplifiers is desirable, since it
sulated. Insulated couplings or· extension where fl = known frequency applied to is convenient to 'have a means for
shafts should be used. horizontal plates adjusting the voltages applied to the
The tube should· be protected from f2 = unknown frequency applied deflection plates to secure a suitable
stray magnetic fields, either by enciosing it to vertical plates pattern /iize.
in an iron or steel box or by using one of nl = number of loops along a
the special CR tube shields available. If vertical edge A Tester for FET and Bipolar Transistors
the heater transformer (or other trans- n2 = number of loops along a The circuit shown is intended solely as a
former) is mounted in the same cabinet, horizontal edge. tester for npn and pnp transistors,
care must be used to place \t so the stray An important application of Lissajous junction FETs, and dual-gate MOSETs.
field around it does not deflect the spot. figures is in the - cl\libration of audio- This equipment is not for use in checking
The spot cannot be focused to a fine point frequency signal generators. For very low audio or high-power rf transistors.
when influenced by a transformer field. frequencies the 6O-Hz power-line frequency The circuit of Fig. 53 is an oscillator
The heater transformer· must be well is held accurately enough to be used which is wired so that it wiII test various
insulated, and one side of the heater as a standard in most localities. The small-signal transistors by switching the
should be connected to the cathode. The medium audio-frequency range can be battery polarity and bias voltage. A
high-voltage dc can be taken from the cover~ by comparison with the 440- and crystal for the upper range of the hf spec-
transmitter plate supply; the current 600-Hz modulation on the WWV trans- trum is wired into the circuit permanently,
required is negligible. missions. It is possible to calibrate over a bot could be installed in a crystal socket if
Methods for connecting the oscil- 10: 1 range, both upward and downward the builder so desires. A 20-MHz crystal
loscope to a transmitter for checking or from each of the latter frequencies and was chosen for this model. Any hf crystal
Test Equipment and Measurements 16~25
cut for fundamental mode operation can
be used.
When testing FETs the bias switch, S3,
is placed in the FET position, thus
removing R2 from the circuit. However,
when testing bipolar transistors the switch
position must be changed to BIPOL so
that forward bias can be applied to the
base of the bipolar transistor under test.
R 1 is always in the circuit, and 'serves as a
gate-leak resistor for FETs being evalua-
ted. It becomes part of the bias network
when bipolars are under test. Cl is use'd
for feedback in combination with the
internal capacitances of the transistors
being checked. Its value may have to be
changed experimentally if crystals for
lower frequencies are utilized in the
circuit. Generally speaking, the lower the
crystal frequency, the greater the amount
of capacitance needed to assure oscil-
lation. Use only that amount necessary to
provide quick starting of the oscillator.
Components R3and R4 are used as a
voltage divider to provide bias for
dual-gate MOSFETs. C2 is kept small in Fig. 52 - t:xterior view of the transistor/FET Fig. 54 - The completed crystal/transistor
tester. checker.
value to minimize loading of the oscillator
.by the low-impedance voltage doubler,
D 1 and D2. Rectified rf from the
oscillator is monitored on M 1. Meter
deflection is regulated manually by means
of control R5. S 1 is used to select the
desired supply voltage polarity - nega-
tive ground for testing n-channel FETs
and npn bipolars, and a positive ground 0.001
when working with p-channel and pnp
devices.
When testing MOSFETs which are not
gate protected (3N140 for one), make
certain that the transistor leads are
shorted together until the device is seated
in the test socket. Static charges on one's
R3
hands can be sufficiently great to damage 330k
the insulation within the transistor. Use a
~
single strand of wire from some no. 22 or
24 stranded hookup wire, wrapping it two
or three times around the pigtails of the
FET as close to the transistor body as
" RFCI
2.SmH
SIA NOPN
_ _..J
The meter indication is significant in
checking any type of transistor. If the EXCEPT AS INDICATED, DECIMAL
device is o~en, shorted, or extremely VALUES OF CAPAC ITANCE ARE
leaky, no oscillation will take place, and IN MICROFARADS (JlF) ; OTHERS
ARE IN PICOFARADS ( pF OR JlJIF I;
the meter will not deUect. The higher the RESISTANCES ARE IN OHMS;
meter reading, the greater the vigor of the k 01000. M-1,OOO,OOO.
transistor at the operating frequency.
High meter readings suggest that the
transistor is made for vhf or uhf service,
Fig. 53 - Schematic diagram of the transistor tester. Capacitors are disc ceramic or mica.
and that its beta is medium to high. Lower Resistors are 1/2- or 1/4-watt composition except for R5. Numbered components not appearing in
readings may indicate that the transistor is parts list are so designated for text discussion. . .
designed for hf use, or that it has very low Bn - Small 9-V transistor-radio battery. with switch.
gain. Transistors that are known to be 01,02 - 1N34A germanium diode or equiv. RFC1 - 2.5-mH rf choke.
J1 - Four~terminal transistor socket. S1 - Two-pole double-throw miniature toggle.
good but will not cause the circuit to M1 - Microammeter. Calectro 01-910 S2 - Part of R5.
oscillate are most likely made for low- used here. S3 - Spst miniature toggle.
frequency o~ audio applications. R5 - 25 kO linear-taper composition control Y1 - Surplus crystal (see text).
18-28 Chapter 18
capacitance required.
, CRYSTALS EXCEPT AS INDICATED, DECIMAL VALUES OF
A transistor can be checked by plugging
CAPACITANCE ARE IN 101 ICROFARADS ( jlF ) ; the unknown type into the panel socket
OTHERS 'ARE IN PICOFARADS I pF OR jljlFl; while using a crystal of known frequency
RESISTANCES ARE IN OHMS;
k.l 000. 101'1.000,000. and condition. Both testers can be used as
calibrators by inserting crystals for band-
edge checking. The frequencies of un-
known crystals can, be checked by
listening to the output from the test
oscillators on a' calibrated receiver or
while using a frequency counter connected
to the designated test point.
F our crystal sockets are provided in the
model shown here. Ji through J4 provide
for testing of FT-243, HC-6/U, HC-17,
and HC-25 crystals, the most popular
O-!\O holder styles in use today. Other types can
be added by the builder if desired.
~
OS'
GND + This circuit provides a simple low-cost
-=..BTI
9V method to optimize a converter or
. receiver for best noise figure. The sim-'
S3B
plicity of this system makes effective
tune-up possible without a lot of test
equipment.
Fig. 55 - Schematic diagram of the crystal tester. Capacitors are disc ceramic. Fixed-value
resistors are 1/2· or 1/4-walt composition_ . Numerous articles have described units
BT1 - Small 9·volt transistor-radio battery Q1 - Vhf npn bipolar, 2N4124, MPS3563 or where noise-figure tests may be made.
D1, D2 - 1N34A germanium diode or equiv. equiv. With the exception of certain thermal-
J1-J4. incl. - Crystal socket of builder's • S1 - Single-pole three-position phenolic rotary limited diodes (5722, for example), an
choice. wafer type, miniature.
R1 - 2 kG linear-taper composition control S2 - Part of R1. absolute value of noise figure is not
with switch. S3 - Double-pole double-throw miniature obtainable with these units; this device is
RFC1 - 2.5-mH rf choke. toggle. no exception.
Anyone using a classic noise-figure
meter soon learns that the tune-up of a
.ONrtil~_-Q
EXCEPT AS INDICATED, DECIMAL
VALUES OF CAPACITANCE ARE
system is a cut-and-try procedure where
. -=I~:f _ BT2
9V
SIB
IN MICROFARADS I.llF I ; OTHERS
ARE IN 'PICOFARA OS I pF OR )i.llF);
RESISTANCES ARE IN OHMS;
an adjustment is made and its influence is
observed by calibrating the system. Then
the excess-noise source is applied and the
k -t000. M*1,OOO,OOo.
+BTI effect evaluated. This is basically an
=-9V
;F, ALL RESISTORS 1/2 WATT CARBON after-the-fact method of testing after an
COMPOSITION TYPE
Q3 adjustment is made, and is consequently
2N4037
time consuming.
The gated noise source doesn't require a
OPTIONAL special detector or any detector at all,
CONTROL
--..,SEE TEXT other than your ear. By turning the noise
'">I source on and off at an audio rate, the
" 2000 ratio of noise contributed by the system to
/.!
noise of the system plus excess noise
appears as an audio note. The louder the
note, the greater the differential in levels
and hence the greater the influence of the
excess noise or the better the noise figure.
If greater precision is desired than
Fig. 56 - Schematic diagram of the gated noise source. subjectively listening to the signal, an
B1, B2 - 9-volt battery, Eveready 216 or Sprague BH-340. oscilloscope may be used. Hook the scope
equivalent. S1 - Double-pole, single-throw miniature vertical input to any point in the audio
C5 - O_OO1",F feedthrough capacitor, toggle.
system of the receiver, such as the speaker
terminals. Adjust the scope for a display
, of several multiples of the train of square
A Tester for Crystals and Bipolar sistorfor use at Ql (when testing crystals) pulses. Proceed by adjusting the device(s)
Transiston is the 2N4124, MPS3563 or equivalent. being tested for greatest vertical de-
All have fT ratings well into the vhf spec- flection. '
The circuit of Fig. 55 is intended trum, and each has reasonably high beta. . The result of an adjustment is instantly
primarily to test surplus crystals and The two characteristics make the devices visible as an increase or decrease in the
bipolar transistors. It uses a Pierce ideal as general-purpose oscillators. recovered audio. This method of noise
oscillator. Battery polarity can be This tester will work well from the evaluation is by no means new. Most
switched to allow testing of npn or pnp upper hf range down to at least 455 kHz. modern automatic noise-figure meters
transistors. Crystal quality is indicated.on S I is used to change the value of feedback turn the excess-noise source on and off
Ml. The greater the crystal activity, the capacitance. The lower the frequency of and then, through rather sophisticated
higher the meter reading. A suitable tran- operation, the greater the amount of' methods, evaluate the results. This tech~
Test Equipment and Measurements 16-27
nique is sometimes called "Y" "parameter cess noise in the region of 50-300 MHz.
testing. ' This unit was originally described by
While the method and circuit described Hartsen in January 1977 QST.
here are not exceptional, they represent a
fresh approach to noise evaluation. This RF Impedance Bridge for Coax Lines
approach does not require long-term The bridge show'n in Figs. 58 through
integrating detectors and tedious "twice- 60 may be used to measure unknown com-
power" measurements which, without plex impedances at frequencies below 30
Fig, 57 - A simple detector which can be MHz. Measured values are of equivalent
absolute calibration, can result in no more used when aligning ssb and 1m receivers (see
than simply optimizing the system. text for details).
series form, R + jX. The useful range of
In some cases the' available noise the instrument is from about 5 to 400
generated by this unit may be too great. ohms if the unknown load is purely
The output may be reduced by inserting the terms ofthe detection (the noise) is not resistive, or 10 to 15.0 ohms resistive com-
attenuators between the generator output coherent. ponent in the presence of reactance. The
and the device under test or by adding a The "scope" jack on most receivers is reactance range is from 0 to approximate-
2000-ohm potentiometer at the point loosely coupled to the i-f amplifier, ly 100 ohms for either inductive or
marked in the circuit (see Fig. 56). The use preceding the detector. A wide-band capacitive loads. Although the instrument
of an attenuator is preferred because it scope connected to this point will show cannot indicate impedances with the ac-
reduces the apparent output VSWR of the the train of pulses and eliminate the need curacy of a laboratory type of bridge, its
generator by increasing the return loss. If ,for aural detection. The alignment of the readings are quite adequate for the
a control is used it must be returned to its later i-f stages of a system should have the measurement and adjustment of antenna
minimum insertion-loss position when least impact on the noise performance, systems for amateur use, including the
starting a test or no signal may be heard. and maximum signal response will always taking of line lengths into account with a
Smith chart or Smith transmission-line
This circuit uses readily availab'le occur at the same setting. With this
thought in mind, the simple prototype calculator.
junk-box parts and may be easily du-
The bridge incorporates a differential
plicated. The lead placement in and detector will generally work for aural a-m
detection. Connect point A to the last i-f capacitor, C 1, to obtain an adjustable
around the diode itself should follow good
ratio for measurement of the resistive
vhf practices with 'short leads and direct amplifier plate or collector. Connect point
placement. B to the audio amplifier, at or near the component of the load. The capacitor
volume control and ground point C. With consists of two identical sections on the
this arrangement the normal detector , same frame, arranged so that when the
Theory 0/ Operation output is turned down with the volume shaft is rotated to increase the capacitance
Q I and Q2 are used in a cross-coupled control, and the temporary detector of one section, the capacitance of the
multivibrator circuit, operating at ap- provides a-m detection. other section decreases. The capacitor is
proximately 700 Hz. The value of CI is The gated noise source has been used adjusted for a null reading on MI, and its
greater than C2 to cause the duty cycle to for literally hundreds of applications and settings are calibrated in terms of
favor the conduction of Q2 slightly. When has proved to be a powerful yet simple resistance at 13 so the unknown value can
Q2 conducts, the pulse is coupled to Q3 addition to the test bench. While no guar- be read off the calibration. A coil-and-
via C3, turning on Q3 and causing current antee of duplication may be made, these capacitor combination is used to deter-
flow through R7, CRI and R8. units develop approximately 18 dB of ex- mine the amount and type of reactan~e,
The diode generates broadband noise
which is passed through R9 to the output.
R 7, C4 and C5 form a low-pass filter to
prevent high-()rder harmonics of the
switching pulses from ,appearipg in the
output.
The influence of stray rf signals entering
the device under test through the genera-
tor may be minimized by shielding the
components shown: A simple box may be
built by using pc-board scraps. For best
match, this source should be connected
directly to the input of the device under
test; therefore, the unit is equipped with a
male connector. This matching becomes a
greater consideration as the frequency of
interest increases.
Addendum
The gated noise source was first
developed in November, 1975. Sub-
sequently, some interesting things have·
been learned regarding its application.
Some contemporary receivers and trans-
ceivers cannot be operated in the a-m Fig. 58 - An RCL bridge for measuring unknown val Lies of complex impedances. A plug-in coil is
mode, and consequently the noise source used for each frequency band. The bridge operates at an rf input level of about 5 volt.S; pickup-link
seems not to operate. The detection of assemblies for use with a grid·dip oscillator are shown. Before measurements are made, the
noise is the process by which the noise bridge must be balanced witiJ a nonreactive load connected at its measurement terminals. This
load consists of a reSistor mounted inside a coaxial plug, shown in front 01 the instrument at the
source operates; therefore, it will not work left. The aluminum box measures 4-1/4 x 10-3/4 x 6-1/8 inches (108 x 273 x 156 mm) and is,
through an fm detector, nor will it work fitted with a carrying handle on the left end and self-sticking rubber feet on the right end and
through a product detector since one of bottom. Dials are Millen no. 10009 with skirts reversed and calibrations added.
18·28 Chapter 18
Fig. 59 - Schematic diagram of the im-
pedance bridge. Capacitance is in microfarads;
resistances are in ohms. Resistors are 1/2-W,
10-percent tolerance unless otherwise in-
dicated.
Cl -r Differential capacitor, 11-161 pF per
section, Millen 28801.
C2 - 17.5-327 pF with straight-line
capacitance characteristic, Hammarlund
RMC-325-S.
01,02 - Germanium diode, high back
resistance.
Jl, J3 - Coaxial connectors, chassis type.
J2 -'To mate plug of Ll, ceramic.
J4 - Phone jack, disconnecting type.
L1 - See text and Table 1. Fig. 60 - All components except"the meter are mounted on the top of the box. Cl is visable in-
Ml - 0-50,.A dc, Simpson Model 1223 Bold- side the shield at the left, with C2 at the right and ,J2 mounted between them. Jl is hidden
Vue, Cat. No. 15560 or equiv. beneath Cl in this view; a part of J3 may be seen in the lower right corner of the box. Com-
Rl - For text reference. ponents for the dc metering circuit are mounted on a tie-point strip which is affixed to the sheild
RFCl - Subminiature rf choke, Miller wall for Cl; all other components are interconnected with very short leads. The 4700-ohm input
70Fl03Al or equiv. resistor is connected across J 1.
18·30 Chapter 18
INPUT OUTPUT
J1 J2
------~-----x
3: r--~--M~--~~
fN34A
T1 PRIi-'
W 1 1N34A
C2 SPF .
C3 22 22 C4
SAMPLING UNIT
METERING UNIT
0·001 0.001
NORM SlA NORM 0 S18 Fig. 64 - This is a photograph of the sampling
;h O
PEAK PEAK ;h unit with the shield removed. The braid of the
.--------6-8~"Fl+ + 1 6.S.uF
coaxial is grounded at only one end.
1""OV;L rL1"O"V
R1
2S K
Table 3
Wattmeter Watts Watts
reading (Iow·power (high·power
scale) scale)
14 20 200
Fig. 63 - Schematic diagram of the watt/meter VSWR indicator. Parts designations called out in 20 40 400
the diagram, but not appearing in the parts list, are for text reference only. ' 25 60 600
30 80 800
C1, C2 ~ 5 pF, silver mica. RFC1, RFC2"- 4 turns no. 22 enameled wire
35 100 1000
C3, C4 - 170·780 pF trimmer capacitor. on a 3/8·in. 00 ferrite bead (950,,).
38 120 1200
Elmenco 469 or equiv. S1 - Rotary switch, two pole, two position.
41 140 1400
J1, J2 - Coaxial connectors. Builder's choice. S2 - Rotary switch, two pole, three position.
44 160 1600
M1, M2 - 0.50 "A dc meter. T1 - Primary: 'see text; secondary: 40 turns
47 180 1800
R3 - Dual 50·0 potentiometer, panel mount. no. 22 enameled wire on T80·2 core.
50 200 2000
6.ak
rh 9V
RESISTANCES ARE IN OHMS;
1\-1000. M-IOOO 000.
6,ak
,., 02 0.01
IN914
.#'
~l
PC BOARD.
16·32 Chapter 16
umt In series with the 50-o.hm resisto.r.
Again null the bridge and make a calibra-
tio.n mark fo.r 40 pF. Co.ntinue o.n in a
similar manner until that half 0. ft he dial is
co.mpletely calibrated.
MOUNT To. calibrate the o.ther half o.f the scale,
·1 the same capacito.rs may be used. This
time they must be placed tempo.rarily in
parallel with C2. Connect the 50-o.hm
resisto.r to. the UNKNOWN terminal and the
20-pF capacitor in parallel with C2. Null
the bridge and place a calibratio.n mark o.n
the panel. Remo.ve the 20-pF unit and
tempo.rarily install the 40-pF capacito.r.
Again null the bridge and make a calibra-
tio.n mark at that po.int. Co.ntinue this
pro.cedure until the capacitance dial is
co.mpletely calibrated. It sho.uld be
po.inted o.ut that the exact resistance and
Mo.UNT capacitance values used fo.r calibratio.n
can be determined by the builder. If
Fig. 69 - Parts-placement guide as viewed from the component side qf the board. resistance values o.f 20, 40, 60, 80, 100
o.hms and so. o.n are mo.re in line with the
builder's . needs, the scale may be
calibrated in tho.se terms. The same is true
fo.r the capacitance dial. The accuracy o.f
the bridge is determined by the co.m-
po.nents that are used in the calibratio.n
pro.cess.
Operation
The resistance dial is calibrated directly
in o.hms, but the capacitance dial is
calibrated ·in terms o.f pF o.r capacitance.
The + C half o.f the dial indicates that the
lo.ad is capacitive and the - C po.rtio.n is
fo.r inductive lo.ads. To. find the reactance
o.f the lo.ad, the dial ·setting must be ap-
plied to. the standard capacitive reactance
Fig. 70 - Etching pattern for the pc board. This is the pattern for the bottom side of the board. fo.rmula:
The top side of the board is a complete ground plane with a small amount of copper removed
from around the component holes. Mounting holes are located in two corners of the board. x. =2nfC -1 -
The result will be a capacitive reactance
ward and requires no special instruments. the UNKNOWN terminal and the resistance fo.r readings in the + C area o.f the dial
A receiver tuned to any portion o.f the co.ntro.l is adjusted fo.r a null. Next, the and inductive reactance fo.r - C po.rtio.ns.
15-meter band is co.nnected to. the reactance dial is adjusted fo.r a nuil and its When using the bridge remember that
RCVR terminal o.f the bridge. The po.wer is po.sitio.n is no.ted. If this setting is the instrument measures the impedance o.f'
switched o.n and a bro.adbandno.ise with a significantly different than the half- lo.ads as co.nnected at the UNKNOWN ter-
l000-Hz no.te sho.uld be heard in the meshed po.sitio.n the value o.f C2 will need minal. This means that the actual lo.ad to.
receiver. Calibratio.n of the resistance dial to. be changed. UniHo.-unit value varia- be measured must be directly at the co.n-
sho.uld be perfo.rmed first. This is ac- tio.ns o.f 120-pF capacito.rs may be suffi- necto.r rather than being attached to. the
co.mplished by inserting small co.mpo.si- cient to. pro.vide a suitable unit. Alter- bridge by a length o.f co.axial cable. Even a
tio.n resistdrs o.f appro.priate values acro.ss natively, o.ther values can be co.nnected in sho.rt .length o.f cable will transfo.rm the
the UNKNOWN co.nnecto.r o.f the bridge. series o.r parallel and tried in place o.f the lo.ad impedance to. so.me o.ther value.
The resistors sho.uld have the sho.rtest lead 120-pF capacito.r. The idea is to. have the Unless the electrical length o.f line is
lengths po.ssible in o.rder to. mate with the capacitance dial null as clo.se as po.ssible to. ~no.wn and taken into. acco.unt, it is.
co.nnecto.r. Start with 25 o.hms o.f the half-meshed po.sitio.n o.f Cl. necessary to. place the bridge at the lo.ad.
resistance (this may be made up o.f series Once the .final value o.f C2 has been An exceptio.n to. this wo.uld be if the anten-
or parallel co.nnected units). Adjust the determined and the appropriate co.mpo.- na were to. be matched to. the
capacitance and resistance dials fo.r a null nent installed in the circuit, the bridge characteristic impedance o.f the cable. In
o.f the signal as heard in the receiver. Place sho.uld be adjusted fo.r a null. The 0 this case the bridge co.ntro.ls may be preset
a calibratio.n mark o.n the fro.nt panel at capacitance po.int can be marked o.n the fo.r 50-o.hms resistance and O-pF
that lo.catio.n o.f the resistance dial. face o.f the unit. The next step is to. place a capacitance. With the bridge placed at any
Remo.ve the 25-o.hm resisto.r and insert a 20-pF capacito.r in series with the 50-o.hm po.int alo.ng the co.axial line, the lo.ad
50-o.hm resisto.r, l00-o.hm unit and so. o.n lo.ad resisto.r. Use. a go.o.d grade o.f (antenna) may be adjusted until a null is
until the dial is co.mpletely calibrated. capacito.r, such as a silver-mica type and o.btained. If the length o.f line is kno.wn to.
The capacitance dial is calibrated in a keep the leads as sho.rt as po.ssible. Null be an even multiple o.f a half-wavelength
similar manner. Initially, this dial is set so. with the capacitance dial and make a at the frequency o.f interest, the readings
that the plates o.f Cl are exactly half calibratio.n mark at that po.int. Remo.ve o.btained fro.m the bridge will be accurate.
meshed. A 50-o.hm resisto.r is co.nnected to. the 20-pF capacito.r and insert a 4O-pF Co.nsult the Transmissio.n Line chapter o.f
18·34 Chapter 18
PICK UP BOX
r --------- -- ------ -
I JI I N P U T ·
-------1I EXCEPT AS INDICATED. DECIMAL VALUES OF
®L~02~2[J . ~N~~Ei
OUTPUT CAPACITANCE ARE IN MICROFARADS (jJF I ;
________________
,@) J2 OTHERS ARE IN PICOFARADS ( pF OR jJjJFI;
RESISTANCES ARE IN OHMS;
k' 1000, ",-1000 000.
J3
EXTERNAL DETECTOR
RI5 RI6
C2 X SET
RIO
500
NOTES:
1. POTENTIOMETER HOUSING IS NOT GROUNDED
(SEE TEXT) .
ECG222
2. BOX CONSTANT TAB (2 PLACES)
3N211
40673
3N187
I
-(±) T
0.375
------~.I
t-ol-;-.---1.500
from the R and X dials. There is a little in-
1...... . . - - - - - 2 . 6 2 5 teraction between the R and X calibrating
controls, but the box constant tabs are the
only controls that greatly affect both the
(e) (2 REQU IRED) (D) (2 REQUIRED) ·resistance and reactance settings. After
•
PICKUP BOx TOP FLANGE PICKUP .BOX SIDE FLANGE the calibration is completed, the box-
MATERIAL, SAME AS (A) MATERIAL; SAME AS (A) constant tabs should be soldered in place.
Since the OIB is calibrated at to MHz, a
C- 0.062 DlA BRASS correction must be made to the reactance
FRONT PANEL
D- 0.031 DIA BRASS dial reading when operating at other fre-
,r-- quencies. All one need do is express the
operating frequency as a multiple of to
0.718 I
I
I
" MHz and multiply the dial reading by that
number.
When a length of cable is used between
the OIB and the load, the cable will act as
. .,~ an impedance transformer unless it is
0.250 SLOT) ~O...,1 perfectly matched to the load. Therefore,
TO CLEAR graphical or analytical methods must be
6-32 SCREW NYLON
WASHERS
used ·to determine the actual load im-
(6 PLACES) TAB ·pedance. Don't forget .to consider the
(E) • (2 REQUIRED) (2 PLACES) velocity factor of the cable. As the
6-32X 3/4 FLAT HD (3PLACES) operating frequency is increased, the ef-
BOX CONSTANT TAB
A·0.450 B =0.450 fect of the pickup box becomes greater.
MATERIAL: BRASS OR COPPER
One must add half the length .of the
(F) PICKUP BOX pickup box to the total length of the
LAYOUT
transmission line used between the bridge
and the load.
Fig. 74 - Mechanical drawings for the pickup box. Dimensions are in inches.(mm = in. x 25.4). When this operating impedance bridge
is used in an rf power transmission line,
high voltages may appear on the unit if
located symmetrically with respect to the tance. The instrument should stay nulled. the coax shield is broken, disconnected or
transmission lines, the instrument will The resistance dial is easier to calibrate. improperly grounded to the total
display unequal readings for inductive and First, adjust the RA potentiometer s9 that transmitting system. Antenna currents on
capacitive reactances of equal magnitude. about 85 percent of its resistance is be- the shield, caused by an unbalanced con-
A capacitive reactance other than 200 tween RI and the wiper of RA. Terminate dition at the load, are another possible
ohms can be used, but remember that the the pickup' box with a shorted connector, source of high voltage on the case. Always
capacitances of the connectors and leads and with RI set to midrange, adjust the be alert to these situations to avoid rf
must be included. A good check on the Rset control for a null indication. This null burns.
proper setting of the tabs is to obtain a indic:ation locates the zero mark for the
null, transpose the input and output con- resistance dial. Now adjust the R control Signal Generators for Receiver Testing
nections and set the Lie switch to induc- to 200 ohms (as indicated by the dial), and Here are two simple signal generators
16·36 Chapter 16
::fJ:~"O"
"""-B:~
• 20+ J1L ) •
15
10
+ 5 +
ZL=R:!:.JO + (~bf
+9V
_ 9V Fig. 78 - Reduced dial escutcheon for the 0 lB.
Send s.a.s.e. to ARRL for full·scale· template.
INPUT M1
Fig. 76 - Parts·placement guide for the null detector pc board. The component side is shown
with an X·ray view of the copper foil. R10 is mounted with short leads on the foil side.
Fig. 77 - Ful'·scale etching pattern for the null detector pc board. The foil side is shown; black
represents copper.
that can be used for receiver performance at a fixed level, it is assumed the user will
measurements. Since many receiver tests use a step attenuator to lower the output
can be carried out at a fixed frequency or level to appropriate values. Construc- Fig. 80 - Interior view of one of the signal
frequencies, two International Crystal tional information on step attenuators can generators. To the right of the oscillator
module is the 7·element Chebyshev low.pass
Mfg. Co., Inc. DE-lO oscillators make be found elsewhere in this chapter. filter network. Notice the use of the feed.
simple yet effective generators. The units , through capacitor with additional low·
described here were designed for 14.040 A Hybrid Combiner for Signal Generators frequency bypassing for the dc lead.
and 14.060 MHz and each provides an Many receiver pei-formance measure-
output power of - 5 dBm. A 7-pole ments require two signal generators to be described here provide 40 to 50 dB of
Chebyshev low-pass filter is contained in attached simultaneously to a receiver. A isolation between ports while attenuating
each oscillator assembly since the har- combiner that isolates the two signal the desired signal paths (each input to out-
monics of the oscillator alone are down generators is necessary to keep one put) by 6 dB. A second feature of these
only 10 to 30 dB up throu~h the 9th har- generator from being frequency- or phase- combiners is that of maintaining the
monic. This magnitude of harmonic modulated by the other. The combiners 50-ohm impedance of the system - very
energy will cause significant errors when
measuring receiver intermodulation per-
formance. A spectral photograph of the
output of the oscillator and 7-element Ll L2 L3 L4
low-pass filter is shown in Fig. 82.
-5dbm
The oscillator assemblies are housed in OUTPUT
chassis made from double-sided circuit-
.board material. The circuit-board panels
are soldered along each seam to construct
an "rf-tight" enclosure. This reduces the
amount of leakage from the box, which is
important when measuring the receiver 1000
noise floor. A feedthrough type of capaci- lN4001 "'F.T."
tor with additional low-frequency bypass- '--.,..r--.....--1>---<:>--o--U + 12 Vd c NOM.
::~ixI
ing is used to feed dc to the oscillator.
This is shown in the schematic diagram
and accompanying photographs. Reac,
tance values for the filter are shown in
Table 5 so that filters may be constructed
for any frequency range. The nearest Fig. 81 - Schematic diagram 0.1 the Signal generator assembly. Reactance values for the filter com·.
standard-value capacitor is sufficient. ponents are given. From this information, filters can be built for any frequency range which the user
Since the output of these generators is may need. .
L1 L2 L3 L4 C1 C2 C3
(0.1 dB ripple) 9.4 16.68 16.68 9.4 4529 5008 4529
(1 dB ripple) 17.24 2(62 24.62 17.24 252.7 266.8 252.7
Inductance values are in "H and capacitance values are in pF.
end protection, and as a calibrated at- required, along with a hybrid combiner, a
tenuator for receiver performance evalua- step attenuator and an ac voltmeter. A
tion. This attenuator uses double-pole, hybrid combiner is essentially a unit with
double-throw toggle switches to select dif- three ports. The device is used to combine
ferent amounts of attenuation. Coaxial the signals from a pair of generators. This
fittings are used at each end of the at- box has the characteristic that signals ap-
tenuator. plied at ports 1 or 2 appear at port 3, and
This unit provides 0-147 dB of attenua- are attenuated by 6 dB. However, a signal
tion in I-dB steps. Careful attention must from port 1 is attenuated 30 or 40 dB
be paid to circuit layout, with good when sampled at port 2. Similarly, signals
shielding between each of the attenuator applied at port 2 are isolated from port 1
sections being essential. A suitable some 30 to 40 dB. The isolating properties
enclosure can be made from double-sided, of the box prevent one signal generator HYBRID COMBINER
printed-circuit board material with in- from being frequency- or phase-
dividual compartments for each section. modulated by the other. A second feature Fig. 84 - Schematic diagram of the hybrid
The resistances shown in Fig. 85 are the of an hybrid combiner is that a 50-ohm combiners. For the 1 to 50 MHz model, T1 is 10
nearest standard values to those impedance level is maintained throughout turns no. 30 bifilar wound on an FT·23·72 ferrite
core. For the 50 to 500 MHz model, T1 consists
resistances apperaring in Tables' 6 and 7. the system. A commercial example of of 10 turns no. 30 bifilar wire wound on an
Although some of the values are a few anhybrid coupler of this kind is an HP- FT·23·63 ferrite core. Keep all leads as short as
ohms off, they should be more than ac- 8721A. possible when constructing these units.
II _______ -L
IL ______ i I ________ ....L II ________ .L
II ______ ---l1
I - I r7 rh r h lJ
r7
_ _ _ _ _ _ _ J_______
Fig. 85 - Schematic diagram of the 0 to 147 dB step attenuator. Resistance values for other amounts of attenuation are given in Tables 6 and 7.
51·58 are dpdt toggle switches.
16·38 Chapter 16
the output of the generator is discon-
Table 6 nected, no signal should be detected at the Table 7
PI·Network Resistive Attenuator operating frequen\=y with a sensitive T·Network Resistive Attenuator
dB Atten. R1 (Ohms) R2 (Ohms) receiver. Ideally, at least one of the signal dB Atten. R1 (Ohms) R2 (Ohms)
1 870.0 5.8 generators should be capable of amplitude 1 2.9 433.3
2 ..436.0 11.6 modulation. A suitable lab-quality piece 2 5.7 215.2
3 292.0 17.6 would be the HP-8640B. 3 8.5 132.0
4 \ 221.0 23.8 While most signal generators are 4 11.3 104.8
5 178.6 30.4 5 14.0 82.2
6 150.5 37.3 calibrated in terms of microvolts, the real 6 16.6 . ,66.9
7 130.7 44.8 concern is not with the voltage from the 7 19.0 55.8
8 116.0 52.8 generator but with the power available. 8 21.5 47.3
9 105.0. 61.6~
The fundamental unit of power is the 9 23.8 40.6
10 96.2 70.7 10 26.0 35.0
11 89.2 81.6 watt. However, the unit which is used for 11 28.0 30.6
12 83.5 93.2 most low-level rf work is the milliwatt, 12 30.0 26.8
13 78.8 106.0 and power is often specified in dB with 13 31.7 23.5
14 74.9 120.3 respect to one milliwatt (dBm). Hence, a 0 14 33.3 20.8
15 71.6 136'1 15 35.0 18.4
16 68.8 163. dBm would be one milliwatt. Tables 8 and 16 36.3 16.2
17 66.4 173.4 9 illustrate the appropriate values in dBm 17 37.6 14.4
18 64.4 195.4 for some common levels expressed in 18 38.8 12.8
19 62.6 220.0 19 40.0 11.4
microvolts. The dBm level can be W ~~ 1QO
20 61.0 247.5
21 59.7 278.2 calcldated with the aid of the following 21 41.8 9.0
22 58.6 312.7 equation 22 42.6 7.8
23 57.6 348.0 23 43.4 7.1
24 56.7 394.6 24 44.0 6.3
25 56.0 443.1 25 44.7 5.6
30 53.2 789.7 30 47.0 3.2
35 51.8 1406.1 where dBm is the power with respect to 35 48.2 1.8
40 51.0 2500.0 40 49.0 1.0
45
50
55
50.5
50.3
50.2
4442.7
7904.3
14,061.5
one milliwatt and V is the rms voltage
available at the output of the signal
generator.
~
50
55
.7
.4
49.8
Q58
Q~
0.18
60 50.1 25,000.0 The convenience of a logarithmic power 60 49.9 0,10
R2 unit like the dBm becomes apparent When RI RI
0 V\/\/' 0 signals are amplified or attenuated. For
RECEIVER
HYBRID STEP UNDER - AUDIO
COMBINER ATTENUATOR VOLTMETER
TEST
SOA
UT 1 CORRECTIONS
." tone is approximately 45 seconds. A 5OO-Hz
.........:. . .- OME6A tone is broadcast during even alternate
fOfll ADDITIONAL INFORMATION CONTACT IIEPOIITS
" •• RADIO ITATION WWV
minutes unless voice announcements or silent
2000 EAST COUNTY 110 I.
GEO ALERTS periods are scheduled. A 440-Hz tone is broad-
cast beginning one minute after the hour by
WWVH and two minutes after the hour by
WWV. The 440-Hz tone period is omitted during
the first hour of the UTC day.
• BEGINNING OF EACH HOUR IS IDENTIFIED BY
O.I·SECONO LONG. ISOO·H, TONE.
Transmitted frequencies from the two sta-
tions are accurate to % 1 part in 10. Atomic fre'
• BEGINNING OF IACH MINUTE IS IDENTIFIED IV
O.I·SECONO LONG. 1000·H, TONE. quency standards are used to maintain this ac·
• THE 21th' 51th SECOND PULSE OF EACH MINUTE IS OMITTED.
curacy. Voice announcements of the time, in
English, are given every minute. WWV utilizes
a male voice, and WWVH features a female
voice to distinguish between the two stations:
WWV time and frequency broadcasts can be
heard by telephone also. The number to cali is
303-499-7111, Boulder, CO.
All official announcements are made by
voice. Time announcements are in UTC (Univer-
WWVH BROADCAST FORMAT sal Coordinated Time). One-seconll markers
VIA TElEPHONE. IB08' 335-4383 are transmitted throughout all programs except
lNOT A TOll-FREE NUMBERI
that the 29th and 59th markers of each minute
are omitted. Detailed information on hourly
broadcast SChedules is given in the accom-
panying format chart. Cqmplete Information on
the services can be found in NBS Special
OMEGA
REPORTS
LOCATION
21· •• ·2•. 0 .. N; "'"."OO.O"W
i
AHD R ... DIATED !"OWEII
UT 1 COlIlIlcnOH,
16·42 Chapter 16
CHU DATA TRANSMISSION SEQUENCE
\1
48 49 50 60 I 27 28 29 32 33 40 41 50 60
I I I I I I I I I I I I
I I I I I I I I I I I
I I I I
fVOiCi"FiiECORi:;;;;;G\ I
~1~~~~~~A~~rt
(INCL)
300 CYCLES (0.3 SEC) DUT·I INDICATOR NRC FSK CODE
OF 1000 CYCLE TONE (SPLIT PULSES)
( EXAMPLe: ) 1000 ·CYCLES ( I SEC)
a CYCLES OF 1000 CYCLE TONE
OF 1000 CYCLE TONE
500 CYCLES 291h PULSE OMITTED
(0.5 SEC) OF
1000 CYCLE TONE
Con'stru'c.tion Practices
and Data Tables
,W hile a better job can be done with a Tapered reamer, T·handle, II2-inch
variety of tools, by taking a little care it is maximum width
will save a lot of time, they are not essen-
tial.
possible to turn out a fine piece of equip- Bench vise, 4-inch jaws or larger
ment with only a few common hand tools, Medium-weight machine oil Twist Drills
A list of tools which are indispensable in Tin shears, lO-inch size Twist drills are made of either high-
. the construction of electronic equipment Motor-driven emery wheel for grinding speed steel or carbon steel. The latter type
is found on this page, To convert English Solder, rosin core only is more common and will usually be sup-
dimensions in the list to millimeters, . Contact cleaner, liquid or spray can plied unless specific request is made for
multiply inches X 25.4, With these tools Duco cement or equivalent high-speed drills. The carbon drill will suf-
it should be possible to prepare panels and Electrical tape, vinyl plastic fice for most ordinary equipment con-
metal chassis for assembly and wiring. It Radio-supply houses, mail-order retail struction work and costs less than the
is an excellent idea for the amateur who stores and most hardware stores carry the high-speed type.
builds gear to add to his supply of tools various tools required for building or ser- While twist drills are available in a
from time to time as finances permit. vicing amateur radio equipment. While number of sizes, those listed in bold type
power tools (electric drill or dri\l press, in Table I will be most commonly used in
Recommended Tools and Materials grinding wheel, etc.) are very useful and construction of amateur equipment. It is
Long-nose pliers, 6- and, 4-inch
Diagonal cutters, 6- and 4-inch
Combination pliers, 6-inch
Screwdriver, 6- to 7-inch, 114-inch blade Table 1
Screwdriver, 4- to 5-inch, 118-inch blade Numbered Drill Size.
Phillips screwdriver, 6- to 7-inch
Phillips screwdriver, 3- to 4-inch • Drilled for Drilled for
Long-shank screwdriver with holding clip Diameter Will Clear Tapping from Diameter Will Clear Tapping from
on blade No. (MilS) Screw Steel or Brass No. (MilS) Screw Steel or Brass
Scratch awl or scriber for marking metal 1 228.0 12·24 28 140.0 8·32
Combination square, 12-inch, for layout 2 221.0 28 138.0 8·32
3 213.0 14·24 30 128.5
work 31 120.0
4 209.0 12·20
Hand drill, 114-inch chuck or larger 5 205.0 32 116.0
Soldering pencil, 30-watt, 1/8-inch tip 6 204.0 33 113.0 4-40
Soldering iron, 2oo-watt, 5/8-inch tip 7 201.0 34 111.0
8 199.0 35 110.0
Hacksaw and 12-ineh blades 196.0 38 108.5 8·32
9
Hand nibbling tool, for chassis-hole 10 193.5 37 104.0
cutting 11 181.0 10·24, 10·32 38 101.5
Hammer, ball-peen I-lb. head 12 189.0 38 088.5 3-48
13 185.0 40 098.0
Heavy-duty jackknife 14 182.0 41 096.0
File set, flat, round, half-round, and 15 180.0 42 083.5
triangular. Large and miniature types 16 177.0 12·24 43 088.0 4-40
recommended. 17 173.0 44 088.0 2·58
18 169.5 45 082.0
High-speed drill bits, no.. 60 through 3/8- 18 188.0 46 081.0
8·32 12·20
inch diameter. 20 161.0 47 078.5 3-48
Set of "Spintite" socket wrenches for hex 21 158.0 10·32 48 076.0
nuts 22 157.0 49 073.0
23 154.0 50 070.0 2·58
. Adjustable wrenches, 6- and lO-inch 24 152.0 51 067.0
Machine-screw taps, 4-40 through 10-32 25 149.5 10·24 52 063.5
thread 26 147.0 53 059.5
Socket punches, 1/2", 5/8", 3/4", 27 144.0 • 54 055.0
1-118", 1-1/4", and 1-112"
17·1 Chapter 17
Fig. 2 - View of the chassiS underside with
Fig. 1 - A compact assembly of commonly availabie items, this soldering station sanitizes the the bottom plate removed. No. 24 hookup wire
electronics assembly process. Miniature toggle switches are used because of the minimal force is adequate for all connections. Use sleeving
required to manipulate them. The force required to operate standard-size switches co.uld wherever the possibility of a short circuit ex-
destabilize the unit. ists. The diode may be installed in either direction.
usually desirable to purchase several of should be filed until smooth and bright, chased in quantities of a gross. Many of
each of the commonly used sizes rather and then tinned immediately by dipping it the radio-supply stores sell small quan-
than a standard set, most of which will be in solder. Most modern soldering iron tips tities and assortments that come in handy.
used infrequently, if at all. are iron-clad and cannot be filed.
Although Table I lists drills down to A Deluxe Soldering Station
no. 54, the series extends to no. 80. No. 68 Useful Materials The simple device shown on this page
and no. 70 are useful for drilling printed- Small stocks of various misceJlaneous can enhance the versatility of longevity of
circuit boards for component leads. materials will be required in constructing a soldering iron as'well as make electronic
radio apparatus. Most of these are assembly more convenient. Fig. I depicts
Care 0/ Tools available from hardware or radio-supply the obvious convenience features - a pro-
The proper care of tools is not only a stores. A representative list follows: tective heat sink. and cage, and a tip-
matter of pride to a good worker. He also Sheet aluminum, solid and perforated, cleaning sponge rigidly attached to a
recognizes the energy saved and the an- . 16 or 18 gauge, for brackets and shielding. sturdy base for efficient one-handed
noyance avoided by possessing a full kit of 112 X II2-inch (12 X 13-mm) alumi- operation. Inside the chassis are some
well-kept, sharp-edged tools. num angle stock. electrical refinements that justify the
Qrills should be sharpened at frequent 1/4-inch (6~mm) diameter round brass sophisticated name "soldering station."
intervals so that grinding is kept at a or aluminum rod for shaft extensions. Soldering iron tips and heating elements
•
minimum each time. This makes it easier Machine screws: Round-head and flat last longer if operated at lower-than-
to maintain the rather critical surface head, with nuts to fit. Most useful sizes: maximum temperature when idling. Many
angles required for best cutting with least 4-40, 6-32 and 8-32, in lengths from solder connections can be made satisfac-
wear. Occasional oilstoning of the cutting 114-inch (6-mm) to 1-112 inches (38 mm). torily with reduced heat, and some smaJl
edges of a drill or reamer will extend the (Nickel-plated iron will be found satisfac- semiconductor devices require lower
time between grindings. tory except in strong rf fields, where brass temperatures to avoid junction damage.
The soldering iron can be kept in good should be used.) In the unit described here temperature
condition by keeping the tip well tinned Bakelite, Lucite, polystyrene and reduction is accomplished by halving the
with sold~r and not allowing it to run at copper-clad pc-board scraps. duty cycle of the applied ac voltage, DI in
full voltage for long periods when it is not Soldering lugs, panel bearings, rubber Fig. 3 conducts only when the "hot" ac
being used. After each period of use, the grommets, terminal-lug wiring strips, line is positive with respect to neutral. If
tip should be removed and cleaned of any varnished-cambric insulating tubing, heat- the diode were reversed, the soldering iron
scale which may have accumulated. An shrinkable tubing. would be heated only on the negative half
oxidized tip may be cleaned by dipping it '. Shielded and unshielded wire. cycles, but the result would be the same.
in sal ammoniac (ammonium chloride) Tinned bare wire, nos. 22, 14 and 12. (This is one of the rare applications of rec-
While hot and then wiping it clean with a Machine screws, nuts, washers, solder- tifierdiodes where the polarity is not im-
rag. If a copper tip becomes pitted it ing lugs, etc., are most reasonably pur- portant.) With current flowing oniy in one
~
GREEN and move .them about until a satisfactory
AC
arrangement has been found, keeping in
mind any parts which are to be mounted
NE2 underneath, so interference in mounting
can be avoided. Place capacitors and
WHITE
WHITE "other parts with shafts extending through
the panel first, and arrange them so that
the controls will form the desired pattern
of the panel. Be sure to line up the shafts
Fig. 3 - Schematic diagram of the soldering station. 01 is a silicon power rectifier, 1·A, 400·PRV. squarely with the chassis front. Locate
S1 and S2 are miniature spst toggle switches rated 3 A at 125 V. This circuit is satisfactory for
use with irons having power ratings up to 100 W. any parti~ion shields and panel brackets
next, and "then the tube sockets and any
other parts, marking the mounting-hole
direction, only one electrode of the neon into the electrical components below the centers of each accurately on the paper.
bulb will appear to glow. Closing SI chassis. RTV compound was used for this Watch out for capacitors whose shafts are
short-circuits the diode and applies full purpose in the unit pictured. off center and do not line up with the
power to the soldering iron, igniting both Purchase a separate 3-wire cord for the mounting holes. Do not forget to mark
bulb electrodes brightly. power input. Merely splicing the soldering the centers of socket holes and holes for
CMOS ICs are prone to 'damage by station into the existing soldering iron wiring leads. The small holes for 'socket-
static charges, so they should ~e soldered cord will shorten the operating radius of mounting screws are best located and
with an iron having a grounded tip. This the iron and make it awkward to use. center-punched, using the socket itself as
requirement is fulfilled by most irons Heyco bushings were used to anchor both a template, after the main center hole has
having 3-wire power cords. Unfortu- cords" in the unit described, but if these been cut.
nately, a grounded tip precludes using the aren't available, grommets and cable By means of the square, lines indicating
iron on a live circuit. If the potentials are clamps will. work as well. Knotting the "accurately the centers of shafts should be
low (less than about 2.5 volts) and the cords inside the chassis is a simple expe- extended to the chassis front !lnd marked
operator is thoroughly familiar with the dient that sometimes provides adequate on the panel at the chassis line, the panel
circuit, a soldering iron may facilitate ex- strain relief. being fastened on temporarily. The hole
perimentation or trouble-shooting. This The underchassis assembly is shown in centers may then be punched in the chassis
technique should be used only when com- Fig. 2. The neon bulb is forced through a with the center punch. After drilling, the
plete safety is assured. The simplest way 3/16-inch (5-mm) ID grommet. The leads parts which requir.e mounting underneath
to increase the usefulness of a 3-wire are sleeved to prevent short circuits. If may be located and the mounting holes
soldering iron is to install a switch in the you mount the bulb in a fixture or socket, drilled, making sure by trial that no in-
ground lead. S2 iiI Fig. 3 serves this pur- use a clear lens to ensure that the elec- terferences exist with parts mounted on
pose. Before clipping the cord on your trodes are distinctly visible. Fit a cover to top. Mounting holes along the front edge
soldering iron, be certain that the tip is the bottom of the chassis to prevent ac- of the chassis should be transferred to the
common to the ground prong and isolated cidental contact with the live ac wiring. panel by once again fastening the panel to
from the other ac prongs. Stick-on rubber feet will ensure a skid-free the chassis and marking it from the rear.
The base for the unit is a 2 X 6 X unit that won't mar your work surface. Next, mount on the chassis the
4-inch (51 X 152 X 102 mm) (HWD) The total cost of this project with all capacitors and any other parts with shafts
aluminum chassis (Bud AC-43 I or new parts (including sardines) was $28. extending to the panel, and measure ac-
equivalent). An Ungar model 8000 solder- One could trim that figure considerably curately the height of the center of each
ing iron holder fits neatly on the chassis with even a modest junk box. The solder- shaft above the chassis, as illustrated in
top. The holder has two mounting holes in ing iron, an Ungar model 127, represents Fig. 4. The horizontal displacement of
each foot. A sponge tray nests between half of the investment.
the feet and the cage. In this model a sar-
dine tin is used for the sponge tray, Chassis Working"
although a suitable watertight enclosure With a ,few essential tools and proper
can also be fabricated from strips of procedure, building radio gear on a metal
copper-clad circuit-board material. The chassis is a relatively simple matter.
tray and iron holder are secured to the Aluminum is preferred to steel, not only
chassis by 6-32 X· II2-inch pan head because it is a superior shielding material,
machine screws and nuts, with flat but because it is much easier to work and
washers under the screw heads (sponge provides good chassis contacts.
o
tray) and lock washers under the nuts The placement of components on the CHASSIS
. (chassis underside). One of these nuts chassis is shown quite clearly in the
fastens a 6-lug tie point strip to the chassis photographs in this Handbook. Aside
bottom. Use the soldering iron holder from certain essential dimensions, which
base as a template for drilling the chassis usually are given in the text, exact duplica-
and sponge tray. The floor of the sponge tion is not necessary. Fig. 4 - Method of measuring the heights of
capacitor shafts. If the square is adjustable. the
tray must be sealed around the screw " Much trouble and energy can be saved end of the scale should be set flush with the face
heads to prevent moisture from leaking by spending sufficient time in planning of the head.
17-3 Chapter 17
(A) (B) (e) (0)
HACKSAW BLADE
1116" STOCK
(1.6mm)
I
I
I
I
tID
"_ ... ' I
I"
COMPLETED
UNIT
(AI
Fig. 6 - D~talls for forming channel·type heat Sinks.
r---
shaded portion of B, making it possible to start THEN FILE UNTIL Ql
the hacksaw blade along the cutting line. A HAS SNUG FIT
shows how a single-ended handle may be 1-3/4" (44mm)
constructed for a hacksaw blade.
I/B"'3mm)
ALUM. ANGLE STOCK
shafts having already been marked on the
chassis line on the panel, the vertical
displacement can be measured from this
"-- y
line, The shaft centers may now be marked
on the back of the panel, and the holes
drilled, Holes for any other panel equip-
ment coming above the chassis line may
then be marked and drilled, and the re-
mainder of the apparatus mounted. Holes 16 - GAUGE BRASS
OR ALUMINUM STOCK
for terminals and other parts of the rear 5/16"DIA. /
edge of the chassis should be marked and (7.9mm) /
drilled at the same time that they are done BOLT TO ALUM. ANGLE HEAT SINK DETAILS
WITH 4 - 40 HARDWARE
for the top.
Drilling and Cutting Holes Fig. 7 - Layout and assembly details of another homemade heat sink. The completed assembly
When drilling holes in metal with a can be insulated from the main chassis of the transmitter by using insulating washers.
hand drill it is important that the centers
first be located with a center punch, so
that the drill point will not "walk" away
- .
turned to squeeze the punch through the Semiconductor Heat Sinks
from the center when starting the hole. chassis. The threads of the bolt should be Homemade heat sinks can be fashioned
When the drill starts to break through, oiled occasionally. from brass, copper or aluminum stock by
special care must be used. Often it is an Large holes in steel panels or chassis are employing ordinary workshop tools. The
advantage to shift a two-speed drill to low best cut with an adjustable circle cutter. dimensions of the heat sink will depend
gear at this point. Holes more than Occasional application of machine oil in upon the type of transistor used and the
1/4-inch (6-mm) in diameter should be the cutting groove will help. The cutter amount of heat that must be conducted
started with a smaller drill and reamed out first should be tried out on a block of away from the body of the semiconduc-
with the larger drill. wood, to make sure that it is set for th'e tor.
The chuck on the usual type of hand right diameter. Fig. 6 shows the order of progression
drill is limited to I14-inch (6-mm) drills. The burrs or rough edges which usually for forming a large heat sink from
The I14-inch (6-mm) hole may be filed result after drilling or cutting holes may be aluminum or brass channels of near-equal
out to larger diameters with round files. removed with a file, or sometimes more height and depth. The width is lessened in
Another method possible with limited conveniently with a sharp knife or chisel. parts Band C so that each channel will fit
tools is to drill a series of small holes with It is a good idea to keep an old wood into the preceding one as shown in the
the hand drill along the inside of the cir- chisel sharpened and available for this completed model at D. The three pieces
cumference of the large hole, placing the purpose. are bolted together with 8-32 screws and
holes as close together as possible. The nuts. Dimensions given are for illustrative
Rectangular Holes
center may then be knocked out with a purposes only.
cold chisel and the edges smoothed with a Square or rectangular holes may be cut Heat sinks for smaller transistors can be
file. Taper reamers which fit into the out by making a row of small holes as fabricated as shdwn in Fig. 8. Select a drill
carpenter's brace will make the job easier. previously described, but is more easily bit that is one size smaller than the
A large rat-tail file clamped in the brace done by drilling a II2-inch (l3-mm) hole diameter of the transistor case and form
makes a very good reamer for holes up to inside each corner, as illustrated in Fig. 5, the heat sink from 1!16-inch (1.6-mm)
the diameter of the file. and using these holes for starting and thick brass, copper or aluminum stock as
For socket holes and other large holes turning the hacksaw. The socket-hole shown in steps A, Band C. Form the
in an aluminum chassis, socket-hole punch and the square punches which are stock around the drill bit by compressing
punches should be used. They require first now available also may be of considerable it in a vise (A). The completed heat sink is
drilling a guide hole to pass the bolt that is assistance in cutting
.
out large openings.. press-fitted over the body of the semicon-
Con.tructlon Practlc•• and Data Tabl.. 17-4
here. In the example given in Fig. 6, the
HEAT SINK
grease should be applied between the three
channels before they are bolted together,
DRILL BIT~ as well as between the transistor and the
channel it contacts.
Construction Notes
If a control shaft must be extended or
insulated, a flexible shaft coupling with
(B) adequate insulation should be used.
Satisfactory support for the shaft exten-
sion, as well as electrical contact for
safety, can be provided by means of a
metal panel bearing made for the purpose.
These can be obtained singly for use with
existing shafts, or they can be bought with
HEAT SINK
a captive extension shaft included. In
1I~~~1 ___ - INSTALLED either case the panel bearing gives a
ON TRANSISTOR
"solid" feel to the control. The use of
~l
fiber washers between ceramic insulation
"., , )
COMPLETED HEAT
SINK -----UU (C
(0)
and metal brackets, screws or nuts will
prevent'the ceramic parts from breaking.
Cutting and Bending Sheet Metal
If a metal sheet is too large to be cut
Fig. 8 - Steps used in constructing heat sinks for small transistors. conveniently with a hacksaw, it may be
marked with scratches as deep as possible
along the line of. the cut on both sides of
the sheet, and then clamped in a vise and
Table 2 worked back and forth until the sheet
breaks at the line. Do not carry the
Standard Metal Gauges
bending too far until the break begins to
weaken; otherwise the edge of the sheet
Gauge American U.S. Birmingham Gauge American U.S. Birmingham
No. or BS' Standard' or Stubs' No. or BS' Standard' or Stubs' may become bent. A pair of iron bars or
pieces of heavy angle stock, as long or
1 0.2893 0.28125 0.300 24 0.02010 0.025 0.022
2 0.2576 0.265625 0.284 25 0.01790 0.021875 0.020 longer than the width of the sheet, to hold
3 0.229" 0.25 0.259 26 0.01594 0:01875 0,018 it in the vise, will make the job easier. C
4 0.2043 0.234375 0.238 27 0.01420 0.0171875 0,016 clamps may be used to keep the bars from
5 0.1819 0.21875 0.220 28 0.01264 0.015625 0.014 spreading at the ends. The rough edges
6 0.1620 0.203125 0.203 29 0.01126 0.0140625 0.013
7 0.1443 0.1875 0.180 30 0.01003 0.0125 0.012 may be smoothed with a file or by placing
8. 0.1285 0.171875 0.165 31 0.008928 0.0109375 0.010 a large piece of emery cloth or sandpaper
9 0.1144 0.15625 0.148 32 0.007950 0.01015625 0.009 on a flat surface and running the edge of
10 0.1019 0.140625 0.134 33 0.007080 0.009375 0.008 the metal back and forth over the sheet.
11 0.09074 0.125 0.120 34 0.006350 0.00859375 0.007
12 0.08081 0.109375 0.109 35 0.005615 0.0078125 0.005
Bends may be made similarly.
13 0.07196 0.09375 0.095 36 0.005000 0.00703125 0.004 Today much of the tedium of sheet
14 0.06408 0.078125 0.083 37 0.004453 0.006640626 metal work can be relieved by using
15 0.05707 0.0703125 0.072 38 0.003965 0.00625 copper-clad printed-circuit board material
16 0.05082 0.0625 0.065 39 0.003531
17 0.04526 0.05625 0.058 40 0.003145 wherever possible. Copper-clad stock is
18 0.04030 0.05 0.049 manufactured with phenolic, 0-10
19 0:03589 0.04375 0.042 'Used for aluminum, copper, brass and nonfer· fiberglass and Teflon base materials in
20 0.03196 0.0375 0.035 rous alloy sheets, wire and rods.
'Used for iron, steel, nickel and ferrous alloy thicknesses up to 118 inch (3 mm). While
21 0.02846 0.034375 0.032
22 0.02535 0.03125 0.028 sheets, wire and rods. it is manufactured in large sheets for in-
23 0.02257 0.028125 0.025 'Used for seamless tubes;. also by some dustrial use, some hobby electronics
manufacturers for copper and brass. stores and surplus outlets market usable
scraps at reasonable prices. Pc-board
stock is easily cut with a small hacksaw.
ductor as illustrated at D. The larger the most hardware stores. A hole is bored in Because the nonmetallic base material
heat sink area, the greater will be the the angle stock to allow the transistor case isn't malleable, it can't be bent in the
amount of heat conducted away from the to fit snugly into it. The transistor is held usual way. However, corners are easily
transistor body. In some applications, the in place by a small metal plate whose formed by holding two pieces at right
heat sinks shown in Fig. 8 may be two or center hole is slightly smaller in diameter angles and soldering the seam. Excellent
three inches in height (power transistor than the case of the transistor. Details are rf-tight enclosures can be fabricated in
stages) . given in Fig. 7. this manner. Many projects in this Hand-
. Another technique for making heat A thin coating of siliGone grease, book were constructed using this tech-
sinks for TO-5 type transistors and larger available from most electronics supply nique. If mechanical rigidity is required of
models is shown in Fig. 7. This style of houses, can be applied between the case of a hlrge copper-clad surface, stiffening ribs
heat sink will dissipate considerably more the transistor and the part of the heat sink may be soldered at right angles to the
heat than will the type shown in Fig. 8. ~ith which it comes in contact. The sheet.
The main body of the sink is fashioned silicone grease will aid the transfer of heat
from a piece of 118-inch (3-mm) thick from the transistor to the sink. This prac- Finishing Aluminum
aluminum angle bracket - available from tice can be applied to all models shown Aluminum chassis, panels and parts
17-5 Chapter 17
may be given a sheen finish by treating soldering to the pins of coil forms or male should be made as inaccessible as possible
them in a caustic bath. An enameled or cable plugs. It helps if the pins are first to accidental contact or short-Circuit.
plastic container, such as a dishpan or in- cleaned on the inside with a suitable twist Where shielded wire is called for and
fant's bathtub, should be used for the drill and then tinned by flowing rosin-core capacitance to grolind is 'not a, factor,
solution. Dissolve ordinary household lye solder into them. Immediately clear the Belden type 8885 shielded grid wire may
in cold water in a proportion of one- surplus solder from each hot pin by a be used. If capacitance must be mini-
quarter to one-half can of lye per gallon whipping motion or by blowing through mized, it may be necessary to use a piece
of water. The stronger solution will do the the pin from the inside of the form or of car-radio low-capacitance lead-in wire
job more rapidly. Stir the solution with a plug. Before inserting the wire in the pin, or coaxial cable.
stick of wood until the lye crystals are file the nickel plate from the tip. After For wiring high-frequency circuits, rigid
completely dissolved. Be very careful to soldering, round the solder tip off with a wire is often used. Bare soft-drawn tinned
avoid skin contact with the solution. It is file. wire, size 22 to 12 (depending on
also harmful to clothing. Sufficient solu- When soldering to the pins of polysty- mechanical requirements) is suitable.
tion should be prepared to cover the piece ren'e coil forms, hold the pin to be Kinks can be removed by stretching a
completely. When the aluminum is im- soldered with a pair of heavy pliers to piece of 10 or 15 feet (3 or 4.5 m) long and
mersed, a very pronounced bubbling takes form a "heat sink" and insure that the pin then cutting it into short lengths that can
place and ventilation should be provided does not heat enough in the coil form to be handled conveniently. Rf wiring should
to disperse the escaping gas. A half hour loosen and become misaligned. be run directly from point to point with a
to two hours in the solution should be suf- Some connections carrying very high minimum of sharp bends and the wire
ficient, depending upon the strength of current can't be made with otdinary tin- kept well spaced from the chassis or other
the solution and the desired surface. lead solder because the heat generated by grounded metal surfaces. Where the
, Remove the aluminum from the solu- the joint resistance would melt the solder. wiring must pass through the chassis or a
tion with sticks and rinse thoroughly in Automotive starter brushes and uhf partition, a clearance hole should be cut
cold water while swabbing with a rag to transmitter tank circuits are two cases in and lined with a rubber grommet. In case
remove the black deposit. When dry, which this situation can occur. Silver insulation becomes necessary, varnished
finish by spraying on a light coat of clear solder prevents this condition in two cambric tubing (spaghetti) can be sUpped
lacquer. ways: It melts at a significantly higher over the wire.
Raw aluminum can be prepared for temperature than tin-lead solder (about In transmitters where the peak voltage
painting by abrading the surface with 600 0 F or 315 0 C) and generates less heat does not exceed 2500, the shielded grid
medium-grade sandpaper, making certain because of its superior conductivity. A wire mentioned above should be satisfac-
the strokes are applied in the same direc- propane torch may be necessary for large tory for power circuits. For higher
tion (not circular or random). This pro- silver soldering jobs. The special flux used vQltages, Belden type 8656, Birnbach type
cess will create tiny grooves on the other- with silver solder releases toxic fumes, so 1820, or shielded ignition cable can be
wise smooth surface. As a result, paint or follow the manufacturer's instructions used. In the case of filament circuits carry-
lacquer will adhere well. Before painting, carefully and work only in a well- ing heavy current, it may be necessary to
wash the abraded aluminum with soap ventilated area. use no. 10 or 12 bare or enameled wire,
and hot water, dry thoroughly: Avoid slipped through spaghetti, and then
touching the prepared surface before Wiring covered with copper braid pulled tightly
painting it. The wire used in connecting amateur over the spaghetti. If the shielding is
equipment should be selected by con- simply slid back over the insulation and
Soldering sidering both the maximum current it will solder flowed into the end of the braid,
The secret of good soldering is to use be called upon to handle and the voltage the braid usually will stay in place without
the right amount of heat. Too little heat its insulation must stand without break- the necessity for cutting it back or binding
will produce a "cold-soldered joint"; too' down. Also, from the consideration of it in place. The braid should be cleaned
much may injure a component. The iron TVI, the power wiring of all transmitters first so that solder will take with a
and the solder should be applied simulta- should be done with wire that has a braid- minimum of heat. Rf wiring in transmit-
neously to the joint. Keep the iron clean ed shielding cover. Receiver and audio cir- ters usually follows the method described
by brushing the hot tip with a paper towel cuits may also require the use of shielded above for receivers, with due respect to
or a moist sponge, as illustrated in the wire at some points for stability or the the voltages involved.
soldering station described earlier in this elimination of hum.
chapter. Always use rosin-core solder;' No. 20 stranded wire is commonly used
never acid-core. Solders have different for most receiver wiring (except for the
melting points, depending upon the ratio high-frequency circuits) where the current
of tin to lead. A 50-50 solder melts at 425 0 does not exceed 2 or 3 amperes. For
F (218 0 C), while 60-40 melts at 371 0 F higher-current heater circuits, no. 18 is
(188 0 C). When it is desirable to protect available. Wire with cellulose acetate in- (A) WRONG
from excessive heat the components being sulation is good for voltages up to about
soldered, the 60-40 solder is preferable to 500. For higher voltages, Teflon-insulated II
the 50-50. (A less-common solder, 63-37, or ot)1er special HV wire should be used.
melts at 361 0 For 182 0 C.) Inexpensive wire strippers that make the
Ii
When sold,ering transistors, crystal removal of insulation from hookup wire (B) RIGHT
diodes or small resistors, the lead should an easy job are available on the market.
be gripped with a pair of pliers up close to When power leads have several
the unit so that the heat will be conducted branches in the chassis, it is convenient to
away. Overheating of a transistor or diode use fiber-in'sulated mUltiple tie points as (e) RIGHT
while soldering Can cause permanent anchorages or junction points. Strips of
damage. Also, mechanical stress will have this type are also useful as insulated sup-
a similar effect. Therefore, a small unit ports for resistors, rf chokes and Fig. 9 - Meth<>ds of lacing cables. The
should be,mounted so that there is no ap- capacitors. Exposed points of high- method shown at C is more secure, but takes
more time than the method of B. The latter is
preciable mechanical strain on the leads. voltage wiring should be held to a usually adequate for most amateur reo
Trouble is sometimes experienced in minimum; those which cannot be avoided qUirements.
~
that they are parallel to the panel or sides
of the chassis.
~TAND
PUSH - PULL AMP.
Winding Coils (Al (BI'
WRONG
TAhO;;- SCRAPE ,
Close-wound coils are readily wound on (el (Ol
the specified form by anchoring one end
of the length of wire (in a vise or to a
Fig. 12 - The suggested winding method for a Fig. 13 - The view at A shows how the turns
doorknob) and the other end to the coil single·layer toroid as shown at A. A \30' gap is on a toroid should be counted. The large black
form. Straighten .any kinks in the wire and recommended (see text). Wrong methods are dots in the diagram at B are used to indicate
then pull to keep the wire under slight ten- shown at Band C. At 0 is a method for placing the polarity of the windings (phasing).
sion. Wind the coil to the required a tap on the coil.
number of turns while walking toward the
anchor, always maintaining a slight ten-
sion on the wire.
To 'space-wind the coil, wind the coil space-wound coil by judicious applica- theoretical and practical aspects of
simultaneously with a suitable spacing tions of Duco cement to hold the turns in toroids. Figs. 12 and 13 illustrate the
medium (heavy thread, string or wire) in place. proper way to wind and count the turns
the manner described above. When the The "cold" end of a coil is the end at or on a toroidal core.
winding is complete, secure the end of the close to chassis or ground potential. A bifilar winding is one which has two
coil to the coil-form terminal and then Coupling links should be wound on the identical lengths of wire, which when
carefully unwind'the spacing material. If cold end of a coil to minimize capacitive placed on the core result in the same
the coil is wound under suitable tension, coupling. number of turns for each wire. The two·
the spacing material can be easily removed Rf chokes must often present a high im- wires can be put on the core side by side at
without disturbing the winding. Finish the, pedance over a broad frequency range. the same time, just as if Sl single winding
This requirement calls for the avoidance were being applied. An easier and more
of series resonances within the range. popular method is to twist the two wires (8
Such resonances can be avoided in single- to. 15 times per inch or 1-112 to 3 mm per
layer solenoids by separating the winding "twist" will suffice), then wind the pair
into progressively shorter sections. A on the core. The wires can be twisted han-
practical choke suitable for hf amateur dily by placing one end of the length of
service at plate impedances up to 5 kO and two wires in a bench vise. The remaining
currents up to 600 mA is shown in Fig. 1Q. wire-pair ends are tightened into the
Another way to build. a broad-band chuck of a small hand drill,' and the
choke is to wind a small number of turns twisting is done.
on a high-permeability ferrite rod (such as A (riji/ar winding has three wires and a
SPACES
BETWEEN
used for antennas in some portable quadriji/ar winding has four. The pro-
WINDINGS
ARE 1/4 "(6 mml
radios). The magnetic core supplies a large cedure for preparation and winding is
inductance with a small winding. Keeping otherwise the same as for a bifilar
B"
(200mml
the number of turns small reduces the winding. Fig. 14 shows a bifilar type of
distributed capacitance and raises the self-
resonant frequency of the choke. Ferrite
chokes are best suited to low-impedance
::::
~
.
o c
applications. A bifilar (this term is ex-
FILTER
CAPACITOR
BRACKET
plained in the following paragraphs) fila-
ment choke for grounded-grid kilowatt hf
amplifiers appears in Fig. 11.
Toroidal inductors and transformers
RED '.
d
'RED~'3 d~RN~~~:::.
BIFILAR GRN
are specified for many projects in this BIFILAR
Handbook. The advantages of this type of (Al (Bl
Fig. 10 - Pictorial diagram of a heavy·duty ·winding include compactness and a self-
plate choke for the hf bands. No. 26 enameled
wire is used for the windings. Low·loss shielding property. June 1979 QST con- Fig. 14 - Schematic and pictorial presentation
material should be used for the form. tains a comprehensive treatment of the of a bifilar·wound toroidal transformer.
17-7 Chapter 17
toroid in schematic and pictorial form. paper. The job can be made a bit easier by
The wires have been twisted together prior tracing over the original layout with a
to placing them on the core. It is helpful, ballpoint pen and carbon paper w,hile the
though by no means essential, to use wires pattern is taped to the copper side of the
of different color when multi filar-winding unetched circuit board. The carbon paper
a core. The more wires used, the more is placed between the pattern and the cir-
perplexing it ·is to identify the end of the cuit board: After the paint has been ap-
windings correctly once the core has been plied, it should be allowed to dry for at
wound. There are various colors of least 24 hours prior to the etching process.
enamel insulation available, but it is not The Vector Company produces a rub-on
easy for amateurs to find this wire locally transfer material that can also be used as
or in small-quantity lots. This problem etch-resist when laying Qut circuit-board
can be solved by taking lengths of wire patterns. Thin strips of ordinary masking
(enameled magnet wire), cleaning them to tape, cut to size and firmly applied, serve
remove dirt and grease, then spray- nicely as etch-resist material too.
painting them. Ordinary aerosol-can When making "one-shot" pc boards it
spray enamel works fine. Spray lacquer is is convenient to cover the copper surface
not as satisfactory because 'it is brittle with masking tape, transfer the circuit
when dry and tends to flake off the wire. pattern by means of carbon paper, then
The winding sense of a multifilar cut out and remove th'e sections of mask-
toroidal transformer is important in most ing tape where the copper is to be etched
circuits. Fig. 13B illustrates this principle. away. An X-acto hobby knife is excellent Fig. 15 - A homemade stand for processing
The black dots (called phasing dots) at the for the purpose. Maskirig tape, securely etched·circuit boards. The heat lamp maintains
top of the T1 windings indicate polarity. applied, serves as a fine etch-resist the etchant-bath temperature between 90 and
That is, points a and c are both start or material. 115 0 F (32 and 46 0 C) and is mounted on an
adjustable arm. The tray for the bath is raised
finish ends of their respective windings. In Many magazine articles feature printed- and lowered at one end by the action of a
this example, points a and d are of op- circuit layouts. The more-complex pat- motor-driven eccentric disc, providing the
posite phase (180 0 phase difference) to terns (those containing ICs and high com- necessary agitation of the chemical solution. A
provide push-pull voltage feed to QI and Q2. ponent densities) are difficult to duplicate darkroom thermometer monitors the
temperature of the bath.
" accurately by hand. A photographic pro-
Circuit-Board Fabrication cess is the most efficient way to transfer a
Modern-day builders prefer the layout from a magazine page to a circuit
neatness and miniaturization made possi- board. A Thermofax transparency- quired with this bath. Ready-made solu-
ble by the use of etched or printed circuit producing machine (most schools have tions (one-pint and one-gallon sizes) are
boards. There are additional benefits to these) will copy the circuit on a clear available through some mail-order houses
be realized from the use of circuit boards: plastic sheet for use as a negative. Pressing at low cost. They are manufactured by
Low lead inductances, excellent physical this negative· against a photosensitive Kepro Company and carry stock numbers
stability of the components and intercon- copper-clad board with a piece of glass E-IPT and E-IG, respectively.
necting leads, and good repeatability of arid exposing the assembly to sunlight for Etchant solutions become exhausted
the basic layout of a given project. The about 90 seconds will deactivate the after a certain amount of copper has been
latter attribute makes the use of circuit etchant resist on the exposed part of the processed. Therefore, it is wise to keep a
boards ideal for group projects. board. The portion of the copper that is quantity of the bath on hand.if frequent
shielded from the light by the negative will use is anticipated_ With either chemical
Planning and Layout resist etching. This process is described in bath, the working solution should be
The constructor should first plan the detail by Taylor, W4POS, in August 1979 maintained at a temperature between 90
physical layout of the circuit by sketching QST. Photosensitive pc-board material is and 115 0 F (32 and 46 0 C). A heat lamp
a pictorial diagram on paper, drawing it to manufactured by Kepro Company. can be directed toward the bath during the
scale. Once this has been done, the inter- etching period, its distance set to maintain
connecting leads can be inked in to repre- The Etching Process the required temperature. A darkroom
sent the copper strips that will remain on Almost any strong acid bath will serve thermometer is handy for monitoring the
the etched board. Toe Vector Company as an etehant, but the two chemical temperature of the bath.
sells layout paper for this purpose. It is preparations recommended here are the While the circuit board is immersed in
marked with the same patterns that are safest to use. A bath can be prepared by the solution, it should be agitated con-
used on their perforated boards. mixing one part ammonium persulphate tinuously to permit uniform reaction to
After the basic etched-circuit design has crystals with twO' parts clear water. A nor- the chemicals. This action will also speed
been completed the designer should go mal quantity of working solution for most up the etching process somewhat. Nor-
over the proposed layout several times to amateur radio applications is composed of mally, the circuit board should be placed
insure against errors. When the foregoing one cup of crystals and two cups of water. in the bath with the copper side facing
has been done, the pattern can be painted To this mixture add 114 teaspoon of mer- down, toward the bottom of the tray. The
on the copper surface of the board to be curic chloride crystals. The latter serves as tray should be non-metallic preferably a
etched. Etch-resistant solutions are an activator for the bath. Ready-made Pyrex dish or a photographic darkroom
available from commercial suppliers and etchant kits which use these chemicals are tray.
can be selected from their catalogs. Some available from Vector. Complete kits The photograph, Fig. 15,shows a
builders prefer to use India ink for this which contain circuit boards, etchant homemade etching stand made up from a
purpose. Perhaps the most readily powders, etch-resist transfers, layout heat lamp, some lumber, and afl 8-rpm
available material for use in etch-resist ap- paper, and plastic etchant bags are also motor. An eccentric disc has been
plications is ordinary exterior enamel available from Vector at moderate prices. mounted on the motor shaft and butts
paint. The portions of the board to be re- Another chemical bath that works against the bottom of the etchant tray. As
tained are covered with a layer of paint, satisfactorily for copper etching is made the motor turns, the eccentric disc raises
applied with an artist's brush, duplicating up from one part ferric chloride crystals and lowers one end of the tray, thus pro-
the pattern that was drawn on the layout and two parts water. No activator is re- viding continuous agitation of the
Construction Practlc.. and Data Tabl.s 17-8
solution. The heat lamp is mounted on an
adjustable, slotted wooden arm. Its height Table 4
above the solution tray is .adjusted to pro- Resistor·Capacltor Color Code
vide the desired bath temperature.
Because the etching process takes between Color Significant Figure Decimal Multiplier Tolerance (%J Voltage Rating·
15 minutes and one hour - dependent
upon the strength and temperature of the
Black o 1
Brown 1 10 1· 100
bath - such an accessory is convenient. Red 2 100 2· 200
After the etching process is comple~ed, Orange 3 1,000 3· 300
the board is removed from the tray and Yellow 4 10,000 4· 400
Green 5 100,000 5· 500
washed thoroughly with fresh, clear Blue 6 1,000,000 6· 600
water. The etch-resist material can then be Violet 7 10,000,000 7· 700
rubbed off by applying a few brisk strokes Gray 8 100,000,000 8· 800
White g 1,000,000,000 g. gOO
with medium-grade, steel wool. W ARN-
Gold 0.1 5 1000
INa: Always use rubber gloves when Silver 0.01 10 2000
working with etchant powders and solu- No color 20 500
tions. Should the acid bath come in con- ·Applies to capacitors only.
tact with the body, immediately wash the
affected area with clear water. Protect the
eyes when using acid baths. A construction technique that is prcfc- Handbook, it is to be understood that
tically indistinguishable from true printed when no tolerance is specified the largest
Alternative Construction Methods circuitry is excavating a copper-clad board tolerance available in that value will be
, Some would-be builders express revul- with a hand-held grinding tool, such as satisfactory.
sion at the prospect of pc-board fabrica- the Moto-tool manufactured by the Values that do not easily fit into the
tion. The distaste for chemical processes Dremel Company. The simpler circuits preferred-number system (such as 500,
should not deter a person, however, for can be cut out of the board with an X-acto 25,000) can be substituted. It is obvious,
several alternatives exist. knife. for example, that a 50oo-ohm resistor falls
Practically all designs are "bread- Perhaps the least complicated approach well within the tolerance range of the
boarded" before. being committed to a to circuit-board fabrication is the use of 47oo-ohm 20-percent resistor used in the
printed circuit. The fact that these proto- unclad perforated board into which a example above. It would not, however, be
types work proves that etched circuit number of push-in terminals have been in- usable if the tolerance were specified as 5
boards aren't an absolute necessity. stalled. The perforated board can be ob- percent.
Where a ground plane is required, Teflon tained with one of many hole patterns,
terminals pushed through holes in a dependent upon the needs of the builder. Color Codes
copper-clad board allow neat and rigid Perforated terminal boards are manufac- Standardized color codes are used to
component mounting. High-value resis- tured by several companies. Their pro- mark values on small components such as
tors with one end soldered to the ground ducts are available from most mail-order composition resistors and mica
plane can also be used for standoff ter- houses. capacitors, and to identify leads from
minals. Once the builder plots the layout of his transformers and other large components.
Low- and medium-speed digital circuits circuit on paper, push-in terminals can be The resistor-capacitor number color code
are often assembled on a wire wrap board. installed in the "perf" board to match the is given in Table 4.
The IC sockets have long pins around layout which was done on paper. The ter-
which small solid wires are wrapped. An minals serve as tie points and provide Fixed- Value Capacitors
electric or pneumatic "gun" is used to secure mounting-post anchors for the The methods of marking "postage-
make the connections in industry, but a various components. Selected terminals stamp" mica capacitors, molded paper
manual wrapping tool can be used when can be wired together to provide ground capacitors and tubular ceramic capacitors
time is not of the essence. Toltec Corpora- and B + lines. Although this technique is are shown in Fig. 16.
tion, listed in Table 11, sells wire- the most basic of the methods, it is entire- Capacitors made to American War
wrapping supplies in small quantities to ly practical. ' Standards (A WS) or Joint Army-Navy
amateurs. (JAN) specifications are marked with the
Radio Shack and the Vector Company Component Values six-dot code shown at the top. Practically
produce a variety of breadboarding fix- Values of composition resistors and all surplus capacitors are in this category.
tures that can also be used in permanent small capacitors (mica and ceramic) are The three-dot EIA code is used for
assemblies. The deluxe models feature specified throughout this Handbook in capacitors having a rating of 500 volts and
several power and ground buses, as well as terms of "preferred values." 'In the ± 20 percent tolerance only; other ratings
IC hole patterns. preferred-number system, aU values repre- and tolerances are covered by the six-dot
sent (approximately) a constant-per- EIA code.
Table 3 centage increase over the next lower value. Example: A capacitor with a six-dot
The base of the system is the number 10. code has the following markings: Top
Approximate Series· Resonant Fre·
quencles of Disc Ceramic Bypass
Only two significant figures are used. row, left to right, black, yellow, violet;
Capacitors "Tolerance" means that a variation of bottom row, right to left, brown, silver,
plus or minus the percentage given is con- red. Since the first color in the top row is
Capacitance Freq.' Freq.' sidered satisfactory. For example, the ac- black (significant figure zero) this is the
0.Q1 "F 13 MHz 15 MHz tual resistance of a "47oo-ohm" A WS code and the capacitor has mica di-
0.0047 18 22 20-percent resistor can lie anywhere be- electric. The significant figures are 4 and
0.002 31 38
0.001 46 55 tween 3700 and 5600 ohms,approximate- 7, the decimal multiplier 10 (brown, at
0.0005 65 80 Iy. The permissible variation in the same right of second row), so the ca'pacitance is
0.0001 135 165 resistance value with 5-percent tolerance 470 pF. The tolerance is ± 10 percent.
'Total lead length of 1 inch (25·mm) would be in the range' from 4500 to 4900 The final color, the characteristic, deals
'Total lead length of 1I2·inch (13·mm) ohms, approximately. with temperature coefficients and
In the component specifications in this methods of testing (see Table 6).
17·9 Chapter 17
FIRST
TableS
SIGNIFlCANT FIGURE
SECOND SI,GNIFICANT FIGURE
Color Code (or Ceramic Capacitors
MICA CAPACITORS - BLACK FIRST THIRD Capacitance
(AWS PAPER FIGURE Tolerance
CAPACITORS-
SECOND Temp.
SILVER)
SIGNIFICANT FIGURE
'Deci More Less Coeff.
SIgni- mal than than ppm!
ficant Multi- 10pF 10 pF deg.
Co/~r Figure pl/er (in %J (in pFJ C.
Black o 1 ±20 2.0 o
Brown 1 10 ± 1 - 30
VOLT AGE RATING DECIMAL MULTIPLIER
Red 2 100 ± 2 - 80
TOLERANCE
Orange 3 1000 -150
OECIMAL MULTIPLIER
Yellow 4 -220
CHARACTERISTIC Green 5 -330
. TOLERANCE Blue 6 ± 5 0.5 -470
Violet 7 -750
AWS AND JAN FIXED CAPACITORS Gray 8 0.01 0.25 30
A- FIRST SIGNIFICANT FIGURE
White 9 0.1 ±10 1.0 500
B- SECOND SIGNIFICANT
FIGURE
C- DECIMAL MULTIPLIER
TEMPERATURE ~
Table 6
COEFFICIENT ~ , D - CAPACITANCE
'\ \ , - - TOLERANCE Capacitor Characteristic Code
9~tI....mII.....n,;IIr----,F= Color Temperature
DECIMAL MULTIPLIER
Sixth CoeffiCient Capacitance
FIRST SIGNIFICANT
FIGURE Dot ppm!deg. C. Drift
A capacitor with a three-dot code has wire-wound units molded in cases iden- Examples: A resistor of the type shown
the following colors; left to right: brown, tical with the composition type) are color- in the iower drawing of Fig. 17 has the
black, red. The significant figures are I, 0 coded as shown in Fig. 17. Colored bands following color bands: A, red; B, red; C,
(10) and the multiplier is 100. The capaci- are used on resistors having axial leads; on orange; D, no color. The significant
tance is therefore 100 pF. radial-lead resistors the colors are placed figures are 2, 2 (22) and the decimal
A capaCitor with a six-dot code has the as shown in the drawing. When bands are multiplier is 1000. The value of resistance
following markings: Top' row, left to used for color coding the body color has is therefore 22,000 ohms and the tolerance
right, brown, black, black; bottom row, no significance. is ± 20 percent.
right to left, black, gold, blue. Since the
first color in the top row is neither black
nor silver, this is the EIA code. The
MIL SPEC. IDENT.
significant figures are'l, 0, 0 (100) and the (SILVER)
decimal multiplier is I (black). The
capacitance is therefore 100 pF. The gold
dot shows that the tolerance is ± 5 percent
and the blue dot indicates 600-volt rating;
FIRST FIG. (GRAY) DECIMAL ""SECOND FIG.
Ceramic Capacitors (GOLD) (RED)
\
tolerance is ±5 percent.
Table 7
83.J.;8FCP
L\RB
~
~
~
FlAIISE 10DY
,
.
CENTER
CONTACT
TIP
2. Screw body on cable. Solder braid through solder holes. Solder
conductor to center contact.
1. Strip Jacket. Don't nick braid. Slide coupling ring and adapter on
cable. Note - use 83·168 adapter for RG·581U and 83·185 for RG·
3. Slide nut over body. Grasp cable with hand and push ferrule over 59/U.
barb until braid Is captured between ferrule and body flange.
Squeeze crimp tip only of center contact with pliers; alternate·solder
tip.
2. Fan braid Slightly, fold back over adapter and trim to 318". Strip
dielectric and tin exposed conductor. Don't nick conductor.
Fig. 20 - Cable stripping dimensions and assembly instructions for several popular coaxial cable connectors. This material courtesy of Amphenol*
Electronic Components, RF Division, Bunker Ramo Corp. (Dimensions on this drawing are In Engllsh'lnches. Multiply Inches Xc 25.4 to obtain mm).
17-11 Chapter 17
Table 8 Table 9
Pllot·Lamp Data Frequency·Spectrum Reference Chart of nonamateur channel assignments and other
frequency data.
Lamp Bead Base Bulb-- Rating
No. Color (Miniature) Type Volts Amp. Frequency (kHz) 15.0 (0.5)"': WWV
26.965-26.985: Citizens Band chan. 1-3 (10-kHz
40 Brown Screw T-31/4 6-8 0.15 15.734264 ± 0.000044: TV hor. scan freq. sep.)
40A' Brown Bayonet T-31/4 6-8 0.15 17.8(0.5)': NAA Cutler, Maine 27.005-27.035: CB chan. 4-7
41 White Screw T-31/4 2.5 0.5 18.6(0.5)': NPG/NLK Jim Creek, Washington 27.055-27.085: CB chan. 8-11
42 Green Screw T-3 1/4 3.2 •• 21.4(0.5)': NSS Annapolis, Maryland 27.105-27.135: CB chan. 12-15
43 White Bayonet T-3114 2.5 0.5 24.0(0.5)': NBA Balboa, Panama, Canal Zone 27.155-27.185: CB chan. 16-19
44 Blue Bayonet T·3114 6-8 0.25 26.1(0.5)': NPM, Hawaii 21.205-27.225: CB chan. 20-22
45 Bayonet T·3114 3.2 •• 60.0(0.5)"': WWVB Ft. Collins, Colorado 27.235-27.255: CB chan. 24, 25, 23
46' Blue Screw T-31/4 6-8 0.25 85: Receiver i-f (command set of "Q5er") 27.265-27.405: CB chan. 26-40
47' Brown Bayonet T-31/4 6-9 0.15 100.0(0.5)': Loran C (regional) 41.25: TV sound carrier (location in receiver i-f)
48 Pink Screw T-3 1/4 2.0 0.06 179: WGU20 CD Station, East Coast, Bc of WX 42.17: TV color subcarrier (location in receiver
49' Pink Bayonet T·3114 2.0 0.06 and time (a-m) i-f)
49A' White Bayonet T·3114 2.1 0.12 285-325: Marine RDF bnad. Two cw tones 1020-Hz 45.75i TV picture carrier (location in receiver i-f)
50 White Screw G-31/2 6-8 0.2 apart 54-72: TV chan. 2-4. (Three 6-MHz chans. start-
51' White Bayonet G-3112 6-8 0.2 285-405: Aero RDF; aero WX (a-m) 325-405. ing from 54)
53 Bayonet G·3112 14.4 0.12 415-490: Marine (cw) 72, 75: RC chans.
55 White Bayonet G-41/2 6-8 0.4 455: Receiver i-f/mech. filters (Collins) 76-88: TV chan. 5·6
292' White Screw T:3114 2.9 0.17 535-1605: bc (a-m), 107 chans. every 10 kHz from 88.1-107.9: Bc (fm) 100 chan. from 88.1 (carrier)
292A' White Bayonet T·31/4 2.9 0.17 540 (carrier) in 200-kHz steps
1455 Brown Screw G-5 18.0 0.25 120-130: Aero, RDF WX
1455A Brown Bayonet G·5 18.0 0.25 Frequency (MHz) 137.5,137.62: WX Sat. (A4). Ref. W1AW Bul. for
1487 Screw T-3 1/4 12·16 0.20 orbital data
1488 Bayonet T·3114 14 0.15 1.8-2.0: Loran A (pulse transmission) 162.4: Marine WX bc (fm, regional)
1813 Bayonet T·3114 14.4 0.10 2.5 (0.5)',': WWV, Ft. Collins, Colorado, WWVH 174-216: TV chan. 7-13
1815 Bayonet T-31/4 12·16 0.20 Kekaha, Hawaii 470-890: TV chan. 14-83 (70 chan. 6-MHz wide)
'40A and 47 are interchangeable. 3.33 (50)',': CHU, Ottawa, Canada
'Have frosted bulbs. 3.395: Transceiver i-f (Heath, Kenwood) 'Standard-frequency transmission figur~ in .
'49 and 49A are interchangeable. 3.579545 ± 10-5: TV chrominance subcarrier brackets is error in parts 10" (Electronics
'Replace with no. 48. 5.0 (0.5)',': WWV, WWVH Engineers' Handbook, McGraw Hill, pp. 1-48).
. 'Use in 2.5-volt sets where regular bulb burns 5.645: receiver i-f (Drake) 'Standard time station. A3 transmissions in:
out too frequently. 7.335 (50)',': CHU 'clude time, weather and propagation on
'White in G.E. and Sylvania; green in National 9.0: Xtal filters (KVG) WWVIWWVH. A3 time transmission on CHU
Union, Raytheon and Tung·Sol. 19.0 (0.5)"': WWV, WWVH (English/French). WWVB has no A3; info in
"0.35 in G.E. and Sylvania; 0.5 in National 10.7: Receiver i-f (fm bc) BCD format generated by reducing carrier by
Union, Raytheon and Tung-Sol. 14.67 (50)',': CHU 10-dB (binary 0).
PLtJG BODY
BNC CONNECTORS (STANDARD CLAMP)
m
~lZ IRB·58/UI 1--1
"/" IRB·59/UI
1. Strip Jacket. Fray braid and strip dielectric. Don't nick braid or
5. Push assembly Into body. Screw nut Into body with wrench until
conductor. Tin conductor.
tight. Don't rotate body on cable to tighten.
~"9/64
2. Taper braid. Slide nut, washer, gasket and clamp over braid.
Clamp inner shoulder should fit squarely against end of Jacket.
~3/16
~ ~
GASKET CLAMP
~
.
3. With clamp in place, comb out braid, fold back smooth as shown.
Trim 3/32" from end.
~
~• SOLDER HOLE
4. Solder contact on cQnductor through solder hole. Contact should 1. Follow 1, 2, 3 and 4 In BNC connectors (standard clamp) except
butt against dielectric. Remove excess solder from outside of con· as noted. Strip cable as shown. Slide gasket on cable with groove
tact. Avoid excess heat to prevent swollen dielectric which would facing clamp. Slide clamp on cable'wlth sharp edge facing gasket.
interfere with connector body. . , Clamp should cut gasket to seal properly.'
Table 10
Copper·Wlre Table
Cont.-duty Current
Cont.-duty current' Feet Carrying
Wire L current' wires or per Ohms Capacity' Nearest
Size Dlam. Circular Single wire . cables in Pound per ,at Diam. British
-A. W. G. in Mil Turns per Linear Inch (25.4-mm), in conduits (0.45 kg) 1000 ft. 700 C.M. in S.W.G.
(B&S) MllsJ Area Enamel S.C.E. D.C.C. open air or bundles Bare 25 0 C per Amp. mm. No.
1 289.3 83690 3.947 .1264 119.6 7.348 1
2 257.6 66370 4.977 .1593 94.8 6.544 3
3 229.4 52640 1?276 .2009 75.2 5.827 4
4 204.3 41740 7.914 .2533 59.6 5.189 5
5 181.9 33100 9.980 .3195 47.3 4.621 7
6 162.0 26250 12.58 .4028 37.5 4,115 8
7 144.3 20820 15.87 .5080 29.7 3.665 9
8 128.5 16510 7.6 7,1 73 46 20.01 .6405 23.6 3.264 ' 10
9 114.4 13090 8.6 7.8 25.23> .8077 18.7 2.906 11
10 101.9 10380 9.6 9.1 8.9 55 33 31.82 1.018 14,8 2,588 12
11 90.7 8234 10.~ 9.8 40.12 1.284 11.8 2.305 13
12 80.8 6530 12.0 11.3 10.9 41 23 50.59 1.619 9.33 2.053 14
13 72.0 5178 13.5 12.8 63.80 2.042 7.40 1.828 15
14 64.1 4107 15.0 14.0 13.8 32 17 80.44 2.575 5.87 1.628 16
15 57.1 3257 16.8, 14.7 101.4 3.247 4.65 1.450 17
16 50.8 2583 18.9 17.3 16.4 22 13 127.9 4.094 3,69 1.291 18
17 45.3 2048 21.2 18.1 161.3 5.163 2.93 1.150 18
18 '40.3 1624 23.6 21.2 19.8 16 10 203.4 6.510 2.32 1.024 19
19 35.9 1288 26.4 21.8 256.5 8.210 1.84 ,912 20
20 3,2.0 1022 29.4 25.8 23.8 11 7.5 323.4 10.35 1.46 .812 21
21 28.5 810 33.1 26.0 407.8 13.05 1.16 _ .723 22
22 25.3 642 37.0 31.3 30.0 5 514.2 16.46 .918 ,644 23
23 22.6 510 41.3 37.6 648.4 20.76 .728 ,573 24
24 20.1 404 46.3 37.6 35.6 817.7 26.17 .577 .511 25
25 17.9 320 51.7 38.6 1031 33.00 .458 .455 26
26 15.9 254 58.0 46.1 41.8 1300 41.62 ,363 .405 27
27 14.2 202 64.9 45.0 1639 52.48 .288 .361 29
28 12.6 160 72.7 54.6 48,5 2067 66.17 .228 .321 30
29 11.3 127 81.6 51.8 2607 83.44 .181 .286 31
30 10.0 101 90.5 64.1 55.5 3287 105.2 ,144 ,255 33
31 8.9 80 101 59.2 4145 132.7 .114 .227 34
32 8.0 63 113 74.1 61.6 5227 167.3 .090 .202 36
33 7.1 50 127 66.3 6591 211.0 .072 ,180 37
34 6.3 40 143 86.2 70.0 8310 266.0 .057 .160 38
35 5.6 32 158 73.5 10480 335 .045 .143 38-39
36 5.0 25 175 103.1 77.0 13210 423 .036 .127 39-40
37 4.5 20 198 80.3 16660 533 ,028 .113 41
38 4.0 16 224 116.3 83.6 21010 673 .022 .101 42
39 3.5 12 248 86.6 26500 848 .018 .090 43
40 3.1 10 282 131.6 89.7 33410 1070 .014 .080 44
'A mil is 0.001 inCh.
'Figures given are approximate only; insulation thickness varies with manufacturer.
'Max. wire temp. of 212 0 F (100 0 C) and max. ambient temp. of 135 0 F (57 0 C).
'700 circular mils per ampere is a satisfactory design figure for small transformers, but values from 500 to 1000 c.m. are commonly used.
17·13 Chapter 17
Table 11 Fox-Tango Club Piezo Technology', Inc,
Box 15944 p, O'. Box 7859
ARRL Parts Supplier List W. Palm Beach, FL 33406 O'rlando, FL 32854
F G and C Communications E, M Poly Paks
730 Cottonwood • Free Box 942
Lincoln, NE 68510 Lynnfield, MA 01940
A, E,I, L, M Adva Electronics Paul A Greenfeld (Manuals) Ouest ElectroniCs
E,I, L, M, U
• 28~ stamp Box 4181 9715 Tulsemere Rd. P. O'. Box 4430
•• $10 Woodside, CA 94062 Randallstown, MD 21133 Santa Clara, CA 95054
H Airway Products P Gregory Electronics Corp. A,B, 0, E,F,G, Radiokit
;,p. O'. Box 337A 249 Route 46 , H, J, K, L P. O'. Box 411
Princeton, MN 55371 Saddle Brook, NJ 07662 ·Free Greenville, NH 03048
A, E, L, M 'Aldelco E Hamilton Avnet W RTC Electronics
• stamp 2789 Milburn Ave. 4445 Si~ma Rd. • S.a.5.e . P. O'. Box 2514
•• $10 Baldwin, NY 11510 Dallas, X 75240 Lincoln, NE 68502
A,C,I, K, L Allied Electronics Hammarlund Manufacturing Co.
Div. of Pax Mfg. Corp. E S. D. Sales
• $1 401 East 8th Street P.O'. Box 28110
•• $15 FI. Worth, TX 75102 100 E. Montauk Highway
Lindenhurst, NY 11757 Dallas, TX 75228
Depl. OST
(manuals - $15, schematics 0 Savoy Electronics, Inc.
A, E, H, J, L, M, U Alpha Electronic ~aboratories - $5) Box 5727
2302 O'kland Gravel Road FI. Lauderdale, FL 33310
Columbia, MO' 65201 I, K Hammond Mfg, Ltd.
Attn. Don Manson, K0TVO' 394 Edinburg Rd. N. E Semiconductor Surplus
314-874-1514 Guelph, O'N NIH 1E5 CANADA 2822 N. 32nd SI. Unit 1
'Hammond Mfg. Co., Inc. (US) . Phoenix, AZ 85018
B,C Amidon Associates 1690 Walden Ave.
12033 O'tsego Street 0 Sentry Mfg. Co.
Buffalo, NY 14225 Crystal Park
N. Hollywood, CA 91607
L Harrison Radio Chickasha, O'K 73108
0' Atlantic Surplus Sales 20 Smith St.
(Facsimile Equipment) 0, J Sherwood Engineering, Inc.
Farmingdale, NY 11713 1268 S. O'lbden SI.
3730 Nautilus Ave.
Brooklyn, NY 11224 B, I, K, M, N, T Herbach and Rademan, Inc. Denver, C 80210
401 E. Erie Ave. F, V Skylane Products
A, E, I, T, U ATV Research Philadelphia, PA 19134
13th and Broadway 406 Bon Air Ave.
Dakota City, NE 68731 L HI, Inc. Temple Terrace, FL 33617
• 25~ Box 864 •• $5 Small Parts, Inc .
E, L Babylon Electronics •• $1 Council Bluffs, IA,51502 P. O'. Box 381736
Box 41778 (send 25~ in coin for manual Miami, FL 33138
Sacramento, CA 95841 list) (mechanical components and
A,B,D,F,G Barker and Williamson, Inc. C, E, I, L, M, N, U Hobbyworld metal stock)
• 10 Canal Street •• $10 19511 Business Center Dr. A, P Spectronics
Bristol, PA 19007 Northridge, CA 91324 1009 Garfield SI.
M,N BUdf<et Electronics E Integrated Circuits Unlimited O'ak Park, I L 60304
• Stamp 2 04 W. North Ave. 7889 Clairemont Mesa Blvd . A, 0, J Spectrum International
•• $10 Chicago, IL 60647 San Diego, CA 92111 P. O'. Box 1084
E,I, M Bullet Electronics 0 International Crystal Co. Concord, MA 01742
P. O'. Box 401244 •• $5 10 North Lee St. M,N Star Tronics
Garland, TX 75040 O'klahoma City, O'K 73102 •• $4 P. O'. Box 683
B Cad del Coil Corp. A,E,M,U Jameco Electronics McMinnville, O'R 97128
35 Main SI. •• $10 1355 Shoreway Rd. E, F, H, I, M, N Surplus Electronics
Poultney, VT 05764 San Carlos, CA 94002 •• $15 7294 N.W .. 54th SI.
(coils for ARRL projects) Miami, FL 33166
0 JAN Crystals
A,H,L Cambridge Thermionic Corp. 2400 Crystal Dr. - P. O'. Box B,M,O',U Teleprinter Corporation of
• Free 445 Concord Ave. 06017 •• "0 America
•• $10 Cambridge, MA 02138 Ft. Myers, FL 33907 550 Springfield Ave.
H Caywood Electronics A,E,F,H,M,N Marlin P. Jones and Assoc. Berkeley Heights, NJ 07922
(Millen Capacitors) P. O'. Box 12685 K Ten-Tec, Inc.
67 Maplewood SI., P. O'. Box U Lake Park, FL 33403 Highway 4,11, E.
Malden, MA 02148 Sevierville, TN 37862
C Kepro Circuit Systems, Inc.
A,C Circuit Board Specialists 630 Axminister Dr. F, V Texas Towers
(circuit boards for ARRL Fenton, MO' 63026 1309 Summit Ave.
projects, kits) Plano, TX 75074
P. O'. Box 969 A, E, M, U Key Electronics
P. O'. Box 3506 Toltec Corp.
Pueblo, CO' 81002 21342 Washington, N.E.
Schenectady, NY 12303
A, E Circuit Specialists Albuquerque, NM 87110
P. O'. Box 3047 F Kirk Electronics (wire-wrapping supplies)
Scottsdale, AZ 85257 •• $10 73 Ferry Rd.
Chester, CT 06412 0' Typetronics
A,B,D,H,L o and V Radio Parts * S.a.8.e. Box 8873
• stamp 12805 W. Sarle H,G,L Leeds Radio Ft. Lauderdale, FL
Freeland, MI 48623 57 Warren St.
New York, NY 10007 E, M Weinschenker
0,1, M Peter W. Dahl Box 353
4007 Fort Blvd. K, L MFJ Enterprises Irwin, PA 15642
EI Paso, TX 79930 P. O'. Box 494
Mississippi State, MS 39762 R Western Nebraska Electronics
E, B, 0, M, N,U Di~ital Research Corp. Rte.l - Box 1
. O'. Box 401247 E, M M HZ Electronics Potter, NE 69156
Garland, TX 75040 2111 W. Camelback Rd.
Phoenix, AZ 85015 E, G, 0 Workman Electronic Products, Inc
L Dominion Radio and Elecl. Co. Box 3828
535 Yonge SI. A,B,H J. W. Miller Div., Bell Industries Sarasota, FL 33578
Toronto, O'N CANADA 19070 Reyes Aves 813-371-4242
Compton, CA 90224 (will refer customer to the
C Dynaclad
P. O'. Box 296 A, B, 0, 0, E, F, G, Milo Associates nearest distrib.)
Meadowland, PA 15347 H, K, M Box 2323 Chart Coding
• S.a.5.8. Indianapolis, IN 46206 A - New Components
A, B, C, 0, F, G, H, Electronics Emporium Inter- •• $15
I, J, K, L, M, P, national B - Toroids and Ferrites
U,V 7889-95 Clairemont Mesa Blvd. L Modern Radio Laboratories C - Etched Circuit Board Materials
San Diego, CA 92111 '5~ P. O'. Box 1477-0 o - TranSmitlin~ and Receiving Materials
Garden Grove, CA 92642 E - Solid State evices
A,G, L Electronic Instrument and F - Antenna Hardware
Specially Corp., MC Division E, F,I, M, N O'lson Electronics
260 South Forge St. G - Dials and Knobs
(Millen parts except capaCitors) H - Variable CapaCitors
5 Lowell Ave. Akron, O'H 44327
I - Transformers
Winchester, MA 01890 B Palomar Engineers J - I-f Filters
L Electro-Sonic, Inc. Box 455 K - Cabinets and Boxes
• S10 1100 Gordon Baker Rd. Escondido, CA 92025 L - General Supplier
Willowdale, O'N M2N 3B3 C PBI Electronics M - Surplus Parts
CANADA 416-494-1555 P. O'. Box 5 N - SU~IUS Assemblies
C, 0, E, G, H, I, M Etco Electronics Azusa, CA 91702 0' - RT Y Equipment & Parts
North Country Shopping Center P - Surplus FM Gear and Parts
T,U,W P. C. Electronics R - Service of Collins Equipment
Rte. '9 North 2522 Paxson
Plattsburgh, NY 12901 T - Amateur TV Cameras & Components
Arcadia, CA 91006 U - Mierocomputer Peripheral Equipment
M,N Fair Radio Sales L C. M. Peterson Co., Ltd. V - Towers
•• $5 Box 1105. 220 Adelaide SI. N. W - Ready-made Printed Circuit Boards
Lima, O'H 45802 London, O'N N6B 3H4 CANADA
'Catalog price
"Minimum O'rder
To the best of our knowledge the suppliers shown are willing to sell components to amateurs in small quantities by mail. This listing does not necessarily indicate that these
firms have the approval of ARRL.
Wave Propagation
Though great advances have been made. Sunspot Cycles any appreciable period, while others have
in recent years in understanding the many Even before their correlation with radio had severl\1 months of little or no activity.
modes of propagation of radio waves, propagation variations was well-known, Sunspot cycles do not have a sine-wave
variables affecting long-distance com- the periodic rise and fall of sunspot shape. The rise is shorter than the decay,
munication are very complex, and not numbers had been studied for many years. but neither is clearly defined. October
entirely predictable. Amateur attempts to These cycles average roughly 11 yel;lrs in 1974 had a solar flux range of 73 to 144.
sc.hedule operating time and frequencies length, but have been as short as 9 and as June 1976, the last month of Cycle 20, had
for optimum results may not always long as 13 years. The highs and lows'ofthe several quiet-sun ·days (solar flux 66), but
succeed, but familiarity with the nature of cycles also vary greatly. Cycle 19 peaked April and August had readings in the 80s.
radio propagation can reduce the margin in 1958 with a sunspot number of over November 1979, peak month of Cycle 21,
of failure and add greatly to one's 200. Cycle 20, of nearer average intensity, had 383 (equalling the Cycle 19 record
enjoyment of the pursuit of any kind of reached 120 in 1969. By contrast, one of high) and 154, within 18 days.
DX. the lowest, Cycle 14, peaked at only 60 in
The sun, ultimate source of life and 1907. Several cycle lows have not Sola.r Radiation
energy on earth, influences all radio reached zero levels on the Zurich scale for Insofar as it affects most radio pro-
communication beyond the local range. pagation, solar radiation is of two
Conditions vary with such obvious sun- principal kinds: ultraviolet light and
"The DXer's Crystal Ball," QST, June,
related earthly cycles as time of day and 'Tilton,
August and September 1975. The WWV bulletin charged particles. The first travels at just
season of the year. Since these differ for form changed in 1976, but basic principles apply. under 300,000,000 meters (186,000 miles)
appreciable changes in latitude and
longitude, almost every communications
circuit is unique in some respects. There 240 J
are also short- and long-term solar cycles /
which influence propagation in less 230
obvious ways. Furthermore, the state of 220
V
the sun at a given moment is critical to - V
long-distance communication, so it is 210
understandable that propagation fore- 200
V
casting is still a rather inexact science. V
With every part of the radio spectrum 190
open to our use differing in its response to
V
180
solar phenomena, amateurs have been, V
170 /
and still are, in a position to contribute to
advancement of the art, both by accident x
.::0 160 /
..J
and by careful investigation. "- /
II:: 150
<t
SOLAR PHENOMENA ..J
0
U) 140 /
Man's interest in the sun is older than
recorded history. Records of sunspot
e
", 130 / (
18;.1 Chapter 18
on the projection surface: Put a baffle on earth's surface where free ions and
the scope to enlarge the shaded area and electrons exist in sufficient quantity to
adjust the focus to give a sharp-edged affect the direction of wave travel.
image of the solar disk. If there are any Ionization of the upper atmosphere is
sunspots you will see them now. Draw a attributed to ultraviolet radiation from the
rough sketch of what you see, every time sun. The result is not a single region, but
an observation is made, and keep it with several layers of varying densities at
your record of propagation observations. various heights surrounding the earth.
Spots move across the image from left Each layer has a central region of
to right, on the projected image, as it is relatively dense ionization that tapers off
viewed with the sun at the observer's both above and below.
back. The line of movement is parallel to
the solar equator. Not all activity capable Ionospheric Layers
of affecting propagation can be seen, but The lowest useful region of the ionos-
any spots !leen have significance. Active phere is called the E layer. Its average
areas may develop before spots are visible height of maximum ionization is about 70
and may persist after spots associated miles. The atmosphere here is still dense
with them are gone, but once identified by enough so that ions and electrons set free
date they are likely to recur about 27 days by solar radiation do not have to travel far
later, emphasizing the worth of detailed before they meet and recombine to form
• Fig. 2 - A simple sunspot projection system, records. neutral particies: The layer can maintain
demonstrated by ARRL staffer AC1Y.' The card·
board baffle at the top provides a shaded area Vhf or uhf arrays capable of movement its ability to bend radio waves only when
for viewing the sun's image. Sunspots large in elevation as well as azimuth are useful continuously in sunlight. Ionization is
enough to affect radiowave propagation can be for solar noise monitoring. ,With a good thus greatest around local noon, .and it
seen quite easily with this simple setup. The system, the "quiet sun" can be "heard" at practically disappears after sundown.
best time to look .for sunspots is the early In the daylight hours there is a still
morning. a low level. 3 Bursts that can be many dB
higher indicate the start of a major event, lower area called the D region where
such as a solar flare capable of producing ionization is proportional to the height of
per second, as does all electromagnetic an hf blackout and possibly vhf auroral the sun. Wave energy in the two lowest
.radiation, so UV effects on wave pro- propagation. frequency amateur bands, 1.8 and 3.5
pagation develop simultaneously with MHz, is almost completely absorbed by
increases in observed solar noise, ap- TYPES OF PROPAGATION this layer. Only the highest-angle radiation
proximately eight minutes after the actual Depending on the means of pro- passes through it and is reflected back to
solar event. Particle radiation moves more pagation, radio waves can be classified as earth by the E layer. Communication on
slowly, and by varying routes, so it may ionospheric. tropospheric, or ground waves. these bands in daylight is thus limited to
take up to 40 hours· to affect radio The ionospheric or skywave is that main short distances, as the lower-angle radia-
propagation. Its principal effects are high portion of the total radiation leaving the tion needed for longer distances trav-
absorption of radio energy and the antenna at angles somewhat above the els farther in the D region and is absorb-
production of auroras, both visual and the horizontal. Except for the reflecting ed.
radio variety. . qualities of the ionosphere, it. would be The region of ionization mainly respon-
Variations in the level of solar radiation lost in .space. The tropospheric wave is sible for long-distance communication is
can be gradual, as with the passage of that portion of the radiation kept close to called the F layer. At its altitude, about
some sunspot groups and other long-lived the earth's surface as the result of bending 175 miles at night, the air is so thin that
activity centers across the solar disk, or in the lower atmosphere. The ground recombination takes place very slowly.
sudden, as with solar flares. An important wave is that portion of the radiation Ionization decreases slowly after sun-
clue for anticipating variations in solar directly affected by the surface of the down,reaching a minimum just before
radiation levels and radio propagation earth. It has two components, an earth- sunrise. The obvious effect of this change
changes resulting from them is the guided surface wave, and the space wave. is the early disappearance of long-distance
rotational period of the sun, approxi- The latter is the resultant' of two signals on the highest frequency that was
mately 27 days. Sudden events (flares) components, direct and ground-reflected. usable that day, followed by loss of
may be short-lived, but active areas The terms "tropospheric wave" and communication on progressively lower
capable of influencing radio propagation "ground wave" are often used inter- frequencies during the night. In the
may recur at four-week intervals for four changeably, though this is not strictly daytime the F layer splits into two parts,
or five solar rotations. Evidence of the correct. FI and F2, having heights of about 140'
"27-day cycle" is most marked during and 200 mi1es~ respectively. They merge
years of low solar activity. THE IONOSPHERE again at sunset.
Solar activity can be observed quite Long-distance communicati.on and much Scattered patches of relatively dense
easily. Simple projection of the sun's ionization develop seasonally aLE-layer
•
over shorter distances, on frequencies
image, as in Fig. 2, is most useful in the below 30 MHz, is the result of bending of height. Such sporadic E is most prevalent
low years o{ the" II-year" cycle. In times the wave in the ionosphere, a region be- in the equatorial regions, but it is common
of high activity the visible evidence may be tween about 60 and 200 miles above the in the temperate latitudes in late spring
difficult to sort out, unless observations· and early summer, and to a lesser degree
are made daily and the results are record- in early winter. Its effects become
ed with care. Enough definition for our confused with those of other ionization on
purposes is possible with the simplest the lower amateur frequencies, but they
'Projection of the sun and interpretation of results
telescopes. Low-cost instruments, 10- to are discussed in reference I, and in QST. December stand out above 21 MHz, especially in the
30-power, are adequate. A principal re- 1974, p. 83; January 1975, p. 84 ana October 1976, low-activity years of the solar cycle, when
quirement is provision for mounting on a p. II. A black-box viewing aevice (Tomcik, other forms of DX are not consistently
K40U) for sun projection is shown in July 1964,
tripod having a pan-tilt head. 2 QST. (Photocopy from ARRL, 75 cents and "- available.
Adjust the aiming to give a circular stamped envelope.) Duration of openings decreases and the
shadow of the scope body, then move the 'Bray and Kirchner, "Antenna Yatterns [rom toe length of skip increases with progressively
Sun," QST, July 1960. Wilson, "432-MHz Solar
scope slowly until a bright spot appears Patrol," QST, August 1967. higher frequencies. Skip distance is com-
monly.a few hundred miles 'on 21 or 2S Virtual Height explanatqry purposes. .But an under-
MHz, but mUltiple hop propagat.ion can An ionospheric layer is a region of standing of long-distance .propagation
extend the range to 2500 miles or more. considerable depth, but for practical must take the earth's magnetic field into
June and July are the peak months in the purposes it is convenient to think of it as account. Because of it, the ionosphere is a
northern hemisphere. Es propagation is having finite height, from which a simple birefringent medium (double refracting)
most common in midmorning and early reflection would give the same effects which breaks up plane-polarized waves
evening, but may extend almost around (observed from the ground) as result from into what are known as the .ordinary and
the clock at times. The highest frequency the gradual bending that actually takes extraordinary waves, f oF2 and fxF2 in the
for Es is not known, but the number of place. It is given several names, such as ionogram. This helps to explain the
opportunities for using the mode drops group height, equivalent height, and virtual dispersal of plane polarization encoun-
off rapidly between the amateur 50- and height. tered . in most ionospheric communica-
144-MHz bands, whereas 28 and 50 MHz The virtual height of an ionospheric tion.'
are quite similar. layer for various frequencies and vertical Sudden marked increases in solar
The greater the intensity of ionization incidence is determined with a variable- radiation, such as with solar flares, trigger
ina layer, the more the wavepath is bent. . frequency sounding device that directs instantaneous effects in the F, E and D
The bending also qepends on wavelength; pulses of energy vertically and measures regions; slightly delayed effects, mainly in
the longer the wave the more its path is the time required for the round-trip path the polar areas; and geomagnetic effects,
modified for a given degree of ionization. shown in Fig. 3. As the frequency delayed up to 40 hours.
Thus, for a given level of solar radiation, rises, a point is reached where no Onset of the D-region absorption is
ionospheric communication is available energy is returned vertically. This is usually sudden, lasting a few minutes to
for a longer period of time on the known as the critical frequency for the several hours, leading to use of the term
lower-frequency amateur bands than on layer under consideration. A representa- SID (sudden ionospheric disturbance).
those near. the upper limit of hf spectrum. tion of a typical ionogram is shown in Fig. Shortwave fadeouts (SWFs) and SIDs
The intensity and character of solar 4.4 In this sounding the virtual height for exhibit wide variations in intensity, dura-
radiation are subject to many short-term 3.5 to 4 MHz was 400 km. Because the tion and number of events, all tending to
and long-term variables, the former still ionogram is a graphical presentation of be greater in periods of high solar activity.
predictable with only partial success. wave travel time, double-hop propagation Radiation Angle and Skip Distance
Absorption appears as an 800-km return for the same
frequency. The critical frequency was just The lower the angle above the horizon
In traveling through the ionosphere, a over 5 MHz on this occasion. Such a clear at which a wave leaves the antenna, the
radio wave gives up some of its energy by F-Iayer ionogram is possible only under less refraction in the ionosphere or
setting the ionized particles in motion. magnetically quiet conditions, and at troposphere is required to bring it back,
When moving particles collide with night, when little or no E- and D-Iayer or to maintain useful signal levels in the
others, this energy is lost. Such absorption ionization is present. case of tropospheric bending. This results
is greater at lower frequencies. It also in the emphasis on low radiation angles in
increases with the intensity of ionization, Effects of the Earth's Magnetic Field the pursuit of OX, on the hror vhf bands.
and with the density of the atIllosphere. The ionosphere has been discussed thus Some of the effects of radiation angle
This leads to a propagation factor often far in terms of simple bending, or are illustrated in Fig. 3. The high-angle
not fully appreciated: Signal levels and refraction, a concept useful for some wave at the left is bent only slightly in the
quality tend to be best when the operating ionosphere, and so goes through it. The
freque,ncy is near the maximum that is 'Davies, "Ionospheric Radio Propagation," NBS
Monograph 80, out of print. Available in some wave at the somewhat lower angle is just
reflected back to earth at the time. technical libraries. ' capable of being returned by the iono-
sphere. In daylight it might be returned
via the E layer. Its area of return from the
F layer, R2, is closer to the transmittting
point, T, than is that of the lowest-angle
wave. If R2 is at the shortest distance
where returned energy is usable, the area
between R 1 and the outer reaches of the
ground wave, near the transmitter, is
called the skip zone. The distance between
R2 and T is called the skip distance. The
distances to both Rl and R2 depend on
the ionization density, the radiation angle
at T, and the frequency in use~ The
maximum distance for single-hop propa-
gation via the F layer is about 2500 miles
R2
(4000 kilometers). The maximum E-Iayer
F2 totAX
single hop is about 1250 miles (2000
F 4000 KM
kilometers).
The maximum usable frequency (muf)
E for E-Iayer communication is about three
times the critical frequency for vertical
EARTH return, as in Fig .. 3. For F-Iayer propaga-
tion it is, about five times.
Fig. 3 - Three types of ionospheric propagation. Sounder, left measures virtual height and Multiple-Hop Propagation
critical frequency of F2 layer. Transmitter T is shown radiating at three different angtes. Highest On its return to earth, the i~nospherical-
passes through the ionosphere after slight refraction. Lower-angle wave is returned to earth by
the E layer, if frequency is low enough, at a maximum distance of 2000 km. The F-Iayer reflection
returns at a maximum distance of about 4000 km, depending on the radiation angle. It is shown
traversing a second path (double hop) from R2 to R4, the latter beyond single-hop range. The
lowest-angle wave reaches the maximum practical single-hop distance at R3. 'See reference 4.
--
10 F2 I . F2.
interesting adjunct. to amateur communi·
<It""
E 400 - - I-hop V V cation at 21 MHz and higher, espec-
"" F- ially in periods of low solar activity.
It is at its best during major me-
200 teor showers. 6
Backscatter
o A complex form of scatter is readily
observed when working near the maximum
2 3 4 5· usable frequency for the F layer at the
MHz time. The transmitted wave is refracted
back to earth at some distant point, which
Fig. 4 - F·layer ionogram taken at night during magnetically quiet conditions. The traces show may be an ocean area or a land mass
the breaking up into ordinary and extraordinary waves. Because it required twice the travel time, where there is no use of the frequency in
the double:.t\op return appeJl,rs as having come from twice the height of the single-hop. question ,at ,the time. A small part of the
energy is scattered back to the skip zone
ly progagated wave can be reflected than' single-sideband signals with sup- of the transmitter via the ionospheric
upward near Rl or R2, travel again to the pressed carrier do. route.
ionosphere, and be refracted to earth_ Backscatter signals are generally rather
This process can be repeated several times THE SCATTER MODES weak, and subject to some distortion from
under ideal propagation conditions; multipath effects. But with optimum
leading to communication beyond halfway Much long·distance propagation can be equipment they are usable at distances
around the world. Ordinarily ionospheric described in terms of discrete reflection, from just beyond the reliable local range
absorption and ground-reflection losses though the analogy is never precise since out to several hundred miles. Under ideal
exact tolls in signal level and quality, so true reflection would be possible only with conditions backscatter communication is
multiple-hop propagation usually yields perfect mirrors, and in a vacuum. All possible over 3000 miles or more, though
lower signal levels and more distorted electromagnetic wave propagation is sub- the term "sidescatter" is more descriptive
modulation than single-hop. This is not ject to scattering influences which alter of what probably happens on such long
always the case, and under ideal condi- idealized patterns to a great degree. The paths,
tions even long-way· around communica- earth's atmosphere and ionospheric layers The scatter modes contribute to the
tion is possible with good signals. There is are scattering media, as are most objects usefulness of the higher parts of the OX
evidence to sUPPQrt the theory that that intervene in the wave path as it leaves spectrum, especially during periods of low
signals for such communications, rather the earth. Strong returns are thought of as solar activity when the normal ionos·
than hopping, may be ducted through reflections and weaker ones as scattering, pheric modes are less often available.
the ionosphere for a good part· of but both influences prevail. Scatter modes
the distance. have become useful tools in many kinds of MF AND HF PROPAGATION
communication.
The 1.8·MHz band offers reliable
Fading Forward Scatter communication over distances up to at
Two or more parts of the wave may We describe a skip zone as if there were least 50 miles during daylight. On winter
follow different paths, causing phase no signal heard between the end of useful nights ranges up to several thousand miles
d~ferences between wave components at ground·wave range and the points R 1 or are possible.
the receiving end. Total field strength may R2 of Fig. 3, but actually the transmitted The 3.5·MHz band is seldom usable
be greater or smaller than that of one signal can be detected over much of the beyond 200 miles in daylight, but long
component. Fluctuating signal levels also skip zone, with sufficiently sensitive' distances are not unusual at night,
result from the changing nature of the devices and methods. A small portion of especially in years of low solar activity.
wave path, as in the case of moving the transmitted energy is scattered back to Atmospheric noise tends to be high in the
air-mass boundaries, in tropospheric pro· earth in several ways, depending on the summer months on both 3.5 and 1.8
pagation on the higher frequencies. frequency in use. MHz. '
Changes in signal level, lumped under the Tropospheric scatter extends the local The 7·MHz band has characteristics
term jading, arise from a variety of communications range to an increasing similar to 3.5 MHz, except that much
phenomena; some, natural, some man- degree with frequency, above about 20 greater distances are possible in daylight,
made. Reflections from aircraft, and MHz, becoming most useful in the vhf and more often at night. In winter dawn
ionospheric "holes" produced by the ex- range. Ionospheric scatter, mostly from and dusk periods it is possible tO,work the
haust from large rocket engines, are in the the height of the E region, is most marked other side of the world, as signals follow
latter category. at frequencies up to about 60 or 70 MHz. the darkness path.
Under some circumstances the wave Vhf tropospheric scatter is usable within The 14·MHz band is the most widely
path may vary with very small changes in the limits of amateur power levels and used OX band. In the peak years of the
frequency, so that modulation sidebands antenna techniques, out to nearly 500 solar cycle it is open to distant parts of the
arrive at the receiver out of phase, causing miles. Ionospheric forward scatter is world almost continuously. During low
distortion that may be mild or severe. discernible in the skip zone at distances up solar activity it is open mainly in the
Called selective jading, this problem to 1200 miles or so. .
increases with signal bandwidth. Double- A major component of ionospheric 'Bain, "VHF Propagation by Metcor Trail Ioniza-
tion," QST, May 1974. Table of major metcor
sideband a-m signals suffer much more scatter is that contributed by short-lived showers, Radio Amateur's VHF Manual, Ch. 2.
18-5 Chapter 18
Tropospheric Bending distances are 600 to 1200 miles. consideration. This job is made easier if
Backscatter, common on lower fre- we understand the causes of the ups and
An easily anticipated extension of nor- quencies, is observed on 50 MHz during downs, so familiarity with basic in-
mal vhf coverage results from abrupt ionospheric propagation, mainly of the F2 formation given earlier in this chapter is
changes in the refractive index of the atmos-variety. Conditions for 50-MHz backscat- helpful.
phere, at bOJ.lndaries between air masses ter are similar to those for the hf bands, What frequencies are "open," and
of differing temperature and humidity detailed earlier in this chapter. where the cutoff in ionospheric pro-
characteristics. Such warm-dry over coo.l- Scatter from meteor trails in the E pagation lies in the spectrum can be
moist boundaries often lie along the .region can cause signal enhancement, or determined quite .readily by tuning up-
southern and western edges of stable slow- isolated bursts of signal from a station not ward in frequency with a general-coverage
moving areas· of fair weather and high baro- otherwise heard. Strength and duration of receiver, until ionospherically propagated
metric presure. Tropospheric bending meteor bursts decrease with increasing signals are no longer heard. The muf for
can increase signal levels from within signal frequency, but the mode is popular the day and the times that a given fre-
the normal working range, or bring for marginal communication in the 50- quency band opens or closes can be found
in more distant stations, not norm- and 144-MHz bands. It has been used on in this way. A daily log will show if con-
ally heard. 220 MHz, and, more marginally, on 432 ditions are improving or deteriorating.
A condition known as ducting or MHz. Listening in the amateur bands and on
trapping may simulate propagation within Random meteor bursts can be heard by immediately adjacent frequencies may be
a waveguide, causing vhf waves to follow cooperating vhf stations at any time or the only way to do this, if the receiver is
earth curvature for hundreds or even season, but early-morning hours are the amateur-bands-only variety. Most DX
thousands of miles. Ducting incidence preferred. Major meteor showers (August bands are narrower in other parts of the
increases with frequency. It is rare on 50 Perseids and December Gelninids) pro- world than in the Americas, so there is no
MHz, fairly common on 144, and more so vide frequent bur~ts.' Some other showers lack of round-the-clock occupancy by
on higher frequencies. ,It occurs most have various periods, and may show other services, ordinarily. Most receivers
often in temperate or low latitudes. It was phenomenal burst counts in peak years. also cover somewhat more than the actual
the medium for the W6NLZ-KH6UK Distances are similar to other E-layer com- amateur assignments, at their widest, so
work on 144, 220 and 432 MHz, over a munication. some commercial, and governmental sig-
2540-mile path. Gulf Coast states see it All scatter communication requires nals can be found close by our band edges.
often, the Atlantic Seaboard, Great Lakes good equipment and optimum operating A worldwide listing of stations, by fre-
and Mississippi Valley areas occasionally, methods. The narrow-band modes are quency, is useful in,identifying signals for
usually in September and October. superior to wide-band systems. propagation monitoring purpose. Don't
Many local conditions contribute to . overlook WIAW; frequencies and sche-
tropospheric bending. Convection in coastal PROPAGATION PREDICTION dule are listed in every QST.
areas in warm weather; rapid cooling of Ability to tune to 5 MHz and multiples
Information on the prediction of maxi-
the earth after a hot day, with upper air thereof, to receive the standard time-
mum usable frequencies (mut) and opti-
cooling more slowly; warming of air aloft and-frequency stations now operating in
mum working frequencies for F-layer pro-
with the summer sunrise; subsidence of pagation was formerly available from the many parts of the world, is a great aid. See
cool moist air into valleys on calm U.S. Government Printing Office. The Table 1. Most such stations operate
summer evenings - these familiar situa- material took several forms, as methods continuously, with appreciable power and
tions create upper-air conditions which omnidirectional antennas. WWV and
developed for military communications
can extend normal vhf coverage. WWVH are excellent indicators, at any
use were adapted to worldwide civilian
The alert vhf enthusiast soon learns to suitable distance from Colorado or Hawaii.
needs. Though the service was terminated
correlate various weather signs and Their signal behavior can tell the expe-
in 1975, the basic methods are still of
propagation patterns. Temperature and rienced observer at least as much about
interest. A full description may be found
barometric-pressure trends, changing cloud in QST for March, 1972. 8 The govern- propagation - at the moment - as does
formations, wind direction, visibility and the content of their propagation bulletins.
ment information is available in some
other natural indicators can give him clues Many receivers can be made to tune some
technical libraries. The propagation charts
as to what is in store in the way of tropos- which appear regularly in QST are of these frequencies by detuning their
pheric propagation. front-panel tracking controls. See Septem-
computer-derived from information simi-
The 50-MHz band is more responsive lar to that described in reference 8. ber 1975 QST, page .23, for suggestions.
to weather effects than 28, and 144 MHz Other means are available to amateurs Simple crystal-controlled converters for
is much more active than 50. This trend who wish to make their own predictions, the standard frequencies offer another
continues into the microwave region, as both short- and long-term. An appreciable possibility (June 1976 QST, page 25).
evidenced by tropospheric records on all amount of observing and record-keeping
our bands, up to and including work over time is involved at first, but the work can Recurring Phenomena
many long paths on 10,000 MHz. be streamlined with practice. Many Because the sun is responsible for all
amateurs who try it find the task almost ,radio-propagation variables, their rhyth-
The Scatter Modes mic qualities are related to time, season
as interesting as any operational suc-
Though they provide signal levels too cess that may result from it. Properly and other sun-earth factors. Some are
low for routine communication, several ,organized, data collection and propa- obvious. Others, particularly the rota-
scatter modes attract the advanced vhf gation prediction can become an ideal tional period of the sun, about 27.5 days,
operator. group project. show best in long-term chart records kept
Tropospheric scatter offers marginal on a monthly or four-week basis. Recur-
communication up to 500 miles or so, Getting Started rence data are used in nearly all prediction
almost regardless of conditions and Because most factors have well-defined work done presently, and the data can
frequenc):, when optimum equipment and cyclical trends, the first step in pro- yield fair accuracy.
methods are used. pagation prediction is to become familiar If the muf is high and conditions are
Ionospheric scatter is useful mainly on with the rhythm of these trends for the generally good for several days, a similar
50 MHz, where it usually is a composite of geographical location and season under condition is likely to prevail four weeks
meteor bursts and a weak residual scatter later, when the same area of the sun will
signal. The latter may be heard only when 'Hall, "High-Frequency Propagation Estimations be in view from the earth. Ionospheric
optimum conditions prevail. The best for the Radio Amateur," QST, March 1972. disturbances also generally follow the
W.v. Prop8g8tlon 18·8
cies at o.ther o.bservato.ries. The 2800-MHz
flux is given because this value co.rrelates
Table 1
well with the io.nizatio.n density o.f the
Some time and frequency stations useful for propagation monitoring. io.no.spheric F regio.n, with the Zurich
Call Frequency (kHz) Location sunspo.t number (see Fig. 1), and with the
WWV 2500, 5000, 10,000, 15,000, 20,000 Ft. Collins, Colorado maximum usable frequency, fo.r lo.ng-
WWVH Same as WWV, but no 20,000 Kekaha, Kauai, Hawaii distance co.mmunicatio.n.
CHU 3330, 7335, 14,670 Ottawa, Ontario, Canada • A so.lar flux value o.f 66 represents
RID 5004,10,004,15,004 Irkutsk, USSR·
RWM 4996,9996, 14,996 Novosibirsk, USSR· "quiet sun" co.nditio.ns. It will be heard
tuo 2500, 5000 Pretoria, South Africa o.nly in the years near the minimum o.f the
VNG 7500 Lyndhurst, Australia so.lar cycle, mo.st recently 1975 and 1976.
BPV 5000, 10,000, 15,000 Shanghai, China At such times any flux variatio.n is wo.rthy
JJY 2500,5000, 10,000, 15,000 Tokyo, Japan
LOL 5000, 10,000, 15,000 Buenos Aires, Argent!na o.f no.te, as it will pro.duce an o.bservable
change in lo.ng-distance pro.pagatio.n in
·Call, from international table, may not check with actual reception. Locations and frequencies the hf range. Large increases in flux values
appear to be as given. co.me fro.m large new active areas o.n the
sun, which will be apparent to' the regular
viewer o.f the sun. The sho.rt-term effects
27-day cycle, tho.ugh there may be marked anno.uncement at 18 minutes after each o.f flares do. no.t appear in the so.lar flux
differences in level fro.m o.ne perio.d to. the ho.ur. values, unless they are very large and
next. The solar flux is a measure o.f the sun's numero.us. Flares are men tiD ned in the
So.me so.lar-activity centers are sho.rt- radiatio.n at 2800 MHz, taken daily at bulletins, if they are majo.r events capable
lived, lasting less than a full ro.tatio.n. 1700 UTC in Ottawa. Similar measure- o.f affecting radio. pro.pagatio.n.
Others go. o.n and o.n, reco.gnizable fro.m ments are made o.n many different frequen- A steady rise Dr fall in so.lar flux will
their pro.pagatio.n effects fo.r a year Dr
mo.re. Recurring pheno.mena are mo.re
apparent in the lo.w-activity years o.f the 235
so.lar· cycle, mo.st o.f them being far enD ugh . 230
apart to. be clearly identifiable. In April 225
and May, 1976, fo.r example. there were
three well-separated areas affecting radio.
propagatio.n. All were o.f "the Did cycle."
There were also. three new-cycle areas seen
briefly, but with no. reco.gnizable radio.- 200
2'5
The Natio.nal Bureau o.f Standards
radio. statio.n WWV, Ft. Co.llins, CO,
transmits ho.urly bulletins o.n so.lar activity
and the co.nditio.n o.f the earth's magnetic
field. This info.rmlj.tio.n is essentially
current, and thus .invaluable to. any
student o.f radio. pro.pagatio.n. Putting it
into. simple chart fo.rm and using it in
co.njunctio.n with visual o.bservatio.n o.f the
sun. pro.vides an excellent base fo.r
pro.pagatio.n understanding and predic-
'85
tio.n. The WWV (and WWVH) signals
'80
themselves are also. very useful in gather- 175
ing pro.pagatio.n data, as the statio.ns are 170
o.n the air co.ntinuo.usly, using co.nstant
po.wer levels and o.mnidirectio.nal anten-
nas. It sho.uld be stated here that WWVH 215
(Kekaha, HI) do.es no.t carry the bul- 210 FEB.---j
letin service, but its signals pro.vide 205
valuable pro.pagatio.n evidence. Other 200
time-and-frequency statio.ns, so.me o.f '95
similarly. '85
'80
In o.rder o.f their I'resentatio.n, the
'75
WWV bulletins give the so.lar flux and A-
170
index fo.r the previo.us Universal Time
'65
(UT) day; the Bo.ulder K-index (changed 27 7 "'3 15 17 ,9 2' 23 25 27 29 3'
every 6 ho.urs); the state o.f so.lar activity
and the co.nditio.n o.f the earth's magnetic
Fig. 5 - Graphs of 2800·MHz solar flux values for the first three months of 1979, approaching the
field for the previo.us 24 hours, and the peak of Cycle 21. The same period in 1976, near the bottom of Cyc,le 20, had values between 68
same facto.rs predicted fo.r the co.ming 24 and 85. The dates are arranged so that days one solar rotation (27.5 days, average) apart line up
ho.urs. The bulletin starts after the time vertically. Recurring solar phenomena are clearly apparent.
18·7 Chapter 18
show clearly in radio propagation effects, relates to fast-changing conditions that active. These three words ratheraptiy
and also in the observer's view of sunspot affect propagation a..dversely. Activity is describe the propagation conditions as-
activity. Trends either way are important given as very low, low, moderate, high or sociated with them and with the K indices.
to the propagation student. They often very high. Often any K value of 4 or higher will be
run for several days, during which the Geomagnetic field conditions are termed associated with a "geomagnetic storm,"
associated changes in muf, and in the quiet, unsettled or active. These relate usually described as "minor" or "major."
duration and' geographical distribution of roughly to K indices of 0 to 1 for quiet, 1 Either is bad news for the amateur interest-
openings on frequencies above about 15 to 3 for unsettled, and 4 or higher for ed in high-latitude hf communication.
MHz, are easily observed. Flux readings
of 80 or higher will make'the 21- and 32
28-MHz bands come alive, even near the
bottom of the solar ·cycle.
28
24
r- m ,-
In the intermediate years of the cycle, as 20
usable at all. 50
The K-index is similar to the A-index, 46
Boulder's. 40 APRIL~
~
28
1800 UT. It represents the conditions
24
during the last three hours just before the 20
bulletin's time of issue. Thus, it is close t6
'6
a now statement ·of a factor of vital 12
importance to any user of the hf radio
spectrum. It may interest vhf operators as 4
well, When the values go above 3. The 0
24
trend is important - a rising trend means
degraded hf propagation; values of 4 and
up may mean auroral conditions on the Fig 6 - A-index information transmitted by WWV for the first four months of 1979, arranged to show
effects of recurring solar phenomena, as in' Fig. 5. Recurrence dates vary because of the varying
vhf bands. travel time of charged particles from the sun that cause geomagnetic disturbances on Earth.
Solar Activity, as the term is used in the A·indices of 10 and lower go with good hf propagation. Peaks above 20 nearly always are accom·
final portion of the WWV bulletins, panied by high signal absorption: Auroral effects are common in high latitudes at these times.
Transmission Lines
TransmisSion lines, and the theory section along the line), only two para- circuitry, it is convenient to be able to
behind them, play an important role in meters are required to express the line express the input impedance of a line in
many phases of radio communication. properties completely. These ar~ the phase terms of the output or load impedance. A
This is because the basic principles velocity, Vp and the characteristic impe- line treated this way is then similar to a
involved can be applied to a wide variety dance, Zo. If the line can be considered filter or matching network with a given
of problems. Types of transmission lines lossless as well, Zo becomes a pure resis- load impedance. One caution should be
include simple two-conductor configura- tance, Ro. kept in mind in applying such relations
tions such as the familiar coaxial cable Assume that a very short burst of power and that is the manner in which the source
and TV p~rallel-wire line. Such lines are is emitted from the source. This is and load are connected to the line can be
useful from power frequencies to well up represented by the vertical line at the left important. There are always some "para-
into the microwave region and form of the series of lines in Fig. 2. As the pulse sitic" effects arising from connectors and
perhaps the most important class. The voltage appears across the load Za, all the post-connector circuit configuration that
waveguide is representative of a second energy may be absorbed or part of it may may cause the line to "see" a different
type. Here, the conductor configuration is be reflected in much the same manner impedance than if measurements were
rather complex and ordinary concepts energy in a wave in water is reflected as made at the load terminals directly. This is
such as voltage and current tend to' the wave hits a steep breakwater or the indicated by the abrupt change in line
become obscure. As a consequence, end of a container. This reflected wave is dimensions at points a and g in Fig. 1.
.various parameters are expressed in terms represented by the second line in the series Even thougq the short line connecting the
of the electric and magnetic fields and the arrow above indicates the generator to the main transmission line
associated with the line. Finally, the direction of travel. As the latter wave (and the one connecting the load to the
propagation of electromagnetic energy reaches the source, the process is again line) might have the same characteristic
through space itself is closely related to repeated with either all of the energy being
similar phenomena in wave guides and absorbed or partially reflected.
transmission lines. In fact, the only The back-and-forth cycle is actually an
significant physical difference is that the infinite one but after a few reflections, the
Zg
power density in a wave propagated in intensity of the wave becomes very small.
space decreases with increasing distance If, instead of a short 1JUlse, a continuous
while it is possible to transmit power over voltage is applied to the terminals of a
Za
long distances with conventional lines transmission line, the voltage at any point
with little attenuation. This is because along the line will consist of a sum of
p~wer flow is essentially confined to one voltages of the composite of waves
dimension in the latter lease while the traveling toward the right and a com-
three-dimensional aspect of space does posite of waves traveling toward the left.
not permit such confinement. The total sum of the waves traveling Fig. 1 - Source and load connected by means of
toward the right is called the forward wave a transmission line.
Transmission Lines and Circuits or incident wave while the one traveling
A transmission line differs from an toward the left is called the reflected wave.
ordinary circuit in one very important Provided certain conditions concerning Z a
aspect. Delay effects associated with the are met, there will be a net flow of energy
finite propagation time of electromagnetic from the source to the load, with a frac-
energy are often neglected in network tion of the energy being stored in the
design since the dimensions involved are "standing" waves on the line. This
normally small compared to the wave- phenomenon is identical to the case of a 5
length of any frequencies present in the coupled resonator with ordinary circuit
circuit. This is not true in transmission- elements. Sections of transmission line are T·IIVP.
line considerations. The finite propaga-
tion time becomes a factor of paramount
importance. This can be illustrated with
often used for this purpose, especially in
the vhf/uhf region. The duplexer found in
many vhf repeaters is a common example.
0
0 T 2"T
i rTT-
3f 4T 5T Sf 7T aT"
19-1 Chapter 19
impedance, if the sizes are different a Fortunately, it is possible to determine Since the foregoing transmission-line
mismatch will still occur. Normally, this the input resistance and reactance of a equations are somewhat awkward to work
effect can be neglected at hf but becomes terminated line if the load resistance and with, various plots have been devised that
important as the frequency of operation is reactance are known, along with the line permit a graphical solution. However,
extended into the vhf region and above. length and characteristic impedance. (With with modern programmable calculators,
In referring to the previous example actual lines, the physical length must be even those in a moderate price class, it
shown in Fig., 2, the ratio of the voltage in divided by the velocity factor of the cable takes approximately four secol').ds to solve
the reflected wave to that of the voltage in which gives the value of I in the following both equations. The plots shown in Fig.
the incident wave is defined as the voltage formula.) The equations are 3A and Fig. 3B were computed in this
reflection coefficient designated by the manner. The curves are for rin and Xin for
Greek letter, r , or by p. The relation r.(1 + tan 2f31) various values of ra (Xa equal to 0) and liQ,e
between the output resistance, Ra, the r in = "(1'--x--:-ta-n-:fJ;:;'I):";2-:+-:(:-r.-:-ta-n-:fJ;:;'\):';'2 length in degrees. Note that 90 degrees
output reactance, Xa, the line impedance, . a
appears to be a "critical" value and
Zo, .and the magnitude of the reflection co- . represents a line length bf a quarter
efficient is wavelength. As this value is approached,
the transmission-line equations can be
approximated by the formulas:
(Ra - RoP +Xa 2
r= (Ro +Rap+Xa 2 for a I-ohm line. Equations are often
"normalized" this way in order to make
Note that if Ra is equal to R o ' and if Xa is
o the reflection coefficient is 0, which
universal tables or plots that cover a wide
range of values. If characteristic impe-. x·
-x.
=---
represents "matched" conditions. All the
In r/ +xl
dances (Zo) other than 1.0 are to be used, ,
energy in the incident wave is transferred the following set of conversions apply If Xa is zero, the formula for a quarter-
to the load. In effect, it was as if there where Ra and Xa are the load resistance wavelength transformer is obtained:
were an .infinite line of characteristic and re.actance and Rin and Xin represent
impedance Zo connected at a. On the the resistance and reactance at the input
other hand, if Ra is 0, regardless of the end of the line. .
value ofXa the reflection coefficient is 1.0.
This means all the power is reflected in ,Quite often, it is mistakenly assumed that
much the same manner as radiant energy power reflected from a load represents
is reflected from a mirror. . power "lost" in some way. This is only
If there are no reflections from the load, true if there is considerable loss in the line
the voltage distribution along the line is itself and the power is dissipated on the
constant or "flat" while if reflections way back to the source. On the other
. exist, a standing-wave pattern will result. hand, the quarter-wavelength transformer
The ratio of the maximum voltage on the is an example where reflections on a
line to the n1inimum value (provided the In order to determine the value of the lossless line can actually be used to
line is longer than a quarter wavelength) is tangent functio~, either the line length in advantage in matching a load impedance
defined as the voltage standing-wave ratio meters or feet, along with.the frequency in that is different from the source im-
(VSWR). The VSWR is related to the MHz, can be substituted into the fol- pedance.
reflection coefficient by lowing expressions: If the terminating resistance is zero, the
input resistance is also zero. In effect, the
line and load act as a pure reactance
VSWR = ..!....±.L
I - r fJl(degrees)= l.2f(MHz) X I (melers) which is given by the formula:
x.+tanf31
This latter definition is a more general· 131 (degrees)= 0.367f(MHZ)?< I(reel) l-x.tanfJl
one, valid for any line length. Quite often,
the actual load impedance is unknown.
An alternate way of expressing the
100
reflection coefficient is
10
where
Pr = the power in the reflected wave and XIN
•
measure with power meters available
commercially or with homemade designs. 0.1
However, it is obvious there can be no
other power sources at the load if the
foregoing definition is to hold. For
0.01+--,--~-~-,-----.----,--+ 0.01 -+-'::::::.....,--'-._---,-_.------._---,-_+1
instance, the reflection coefficient of the 20 30 40 50 60 70 eo 90 20 30 40 50 60 70 eo 90
generator in the example shown in Fig. 2
is 0.9. This -value could have been ob- /I..(DEGREES) ,e..(OEGREES)
1.6x
3.41r
3.2r
.appears as a capacitance. Shortest wavelength
before next mode
Waveguides becomes possible 1.1x aSr
A waveguide is a conducting tube ELECTRIC FIELD INTENSITY
?-/2
MATCHING 1
SECTION......... 1../4
1
TRANS. LINE
Z=~
CANCELLING PLATE
where ZI = antenn\l impedance
?-/4 STUB Zo= characteristic impedance of the
line to which it is to be matched.
Fig. 6 - At its cutoff frequency a rectangular waveguide can be analyzed as a parallel two· Example: To match a 600-'Ohm line to
conductor transmission line supported from top and bottom by infinitely many quarter·wavelength an antenna presenting a 72-ohm load, the
stubs.
quarter-wave matching section would
require a characteristic impedance of
3
/r/
//
V /""
.-"
V
self-resonant at the median operating' / 1/ y L-L- >--
I-or·<-----~-- ),./2 -------1 frequency. The length is usually,not highly
critical, because a folded dipole tends to
2 ,/
V V"'"
5
I I :::: t::: l-
I-
have the characteristics of a "thick" :;;
~ 1.0
- t-; RAtIO-4
~~I::
r "
11 is based on the fact that the impedance 0.2
between' any two points along a resonant "-
II in
I"
antenna is resistive, and has a value which T
depends on the spacing between the two o.1 I FEED POINT
(C)
1.5 2 3 4 5 7.5 10 20
points. It is therefore possible to choose a S/d2
into two' or more conductors. drawing) forms with the antenna con- 0.3
t-
The ratio by which the input impedance ductor opposite it a short section of 4 d2 h 6
of the antenna is stepped up depends not
0.2 r 1
transmission line. Each of these transmis- +-====, f==4 dl
=i
only on the number of conductors in the
folded dipole but also on their relative
sion-line sections can be considered to be
terminated in the impedance that exists at 2
i.L
3
EEED POINT
4 5 8 10
d2
15 20
130 50
diameters, since the distribution of cur- the point of connection to the antenna. S/d2
rent between conductors is a function of Thus the part of the antenna between the
,. their diameters. (When one conductor is two points carries a transmission-line Fig. 10 - Impedance transformation ratio,
larger than the other, as in Fig. 8, the current in addition to the normal antenna three·conductor folded dipole. The dimensions
larger one carries the greater current.) The current. The two transmission-line d1, d2 and sare shown on the inset drawing.
ratio also depends, in general, on the Curves show the ratio of the impedance
matching sections are in series, as seen by (resistive) seen by the transmission line to the
spacing between the conductors,. as shown the main transmission .line. radiation resistance of the resonant antenna
by the graphs of Figs. 9and 10. An impor- If the antenna by itself is resonant at the system.
tant special case is the two-conductor operating frequency, its impedance will be
dipole with conductors of equal diameter; purely resistive. In such case the mat-
as a simple antenna, not a part of a direc- ching-section lilles are terminated in a ·1
tive array, it has an input impedance close •I ),./2
tained from Fig. 9. Similar information well as resistive. This prevents a perfect (A)
for a three-conductor .dipole is given in match to the main transmission line, since
Fig. 10. This graph applies if all three con- its load must be a pure resistance for I· ),./2 -I
ductors are in the same plane. The two
conductors not connected to the transmis-
sion line must be equally spaced from the
perfect matching. The reactive component
of the input impedance must be tuned out
before a proper match can be secured.
~'3:
LtNE . T
fed conductor, and must have equal One way to do this is to detune the
diameters. The fed conductor may have a antenna just enough, by changing its (B)
different diameter, however. The length, to cause reactance of the opposite
unequal-conductor method has been kind to be reflected to the input terminals Fig. 11 - The "T" match and "gamma" match.
19-5 Chapter 19
of the matching section, thus cancelling
the reactance introduced by the latter. i'4"-----}../2 --c---~
.. I
Another method, which is considerably
ea~ier to adjust, is to insert a variable
capacitor in series with the matching
section wher!l it connects to the transmis- 14-1"
~~~-}"/2---~'1
sion line, as shown in chapter 20. The.
capacitor must be protected from the wea-
A-COAX~AL
ther. I
LINE SHORTED TO
The method of adjustment commonly cOAX OUTER
CONDUCTOR
used is to cut the antenna for approximate HERE
resonance and then make the spacing x
(A) (8).
some value that is convenient construc-
tionally. The distance y is 'then adjusted, 1-01"r-----}../2----..,. 1
'~'----~ZO=2~ZI_,~.",
doing this are called baluns. The types
used· between the antenna and transmis-
sion line are generally linear, coflsisting of
transmission-line sections.
The need for baluns also arises in
coupling a transmitter to a balanced A
transmission line, since the output circuits B
of most transmitters have one side PARALLEL-
CONDUCTOR
grounded. (This type of output circuit is LINE
desirable for a number of reasons, C
including TVI reduction.) The. most D
19·9 Chapter 19
balanced transmission line are shown in
Fig. 20. Thl: inductor Ll should have a 2CS
TO TO
reactance of about 60 ohms when ad- TRANS, TRANS.
justable inductive coupling is used (Figs.
20A and20B). When a variable series
LOW Z
capacitor is used, L I should have a reac-
tance of about 120 ohms. The variable
capacitor, C I, should have a reactance at
maximum capacitance of about 100 ohms.
On the secondary side, Ls and C s should (AJ (BJ
be capable of being tuned to resonance at
about 80 percent of the operating fre- 2CS
quency. In the series-tuned circuits, for a TO TO
'''~ "1)
given low-impedance load, looser coup-
ling can be used between L I and Ls as the
Ls-to-Cs ratio is increased. In the parallel- LS LOW Z LP CP
HIGH
Z
tuned circuits, for a given high-impedance Ct c.
load looser coupling can be used between
Ll and Lp as the Cp-to-Lp ratio is in-
creased. The constants are not critical; the
rules of thumb are mentioned to assist in
correcting a marginal condition where (CJ (OJ
sufficient transmitter loading cannot be
obtained. Fig, 20 - Simple circuits for coupling a transmitter to a balanced line that presents a load dif-
Coupling to a coaxial line that has a ferent than the transmitter output impedance. (A) and (8) respectively are series- and parallel-
high SWR, and that consequently may tuned circuits using variable inductive coupling between coils, and (C) and (0) are similar but use
present a transmitter with a load it cannot fixed inductive coupling and a variable series capaCitor, C1. A series-tuned cfrcuit works well with
a low-impedance load; the parallel circuit is better witlrhigh-impedance loads (several hundred
couple to, is done with an unbalanced ver- ohms or more).
sion of the series-tuned circuit, as shown
in Fig. 21. The rule given above for
coupling ease and Ls-to-C s ratio applies to
these circuits as well.
The most satisfactory way to set up ini-
tially any of the circuits of Fig. 20 or 21 is
to connect a coaxial SWR bridge in the
line to the transmitter, as shown in Fig.
21. The "Monimatch" type of bridge,
which can handle the full transmitter
power and may be left in the line for con- (A) (8)
tinuous monitoring, is excellent for this
purpose. However, a simple resistance
bri,Pge is perfectly adequate, requiring Fig. 21 - Coupling from a transmitter designed for 50- to 75-ohm output to a coaxial line with a 3 '
only that the transmitter output be re- or 4:1 SWR is readily accomplished with these circuits. Essential difference between the circuits
duced to a very low value so that the is (A) adjustable inductive coupling and (B) fixed inductive coupling with variable series capaCitor.
bridge will not be overloaded. To adjust In either case the circuit can be adjusted to give a 1:1 SWR on the meter in the line to the
. the circuit, make a trial setting of the transmitter. The coil ends marked "x" should be adjacent, for minimum capacitive coupling .
coupling (coil 'spacing in Figs. 20A and B
and 21A, CI setting in others) and adjust
Csor Cp fo'r minimum SWR as indicated tuning that is critical with frequency. The could be used to give the transmitter a
by the bridge. If the SWR is not close to bridge method is simple and gives the 50-ohm load, even though a significant
practically I: I, readjust the coupling and optimum operating conditions quickly mismatch was present at the antenna feed
return Cs or Cp ' continuing this procedure and with certainty. point. It is important to remember that
until the SWR is practically I: 1. The set- the Transmatch will not correct the actual
tings may then be logged for future A Transmatch for Balanced or SWR condition; it only conceals it as far
reference. Unbalanced Lines as the transmitter is concerned. A
In the series-tuned circuits of Figs. 20A Most modern transmitters are designed Transmatch is useful also when using a
and 20C, the two capacitors should be set to operate into loads of approximately 50 single-wire antenna for multiband use. By
at similar settings. The "2Cs'" indicates .ohms. Solid-state transmitters produce means of a balun at the Transmatch out-,
that a balanceq series-tuned coupler re- progressively lower output power as the put it is possible to operate the transmitter
quires twice the capacitance in each oftwo SWR on the transmission line increases, into a balanced transmission line, such as
capacitors as does an unbalanced series- owing to the built-in SWR protection cir- a 300- or 6OO-ohm feed system of the type
tuned pircuit, all other things being equal. cuits. Therefore, it is useful to employ a that would be used with a multi band
It is possible to use circuits of this type matching network between the transmitter tuned dipole, V beam or rhombic anten-
without initially setting them up with an and' the antenna feeder when antennas na.
SWR bridge. In such a case it is a matter with complex impedances are'used. One A secondary benefit can be realized
of cut-and-try until adequate power example of this need can be seen in the from Transmatches of certain varieties:
transfer between the amplifier and main case of an 80-meter, coax-fed dipole The matching network can, if it has a
transmission line is secured. However, this antenna which has been cut for resonance bandpass response, attenuate harmonics
method frequently results in a high SWR at, say, 3.6 MHz. If this antenna were from the transmitter. The amount of at-
in the link, with consequent power loss, used in the 75-meter phone band, the tenuation is dependent upon the loaded Q
"hot spots" in the coaxial cable, and SWR would be fairly higp. A Transmatch (QL) of the network after the impedance
Transmission Lines 19-10
J1
J3
DUMMY
LOAD
XMTR
(50/\.) 200
•
SAL.
(8)
has been matched. The higher the QL the cuit of Fig. 23, it can be seen that at high Fig. 26. The example at A shows the se-
greater the attenuation. Some Trans- load impedances, the Ultimate cond harmonic down only 14 dB from the
matches, such as the "Ultimate Transmatch will have minimal effective fundamental when looking into 1000
Transmatch" of Fig. 23, can exhibit a output capacitance in shunt with the in- ohms with the Ultimate. The display at B
high-pass response (undesirable), depend- ductor, giving rise to a high-pass response. of Fig. 26 shows the SPC response at 1000
ing upon the transformation ratio they Another advantage of the SPC ohms with the second harmonic down 28
are adjusted to accommodate. In a worst- Transmatch is its greater frequency range dB from the fundamental energy. The
case condition the attenuation of har- with the same component values used in response at A resembles a high-pass
monic currents may be as low as 3 to 6 dB. the Ultimate Transmatch. The circuit of characteristic.
Under different conditions of impedance Fig. 24 operates from 1.8 to 30 MHz with
transformation (better) the attenuation the. values shown. Only 3/4 of the avail- Construction
can be as great as 20 to 25 dB. able inductance of Ll is needed on 160 Figs. 22 and 25 show the structural
The "SPC Transmatch" described here meters. details of the Transmatch. 'fhe cabinet is
was developed to' correct for the The notable difference in outward per- homemade from 16-gauge aluminum
sometimes poor harmonic attenuation of formance over the circuit in Fig. 23 is sheeting. L brackets are affixed to the
the T network which is used in the somewhat sharper tuning. This is because right and left sides of the lower part of the
Ultimate Transmatch. The SPC (series- of the increased . network Q. This is cabinet to permit attachment of the
parallel capacitance) circuit maintains a especially prominent at 40, 80 and 160 U-shaped cover.
bandpass response under load conditions meters. For this reason there are vernier- The conductors which join the com-
of less than 25 ohms to more than 1000 drive dials 'on Cl and C2. They are also ponents should be of heavy-gauge
ohms (from a 50-ohm transmitter). This is useful in logging the dial settings for material to minimize stray inductance and
because a substantial amount of changing bands or antennas. heating. Wide strips of flashing copper are
capacitance is always in parallel with the Spectrographs of an Ultimate suitable for the conductor straps. The
rotary inductor (C2B andLl of Fig. 24). Transmatch and the SPC Transmatch center conductor and insulation from RG-
In comparison with the "Ultimate" cir- bandpass characteristics are shown in 59/U polyfoam coaxial cable is used in
19·11 Chapter 19
Fig. 25 - Interior view of the W1 FS SPC
Transmatch. L2 is mounted on the rear wall by
means of two ceramic standoff insulators. C1
is on the left and C2 is at the right. The coaxial
connectors, ground post and J3 are on the
lower part of the rear panel.
Fig. 27 - Exterior view of the band-switched link coupler. Alligator clips are used to select th~
proper
. tap positions
,
of the coil.
All leads should be kept as short as between 12 and the arm of SIA (Fig.
possible to help prevent degradation of 24A). If this is done there'should be room
the circuit Q. The stators of Cl and C2 for an edgewise panel meter above the ver-
should face toward the cabinet cover to nier drive for C2.
minimize the siray capacitance between Initial adjustment should be done with
the capacitor plates and the bottom of the a dummy load attached to J I, and with S1
cabinet (important at the upper end of the in the D position. This will prevent in-
Transmatch frequency range). Insulated terference which could otherwise occur if
ceramic shaft couplings are used between tuning is done "on the air." Set Cl and
the vernier drives and Cl and C2, since C2 at midrange. Adjust Ll for a decrease
the rotors of both capacitors are in reflected power. Then adjusrCl and C2 •
"floating" in this circuit. Cl add C2 are alternately for the lowest possible SWR
supported above the bottom plate on condition. If the SWR can not be brought
steatite cone insulators. S 1 is attached to to 1: I, adjust Ll slightly and repeat the
the rear apron of the cabinet by means of tuning procedure until an SWR of 1 is ob-
two metal standoff posts. tained. Next, reduce transmitter power to
a few watts, switch SI to T (Trans match)
Qperation and readjust the network for an SWR of 1
The SPC Transmatch is designed to with the antenna connected. Make sure
handle the output from transmitters the frequency is not in use during this
which operate up to 2 kW PEP. L2 has tuneup period, and identify your1station.
been added to improve the circuit Q at 10 The operating power can now be in-
and 15 meters. However, it may be omit- creased to the normal level.
Fig. 26 - Spectrographs of the response
characteristics of the Ultimate Transmatch (A)
ted from the circuit if the rotary inductor The air-wound balun of Fig. 24B can be
and the SPC Transmatch (B) looking into a (L 1) has a tapered pitch at the minimum- used outboard from the Transmatch if a
1000-ohm termination from a 50-ohm signal inductance end. It may be necessary to low-impedance balanced feeder is con-
source. Greater harmonic attenuation is ob- omit L2 if the stray wiring inductance of templated. Ferrite or powdered-iron core
tained with the SPC Transmatch (see text). The
scale divisions are 2 MHz horizontal and 10 dB
the builder's version is high. Otherwise, it material is, not used in the interest of
vertical. The fundamental frequency is 8 MHz. may be impossible to obtain a matched avoiding TVI and harmonics which can
condition at 28 MHz with certain loads. result from core saturation.
this model for the wiring between the An SWR indicator is used between the The B position of S,1 permits switched-
switch and the related components. The transmitter and the Transmatch to show through operation when the Transmatch
insulation is sufficient to ,prevent when a matched condition is achieved. is not needed. The G position is used for
breakdown and arcing at 2 kW PEP input The builder may want to integrate an grounding the antenna system, as
to the transmitter. SWR meter in the Transmatch circuit neces&ary; a quality earth ground should
Transmission Lln.s 19·12
be attached at all times to the Transmatch
. chassis. TO BALANCED FEED LINE
Final Comments
Surplus coils and capacitors are okay in
ihis circuit. .L1 should have at least 25 uH
of inductance, and the tuning capacitors
need to have 150 pF or more of
capacitance per section. Insertion loss
through this Transmatch was measured at
less than 0.5 dB at 600 watts of rf power
on 7'MHz.
19·13 Chapter 19
Table 1
Characteristics of Commonly Used Transmission Lines
pF Max
Zo Vel per Dlel. Operating Volts
Type of line Ohms % foot 00 Material (RMS)
RG·8/U 52.0 66 29.5 .405 PE 4,000
RG-8/U Foam 50.0 80 25.4 .405 Foam PE 1,500
RG·8A1U 52.0 66 .29.5 .405 PE 5,000 TUlING
0'"
RG·9/U 51.0 66 30.0 .420 PE 4,000
RG·9A/U 51.0 66 30.0 .420 PE 4,000
RG·9B/U 50.0 66 30.8 .420 PE 5,000
RG·11/U 75.0 66 20.6 .405 PE '4,000
RG·11/U Foam 75.0 80 16.9 .405 Foam PE 1,600
RG·11A/U 75.0 66 20.6 .405 PE 5,000
RG·12/U 75.0 66 20.6 .475 PE 4,000
RG·12A/U 75.0 66 20.6 .475 PE 5,000
RG·17IU 52.0 66 29.5 ,870 PE 11,000
2 3 4~67119tO
RG·17A/U 52.0 66 29.5 .870 PE 11,000
RG·55/U 53.5 66 28.5 .216 PE 1,900 CEN~"-TO-CENT£" SPAClIHI (·IIICHIS'
RG·55A/U 50.0 66 30.8 .216 PE 1,900
RG·55B/U 53.5 66 28.5 .216 PE 1,900
RG·58/U 53.5 66 28.5 .195 PE 1,900 Fig. 30 - Chart showing the characteristic im-
RG·58/U Foam 53.5 79 28.5 .195 Foam PE 600 pedance of spaced-conductor parallel trans-
RG·58A/U 53.5 66 28.5 .195 PE 1,900 mission lines with air dielectric. Tubing sizes
RG·58B/U 53.5 66 28.5 .195 PE 1,900 given for outside diameters.
RG·58C/U 50.0 66 30.8 .195 PE 1,900
RG·59/U 73.0 66 21.0 .242 PE 2,300
RG·59/U Foam 75.0 79 16.9 .242 Foam PE 800
RG·59A/U 73.0 66 21.0 .242 PE 2,300
RG-62/U 93.0 86 13.5' .242 Air Space PE 750 transmissIOn lines used by amateurs'.
RG-62/U Foam 95.0 79 13.4 .242 Foam PE 700 Open-wire line has a velocity factor of
RG-62A1U 93.0 86 13.5 .242 Air Space PE 750 essential unity because it lacks a substan-
RG·62BIU 93.0 86 13.5 .242 Air Space PE 750
RG·133A/U 95.0 66 16.2 .405 PE 4,000
tial amount of solid insulating material.
RG·141/U 50.0 70 29.4 .190 PTFE 1,900 . Conversely, molded 300-ohm TV line has
RG·141A/U 50.0 70 29.4 .190 PTFE 1,900 a velocity factor of 0.80 to 0.82. The
RG·142/U 50.0 70 29.4 .206 PTFE 1,900 higher cost of the larger coaxial lines is
RG·142A/U 50.0 70 29.4 .206 PTFE 1,900
RG·142B/U 50.0 70 29.4 .195 PTFE 1,900
often worth the expenditure in terms of
RG·174/U 50.0 66 30.8 .1 PE 1,500 reduced feeder losses.
RG·213/U 50.0 66 30.8 .405 PE 5,000 Amateurs can construct their own
RG·215/U 50.0 66 30.8 .475 PE 5,000 parallel transmission lines by following .
RG·216/U 75.0 66 20.6 .425 PE 5,000 the chart contained in Fig. 30. When using
Aluminum Jacket wire conductors it is an easy matter to
Foam Dielectric fabricate open-wire feed lines. Spacers
1/2 inch 50.0 81 25.0 .5 2,500
made of high-dielectric material need to
3/4 inch 50.0 81 25.0 .75 4,000 be affixed to the conductors at ap-
7/8 inch 50.0 81 25.0 .875 4,500 propriate distances apart to maintain the
1/2 inch 75.0 81 16.7 .5 2,~ spacing between the wires (constant im-
3/4 inch 75.0 81 16.7 .75 3,500
7/8 inch 75.0 81 16.7 .875 4,000
pedance) and to prevent shorting of the
Open wire 97 conductors.
75-ohm trans·
mitting twin lead 75.0 67 19.0 VSWR Nomograph and VSWR/Loss
300·ohm twin lead 300.0 82 5.8 Chart
300-ohm tubular 300.0 80 4.6
The graph displayed in Fig. 32 can be
Open wire, TV type used to determine the VSWR on a
1/2 inch 300.0 95 transmission line when a forward and
1 inch 450.0 95 reflected power measurement are known.
As an example, suppose a forward power
Dielectric Designation Na"le Temperature Limits
measurement is taken as 100 watts and the
PE Polyethylene -65· to +80· C reflected power measurement is 11 watts.
Foam PE Foamed Polyethylene -65· to +80· C
PTFE Polytetrafluoroethylene - 250· to + 250· C The l00-watt line on the horizontal axis is
(Teflon) located and the II-watt line Qn the vertical
axis is noted. The intersection of these two
lines on the graph is at the 2 to I VSWR
line. Therefore the VSWR for this set of
The circuit described here is intended must be taken into account when cutting a conditions is 2 to I.
for power levels up to roughly 200 watts. transmission line to a specific part of a The graph at Fig. 33 provides a con-
Balance was checked by means of two rf wavelength. - such as with a quarter- venient means of determining total losses
ammeters, one in each leg of the feed line. wavelength. transformer. For example, if if the VSWR at either the input or the load
Results showed the balance to be well RG-8A/U were employed to make a is known and if the loss in the line without.
within 1 dB. quarter-wavelength line at 3.5 MHz, the standing waves is known. (This latter fac-
line dimension should be 234M(MHz) X tor may be obtained from Table I of this
Transmission Line Characteristics 0.66. Thus, the line would be 44.1 feet chapter or the manufacturer's literature).
Each type of line has a characteristic (13.44 m) long instead of the free-space Conversely, if the VSWR values at the in-
velocity factor, owing to the insulating- length of 66.85 feet (20.37 m). Table 1 put to the line and at the load are
material properties. The velocity factor shows various velocity factors for the measured with a reliable instrument, the
S
V V~ ~'O ..-< k(Q ~". /
....
V V
/ . . . V VV fe;~ ..;'< A V V
4
r.-, .
3
V
1>'
,1 AV
vV' V
V . . . . V ......
t:1:(.'9V h~Vt:;:/ ~-....
...........:: ~
......V ~ V
V
V
/
V
2. S
.....e; L>- ~G Vv
~<j
~
~ ~
~ V ~~
V VV V
V
1.0 V iA
VV
....e;-jJ.
~ I~
V ....e; &
~ ~ 6
Bs1
~ ~ '~
0.9 k::' ~ ty' ./ /' 1./ 7' / ' / ./.
0.8
o.7
,
V
........-/ -:/' 1>1/
I:/'" 1'~
.. / Il..
:,...--
V V V· V ,;; t7~........ V [,t>S
V V Vy ~ ~ _"t:
~ % 6J;:. ,I>V V -'" ./ t/ V vb:= 'A'? .. Ill.
0.6
o. S
"
~ V V 'l-t/ V' V V V~ v::.V k:: ~y V~
~~
, /V / V
V
,,~
~
V Vi.-- /'
V ~ f::/ V vI:?
0.4
P
~,~ ~ ~ ~ ~ ~ VA l4
V
~
~
V :,...--
~ V
0.3
0.25
,'0. '0 ~
0).
,/ V V / V
v~ V~ ~ ~ && ~
~'
V
0.2
0.1
St:/ y
,
........
V ~ ~ p- k;::b:= ~ ~ I~
V ~.
I?~~ V
V
rV
~-
~,
,
V ~ ~ V V V" ,A
.1
I
.... V
1.S 2 2.S
~ ~ ~ ~V V
3
V
4
,/'
5678910 IS
/ 20 25 30
V 40 50 60 70 80 90100 150 200 250 300 400 500 600 700 900
800 1000
FREQUENCY (MHz)
Fig. 31 - This graph displays the attenuation in dB per 100-foot lengths of many popular transmission lines. The vertical axis represents attenuation
and horizontal axis is frequency, 1 - 1000 MHz.
VSWR NOMOGRAPH
100 + L..-.L
,-- L T t,
-,
-- 7
~ / /
-- REFLECTED POWER;
- 1+ ) 1/ / / 1/
- FORWARD POWER i i 7 7 7 / /
_VSWR = - I / V / / V
[7
V II V
- ,~tf-' l.I 7 7 I....
- 1_ REFLECTED POWER ' ; ' , ' - - f.......cL~L / 1I V/
V '/ V V II
'w - FORWARD POWER)V , /V ,
~
~
« 10 ~ +:-~~t. H'J ji" I , , .. ,,0,
Or'"
~ Vv v ~
V V 1// II
'
l/ 11
~
I
f- ,
1
- - .'"
,H---H=lt
, ., i
tit"
---. -I t~
t ~C:J
(/j ~ ~ ~
'
, ./ v / /
)
1/ ./
v V 17
V
,-olf
0: 0 1/ f./' V
.~·~V
I .
11/
UJ
~
0
1---
f--
!
-i-tt--t
++
1
, -I/.lIIi.,,·
--+--+-
V i7 v
0 VlZf-y 11-.'
f) -/ l71/ l/ j) v
t--- ..:~ , , :t~t- V
I ' i 1/ 1/ / 1/
D- -t--tz, tIL
1--1-
. ~~
... G:> v 1/
. ./ /
o 1/ 1/ / 1/ . ...
..... fJ II V/ l/ V 171/
~
II
UJ
~
0 1.0
I
I/I/V V V Jl 11 VjV 1I1/V,. ,:~ ~ ~ VV. II~ 171/
-~ t- ...~ q"
UJ
...J - r / / ~/I-)L / ./V "~
. ~
1/
IL
UJ 1/ 1/ v / V 17 ' " ...~O/
0: l/ 1/ / / .... 1/ V V VI/ 1/ ....,~
/
1/ / / /
1/ 1/ V 1/ 1/ /
II V II / bIIIf-IL VI/ / / 1I v
V V / 1/ V / V V II V/ V V V II
19·15 Chapter19
total line loss and the loss of tlie line
without standing waves (matched loss)
may be determined from the graph.
10
V, 1I ~ IJ
if [7 /).'
V
The horizontal axis of .the graph is
calibr.ated in' values representing the
VSWR at the load, while the vertical axis 9 -~
V J /
V '7 II "
17
)
II V /
7 [L1
V / / I'
represents total loss of the line in decibels.
The curves that are predominantly vertical ,I/ 1I II II / J
(dotted lines) in the body of the graph
represent the VSWR value at the line in-
put, and the curves that are predominant- 8
.-!- Vt /
V /j /
I
I
IA
1/
'I
~
",'
/
I
v~
I /
ly horizontal in the lower portion of the
graph (solid lines) represent the matched-
V I fJ l0 "
0'" I I
line loss. Interpolation of values may be
made between curves, and the curves,are .3~ v,/ l( ! J
[7 /~y
I
I
LI
~
7
1/ V /
I
1/ / 1/ 11 / ~
interrelated so that each set or family may /
be considered as another "axis" 'of the II j ~J / J /
~ / V I
overall graph.
A couple of examples best illustrate use
6
..-""': / / V r/ / /
/
1/ /
)
V /
l/'~'
of the graph. Suppose 100 feet of RG-8
feed line connects a IS-meter transmitter
I 1.1
A
V V V .
~/
.:) I
and its load. If the VSWR as measured at
the load is 3:1, what is the total line loss
and what is the VSWR at the line input?
ID
'It
I
.J... V /,.3 y
I /
VI
" [V I
W vI
1.,1
I
{-
I
:
5
(I)
11 ~~
1//~J1.5 1>'11
(I)
First, it is necessary to determine the o
matched loss of the length of the coaxial oJ
V .),"'/
.... 1
7 ;'
I
1-Vi
cable. Table 1 indicates that the line has a I V I
loss of 1.0 dB per hundred feet at 21 ...'"
~ 4
I II 1/ /2 // /
MHz. Proceed by- running a finger along
the scale at the bottom of the graph and
locate the value of 3, representing the 3:1
I
I
I
I 3)
V/
V / /
V /
/3
14
I
/ ,./ /
VSWR at the load. Follow the v~rtical
"3" line up until it intersects with the" 1" 3 ,-~I
I /.V
II 7
/
I
/
1I /5 1
6 1/
,,
solid-line curve which represents 1 dB of
matched-line loss. ,The calibration scale at
I
I
I I 2
V -1 J / ./.
LY I
I
/
A
'Y I
1
V l' /
I
between the dotted-line curves, it may be I "L /
{ /j
I I.'/(
seen that the VSWR at the line input is ap- ! "
I L.J-- --;
~-
VI VI 112
proximately 2.3: 1. V I
/
~ I ) ~ I II
: I,
J ~
As another example, assume the use of
~ ~
.LC-!- ~ .
I
.!- ~..-J
-
a line with a matched loss of 3 dB and that
0.2!.,!... ~
~ /
the measured VSWR at the line input is ~
2: 1. What is the total line loss and the I' 1- /
7 /0 'itT
I - 7' I
-'I 1 I I
VSWR at the load? In this case begin as
o
II il II V
before, by running a finger along the scale 1 2 3 4 6 8 10 12 15 '20
at the bottom of the graph until the value VSWR AT LOAD
of 2 is located (it will, be the same at the
line input if theJine is lossless, or has 0 dB
of matched loss). This time, however, Fig: 33 - Transmission loss as a function of source and load VSWR. See text for applications.
proceed by following the "2" dotted line
up and slightly to the right, as this dotted
line represents a 2: 1 VSWR at the line in- the total line loss is just a bit more than S were known, it would have been possible
put. Follow this line until it intersects with dB, and that the VSWR at the load (scale to determine the matched loss and the
the 3 solid curve, representing the at the bottom) is S: 1. Had the attenuation t,otal loss from the intersection of the S
matched loss of the line. From this in- values not been known in this example but VSWR-at-Ioad vertical line and the 2
tersection it may be read (scal~ at left) that the VSWR at both the load and line input VSWR-at-input curve. '
Antenna,s
for High Frequency
A n antenna system is comprised of all in urban areas have .frequent need to though circular polarization is possible,
the components which are used between accept a compromise type of antenna for just as it is at vhf and uhf. The
the transmitter or receiver and the actual the hf bands because the city lot won't polarization is determined by the position
radiator. Therefore, such items as the' accommodate full-size wire dipoles, end- of the radiating element or wire with
antenna proper, transmission line, fed systems or high supporting structures. respect to earth. Thus, a radiator that is
matching transformers, baluns and' Other constrictions are imposed by the parallel to earth radiates horizontally,
Transmatch qualify as parts of an antenna amount of money available for an while an antenna at a right angle to earth
system. antenna system (inclusive of supporting (vertical) radiates a ver.tical wave. If a wire
Only the antenna does the radiating in a hardware), the number of amateur bands antenna is slanted above earth, it radiates
well-designed system. It is noteworthy to be worked, and local zoning or- waves which are between vertical and
that any type of feed line can be utilized dinances. Finally, the operational ob- horizontal in natuce. During line-of-sight
with a given antenna, provided a suitable jective comes into play: To dedicate one's communications, maximum signal strength
matching device is used to ensure a low self to DXing, or settle for a general type will be realized when the antennas at both
standing-wave ratio (SWR) between the of operation that will yield short- and ends of the circuit have the same polarity.
feed line and the antenna, and again long-haul QSOs during periods of good Cross polarization results in many deci-
between the feed line and the transmitter propagation. Because of the foregoing bels of signal reduction. However, during
and/or the receiver. Some antennas influences, it is impossible to suggest one propagation via the ionosphere (sky wave)
possess a characteristic impedance at the type of antenna system over another. it is not essential to have the same
feed point close to that of certain transmis- Perhaps a general rule of thumb might be polarization as the station at the opposite
sion lines. For example. a half-wavelength to erect the biggest and best antenna end of the circuit. This is because the
center-fed dipole, placed a correct height collection that space and finances wnt radiated wave is bent and tumbled
above ground, will have a feed impedance allow. If a modest system is the order of considerably during .its travel from the
of approximately 75 ohms. In such a case the day, then use whatever is practical and atmospheric layer from which it is
it is practieal to use 75-ohm .coaxial or accept a tradeoff between elaboration and refracted. At the far end of the com-
balanced line to feed the antenna. But few performance. Practically any radiator will munications path the wave may be
amateur half-wavelength dipoles actually enable the operator to make good horizontal, vertical or somewh6re in
exhibit a 75-ohm impedance. This is contacts under some conditions of pro- between at a given instant. On multihop
because at the lower end of the high- pagation, assuming that the radiator is transmissions, where the signal is refract-
frequency spectrum the typical height able to accept power and radiate it at ed more than once from the atmosphere,
above ground is rarely more than 1/4 some useful angle respective to earth. and similarly reflected from the earth's
wavelength. The 75-ohm characteristic is In general, the height of the antenna surface during its travel (skip), con-
.most liRely to be realized in a practical above ground is the most critical factor at siderable polarization shift will occur.
installation when the horizontal dipole is the higher end of the hf spectrum - 20, 15 Therefore, the inain consideration for a
approximately one-half, three-quarters or and 10 meters. This is because the antenna good DX antenna is a low angle of
one wavelength above ground. At other should be clear of conductive objects such
heights the feed impedance may vary from as power lines, phone wires, gutters and
roughly 58 to 95 ohms (see Fig. 1). This the like, plus high enough to have a low
gerieral principle applies to nearly all radiation angle. This is not.· nearly as t20 I
>../2 HORIZONTAL
antennas which are erected horizontally important at 160, 80 and 40 meters, but it tOO
above an earth ground. Furthermore, the
precise conductivity of the earth at one
is still recommended that lower-frequency
antennas be well away from conductive ~ 80
It ~ ~
~V ~ fIC V
:t II
location may differ markedly at another objects and ,as high above ground as o
- 60
site, and this phenoinenon has a direct possible in. the interest of good per- 0:
r./2 VERTICAL
effect on the electrical height of the formance. The exception is a ground- ~
0:
40 /
radiator above ground. A curve (dotted mounted vertical antenna. Ground-plane 20 /
line) is included in Fig. J to demonstrate verticals, however, should be installed as
the radiation resistance of a vertical dipole high above ground as possible so that o
o 0.2 0.4 0.6 0.8 1.0 t.2 1.4 t.6
at various feed-point heights above ground. their performance will not be degraded by
HEIGHT (>..)
conductive objects. (ANTENNA CENTER)
The Antenna Choice
Paramount among the factors to con- Antenna Polllrlzation Fig. 1 - Curves showing the radiation resistance
sider when selecting an antenna is the Most hf-band antennas are either of vertical and horizontal half-wavelength dipoles
matter of available space. Those who live vertically or horizontally polarized, al- at various heights above ground.
20~1 Chapter 20
radiation rather than the polarization. It·
should be said, however, that most DX
antennas' for hf wor.k are horizontally
polarized. The major exception is the
ground-plane vertical and phased vertical
arrays.
Impedance
The impedance at a given point in the
antenna is determined by the ratio of the
voltage to the current at that point. For
example, if there were 100 rf volts and
1.4 amperes of current at a specified point a:
0
'rl
I
H
'/1..
i
I~
in an antenna, the impedance would be l-
60'"
CI)
OJ sharp, Hence, the antenna is capable of diameter tubing. At 28.7 MH~, a half-
wavelength in space is
co
z "" RESISTANCE t; working over a wide frequency range,
4 ,,-
~ 0.9 55 ~ This effect is greater as the diameter is
Q. // z
increased, and is a property of some
5 0 .9 2, 50i'!
492
CI)
importance at the very high frequencies 28.7 = 17.14 ft (5.3m)
"
:Ii 45",
iii
where the wavelength is small,
10 I 50 I 200 I 1000 I 5000 I 0:
20 100 500 2000' 10,000
The Half-Wave Wavelength Antenna from Equation I. The ratio of half
RATIO OF HALF WAVELENGTH A fundamental form of antenna is a wavelength to conductor diameter (chang-
TO CONDUCTOR DIAMETER
single wire whose length is approximately ing wavelength to inches) is
equal to half the transmitting wavelength,
Fig. 3 - Effect of antenna diameter on length for It is the unit from which many more-
half-wavelength resonance, shown as a complex forms of antennas are con-
(17.14 X 12) = 411
multiplying factor, K, to be applied to the free- 0.5.
space, half-wavelength equation (Eq. 1). The structed, It is known as a dipole antenna,
The length of a half-wave in free spac~
effect of conductor diameter on the center
impedance is shown also. is
From Fig. 3, K =
0.97 for this ratio. The
length of the antenna, from Equation 3a
is
Length (ft) = f(~HZ) la
angles smaller than a few degrees is (492 2~ ~.97) = 16.63 ft (5.06 m)
practically impossible to obtain over 150
horizontal ground, Above 15 degrees, Length (m) = f (MHz) Ib
however, the curves are accurate enough or 16 feet 7-1/2 inches, The answer is
for all practical purposes, and may be obtained directly in inches by substitution
The actual length of a half-wavelength
of Equation 3b. '
. taken as indicative of the result to be antenna will not be exactly equal to the
expected at angles between 5 and 15 half-wavelength in space, but depends
degrees, upon the thickness of the conductor in (5905 X 0.97) _ 1996' h (506 )
28.7. - . mc es . m
The effective ground plane - that is, relation to the wavelength as shown in
the plane from which ground reflections Fig, 3, where K is a factor that must be The length of a half-wavelength antenna
can be considered to take place - seldom multipled by the half wavelength in free is affected also by the proximity of the
is the actual surface of the ground but is a space to obtain the resonant antenna dipole ends to nearby conductive and
few feet. below it, depending upon the length, An additional' shortening effect semi conductive objects. In practice, it is
characteristics of the soil. occurs with wire antennas supported by often necessary after cutting the antenna
insulators at the ends because of the to the computed length to do some
Cu"ent and Voltage Distribution capacitance added to the system by the experimental "pruning" of the wire,
When power is fed to an antenna, the insulators (end effect), The following shortening it in increments to obtain a low
current and voltage vary along its length, formula is sufficiently accurate for wjre SWR. This can be done by applyingrf
The current is maximum (loop) at the antennas for frequencies up to 30 MHz, power through an SWR indicator' and
center and nearly Zero (node) at the ends, observing the reflected-power reading,
while the opposite is true ofthe rf voltage, Length of half-wave antenna (ft) = When the lowest SWR is obtained for the
.The current does not actually reach zero desired part of an amateur band, it does
at the current nodes, because of the end 492 X 0.95 468 not necessarily follow that the antenna is
f(MHz) = f(MHz)
2a
effect; similarly, the voltage is not zero at resonant at that frequency, However, a
its node because of the resistance of the matched condition will have been secured
antenna, which consists of both the rf 150 X 0.95 - the basic objective in preparing an
143
f(MHz) = f(MHz)
resistance of the wire (ohmic resist~nce) 2b antenna for use,
a:nd the radiation resistance, The radiation
resistance is an equivalent resistance, a' Radiation Characteristics
Example:
convenient conception to indicate the The classic radiation pattern- of a dipole
A half-wave antenna for 7150 kHz (7.15
radiation properties of an antenna, The antenna is most intense perpendicular to
radiation resistance is the equivalent MHz) is j~185=65,45 ft,or 65 ft 5 in.(l9,9m), the wire, A figure-8 pattern (Fig, 4) can be
resistance that would dissipate the power assumed off the broad side of the antenna
the antenna radiates, with a current Above 30 MHz the following formulas (bidirectional pattern) if the dipole is 1/2
flowing in it equal to the antenna current should be used, particularly for antennas wavelength or greater above earth and not
at a current loop (maximum), The ohmic constructed from rod or tubing, K is tak- degraded by nearby conductive objects,
resistance of a half-wavelength antenna is en from Fig, 3, This assumption is based also on a
ordinarily small' enough, compared with symmetrical feed system, In practice, a
the radiation resistan.ce, to be neglected Length of half-wave antenna (ft) = coaxial feed line will distort this pattern
. for all practical purposes, slightly, Minimum horitontal radiation
492 XK
3a occurs off the ends of the\dipole, The
Conductor Size f(MHz) foregoing relates to a half-wavelength
The impedance of the antenna also , 5905 X K antenna which is parallel to the earth,
depends upon the diameter of the Length (m.) = f (MHz) 3b However, if the dipole is erected verti-
conductor in relation to the wavelength, cally, uniform radiation in a,ll compass
as indicated in Fig, 3, If the diameter of 150 XK 3c directions (a doughnut pattern if it could
the conductor is increased, the ca- Length (m) = f(MHz) be viewed from above the antenna) will
pacitance per unit length increases and the result,
inductance per unit length decreases, Length (mm) = 150,000 X K 3d One of the greatest errors made by
Since the radiation resistance is affected f(MHz) some beginners is to 'assume' that a dipole
20-3 Chapter 20
DX-communications angle of approxi-
90' mately 15 degrees. The higher~angle l.obes
(50°) are useful for short-haul com-
munications and compare favorably in
practice with the lobe angle seen in Fig.
5A. At heights appreciably lower than 1/2
wavelength, the lobe angle becomes
higher, and eventually the two lobes 75-I"l. LINE
20·5 Chapter 20
GUY~_ _ __ FEED
\ Lt POINT
(LO-Zl
Ct C,t
TO
ANCHOR
Fig. 12 - Example of a.trap dipole antenna, L 1 and C1 can be tuned to the desired frequency by
means of a dip meter before they are installed in the antenna,
TO TO
ANCHOR ANCHOR
'BENT DIPOLE
wire lengths between the traps and beyond
(Al will be necessary.
The trap capacitors should be high-
voltage and-current' units (C 1). Transmit-
ting mica capacitors offer good per-
FEED formance. Transmitting ceramic capa-
POI'NT
(LO-Z) citors are usable, but change value with
(Al extreme changes of temperature. There-
fore, they are more suitable for use in
regions where the climate is fairly
constant throughout the year. The coils
(Ll) should /be of reasonably heavy wire
gauge to minimize PR losses. The X Land
, Xc values "in the traps ·are not critical.
Generally the reactance can be on the
BENT INVERTED V order of 100 to 300 ohms. The traps are
(el checked for resonance before they are
installed in the antenna system. This can
Fig, 11 - When limited space is available for a
dipole antenna the ends can be bent downward
be accomplished by means of a dip meter
{A), or back on the radiator as shown at B, The Fig, 13 - A helically wound dipole is illustrated and a calibrated receiver. Weatherproofing
inv~rted Vat C can be erected with the voltage at A. As shown, the radiation resistance will be should be added to the traps as a measure
ends bent parallel with the ground when the very low and will require a broadband matching against detuning and damage from ice
available supporting structure is not high enough transformer, The coupling method seen at B is snow and dirt.
to permit an enclosed angle of approximately 90 more satisfactory for providing a matched
degrees. condition.
Helically Wound Dipoles
~he overalI length of a half-wavelength
dipole can be reduced considerably by em-
20-acre field. This verified the omni- to ensure a low SWR in the favored part ploying helically-wound elements. Fig. 13A
directional property mentioned earlier. of the amateur band for which the shows the general form taken with this
antenna is built. If open-wire feeders and type of antenna. A length of insulating rod
Bent Dipoles and Trap Dipoles a Transmatch are used, the dipole length or tubing (fiberglass or phenolic) is used to
When there is insufficient real estate to will not be a critical factor, provided it is contain the wire turns of the dipole. The
permit the erection of a full-size hori- close to the length required for a fully material should be of high dielectric
zontal or inverted-V dipole, certain extended half-wavelength dipole. Pruning quality. Varnished bamboo has been used
compromises are possible in the interest of will be required if single-band operation successfully by some in lieu of the more
getting an antenna installed~ The voltage with coaxial feed line is planned. expensive materials. A hardwood pole
ends (far ends) of a dipole can be bent Trap dipoles offer one solution to from a lumber yard can be used after
downward toward earth to effect reso- multiband operation with a shortened being coated one or more times with
nance, and the performance will not be radiator. The concept is seen in Fig. 12. In exterior spar varnish.
reduced markedly. Fig. 11 illustrates the this example the dipole is structured for To minimize losses, the wire used
technique under discussion. At A the two-band use. Assuming in this case that should be of the largest diameter prac-
dipole ends are bent downward and, the antenna is made for operation on 80 tical. The turns can be close wound or
secured to anchors by means of guy line. and 40 meters, the overall radiator spaced apart with little difference in
Some pattern distortion will result from (inclusive of the traps) must be resonant at performance. The ends of the helical
bending the ends. The dipole ends can the center of the chose'il. section of the dipole should contain capacitance hats
also be bent back over the wire halves of 80-meter band. The traps add loading to (disks or wire spokes preferred) of the
the antenna, as seen at B in Fig. 1 L This the dipole, so the length from the feed largest size practical. The hats will lower
causes some signal cancellation (more point to the far end of each leg will be the Q of the antenna and broaden its
severe than with· the system of Fig. IIA), somewhat shorter than normal. During response. If no disks are used, extremely
so it is not a preferred technique. 40-meter operation the traps present a high rf voltage can appear at the ends of
Fig. IIC demonstrates a bending high impedance to the signal and "di- the antenn·a. At medium power levels and
technique for inverted Vs when the vorce" the wires beyond the traps. higher the insulating material can burn
available supporting mast or tree is too Therefore, the wire length from the feed when no hats are used. The voltage effect
short to permit normal installation. The .point to each trap is approximately is similar to that of a Tesla coil.
ends of the dipole are guyed off by means what it would be if the dipole were The feed impedance of helical dipoles
of insulators and wires, as shown. cut for just 40 meters, with no traps in the or verticals is quite low. Therefore, it may
Alternatively, but not preferred, is the line. This principle can be extended for be necessary to employ some form of
fold-back method at B in Fig. 11. additional bands, using a new set of traps matching network to interface the anten-
All of the shortening systems high- for each additional band. Since there is na with 50-ohm coaxial cable. A broad-
lighted in Fig. 11 will have an effect on the considerable interaction between the vari- band, variable-impedance transformer is
overall length of the dipole. Therefore, ous segments of a multiband trap dipole, convenient for determining the turns ratio
some cutting and testing will be necessary c~>nsiderable experimentation with the of the final transformer used. The feed
Antennas for High Frequency 20-6
method shown at B of Fig. 13 can be used minor lobes will occur off the broadside of a, complex impedance, not purely a
to secure a matched condition. L2 is the antenna. When a -long wire is resistance. For most amateur work a
wound over L 1, or between the two halves terminated at the far end (Fig. 15B and termination of 400 to 600 ohms will be
of Ll, as illustrated. Cl is adjusted for an C), maximum directivity is off the entirely suitable. The reactance values
SWR 'of 1 at the center of the desired terminated ends, as illustrated in Fig. 18. normally lie between j 150 and j250.
operating range. The bandwidth of this Therefore, the antenna should be pointed The most effective method of termina-
type of antenna is quite narrow. A toward the DX area of interest,. such as ting the long wire is by means of a
40-meter version with an 18-foot (5.5-m) NE from the New England area for buried or above-ground system of radial
overall len~h exhibited a 2: 1 SWR working Europe. The addition of a wires. Although some amateurs have
. bandwidth of 50 kHz. The capacitance terminating resistor does not decrease the simply returned the terminating resistor to
hats on ~hat model were merely I8-inch antenna gain in the favored direction, but a pipe driven into the earth, the ef-
(457-mm) lengths (spikes) of no. 8 it does consume 3 dB of power. Thus, the fectiveness of the ground is poor. Also,
copperweld wire. Greater bandwidth would terminating resistor needs to be capable of the quality of such a ground system will
result with larger capacitance hats. dissipating one half the power fed to the change with the seasons. The same is true
To obtain half-wavelength performance antenna, without excessive heating. Fur-
it is necessary to wind approximately one thermore, the resistor Rt of Fig. 15) must
wavelength of wire on the tubing. Final be noninductive so that a pure resistance
pruning can be accomplished while ob- exists. The optimum value of the termina- _MAXIMUM S I G N A L _
serving an SWR indicator placed in the tion cannot be calculated by means f - - L ~I WAVELENGTH---!
transmission lioe. Proximity to nearby
conductive objects and the earth will have
of a simple equation. The best value
depends on the length of the long wire, the
L---.8 6-----iJ----1=.-
a significant effect on the resonance of the height aboveground and some less L(FEE ) 0 984(N-O.025)
T F(MHz)
antenna. Ideally, final adjustments should significant factors. Generally, the value of NoNO.OF FULL WAVES
be made with the antenna situated where R t is determined experimentally by prob-
it will be during use. Marine spar varnish ing the long wire with an rf ammeter to
should be painted on the elements after all establish' that uniform current distribution
tuning is finished. This will protect the exists. The entire length of the antenna k/4--1
(A)
antenna from the weather and will lock should be probed in this manner. In
the turns in place so that detuning will not practice, the correct termination requires
occur later on. """"" """~
A reasonably linear current and voltage
distribution will result when using a
*
helically wound dipole or vertical. The
3.0
same is not true of center-, mid- or
end-loaded (lumped inductance) dipoles. (/)
i= 2.5 GROUND
The efficiency of this antenna will be TO RADIALS
~
tradeoff, this antenna is capable of good z NETWORK
",,/
~
P -- 1-0--
r- - r-
30
20
zQ.
ell/)
'"a:
iO
Single Long Wire o / V
o
A long-wire antenna can be operated o 2 3 4 5 6 71 8 9 iO
L'ENGTH IN WAVELENGTHS
with or without a resistive termination.
When no terminating resistor is used, the
maximum-response pattern will be bidirec- Fig. 16 - Theoretical gain of a true long-wire antenna over a reference dipole as a function of wire
tional off the ends of the wire. Some length. The angle, with respect to the wire, at which the radiation intensity is maximul'Q is shown also.
20·7 Chapter 20
long-distance communications when Yagi
45 0 90 0 45 0 900 or quad beams are useless. It depends on
\\ 1\\1\\ f\, f(., K~ ~ B
the propagation conditions at a given
\ \ I,\: ~ , ~ ~
~ 80
0
7
period of the day. The major limitation is,
~ 70 0 6
of course, the inability of the operator to
o
l- \ \ 1'\, I" ~ ~ ~ 5
rotate the antenna toward the compass
" '-... ......""'" t-:::: -:::
I- 60 0
'"
<.) 1', 4 point .of interest. If one were able to look
:t
<J>
W
50 0
1\ ", 1'- ----- 3 down upon a long wire and see the
radiation pattern,. the lobes illustrated in
-- -
0; 40 0
" 2
(Al :J:
i'-.. Fig. 18 would be typical. The comparison
~ 30 0
.......... is between a two- and a four-wavelength
~
~ 20
0 r- t-- long wire. It can be seen that the longer
the antenna the more directional it
:i'" 10 0
becomes as the major lobes compress. The
00 12345678 examples are for terminated long wires.
LENGTH IN WAVELENGTHS Fig. 19 shows curves for the wave angle
versus length in wavelengths.
4 WAVELENGTHS
(Bl Fig. 19- Angles from which long-wire radiation Wire Combinations
is maximum (solid curves) and zero (dashed
curves). The major lobe has the power gains Two long-wire antennas can be COIn-
Fig. 17~' Elevation-plane radiation characteristic
given in Fig. 16. Secondary lobes have lower bined as shown in Fig. 20 to form what is
amplitude, but the maxima may exceed the called a V beam. The radiation properties
of a one-wavelength long wire (A) compared to
radiation intensity of a half-wavelength dipole.
that of a four-wavelength long wire. The patterns are similar to those of single long wires,
are for unterminated wires erected over a respective to being terminated or un-
perfectly conducting ground.
terminated. The terminations, Rl and R2,
can be returned to a radial ground system
---'----iI
~
:> V4
L=7>.. Rl _-- or to a 1/4-wavelength wire extension at
~ -~~
the far ends of the V. If a proper
termination is made, the system has a
MATCHING MAXIMUM.
feed-point impedance roughly equivalent
NETWORK SIGNAL = to the termination value. This requires a
matching circuit between the feed point
and the transmission line, or between the
TERMINATED _-'\~:
V-BEAM ~. vI,- transmission line and the station equip-
<
<'R2'
--- ment, to provide a 50-ohm unbalanced
?' load for the transmitter and receiver. A
(Al ~
Transmatch is the better choice for most
installations. The feeder can be a length of
MA~.~
ANTENNA 600-ohm open-wire line to the Trans-
2>-
AXIS match, assuming a 600-ohm termination
(A) MAX.
is used.
-~- Fig. 21 shows the relationship between
the apex (enclosed) angle of a V beam and
the leg size in wavelengths. Two wave
(81 angles (0 and 15 degrees) are provided. A
V .beam is designed primarily for one
band, but it can be used effectively on
Fig. 20 - Two long-wire antennas can be
combined to form a V beam as seen at A. Gr.eater several amateur bands when the leg length
gain and directivity can be had when using this is at least one wavelength on the lowest
scheme. The V can be operated as resonant operating band. However, the apex angle I
(8) (unterminated) for bidirectio,nal response, or will be optimum only for the band it is
terminated (nonresonant) for undirectional
response. Terminations R1 and R2 can connect designed for.
Fig. 18 - Typical radiation patterns (azimuth) for to quarter-wavelength extensions (Fig. 15C) or as
two-wavelength and four-wavelength terminated shOwn by R1' and R2' to an earth ground. The Rhombic Antenn~s
long wires. In practice, the pattern is somewhat azimuthal pattern at B is for an unterm~nated V
distorted by the wir!3 attenuation. beam. Four long-wire antennas can be com-
bined in a rhomboid configuration to
develop a rhombic antenna. The basic
of buried radials, but the effects are not wave angles also. A comparison of the structure is illustrated in Fig. 22. The
nearly so pronounced. Fig. 15B shows an approximate wave angles for an un- system can be used with or without the
illustration of radial-wire termination. An terminated one-wavelength and an un- termination, R t . The decision is based on
alternative which has been applied by terminated four-wavelength long wire is whether or not bidirectional radiation is
some antenna designers is shown in Fig. provided in Fig. 17. It is apparent that the desired. This follows the rule for long
15C, where a 1/4-wavelength extension is greater the number of wavelengths the wires and V beams. A typical terminating
added to the long wire. R t is inserted lower the angle of radiation for the major resistance is 600 ohms for amateur work.
between the antenna and the 1/4- lobes. Owing to the increased number of This makes the system compatible with
wavelengtli extension. It is assumed in this high-angle minor lobes of the longer 600-ohm open-wire transmission line.
case that the long wire consists of a mul- antennas (Fig. 17B), good results can be Sides a, b, c and d all have the same length
tiple of half waves. obtained in the favored antenna direction (Fig. 22). The opposite corner angles are
The theoretical gain of a long-wire under a variety of sky-wave conditions. In equal. In effect, this antenna is two V
antenna is given in terms of wavelengths an actual situation, a long wire of several beams which are joined end to end.
at Fig. 16. Data are provided to show the wavelengths can be capable of effective Maximum directivity is off the terminated
eo it 70
~ ....o I
70 .... 60
60 i'::: ........... u
OJ
Q. 50 \
~ r---
-- -- --
............... <J)
OJ
It 40 I\.
'j
"~
50
r---... ,-0 0 I:
'"z
<l 40
x
OJ
>
15 0 ""-
r--.. r--
~
~ 20
"'
30
...........
-
---
Q.
<l
30 ,
r-- I-- :'1 10
00 1 2 3 4 5 6 7 8 9
LENGTH IN WAVELENGTHS
20-9 Chapter 20
radiation patterns are given in Fig. 27.
The example at A is for a quarter- 14
wavelength radiator over a theoretically
12
ideal ground. The dashed lines show the
qment distribution, inclusive of the image 10
.,.v~
,/'"""
m
portion below ground. The imager can be /
":' 8
equated to one haJf of a dipole antenna, z //.
with the vertical radiator representing the ~ 6
remaining dipole half. The illustration at
V
4 ,
B characterizes the pattern of a half-
wavelength vertical. It can be seen that the 2
radiation angle is somewhat lower than o
1 2 3 4 5
that of the quarter-wavelength version at
A. The lower angles enhance the DX LEG LENGTH-WAVELENGTHS
capability of the antenna. Two half
wavelengths in phase are shown in Fig. 27 at
C and D. From a practical point of view, Fig. 24 - Rhombic.ante~na design chart for Fig. 25 - Theoretical gain of a nonresonant
few amateurs could erect such an antenna any given leg length, the curves showing the rhombic antenna over a half-wave dipole in free
unless it was built for use on the higher hf proper tilt angle to give maximum radiation at space. This curve includes an allowance of 3dB
bands, such as 20, 15 or 10 meters. The the selected wave angle. The broken curve for loss in the terminating resistor.
marked "optimum length" shows the leg
very low radiation angle is excellent for length that gives the maximum possible output
DXing, however. at the selected wave angle. The optimum
length as given by the curves should be
FuU-Size Vertical Antennas multiplied by 0.74 to obtain the leg length for ,
When it is practical to erect a full-size
which the wave angle and main lobe are
, ! i
aligned.
vertical antenna, the forms shown in Fig.
28 are worthy of consideration. The ex~ , I
I
i / N
ample at A is the well-known vertical
,
f\ i
~ ---.. ~:\ R i
ground plane. The ground system consists antenna, the feed impedance would be ,
of four or more above-ground radial wires near 30 ohms. In a practical case, owing to
\
·f \ '" , !
against which the driven element is imperfect ground, the impedance is more , \
worked. The driven element length in feet
is derived from the standard equation
apt to be in the vicinity of 5 to 15 ohms. .~4
,
i !
II
:;1 I ,
I I
A gamma-feed system for a grounded 'I
1/4-wavelength vertical is presented in Fig.
28D. Some rules of thumb for arriving at , III 1
I
I
I
i~
roughly 50 ohms. When the radials are at results can be had when the radiator is
right angles to the radiator (Fig. 28B) the approximately three percent shorter than
feed impedance approaches 30 ohms. The the resonant length. Amateur antenna
major advantage in this type of vertical towers lend themselves well to use as
shunt fed verticals, even though an (B)
antenna over a ground-mounted type is
that the system can be elevated well above hf-band beam antenna may be mounted
nearby conductive objects (power lines, on the tower. The overall system should
(A)
trees, \ buildings, etc.). When drooping be close to resonance at the desired
radials are utilized they can be used as guy operating frequl!ncy if gamma feed is to be
wires for the mast whilZhsupports the used. The hf-band beam will contribute
antenna. The coaxial cable shield braid is somewhat to top loading of the tower.
connected to the radials and the center The natural resonance of such a system
conductor is common to the driven can be checked by dropping a no. 12 or 14
element. wire down from the top of the tower
The Marconi antenna seen in Fig. 28 at (making it common to the tower top) to
C is the classic form taken by a form a folded unipole (Fig. 28E). A four-
(C) (0)
ground-rnounted vertical. It can be ground- or five-turn link can be inserted between
ed at the base and shunt fed, or it can the lower end of the drop wire and the
be isolated from ground, as shown, and ground system, then a dip meter inserted
series fed. This antenna depends upon an in the link to observe the resonant Fig. 27 - Elevation plane responses for a
effective ground system for efficient frequency. If the tower is equipped with quarter-wavelength vertical antenna (A), a 1/2-
wavelength type (B) and two half waites in phase
performance. The subject of ground guy wires, the latter should be broken up (C and D). It can be seen that the examples at B
screens is treated later in this section. If a with strain insulators to prevent unwanted and D provide lower radiation angles than the
perfect ground were located below the loading of the vertical. In such cases version at A.
fl
not able to provide 1/4-wavelength
.Lr1
resonance, portions of the top guy wires
can be used ~s top-loading capacitance. It
will be necessary to experiment with the
guy-wire lengths (using the dip-meter
technique) while determining the proper
11
H'~(
.
RADIATOR
/50[1 ~
I
dimensions.
A folded-unipole type of vertical is RADIALS~ RADIA~3.9 / II! I /~/J/);
GROUND
depicted at E of Fig. 28. This system has
the advantage of increased feed-point
/ A\ , L'''/3,9
~ GROUND PLANE
(Bl
MARCONI
(Cl
impedance. Furthermore, a Transmatch GROUND PLANE
can be connected between the bottom of (Al
the drop wire and the ground system to
permit operation on more than one band.
For example, if the tower is resonant on
80 meters, it can be used as shown on 160
and 40 meters with reasonable results,
even though it is not electrically long
enough on 160. The drop wire need not be
a specific distance from the tower, but
spacings between 12 and 30 inches are
H·"/4
I
C1
r J
NO.8
WIRE
MATCHING
NETWORK
GUY
WIRE_
suggested.
GAMMA- FEED FOLDED UNIPOLE SLANT-WIRE FEED
The method of feed seen at Fig.· 28F is (Dl (El (Fl
commonly referred to as "slant-wire
feed." The guy wires and the tower
combine to provide quarter-wave reso-
nance. A matching network is placed Fig. 28 - Various types of vertical antennas.
between the lower end of one guy wire and
ground and adjusted for an SWR of I. It
does not matter at which level on the at D. This system works well with gamma resonance will drop some 100 kHz.
tower the guy wires, are connected, feed. The loading wires are trimmed to A method for effecting the top-loading
assuming that the Transmatch is capable provide an electrical quarter wavelength of Fig. 29, illustration F, is shown in the
of effecting a match to 50 ohms. for the overall system. This method of drawing of Fig. 30. Pipe section "D is
Physically Short Verticals loading will result in a higher radiation mated with the mast above the hf-band
resiMance and greater band-width than beam antenna. A loading coil is wound on
A group of short vertical radiators is the systems shown in Fig. 29 at A, Band solid Plexiglas rod or phenolic rod (item
presented in Fig. 29. IIlustrations A and B C. If an hf-band or vhf array is atop the C), then clamped inside the collet(B). An
are for top and center loading. A tower, it will simply contribute to the top aluminum slug (part A) is clamped inside
capacitan<;e hat is shown in each example. loading. item B. The top of. part A is bored and
It should be as large as practical to A tri-wire unipole is shown at E of Fig. threaded for a 3/8 inch X 24 thread stud.
increase the radiation resistance of the 29. Two no. 8 drop wires are connected to This will permit a standard 8-foot (2.4 m)
antenna and improve the bandwidth. The the top of the tower and brought to stainless-steel mobfle whip to be threaded
wire in the loading coil is chosen for the ground level. The. wires can be spaced any into item A above the loading coil. The
largest gauge consistent with ease of convenient distance from the tower - capacitance hat (Fig. 30, illustration B)
winding and coil-form size. The larger normally 12 to 30 inches (0.3 to 0.76 m) can be made from a II4-inch (6.3-mm)
wire dIameters will reduce the PR losses from one side. CI is adjusted for an SWR thick brass of aluminum plate. It may be
in the system. The coil-form material of I. This type of vertical has·a fairly nar- round or square. Lengths of 118-inch
should be of the medium- or high- row bandwidth, but because Cl can be (3-mm) brazing rod can be threaded for a
voltage breakdown resistance dielectric motor-driven and controlled from the 6-32 format to permit the rods to be
type. Phenolic or fiberglass tubing is en- operating position, QSYjng is accomplish- screwed into the edge of the aluminum
tirely adequate. ed easily. This technique will not be plate. The plate contains a row of holes
A base-loaded vertical is shown at C of suitable for matching to 50-ohm line along its perimeter, each having been tap-
Fig. 29. Since this is the least effective unless the tower is less than an electrical ped for a 6-32 thread. The capacitance hat
method of loading in terms of antenna quarter wavelength. is affixed to item A by means of the 8-foot
performance, it should be used .only as a A different method for top loading is whip antenna. The whip will increase the
last choice. The primary limitation is that shown at F of Fig. 29. W9UCW described effective height ?f the vertical antenna.
the current portion of the vertical exists in this system in December 1974 QST as
the coil rather than the driven element. "The Minooka Special." An extension is Cables and Control Wires on Towers
With center loading the portion of the used at the top of the tower to effect an Most vertical antennas of the type
antenna below the coil carries current, electrical quarter-wavelength vertical. LI shown in Fig. 29 consist of towers and hf
and with the top-loading version the entire is a loading coil with sufficient inductance or vhf beam antennas. The rotator control
vertical element carries current. Since the to provide antenna resonance. This type wires and the coaxial feeders to the top of
current part of the antenna is responsible of antenna lends itself nicely to operation the tower will not affect antenna perfor-
for most of the radiating, base loading is on 160 meters. LI and the pipe extension mance adversely. In fact, they become a
the lea,st effective of the three methods. above the hf-band beam can be tuned at part of the composite antenn<;t. To prevent
The radiation resistance of the coil-loaded ground level against the ground system. It unwanted rf currents from following the
antennas shown is usually less than 16 should be made resoriant approximately wires into the shack, simply dress
ohms. 100 kHz higher than the desired operating them close to the tower legs and bring
A method for using guy wires to top frequency for use on 160 meters. After it them. to ground level. This de-
load a short vertical is i1Iustrated in Fig. 29 is in place on the tower, the overall system couples the wires at rf. The wires should
20-11 Chapter 20
" 3/S"X24 THREAD
b-D~/82G!f~ FOR 8' STEEL WHIP
ALUMINUM -:-_-:.....
ADAPTOR A
SLUG
H< 1../4
~-'A"L_~CLAMPS
I
i
J;;L;ffl/n)?
TOP LOADING ~~~~~;LOADS---+,I&:=:[:Ik_LOAOING COIL
(A) ROD 20"
I
I LONG
/IIJII~j}/11
CENTER LOADING
(8)
(Al
SA'SE LOADING
r-: '
(C)
TOP LOADING
Fig, 30 - At A are the details for the tubing section of the loading assembly. Illustration B shows the
(0) top hat and its spokes. The longer the spokes, the better,
Jl
H
the ground wherever space will permit.
JJ
FLAT-TOP T
(A)
INVERTED L
(8)
HALF SLOPER
They may run away from the antenna in
only one or two compass directions.
Results will still exceed those of when no
ground system is used.
O' Ie) A single ground rod, or a group of them
Lt ~ TOWER L Z 45-
HF BEAM
bonded together, is seldom as effective as
L2 -TOWER L ~ 30° a collection of random-length radial
wires. In some instances a group of short
Lt, L2 '" 1.045( 234 ~ FT.
f(MHz)] radial wires can, be used in combination
with ground rods driven into the soil near
LI 270' + .....+---<!~~1!----1----l- 90' the base of the antenna. The power-
company ground can be tied in also, and if
a metal fence skirts the property it can
'L2 (D) IE) also be used as part of the ground system.
GUY GUY I 225'
A good tule is to use anything that will
.-obJk..". serve as a ground when developing a
GROUND SYSTEM
180'
radial-ground system.
All radial wires must be connected
together at the base of the vertical
Fig. 31 - Some variations in vertical antennas which offer excellent performance. antenna. The electrical bond needs to be
of low resistance. Best results will be
obtained when the wires are soldered
together at the junction point. When a
grounded vertical is used, the ground
wires should be affixed securely to the
upon the tower height, the beam atop the network that approaches the ideal. It base .of the driven element. A lawn-edging
tower and the ground condition below the might be said that "a poor ground is tool IS excellent for cutting slits in the soil
system. better than no ground at all," and when laying radial wires.
Fig. 31 E shows the relative radiation therefore the amateur should experiment
pattern of the antenna in Fig. 31D. The with whatever is physically possible rather Trap Verticals
maximum radiation is off slope wire L I, than exclude vertical antennas from his or
Although a full-size, single-band anten-
with a minor lobe occurring off the com- her construction plans. It is often possible
na is more effective than a lumped-
pensating wire, L2. The pattern shown to obtain excellent DX results with
constant one, there is justification for
was obtained with a field-strength meter practically no ground system at all.
using trap types of multi band antennas.
placed at the ends of 2-wavelength radials. Although the· matter of ground systems
Without L2 in the system, a single promi. could be debated almost endlessly, some The concept is especially useful to
operators who have limited antenna space
nent lobe prevails. However, the half- practical rules of thumb are in order for
on their property. Multiband "Com-
sloper is otherwise (and for all practical those wishing to erect vertical antennas.
promise" antennas are also appealing to
purposes) omnidirectional. Users may Generally, if the physical height of the
persons who engage in portable operation
want to orient the pattern in some favored vertical is a quarter wavelength, the radial
wires should be of the same length and and are unwilling to transport large
DX direction.
Tests indicate this antenna to be effec- dispersed uniformly from the base of the amounts of antenna hardware to the field.
The trap vertical antenna operates
tive on its harmonics for DX and local tower. In this example approximately 60
much in the same manner as a trap dipole .
work. If harmonic operation is planned, radial wires will suffice. The conductor
or trap-style Vagi. The notable difference
the 50-ohm coaxial feeder should be size of the radials is not especially
is that the vertical is one half of a dipole.
replaced with open-wire line, which is significant. Wire gauges from no. 4 to no.
The radial system (in-ground or above
coupled to the transmitter by means of a 20 have been used successfully by amateurs.
ground) functions as a ground plane for
Transmatch. With open-wire feed, the L2 Copper wire is preferred, but where
the antenna, and represents the missing
compensating wire of Fig. 31 D will not be soil acid or alkali is not high in level
necessary. Best performance will be had aluminum wire can be used. The wire~
half of the dipole. Therefore, the more ef-
fective the ground system the better the
when the base of the tower is well ground- can be bare or insulated, and they can be
antenna performance.
ed. Buried radials are highly recommend- laid on the earth surface or buried a few
Trap verticals are adjusted as quarter-
ed as part of the ground system. Opera- inches below ground. The insulated wires
wavelength radiators. The portion of the
tion on 160 meters can be had by feeding will have greater longevity by virtue of
reduced corrosion and dissolution from antenna below the trap is adjusted as a
the low end of L2. .
quarter-wavelength radiator at the highest
The bandwidth of the system in Fig. 31 soil chemicals.
The shorter the vertical antenna the proposed operating frequency, i.e., a
will be approximately 50 kHz on 160
20/15-meter trap vertical would be a reso-
meters, 100 kHz on 80 meters, 200 kHz on fewer and shorter the radials need be. For'
nant quarter wavelength at 15-meters
40 meters, and so 'on, between the 2: I example, a vertical which is 1/8 wave-
length high will provide maximum field from the feed point to the bottom of the
VSWR points.
strength with 16 to 18 radial wires, and the trap. The trap and that portion of the
Ground Systems wires need be only as long as the vertical is antenna above the trap (plu~ 'the 15-meter
The importance of an effective ground high. If time and expense are not a section below the trap) constitute the com-
system for vertical antennas cannot be prime consideration, the amateur should plete antenna during 20-meter operation.
emphasized too strongly. However, is it bury as much ground wire as possible. But, because the trap is in the circuit the
not always possible to install 'a radial Some operators have literally miles of wire overall' physical length of the vertical
20-13 Chapter 20
Table 1
A
Tubing-Section Lengths for 2-Band Vertical
Band A B C 0 E F C1 L1
(MHz) • (pF)' (approx. /AH)
8----.>0.-,
(5/8"OD) Tubing 21/28 25 16 25 25 25 33 18 1.70
Length 14/21 38 33 37 37 37 33 25 2.25
(Inches) 10/14' 42 42 54 54 54 49 39 3.25
21/28 20 16 21.5 21.5 21.5 33
CLAMP 14/21 33 33 33 33 33 33
Tubing 10/14' 37 37 49.6 49.6 49,6 49
Length
c at Resonance mm = In. x 25.4 dimensions.
______ HOSE 'New WARC·79 band.
(approx.
CLAMP inches)' 'See text.
'Midband
~"
D Dimensions XC1' Xu :::: 300 O.
(3/4"OD)
SIZES SHOWN
trap is adjusted, it can be installed in the suggested for the trap coil. The heavier
antenna, and no further adjustment wili the wire gauge the lower the trap losses
INCHES (") x 25.4 be required. It is easy, however, to be and the higher the Q. The larger wire sizes
= mm misled after the system is assembled: At- will reduce coil neating.
tempts to check the trap with a dip meter
CLAMP F will suggest that the trap has moved much A Practical Two-Band Vertical'
lower in frequency (approximately 5 MHz Fig. 32 contains a pictorial view of a
lower in a 20/15-meter vertical). This is two-band trap vertical (20/15 meters)
because the trap has become absorbed in- which can be collapsed to 39 inches (991
to the overall antenna, and the resultant mm) for easy transportation on holidays,
(1-3/8" OD x 8") resonance is that of the total antenna. Ig- DXpeditions or camping trips. All of the
nore this phenomenon. tubing sections except B telescope
Multiband operation for three or four together to make a compact package. The
WIRE
bands is entirely practical by using the ap- trap and base plate will be separate from
propriate number of traps and tubing sec- the remainder of the antenna during
Fig. 32 - Details of the two· band trap vertical, tions. The construction and adjustment storage or transport.
which telescopes to 39 inches when dis· procedure is the same, irrespective of the If portability is not a requirement, a
mantled. Stainless·steel hose clamps are used number of bands covered. The highest- single section of aluminum tubing can be
to hold the tubing sections together and to af·, frequency section is always closest to the
fix the trap to the tubing. A short length of
used below the trap, although two sec-
flexible wire and a banana plug are connected feed end of the antenna, and the lowest- tions (telescoping) are recommended to
to the base of the antenna for joining the frequency section is always the farthest facilitate adjustment of the 15-meter por-
antenna to the coax connector of Fig. 34. from the feed point. As the operating fre- tion of the system. Similarly, two
quency is progressively lowered, more telescoping tubes can be used above the
traps and more tubing sections become a trap (as shown) to permit adjustment for
antenna will be slightly less than that of a functional part of the antenna. 20-meter operation. ,
single-band, full-size 20-meter vertical. The trap should be weather-proofed to Table 1 contains data on the starting
prevent moisture from detuning it. lengths of the tubing sections, plus ap-
Traps Several coatings of high-dielectric com- proximate dimensions for resonance on a
The trap functions as the name implies: pound, such as Polystyrene Q Dope, are variety of band pairings. Final adjustment
It traps the 15-meter energy and confines effective. Alternatively, a protective sleeve is done for the lowest VSWR attainable in
it to the part of the antenna below the of heat-shrink tubing can be applied to the the chosen part of each band (resonance).
trap. During 20-meter operation it allows coil after completion. The coil form for The adjustment must be done while the
the rf energy to reach all of the antenna. the trap should be of high-dielectric quali- antenna is mounted for use with the
Therefore, the trap should in this example ty and be rugged enough to sustain stress ground system in place.
be tuned as a parallel resonant circuit to during periods of wind. Fig. 33 sh9wS the details of a simple
21 MHz. At this frequency it "divorces" The trap capacitor must be capable of trap with the tubing sections keyed to the
the top section of the vertical from the withstanding the rf voltage developed nomenclature of Fig. 32. The 10 of the
lower section because it presents a high- across it. The amount of voltage present PVC tubing is too small to accommodate
impedance (barrier) at 21 MHz. General- "- will depend on the operating power of the the 112-inch 00 (l3-mm) tubing.
ly, the trap inductor and capacitor have a transmitter. Fixed-value ceramic transmit- Therefore, a hacksaw is used to cut four
reactance of 100 to 300 ohms. Within that ting capacitors are suitable for most slots at the ends of sections Band C so
range it is not ,critical. power levels if they are rated at 5000 to they will compress and fit into the PVC
The trap is built and adjusted separately 10,000 volts. A length of RG-58 or RG-59 tubing. The wooden dowel plugs permit a
from the antel}na. It should be resonated coax cable can be used successfully up to tight borid when the hose clamps are conv
at the center of the portion of the band to 200 watts. (Check to see how many pF per pressed over the ends of the PVC tubing.
be operated. Thus, if one's favorite part foot your cable is before cutting it for the Innovative builders can find other
of the 15-meter band is between 21,000 trap.) RG-8 or. RG-ll cable is recom- methods for mounting the trap in the
and 21,100 kHz, the trap would be tuned mended for the trap capacitor at powers antenna.
to 21,050 kHz. in excess of 200 watts. The advantage of Copper straps G are slid into the PVC
Resonance is checked by using a dip using coax cable is that it can be trimmed tubing to provide electrical contact with
meter and beating the dipper signal easily to adjust the trap capacitance. tubing sections Band C. The straps are
against a calibrated receiver. Once the Large-diameter copper magnet wire is bent (as shown) so they will fit under the
Antennas for High Frequency 20·14
can be changed to help provide a match to
SECTION B
50-ohm feed line. The greater the angle
~--- 5-1/2" "1
~
SAW SLOTS (4)
between the vertical element and the
radial wires, the higher the feed im-
1/4', 2"--....~ , pedance. The feed impedance will be ap-
COPPER proximately 30 ohms when the radials are
STRAP G
at right angles to the vertical element. If
such an installation is contemplated,a
1.6: 1 broadband toroidal transformer can
be used at the feed point to effect a
T~B~~G--"~
(718"00,4") SAW
matched condition.
___ TRAP SLOTS (4) Universal Mounting Plate
Fig. 34 illustrates a mounting plate
which will satisfy a host of conditions one
J~WOOOEN DOWEL might encounter when operating from an
114")1. 2" G 'UJ
Q
unfamiliar place. The hole size and spac-
COPPER ing will depend on the U bolts or muffler
STRAP~
~, , . . " ' " ' " clamps used with the antenna. The lower
ATTACH
set of holes (except the bottom-most tvvo) ANGLE BRACKET
20-15 Chapter 20
ANT. 4 NW
'NORMAL LOBE
N /
t ,;j':2?
~----------~-----------~3SW ANT. 2 SE
~
;. .
OL3 4' 50.0.
~. 50.0.
r+
ment phasing is accomplished with three
90° sections ofRG-8/U cable; DU, DL2 NE
- RELAY
and DL3. The four outputs are at 0°.
- 90°, - 90° and - 180°, respectively in
phase relationship, with power from the e:t,,~
transmitter divided into four equal parts.
Special attention should be paid to pre-
serving symmetry throughout the system,
in the I?ybrids, phase shifters, rf switch-
ing, feeds and antenna placement, in
Fig. 36 - Schematic diagram of the phased array. Relays K1 through K6 are three-pole, double-
order to have the array perform uniformly throw UrNts with iii II three sections of each relay connected in parallel. Each of the 100-ohm
as it is switched between the four resistances are comprised of two 200-ohm noninductive 50-watt resistors manufactured by
headings. Nytronics: The quarter-wavelength lines should be adjusted for 7.1 MHz with the aid of a noise
The switching was done with six relays bridge, RX bridge or similar device. Calculated lengths of line for the velocity factors of 0.81 and
0.66 are 28.1 feet and 22.9 feet respectively. (Feet x 0.3048 = M.)
as shown in Fig. 36 and the accompanying
photograph (Fig. 37). With no voltage ap-
plied to the relay coils, the arms are in the ing/switching system can be transported three sections were connected in parallel
positions shown in Fig. 36, giving north- or worked on as a single unit. See Fig. 38. to assure that the relays would handle the
easterly directivity a convenient Basic layout of this system is shown in transmitter power. A three-wire control
heading for operation from New England. Figs. 37 and 39_ Measured phase .errors system is used to steer the array_ Four dif-
The three Wilkinson power divider due to unequal lead lengths in the switch- ferent commands are required for four
resistor networks along with the switching . ing system are less than 2" _ The builder different directions as shown in Fig_ 36 .
relays are mounted. on a piece of sheet need not follow the layout shown here - An aluminum chassis measuring 10 X
aluminum measuring 10 X 12 inches (254 generally speaking, if the leads are kept 12 X 3 inches (254 X 305 X 76 mm) placed
X 305 mm). Each of the three phasing short and direct the system should func- over the piece of aluminum on the
lines and quarter-wavelength lines tionaccording to specifications listed resistor/relay side protects these com-
associated with the power dividers con- earlier. ponents from the weather. Silicon seal is
nect to coaxial fittings is mounted to the The relays used in this system were used to fill any holes or slots in the
piece of sheet aluminum. The entire phas- surplus 3-pole double-throw units. The chaSSIS.
Fig. 37 - Close·up view of the. relay and Fig. 38 - Photograph of the assembled phase
resistor assembly. system.
DL3..A"
20-17 Chapter 20
(roughly 15 degrees) than the larger lobe /Xl III
of Fig. 41 (approximately 30°). The pat- "± +10 A-I L :;;
:r
B~ B -PA ~
H/4" ALUMINUM 0
tern at B shows that some useful high-
--
I-
A'). , I
/TUBING
angle radiation exists also, and the higher '"z +5
.........
T._
'"
OJ
V '-'
SCREW AND
/SOLDER LUG
lobe is suitable for short-skip contacts
when propagation conditions dictate the
'"
I-
III
0 '"
I -- ,/ -
Z
;:!
~
. .-
0 -5 80 III
t " i-3/8" SCHEDULE .J
I l,/ '"
~"
J
, 4 0 PVC PIPE
need. A free-space azimuth pattern for the '"L;: -10 60
II!
z
same antenna is provided in Fig. 42. The II /
.'".
OJ 0
> -15 1/,
1 1
GROUND back-lobe pattern reveals that most of the
power is concentrated in the forward lobe.
..
>=
.J -20
OJ
/
"- . .-""
./
-RADIATION
RESISTANCE
40
20
>=
0
I i i " GALVANIZED
I ~WATER PIPE The power difference dictates the front- '" -25
,0 0.05 0.1 0.i5 0,20,250.30,350.4
to-back ratio in dB. It is infrequent that ELEMENT SPACING-WAVELENGTHS
LJ two three-element Yagis with different
element spacings will yield the same lobe Fig. 43 - Gain vs. element spaCing for an
patterns. The data in Fig. 42 are given antenna and one parasitic element. The
~I only for illustrative purposes. reference pOint, 0 d9, is the field strength from
a half-wave antenna alone. The greatest gain is
I Parasitic Excitation in the direction A at spacings of less than 0.14
wavelength, and in direction 9 at greater spac-
In most of these arrangements the ings. The front-to,back ratio is the difference in
additional elements receive power by dB between curves A a,nd B. Variation in radia-
i-1I8" TUBING <12') t-i/4"TUBING U2~) tion resistance of the driven element is also
TELESCOPED 2'3"
INTO i-1I4" TUBING
induction or radiation from the driven shown. These curves are for a self-resonant
element, generally called the "antenna," parasitic element. At most spacings the gain
and reradiate it in the proper phase . as a reflector can be increased by slight .
Fig. 40 - Details of element construction and relationship to achieve the desired effect. lengthening of the parasitic element; the gain
mounting method. The water pipe should be These elements are called parasitic ele- as a director can be increased by shortening.
driven into the ground so that approximately This also improves the f~ont-to-back ratio.
one foot of the pipe extends above ground level. ments, as contrasted to the driven
Feet x 0.3048 = m and inches x 25.4 mm. = elements which receive power directly
from the transmitter through the trans- special case where the parasitic element is
mission line. resonant. It is indicative of the perfor-
The parasitic element is called a director mance to be expected under maximum-
when it reinforces radiation on a line gain tuning conditions.
pointing to it from the antenna, and a
reflector when the reverse is the case. Three-Element Beams
20'
Whether the parasitic element is a director A theoretical investigation of the three-
10' 10' or reflector depends upon the parasitic- element case (director, driven element and
O' ...L_ _--=:s.2!lI_;:;:==_~_L o' element tuning, which usually is adjusted reflector) has indicated a maximum gain
by changing its length. of slightly more than 7 dB. A number of
(A)
Gam VS.,Spacing experimental investigations have shown
that the optimum spacing between the
The gain of an antenna with parasitic driven element and reflector is in 'the
elements varies with the spacing and
region of 0.15 to 0,25 wavelength, with
tuning of the elements and thus for any
0.2 wavelength representing probably the
20' 20' given spacing there is a tuning condition
best overall choice. With 0.2-wavelength
that will give maximum gain at this
10' iO' reflector spacing, Fig. 44 shows the gain
o·~__~~~~~~~~-L~
spacing. The maximum front-to-back
variation with director spacing is not
ratio seldom; if ever, occurs at the same
especially critical and that the overall
condition that gives maximum forward
gain. The impedance of the driven length of the array (boom length in the
element also varies with the tuning and case of a rotatable antenna) can be
Fig. 41 - Elevation-plane response of a three-
element Vagi placed 1/2 wavelength above a spacing, and thus the antenna system anywhere between 0.35 and 0.45
perfect ground (A) and the same antenna must be-tuned to its final condition before wavelength with no appreciable difference
spaced one wavelength above ground (9). the match between the line and the in gain.
antenna can be completed. However, the Wide spacing of both elements is
tuning and matching may interlock to desirable not only because it results in
o' high gain but also because adjustment of
sQme extent, and it is usually necessary to
run through the adjustments several times tuning or element length is less critical and
to insure that the best possible tuning has the input resistance of the driven element
been obtained. is higher than with close spacing. The
latter feature improves the efficiency of
Two-Element Beams the antenna and makes a greater band-
270·'f-----~~~oe=:;:....,.--_+90·
width possible. However, a total antenna
A two-element beam is useful where length, director to reflector, of more than
space or other considerations prevent the 0.3 wavelength at frequencies of the order
use of the larger structure required for a of 14 MHz introduces considerable
three"e1ement beam. The general practice difficulty from a constructional stand-
is to tune the parasitic element as a reflec- point. Lengths of 0.25 to 0.3 wavelength
tor and space it about 0.15 'wavelength are therefore frequently used for this
3-ELEMENT YAGI (FREE SPACE)
from the driven element, although some band, even though they. are less -than
AZIMUTH-PLANE PATTERN successful antennas have been built with optimum.
O.I-waveientgh spacing and director tun- In general', the antenna I gain drops off
Fig. 42 -: Azimuth-plane pattern of a ·three ele- ing. Gain vs. element spacing for a two- less rapidly when the reflector length is·
ment Vagi·in free space. element antenna is given in Fig. 43 for the increased beyond the optimum value than
Ant_"" for High FrequencJ 20·11
465
8 f(MHz) ..........
;:;
~
:t
to
455 '" l'"
1/v
r-- r--
""
f(!'IHz )
, r-- +- ..J
0:
V g445
trl f(MHz )
v
'"
0:
o
5
0.10 0.15 0.20
SPACING FROM DIRE'CTOR TO FED ELEMENT IN k
0.25 0.30
;::; 500
435
f(MHz) 0.1
~ f(MHz ) ..........
0.15 0.2 0.25
DIRECTOR SPACING ( .. )
(A)' '" 0.3
Fig. 44 - Gain of three·element Vagi versus director spacin!l, the reflector being fixed at 0.2
wavelength.
:t
l-
to
..J
!5
490
f(MHz )
'" ........
~
I- '-.....r-.,
t;l
..J
::; 480
1'---1'-...
it does for a corresponding decrease below or the S meter of a communication.s 0: f(MHz)O.1 0.15 0.2 0.25 0.3
the optimum value. The opposite is true of receiver, used in conjunction with a dipole REFLECTOR SPACING ( .. )
. (81
a director. It is therefor~ advisable to err, antenna . located at least 10 wavelengths 466
----
_ f(MHz )
if necessary, on the long side for a away and as high as or higher than the t 468 I--
reflector and on the short side for a beam that is being checked. A few watts of --: !(MHz )
O.~ ~ f---
director. This also tends to make the power fed into the antenna will give a i= 470
antenna performance less dependent on useful signal at the observation point, and
the exact frequency at which it is the power input to. the transmitter (and
~ !(MHz )
. 472
~ f(MHz )
V v~ f---
r--
operated, because an increase above the hence the antenna) should be held
~ 474 ~ V t::.-
~ ((MHz )~ v
design frequency has the same effect as constant for all the readings. 0: 476
o !<MHz)O.l
increasing the length of both parasitic Preliminary matching adjustments can 0.15 0.2 0.25 0.3
DIRECTOR SRACING ( .. )
elements, while a decrease in frequency be done on the ground. The beam should (el
has the same effect as shortening both be set up so the reflector element rests on
elements. By making the director slightly earth, with the remaining elements in a
Fig. 45 - Element lengths for a three-eleme(lt
short and the reflector slightly long, there vertical configuration. In other words, the beam. These lengths will hold closely for tub-
will be a greater spread between the upper beam should be aimed straight up. The ing elements supported at or near the center.
and lower frequencies at which the gain matching system is then adjusted for I: 1
starts to show a rapid decrease. SWR between the feed line and driven
When the over all length has been element. When the antenna is raised into The elements in close-spaced (less
decided upon, the element lengths can be its operating height, only slight touch-up than 1I4-wavelength element spacing) ar-
found by referring to Fig. 45. The lengths . of the matching network will be required. rays preferably should be made oftubing
determined by these charts will vary slight- A great deal has been printed about the of 1/2 to I inch (13 to 25-mm) diameter. A
ly in actual practice with the element need for tuning the elements of a Yagi- conductor of large diameter not only has
diameter and the method of supporting type beam. However, experience has less ohmic resistance but also has lower Q;
the dements. The tuning of a beam should shown that lengths given in Fig. 45 and both these factors are important in
always be' checked after installation. Table 2 are close enough to the desired close-spaced arrays because the im-
'However, the lengths obtained by the use length that no further tuning should be re- pedance of the driven element usually is
of the charts will be close to correct in quired. This is true for Yagi arrays made quite low compared to that of a simple
practically all cases, and they can be used from metal tubing. However, In the case dipole antenna. With three- and four-
without checking if the beam is difficult to of quad antennas, made from wire, the element close-spaced arrays the radiation
access. , reflectors and directors should be tuned resistance Of the driven element may be so
In order to make it even easier for the with the antenna in its operating location. low that ohmic losses in the conductor can
Yagi builder, Table 2 can be used to deter- The reason is that it is practically impossi- consume an appreciable fraction of the
mine the element lengths needed. Both cw ble to cut and install wire to the exact power.
and phone lengths are included for the dimensions required for maximum gain or
three bands, 20,.15 and 10 meters. The 0.2 front-to-back. Feeding the Rotary Beam
wavel€ngth spacing will provide greater Any of the usual methods of feed
bandwidth than the 0.15 spacing. Anten- Simple Systems: the If.otary Beam (described later under "Matching the
na gain is essentially the same with either Two- and three-element systems are Antenna to the Line") can be applied to
spacing. The element lengths given will be popular for rotary-beam antennas, where the driven element of a rotary beam. The
the same whether the beam has two, three the entire antenna system is rotated, to popular choices for feeding a beam are the
or four elements. It is recommended that permit its gain and directivity to be gamma match with series capacitor and
"plumber's delight" type construction be utilizedJor any compass direction. They the T match with series capacitors and a
used where all the elements are mounted may be mounted either horizontally (with half-wavelength phasing section, as shown
directly on and grounded to the boom. the plane containing the elements parallel in Fig. 46. These methods are preferred
This puts the entire array at d.c ground to the earth) or vertically .. over any others because they permit ad-
potential, affording better lightning pro- A four-element beam will give still more justment of the matching and the use of
tection. A gamma section can be used for gain than a three-element one, provided coaxial-line feed. The variable capacitors
matching the feed line to the array. the support is sufficient for about 0.2 can be housed in small plastic cups for
wavelength spacing between elements. weatherproofing; receiving types with
Tuning Adjustments The tuning for maximum gain involves close spacing can be used at powers up to
,.
The preferable method for checking the many variables, and complete gain and .a few hundred watts. Maximum
beam is by means of a field-strength meter tuning data are not available. capacitance required is usually 140 pF at
20-19 Chapter 20
Table 2
Element Lengths for 20,15 and 10 Meters,Phone and CW
A B A B A B A B
14,0~ 33' 5-3/8"· 33'8" 35' 2-112" 35' 5-114' 31' 9-3/8" 31' 11-5/8" 31' 1-1/4" 31' 3-518"
14,250 32' 11-3/4" 33' 2-1/4" 34' 8-112" 34' 11-114" 31' 4" 31' 6-3/8" 30'8"- 30' 10-112"
21,050 22'4" 22' 5-5/8" . 23'6" 23' 7-3/4" 21' 2-112" 21' 4" 20: 9-118'; 20' 10-7/8"
21,300 22' 3/4" 22' 2-3/8" 23' 2-518" 23' 4-1/2" 20' 11-112" 21' 1" 20' 6-1/4" 20' 7-3/4"
28,050 16'9" 16' 10-1/4" 17' 7-5/8" 17' 8-7/8" 15' 11" 16' 15'7" 15' 9-112"
28,600 16' 5-1/4" 16' 6-3/8" 17' 3-112" 17' 4-314" 15' 7-1/4" 15' 8-1/2" 15' 3-3/8" 15' 4-112"
9
1°.151°.151 0.15 I
These lengths are for 0.2- or 0.15-wavelength element spacing.
To convert ft to meters multiply ft x 0.3048.
Convert in. to mm by multiplying in. x 25.4.
14 MHz and proportionately less at the other serving as a parasitic element -,-
higher frequencies. usually a reflector. A variation of the
If physically possible, it is better to quad is called the delta loop. The
CENTER OF ~ adjust the matching device after the electrical properties of both antennas are
ORIVEN ELEMENT n~ __;;.J;~ antenna has been installed at its ultimate the same, generally speaking, though
height, since a match made with the some operators report better DX rest11ts
50 OR 751\ antenna near the ground may not hold for with the delta loop. Both antennas are
COAX LINE the same antenna in the air. shown in Fig. 47. They differ mainly in
their physical properties, one being of
Sharpness of Resonance
plumber's delight construction, while the
TO TRANS.
GAMMA MATCH
Peak performance of a multielement other uses insulating support members.
(A) parasitic array depends upon proper' One or more directors can be added to
phasing or tuning of the elements, which either antenna if additional gain and
can be exact for one frequency only. In the directivity is desired, though most
case of close-spaced arrays, which because operators use the two-element ~rrange
of the low radiation resistance, usually are ment.
quite sharp-tuning, the frequency range It is possible to interlace quads or
over which optimum results can be "deltas" for two or more bands, but if
secured is only' of the order of one or two this is done the formulas given in Fig. 47
percent of the resonant frequency, or up may have to be changed slightly to com-
TO TRANS.
to about 500 kHz at 28 MHz. However, pensate for the proximity effect of the sec-
T-MATCH the antenna can be made to work ond antenna. For quads the length of the
(9) satisfactorily over a wider frequency range full-wave loop can be computed from
COAX
by adjusting the director or directors to
~~j~~~~~i~:,
TRANS. give maximum 'gain at the highest fre- 1005
quency to be covered, and by adjusting the Ful!-wave loop (ft)
TOC1
LOOPA- (
=f(MHz)
TO
TO C2 reflector to give. optimum gain at the 306
COAX LINE- TO
lowest frequency. This sacrifices some Full-wave loop (m) =f('MHz)
LOOP A (FEET)' f7;~z) OF D~~~~E:L. gain at all frequencies, but maintains
4 :1 COAX BALUN
more uniform gain over a wider frequency
(C)
range. If multiple arrays are used, each
. The use of large-diameter conductors antenna should be tuned up separately for
will broaden the response curve of an maximum forward gain as noted on a
Fig. 46 - Illustrations of gamma and array because the larger diameter lowers field-strength meter. The reflector stub on
T-matching systems. At A, the gamma rod is
adjusted along with C until the lowest SWR is the Q. This causes the reactances of the the quad should be adjusted for the
obtained. A T-match is shown at B. It is the elements to change rather slowly with foregoing condition. The delta loop
same as two gamma-match rods. The rods and frequency, with the result that the tuning gamma match should be adjusted for a I: 1
C1 and C2 are adj usted alternatElly for a 1: 1 stays near the optimum over a con- SWR. Nil reflector tuning is needed. The
SWR. A coaxial 4:1 balun transformer is shown
at C. A toroidal balun can be used in place of
siderably wider frequency range than is delta loop antenna has a broader fre-
·the case with wire conductors,' quency response than the quad, and holds
the coax model shown. The toroidal version
has a broader frequency range than the epaxial at an SWR of 1.5: 1 or better across the band
one. The T-match is adjusted for 200 ohms and it is cut foI'.
the balun steps this balanced value down to 50 Delta Loops and Quad Beams The resonance of the qua,d antenna can
ohms, unbalanced. Or, the'T-match can be set One of the more effective DX arrays is be found by checking the frequency at
for 300 ohms, and the balun used to step this
down to 75 ohms, unbalanced. Dimensions for called the "cubical quad" or, simply, which the lowest SWR occurs. The
the gamma and T-match rods are not given by "quad" antenna. It consists of two or element length (driven element) can be
formula. Their lengths and spacing will depend more square loops of wire, each supported adjusted for resonance in the most-used
I upon the tubing size used, and the spacing of by a bamboo or fiberglass cross-arm portion of the band by lengthening or
the parasitic elements of the beam. Capacitors
C, Cl and C2 can be 140 pF for 14-MHz beams.
assembly. The loops are a quarter shortening it.
Somewhat less capacitance will be needed at wavelength per side (full wavelength A two-element quad or delta loop
21 and 28 MHz. ' overall) one loop being driven, and the antenna cempares, favorably with a
:~,,,(/ I EFL.
17' 3-112"
23 2-5/8' ~,
1 8'
I i
=t
L 22' 3/4"
{
f
~J
I
16'5-i/4 11 I
OM
DR IVEN
EL EMENT 8'
V
{
r
FI RST
115' 7-1/4"
DELTA LOOP
Fig. 47 - Information on building a quad of a delta·loop antenna. The antennas are electrically
similar, but the delta-loop uses "plumber's delight" construction.
three-element Vagi array in terms of gain diameter. Another feature of this device is
(see QST, May, 1963 and January, 1969, that it permits.the beam to be tilted after it
for additional information). The quad and is mounted in place on the tower, pro-
delta-loop antennas perform very well at viding access to the elements if they need
50 and 144 MHz. A discussion of to be adjusted once the beam has been
radiation patterns and gain, quads vs. mounted on the tower.
Yagis, was presented by Lindsay in May
1968 QST. A Small Yagi for 40 Meters'
A 7-MHz antenna for most amateur in-
An Optimum-Gain Two-Band Yagi Array stallations consists of a half-wavelength
Fig. 49 - The boom-to-mast fixt.ure that holds
If optimum performance is desired dipole attached between two convenient the two 12-foot boom sections together. The
from a Vagi, the dual four-element array supports and fed power at the center with unit is made by Hy-Gain Electronics.
shown in Fig. 48 will be of interest. This coaxial cable. When antenna gain is a re-
antenna consists of four elements on 15 quirement on this frequency, the dimen-
meters interlaced with. the same number sions of the system can become over-
for 10. Wide spacing is used, providing ex- whelming. A full-size three-element Vagi
cellent gain and good bandwidth on both typically would have 68-foot (20.7-m)
bands. Each driven element is fed. elements and a 36-foot (IO.9-m) boom.
separately .with 50-ohm coax; gamma- Accordi'ngly, half-size elements present
matching systems are employed. If some distinct mechanical as' well as
desired, a single feed line can be run to the economical advantages. Reducing the
spacin~ between elements is not recom-
array and then switched by a remotely
controlled relay. mended since it would severely restrict the
The element lengths shown in Fig: 48 bandwidth of operation and make the
are for the phone portions of the band, tuning critical. This array (Fig. 50)
centered at 21,300 and 28,600 kHz. If features good directivity and reasonable
desired, the element lengths can be gain, yet the mechanical design allows the
changed for cwoperation, using the use of a "normal" heavy-duty rotator and
dimensions given in Table 2. The spacing a conventional tower support. Element
of the elements will remain the same for loading is accomplished by lumped induc-
both phone and cwo tance and capacitance hats along the
38-foot (11.6-m) elements. This design
Construction Details concept can be applied on any of the Fig. 50 - The shortened 40-meter Vagi beam
amateur hf bands. . closely approximates the siz~ of a standard
The elements are supported by commer-
20-meter Vagi. it is shown on a SO-foot (1.8-m)
cially made U~bolt assemblies. Muffler telephone pote.
clamps also make excellent element sup- Construction
ports. The boom-to-mast support (Fig. The system described here uses stan-
49) is also a manufactured item that is dard sizes and lengths of aluminum tubing used. All three elements are the same
designed to hold a 2-inch (51-mm) available through most aluminum sup- length; the tuning of the inductor is
diameter boom and that can be used with pliers. For best mechanical and electrical slightly different on each element, how-
mast sizes up to 2-112 inches (63.5 mm) in performance, 6061-T6 alloy should be ever. The two parasitic' elements are
20-21 Chapter 20
grounded at the center with the associated
boom-to-element hardware. A helical
hairpin match is used to provide a proper
match to the split and insulated driven ele-
ment. .Two sections of steel angle stock
are used to reinforce the driven-element
mounting plate since the Plexiglas center
insulating material is not rigid and ele-
ment sag might otherwise result. The 21' ~
on the inner tubing before it is placed 2 U-bolts, TV antenna to mast type, 1 variable capaci"
inside the boom sections. A pair of 3/8 X tor, 150 pF maximum, any type, 1 plastic freezer con-
3-1/2 inch (9.5 X 89 mm) steel bolts are tainer, approximately 5 X 5 X 5 inches, to house
gamma capaCitor.
placed at right angles to each other at
every connection point to secure the Gamma rod, 3/8- to 1/2-inch diameter aluminum tub-
ing, 36 inches long. (Aluminum curtain rod or simi-
boom. Caution: Do not overtighten the lar.)
bolts since this will distort the tubing =
R X 0.3048 m. In. X 25.4 mm. =
making it impossible to pull apart
sections, should the need arise. It is much
better_ to install locking nuts over the
original ones to assure mechanical se- SWR of less' than 2: I was achieved across
curity. Fig. 53 - Each loading coil is. wound on Plex- the entire 40-meter band with the antenna
iglas rod. The capacitance hats for the
The helical hairpin details are given in parasitic elements are mounted next to the mounted atop an 80-foot tower.
Fig. 55. Quarter-inch copper·. tubing is coil, as shown here. The hose clamps com· The tuning of the array can be checked
formed into seven turns approximately 4 press the tub(ng against the Plexiglas rod. by making front-to-back ratio measure-
inches long and 2-114 inches (102 mm) ID. Each capacitance hat consists of two sections. ments across th~ band. With the di-
of tubing and aSSOCiated muffler clamps.
mensions given here, the best figures of
Tuning and Matching front-to-back (approximately 25 to 30
The builder is encouraged to carefully the resonant frequency (capacitive reac- dB) should be noticed in the cw portion of
follow the dimensions given in Fig'. 54. tance) of the element to cancel the effect the band. Should the builder suspect the
Tuning the elements with the aid of a dip of the hairpin inductive reactance. The tuning is incorrect or if the antenna is
oscillator has proved to be somewhat model shown here had capacitance hats mounted at some height greatly different
unreliable and accordingly, no resonant mounted 2-1/2 inches (63.5 mm) out from than 80 feet (24.3 m) retuning of the
frequencies will be given. the ends of the coils (on the driven elements may be necessary.
The hairpin matching system may not element only). An SWR indicator or
resemble the usual form but its operation wattmeter should be installed in series A Three-Band Quad Antenna System
and adjustment are essentially the same. with the feed line at the antenna. The
F or a detailed explanation ofthis network hairpin coil may be spread or compressed Quads have been popular with
see The ARRL Antenna Book, 13th with an insulated tool (or by hand if amateurs during the past few decades
edition. The driven element resonant power is removedl) to provide minimum because of their light weight, relatively
frequency required for the hairpin match reflected power at 7.050 MHz. The small turning radius, and their unique
is determined by the placement of the builder should. not necessarily strive for a ability to provide good -DX performance
capacitance hats with respect to the ends perfect match by changing the' position of when mounted close to the earth. A two-
of the coils. Sliding the capacitance hats the capacitance hats since this may reduce element three-band quad, for instance,
away from the ends of the coils increases the bandwidth ofthe matching system. An with the elements mounted only 35 feet
Antennas for High Frequency 20·22
7 TURNS 1/4" Dill.
( COPPER TUBING*
PVC PIPE (PLASTIC)
WITH INSIDE DIA. TO
SLIP OVER
ALUMINUM