this connotation being dcscripti\'C of the shaping of the
1sed
electron stream brought about by the specialised formation
of the electrode structure which is characteristic of all such
tubes. The distinguishing features of the final form of
power tetrodes arc:
1. The existence of 'aligned grids', brought about by
the turns of the grid and screen being positioned
direcclv in line with each other.
2. The presence of rwo small 'beam forming' plates held
at cathode potential.
3. The use of critical distance' spacing between screen
and place ro suppress the effects of secondarv emis-
sion from the plate.
It is the second of these three tcatures which, because it
had the ctfo.:t of adding a fi.111:hcr electrode, has resulted in
the still unresolved question of whether such tubes should
be classified as tetrodes or pemodes. To further add to the
confusion the following definition was applied bv RCA in
the 1940 Tube Manual RC14:
'A beam power tube is a tetrode or pentode in which
the use is made of directed electron beams to contribute
substantially to its power-handling capability. Such a
tube contains a cathode, a control-grid, a screen, a plate
and, optionally a suppressor grid. \Nhcn a beam power
tube is designed vvirhout an actual suppressor the elec-
trodes arc so spaced that secondary emission from the
plate is suppressed by space-charge effects between the
screen and plate... . In place of the space-charge effect
just described it is also feasible to use an actual suppres-
sor to repel the sccondarv clccrrons. Examples of beam
power tubes using an actual suppressor arc the 6 V 6 and
6GG'.
A twenty-year history of the 6L6.
The author must admit ro being particularly confi.
over the last sentence as he has never seen a 6V6 with a
suppressor grid nor ever heard of the 6G6G as being other
than a true pcntodc. Bec.rnse the sentence in question was
withdrawn in subsequent editions of the Manual it seems
likclv that the information was incorrect.
But, to continue with our story-the world's first beam
power tetrode was the type 6L6, released by RCA in July
1936.1.' Presumablv because RCA's efforts were being con-
centrated on the production of metal tubes at the time, the
6L6 was first issued in this form. In the following year a
glass version, the 6L6G, was issued in an ST-16 size bulb
whilst later still other variants followed. The final version,
type 6L6GC, was fitted with a tubular T-12 bulb and also
carried an increased plate dissipation of 30 watts compared
with 19 watts for the earlier versions.
Although listed as a receiving type the 6L6 was never
widclv used in this application because the output of a
push-pull pair even when used in class A mode was exces-
sive for all but the largest and most powerful receivers. In
any case the 6L6 \Vas primarilv designed to operate in
Class AB mode as its characteristics were specifically suited
to_ this class of service. Thus the tube found its widest ap-
plication in high power amplifiers and also became pop-
ular with amateur transmitters. Not until a lower power
version appeared did the beam power tube become a rival
to the well-established power pencodc.
As the originator of the species nor surprisingly it was
RCA who developed the first lower power version which,
like its bigger brother, was initially issued in 1nctal form.
Towards the end of 1937 the type 6V6 (metal) and the
type 6V6G (glass) both appeared at much the same timc. 14
The final version, type 6V6GT, was introduced in 1939. 15
71
In this form the tube \\'as to remain in continuous produc-
tion for mTr 40 vcars and must be cou 11tc.:d ;ls one of the:
most successful desig ns of an~· class or tube c.:\·er produced.
In tum its characteristics wen: incorporated into the lokral-
based 7C5 of I 939 ~rnd the miniature tvpl· 6r\Q5 of 1947.
following thl' introduction of bl·am retrOlks no further
.~
new output penrmks were de,·dopc.:d with the exception
of bartcrv-operarcd types. In this c.1se the fan that the
emitter (tilament) \\'as spread out O\'lT .1 greater area, b~·
I
comparison with an c~1ui\·a knt indirecrl~·- heated cathode
tube. made cl1l' application of the be;1111-forming f)rincipk
somewhat rnorl· ditlicult to impkment. B~· 1940. ho\\'-
e,·er. RCA had p rod uced nm 1.4-,·olt r~-pe!> ( 1QSGT .rnd
1TSGT). as wdl as one dual-mirage 1A./2.8-rnlr n·pc
(3Q5GT ).
The 25-year history of the 6V6.
•
In the manufacture o f these and subsequent simibr n ·pcs
the time-honoured , ·ce filament had to be abandoned ;1nd
a Ill'\\' design used \\'here in the tila111enr rook the form or
t\\'O paralkl \\'ire11. Because this st~· k of fila111cnr \\·as nor
adaptable to miniature rubes these had perforce to rrnuin
as pcntocks.
-
Britis/J Al[11ncd-Grid Tctrorics

Apart from th1.· Hi,·ac-I-brries output tetrode prel'i-

o usly mentioned apparently no furth1.:r British de,·dop-
ment "·ork took place, o r ar kasr none was ,·isibk unril
rnid- 1937. At this time the Marconipho nc Co. and GEC
Ltd. jointly launched their so-calkd ' International Octal'
r:inge of , ·ahTs. These \\'ere similar, and in some cases
identical, to the American octal-based g lass series and ;1p-
pear to hal'e been de\'cloped \\'ith the acri,·e co-oper;1rion Osram KT63 , the later version is on the right.
of RCA who at that rime had an interest in the Marconi-
phone Co.
The tirst beam tetrodcs rckascd in the U.K. \\'ere n ·pcs
KT32. KT63, :111d KT66; the letters KT in the: rvpe m1111-

-
bers srnnding for 'kinkkss tctr0tk'. Of the three, on l)' rhe
KT32 W<lS direcr l~· eqtti\·aknt to an~' Amcricm t~·pc, in
this case the 25L6G . The KT63 was similar to the 6\'6G
bur had ratings and d 1aractcristics identical to the 6F6G
prntode. T he ~KT66 was comparabll' co the 6L6G bur had
a somewhat higher plate dissipatio n. Like the 6L6 ir also
underwent de,·doprncntal changes m ·cr the ~·cars ending
up e;lrrying ma,ximum plate and scrern ratings of 500 rnlt:s
when b~' this time a Class AB2 pair was ...:apabk of 50
watts o utput.
Initialh· the de\'dopmrnt ofkinkkss t1.·rrodes b~· M-0.\'.
\\'aS confined to the production of output types bur b~·
1938 Rf types had been added to both rhe nc\\' Interna-
tional range as well as the older 4 -,·olt :-;cries. This dl'\·cl-
o pmenr was without precedent either i11 the U. K. or in
o ther countries and remained unique to 1\tlarconi-Osr.1111 .
Coinciding "·irh the production of the tirsr omput tcr- Marconi-Osram KT66 power tetrodes. The earlier version is on
rodcs in 1937 M-0.\'. adopted a po l ic~· ofgraduall~· phas- the right.

72
ing o ut all their existing ompm pemodes, indud ing b;u-
tery types. and replac ing them ,,·irh equi,·alent kinkless
tetrodcs. In this connectio n it is interesting to note th.it
the production of be;\m-typc battery tl'tr~>dcs occurn:d
so1111: thrc:c \'cars earl i<:r th;\11 in the U .S.
lh 1938 other \·al\'!: makers. initialh- Cossor and Nb :t.lb.
had .marketed o utput tetrodes \\'h ilc H i\'ac continued pro-
d uction of their Harries design. In t he same vear Brimar
produced t he American t~·pes 6V6G, 6L6G. an~i 26L6(; in
their ne\\'I\' introd uced I nrcrnarional Octal range. These
same three. t~·p1:s \l'Cre ;1lso indudcd among the va'h·es listed
b~· Cossor, Mull;mi, and Tungsram in 1940.
Apart from the marketing of the three Arnerican-t~' pe
valves 1111:mio111..:d this W;\ S Mullard's o nlv ackno\\'lcd u;menr
of the th r1:at to the su prem;1cy of the pe;1tode as an <~utput
,·alve. T his \\'as o b\·io uslv because their parent compa n ~' ,
Philips. as o wners of the basic pcnrode patent co uld more
pro ti tab l ~· rnntinue alo ng establ ished lines.
Co11ti11e11tnl Del'clop111e11ts
In conti nental Europe the o utput tetrode made little
impression on thl· scene and always remained a relatively
ran: bre1:d of nib<.:. T he kno\\'n mam1facture rs of t his class
of tubc- Fivre, Loewe, Telcti.mkrn, and Philips- Va lvo-
berwccn them produced o nlv about half •l dozen types
\\'hich, ;1p;1rt from those made b~, Fivre, were all in t he
fo rm o f multiple tubes.
The f'i rst conrinenr;1J use of ;\ tetrode output tube oc-
curred in L938 \\'hen Telefunken produced a dual tetrode,
type VEL LL, for use in one model of the ~azi-insp i red
German 'Peoples Receiver' kn0\\'11 as the Kkinemfanger.'"
The tirst section of che VEL 11 operated as a screened-grid
detecto r \\'hilst the second section \\'as ,1 po\\'er t1:trotk
capable of 2 \\';ms o utput. A second Tclefonkcn multiple
tube, type VCL 11, consisted o f a triode-cum-tetrode hav·
ing an o utput o f 0.8 watts. Two somewhat similar tubes,
types ECLL 11• and UCLl 1, followed in 1939; these lmi a
much higher power o utput of 4 .2 watts.
An ind ication o f the rclat i\'ely unimportant position oc-
cupied b~· the output tetrode in continental Europe can be
obtained from a smdv of Brans' Vndc 1Wecu111. For exam-
ple. in the 1948 edition of this \\'Ork there is listed onlv
one manufacturer, Fine. who made am· American-typ~
beam power tetrodes while amongst t he tew manufac-
tur1:rs \\·ho d id mak1: other types of o utput tecrodcs none
rnatk sing le-sectio n, i.e., no n-multiple, types.
REFERENCES
l. Sec ad\'t . C.E. Mfg. Co. Inc., Rn.dio Ne111s, June 1928,
p. 1367.
2. New A.C. Tubes Developed, Rndio News, April 1929,
p. 946.
3. Recent Advances in Radio T u be Design, Rndio Cm.ft,
May 193 1, p. 599.
4. Ne\\' Types o f Receiving T ubes, Radio Craft, May
193 L, p. 686.
5. W hat's N ew at the Trade Sho \\', Radio Nc111s, Julv
193 1, p. 5 1. .
6. Ririe1-'s Pe1pct11nl Tro11/Jlcs/Jootc1.l_f /vim111nl, Vol. II, p.
8, 'ERLA' model 636 schematic dared April 20, 193 l.
7. Ne\\' Tvpes of Receiving Tubes, Rnriio Cm.fr, April
1931, p. 663.
8. Sec R adio Retailil~q. Ma~· 1932, p. 33.
9. Still More New T ubes, Radio Craft, Sept. 1932, p.
142.
10. N ew Tube Ann oun ce men t~ , f{ iidio Craft, Jan. 1933,
p. 398.
11. New Tube Announcements, /fotfio Craft, Dec. 1932,
p. 334.
12. J.H . Owen H arries, Wireless World, Aug. 2. 1935,
pp. 105- 106.
13. The New Beam Po\\'er Tube, Rnriio Crnft, July 1936.
p. 12.
14. T he 1\farch of Tubes, Rnriio Craft, April 193 7, p.
636.
15. A Guide to the New T u bes, Ac1·0J1ox R esearch Worl?-
cr, July 1939, p. 2.
16. The Peo ple's Set, Wird1:ss World, M arch 1939.
17. Sec Philips' publicatio n T/Jc Brit{rre to H tqhci- Radio
£11te~·tai1111unt dated Sept. 1939, p. l 0.
73
 Cfzapter 'Fen

Double-Filament and Multiple Tubes

A nm-iar.n DcPclopnicnts
The idea of titting a tube with more than one filamenr
goes back to the earliest days for it was in 1909 that th<.:
first do uble-filament De.: f o rest Audions were produced .
Se,·eral ~·cars later other manufacturers rook up the idea,
the Cunningham AudioTron and the Moorhead Electron
Relay being rhc best-kt10 \\·n examp les of unbased doubk-
tilamenc tu bes of the period.
T he first American double-filamcnr cube to be tittt:d
with a standardised base was a type made by Moorhead in
1922 which was known as the A-P Two-in-One. It con-
sisted o f a dual c~'lindrical dt:ctrode assembly joined 'Sia-
mese twin' fashion by m<.:ans of a o ne-piece anode formed
in a fi g ure-o~:eight pattern . The anodes and grids were
effectively in parallel while the two axial filaments wen.:
joined in series and intended for operation from a St<\11-
The New
dard 5-volt suppl~' · H owever, as the conunon point of the
rwo tilarncnrs was connected to the brass base-shell it was
possible to us<.: o ne filament at a rime fro m a 2-\'olr supply.
Apart from a slightly later tube known as the Apco, no
otht:r d o uble-filament cubes appeared until 1940 and by
then they had been developed for an entirely ditkrc.:m
reason. Originally the reason for the fitting of a second
ti lament \\·as to allow an extension of the tube's opn.iring
liti:time, as after rhc first fil~unent h<ld burnt o ut th<.: sec-
ond o n<.: could be brought into use. In the C<lSe o r mo re
rccrnr tubt:s, hmve\'er, the fitting of a double filament
allowt:d eith<.:r series or parallel operation as desired .
Th<.: need for series/parallel filament operation had
arisen in 1940 with the production of mains/battery por-
table receivers which used 1.4-volt drv-cdl rubes. As a
matter of design co1wenicnce when these receivers were
op<.:rating fron~ mains power all rube fi laments were c.:on-
ncc.: red in series, but as the fi lament current of mosr o urpur
rubes was twice that of other tubes in the set it bec1111e
11<.:c.:essary to produce an o utp ut tube having two tilamrnts
which could be operated either in series or parallel as
required. The first standardised double-filament t~' pe, tvpt·
3QSGT, was produced expressly ro meet this requin.: ment
and thereafter double-filamcm output rubes remained in
use fr>r as long as rube-operated portable recei\'ers \\'Cre in
production.

Britisl; Dcvclop1J1mcs

One of the earliest-kno\\'n examples of a British -made

do uble-filament valv<.: was ad\'crtiscd in 1923 1 b~' Phillips
Valves Ltd. of Southall, Middlesex, a small independent
(·Pt;l{A.l'I NO 01\ T A
Filnmc11l V olt:.! 5.0 who also made c.:onvcncio nal single-filament t~1 p<.:s. T he
Fll;11ncnt ~\ mrcr-.1t.1 • 0 . .ts
Pl1•h) Von ' • .,.0 tu zoo
Phillips doubk-tilarnent valve was a bright emitrer typt:
whi1.:h was unique in being fitted with a composition base
at a time when all other makes still h;1d metal-shell b;1ses.
A small link-arm on the undersid e of rhe base allowed the
A=P Radio Laboratories
6~& Mission Strocl San Francisc.o , Cal., U.S.A. second filament to be b rought into use after rhe tirsr one
had fa iled.

74
 An unusual , imked unique, valve known as the 'Nelson
Multi', was produced b~, the Nelson Electric Co. Ltd.
towards the end of 1925. 2 T his valve had no less than three
filamcllts, any one of which could be brought into use
separate Iv bv means of a scissors rvpc switch on the under-
side of the base. In :iddirion ir \\'as possible to adjust the
switch to allow two tilamcms to be run in parallel and so
use the \'alvc as a pO\\'Cr amplifier. Two versions, known
as tvpes A and DEA, were otfrred initially and early in
1926 a further two, types DE2 and DE06, were added to
the range. Shortly afrer this, however, the Nelson Electric
Co. ceased advertising and nothing further was heard of
the company's products.
As in the U.S.A., several years elapsed before doublc-
filament valves again appeared on the scene and when this
occurred they were not of British design. In May 1940
only two manufacturers, Brimar and Tungsram, listed any
such valves and in borh cases it \Vas the American type
3QSGT. By then, however, World War II had been in
progress for some six mo nths and British valve makers
were becoming too occupied with catering for military
demands to have time to develop new types for commer-
cial use, so it was no t until the pose-war ~'cars that other
makers engaged in the production of the 3QSGT and
similar types.
Enr~}' Multiple Tubes
Sooner or beer it was inevitable that someone would
get the idea of purring two o r more identical electrode
Loewe receiver using 3NFB tube.
assemblies inside a single bulb, chus creating a two-in-one
or three-in-one tube. Superficially at least the idea was
attractive as it would appear to offer economics in produc-
tion costs wh ich, if it did not result in two tubes for the
price o f o ne, did make it cheaper than using separate tubes.
It was in German~· that multiple tubes were first devel-
oped serio usly and in chat country the firm of Loewe
Radio AG became the undisputed leader, not only in the
manufacture of multiple rubes but also in the production
of small receivers inco rporating chcm. 3'' Loewe had previ-
ously made l.'.Onvcntio nal rubes under the name ' Loewe
Audi on' but from 1926 onwards specialised in the produc-
tion of 2-in-l :rnd 3-in-l types.
The Loewe 3-in- l was unique in that it contained within
the bulb not o nly rhn.:e separate electrode assemblies but
also all the associated resistors and capacitors necessary for
a 3-tube receiver; only the tuning circuits were, of neces-
sity, o utside the bulb. To prevent contamination of rhe
vacuum within rhe bulb each capacitor and resistor was
indi viduall~, scaled inside a glass phial, and the completed
assembly of electrodes and components was supported by
an intricate arrangement of glass rods and beads. Although
German glassblowers have always been renowned for their
workmanship rhc overall effect in this case was of German
ingenuity gone mad. Even with the most proficient glass
workers such a form of construction must han': been ex-
pensive and in addition was far from robust. Another
rather obvious drawback of the Loewe tube was that in
the eYent of a filament burn-out a major part of the re-
ceiver had to be discarded, though in the Vaterland this
75
was overcome bv the provision of :l repair service \\'hereby At much the same time as the Locwe multiple rubes
new filamenrs could be fitted to a burnr-ouc rube. fi rst appeared, o r a little earlier if the absence of any visi-
In later productions of the Loewe multiple rubes mosr ble gcttering is :mv indicatio n, another Germ.an firm also
of the complicated glass work was dom: away \\'ith and brought o u t 2 -in- l and 3-in· l types. These were produced
use was m:tde of mica sp acers :rnd supporrs. As if ash:urn:d by the Suddeutschc Tcldo n Apparate, Kabel & Draht
of the changed internal appe:i.ra1m: of these latn pro· Wcrke and were sold unde r the brandname TKD o r Tc Ka
ducrio ns Loewe sprayed the bulbs with aluminium paint De. Bo th tvpes were noteworthy for their compact con·
which had no functional purpose and served merd y ro
hide the internal structure fro m view.
In its o riginal form the Loewe recciver consisted of a p~~}}}lll~??ld ~//{{{{(\\\\~
non- rcgcnerati\'e detector followed by a rwo-stagc resis·
tance coupled amplifier bur later productions \\'ere mo di- NELSON MULTI
fied to incorporate regeneratio n in the: d etector circuit. This
necessitated an additional external connectio n on the tube
base, the SC\'Cnth contact being located in the centre of the
existing pin circle. The o riginal 6 -conract ru bes were des·
ignatcd 3NF, while the later 7-co nract rv pes were known
as 31 rB in Germany ;llld RNF7 in England . In both cases
the letters NF indicated N ieder Frequcnz (lo\\' frequency
or AF) .
Ano ther Lo ewe ru be, tho ugh one with completely di f·
fcrenc functions, was also produced at this time. It con·
sisted o f two screen-grid tetrodcs arr;mgcd as a rn·o-stagc
RF ampl ifier ;rnd was known as type 2HF ( HF = H och Fre· A
quenz). The same typc of 6-concact base as used on the VALVE
3NF was titted to the 2HF and a similar sized bulb was WITH
used.
THREE
/,
Fl LAMEN1 '",
II/VO
THE i.JFEOF
THRE€ VALVES I
MOTE THF. SCISSOR SWITCH IK THE DASE CAP.
WHJCH WHl!N CLOSF.I) CAN BF. PL~CED IN
CONTACT Wll'll AN\' DNF. Fll.,UIF.NT. WHEN
OPEN ANO 11' CONTACT WITH A:-1\' T\\'O,
i\' i\JAl-\ES TUE V.-\L\ 'li A POW ER AMPLIFIER.

!'flu& ~
NELSONHuLT I
THREE FILAMENT
'0'~!S'0'~ $

oJM in aaclt dMJta;nd, •1~""'11

-·---
TYPE A . P rlco 9 {6.
BRIC.HT VALVE · Ftramu1 Vetth •lo S. Amps. o-s
DETECTOR · • Plate Volts, lO to 60
AMPLIFIER · • • • . P1·:1tt Volts, so to 100
PO\NER AMPLIFI E R • ·Plate Volts, 80 to 120
(TV'llfo Fiiaments. ) .. Grid 8iu1 J to 6 wolts
Loewe multiple tubes. the centre tube was originally coated TYPE D.E. A. P rice 15/ • .
with aluminum paint. DULL. EMITTER - .. filament Volts 2·6, Amps.0·18
D E TECTOR · · • • PlaltVol!s,301060
AMPLIFIER • • • • - PlateVoltt,6Dfo80
PO\NER AMPLI F I ER • ·PlatoVolts,8010120
(T\NO Fllamont.e. ) -Grid Biu l to 6 Volh
In later ~1cars Loewe went o n to make: AC versions of We I ra Uhlbltint at rhe Wlrcleu Exhibition,
Horticultural Holl, STAND 7&. C•U •ndseeU1em.
their multiple tubes and by 1935 four d iffrrent types were If wurblc lo obtain from your local supplier, write direct to
in productio n. These included such complex types as an
RF penro de, detector triode and output pt:nto de corn bina·
THE NELSON ELECTRIC CO. LTD.
tion which required 13 external connections. T he: prod m:· MERTON PARK
tion of such rubes remained unique to Lo cwt: :tnd was s.w. 19. OCTOBER jTH. 1925.

never attempted elsewhere.

76
struction and small physical size; the 2-in-1 type VT126
having a maximum diameter of onlv 40 mm, whik the
3-in-l t~·pc VT139, which used the same sized electrodes,
mcast;:ed 50 mm. These and other TKD tubes arc notable
for the use of one-piece punched sheetmeral grids-the
so-called ladder grid-as well as for the use of oxide-
coated filaments.
The only early American multiple tube was one which
made a brief appearance during 1927. In March of that
vcar a company named the Emerson Rad Val Corp. of
New York marketed a 3-i.n- l tube known as the Emerson
Multivah·e type E.M.V. -3A. 5 This tube had three small
cylindrical electrode assemblies each having an axial fila-
ment. T hc filaments were series-connected, the terminal
voltage bcing the standard American 5-volt rating. A stan-
dard S-14 size bulb was used, together w ith a standard
4-pin UX base. The base was modificd bv the addition of a
flange or collar which carried four screw terminals, giving
a total of eight external connections. The Multivalve was
almost certainly made by Cleartron as the bulb had the
characteristic flattened spike on the seal-off tip which was
a hallmark of Clcartron tubes.
In 1927 the Multivalve tube is known to have been used
in a receiver marketed under the name 'Babv Emerson' and
also in one made by the Standard Radio Corp. of v\Torces-
ter, Mass., under che name 'Standardyne'."
T he firsr British multiple vah·c appeared late in 1927,
being produced by Ediswan as tvpc ES220. In essence it
consisted of two triodes but these were not constructed in
the form of separate assemblies; instc1d one section was
built inside the other, resulting in the most extraordinary
electrode formation imaginabl~. The £5220 appears t~:>
have been in the nature of an experimental or prototype
valve although an Ediswan receiver known <lS the 'One-
Emerson Multivalve and carton (1927).
TeKaDe 3-in-1 (L.) and 2-in-1 (R.) multiple tubes.
Der' was produced in 1928. Like the valve irsclt~ the re-
ceivers also seem to have been pro totypes on ly.
A year later B.T-H produced a two-in-one rnlve not
identified bv anv tvpe number but it was a fairly co1wcn-
tional reprcscntati\'e of the genre in that it had two inde-
pendent sections provided with a common filament. 7 A
standard British 4-pin base was used with two additional
connections provided by means of two terminals mounted
on opposite sides. B.T-H marketed two models of re-
ceivers using Duplex valves, in fact as far as can be ascer-
tained, the valves were developed solely for use in B.T-H
receivers and were not marketed separately.
Later Twin Triodes
Except in the case of Loewe who continued the produc-
tion of multiple tubes, no further developments occurred
until 1933 when the first of the so-called 'twin triodes',
intended for Class B output stage use, appeared in the
U .5. These tubes were first produced as battery-operated
types and it was in this application that they were most
widclv used. T his was because the inherent efficiency and
econ~my of Class B operation was of g reater impo1:tance
in the case of batterv-operated receivers. The type 19,
which was in the same 2-volt range as the '30' Air CeU
series, was released early in 1933&and found ready accep-
tance by recei\'er manufacturers. It was capable of 2 watts
output under zero bias conditions with 135 volts on the
plate. Under these conditions the static plate current was
I 0 mA, though for greater economy this could be reduced
ro 1 mA by the application of - 6 volts bias, when the
output kll to 1.6 watts. Either way this was a remarkable
77

~:.~!':fl~ ~~~U~tdl:;;::&£~~lDio 6*~
V•h'&. I
~ -~~~~~~~~·-
Ediswan ES220 multiple valve
(1927).
performance when compared w ith the type 33 penrode
operating in Cbss A mode. An octal-based version of the
19 was released in 1936 and a lower-power version, type
1G6G, followed in 1939.
In mid-1933 a 2.5-volt AC version, type 53, and a 6.3-
volt versio n, type 79, were relcased,9 bur as Class B opera-
tion in mains-operated receivers and car radios was littl<.:
used these two rubes were relatively unimportant commer-
cially. Even so there was apparently sufficient demand to
warrant the development o f further types, examples o f
which arc types 6A6 and 6N7, though these were proba-
bly more often used as other than output tubes-for in -
stance as phase inverters or audio mixers.
Meanwhi le, on the ocher side of the Achrntic simibr
developments had been taking p.lacc. It is probably tnie to
say that Class B ope ration of the output stages of battery
sets was even more cnthusiastically taken up in England
than it was in the U.S. This was because of the propor-
tionately greater number of battery-operated receivers in
use in Britain and because economv. in batten· - current w:is
more soughr afrcr.
Bv the end o f 1933 all \'alve makers had at least one
Class B twin triode in production whilst some offered two
different types. As in the U .5., the filament voltage was
standardised at two volts and most valves had characteris-
tics which were.: quite similar to the American type 19.
Examples o f some of the first issues arc: Cosso r 220B,
78
O sram B21, Mazda PD220, and Mullard PM2B . An tUlU -
sual, indeed unique, Class B V<llvc was marketed briefly by
Hi,·ac towards the end of 1934. '0 It was known as type
DB240 and included a dri,·er triode.: in the same bulb as
the output pair. As far as can be determined this Hi,·ac
valve had no counterpart an~'where else in the world and it
is mentioned here just 'for the rccord'.
In spite of the initial enthusiastic reception accorded the
Class B output stage it did not enjo~' as long-hed popu-
larity as in America. One reason for this was that a some-
what similar type of ourpur stage known as quiescent push-
pull (QPP) had always bcen po pular bccausc it did not
require the driving power needed in Class B operation.
Apart fro m this there had been continuing and intensive
development of highly eflicicnt battery-operated output
triodes and pentodes w hich had no counterparts amongst
American rubes and which had the effect of making C lass
B less attractive.
On the European continent thc adoptio n of Class B
appears to ha,·e taken place rather later and to a lesser
extent than occi.JtTed in Britain and the U.S., judging by
the small number of manufacturers who produced suitable
rubes. The first offering from Philips, the type B240, did
nor appear until 1935. T his was followed in 1937 by the
KD D l in the 2-volt 'Goldcn' series of side-contact types.
Examples from other makers arc: Dario TB402, T elefun-
ken RE402B, and T ungsram C l3220.
Class B operation of output stages in mains-operated
n:ccivcrs was almost unheard o f in Europe with the result
that there were no valves developed for this class of ser-
vice. H owever, by 1940 some of those British manufac-
turers making American type valves listed certain Class B
types though no single maker listed all types. O sram was
the only maker to assign a private type number ro any
such rube, their type B63 being the direct equivalent of
the 6N7.
The fi rst tw in-triode voltage amplifiers were the types
6C8G and 6SC7 which appeared in l 940 n These were
followed in 1945 by the better-known 6SN7Gt and the
lesser-known 7F8. When in 1948 RCA introduced noval-
based miniature tubes it so happened that the first two
types issued were twin-triodes-types 12AU7 and 12AX7.
They were fo llowed by the l 2AT7, the three types being
made by all American manufacturers.
Dircct-Co11plcd Tubes
A peculiar!~· American breed of rube, the direct-coupled
amplifier, first appeared early in 1932 \\'hen the Cable
Tube Corp. of Brooklyn, N . Y. announced their Speed
'Triple Twin' type 295 in March of that year. 12 The term
T riple Twin was explained as indicating that the rube had
three times the o utput o f the type 45 triode and twice that
o f the type 47 pcntode. That this claim was slightly exag-
gcrated can be confirmed if the published figure of 4 .5
watts t(x the 295 is compan:d with the tigun:s for the
other two tll bes.
It is probably no t stretching the imag inatio n too much
to suggest that, coming at the ti me when it did, the tube's
prod uctio n was an attempt to ..:apit;1lis1.: o n the publicity
which had surrounded the Loftin vVhitc direct-coupled
amplifier during 1930-31 . Be t hat as it nm-, the Speed
295 pro,-cd to be the progeni tor of a line of si milar rubes,
the de,·d o pmcnt of "·hich rnnrin ued fo r the next fi,·c years
or so.
The Triple T"·in was a two-in-one rube consisting of
an indirectly-heated d ri,·cr t riode din:..:tly ..:oupled ro a
di rectl~·-heated o utpllt triode and was the first American
mult iple tube to emplov two dissim ilar triodt.:s. The type
295 carried a 2.5-volt he:m:r/f'i lamem rating and was soon
fo llo wed bv a 6.3-volt versio n, type 293 and <lll AC/DC
,·crsio n, type 291. As befitted their applications the last
rwo types had much smalkr power o utput capabilities,
though were o therw ise identical in constructio n apart fro m
having smaller bulbs.
Direct-coupled output triodes 1932- 1937.
In spite of ha,·ing had rhe ac.h-anrage of extensive pu b-
lici~· in technical journals of rhe day, cou pled w ith plen-
tiful a<.h-crt ising, the Triple Tw111 tubes were not even
modcratclv successti.i l, as to the \\"ricer's knowled ge they
were never used in a1w commerciallv built recci\'er or
amplifier. The tube's main d rawbai:k \\;as no r so much the
result of an~· shortcomings in irs design ;is the need fo r
rather complicated input and biassing circuitr~'. For CX<\ll1·
pie, a special AF choke was needed in the inpur secrio n's
catho de circuit and such a compo ncnt w;is an anathema in
the davs of resistance-couplcd audio stag1.:s.
For some reason .Speed, in spite of being the originators
of d irect-coupled tubes, d id not continue with their devd-
o pment and it was kfr to two other compan ies, Arcturus
and Triad, to carry o n fu rther developmental work. So it
was that towards the end of l 933, some 18 months after
the arrival of the 295, Arcturus an nounced the type 2B6. u
T he main d ifference between the new rube and its prede-
cessor lay in the use of an indi rectly- heated o utput section,
resulting in a considerably reduced heater wattagc . Othcr
d itfrrcnces were the abo li ti<.m o f thc top-cap co nne..:tio n
and the use ofa large 7-pin base and an ST16 bu lb.
Coinciding with the rdcase of the 2B6 was rhc rdease
of receivers by two companies, Lincoln and Mc1\i\urdo
Siln :r. using the new tube. HO\\"ever, apart from its use by
o ne o r two small manufacturers of public address ampli-
fiers, this appears to h:l\"e been the only occasion when the
2B6 was used in a commercially built receiver. In spitc of
this the 2 R6 had m ade its mark and fu rtht.:r dcvclopmems
were soon to follo w.
T he first direct-coupled tube to achieve real commcrcial
success was the type 6BS anno unced bv Triad in April
1935. 1; Other manufacturers who made this tube were Na-
tional U nion, Ravtheo n, Champio n, and Tung Sol. Unlike
earlier direct-coupled nibcs the 6 B5 found ready accep-
tance amongst manufacturers of both ho me and auromo-
Speed 'Triple-Twin" direct-coupled output triodes .
bile receivers, it being used by such large companies as
C rask\'. and U nited Motors. Undoubtcdh• . the main rea-
son for the success of this new tube was irs abilirv to
replace the common(~· used type 42 output pcntode si~ply
by o mitting the usual biassing circuitry. Bt.:causc th<.: o pcr-
ating conditions of the 6B5 matchcd those of the 42 q uite
closdv no circuit redesig n o r o ther alteratio ns wcre rc-
quired in the receiver, even the base conncctio ns w<.:re the
same. \Nhat more could anv radio manufacturer ;1sk? The
din:ct-couplcd tube had arrived !
Compared with earlier designs the 6B5 had the advan-
tage of req uiring no external biassing arrangemcnts and
in additio n bad somewhat improved characteristics. Ir dif-
fered from earlier tu bes in having an internally mounted
resistor connected between the two sect ions, the stated
79
 RADIO ENGINEERING MARCH, 1932

$1>1! I! I>
TStll>'LB ·TWIN
a 1nasterachieve1nent
from the
Speed Laboratories
The SPEED "TRIPLE-TWIN," a combination
power output and detector tube , far surpasses
all recent developments and is compar;.'-'IP
in importance only t o th~ invon+inn r,f ~he
vac u•.•r> + 11 ~P itself.

Its features are outstanding: Applications

• Ra d io Broddcd st Re ceivers
• Triple th e 245's output and doub le th e 247's with . • Spec ia l Receive rs
out incre ased plate voltages. { Arm~ . Navy, Po lice . Aircraft)
• Te le vision
• One "Triple· Twin" supplants complete DETEC-
• Sound Proje ctors <1nd Recorders
TOR and AUDIO Syste m.
• Thedlre Sound Eq .. ipment
• Super sensitivity allows e limination of pre-stages • Public Address Syst e ms
in spe cial applica tions. • C e ntra lized Radi o

• Flat fre que ncy re sponse, 30 t·o 50,000 cycles. A

• C11rrier Curre nl· Systems
• Communicatio n Repeater Systems
boon to t elevision.
• Sroc)dcast Transm it+er$
• Economy tn chassis construction . • Automatic Phonographs
• Industria l Applic<ltions

CABLE RADIO TUBE CORP.

230 North Ninth Street
BROOKLYN, N. Y.
purpose of which was to prevent warm-u p surges. lkc n1sc
the o utput section o perated with a positive bias it had to
have a high-mu characteristic (to prevent excessive place
current) and this was achieved by using two separate c:on-
trol grids connected in parallel.
Following the introduction of octal-based tubes the 6 13 5
was issued in this form as type 6N6G in 1937. A mctal-
glass version, type 6N6MG, \\'as made by National Union,
Ra~'theon and T riad. Triad also originated rwo lower
power versions, types 6AB6G and 6AC6G, as well as an
AC/DC version, type 25N6G.
In summary it may be said that althoug h the dirccr-
couplcd tube in its later forms enjoyed a certain :imount
of success it nc,·cr became a serio us threat to the wdl-
entrenched pcntodc or beam tetrode output tubes. Dircct-
coupkd output tubes were largely unknown outside the
U.S. tho ug h types 6BS and 6 N 6G \\'ere made, o r at least
listed, by Brimar from 1938 o nwards. \1Vhilst these were
probably intended as replacements for use in impo rted
American receivers it is known that at least two British set
makers, Pilot and Halford, used 6N6G valves in cerr:un
models of receivers made during the late 1930s.
Other Multiple T11bu
A type of multiple tube which originated in the U .$. was
an output pentode cum rect ifier combination intended for
use in small AC/DC recei,·crs. The fi rst of such ~·pes, ~·pc
12A7, was released late in 1933 and it was followed in
1939 by the 25A7G. In 1940 RCA first issued beam o ut-
put tetrodcs combined with rectifiers, examples of \\'hich
are types 32L7GT and 70L7GT.
Another earlv American combinatio n \Vas a v:iri-mu
pentode combined with a triock, the type 6F7 issued in
1933. This tube rcn1aincd the sole representative of its
d;1ss in the 6.3-V . range tho ugh in 1939 a tew manufac-
turers produced a similar tube with a 12.6-V. heater rating
known as type 12B8GT.
The final combinatio n tube issued prior to World War
II was the type 6A D7G which appeared in 1940. It con-
sisted of an output pentodc combined with a driver triode
and it remained the sole representative of this class of
tube.
REFERENCES
1. Sec advt. Phillips Vah-es Ltd., Amateur Wil·eless, Sept.
15, 1923.
2. Sec advt. Nelson Electric Co., Wireless World, Oct. 7,
1925, p. ad. 12.
3. Tubes W ithin T ubes, Radio News, July 1927, pp. 30-
31.
4. Loewe Type R0433 receiver, Wireless World, March
12, 1930, p. 286.
5. Sec advt. Emerson Radval Corp. , Radio NL'WS, March
1927, p. 1081.
6. T rend of Development, Wireless World, Oct. 5, 1927,
p. 482.
7. Show Report, Wireless World, Sept. 26, 1928, p. 401.
8. New Tube De"clopments, Radio Cmft, Jan. 1933, p.
398.
9. Radio Tube Progress, Radio E11gi11eering, June 1933,
p. 28.
10. Sec Wireless Wol'lrl Valve Data, Nov. 30, 1934.
11. Sec RC.A Tube Mrmual, R C14, July 1940.
12. Sec advt. Cable Tube Corp., R adio Engineei·ing,
March 1932.
13. C.H. Stro hmeyer, Audio System with the new 2B6
tube, Radin £11..r1i11cering, August 1933, pp. 11- 12.
14. Cliffo rd E. Denton, The New 6B5 D ynarnic-coupkd
A.F. Tube, Radio Cm.ft, April 1935, p. 590.
81
 Cfiapter 'Eleven
The Return o i the Diode
America11 Dcvclop111c11tS
Fo llowing the invention and tirst use of the diode from
1905 o nwards, a period which spanned little more than
a decade, the progenitor of all radio tubes was guickly
eclipsed by the arrival of the triode detector and manv
years were to pass before the d io de once again came into
its own. Aided by rhc use of regeneration the superior
performance of the triode detector ensured its dominance
for the next 15 years . Not that the triode was without irs
d rawbacks; drawbacks which became increasinglv impor-
tant as developments in recei\'er circuitry took place .
One o f the drawbacks exhibited bv the triode when
used in the original leaky-grid mode ~vas its inabilit\' to
handle large signal inputs without introducing distortion.
In the d ays when pre-detecto r amplificatio n was either
smaJJ or non-existent this drawback \\'as of no signifi-
cance, but following the in\'ention of the screen-grid mbe
with its inherently high ampli fication factor. the increased
pre-detector stage gain led to the detector becoming a
bottleneck. This weakness \\'aS further emphasised with
the coming into use of the supcrhctcrodyne circuit \\'ith its
extrcmclv hig h gain.
As a result receiver manufacturers were soon forced to
seek a better detector and although both triode and scrcen -
grid rubes were used in the so-called 'biassed detector· o r
'anode bend' mode for a fc\\' years, it soon became appar-
ent that further impro,·ement \\'aS desirable. The short-
comings of amplifying rubes when used as detectors kd
directly back ro the diode, \\'hich \\'hen resurrected \\'as
found to be as near ideal a detector as could be dc,·ised.
Once this fact had been established it o nlv remained to
put it into practice, bur at the time there " ·as no such
things as diode detector tu bes in productio n. Recei\'cr
manufacturers were therefore forced to ado pt the make-
shift practice of using triodes as diodes. This \\·as simply
done by using the grid as the anode \\'hik the true anode
was usuallv earthed and thus functioned solclv as a shield.
Connected in this manner the type 27 rube .was success-
fullv
. used bv . several American manufacturers until the
arri,·aJ of specialised detector diodes.
82
The earliest-known commercial use of diode detection
carried o ut in this manner occurs in the Edison models R4
and RS receivers marketed in mid -1929. 1 In th<: fo llo wing
vcar Philco sirnilarlv used d io de detectio n but in this case
~vith the addition (;f auromaric gai n contro l (A VC) , en:n
though vari-mu tubes had nor then made their appearance.
In both cases rhe receivers were of the TRf v:triery bur in
1931 the Philco model 90 (bre) appears to ha\·c been rhe
first supcrhcterodvne ro incorporate diode detection. In
the same \'car Grigsby·Gruno\\' \\Tre the first ro use full-
wave detectio n using two t~'pc 27 rubes, connected as
diodes, in their mode l 25 recci,·er. 1
Duo-diodes 6H66 and GH6 The first duo-diodes
(1936). (1931).
Finalh'. afrer more than cwo ,·cars of these makeshift
arrangements. the first ' modern· diode detector tube was
put into production. E\'cn then it \\'as nor a\·aibblc l()r
general sale as the tube in question " ·as made b~· a recei,·er
manufacturer solely for use in the company's O\\'n sets. In
December 193 1 the (~ri gsby-G nmo"· Co. rde.\sed rheir
model 200 recc i,·cr which used the ne\\'ly de,·d oped Ma-
jestic type G-2-S duo-dio d e rube. This rube, rogerhcr \\'ith
the later G-4-S, rcm:tined uniquc o n the Arnericm scene
until the ad,·enr of the rype 6H6 metal rube in 1935.
The arrival of the G-2-S ma~· be said ro ha,·e sign.lllcd
the Start of 3 \\·orld-\\'ide S\\'ing. to\\'ards the USC of diode
detection using rubes specially designed for the purpme.
E,·cn so triodes \\'ere still being pressed into scn·il.'.e a~
diodes as late as 1933 by several well-known manufactur·
ers, such as Crosley and Philco, although by this time
duo-diodes in combination with triodes or pcntodcs had
come into general use.
Another factor of perhaps equal importance in bringing
about the general use of diode detection was the invention
of the so-called 'automatic vo lume control' (AVC), a fra-
ture which was greatly facilitated by the use of diode de-
tection. Prior to this A VC could o nly be obtained by the
use of a separate tube, often needing a rather complex cir-
cuit in order to obtain the requisite control voltage. When
diode detection was used it was a simple matter to obtain
' free' AVC by using the rectified signal voltage. So it was
that within the space of two or thn:e years diode detection
came to dominate the scene <md has retained its popularity
to this very day, even though vacuum tube bas been super-
ceded by crystal. Norman E. Wundorlich

Because o f the exceptionally small emission require-

ments of a signal diode it was soon realised that it was
possible to incorporate a couple of diodes into the same
envelope with an existing triode without the need to in-
crease either the heater wattage or the size of the electrode
assembly. Furthermore, the triode could directly follow
the diode in receiver tube sequence as a first stage audio
amplifier so that the combination of the two stages witl1in
a single bulb was a logical step. T his led to the develop-
ment of the so-called duo-diode-triode tube wherein two
tiny diode plates were incorporated into the structure of
a triode and shared a common cathode. The fi rsc of such
tubes was the type 55 which was announced in July 1932.
Much has been written about the Arcturus 'vVunderlich'
tube, a conremporary of the 55, which was announced in
Ma\' 1932.4 Like the 55 the Wunderlich was essential!\· a
form of diode-triode but because of the unusual meth.od
of combining the functions of detection and amplification
in one electron stream (unlike the three-stream 55 ) ir de-
serves special mention. Firstly rhc German-sounding name
was that of the rube's inventor, Norman E. Wunde rlich, a
former e ngineer of the old Radio- Victor Corp. who ap-
pears to have invented and patented the tube befo re sell -
ing the rights to the Arcturus Radio Tube Co .
The distinguishing feature of the Wunderlich lay in the
structure of its cwin grids. These were arranged in co-
cylindrical, or to be more precise, co-planar fashion around
a common cat hode and in turn were surrounded bv a
The Wunderlich tube and its inventor.
common (triode) plate. Functionallv the nibc could be
regarded as either a leaky-grid triode having two grids, or
as a rube having t wo open-mesh diode plates which al-
lowed the electron stream to pass through and n.:ach the
outer plate. In practice the Wunderlich was always used as
a full-wave detector and in fac t this mode o f operation was
inherent in its design . As with the 55 it was a simple
matter to obtain an AVC control voltage from the rc<.:ti-
tied signal, though the fu ll-wave circuit configuration wns
somewhat inflexible when it came to other than 'simple'
AVC. To obviate tl1is shortcoming a finaJ. form of the tube
known as tvpe ' B' was announced in March 1933 wherein
an entirely separate diode plate was added to the existing
assembly. H owever, due to its comparati\-c late arrival this
particular m be was destined to become little more than a
curiositv
. as it was neve r used commcrciallv.
.
Historically the Wunderlich was a dead-end tube; its
invention did not g ive rise ro a new generation of similar
tubes and within the space o f a few vears it was all but
II.

Radiotron 85 (6.3V) and 55

(2 .SV).

83
forgotten. The original 2.5-V Wunderlich tube was not
allocated any type number but after the production of a 6.3-
V car radio version, identified as ' A-Auto', the 2.5-V ver-
sion became known as type A.
Contrary to a widely-held belief~ the vVunderlich design
was not unique as at much the same time two other tube
makers had released similar types. These diffe red only in
that they were two-stream variants, but the principle of
operation remained the same. \Nhereas the Arcturus Wun-
derlich used a conunon cathode, tubes made by Ken-Rad
and Sylvania used twin cathodes each with its own grid,
the twin assemblies being surrounded by a common plate.
Examples arc: Sylvania types 29 and 69;' Ken-Rad KR20
and KR.22.~ Although the \Nunderlich was used by several
receiver manufacturers during 1932-33 there is no record
of any conunercial use of either the Ken-Rad o r Sylvania L. Mu/lard duo-diodes for 2 -volt battery operation. R. Mazda
versions. V914 (formerly AC/ DD) 4-volt AC.
For the next forty years the duo-diode triode was to
reign supreme as the most commonlv used multi-function atclv superseded by the type V. 914, the.: carlicr type being
radio tube and even though diodes were later combined no longer listed aftcr 1934.
with either AF or RF pentodcs these particular combina- Initially British valve makers offered plain duo-diodes
tions never became as widely used as the DDT. onlv and not until later in 1933 did the.: fi rst duo -diode
By 1933 duo-diode pentodes in both 2.5-and 6.3-volt tri~des appear. In addi tio n tO producing duo-diodcs t<x
versions (types 2B7 and 6B7) had appeared and by 1937 AC <llld AC/DC operation British and contincnral manu-
they were joined by a 2-volt battery version, type 1F6. lt facwrers also made battery-operated types, a class of tube
should be me ntioned here that the combining of diodes completclv unknown in the U nited Statcs. Of the last n:uned
with oucput pentodcs or tetrodes was unknown in the the first British example was the Cossor 22000 which
U.S. in spite of its popularity in Euro pe. When the first appeared in 1934.8 A variation also unknown in the.: U.S.
1.4-volt miniature battery tubes were released in 1940 it is was a 2-volt battery type having an indin.:ctly-heared cath-
interesting tO note that a duo-diode triode was not in- ode. This particular type o f valve.: was produccd on l~' by
cluded in the range, only a diode-pentode, tvpe 155. In Philips-Mullard and was intended for use in their own
the case of the 7-pin indirectly- heated miniatures it was recei\·ers. T he Mul)ard 202 and the Philips Kl32 wcrc
impossible to produce a duo-diode pentode due to the first listed in 1936. In spite of being indirecrl ~· hcned
limitation of the number of base pins. these valves consumed only 0.09 amps, a figure about half
During World War II the first American miniaturised that of most ordinary 2-volt types.
twin diode, tvpe 6AL5, was introduced bv Syh·ania. T his During its heyday the plain duo-diode was much mo re
tube quick!~' became an industry standard, made by all widclv used in Britain than in America, which ma~· seem
other makers. Prior to its advent it had been necessarv to surprising in Yiew of the fact that British receivers of the
use a triode strapped as a diode when a miniature tvpe was period generally used fewe r valves than their American
required. It may be recalled that a similar siniation had
arisen manv vcars earlie r in the case of standard sized
tubes. The 6AL5 was announced in August 1944 and was, Cl
of course, intended for milirarv applications. Ati:er the war
it found extensive use in FM and television recei\'ers.

British Devclopmeuts

Following American practice the use of diode detection,

together w ith AVC became commonplace in Britain from
1933 onwards. The first British receiver known to have
ii
JJ1'
utilised both these features, the M urphv model AS, was
being advertised in November 1932. An article in Wireless
World stated that a Mazd a type AC/DD valve was used.- Philips and Mu/lard duo-diodes . They came in various shapes
For some reason this particular valve was almost inunedi- and sizes.

84
counterparts. The n.:ason for the popularity of such ,·ah-cs
with British set makers mav have been due to a drawback
connected with the use of some types of diode combin:i-
tions which m:initcstcd itself as an unwanted coupling be·
tween the sections. Such coupling could be complctcl~,
eliminated only b~· the use of t\\'O separate vah-cs for the
func tions of detectio n and AF amplificatio n. Be that as it
may. the duo-diode retained its populariry in the U.K. up
to the cessation of domestic rccei,·cr p roduction earl\· in
World War II.
From the rnon.: conservative British standpoint the use
of a separate valve solely for the purpose of detection and
AVC could be considered wasteful. That the overall mun·
bcr of valves was not necessari1~· increased by this proce-
dure can be n:phlincd by the fact that in such cases the
use of an extremely sensitive output pentodc, plus a heavily
delayed A VC s~1s tern, allowed a penultimate stage to be
dispensed with.
The first British duo-diode triodes appeared during
1933, one of the earliest being the Marconi-Osram MHD4
announced in ;..i1a~· of that year. 9 By comparison the first
.Mullard combination valve, type SD4, d id not appear until
the end of 1933. This ,·ah·e was a single-diode tetrode and
was not marketed scpararely bur formed part of the com·
pkrncnt of a British-built Philips receiver, model 634A.
Philips £444 'Binode' fitted with American 6-pin base.
The equ ivalent Philips 'Binode' type £444, was used in
continental Europe.
By the end of 1933 most valve makers were producing a
variety of duo-diode triodes suitable for operat ion in all

A group of dual-grid detector tubes (1932).

85
 The 6ALS fills the need for a high perveance twin diode with the
low voltage drop required for m any special r.f. circuit applications.
WPB and the Services consider diode connection of the 6]6 twin
triode (and other triodes) to be a wasteful misuse. With minor
changes of socket wiring, the 6ALS easily replaces the diode-
connected 6]6.
Specifically manufactured and rated as a diode, the 6ALS is
tested as a diode. Close production control keeps within a narrow
range the cutoff characteristic in the contact potential region. De-
signed throughout for efficiency on high and very-high radio
frequer.cies, the 6ALS ha s a separately connected shield which
may be grounded to isola te the two diodes and their associated
circuits. A midget miniature bulb permits extra space savings.
Possible uses include: Detector and AVC, clipper, limiter,
FM frequency discriminator, special high-frequency diode,
power rectifier.

TYPE 6AL5
HYTRON twin Diode
... 1\.frequene)'
Very-" 19 lllSllCS
llllCAL CHAllACll ... 6.3 volts
£LlC DC) . .... . ... . .. . . 0.3 amperes
H ter potential (AC or . . . . .. . . . . . . .. . ·460maic. volts
H~~ter current· te~ti;it : : . .. . . · · · · · · ·. ·. ·. ·. 350 max. v~':.
Peak inverse po otentialt .. . . .. . . . . ·. ... ... 60 max. ma.
Heater-cathode P t per p\atet · · · · · · · .. 10 ma:ic. DC
Peak plate curren ent per p\atet ...... .
Average plate curr CtllOt>E CAPACllANCES 0 015 mmf.
. 1NTEltlLl . . . . . . . . . . . . . . . . 2.8 ·mmf.
to plate 2 · · · · · · · · · · . .. .. · · · · · · · · · · · 3,5.mmf.
Plate l d ,. . .. . .... . . .. . .. . . d
Plate to catho ~e . ·. ·. '. . .. . . . ... : .' . ~i~s~-ntting sh1el .
Cathode t<;> all . are averages with
Capac1tancesHYSICAL CHAllACl£1llSTICS T-5~ midget
P . . ... · · ·. ·: · · · · button 7·P10
......... Nhniature . h ma:ic.
B lb . .. ··· ·· · · ·· . . . .. · · · · · l.82 1nc es ·
u ........ " . . . . . . .. .. O 75 inch ms1t.
Base · · · · ··· ·
Height overall ... . .. . . : : : .... ... . . . : . .,;,~"imum• ~hould be
Diameter . ..... · .. ·~~~n urc (}bso\ut~; ~i~~g.!'o\ta~e "ariat,ons.
t Mrucimum ro~•of~o/o lower to ollo.J:odc scctioOS· i .
appro~irnnte y co{ tbe t wo t w'n l
• Va\uc \t\ for on f

tUBE5

BUY ANOTHER WAR BOND

classes of service. As in the U.S. , diodc-pentodc combina-
tions were less common and those that were available dit:
te red in that the diode units were usuallv combined with
output pentodes. American practice was to combine only
voltage-amplifying pentodcs with d iodes. One o r two ex-
ceptions to the British no rm may oc mentioned: in 1933
Cossor introduced their DD/Pen in which the pentode
unit was d esigned co operate as an A VC amplifier, while
in the same year Lissen produced a vari-mu penrode com-
bined with a single diode. A similar valve incorporating
two diodes, the Marconi-Osram 'ND40, was marketed in
1935. Apart from the Mullard example mentioned earlier,
no o ther British m:mufacturers made diode-tetrode com-
binations although they they were fairly common on the
continent and were marketed in the U.K. by Triotron and
Tungsram.
The first duo-diode output pcnrode to appear was Maz-
da's AC/PenDD, marketed in 1934. 10 The pentode section
of this valve was of the hig h sensitivity variety, identical
to the AC2/Pcn. Fo llowing this came Fcrrami's PT4D in
1935, and by 1936 other manufacturers had similar types
in production. Surprisingly, Mullard's Pen4DD was not
listed until 1937 .
ln 1934 Mazda became the world's first valve maker
to issue a combination triode-triple-diode valve known as
tvpc AC/HLDDD. It was one o f the first valves to use the
newly in troduced 9-pin base.
Following this came the Philips-Mullard triple diode,
type EAB 1. 11 This class of valve was quire a raritv both at
the time and subsequently, and was intended mainly for
the equipping of Philips and Mullard receivers which used
an elaborate system of disrorrionlcss dcla~icd A VC. T he
EAB l was <Umounccd in 1938 and no similar valve was
produced by any other British manufacturer. In the U.S.
no sim ilar tube was seen until 1953 when the type 6BC7
appeared.
With the imroductio n of FM broadcasting in the U.K.
in the post-war years the need arose for a triple d iode
for use in combined AM/FM receivers. In this case three
diodes were combined with a high-mu trio de, the bcst-
known example o f which was type EABC80.
Before leaving the subject of diode detectors a pecul-
iarly British contribution to the genre remains to be dis-
cussed. This is the 'Wcstcctor', a small radio version of the
Westinghouse copper-oxide rectifier better known for its
battery charging and radio HT supply applications. Al-
though a solid-state device no apology is offered for in-
cluding rcfcn:nce to it here for none other than the august
Wireless World published derails of Westectors year after
year in its valve data pages. And what is good enough for
Wireless World is good enough for the author!
Wcstectors were a pro duct of the Westinghouse Brake
& Saxby Signal Co., being first marketed in 1933. 12 Due
to their inherently high self-capacity the earliest types
were limited to use as second detectors in superhctcro-
dyncs where the intermediate fregucncy did no t exceed
about 150 Kc. Il}' 1934 the upper frequency limit had
been extended to 1500 Kc, allowing their use at higher
intermediate frequencies or even in TRF receivers. In spite
of this Wcstectors found onlv limited acceptance in the
pre-war years and at no time were they able ro challenge
the superiority of the thermionic d iode. In the years fol-
lowing World War II product ion of Westinghouse coppcr-
oxide diodes was continued for several years in spite of the
coming into use of germanium and other types of diodes
which, incidentally, the company also produced.

REFERENCES
WESTECTORS
1. Sec Ride1,.,s, Vol. 1, p. 7, Edison models R4, RS,
HIGH FREQUENCY METAL RECTIFIERS dated Mav 1929.
2. Sec Ride1,.,s, Vol. 2, p. 8, Majestic model 25, dated
Nov. 11 , 1929.
3. Service, July 1932, p. 160.
4. More New Tubes, Radio Craft, May 1932, p. 654.
5. Still More New Tubes, R adio Craft, Sept. 1932, p.
142.
6. Sec Km-Rad Data Sheet, dated August 1932.
7. Practical Automatic Volume Control, Wireless World,
TYPE WM.24 OR WM.26.
Actual Size. Jan. 6, 1933 .
8. Sec Wireless Wo1'ld Valve Data, Nov. 30, 1934.
9. The Double Diode Triode, Wireless World, Mav 19,
Model. Type. Lenglh. Maximum Safe
I nput Voltage.
rl M~m~
Current
Output.
1933, p. 355.
10. New Diode Pcnrodc Ourput Valve, Wtrdess World,
.

W.4.
W.6.
WM.24.
.. Half-wave..
Half-wave.
Full- wave centre
Ir
I
3•
•
24 v. peak carrier
36 v. pe.a k carrier.
24 v. each side of
o · 25 m/ a
o ·25 m/ a
March 2, 1934.
11. The Wcstector, Winless Wodd, Jvfarch 3 , 1933, pp.
tapped. centre tap.
173-74.
0.5 m/a
WM.26. Full-wave centre 3• 36 v. each side of 12. New Range of Valves, Wireless World, June 2, 1938,
tapped. centre tap. 0·5 m/a
p. 499.

87
 Chapter 'Iwelve
Frequency Changers
Pentagrid H eptodes
In order to trace the origin of tubes designed to incor-
porate the functions of oscillator and mixer in one unit
(for use in superheterodync circuits) it is necessary to
know something of circumst<Ulces surrounding their in-
troduction. The earliest supcrhctcrodync receivers made
use of a separate oscillator tube coupled to another tube
which rectified (detected) the incoming signal. This was
the so-called 'additive' svstcm in which the mixer tube
became known as the ' first detector' co distinguish it from
the normal detector, now known as the 'second detector'.
This system was quite satisfactory and in some cases n.:-
maincd in use for manv vcars after the introduction of
special multiple tubes.
Initially at least, it seems to have been economic rather
than technical considerations that led to the development
of special tubes designed to combine the functions of os-
cillator and mi.xer in one unit. Because the superber circuit
came into general use at a time when the prevailing eco-
nomic depression had forced most receiver manufacturers
to produce small low-priced models, it was conscqucntlv
necessary to devise wavs of making a cheaper supcrhctcr-
odyne before this class of receiver could compete with the
TRF type. The first step in pruning production costs was
achieved by dispensing with an RF stage and using in its
place a pre-selector (bandpass) input circuit, while later
the use of a higher intermediate frequcncv enabled even
this to be dispensed with. The second step consisted in
using a self-oscillating mixer tube thus avoiding the need
for a separate oscillator tube. Coupled with o ther econo-
mics these two steps made possibk the production of a
highly efficient five tube receiver which remained the
mainstay of the industn1 for many years. So it was that the
single-rube mixer circuit, known as the ' aurodync', came
into use, but within the space of two years it had been
rendered obsolete by the coming of combination mixcr-
oscillator tubes.
In the case of battery-operated receivers although it was
possible to use the autodyne circuit practical difficulties
88
made it less popular than in mains-operated sets. Because
the autodyne mode of operation called for rhe tube's (fila-
mentary) cathode to be at an Rf potential with respect to
13- or earth it was necessary ro isolan: it b~· using Rf
chokes in the filament leads. As a result of this complica-
tion many receiver manufacturers prdcrrcd to use a sepa-
rate oscillator tube coupled to the first detector. Late
in 1932 a unique indirectly-heated 2-volt b:mcry tetrode,
type 15, 1 was released b,· Svlvania to facilitate the use of
the autodvne circuit. So efficient and reliable was the 15
that it remairn:d popular even after the introduction of
battery-opcr:ited penragrid rubes in the fo llowing year.
Whi lst the self-oscillating mixer d id save the use o f one
tube this was just about the o nly virtue it had. Amo ngst
its drawbacks was the fact that because, like its predeces-
sor, it had to rccti(v the incoming sig1uJ this nude it
impossible to apply an~' fr:mn of gain contro l to rhe mixer
stage. Because, as an c.x:onomy measure, most small n.:cciv-
crs had no Rf stagc, a satisfacton· control of gain was
difficult to achie,·c when irs application was confined to
the IF stage alo ne. T he nced fo r a single rube which could
combine the functio ns of oscillator ;rnd mixer ;tnd ar rhc
same time allow the application of an AVC volragc was
never greater.
In April 1933 RCA announced the rypc 2A7, :md the
pcntagrid converter was bom .2 Altho ug h this tube w:is not
strictlv the tirst in which the functio ns of osci llator and
mixer h:id been combined- the French bi-grille rubes had
earned this distincrion- ir was the tirst co achieve the so-
callcd 'multiplicative' mixing. The 2A7 was single-stream
tube with all rhc clcmenrs disposed radially around the
cathode, with the two innermost grids forming rhc osci l-
lator section. The next electrode in sequence dcscr\'cs spe-
cial mention as it actcd as a space-charge g rid ;md formed
a so-called '"irtual cathode' for the control grid (G4). It
was this lcature which allowed the control g rid to be given
\'Jri-mu characteristics thus permitting the application of
gain contro l to rhc mixer stage.
So successfu l was this design th:-ir it remained an indus-
try standard fr)r many years and later termed the basis for
the development of 'ocrode' converters. Subsequent ver-
sions of the 2A7 differed only in heater voltage or type·
of base, the sequence being 6A7, 6A8, 6A8G, 6A8GT,
12A8GT, 7B8. Battery-operated equivalents of the 2A7
were not long in making their appearance, the first of
which was type 1A6.-; Unfornmatcly, due to the minimal
filament current used, the 1A6 was somewhat deficient in
performance. In particular the mutual conductance of the
oscillator section was rather on the low side resulting in
the 1A6 being unsuitable for short-wave working; in fact
the tube was even prone to stop oscillating on the broad-
cast band. As an expedient to overcome this problem,
pending the arrival of an improved rype of tube, receiver
manufocnirers sometimes wired a separate triode in paral-
lel with the oscillator section, a practice which led some
wag to dub the 1A6 'The tube with an outboard mutual
conductance'.
An improved version of the l.A.6, known as type 1C6,
appeared in August 1934,' and due to an increased em.is-
sion (obtained by <U1 increase in filament consumption)
and improved mutual conductance of the oscillaror sec-
tion, no further troubles were experienced.
~·
'
: • ·1-r , 1
!. ~ , .
te• ~ #
Early American frequency changers. The tube on the left was a
2-volt battery tetrode with an indirectly-heated cathode.
Little time was to pass before the pentagrid crossed the
Atlantic (where it became known as the heptode), for by
the end of 1933 Ferranti had produced the first British
version, rype VI-:lT4.; Because of the short time lapse be-
tween the appearance of the American and British types,
coupled with the similarity of the ferranti version, it
seems likely that Ferranti Ltd. with their on-the-spot rep-
resentation in the form of Ferranti Inc. of New York,
must have obtained a manufacturing license from RCA to
enable them to be first in the field in Britain. This sug-
gestion is given credence by the fact that Ferranti Ltd.
was a newcomer to valve production, having been engaged
in valve making for only one year at this time.
Nearly a year was to elapse before other British valve
makers marketed heptode converters, but by the end of
1934 all except Mazda and Mullard had foll ranges in
production. The fact that two of the largest makers did
not produce heptodcs may be taken as an indication that
this class of valves never dominated the British scene in
the way that the pentagrid dominated the American mar-
ket. \Vithin a year of the heptode's British debut other
types of frequency changers were on the market and had
found ready acceptance by receiver manufacturers.
Hexodes and Octodes
Apart from the pemagrid the only other early American
multiple tube capable of being used as a converter was the
type 6F7, a triode-pentode originated by RCA in 1933. 6
Unlike European rypes the 6F7 was not specifically de-
signed as a frequency changer and was in fact more often
used in other applications as evidenced by a contemporary
magazine article entitled 'Six Uses for the 6F7'. The two
sections of the 6F7 were electrically independent apart
from the use of a common cathode, the need for which
was dictated by the maximum number of base pins (7)
available.
At a slightly later date the triode-penrode mixer enjoyed
a limited degree of populariry in both Britain and the
continent. In 1934 Mazda produced the first British ver-
ll
~·
.1}
.
.
.
Ii:- ".~'~ ,..
RCA 6F7 triode-pentode (L.) and RCA 2A 7 pentagrid converter
(R.) 1933.
89
 THE WIRELESS WORLD ADVERTISEMENTS.

UNDREDS of thousand~ o f Ferranti Valves are at

H present in use, giving daily dem onstration of
th eir efficiency and reliability. Wh at finer testimony
than the acclamation of even the n.ost critical of Radio
experimenters?

Ferranti Ltd. have h ad great experience in th e precision

manufacture of fine M echanica l and El ectrical detai ls.
This , especially applied to the design and manufacture
of valves, has resulted in a very high standard of quality
and remarkable uniformity of valve performance. Every
Ferranti V a lve is subj ect not only to normal valve
t est s but also is tested in a r eceiver under norma l working
conditions.

D esigned and constructed entirely in the famous

Hollinwood works, Ferranti Valv es have taken th eir
place in the forefront of Progress towards true R ad'io
reproduction.

The valve illustrated is th e F erranti H eptode V.H.T. 4,

the first valve of its kind made in England, which
combines in one valve the function of both oscillator
and modulator and in addition is a variable Mu type
enabling full A .V.C. to be obtained in sets with only
one l.F . stage.

There are A.G. Mains, Battery and Universal valves

available.

Price 20 1-
Write for Valve Catalogue to
FERRANTI LTD., HOLLINWOOD, LANCASHIRE

Me11tio 11 oj "The ll'i r de; s W ot ld," w lte11 wrili11g t o advert ise.-s, :l'ill e11s11r~ prom{'/ at/e11ti.;n.