EPElectronics 2000 03

Volume 2 Issue 3
March 2000
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, Febuary 1999 - www.epemag.com - XXX
Maxfield & Montrose Interactive Inc
Copyright  2000, Wimborne Publishing Ltd
and Maxfield & Montrose Interactive Inc.,
PO Box 857, Madison, Alabama 35758, USA
All rights reserved.
WARNING!
The materials and works contained within EPE Online — which are made available
by Wimborne Publishing Ltd and Maxfield & Montrose Interactive Inc — are
copyrighted. You are permitted to download locally these materials and works and to
make one (1) hard copy of such materials and works for your personal use. International
copyright laws, however, prohibit any further copying or reproduction of such materials
and works, or any republication of any kind.
Maxfield & Montrose Interactive Inc and Wimborne Publishing Ltd have used their
best efforts in preparing these materials and works. However, Maxfield & Montrose
Interactive Inc and Wimborne Publishing Ltd make no warranties of any kind, expressed
or implied, with regard to the documentation or data contained herein, and specifically
disclaim, without limitation, any implied warranties of merchantability and fitness for a
particular purpose. Because of possible variances in the quality and condition of
materials and workmanship used by readers, EPE Online, its publishers and agents
disclaim any responsibility for the safe and proper functioning of reader-constructed
projects based on or from information published in these materials and works.
In no event shall Maxfield & Montrose Interactive Inc or Wimborne Publishing Ltd be
responsible or liable for any loss of profit or any other commercial damages, including
but not limited to special, incidental, consequential, or any other damages in connection
with or arising out of furnishing, performance, or use of these materials and works.
Copyright © 2000 Wimborne Publishing Ltd and EPE Online, March 2000 - www.epemag.com - 164
PROJECTS AND CIRCUITS
PARKING WARNING SYSTEM - by Tom Webb 167
Avoid having an unwanted rear entrance to your garage!
HIGH PERFORMANCE REGENERATIVE RECEIVER - Part 1 173

by Raymond Haigh
Beautifully designed receiver covering 130kHz to 30MHz global broadcast and amateur bands
AUTOMATIC TRAIN SIGNAL - by Robert Penfold 179

Easy-to-build, low-cost Starter Project for model railways
INGENUITY UNLIMITED - hosted by Alan Winstanley 194

Delay-On Timer, 555 Power Supply, PIC Adapter Socket; Shaky Dice, ...
EPE ICEBREAKER - by Mark Stuart 183

Real-time PIC In-Circuit Emulator, programmer, debugger, and development system
SERIES AND FEATURES

NEW TECHNOLOGY UPDATE - by Ian Poole 196
Magnetic memories threaten conventional and Flash devices
TECHNOLOGY TIMELINES - Part 2 Days of Later Yore, plus Fundamental 198

20th Centuary electronics by Clive “Max” Maxfield & Alvin Brown
Who, what, where, and when - the fascinating story of how technology
developed in the past millennium.
TEACH-IN 2000 - Part 5 - Waveforms, Frequency and Time 208

by John Becker
Essential info for the electronics novice, with breadboard experiments and interactive
computer simulations.
NET WORK - THE INTERNET PAGE surfed by Alan Winstanley 221

Radio Bygones Message Board; Search Engines; Disappearing URLs
CIRCUIT SURGERY - by Alan Winstanley 223

Opamp Differentials; Hot Regulator; Conventional Current Flow
PRACTICALLY SPEAKING - by Robert Penfold 228

A novice’s guide to resistors and potentiometers
REGULARS AND SERVICES

EDITORIAL 165
NEWS - Barry Fox highlights technology’s leading edge. Plus everyday 231
news from the world of electronics.
READOUT - John Becker addresses general points arising. 235

SHOPTALK - with David Barrington The essential guide to component 240
buying for EPE Online projects.
Copyright © 2000 Wimborne Publishing Ltd and EPE

EPEOnline,
Online,March
March2000
2000- -www.epemag.com
www.epemag.com- -164
84
BASH, WIND, WIRE
Occasionally a project comes along that makes you realize just how clever some of our modern ICs
are. One such project is the Micro-PICscope to be featured next month (see the Next Month page for
more details). When the editor of the hard copy edition of EPE – Mike Kenward – was an apprentice with
the Ministry of Aviation (now MoD), back in the distant past, he built a Wireless World Oscilloscope – in
those days Wireless World (now Electronics World) actually published constructional projects.
To build the ‘scope they first spent a week “chassis bashing”, and they also had to weld up the sides of
the aluminum chassis (and if you have ever tried welding aluminum you will know that many of them had to
start again when the prized chassis fell into holes under the welding torch!) Then they fitted the
transformers, valve bases, switches, potentiometers, tagstrips, and the oscilloscope tube, inside its mu-
metal screen; after which they could start wiring it all up. Oh, by the way, they also had to wind their own
mains transformer and EHT transformer before we started on the electronics.
By the time the unit was finished it weighed about 10kg and measured around 500mm x 300mm x
250mm, plus the performance was rather limited and it took five minutes to warm up! Now just 30 years on
we can do roughly the same thing with two chips and a liquid crystal display (LCD), put it in your pocket,
power it from a battery and build it for under 20 UK Pounds. Accepted the performance is very limited and
so, too, is the display, but it is a useful little tool for any hobbyist. We expect it to be very popular and there
are a couple of other PIC-based items of test gear in the pipeline.
TERRY
We dedicate this issue to Terry Farmiloe, the Typesetting Manager of the printed edition of EPE, who
died on Jan. 2nd, aged 61. Terry had a gruff exterior that hid a heart of gold. While his name will not be
known to readers, he has been responsible for running the EPE typesetting department, and thus the
production of EPE and other publications, for the last 10 years.
Good luck on your onward journey, Terry, we miss you greatly. Our sympathy goes to your loving
family.
Copyright © 2000 Wimborne Publishing Ltd and EPE

EPEOnline,
Online,March
March2000
2000--www.epemag.com
www.epemag.com--XXX
165
MICRO-PICSCOPE
It’s astonishing what opportunities are continuing to be revealed for the recently introduced
PIC16F87x series of microcontrollers. The Micro-PICscope is a prime example of a design idea
whose implementation was greatly simplified by using one of these devices.
The Micro-PICscope is a handy little item of test gear and of benefit to anyone’s workshop. Us-
ing an alphanumeric liquid crystal display, it is basically a signal tracer, but one with the great ad-
vantage that it shows a representation of the signal waveform being traced. This is shown across
eight of the LCD character cells and is a real-time trace of the monitored waveform.
Not only that, the display also shows the frequency of the signal being monitored, and its peak-
to-peak voltage. The frequency range covered is basically for audio, but frequencies well to either
side of this range can be traced.
Several ranges of control are offered by push-button selection, covering the sampling rate, and
synchronization on/off for the ‘scope display. The signal input is switchable to provide different
maximum peak voltage monitoring ranges. Selection of AC or DC input is provided.
The entire design requires only two ICs, a PIC microcontroller and an opamp, plus a 2-line by
16-character intelligent LCD. Probably the simplest and cheapest ’scope ever.
FLASH SLAVE
Cameras have undoubtedly increased in sophistication over the last ten years or so, with fea-
tures such as auto-focus and built-in flashguns now being commonplace. On the other hand, a few
“standard” features seem to have become rarities that are featured on little more than a few up-
market cameras. The humble flash socket certainly falls into this category.
For most users, this lack of an external flash connector is probably of little consequence, but it
is a major drawback for anyone wishing to go beyond simple “point and shoot” flash photography.
This easy-to-build, inexpensive little unit will fire a secondary flash without any connections to the
camera, thus overcoming the problem.
GARAGE LINK
This circuit helps to prevent the garage door (or either door in the case of a double garage hav-
ing twin doors) being left open all night. It works by establishing a radio link between the garage
and some point inside the house. The unit indoors then provides an audible warning in the form of
a short bleep every 45 seconds. It could also be used to monitor a range of other things around the
home.
PARKING WARNING SYSTEM
by TOM WEBB
How to avoid having an unwanted rear entrance to your garage!
This is a device to aid you in INFRA-RED INFRA-RED BUZZER

parkiar in a garage by providing a ENCODER &
TRANSMITTER
RECEIVER &
DECODER
MONOSTABLE AND/OR
L.E.D.s
visual and audible warning. It is
easily set up by mounting it onto
the wall at the end of the garage.
The device produces a coded
infrared (IR) beam which detects CAR
the proximity of the vehicle by

bouncing IR off it as it Fig.1. Parking Warning System block diagram.
approaches, without being
confused by other IR sources. influenced by stray background removed. An HT12D decoder
When the vehicle is within the IR emission from lights etc. A then decodes the signal to give
preset range, an audible warning coded IR signal is better since a steady output.
is given and a group of light the receiver can be set up to
emitting diodes (LEDs) are turned only accept a specific code.
on. CODED
There are a number of
The block diagram in Fig.1 encoding and decoding ICs TRANSMITTER
shows how the circuit is split up available, but two from Holtek Either the HT12A or HT12B
into separate sections. are used for this circuit. The transmitter devices may be
HT12B transmitter encodes the used in the coded transmission
signal and adds a 38kHz carrier circuit. They work in exactly the
INFRARED CODING signal for greater reliability. A same way except the four data
A system based on a separate demodulating sensor outputs of the HT12A are
continuous IR signal would fail in detects the coded signal and inverted as compared with the
this type of application, since the provides a clean output HT12B. However, since these
receiving circuit would be heavily waveform with the 38kHz carrier outputs are not used in this
circuit, this is of no importance.
Referring to the full circuit
diagram in Fig.2, pins A0 to A7
of the transmitter IC2 set the
coded signal for the IR
transmission, which can only be
accepted by a decoder chip
(IC3 in Fig.2) with the same
settings. The printed circuit
board is designed so that pins
A0 and A1 are connected to the
0V supply line, pins A2 to A7
being left unconnected.
Pin 9 of IC2 is connected to
Fig.1. Parking Warning System block diagram. 0V and pin 18 connected to the
Constructional Project
R1
100Ω
OUT IC4 IN
7805
a
COM D3
R3
10k k
a
D4
+ VE TR1 VR2
R2 BC184L
c
1M k +
47k WD1 S1
b a 12V
IC1 OUT D5
PIC-26043S
e k
GND IC5a IC5b
14 4001 C4 4001 k + 12V D.C.
2 47µ 6 R10 D2
1 18 3 4 330Ω 1N4001 POWER
A0 VDD 5 a
1 INPUT
2 17 R4 (SEE TEXT)
A1 VT 47k 7
3 16
A2 OSC1 0V
4 15
A3 OSC2
TR3
5
A4 IC3 DIN
14
R9 BC184L
HT12D c
6 13 4k7 C5 C6
A5 D11 b
100n 1000µ
C1 7 12
100µ A6 D10
e
8 11 IC5c IC5d
A7 D9
R7 4001 4001
9 10 9 13
VSS D8 100k 10 11
8 N.C. 12 N.C.
a
D1
IR
R5 k
4k7
1 18
A0 VDD R8
2 17 C2 100Ω
A1 DOUT 100p
3 16
4
A2
A3
X1
X2
15
VR1
Fig.2. Complete circuit diagram for
5
6
A4 IC2
HT12B
L/MB
14
13
X1
455kHZ
R6
10M C3
100p
TR2
470Ω
the Parking Warning System. The
A5 D11 TIP122 OR
7
A6 D10
12
TIP121
c
transmitter is the lower section
11
(IC2/TR2) of the circuit.
8 b
A7 D9
9 10
VSS D8
e
positive supply, which should A7. Since pins A0 and A1 on

not exceed +5V. IC2 pins 11 to INFRARED transmitter IC2 are connected to
14 are not used. The pins 0V, the same pins on IC3 are also
labeled X1 and X2 are the RECEIVER connected to 0V.
oscillator control pins, and The IR sensor/amplifier/ Resistor R4 sets the
require a 455kHz ceramic demodulator, IC1, is housed in oscillation frequency for IC3 to
resonator (component X1) along a package resembling a small the required 150kHz. The value is
with resistor R6 and two power transistor. The receiver chosen to suit a power supply of
capacitors, C2 and C3. rejects all IR transmissions between 4××5V and 5V. When IC3
The Dout pin provides a except the required 38kHz receives a correctly coded signal,
coded output superimposed on signal, and provides a clean its pin 17 (VT) goes high. This
a carrier signal of 38kHz which, output (easily viewed on an triggers the monostable formed
with the aid of Darlington oscilloscope). There are three by IC5a and IC5b. When the VT
transistor amplifier TR2, possible receivers that perform pin of IC3 is open circuit (no
operates the IR light emitting the functions required, but in signal being received), resistor
diode D1. tests the best performer for this R7 draws the input pin of the
circuit was the PIC26043S (not monostable to 0V. Once
Potentiometer VR1 allows
a PIC microcontroller!). triggered, the monostable’s
the transmission power to be
varied. Ballast resistor R8 When the detector detects a output remains high for a period
prevents a power supply short signal having a frequency of set by the values of C4 and VR2.
circuit through D1 and TR2 38kHz, its output goes high. The formula used to calculate
when VR1 is set to minimum Transistor TR1 inverts this level this period is T = 0××7 x R x C.
resistance. and supplies it to the decoder With VR2 set to 100kW, the
IC3 at DIN (pin 14).
IC2 pin 10 is connected to period will be 0××7 x 0××1MW x 47uF
ground to hold the transmitter The code to which IC3 = 3××29 seconds.
perpetually triggered. responds is set by its pins A0 to
COMPONENTS from the buzzer which might
otherwise damage the circuit.
Note that on the IR diode,
D1, the long leg is likely to be
Resistors R10 is a ballast resistor to limit the cathode (k), but check this
R1, R8 100 ohms (2 off) the current through the LEDs. with the component supplier’s
R2, R4 47k (2 off) catalog.
R3 10k
R5, R9 4k7 (2 off) POWER SUPPLY Infrared receiver IC1 has a
R6 10M “dome” on its sensitive side,
R7 100k Power to the circuit is which should face outwards
R10 330 ohms intended to be from a 12V mains from the PCB. Once soldered
Potentiometers adapter as the circuit will need to in, IC1 should be bent back to
VR1 470 ohm miniature horizontal
be left switched on for long so that the dome is facing
skeleton preset
VR2 1M miniature horizontal periods of time. A supply of 12V upwards.
skeleton preset is required in order to power the
buzzer. The power supply is Use IC sockets for IC2, IC3
Capacitors and IC5. Do not insert the dual-
C1 100u radial electrolytic, 25V regulated down to 5V by IC4 to
C2, C3 100p (2 off) suit the rest of the circuit. in-line (DIL) ICs until
C4 47u radial electrolytic, 25V construction has been
C5 100n ceramic If a buzzer is not being used, completed and fully checked.
C6 1000u radial electrolytic, 25V then diode D2 can be omitted and
Semiconductors a supply of 5V (or 4××5V) could be
D1 IR diode used by inserting a wire link in the CASING
D2 1N4001 rectifier diode place of regulator IC4 (between
D3 to D5 red LEDs (3mm or 5mm)
Two plastic cases will be
its In and Out pins). However, in needed as the LEDs need their
TR1, TR3 BC184L npn
transistors (2 off) this case, the value of LED own separate case in order to
TR2 TOP122 (or TIP121) npn ballast resistor R10 should be be seen through the rear
Darlington transistor reduced to about 180 ohms. This windscreen of the car.
IC1 PIC26043S IR receiver also means that batteries could
IC2 HT12B (or HT12A) encoder The circuit board is
be used as the standby current is
IC3 HT12D decoder mounted in its own case on
IC4 78L05 +5V 100mA regulator less than 10mA.
small PCB supports which firmly
IC5 4001B quad NOR gate Capacitors C5 and C6 secure it in place, see Fig.4.
Miscellaneous decouple the power fed to IC4. Drill holes in the case to suit the
S1 s.p.s.t. toggle switch (optional) Capacitor C1 and resistor R1
WD1 buzzer, 12V
positions of the IR receiver and
X1 455kHz resonator
smooth out the voltage supplied IR diode, see photographs. The
to the receiver device, IC1. hole for the IR receiver should
Printed circuit board available not be too small otherwise the
from the EPE Online Store, code range will be reduced. If
7000258 (www.epemag.com); CONSTRUCTION maximum range is required
plastic case to suit (2 off, see text);
14-pin DIL socket; 18-pin DIL Apart from the buzzer and then the IR receiver should be
socket (2 off); connectors for LEDs, all the components are positioned right by the hole.
power and LED cables (see text); contained on a single printed
PCB pillars (4 off); connecting
If you prefer to have
circuit board (PCB). The topside
wire, solder, etc. plugged connections for the
component layout and full size
power supply input and for the
See also the underside copper foil master are
output to the LEDs, suitable
SHOP TALK Page! shown in Fig.3. This board is
holes should also be drilled for
available from the EPE Online
their sockets. You also need a
Approx. Cost Store (code 7000258) at
Guidance Only $29 www.epemag.com
hole for the power on/off switch
if you decide to use one,
Begin construction by although one was not used on
The output from the
soldering in the resistors and the the prototype.
monostable, at IC5b pin 4, is
four wire links. Ensure the correct Additionally, two holes are
fed to transistor TR3 via resistor
orientation in the PCB for required to allow adjustment
R9. When the output level is
components C1, C4, C6, TR1 to access to the two preset
high, TR3 is turned on and
TR3, D1 and D2. Capacitors C2, potentiometers, using a small
drives the warning buzzer WD1
C3 and C5 may be connected screwdriver. All holes should be
and turns on LEDs D3 to D5.
either way round. drilled accurately to correspond
Diode D2 prevents back EMF
+
C4
+
C1
+VE GND OUT
R
9
D3 D4 D5
R a k a k a k
e c b e
R 1
VR2 3 c IC1 R
TR3 b 2
D2 TR1 OUT IC4 L.E.D.S CONNECTED
IN SERIES
k a
D5(k) COM
R10 IN
D3(a) C5
+
SCREENED
WD1 C6 R CABLE
R4 7
WD1+
TR2
R5 b
c +
C2 a
e V+
R OPTIONAL
6 X1 D1 VIA S1 JACK PLUG
C3 R
k 8
0V
VR1 0V
Fig.4 (below). Wiring from the circuit

board to the optional LED jack
socket and power connector.
OPTIONAL JACK
SOCKET TO CONNECT
TO L.E.D. BOX
258 POWER
+
HOLE TO
ADJUST
RANGE
OF IR
OPTIONAL
POWER
+ CONNECTOR
BUZZER
+
HOLE TO ADJUST
TIME BUZZER OR INSIDE VIEW
L.E.D.S ARE ON
Fig.5 (above). The LEDs mounted in

a separate case and connected via
screened cable and jack plug to the
Fig.3. Printed circuit board topside compo- Fig.5. the remaining chips, correctly
nent layout and (top right) approximately orientated.
full-size underside copper foil master. TESTING Testing of the IR modules
presents a problem as if one
The first check doesn’t work then the other will
with their respective is to make sure the voltage seem not to be working as well.
components. regulator IC4 is the correct way If in doubt use a voltmeter or
around. Connect the circuit to oscilloscope as follows:
The LEDs mounted in their the 12V power supply and then
separate case can be connected check that 5V is present on the Test the voltage on the VT
to the circuit using single output pin of IC4. If it is, then pin (pin 17) on IC3 of the
screened wire, as shown in disconnect the power and insert receiver module. It should
know if the receiver IC is
working, and hence determine if
the fault lies in the transmitter or
receiver or both.
If the output from IC1 is
working, test the signal at pin 14
(Din) of IC3 on the receiver
module. It should be at about
0V when no signal is received,
rising to about 1××3V (as seen on
a voltmeter) when a signal is
received. Again, an oscilloscope
will show that the signal actually
pulses to about 5V.
If the VT pin on the receiver
is working then simple voltmeter
tests should establish the
position of any other faults.
Completed circuit board mounted inside its box.
If the circuit is triggered
normally be at 0V but change to oscillating, a voltmeter provides straight away then IC1 may be
about 5V when a signal is a rather approximate guide to receiving IR straight from the IR
received. Now check the voltage. If an oscilloscope is diode D1, through stray
voltage on the pins of IC1. Pin 3 available it should be possible reflection inside the case. If this
should be at 5V and pin 2 at 0V. to view the encoded signal, in happens the transmitter should
When a signal is not being which case the trace will rise be surrounded by a rolled piece
received, pin 1 (the output pin) and fall between 4V and 0V. If of black card.
should be at just under 4V. this test fails then try sending a
When a signal is received this signal from a TV remote control
voltage should fall by about 1V. unit. The signal will not be
SETTING UP
decoded, but you will at least The presets VR1 and VR2
Note that as the signal is
can be adjusted to suit the
user’s own particular needs. The
following is a summary of their
functions:
VR1: Adjusts the range of
the IR beam by decreasing or
increasing the power going
through IR diode D1. Reducing
the resistance extends the
range.
VR2: Sets the time the
buzzer and LEDs stay on by
controlling how much
recharging current is input to the
monostable. Reducing the
resistance reduces the time.
COMMON
The transmitter/receiver case and the smaller LED
box with their lids removed. Note the “tube” of black PROBLEMS
card to stop stray reflections from reaching the IR Typical mistakes include dry
receiver chip. joints and bridged pads, i.e.
adjacent pads accidentally
Completed remote LED

warning box.
joined together with solder. This Parking Warning
Other problems include failure System should be set up with
to insert wire links. Also check the IR sensors lining up with the
that the components are extremity of the car, e.g.
correctly placed, and the correct bumper. The LED box should
way round. Note again that be positioned so as to be seen
some IR LEDs are unusual in through the rear windscreen.
that the longer lead denotes The time the LEDs and buzzer
cathode (k). are on, and the range of the IR
can easily be changed using a
screwdriver to adjust the presets
IN USE VR1 and VR2.
HIGH PERFORMANCE
REGENERATIVE
RECEIVER by RAYMOND HAIGH
developed across it.
Provides continuous coverage from 130kHz to
30MHz. Capable of receiving broadcast and amateur Armstrong (and others)
discovered that, by connecting
stations from around the world. a triode valve to the tuned
“regeneration”, the technique circuit and feeding back a tiny
ORIGINS OF portion of the amplified signal to
produces a truly dramatic
REGENERATION increase in receiver sensitivity the coil, its Q can be
Almost a hundred years ago, and selectivity. dramatically increased. By this
scientists and engineers in means, Q factors of several
Armstrong filed his patent in
Europe and America were trying thousand can be achieved
October 1913, just two months
to develop more sensitive circuits before the onset of oscillation,
before his 23rd birthday. At this
for the reception of radio signals. and the wanted signal is greatly
amazingly young age he had
amplified.
C. S. Franklin in England and pushed forward the frontiers of
A. Meissner in Germany were technology and made man’s It is this phenomenon which
both working on similar lines, but dream of long-distance radio imparts such a high degree of
the credit for discovering the reception a reality. sensitivity and selectivity to
benefits of applying positive simple regenerative receivers.
feedback to a tuned circuit is
generally attributed to that great
HOW IT WORKS
Tuned circuits, formed by
POPULARITY
American radio pioneer, E. H.
Armstrong. Known as an inductor (coil) and a Regenerative radio sets
capacitor, are crucial to the were produced in large numbers
working of radio receivers. By throughout the ’twenties. Skill is
varying one of the required to get the best out of
components (usually the radios of this kind: in particular,
capacitor), the circuit can be the regeneration control has to
tuned to resonate at a be carefully adjusted when
particular frequency. receiving weak signals. Largely
because of this, the easily
This combination
operated superhet (also
magnifies a signal to which it
invented by Armstrong) began
is tuned. The degree of
to challenge the popularity of
magnification is dependant on
the regen’ in the ’thirties.
the quality of the tuned circuit,
and this is defined by a figure During the Second World
of merit known as the Q- War, Germany manufactured
factor. A figure of 100 is regenerative sets for military
common. If a signal of 1mV is use, and the British
applied to a tuned circuit with incorporated circuits of this kind
a “Q’’ of 100, a voltage of 100 into clandestine transceivers.
x 1mV, or 0××1V will be Manufacture for domestic
listeners continued almost to the
end of the valve era, with Ever-
S1A
Ready producing a two-valve
9V TO 12V
BC239C
e b c
BC237
BC238
0V
B1
battery-operated set (their
TO VR8
TO VR8
ON/OFF
Model H) during the ’fifties.
OUTPUT
AUDIO
LEADOUTS FOR ALTERNATIVE TRANSISTORS

+
3SK81
BF981
BF960
BF961
C19
1m
AVOIDING
C18
10n
d
PROBLEMS
+
100m
C17
g2
s
Regenerative receivers are
e
c
g1
easily overloaded by powerful
220W
R12
470W
10k
R11
R14
BC239C
TR4
2M2
signals. They are also affected
R13
2N3819
+
d g s
by aerial characteristics.
C16
1m
C15
When an aerial system,
1m
+
which is directly connected to
BC557
BC212
e b c
10n
C14
the tuned circuit, is resonant at
the reception frequency (or a
100n
C13
harmonic), it absorbs energy
d
and inhibits regeneration.
s
100W
R10
10k
Known as “suck-out”, the
R9
BF981
TR3
g2
g1
phenomenon manifests itself as
C8 *
dead spots in the tuning range.
REGEN
VR3
100k
Overload and “suck-out”,
VR4 *
470W
22k
R8
together with an erratic

VR2
10k
+
feedback control, can ruin the
C12
1m
performance of regenerative
+
1M
radios. They are avoided in this
R7
100m
C11
design.
D
100n
C10
10n
C9
WAVE TRAP
VC2*
d
TUNE
FINE
25p
10k
4k7
R5
R4
2N3819
Powerful local radio

TR2
transmitters can swamp

R6 *
C8 *
VC1*
365p
100n
TUNE
C6
regenerative receivers (they

C
even cause problems with

+
C7 *
4 m7
C5
superhets of advanced design).

The answer to this is the
SEE TEXT AND TABLES FOR
inclusion of what is known as a

A
B
SK3/PL3
*
COMPONENTS MARKED
TABLE 2)
“wave trap”.
L2 *
(SEE
An inductor L1 and
100W
capacitor C1 form a parallel

100k
22k
R2
R3
R1
cc
tuned circuit, which presents a

b
high impedance at resonance,

100n
ee
C4
ATTN.
LOG.
TR1
BC557
VR1
R.F.
1k
see Fig.1. When the inductor/

w
SK2/PL2
capacitor combination is set to

EARTH
100n
AERIAL
C3
the frequency of the offending

transmitter it blocks it out.
C1*
+
220p
33p
TO
100m
C2
The problem is invariably

SK1/PL1
encountered on Medium
331208NO
L1*
RWR
Waves, and suitable component

values to tackle this problem,
should it arise, are scheduled in Fig.1. Circuit diagram of the High Performance
Table 1. Regenerative Receiver.
k
TO AUDIO
+9V TO +12V C23
BB405B
INPUT 100n + S1B
FROM
BANDSPREAD VR5 RECEIVER C24
ON/OFF
a
1k
TO
IC1 220m
C19 TDA7052
1 B2
BANDSET VR6 SK1 OUTPUT 6V TO 9V
KV1236 D1 100k VR8 2 5
1/2 OF 4k7
KV1236 LOG.
PHONES
3
a k R15 VOLUME 8 LS1
C22 8W
470k 6
100n
C20
TUNING 1n C21
100n 0V
CAPACITOR
TO TR3 g1 k VR7
D1
1/2 OF
100k
Fig 3. The audio power amplifier stage.
KV1236
a
OV (GND)
Fig.2. Alternative electronic tuning sys- Frequency (kHz) Frequency (kHz)

tem. For fine tuning only, delete VR5 and C1 pF at max inductance at min inductance
(core fully in) (core fully out)
C22, use a BB405B varicap diode, and
33 1300 1700
47 1100 1400
68 900 1200
CIRCUIT DETAILS 120 700 900
The circuit diagram of the 220 550 700
High Performance Regenerative
Receiver is shown in Fig.1. With the use of transistors, better frmoscii.htm, Ed.)
Grounded-base transistor, TR1, results, without recourse to Preset potentiometer VR4 is
acts as a radio frequency (RF) specially designed coils, can be included on the printed circuit
amplifier. Whilst its most achieved by separating them. board (PCB) for use during the
important function is to isolate
Field effect transistor TR2, setting-up process, after which it
the regenerative stage from the
biased by resistor R5 into the is shorted out and replaced by
aerial, it also provides a useful
non-linear region of its fixed resistor R6. Bypass
amount of gain.
characteristic curve, functions as capacitor C8 assists
Signal input is fed to the a sensitive, drain-bend detector. regeneration when the feedback
emitter (e) of TR1, and winding is comparatively small.
Source decoupling at RF and
potentiometer VR1 acts as an It is not required on all coil
audio frequencies (AF) is
attenuator: an essential feature ranges.
provided by capacitors C5 and
that prevents overload on
C6. The output of TR2 is Feedback, or regeneration,
strong signals. Bias is fixed by
developed across drain load is controlled by potentiometer
resistors R2 and R3, and C4 is
resistor R4 and C9, R8 and C14 VR2, which adjusts the voltage
the base (b) bypass capacitor.
remove residual RF. on gate 2 of TR3, thereby
The RF stage is decoupled from
varying its gain. Preset VR3
the supply rail by R1, C2, and
fixes the range of control,
C3. Q-FACTOR capacitor C12 decouples gate 2
The output impedance of a Dual-gate MOSFET TR3 and eliminates potentiometer
grounded-base stage is high provides the modest amount of noise, and resistor R10 and
enough for TR1 to be connected RF gain required for regeneration capacitor C13 decouple the
directly to the tuned circuit, and or Q multiplication. Arranged as a stage from the supply rail.
the use of a pnp device enables Hartley oscillator, feedback from When the tuning coil L2 is
its collector (c) to be taken to TR3 source (s) is connected to a removed for band changing, the
supply negative via the coil L2. tapping on coil L2, via bias signal gates of TR2 and TR3
components resistor R6 and are kept at 0V by resistor R7.
DETECTOR capacitor C8. (Hartley oscillators
were introduced in detail in the
Old valve receivers July 1999 installment of our six- TUNED CIRCUIT
invariably combined the part series on oscillators. For The receiver is tuned by
functions of signal detection and more details bounce over to inductor (coil) L2 and variable
regeneration (or Q www.epemag.com/ capacitors VC1 and VC2. The
multiplication) in a single stage.
COMPONENTS
Resistors Capacitors (continued) Miscellaneous
R1, R9 100 ohms (2 off) C5 4u7 radial electrolytic L1 RWR331208NO inductor (TOKO),
R2 22K C7 axial polystyrene, see text only required if "wave trap" is
R3 100k and Table 2 (next month) needed (see text)
R4, R10, R12 10k (3 off) C8 ceramic, see Table 2 (next month) L2 tuning band coils (TOKO), (8 off)
R5 47k C9, C14, C18 10n disc ceramic (3 off) see text and Table 2 (next month)
R6 various values (see text and C12, C15, C16, C19 1u radial PL1 to PL8 9-pin D-type plugs for L2
Table 2 next month) electrolytic (4 off) (8 off) see Table 2 for other components
R7 1M *C20 1n (1000p) or 50p polystyrene S1 d.p.d.t. toggle switch
R8, R14 470 ohms (2 off) (see Fig.2) SK1, SK2 screw terminal post
R11 220 ohms C24 220uF radial electrolytic (Aerial and Earth)
R13 2M2 VC1 365p Jackson O-type air-spaced SK3 9-pin D-type socket (for plug-in
*R15 470k tuning capacitor (see text) tuning coils)
All 0.25W 5% carbon film VC2 25p Jackson C804-type air-spaced SK4 switched stereo jack socket
tuning capacitor (see text) B1 9V to 12V battery pack
Potentiometers B2 6V to 9V battery pack
All capacitors 12V working or greater
VR1 1k rotary carbon (logarithmic
law if obtainable) Printed circuit boards available from the
Semiconductors
VR2 10k rotary carbon, linear *D1 KV1236, KV1235, or BB405B EPE Online Store, codes 7000254
VR3, *VR7 100k enclosed horizontal (receiver), 7000255 (Electronic Tuning),
varicap diode (see text)
preset (2 off) and 7000256 (Amplifier); 9-pin D-type
TR1 BC557 pnp silicon transistor
VR4 22k enclosed horizontal preset plugs (8 off for tuning coils); aluminum or
TR2 2N3819 n-channel field effect
*VR5 1k rotary carbon, linear diecast box; 8-pin DIL socket; plastic
transistor
*VR6 100K rotary carbon, linear control knobs (4 small, 1 large); reduction
TR3 BF981 n-channel dual-gate
VR8 4k7 rotary carbon, logarithmic drive for tuning capacitor; multistrand
MOSFET
Capacitors connecting wire; card for tuning dial; nuts,
TR4 BC239C npn silicon transistor
bolts, washers, and stand-offs; solder
C1 axial polystyrene, See Table 1 IC1 TDA7052 low voltage 1W
pins, solder, etc.
C2, C11, C17 100 uF radial power amplifier
electrolytic (3 off) Note: All components marked with an
C3, C4, C6, C10, C13, *C21, *C22, C23 See also the asterisk (*) are for the optional electronic
100nF disc ceramic (8 off) SHOP TALK Page! tuning system.
Approx. Cost
Guidance Only
(Excluding batteries and tuning capacitors) $56
larger of the two capacitors, bias of audio amplifier, TR4, are
VC1, acts as a coarse (Bandset) fixed by resistors R13 and R14. ELECTRONIC
tuning control. The smaller one, Signal output is developed
VC2, provides fine across collector (c) load resistor TUNING
(Bandspread) tuning. These R12; and R11 and C17 The use of a separate Q-
components are discussed later. decouple the stage from the multiplier stage (TR3) makes
Fixed capacitor C7 limits the supply. the receiver tolerant of
maximum value of VC1 on the The low value of emitter electronic tuning. (The
shortwave ranges. The reduced bypass capacitor C16 results in somewhat modest Q of high
swing makes tuning less critical gain-reducing negative capacitance varicap diodes
and consistent regeneration feedback at the lower audio inhibits the operation of most
easier to achieve. frequencies. This improves regenerative sets).
Details of the coverage clarity. Coupling and DC A suitable, add-on,
obtained with a range of Toko blocking capacitors C15 and electronic tuning circuit is given
coils, together with the C19 have a low value for the in Fig.2. Potentiometer VR6
associated values of C7, R6, same reason. controls the reverse bias on
and C8, are given in Table 2 Response to the higher varicap diode D1 and varies its
(next month). audio frequencies is curtailed by junction capacitance. This forms
capacitor C18. Constructors the coarse, or Bandset, tuning
who find the tone too “bright” control.
AUDIO AMPLIFIER
should increase the value of this Potentiometer VR5 permits
The base (b) and emitter (e) component to 47nF or 100nF. a small adjustment of the bias
voltage, and acts as the fine, or operated Bandspread control in are mains powered).
Bandspread, control. Preset a position remote from the The power amplifier current
VR7 fixes the lowest level the tuned circuit. The prototype swings between 6mA and 60mA
bias voltage can fall to, thereby Receiver, shown in the or more when it is being driven
determining the maximum photographs, incorporates this hard. The resulting supply
value of the tuning capacitance. arrangement. voltage fluctuations would
(Diode junction capacitance disturb the operation of the Q-
increases as the reverse bias is multiplier, despite heavy
reduced.)
POWER AMPLIFIER
decoupling.
The varicap diode D1 is The circuit diagram of the
additional, single chip, audio Separate battery supplies
coupled into the main circuit via for the Receiver and Power
DC blocking capacitor C20 and power amplifier stage is given in
Fig.3. This amplifier has its own Amplifier sections are,
resistor R15 isolates the signal therefore, strongly
path from the potentiometer 6V to 9V power supply to avoid
any possible interaction with the recommended. They are
chain. Potentiometer noise is essential when electronic tuning
prevented by capacitors C21 receiver section. Designed
around a TDA7052 low voltage is adopted. A double-pole toggle
and C22. switch, S1a and S1b, connects
power amp IC, the only external
High value varicap diodes components are capacitors C23 the two separate battery packs
have a relatively large minimum and C24 which ensure the into circuit.
capacitance, and an additional stability of the device.
coil may be needed in order to
secure continuous coverage.
Potentiometer VR8 acts as the COMPONENTS
volume, or AF gain, control.
Furthermore, performance Before we commence
above 20MHz or so is not quite The power amplifier IC1 is construction, a few words now
as satisfactory as that afforded short-circuit protected, requires on choice of components may
by a traditional variable no heatsink and can deliver a help. Readers are also directed
capacitor. clean 1W of audio into an 8 to our Shoptalk page for details
ohm speaker with a 6V supply. of possible suppliers for some
These disadvantages do not It is also claimed that there are
apply when the electronic tuning off those “hard to find” items.
no switch-on or switch-off clicks
circuit is used with a VHF diode with this device.
solely to provide fine tuning
(VR5 is omitted and the top end
of VR6 is connected directly to POWER
the positive supply rail). This SUPPLIES
arrangement has the advantage
of low cost and conveys a Current drain is
freedom to locate the DC extremely modest, being
only 2mA for the radio
section and 50mA for
the power amplifier when
it is delivering a good
speaker volume (5mA
when ’phones are used).
Battery supplies are,
therefore, eminently
suitable, and any
possibility of hum and
interference from the The chassis of the prototype was fab-
mains is avoided ricated from aluminum and a wooden
(regenerative receivers case with hinged lid holding the loud-
are very susceptible to speaker made to house the receiver.
this and require a The lid can be raised and held up by a
carefully designed
hinged wire frame (shown above)
supply unit when they
when in use.
Coils gate MOSFETS, including the rate which is too fast, connect a
40673 and the MFE201. The 10pF or 5pF polystyrene
All of the inductors used in 3N201 was not tried, but it capacitor in series with it to
this Receiver are from the Toko should prove satisfactory. reduce its swing.
range. Their frequency
coverage is shown in Table 1 A 2N2905 pnp transistor Inexpensive, polythene
and Table 2 (next month) worked well in the RF stage, dielectric variables, of the kind
together with suitable tuning and a 2N5827 or a 2N5828 used in transistor portables, can
capacitor values. should be suitable for TR4. also be used. Some of these
The alternative devices have comparatively low values,
Coils can also be hand and both sections may need
wound. As a very rough guide, mentioned here have different
case styles to those depicted in connecting in parallel to obtain
when 20mm to 25mm diameter the required tuning range. (A
formers are used, feedback Fig.1, and the lead-outs must be
checked. swing of at least a 10pF to
tappings should be about 10 200pF is needed to give
turns up from the “earthy” end continuous coverage from
on Long waves, 5 turns on
Tuning Capacitors
150kHz to 30MHz with the coils
Medium waves, and 2 or 3 turns A Jackson 365pF O-type listed in Table 2). The 25pF FM
on Shortwaves. air-spaced tuning capacitor is tuning section of one of these
the preferred component for capacitors can act as the
Transistors bandset control VC1, and a bandspread control VC2.
Transistor types are not 25pF Jackson C804 type is
ideal for VC2, the Bandspread If salvaged tuning
critical. The Q-multiplier circuit capacitors are used, make sure
works well with a range of dual- control. If this latter value
produces a bandspread tuning that they are clean and dry, that
the rotor contacts are
satisfactory, and that the vanes
are not shorting.
Varicap diodes are retailed
by a number of suppliers and
should not be too hard to find.
Any 450pF varicap designed for
9V bias, should be suitable for
full electronic tuning.
NEXT MONTH
In Part 2 next month we’ll
go over the constructional
details for this project.
AUTOMATIC TRAIN SIGNAL
by ROBERT PENFOLD
An easy-to-build, low cost starter project for your wave rectifier circuit. The
model radio system. voltage on the track is a DC
signal, but its polarity depends
This very simple project, this is of no practical on the direction of the train.
suitable for beginners, is a two- importance, as the state of the
To operate the main circuit
color (red/green) reliably it is important
signal for a model that the input signal
railway. It uses a has the correct
TRESHOLD
simple form of GREEN
polarity, and the
automatic operation, FULL-WAVE LOWPASS VOLTAGE purpose of the rectifier
RECTIFIER FILTER DETECTOR RED
and if you stop the is to ensure that the
train in front of the main circuit is fed with
signal it automatically a positive signal
switches from “green” regardless of the
to “red”. When the Fig.1. Block diagram for the Automatic Train Signal. train’s direction. The
train is restarted the output of the rectifier
signal automatically is fed to a potentiometer that
switches to “green” again. signal is irrelevant except when
the train is approaching it. enables the output voltage to be
To an onlooker it appears as reduced. This enables the user
though the signal is changing to adjust the threshold voltage
color and the train is responding SYSTEM OPERATION at which the signal changes
to the change. In reality the train state.
The block diagram of Fig.1
and the signal are both The threshold level used is
helps to explain the way in
responding to changes in the not critical, but the signal should
which the Automatic Train
track voltage. The signal will, in not go to red while the train is
Signal functions. The voltage
fact, go to “red” wherever the still moving. On the other hand,
from the track is fed to a full-
train is stopped on the layout, but some types of train controller
S1
ON/OFF
D5 a
GREEN
L.E.D. k
C2 R3
100n 39k
R5
SK1 1k5
7
3
D1 TO D4 B1
ALL 1N4004 D4 D1 R1 R2 IC1 6
9V
10k 2k2 LF351N
2
(6 x AA)
4
D2 D3
R6
1k5
SK2 VR1 R4
C1
22k 22µ 10k
D6 a
THRESHOLD RED
L.E.D. k
Fig.2. Complete circuit diagram for the Automatic Train Signal

circuit activates the red signal
LED, but with higher input COMPONENTS
potentials it switches on the Resistors
green LED instead. R1, R4 10k (2 off)
CIRCUIT OPERATION R2 2k2
R3 39k
The full circuit diagram for R5, R6 1k5 (2 off)
the Automatic Train Signal is All 0.25W 5% carbon film
shown in Fig.2. The voltage Potentiometer
from the rail tracks is connected VR1 22k rotary carbon, linear
to sockets SK1 and SK2, which Capacitors
feed into a full-wave bridge C1 22u radial electrolytic, 25V
rectifier (D1 to D4). The positive C2 100n ceramic
DC output signal from the Semiconductors
D1 to D4 1N4004 rectifier
rectifier circuit is fed to a diodes (4 off)
volume control style variable D5 green LED, 3mm or 5mm
attenuator (VR1) and then to a diameter (see text)
simple lowpass filter comprised D6 red LED, 3mm or 5mm
of resistor R1 and capacitor C1. diameter (see text)
IC1 LF351N opamp
The cut-off frequency of this
filter is low enough to ensure Miscellaneous
never produce an output level B1 9V battery pack (6 x AA cells
that there are no problems with in holder)
that is right down at zero volts, flickering of the signal lights S1 s.p.s.t. miniature toggle switch
and the threshold level must be when the track voltage is near SK1, SK2 4mm socket (2 off)
high enough to ensure that the the threshold level. On the other
signal does go to red when the hand, it is not so low that the Medium size plastic case (see text);
train stops. 0.1 inch pitch stripboard, size 20
unit is slow responding to holes x 20 strips; 8-pin DIL socket;
It cannot be safely assumed changes in track voltage. control knob; PP3 battery clip;
that the signal across the tracks An operational amplifier, multistrand connecting wire; single-
is a steady DC potential. The sided solder pins, solder, etc.
IC1, is used here as a voltage
motor in the train is likely to comparator. Resistors R3 and See also the
introduce large amounts of R4 form a potential divider that SHOP TALK Page!
noise onto the track voltage, biases the non-inverting input of
which might not be a simple DC IC1 (pin 3) to about 1××8V. The Approx. Cost
signal anyway. Many train
controllers use some form of
output of IC1 at pin 6 will go Guidance Only $11
high if the inverting input (pin 2) (Excluding Batteries)
pulsed output signal, where the
is taken below this potential, or
motor is controlled by varying to their original states when the
low if it is taken above the
the average output signal. train is stopped again, with the
reference level.
Others use the rectified but non- red LED switched on.
smoothed output from a mains The voltage fed to the
transformer. inverting input will be very low
In order to avoid problems
with the train stationary, sending ON TRACK
the output of IC1 high. As a The current consumption of
with noise on the input signal,
result red LED D6 is switched the circuit is about 7mA. A PP3
and to accommodate pulsed
on, but green LED D5 is size battery is just about
controllers, the output from the
switched off. adequate to supply this, but a
threshold control is fed to a
lowpass filter. This provides a When the train is started, battery pack consisting of six AA
reasonably smooth DC output the voltage fed to the inverting size cells in a holder will provide
signal at a potential that is equal input rises, and eventually cheaper running costs.
to the average input voltage. becomes greater than the Operation from a mains
reference level at the non- power supply unit is made slightly
Finally, this signal is applied
inverting input. The output of awkward by the fact that neither
to a simple voltage detector
IC1 then switches to the low supply rail can be earthed. This is
circuit. With an input voltage of
state, switching off D6 and because one of the input lines
up to about 1××8V the detector
switching on D5. Things revert
FLAT
D5
k a
Fig.3. Stripboard component layout, inter-
2 4 6 8 10 12 14 16 18 20
1 3 5 7 9 11 17 19
wiring to off-board components and details
A
B of breaks required in the underside
C
D ON/OFF copper strips.
E
F
SK1
G S1
H
R 20 18 16 14 12 10 8 6 4 2
I 3 C
R 19 17 15 13 11 9 7 5 3 1
J k 2
5
K A
D1 k
L a B
D3 R C
2
a D
N
R E
O k 6 RED
+ R TO F
R 4 B1
Q k D4 BLACK G
1 C1
a H
R D2
S a I
T J
K
L
SK2 k a M
D6 N
FLAT O
VR1 P
Q
R
S
T
THRESHOLD
might be earthed, and neither of mounting bolts. There are just six of them might be too long to
these lines connects to a supply breaks in the copper strips. These permit this. They will then have
rail of the signal circuit. can be made using a special tool to be made from 22s.w.g. or
Earthing one rail of the or by using a small hand-held 24s.w.g. tinned copper wire. Fit
signal circuit could produce an twist drill bit of about 5mm single-sided solder pins at the
unwanted connection that would diameter. points where connections will be
prevent the unit from working, The board is now ready for made to the controls, LEDs, and
and could result in a heavy the components and the three sockets.
current flowing through the input link-wires to be added. It is
rectifier circuit. The most generally considered best to start CASING UP
practical solution is to use a 9V with the small components and
or 12V regulated battery work up to the largest, but in this If the unit is powered from a
eliminator. These use double case the components are all quite PP3 size battery it should be
insulation and have neither small. possible to fit it into practically
supply rail earthed. any small plastic box. A
It is probably best to work medium size case about
across the board methodically, 150mm or so long will have to
CONSTRUCTION being careful to get everything in be used if an AA battery pack is
the right place. In the cases of to be accommodated.
The Automatic Train Signal IC1, C1, and the four rectifier
circuit is built up on a piece of diodes (D1-D4) you must also be Threshold control VR1 and
stripboard containing 20 holes careful to fit them the right way On/Off switch S2 are mounted
by 20 copper tracks. The round. The LF351N used for
component layout, together with IC1 is not a static sensitive
details of breaks required in the component, but as with any
copper strips, is shown in Fig.3. DIL integrated circuit it is still
Construction follows along advisable to mount it on the
the normal lines with a standard board via a holder.
size board being cut down to the It might be possible to
correct size using a hacksaw. make the link-wires using
Next drill the two mounting the wire trimmed from the
holes, which have a diameter of resistor leads, but one or two
3mm and accept Metric M2××5
Layout of components inside the small plastic box. Note, the

input sockets must be fed with the track voltage.
on the front panel, while input the model railway, as must the Alternatively, by simply
sockets SK1 and SK2 are size of the two LEDs. For the leaving some bare wire at the
mounted on one side or at the usual smaller gauges 3mm ends of the leads it might be
rear of the case, see diameter LEDs are the best possible to make connections to
photographs. An exit hole for choice, but for larger gauges the screw or spring connectors
the lead to the signal LEDs is 5mm types would be better. The on the train controller, being
required in one side or the rear LED current is not very high, so careful to leave the connections
of the unit. The circuit board is “high brightness” types are to the track intact. Failing that, it
mounted in any convenient preferable. will be necessary to make up a
space, and it is advisable to use Unlike filament bulbs, LEDs dual supply lead to enable both
some extra nuts or short will only work if they are the signal and the track to be
spacers between the board and connected with the correct fed from the controller.
the case. This avoids any polarity. Having the cathode (k)
tendency for the board to buckle With Threshold control VR1
lead slightly shorter than the at a roughly middle setting the
and break when the mounting anode (a) lead is the normal
nuts are tightened. To complete signal should work quite well,
way in which the polarity of a with “red” and “green” signals
the main unit the hard wiring is LED is indicated. There may
added. This is all shown in Fig.3 being obtained when the train is
also be a “flat” on the cathode respectively stopped and
and is perfectly straightforward. side of the encapsulation. running fast. The signal will
probably be at “red” when the
SIGNAL BOX TESTING train moves very slowly, or
Construction of the “signal” possibly at “green” when the
Input sockets SK1 and SK2 train has stopped. A little
is left to the ingenuity of must be fed with the track
individual constructors. At its experimentation with various
voltage, and the way in which settings for VR1 should soon
most basic the signal can just this is done must be varied to
consist of a very small plastic get the signal switching states at
suit the equipment with which the start/stop track voltage of
box for the two LEDs. However, the signal is used. In most
it should not be too difficult to the train.
cases the easiest way is to
fabricate a more convincing make up a twin lead fitted with Note that the more simple
signal from balsa wood, bits of 4mm plugs to connect to SK1 types of controller tend to
dowel, etc. Provided you know and SK2, and small, insulated provide a start voltage that is
what you are doing, it would covered, crocodile clips at the much higher than the one at
probably be possible to adapt a other end. The power which the train stops. The
ready-made signal “tower” to connectors on the track are threshold voltage of the signal
work with this circuit. often quite crude, and will then has to be set at a
The size of the signal must permit power to be tapped off compromise level somewhere
be varied to suit the gauge of using the crocodile clips. between these two voltages.
ROLL-UP, ROLL-UP! Win a Pico PC-Based Oscilloscope
Ingenuity is our regular round-up of readers' own • 50MSPS Dual Channel Storage
circuits. We pay between $16 and $80 for all material Oscilloscope
published, depending on length and technical merit. • 25MHz Spectrum Analyzer
We're looking for novel applications and circuit tips, not
simply mechanical or electrical ideas. Ideas must be the • Multimeter
reader's own work and must not have been submitted • Frequency Meter
for publication elsewhere. The circuits shown have • Signal Generator
NOT been proven by us. Ingenuity Unlimited is open to
ALL abilities, but items for consideration in this column If you have a novel circuit idea which
should preferably be typed or word-processed, with a would be of use to other readers, then a Pico
brief circuit description (between 100 and 500 words Technology PC based oscilloscope could be
maximum) and full circuit diagram showing all relevant yours.
component values. Please draw all circuit schematics Every six months, Pico Technology will be
as clearly as possible. awarding an ADC200-50 digital storage oscil-
Send your circuit ideas to: Alan Winstanley, Ingenuity loscope for the best IU submission. In addition,
Unlimited, Wimborne Publishing Ltd., Allen House, East two single channel ADC-40s will be presented
Borough, Wimborne, Dorset BH21 1PF. They could earn to the runners up.
you some real cash and a prize!
Zener Diode Tester - A tap on the knee

C1
1000µ
The Purpose of the circuit dia- 35V

+
42V
gram illustrated in Fig.1 is to help FS1

500mA D1
1N4004
D2
1N4004
D3
1N4004
TR1
2N1711
T1
in measuring the value of an a k a k a k c e IN IC1 OUT
230V LM317T
S1
unidentified Zener diode. It is cen- 10V

b
ADJ
C2 + C3 + C5 +
tered around a common LM317 L
N
1000µ
25V
1000µ
50V
R1
680Ω
0.25W
10µ
50V
R2
270Ω
0.25W
voltage regulator (IC1), which is 0V

VR1
22k
connected in its constant-current 0V
4.7mA
mode. The current is set to under E

D4
18V
D5
18V D10 a
RED
5mA, being calculated by the for- k
0.5W
a k
0.5W
a L.E.D. k
mula I = 1××25V/R, and this flows D6

a
1N4148
+
through the Zener diode being D7
k
a VZ
ZENER
UNDER
k
TEST
tested. A voltmeter can be con- trebler circuit, based around C4 +
10µ
50V
1N4148
k
a
– a
nected in parallel with the Zener diodes D1 and D2 and capacitors D8

1N4148
k
EXTERNAL
and the Zener voltage read di- C1 and C2, it is easy to use a D9
1N4148
a VOLTMETER
rectly. By adjusting control poten- much lower secondary voltage

k
tiometer VR1, it is possible to vary (e.g. 10V AC) and raise this to a
the voltage on transistor TR1 useful 42V DC (10V AC x 1××414 x
emitter (e), which provides an ad- 3) which is current limited to about Fig.1. Zener Diode Tester.
justable voltage to the current lim- 35mA. Due care must be taken
iter. VR1 is then adjusted until the when working with these higher for AC Monitoring - PIC
LED D10 illuminates to indicate voltages, and extreme care the value
that current is flowing, when the should be exercised to ensure
test voltage may be read. The cir- The circuit of Fig.2 was de-
that all diodes and electrolytic ca-
cuit operates from a mains power signed in order to interface a
pacitors are correctly polarized.
supply. To measure diodes up to, PIC microcontroller to a stan-
say, 33V DC, it is normally neces- Gianfranco De Dominici dard clip-on DVM (digital volt-
sary to use a transformer of 30V Greenock, Scotland meter) current probe. The PIC
AC output, but by using a voltage Microcontroller Interface device chosen must have an
Ingenuity Unlimited
analogue input and the software to heatsink the MOSFET because VCO.
depends on the application used. of its significant positive tempera- The final stage buffers the
The circuit utilizes low-noise ture co-efficient. triangle wave produced and sets it
LMC6084 operational amplifiers to 6V (24V pk-pk) and a non-
(IC1a and IC1b) as a precision Gerard la Rooy
New Zealand. linear impedance formed by
rectifier and a filter (IC1c) with diodes D18 to D27 and neighbor-
pre-settable gain to give a re- ing resistors, R43 to R47, is used
sponse to the average current. AF Sweep Signal Generator to generate an approximation to a
The interface circuit requires a
low impedance current probe pro-
- More Scope sine wave. The output is also set
to 6V RMS (about 17V pk-pk) us-
viding 1mV per ampere to be con- The purpose of the audio fre- ing VR8. It is best to use a sine
nected to the input. The accuracy quency (AF) Sweep Generator cir- wave for frequency response
of the circuit is about as good as cuit in Fig.4 (shown in stages) is to curves, as square or triangular
the 8-bit ADC used within the mi- provide an audio test signal in the waves contain odd harmonics.
crocontroller when one per cent range of 20Hz to 20kHz. This can
resistors are employed. be set either to a single frequency Power Supply
or a sweep range of up to ten oc- A mains power supply pro-
Gerard la Rooy taves, which can be used to check
New Zealand. vides +15V and -15V, together
the frequency response of an audio
with a 5V reference. For the
amplifier etc.
opamps, LF353s are recom-
Reverse Polarity The design provides Square mended, particularly where the
Indicator-Connection Pro- wave, Triangular wave, and an ap- opamp is used as a comparator
tection - Connection Pro- proximate Sine wave. The output as it has a good slew rate. This is
level is 0V to 6V RMS. also required to produce a rea-
tection sonable triangular wave at 20kHz.
In order to provide reverse On the Ramp
An LF353 gives a fairly sym-
polarity protection for some 12V The first stage of the generator metrical positive and negative out-
and 24V DC equipment, the cir- gives a continuous 0V to 11V to 0V put, although it was found neces-
cuit diagram of Fig.3 was devised. ramp, which is used to sweep the sary to set a voltage differential
This uses a MOSFET power tran- audio frequency output, and may using diodes on the positive and
sistor, TR2, to virtually eliminate also be used as a timebase for an negative supply rails. The result-
the problem of undesirable for- oscilloscope. This should be set to ing 13××6V and 14××2V gives a fairly
ward voltage drop, which would two seconds per cycle using VR1,
good ±12V output swing. On the
be apparent if ordinary rectifiers as the 20Hz to 40Hz octave needs
a certain amount of time to com- prototype, the square, triangular
were used for protection, and the
plete a few cycles. This stage also and sine wave outputs were con-
circuit even outperformed a
provides fixed triangle and square nected to a switch and a 5 kilohm
Schottky diode in this respect.
wave outputs. potentiometer, and then on to
The voltage drop is determined by
phono sockets. The ramp and
the on-resistance of TR2 and the
Sine of the Times square wave output signals were
current flowing into the load.
also brought out to front panel
Since the MOSFET is used in its The second stage produces a
sockets for use as a timebase sig-
reverse-conducting mode it is fea- logarithmic curve so that ten divi-
nal. The ramp can be used as
sible to expect an effective on- sions across an oscilloscope tube
channel B input for X/Y plotting, or
resistance of about 20 per cent correspond to ten octaves: 20Hz,
use a slow timebase (0××1sec/div.),
below the manufacturer's specifi- 40Hz, 80Hz, 160Hz, 320Hz, 640Hz,
cations, which are usually given in 1,280Hz, 2,560Hz, 5,120Hz, and trigger the oscilloscope on the
forward conducting mode. Two 10,240Hz and 20,480Hz. Its two falling or rising edge of the square
light-emitting diodes (LEDs D2 outputs labeled +Vc and -Vc con- wave.
and D3) are incorporated to indi- nect to the third stage, which is a J.D. Gray
cate correct or reversed polarities. voltage-controlled oscillator (VCO). London N10
The circuit can be scaled up The two signal voltages required
by paralleling MOSFETs or by are +7V and -7V for 20kHz, 3××5V
placing a Schottky diode in paral- for 10kHz, 1××75V for 5kHz etc.,
lel with the transistor, to get the down to about 170mV for 20Hz. A
best of both worlds. It is advisable square wave is also provided by the
EPE ICEBREAKER by MARK STUART
A real-time PIC In-Circuit Emulator (ICE), programmer, ging (ICD). This requires a chip
debugger, and development system with special built in hardware
(known as a “Background De-
This project combines Mi- system is called an In-Circuit bugger”) and software which
crochip’s MPLAB development Emulator or “ICE” and profes- can communicate its status via
software with the advanced self- sional systems are available for a serial link.
debugging features of the latest practically every type of micro- The chip is fitted to its work-
PIC chips. The result is a user- controller. ing printed circuit board (PCB)
friendly advanced development The problem with this is and all external hardware is
system for a very low cost. cost. A professional ICE for the connected and active. Code is
PIC series of chips costs a rea- then programmed, run, and de-
sonable 2000 UK Pounds or so bugged under PC control, until it
DEVELOPMENT – not a lot if you are a profes- is running correctly. For Mi-
SYSTEMS sional programmer being paid crochip PIC users, the good
twice that each month – but news is that the PIC16F877 and
Developing projects with mi-
more certainly enough to make its close relatives the ’876, ’874,
crocontrollers is extremely inter-
your eyes water if you are an and ’873 have built in ICD facili-
esting, especially the ability to
amateur! ties and can be used to develop
alter the way hardware responds
just by making simple changes to Just lately a new type of de-
program code. velopment system has ap-
peared called In-Circuit Debug-
The problem is that each
change of code has to be written,
compiled (converted into suitable
+5V
R14
form for loading) and pro- 14
10k
grammed into a chip before it can

OSC 2
33
X1 RB0
be tested. Several low cost sys- 20MHz 13
OSC 1
RB1
34
35
PL1
tems – such as the EPE PICtutor 2

RB2
RB3
36
R12 1
are available (see our web page C2 C1
3
RA0
RA1
RB4
37 10k 6
38
at www.epemag.com for more 10p 10p 4
5
RA2
RB5
39 D7
k
R13
details) and are adequate for the

RB6 5V1 470!
RA3 40 a
6 RB7 9
RA4 IC1
development of simple programs, k 5
7
RA5 PIC16F877 D8 R11
but for larger changes and pro- 11

VDD(+V) MCLR
1
5V1
a
1k
grams it can be tedious, and er- 32

12
VDD(+V)
RE0
8 D6
k
5V1
rors are almost as easy to intro- 31
VSS(GND)
VSS(GND) RE1
9
10
a
duce as to eliminate. RE2

LK1 LK2
15 19
Better methods of testing pro- 16
RC0
RC1
RD0
RD1
20
grams rely on more advanced 17

RC2 RD2
21
18 22
software that can “run” in a virtual 23
RC3 RD3
27
chip on a PC screen and ad-
RC4 RD4
24 28
RC5 RD5
vanced hardware which commu- 25
26
RC6 RD6
29
30
nicates with the program in the RC7 RD7
PC, reads input pins, and 0V (GND)
switches output pins to match the

levels of the virtual chip. Such a Fig.1. Minimum circuit for using the PIC16F877 as a
“background” debugger.
code which can be run in these computers read serial data cor- working code has been devel-
and smaller chips in the range – rectly when 0V to 5V swings are oped and debugged it can be
such as the most popular used. A third connection links loaded and will run alone in any
PIC16F84. the serial port RTS output to the chip in the 16x range – the pro-
VPP or MCLR pin of the chip. tected communication code is re-
PIC IN-CIRCUIT This allows control of the pro- quired only in the chip used for
DEBUGGING gramming voltage debugging.
This article is intended as an (programming at 5V, as op- Connections for suitable se-
easy introduction to ICD with very posed to the 12V normally re- rial leads are shown in Table 1.
simple demonstration programs, quired) and resetting of the chip These are standard connections,
users can then progress to using by the computer. but can be made up with 4-way
the more complicated features of After a Reset, the chip cable (flat telephone cable is
the chips. It is not intended as a checks if pins RB6 and RB7 are ideal) if required. Take care when
programming tutorial, but the op- shorted to 0V. If they are, it ig- making leads to get the pin num-
eration of some programs is de- nores the ICD functions and just bering correct – it is very confus-
scribed in the course of demon- runs the code directly starting ing, the only safe way is to read
strating the ICD hardware. from location 0. Links fitted in the molded numbers on the con-
Simple programs can be the positions marked LK enforce nectors.
loaded, run and debugged without this option. The port connections for the
knowing much about the entire As well as the hardware PIC 16F877/874 and the alterna-
ICD system which is extremely connections, the computer tive connections for the 28-pin
complicated and occupies many needs to run a program to com- PIC16F876/873 versions of the
pages of the PIC data sheets. municate with the chip, and the chip are shown in Fig.2.
The Microchip web site chip must have a program to
(www.microchip.com) provides communicate with the com-
an enormous amount of informa- puter. The program in the chip HARDWARE
tion for those wishing to know is loaded and copy-protected Whilst the minimum system
more. into the upper half of the chip’s could be used, it is unlikely that
8K program memory. any PIC application would oper-
It may seem wasteful to use ate without external hardware.
MINIMUM ICD SET UP Most systems have power sup-
half of the chip’s memory for
The minimum hardware re- protected code, but in practice, plies, input switches, output de-
quired for ICD, using the the 4K remaining is a vast vices and so on. It is irritating and
PIC16F877, is shown in Fig.1. amount of space for the PIC time consuming to have to set up
Communication to a computer program and it is very doubtful these simple hardware require-
serial port is achieved via the that it will ever be filled. Once ments when the object is to get a
Port RB6 and RB7 pins of the
chip, which cannot be used for
other functions. As there are
plenty of other port pins available
this is not a significant limitation.
Port RB6 receives data from
the computer, and RB7 transmits
data to the computer. Both of
these pins operate at simple 0V
to 5V logic levels. Some com-
puter serial port output pins swing
10V positive and negative, and so
limiting resistors and 5××1V Zener
diodes are used for protection.
The serial data sent back to
the computer should also be ca-
pable of swinging 10V, but it has
been found that practically all
Table 1: Serial Lead the PIC could be fitted into an- tion is not recommended for
Connections. other board with other hardware ICD use. Other crystal frequen-
and still debugged via the same cies can be used and the com-
9-way to 9-way
computer lead. puter serial port speed altered
ICE Computer
Voltage regulator IC2 allows accordingly. 20MHz gives the
1 7 fastest communication (38,400
2 3 TxD
a range of power adapters to be
used connected to 2××1mm power BAUD) and is best if there are
3 2 RxD
5 5 GND socket SK5. Links 3 and 4 allow no other special frequency re-
positive inner or outer connec- quirements.
9-way to 25-way tions to be set. If accidental Stepping motor driving is a
ICE Computer power reversal is possible, the very popular PIC application.
1 4 positive link connection can be Transistors TR1 to TR4 and as-
2 2 TxD made using a 1A diode (e.g. sociated resistors R2 to R5 and
3 3 RxD 1N4001) instead of a piece of protection diodes D1 to D4 pro-
5 7 wire. Power is indicated by light- vide four open collector drivers
emitting diode (LED) D5 via resis- for four-phase unipolar motors.
program written and tested.
tor R8. Connectors PL3 and PL4 allow
EPE ICEbreaker was designed
Switch S1 provides an alter- for 2××54mm and 2mm pitch mo-
to include a number of input and
native hardware reset which can tor connectors. Input to the
output devices along with a sol-
be useful for stopping programs drivers is via SK4. The transis-
derless connection system so
quickly and for restarting from tors can also be used individu-
that many applications could be
location 0. ally as simple open-collector
tested from a notebook PC with-
npn switches for driving relays,
out the use of a soldering iron. For ICD operation the PIC lamps and similar loads up to
The full ICEbreaker circuit needs accurate timing. A 20MHz 24V and 400mA.
diagram is shown in Fig.3, crystal X1 together with capaci-
tors C1 and C2 provide the stan- Two other output transistors
whilst Fig.4 gives the PCB de-
dard oscillator components. Alter- are fitted. TR5 is a simple open
tails.
native positions (X2) and (C3, C4) collector npn device and TR6
Resistors R11 to R13, and drives a double-pole
are to be used with 28-pin chips.
R14, Zener diodes D6 to D8 and changeover relay RLA. These
links LK1/2 are the same as the Resistors R15 and R16 allow two devices are useful for bidi-
minimum system. PL2 allows RC oscillator options to be used if rectional control of a DC motor.
the power and computer con- required for testing or running RLA can be wired as a revers-
nections to be extended so that fully debugged code, but as RC ing switch and the motor can be
oscillator stability is poor, this op-
MCLR/VPP/THV 1 40 RB7/PGD
RA0/AN0 2 39 RB6/PGC
RA1/AN1 3 38 RB5
RA2/AN2/VREF 4 37 RB4 MCLR/VPP/THV 1 28 RB7/PGD
RA3/AN3/VREF 5 36 RB3/PGM RA0/AN0 2 27 RB6/PGC
RA4/TOCK1 6 35 RB2 RA1/AN1 3 26 RB5
RA5/AN4/SS 7 34 RB1 RA2/AN2/VREF 4 25 RB4
RE0/RD/AN5 8 33 RB0/INT RA3/AN3/VREF 5 24 RB3/PGM
RE1/WR/AN6 9 32 +VE(VDD) RA4/TOCKI 6 23 RB2
RE2/CS/AN7 10 31 GND/(VSS) RA5/AN4/SS 7 22 RB1
+VE(VDD) 11 30 RD7/PSP7 GND(VSS) 8 21 RB0/INT
GND(VSS) 12 29 RD6/PSP6 OSC1/CLK IN 9 20 +VE(VDD)
OSC1/CLK IN 13 28 RD5/PSP5 OSC2/CLK OUT 10 19 GND/(VSS)
OSC2/CLK OUT 14 27 RD4/PSP4 RC0/T1OSO/T1CKI 11 18 RC7/RX/DT

RC0/T1OSO/T1CKI 15 26 RC7/RX/DT RC1/T1OSI/CCP2 12 17 RC6/TX/CK
RC1/T1OSI/CCP2 16 25 RC6/TX/CK RC2/CCP1 13 16 RC5/SDO
RC2/CCP1 17 24 RC5/SDO RC3/SCK/SCL 14 15 RC4/SDI/SDA
RC3/SCK/SCL 18 23 RC4/SDI/SDA
RD0/PSP0 19 22
PIC16F876 AND 873
RD3/PSP3
RD1/PSP1 20 21 RD2/PSP2
PIC16F877 AND 874
Fig.2. Port connections for the PIC16F877/874 and the alternative 28-pin PIC16F876/873.
turned on and off by TR5. very low prices. A standard
Many applications require Pentium 133 without special from one to the other.
display of information, and an sound, graphics or multimedia
is more than adequate provided
intelligent LCD module is an
it has a spare serial port (COM CONSTRUCTION
ideal display device. X3 has
standard 4- or 8-bit drive capa- 1 – 4). The EPE ICEbreaker
bility, and requires a minimum The software is designed to printed circuit board component
of six output lines for driving. All be run in conjunction with Mi- layout and (approximately) full
pins are available at connector crochip’s MPLAB software. This size copper foil master are
SK1. Preset VR1 allows the is available from many sources shown in Fig.4. This board is
contrast of the LCD to be al- – the Microchip web site is the available from the EPE Online
tered to suit the lighting condi- ideal one as it allows the very Store (code 7000257) from
tions and viewing angle. latest version to be loaded, al- www.epemag.com
Just two input devices are ternatively the Microchip CD- Assembly of the board is
fitted. S2 and S3 are simple ROM is widely available and straightforward. Begin by fitting
single-pole push-to-make good for those without internet seven 12mm pillars with short
switches with pull-up resistors access. M3 screws before adding any
R6 and R7. MPLAB is used in “editor components. Refer to the com-
only” mode to allow assembly ponent layout drawing and then
Whilst the circuit diagram
language source code to be fit plain uninsulated wire links in
seems simple, the PCB layout
written and then assembled to all of the positions shown. Fit
shows that there are far more
produce the necessary .HEX two-way pin headers in the posi-
connection points, and that a
code for programming into the tion for LK1 and LK2 so that two
prototyping area with a solder-
chip. MPLAB also produces a shorting links can be connected
less breadboard is provided.
.COD file which is used by the if required.
Each side of the main inte-
grated circuit (IC) sockets there ICEbreaker software to keep Links LK3 and LK4 provide
are spaces for rows of turned- track of the program execution the facility to set the input power
pin sockets. The inner two rows when debugging, single step- socket for positive or negative
connect to the adjacent IC pins, ping and running the program. inner connection. For positive
whilst the outer two rows are Like many other PC programs, inner fit the links in position B,
power and ground connections. MPLAB has a lot of features for negative fit them in position
that are not regularly used, how- A. As mentioned previously it is
Turned-pin socket strips can
ever the advanced features possible to add a diode in place
be fitted in all rows, but it is
don’t get in the way when using of one of the links to protect
more practical to have a single
it at the simple level required by against polarity reversal. To do
row of sockets each side of the
ICEbreaker. this, fit the cathode of the diode
chip and leave the other spaces
The ICEbreaker software is to the point marked with a +
blank so that pull up or pull
a simple stand-alone application sign, and the anode of the diode
down resistors can be soldered
that can be run directly from a to the appropriate A or B posi-
in position if required. An addi-
floppy disk if necessary. This tion.
tional “patch” area is provided
below IC1 and is ideal for article assumes that the con- Fit the diodes and resistors
adding “permanent” hardware tents of the ICEbreaker disk are next, taking care to identify the
such as LEDs or presets. copied into a new folder type and polarity. Usually the
(directory) on the C-drive, which cathodes are marked with a
has been labeled “icebreak”. black or dark blue band, which
ICEBREAKER The only files required are ice- should be positioned to match
SOFTWARE break.exe and icebreak.ini, but the line on the component lay-
it is also convenient to store out diagram. The transistors
ICEbreaker must be run TR1 to TR6 are all the same
program files in the same direc-
from a PC with at least Win- type and are fitted with their
tory.
dows 95. This helps keep the curved sides as shown in the
software simple, and is not a ICEbreaker and MPLAB
diagram. They should be fitted
serious restriction as PCs that should be run together, and the
close to the board surface so
can run Win95 are available at “Alt” and “Tab” keys or the
that they cannot get bent and
taskbar buttons used to switch
X3 LK3/4
16 x 2 L.C.D. DISPLAY R6 R7 R8 SK5
4k7 4k7 1k PUSH BUTTON
CONNECTIONS +5V OUT IN
14 13 12 11 10 9 8 7 6 5 4 3 2 1 IC2
SK2 7805
SK3 COM. POWER

S2 S3 INPUT
MCLR R14
a 10k
D5 +5V 0V B7 B6
L.E.D. PL1
k PL2
R11
C5 C6

1k
1 6 100n 100n
VR1
10k
R12
10k
11/32(20)
CONTRAST R15*
(R16) 5 9
+V
R1 R13
Copyright © 2000 Wimborne Publishing Ltd and

4k7 470!
*SEE TEXT 40(28)
13(9) RB7
SK1 OSC1/CLK IN
39(27)
RS RB6
R/W IC1
E
D0 PIC16F877/874
D1 X1 (X2) PIC16F876/873 k
D2 20MHz D6
D3 14(10) 5V1
D4 OSC2/CLK OUT a k
D5 LK1 LK2 D7
D6 1(1) 5V1
D7 MCLR a k
D8
0V
L.C.D. CONNECTIONS C1 (C3) 5V1
a
10p 12/31(8/19)
C2 (C4) S1 LINK FOR
10p FREE
RESET RUNNING
LK5 LK6
STEPPING 0V
MOTOR
CONNECTIONS
PL3
SK6
1 C 2 3 C 4 TRANSISTOR TRANSISTOR RELAY
PL4 BASE COLLECTOR DRIVER NC1
RLA1 POLE 1
SK7/TB SK7/TC SK7/RD
k RLA NO1
D9 2
1N4001 NC2
TR1 TR2 TR3 TR4 a RLA2
ZTX451 ZTX451 ZTX451 ZTX451 POLE 2
c c c c NO2
b b b b
Fig.3. Complete circuit diagram for the EPE ICEbreaker.

e k e k e k e k R9 R10
STEPPER MOTOR TR5 TR6
D1 D2 D3 D4 470! 470!
DRIVER INPUTS
ZTX451 ZTX451
30V 30V 30V 30V
a a a a c c
R2 R3 R4 R5 b b
220! 220! 220! 220!
SK4
e e
P4
P3
P2
P1
EPE Online, March 2000 - www.epemag.com - 187

moved around when the board
is handled.
Use turned-pin socket strips
for the 40 and 28-way IC posi-
tions, and position a second row
of sockets alongside. Note that
there is also the option of a
narrow-bodied version of the
28-pin device, and holes have
been drilled to allow for this type
to be used. If required, socket
strips can be fitted for both
types without causing any diffi-
culty. Also fit turned-pin socket
strips for SK1, SK2, SK3, SK4,
SK6, and the three connections
TB, TC, and RD. Also fit two 13-
way strips to the upper and
lower rows of the patch area – Display module removed from the PCB to reveal the regulator
these are the positive and nega- IC mounted underneath.
tive “rails” and make very con-
venient connection points for the board on 16mm long 6BA or Make sure that it is accurately
taking power to the breadboard. M2.5 screws. Fit the four screws positioned (and the right way
Fit pin headers for PL2, PL3 from the track side of the board up) before pressing it firmly into
and PL4 – the holes for these and secure them with nuts. Fit place.
are made tight to give extra four more nuts and position
support and so the pins may them equally so that the LCD
need pressing home against a lies level and approximately TESTING
hard surface. Fit push-switches 10mm from the board. The con- Once the hardware has
S1, S2 and S3, preset VR1, re- nections to the LCD are made been assembled and before fit-
lay RLA and the voltage regula- to allow it to be unplugged for ting IC1, check for dry joints,
tor IC2; an M3 screw and nut access to IC2 and for use in solder bridges and component
should be used to secure the other applications. polarities.
tab. A heatsink is not required Fit a 16-way pin-to-pin con- Once everything looks cor-
for most applications, but there nector to the board, with the rect, connect the power supply.
is space to fit a low profile type, slightly thinner pins upwards. Fit Check that D5 lights and, if a
or even a small piece of alu- (but do not solder) a 16-way meter is available, check the 5V
minum if higher current loads wirewrap turned-pin socket strip regulated supply. The LCD con-
are to be used. to the LCD so that the sockets trast control VR1 should alter
The 20MHz crystal and its face downwards and plug onto the density of a single top row
associated capacitors C1 and the pin-to-pin connector. Make of block characters as the LCD
C2 should be fitted if the (usual) sure the LCD is level, solder the initializes itself for one-line
40-pin device is being used for wire wrap pins to the LCD and mode.
IC1. If a 28-pin version is used cut off the excess. The LCD can Switch off, insert IC1 and
then fit these components to the now be secured by fitting an- connect a suitable lead between
alternative locations X2, C3 and other four nuts. SK1 and the serial port of the
C4. If both types of device may The serial port connector PC, which has MPLAB and the
be used, it is possible to fit two PL2 and power connector SK5 ICEbreaker software (see the
crystals and two pairs of capaci- fit directly onto the board. Make Shoptalk page) installed.
tors. sure that they are pressed fully
home before soldering.
DISPLAY MODULE The solderless breadboard
SOFTWARE
The LCD module fits above is secured to the board simply INITIALISATION
by its self-adhesive backing.
PL1 POWER IN
a a PL2 MCLR
RESET SK5
B6 B7 0V +5V
9 6 S1
D7 D8
5 1
B
LK2
a k R12 k k
D6 LK1 A
R11 R13
R14 A B
LK3/4 X3
LCD DISPLAY
C6
IC2
IC1 IN
COM.
OUT
(28-PIN) C5
IC1 R1
(40-PIN) 14 1
X2
C4
k C R
3 16 VR1 SK1
D5
a SK6 D7
R8 NC2 D6
RLA2 P2 D5
X1
R NO2 D4
15 NC1 RLA D3
C2 C 2
1 RLA1 P1 D2
NO1 D1
S2
D0
COIL
E
k D9 a R/W
TR1 k a
SK3 SK2 PL4 D1 R10 RS
S3 R2 PL3
P1 e b c 1 b
TR2 LK5 c eTR6
P2 C RD
R3 e b c k a R9
R7 P3 2 D2 TB b
TR3 k a c e
P4 R4 e b c 3 D3 TC
R6
SK4 TR4
e b c
C
k a
SK7 TR5
R5 4 D4
ICEBREAKER
257
Fig.4. Printed circuit board component layout and (approximately) full-size copper foil
master pattern.
COMPONENTS
Resistors
R1, R6, R7 4k7 (3 off)
R2 to R5 220 ohms (4 off)
R8, R11 1k (2 off)
R8, R10, R13 470 ohms (3 off)
R12, R14 10k (2 off)
R15, R16 See text
All 0.25W 5% carbon film
Potentiometer
VR1 10k carbon preset
Capacitors
C1 to C4 10p ceramic, 2.5mm pitch (4 off), see text
C5, C6 100n multilayer polyester (2 off)
Layout of components on the completed EPE
Semiconductors
D1 to D4 30V 400mW Zener diodes (4 off)
ICEbreaker PCB.
D5 3mm low-current red LED
D6 to D8 5.1V 400mW Zener diodes (3 off) files called ib1.lst, ib1.hex, ib1.cod, and ib1.err in
D9 1N4001 diode the icebreak directory. The ib1.err file will contain a
TR1 to TR6 ZTX451 npn transistors (6 off)
few warning messages, which can be ignored.
IC1 PIC16F877P20 microcontroller, pre-programmed
IC2 7805 voltage regulator Leave MPLAB running, but minimize it by click-
X1 (X2) 20MHz low-profile crystal (see text) ing on the appropriate box. Open the icebreak file
X3 16x2 alphanumeric LCD module
and double click on icebreak.exe to start the pro-
Miscellaneous gram. The screen will display the main ICEbreaker
Socket strips to make up the following: 11-way (SK1);
1-way (SK2, SK3 -- 2 off); 4-way (SK4); 6-way (SK6) window as shown in Fig.5, and possibly the Watch
SK5 2.1mm PCB power connector and Source windows (Figs. 7 and 8). In the main ICE-
S1 to S3 s.p.s.t. push-to-make switches (3 off) breaker window click on “Options” and then select
RLA d.p.c.o. 5V coil relay (BT47) “Programmer” this will produce the communications
PCB available from the EPE Online Store code 7000257
set up box shown in Fig 6. In this box set up the serial
(www.epemag.com); breadboard; 9-way 90o male D-type COM port that you are using. If a 20MHz crystal is
connector (PL1); 6-way 0.1in. pin header (PL2, PL4 -- 2 off); fitted the Baud box must be set to 38400. Other crys-
6-way pin strip, 2mm pitch (PL3); 2-way 0.1in. pin header tal frequencies can be used and the Baud rate ad-
with DP link plug (2 off -- LK1, LK2); socket strips, 20-way justed proportionally – e.g. a 5MHz crystal would op-
(4 off), 14-way (2 off), 13-way (2 off); 16-way pin-to-pin strip
for LCD; 16-way long-pin socket strip for LCD.
erate at 38400/4 or 9600 Baud. Once set up close the
box by pressing OK.
Hardware: 12mm M3 HEX pillars (7 off); M3 screw x 6mm Back in the main
(7 off); screws CSK (4 off) and 12 nuts (6BA or M2.5) for
LCD mounting.
box select “File” and
“Open”, which will re-
See also the veal a standard file se-
SHOP TALK Page! lect dialog window list-
ing the files in the ice-
break directory. Select
Approx. Cost Guidance Only $60 and load ib1.asm and
then open the source
Run MPLAB, select the “project” tab and then code window by select-
“new project”. In the directory box select c:\ ice- ing ‘Window’ and then
break and set up a project named ib.prj and edit ‘Source’. The window
the project so that it contains the simple test pro- should contain the
gram ib1.asm which is included on the ICEbreaker ib1.asm source file with
disk. Close the “Project Edit” window then select numbered lines as
“File” and ib1.asm. This will open the ib1.asm file shown in Fig.7.
on the MPLAB screen.
Before the program
Next select “Project” and “Make project”. This can be run it must be
will then run the MPASM program and produce Fig.5. Main ICEbreaker sent to IC1 by selecting
window.
tor on the breadboard and
make the two connections
to the turned-pin socket
strips. The row of 13 sock-
ets at the bottom of the
patch area is a good place
to find 0V. Provided the
program has been set up
correctly and loaded into
IC1, the LED should flash
when the program is set to
Fig.6. Setup box. “Run”.
In order to flash the
“Program” and then “Program” LED slowly, the program
in the main ICEbreaker window. has three nested counting
A progress bar appears and loops. To single step
ICEbreaker sends the program
Fig.7. Source file window.
through them would take years,
to the first 4K of the program the “Array” box. and so it is impractical to go
memory in IC1. Once this is right through all of the states of
completed, it should be possible The selector box in the PORTA. The alternative to sin-
to step, run, and reset the code “Watch Add” window allows a gle stepping is to insert a break-
one line at a time using the choice of locations, labels and point and run the program to
“Step” button in the main win- registers to be selected. It is im- there.
dow. In single step mode, at portant to understand that some
of these options are not what Select “Options”,
each step, a highlight line pro- “Breakpoint” from the main ICE-
gresses through the source they seem, for example the W
option returns not the value of breaker window and set the
code window, and the main ICE- value to 24. Fig.10 shows the
breaker window shows the Pro- the W register, but the number
0 which has been assigned to “Breakpoint” setting window. En-
gram Counter, the contents of tering a breakpoint highlights
the W register and the Status the label W in the fifth line of
the source code. the line in the “Source” window.
register bits. “Reset” and then “Run” the pro-
Sometimes the highlight gram and it will now stop at the
does not track the source code TESTING A breakpoint. Press run again and
exactly, and is one line above PROGRAM it will loop again to the same
or below the current line. This is breakpoint – each time incre-
due to the communications be- Once some familiarity has menting the value at PORTA so
tween the computer and IC1 been achieved with the ICE- that the LED on PORTA 0 turns
and depends upon the way breaker windows, it is time to on and off alternately. Try con-
some of the source code is writ- connect some hardware and necting the LED to IC1 pin 3
ten; it is only a minor inconve- see how it operates. As with all PORTA 1 and see that it
nience as the actual line of code good microcontroller hardware switches every other loop.
is easily worked out from the systems, the first thing to do is
Now that a LED can be
program counter in the ICE- flash a LED! The ib1.asm pro-
flashed, it is just a few more
breaker main window. gram counts up through
steps to controlling all sorts of
PORTA, which is set to output
Other registers may be set peripheral devices, and whilst
mode, and so all that is neces-
up in the “Watch” window – se- the PIC16F877 cannot run the
sary is to connect a LED from
lect “Windows”, “Watch” in the proverbial “Power Station” it is
pin 2 of IC1 via a current limit-
main window. Fig.8 shows the capable of an amazing number
ing resistor (anything from
main Watch window and Fig.9 of very complicated feats. The
100W to 2k2) to 0V.
the “Add” window. Registers development of longer pro-
may also be “Watched” by set- Using solid core 1/0·6 con- grams controlling more hard-
ting the first location and enter- necting wire links it is an easy ware is so much easier when it
ing the number of registers in matter to put the LED and resis- is possible to test the programs
quickly in this way. Single step- tor. Connect PORTA 0 to 3 to
ping and watching the program the four stepping motor drive
and data registers allows even sockets P1 to P4 and then fol-
complicated routines to be low the procedures used for
tested and debugged, and sim- ib1.asm to compile, load and
ple changes can be made and run the program. Notes are in-
checked immediately. cluded in the code suggesting
To make a simple change modifications that can be made
to the ib1 program, select to the code for altering speed,
Fig.8. ICEbreaker “Watch” direction, and duration of travel.
“Program”, “Code” from the ice- window.
breaker main menu and then Program ib3.asm runs the
select location 14. The contents LCD. It uses six connections
can be read and should be from PORTC 2 to 8 to connect
30FF, which means MOVLW to RS, E, and D4, 5, 6, and 7 of
FF. Modify the value to 3010 the display in that order. The
which will load the value 10 in- code initializes the display, and
stead of FF and press the then can be set up as a subrou-
“Write” button. Clear the break- tine to write any character to
point, “Run” the program, and any display location. The source
see the change in speed. code has notes to explain the
The program in IC1 has Fig.9. ICEbreaker “Watch operation and to suggest possi-
been modified, but remember Add” window. ble changes for more advanced
that the Source Code has not, applications.
and so will need changing to the
new value once the required COMPLETED
speed has been set. To modify
the source code run MPLAB, PROGRAMS
modify ib1.asm, recompile the Once a program has been
code by selecting “Project”, debugged and is working cor-
“Make project” (or by pressing Fig.10. ICEbreaker rectly, it can be programmed
the appropriate shortcut button) “Breakpoint” window. into another PIC16F877 or any
and then switch back to ICE- other suitable PIC chip using an
breaker . files. Changing register file con-
appropriate programmer (PIC
tents is particularly useful when
Select “File”, “Reopen” and Toolkit Mk2 from the May and
combined with single stepping,
the modified source code will June issues of EPE Online is
as it allows routines to be tested
appear in the ICEbreaker ideal – www.epemag.com/
with a range of values, for ex-
“Source” window. To complete ample a timing loop can be set 0599p1.htm). The ICEbreaker
the operation select “Program”, up with 00 in the loop counting code does not have to be in the
“Program” and the new code will register and single stepped to chip and so any blank chip can
be loaded into IC1. Although the see what happens at the end of be used with this method. The
program in IC1 had already the loop. ICEbreaker board can only pro-
been modified, it is always good gram chips that already contain
practice to reprogram with the Once experience is gained, the special icebreak code – this
newly compiled code to prevent the range of tools available will is necessary because the chip
simple errors creeping in – es- be understood, and it will be- has to communicate with the
pecially when a number of mod- come easy to set up and check PC via the standard serial port
ifications might have been simple routines and combine interface. Chips with icebreak
made. them into full programs. code are readily available – see
As well as changes to the the Shoptalk page.
program memory, the same OTHER PROGRAMS Once programmed, ICE-
procedure can be used to mod- Program ib2.asm is a sim- breaker chips will run normally
ify the EEPROM, and register in other circuits if required to do
ple driver for the stepping mo-
so but it is important to make THE NEXT STEPS EPE ICEbreaker allows pro-
sure that the two pins RB6 and grams that are intended to be
RB7 are connected to 0V. This It is tempting to continue run on much simpler chips to be
is because the ICEbreaker soft- and describe the many features checked and debugged. All that
ware automatically starts to run, of the PIC16F877, but it really is is necessary is to ensure that
and immediately checks for an impossible task because the the ports and register addresses
ground connections on these chip is so powerful (see also our are compatible with the smaller
pins. If it finds that they are PIC16F87x Mini Tutorial in the chips. Applications for the
grounded, the program jumps to Oct ’99 issue of EPE Online). PIC16F84 are particularly suit-
location 0000 and starts running The beauty of the PIC range of able for development using ICE-
the program from there, as a devices is that it is possible to breaker, and so the programs
normal chip would. run the same code on many dif- previously published by EPE
ferent chips. can be used. Note though that
the MPLAB environment uses
MPASM code, and so the PIC
Toolkit Mk2 software will be
necessary to convert the origi-
nal TASM source code to
MPASM assembly language.
ICEbreaker provides an ad-
vanced way to learn program-
ming. Along with the PIC pro-
gramming and data sheets and
back issues of EPE, it will be-
come an indispensable tool for
learning, development, and test-
ing of PIC projects.
IAN POOLE INTRODUCES THE MRAM, WHICH (IT IS HOPED) WILL POSE A REAL
THREAT TO “FLASH” MEMORIES AND OTHERS.
Memory is one of the industry. It is a non-volatile form moment, improvements are

crucial elements in today’s of random access memory anticipated that will give the
technology. There have been claimed to be faster than Flash, new technology a significant
many improvements in disk which will be its main advantage.
based technology over recent competitor when it is launched MRAM has further advantages.
years, but there are similar onto the market. However, as It does not suffer from the wear
levels of progress taking place prices fall and MRAM gains out mechanism experienced
in the memory more directly wider acceptance it is likely that with Flash devices. Although
associated with the CPU itself. it will be used in many more great improvements have been
memory applications. made in this area with Flash
To illustrate this, it was less devices, they still have a limited
than thirty years ago when life and this means that they
ferrite core memory was used. SPECIFICATIONS cannot be used in high usage
Fortunately this was superseded For a given speed and areas of a computer’s
by electronic forms of memory geometry the new MRAM architecture.
when the levels of integration of technology consumes less
circuits grew to a sufficient
degree to allow them to be
power and this is a particularly OPERATION
important factor in today’s
used. technology where many items The operation of the new
Now there are a variety of are battery powered. The power memories is based around a
different forms of memory that reduction also means that the structure known as a magnetic
can be used dependent upon power supply requirements for tunnel junction (MTJ). These
the exact requirement that the unit as a whole may be devices consist of sandwiches
needs to be fulfilled. SRAM, reduced, and this can reflect in of two ferromagnetic layers
DRAM, EPROM, EEPROM and a decrease in costs – all- separated by thin insulating
a variety of others including important in today’s fiercely layers.
Flash memories are available, competitive marketplace. A current can flow across
each having their own forte and As the speed of the new the sandwich arising from a
area where they perform to their devices is faster than Flash, this tunneling action and its
best. too is another selling point as magnitude is dependent upon
However, if it were possible the speed of flash devices can the magnetic moments of the
for only one type of memory to be such that it impacts on the magnetic layers. These layers
be used, then this could lead to operation of the whole unit. can either be the same, when
more efficient use of the Although not much faster at the they are said to be parallel, or in
circuitry, and possible cost opposite directions when they
savings. For this to be feasible, TOP LEAD are said to be antiparallel.
an all-purpose, yet low cost FREE FERROMAGNET
Magnetic tunnel junctions
memory would need to be TUNNEL JUNCTION comprise sandwiches of two
available, and this may be just PINNED FERROMAGNET ferromagnetic (FM) layers
around the corner because a ANTIFERROMAGNET
separated by a thin insulating
new technology is about to hit layer which acts as a tunnel
the market. SEED LAYER barrier (Fig.1). In these
Known as magneto BOTTOM LEAD structures the sense current
resistance random access usually flows parallel to the
memory, or MRAM, the new
SUBSTRATE layers of the structure, while the
memory is creating significant current write is passed
interest in the semiconductor
Fig.1. Structure of an perpendicular to the layers of
MRAM cell. the MTJ sandwich.
New technology Updates
The resistance of the MTJ variation in resistance due to flexibility and tolerance relative
sandwich depends on the magnetoresistive effects were to silicon.
direction of magnetism of the seen. Investigations into the In the next stage of
two ferromagnetic layers. dependence of MR on the development it is hoped to use
Typically, the resistance of the ferromagnetic metals a silicon diode or transistor to
MTJ is lowest when these comprising the electrodes were reduce the fabrication costs.
moments are aligned parallel to made. This will bring the final version
one another, and is highest It was anticipated that the nearer and ensure that its unit
when antiparallel. magnitude of the MR would costs will be such that it can
To set the state of the largely be dependent on the effectively compete with
memory cell a write current is interface between the tunnel existing technologies.
passed through the structure. barrier and the magnetic The new MRAM technology
This is sufficiently high to alter electrodes. It was found that is an exciting development that
the direction of magnetism of thick layers of certain non- could revolutionize current
the thin layer, but not the thicker ferromagnetic metals could be trends in electronic memory.
one. A smaller non-destructive inserted between the tunnel The migration from magnetic
sense current is then used to barrier and the magnetic core memory in the early 1970s
detect the data stored in the electrode without quenching the proved to be a major step
cell. MR effect. However, the MR forwards. Now the introduction
was quenched by incomplete of MRAMs could provide similar
oxidation of the aluminum layer. levels of benefit.
CONSTRUCTION
A wide range of structures
and materials have been
OTHER
investigated to obtain the DEVELOPMENTS
optimum structure. In view of IBM is not the only
the potential of the new manufacturer investigating
technology a number of these structures. Apart from
manufacturers are investigating Motorola, IMEC, Europe’s
different approaches. Motorola, leading research center for the
IBM and many others all believe development and licensing of
there is a future for the new state-of-the-art microelectronics
idea. technologies, is also making
IBM have fabricated significant developments with
junctions using computer- MRAM technology.
controlled placement of up to They have succeeded in
eight different metal shadow developing a demonstration
masks. The masks were MRAM matrix cell array with a
successively placed on any one DRAM style of architecture.
of up to twenty 25mm diameter This brings MRAM technology a
wafers with a placement step closer to the production of
accuracy of approximately ± a viable alternative to the
40um. By using different masks, existing non-volatile memory.
between 10 to 74 junctions of a The memories produced in this
size of approximately 80 x development used a similar
2
80um could be fashioned on structure to that employed by
each wafer. IBM, i.e. two magnetic layers
The tunnel barrier was separated by an insulating
formed by in-situ plasma layer.
oxidation of a thin aluminum Bit-selective addressing
layer deposited at ambient was based around a GaAs
temperature. Using this diode. The GaAs technology
technique, large levels of was selected because of its
TECHNOLOGY TIMELINES
Part 2 - DAYS OF LATER YORE
by Alvin Brown and Clive “Max” Maxfield
Boldly going behind the beyond, behind which no one special words like dozen
has boldly gone behind, beyond, before! (meaning 12) and gross
(meaning 12 x 12). The fact that
we have 12 months in a year
The purpose of this series is and 24 hours in a day (2 x 12)
to review how we came to be DECIMAL NUMBER are also related to these base-
where we are today (technology- SYSTEM 12 systems.
wise), and where we look like
ending up tomorrow. Our first Throughout history, humans However, the number
step is to cast our gaze into the have experimented with a system with which we are most
depths of time to consider the variety of different number familiar is the decimal system,
state-of-the-art as the world was systems. For example, the which is based on ten digits: 0,
poised to enter the 20th Century. ancient Babylonians used a 1, 2, 3, 4, 5, 6, 7, 8 and 9. The
base-60 system, which is why name decimal is derived from
In Part 1 we considered the Latin decam, meaning ten.
we have 60 seconds in a minute
physics, electronics, and As this system uses ten digits, it
and 60 minutes in an hour.
communications prior to 1900. In is said to be base-10 or radix-
Some people used their fingers
this installment we shall 10, where the term radix
first examine the state of comes from the Latin
computing prior to 1900. word meaning root.
We shall then turn our
attention to the key
discoveries in THE ABACUS
fundamental electronics
The first actual
that occurred in the 20th
calculating mechanism
Century.
(at least that we know of)
is the abacus. Some
COMPUTING authorities hold that the
first abacus was
PRIOR TO 1900 invented by the
The first tools used as Babylonians sometime
aids to calculation were between 1,000 BC and
man’s own fingers. Thus, Napier’s Bones, simple multiplication tables in- 500 BC, but others
it is no coincidence that scribed on wood or bone. John Napier went on believe that it was
digit refers to a finger (or to invent logarithms. Courtesy of IBM. actually invented by the
toe) as well as a Chinese.
numerical quantity. Although the abacus
Similarly, small stones or pebbles and their toes for counting, so does not qualify as a
could be used to represent larger they ended up with base-20 mechanical calculator, it
numbers than fingers and could systems, which is why we still certainly stands proud as one of
also store intermediate results for have special words like score, first mechanical aids to
later use. This explains why meaning twenty. calculation.
calculate is derived from the Latin Similarly, some groups
word for pebble (calculus). experimented with base-12
systems, which is why we have THE SLIDE RULE
Special Feature
and theologian, Blaise Pascal
with the invention of the first
operational calculating machine.
In 1642, Pascal introduced his
Arithmetic Machine, which could
add and subtract numbers
(multiplication and division
operations were implemented by
performing a series of additions
A modern construction of a mechanical calculator in- or subtractions).
spired by a drawing in one of Leonardo da Vinci’s note-
books. Courtesy of IBM. STEP RECKONER
In the early 1600s, Scottish mechanical calculator (see Part Mechanical calculation was
mathematician John Napier 1). Working models of a taken a step further in the 1670s
invented a tool called Napier’s mechanical calculator loosely by a German Baron called
Bones. These were simple based on these drawings have Gottfried von Leibniz. After
multiplication tables inscribed since been constructed, receiving his bachelor’s degree at
on strips of wood or bone. although some people believe seventeen years of age, Leibniz
Napier also invented the the reconstruction is far more developed Pascal’s ideas and, in
concept of logarithms, which sophisticated than anything 1671, introduced the Step
were used as the basis for the Leonardo had in mind. Reckoner. In addition to
slide rule by the English performing additions and
mathematician and clergyman subtractions, the Step Reckoner
William Oughtred in 1621. FIRST MECHANICAL could multiply, divide, and
The slide rule was an CALCUL
exceptionally effective tool that ATOR
remained in common use for
Sometim
over three hundred years.
e around
However, like the abacus, the
1625, the
slide rule does not qualify as a
German
mechanical calculator in the
astronomer
modern sense of the word.
and
mathematicia
REDISCOVERED n Wilhelm Blaise Pascal’s Arithmetic Machine. Courtesy of IBM.
Schickard
NOTEBOOKS wrote a letter to a friend stating
Leonardo da Vinci was a that he had built a machine that evaluate square roots.
genius: painter, musician, “… immediately computes the The mechanical calculators
sculptor, architect, engineer, given numbers automatically; created by Pascal and Leibniz
and so forth. It is well known adds, subtracts, multiplies, and were the forebears of today’s
that he sketched concept divides’’. Unfortunately, the desktop computers, and
designs of such futuristic original machine was destroyed derivations of these devices were
devices as tanks and in a fire, but working models in widespread use for over two
helicopters, but until quite have since been constructed hundred years until their
recently there was no indication from Schickard’s notes. electronic counterparts finally
that he had ever turned his became available and affordable
mind to mechanical calculation. in the early 1970s.
However, two of da Vinci’s
ARITHMETIC
notebooks dating from around MACHINE
the 1500s were rediscovered in FIRST MECHANICAL
For reasons unknown, many
1967. These priceless tomes references ignore Schickard’s COMPUTER
contain a wealth of drawings, device and instead credit the In the 1800s, books of
some of which may represent a French mathematician, physicist mathematical tables such as
Special Feature
modern computers, consisted of 4,000 components,
including a processing weighed in at a whopping three
unit that could change tons, and was 10 feet wide and
the flow of a program 6·5 feet tall (3m x 2m).
depending on the results In the early 1990s, more than
of previous one hundred and fifty years after
computations. Babbage its conception, Babbage’s
worked on the Analytical Difference Engine performed its
Engine from around first sequence of calculations and
1830 until he died in returned results to 31 digits of
1871, but sadly it was accuracy, which is far more
never completed in his accurate than most of today’s
lifetime. electronic pocket calculators!
FIRST TABULATING
COMPUTER MACHINES
PROGRAMMER These days we can only
Modern construction of Babbage’s Dif- Augusta Ada imagine the problems besetting
Lovelace, the daughter the American census takers in the
ference Engine. Courtesy of IBM.
of the English poet Lord latter part of the 19th Century,
logarithmic and trigonometric Byron, was one of the because it was estimated that the
functions were in great demand few people who had any clue 1890 census would include more
by navigators, engineers, and what Babbage was talking than 62 million Americans.
so forth. Such tables were about. Ada created a program The problem was that the
generated by teams of to compute a mathematical existing system of making tally
mathematicians working day sequence known as Bernoulli marks in small squares on rolls of
and night on derivatives of the numbers, which would have paper and then adding these
primitive mechanical calculators been extremely interesting had marks together by hand was time-
invented by Pascal and Leibniz. Babbage’s machine ever consuming, expensive, and prone
These mathematicians were actually worked. to error. In fact it was determined
referred to as computers Ada is now credited as that if the existing system were
because they performed all the being the first computer used for the 1890 census, then
computations. programmer, and the ADA the processed data would not be
In 1822, the eccentric programming language was ready until after the 1900 census,
British mathematician and named in her honor in 1979. by which time it would be largely
inventor Charles Babbage In fact, one of Babbage’s worthless!
proposed building a machine earlier Difference Engines was During the 1880s, a lecturer
called the Difference Engine to eventually assembled by a team at MIT called Herman Hollerith
automatically calculate these at London’s Science Museum came up with a solution to this
tables. from his original drawings. The problem, which was to use
Babbage had only partially final machine was constructed punched cards to represent the
completed the Difference from cast iron, bronze and steel, census data, and to then read
Engine when he conceived the
idea of a much more
sophisticated machine called an
Analytical Engine. (This is often
referred to as Babbage’s
Analytical Steam Engine,
because he intended for it to be
powered by steam).
The Analytical Engine
included many concepts that Gottfried von Leibniz’s Step Reckoner of 1671.
would eventually be featured in Courtesy of IBM.
Special Feature
However, as with most things
POISED ON (especially small children),
control is easier to talk about than
THE BRINK it is to achieve. With the
… exception of simple manipulation
and modulation using devices
So now we are such as transformers, or
poised on the brink rectification using crystals, the
of the 20th Century. most sophisticated form of control
It’s 11:59pm on prior to the beginning of the 20th
December 31st Century was the mechanical
1899. Queen switch (or its electromechanical
Victoria is still on counterpart, the relay).
the throne of
England. Light When it comes to coarse
bulbs are control like turning a light bulb on
Hollerith’s Tabulator/Sorter unit. considered to be or off, then a mechanical switch
Courtesy of IBM. amazingly cool (but is definitely worth considering, but
almost no-one has if you’re looking for fine control,
and collate this data using electricity in their homes). The mechanical switches generally
machines. vacuum tube has yet to be leave something to be desired.
The result of Hollerith’s invented. Rudimentary Similarly, a mechanical device is
labours was an automatic telephones are available only to only of use if you only wish to turn
tabulating machine with a large the favored few, and something off every now and
number of clock-like counters “computers” are the ill-used then, but such devices have a
that were used to accumulate mathematicians who furiously maximum capability of only a
and display results. Operators hand-crank their mechanical very few cycles per second.
used switches to instruct the calculators in the dead of night! So one key requirement as
machine to examine each card Tick-tock, tick-tock . . . the we entered the 20th Century
for certain characteristics, such second hand is wending its way (“we” meaning the human race,
as gender, profession, marital towards the beginning of a new not the authors personally you
status, number of children, and century. Who can guess what understand) was for a more
so forth. surprises the future will hold? sophisticated way to control
An electrically controlled electricity.
sorting mechanism detected
any cards that met the specified
criteria and gathered them into
VACUUM TUBES –
a separate container. The ability FLEMING’S DIODES
to quickly and easily collate FUNDAMENTAL As we discussed in Part 1,
data in this way drove
statisticians of the time into a ELECTRONICS IN the American Inventor Thomas
TH Alva Edison demonstrated his
frenzy of excitement. THE 20 first incandescent light bulb in
Following their application 1789 (one year after the English
to the census problem, CENTURY inventor Sir Joseph Wilson Swan
Hollerith’s machines proved For our purposes here, demonstrated his bulb, but let’s
themselves to be extremely electricity may be considered to not delve into that debate again
useful for a wide variety of consist of vast herds of here).
applications, and some of the electrons migrating from one Four years later in 1883, an
techniques they used were to place to another through some engineer working for Edison –
prove significant in the conducting medium like a William Hammer – observed that
development of the digital copper wire. The art of he could detect electrons flowing
computer in the 20th Century. In electronics comes in controlling from the lighted filament to a
fact, in February 1924, these herds: telling them when metal plate mounted inside the
Hollerith’s company changed its to start, when to stop, where to bulb. Even though Hammer
name to International Business go, and what to do when they discovered this phenomena, it
Machines, or IBM! get there.
Special Feature
In fact it wasn’t until the third electrode called the grid into
Edison Effect’s twenty-first his version of a vacuum tube.
birthday in 1904 that the The resulting three-terminal
English electrical engineer, device was called a triode.
John Ambrose Fleming, filed This device was particularly
a patent for the first vacuum cunning, because a small signal
tube device based on this applied to the grid could be used
effect. (Due to the fact that to control a much larger signal
these devices are created flowing between the cathode and
using evacuated glass tubes, the anode. The result was a
they are still referred to simply device that could be used to
as tubes in America. However amplify signals, and de Forest
in England they became more used his triodes to build many of
commonly known as valves, the early radio transmitters (he
because this name – derived also presented the first live opera
from pneumatic and hydraulic broadcast and the first news
valves – better reflected their report on radio).
control function.)
In addition to acting as an
Sir John Fleming (1849-1945), What Fleming had amplifier, de Forest’s triodes
British physicist and inventor of discovered was that the could also be used in the role of
the thermionic valve. Courtesy of electrons in his vacuum tube switches (the presence or
the Science Photo Library. only flowed from the cathode absence of a signal on the grid
(the heated filament) to a terminal could turn the output –
positively charged anode. the anode – on or off). This ability
subsequently became known as Thus, Fleming had created a
the Edison Effect, because to act as switches meant that
form of diode, which is a device vacuum tubes were destined to
Edison was the man in charge. that only conducts electricity in play a significant role in digital
Sad to relate, Edison one direction. computing.
himself did not take the time to This was of particular
investigate the effect any As we shall see in a future
interest, because some installment, early digital
further. This was unfortunate, electrical equipment like radios
because electronics as we now computers (circa 1940) were
will only work with either mechanical or
know it might have taken a unidirectional, or direct current
giant leap forward had he done electromechanical (based on
(DC), but most electrical relays), but they soon came to be
so. supplies are based on constructed from vacuum tube
alternating current switches, because these were
(AC), because this much, much faster.
provides a more
efficient way to Unfortunately, vacuum tubes
transport electricity have a number of disadvantages,
over long distances. not the least that the metal
Fleming’s vacuum tube forming the cathode evaporates
diode could therefore over time causing a performance
be employed in the degradation. Also, in addition to
role of a rectifier to requiring dangerously high
convert AC to DC. voltages, vacuum tubes occupy a
lot of space, they generate a lot
of waste heat, and they are not
LEE DE particularly reliable, which
FOREST’S becomes especially noticeable
when they are used in large
An Audion triode from about 1914.
TRIODES numbers.
Right: An MO Valve Co. triode from In 1907, the For example, the ENIAC
about 1920. Courtesy of Radio Bygones American inventor Lee computer, which was constructed
de Forest introduced a at the University of Pennsylvania
(www.radiobygones.com).
Special Feature
loud alarms). transistor being used in the role of
Prior to an amplifier (the title of the patent
Fleming inventing was the Method and apparatus
his vacuum tube for controlling electric currents).
diode, early radios
relied on the use of SEMICONDUCTORS
crystals for
rectification. At that Some substances facilitate
time no one really the conduction of electricity and
understood how are therefore known as
crystals could conductors. Other materials resist
convert an AC the flow of electricity, and these
signal into its DC are known as insulators. What the
counterpart, and early pioneers didn’t fully
following the understand is that by adding
Part of the ENIAC computer. Courtesy of IBM. advent of the impurities to certain crystalline
vacuum tube most structures, it is possible to create
between 1943 and 1946, used a special class of materials
approximately 18,000 vacuum people really couldn’t care less.
known as semiconductors, which
tubes. This monster was 10 feet However, some scientists,
can exhibit both conducting and
(3m) tall, occupied 1,000 square inventors and engineers did
insulating properties.
feet (93 m2) of floor space, and remain interested in crystals in
required 150 kilowatts of power, general, especially as they Sad to relate, serious
which was enough to light a began to discover more of the research into semiconductors did
small town at that time. special properties associated not really commence until World
with different crystalline War II. At that time it was
However, whilst it was a
structures. For example, in recognized that devices formed
tremendous achievement for its
1907 a letter from Mr. H.J. from semiconductors had
time, ENIAC was painfully
Round was published in the potential as amplifiers and
unreliable due to the vacuum
American Electrical World switches. If it proved possible to
tube technology of the day. In
magazine as follows: create them, these new devices
fact 90 per cent of ENIAC’s
could be used to replace
down-time was attributed to To the editors of Electrical
prevailing vacuum tube
locating and replacing burnt-out World: Sirs – During an
tubes – sometimes as many as investigation of the
50 a day! unsymmetrical passage of
current through a contact of
carborundum and other
CRYSTAL GAZING substances a curious
The fact that certain crystals phenomenon was noted. On
have special properties had applying a potential of 10 volts
been known for a long time. For between two points on a crystal
example, in 1880, the French of carborundum, the crystal
physicist Pierre Curie had gave out a yellowish light.
discovered the piezoelectric
Mr Round went on to note
effect. In this case, certain
that some crystals gave out
crystalline substances produce
green, orange, or blue light.
an electrical charge if they are
This is quite possibly the first
squeezed, and correspondingly
documented reference to the
they change size if an electric
effect upon which light-emitting
current is applied to them. The first ever transistor,
diodes (LEDs) are based.
This effect subsequently created on 23 December
Similarly, as far back as
found many diverse applications 1947. The photo scale is
1926, Dr Julius Edgar Lilienfield
in electronics, from sensors
of New York filed for a patent approximately twice life
(including microphones) to
on what we would now size. Courtesy of Bell Laborato-
actuators (including extremely
recognize as an npn junction ries/Lucent Technologies.
Special Feature
TIMELINES technology, with the advantages semiconductor of choice. (As
that they would be much silicon is the main constituent of
1901: Hubert Booth invents the smaller and lighter, they would sand and one of the most
first vacuum cleaner. consume much less power, and common elements on earth –
1902: Robert Bosch invents the they would be far more reliable. silicon accounts for
first spark plug. approximately 28 percent of the
Earth’s crust – we aren’t in any
1902: America. Millar Hutchin- TRANSISTORS danger of running out of it in the
son invents the first elec-
In the early 1940s, foreseeable future.)
trical hearing aid.
scientists at Bell Labs in the Very quickly after the first
1904: England. John Ambrose United States started to transistors had been developed
Fleming invents the vac- experiment with impure crystals they started to appear in
uum tube diode rectifier. of germanium. The first true commercial products. For
1904: First ultraviolet lamps semiconductor components example, 1952 saw the
are introduced. were two-terminal diode appearance of the transistor-
devices. On the 23rd December based hearing aid, quickly
1904: First practical photoelec-
1947, a team comprising the followed by Sony’s pocket-sized
tric cell is developed. scientist William Bradford transistor radio. It was also
1906: First tungsten-filament Shockley and the theoretical obvious to computer scientists
lamps are introduced. physicists John Bardeen and that they could now make
1907: America. Lee de Forest Walter Brattain succeeded in machines much smaller (the size
creates a three-element creating the first point-contact of a room instead of a house), but
amplifier vacuum tube transistor (whose name was only a very few forward thinkers
(triode). derived from transfer resistor). had any idea as to what was yet
1908: Charles Frederick Cross Like the triode, this was a to come …
invents Cellophane. three terminal device that could
be used both as an amplifier
1909: Leo Baekeland patents and a switch. Once they had
INTEGRATED
an artificial plastic that proved that their creation CIRCUITS
he calls Bakelite. worked, the team broke up to Sometime after the invention
1909: General Electric intro- celebrate the Christmas of the transistor, people began to
duce the world’s first holidays, which is why many think that it would be a good idea
electric toaster. (lesser) references state that the to be able to fabricate entire
1910: First electric washing first transistor did not make an circuits on a single piece of
appearance until 1948. semiconductor. In fact, the first
machines are introduced.
In 1950, Shockley invented public discussion of this concept
1910: France. Georges Claude
a new type of device called a is generally credited to a British
introduces neon lamps.
bipolar junction transistor (BJT), Radar expert, G. W. A. Drummer
1911: Dutch physicist Heike which was more reliable, easier in a paper he presented as far
Kamerlingh Onnes dis- and cheaper to build, and gave back as 1952. However, it was
covers superconductiv- more consistent results than not until 1958 that a young
ity. point contact devices. Then, in engineer called Jack Kilby
1912: The Titanic sinks on its 1962, Steven Hofstein and actually succeeded in creating
maiden voyage. Fredric Heiman at the RCA multiple components as a single
research laboratory at device.
1912: America. Dr Sidney Rus- Princeton, New Jersey, invented
sell invents the electric To a large extent the demand
a new family of devices called for miniaturization was driven by
blanket. field effect transistors (FETs). the requirements of the American
1913: William D. Coolidge in- Although germanium armed forces and also by rocket
vents hot-tungsten fila- exhibits more desirable research. At that time, one
ment X-ray tube. This electrical characteristics, for a technique that was receiving a lot
Coolidge Tube becomes variety of reasons silicon is of attention was the Micro-Module
standard generator for easier to work with, and so by program, which was sponsored by
medical X-rays. the late 1950s silicon had the US Army Signal Corps.
1914: America. Traffic lights replaced germanium as the
Special Feature
are used for the first time
(in Cleveland, Ohio).
1917: Clarence Birdseye pre-
serves food using freez-
ing.
1919: The concept of flip-flop
(memory) circuits is in-
vented.
1919: Walter Schottky invents
the tetrode (first multiple-
grid vacuum tube).
1921: Czech author Karel
Capek coins the term
robot in his play R.U.R.
1921: Albert Hull invents the The first integrated circuit, created by Jack Kilby of
magnetron (a microwave Texas Instruments in 1958.
generator). Courtesy of Texas Instruments.
1921: Canadian-American
Using this technique, all of devices.
John Augustus Larson the components were created
invents the polygraph lie- In a desperate attempt to
with a uniform size and shape, avoid being consigned to working
detector. and the wiring was built into the on Micro-Modules for the rest of
1921: First use of quartz crys- components themselves. These his career, Kilby started
tals to keep radios from Micro-Modules could then be pondering the fact that multiple
wandering off-station. snapped together to form devices such as resistors,
1923: First neon advertising circuits. capacitors and transistors could
sign is introduced. Texas Instruments was be fabricated on a single piece of
1923: First photoelectric cell is working on the Micro-Module semiconductor and connected
introduced. program when Jack Kilby joined together in situ to form a
the company in May 1958. In complete circuit.
1926: America. First “pop-up’’ the middle of the summer, most
bread toaster is intro- As soon as his boss returned
of the plant was to shut down from vacation, Kilby explained his
duced. for a mass vacation, but as a ideas and received permission to
1926: America. Dr Julius Edgar new employee Kilby did not experiment further. On 12th
Lilienfield from New York have any vacation time coming, September 1958, he powered up
filed for a patent on what so he was left to his own his first prototype – a phase shift
we would now recognize oscillator – which
as an npn junction tran- immediately started to
sistor being used in the oscillate at
role of an amplifier. approximately 1××3MHz.
1927: Harold Stephen Black Although
conceives the idea of manufacturing
negative feedback techniques
which, amongst other subsequently took
things, makes hi-fi ampli- different paths to those
fiers possible. used by Kilby, he is still
1927: First five-electrode vac- credited with the
uum tube (the pentode) manufacture of the first
is introduced. true integrated circuit.
1928: Joseph Schick invents The original bipolar
the electric razor. junction transistors
The first microprocessor, Intel’s 4004, of were manufactured
1928: America. First quartz
1971. Courtesy of Intel. using the mesa process
Special Feature
crystal clock is intro- William Shockley invents market.
duced. first bipolar junction tran- 1970: America. Fairchild intro-
1930: America. Sliced bread sistor. duce the first 256-bit
arrives. 1952: England. First public dis- static RAM called the
1936: Fluorescent lighting ar- cussion of integrated cir- 4100.
rives. cuits is credited to a 1970: America. Intel announce
British radar expert, the first 1024-bit dy-
1938: Hungarian Lazlo Biro G.W.A. Dummer.
patents the first ball-point namic RAM, called the
pen. 1954: America. C. A. Swanson 1103.
company markets the 1971: America. Intel creates
1938: Walter Schottky discov- first “TV Dinner’’.
ers the existence of
holes in the band struc- 1955: Velcro is patented.
emitter into the base.
ture of semiconductors 1957: America. Gordon Gould
One of Hoerni’s colleagues,
and explains metal/semi- conceives the idea of the
Robert Noyce, invented a tech-
conductor interface recti- laser.
nique for growing a layer of sili-
fication. 1958: America. Jack Kilby, con dioxide insulator over the
1938: America. Claude E. working for Texas Instru- transistor, and then etching this
Shannon publishes an ments, succeeds in fabri- layer to expose small areas
article (based on his cating multiple compo- over the base and emitter. Thin
Master’s thesis at MIT) nents on a single piece layers of aluminum were subse-
showing how Boolean of semiconductor. quently diffused into these ar-
algebra could be used to 1959: Swiss physicist Jean Ho- eas to create wires. The pro-
design digital circuits. erni invents the planar cesses developed by Hoerni
1939: Light-emitting diodes process, in which optical and Noyce led directly to mod-
were patented by lithographic techniques ern integrated circuits.
Messers Bay and are used to create tran- By 1961, both Fairchild and
Szigeti. sistors. Texas Instruments had an-
1943: German engineer Paul 1959: America. Robert Noyce nounced the availability of the
Eisler patents the printed invents technique for first commercial planar inte-
circuit board. creating microscopic alu- grated circuits, comprising sim-
minum wires on silicon ple logic functions. Only nine
1945: Percy L. Spensor invents years later in 1970, Fairchild
the microwave oven (the (leads to development of
integrated circuits). introduced the first
first units go on sale in semiconductor-based 256-bit
1947). 1960: America. Theodore static RAM, while Intel an-
1947: America. Physicists Maimen creates the first nounced the first 1024-bit dy-
William Shockley, Wal- laser. namic RAM.
ter Brattain, and John 1962: America. Steven Hof- Then in 1971, Ted Hoff et al at
Bardeen create the first stein and Fredric Heiman Intel invented the world’s first
point-contact germanium at RCA Research Lab computer on a chip – the 4004
transistor on the 23rd invent field effect transis- microprocessor. The successors
December. tors (FETs). of this device (the 4040, 8008
1948: First atomic clock con- 1962: America. Unimation in- and 8080) heralded a new area
structed. troduces the first indus- in computing. Systems small
1950: Maurice Karnaugh in- trial robot. enough to fit on a desk could be
vents Karnaugh Maps 1967: America. Fairchild intro- created with more processing
(circa 1950), which duce an integrated circuit power than monsters weighing
quickly become one of called the Micromosaic tens of tons only a decade be-
the mainstays of the (the forerunner of the fore.
logic designer’s tool- modern ASIC). Almost unbelievably, individuals
chest. could now own their own per-
1968: America. First Static
sonal computer. As we shall see
1950: America. Physicist RAM IC reaches the
in future parts, the effects of
Special Feature
these developments are still un- NEXT MONTH modules (MCMs) really deserve
folding, but it is not excessive to an article in their own right. Im-
say that electronics in general, In Part 3, we shall consider the age capture and display technolo-
and digital computers in particu- development of communica- gies like charge-coupled devices
lar, have changed the world tions in the 20th Century. (CCDs), cathode ray tubes
more significantly than almost (CRTs), liquid crystal displays
any other human invention. HUMBLE (LCDs) and so forth are certainly
entitled to be acknowledged.
APOLOGIES Devices and techniques such as
MORE INFO Before you reach for your pen masers, lasers, fiber optics, mag-
The way in which components or computer keyboard to send netrons (leading to both Radar
like transistors and integrated us a stern letter of chastise- and microwave ovens), mercury
circuits perform their magic is ment, we crave your indulgence delay lines and magnetic core
discussed in greater detail in and ask you to accept our hum- stores also deserve a hearing.
Bebop to the Boolean Boogie blest apologies for all of the And of course a tremendous vari-
(An Unconventional Guide to things we had to leave out. The ety of other technologies, ranging
Electronics), while the history of history of cunning devices such from plastics to storage
the early computers is dis- as light-emitting diodes (LEDs), (batteries), have all played a
cussed in more detail in Bebop laser diodes, phototransistors large part in the evolution of elec-
BYTES Back (An Unconven- and suchlike would make fasci- tronics as we know it today.
tional Guide to Computers). By nating reading. Unfortunately, space limitations
some strange quirk of fate, both Who could deny that intercon- meant that when “push came to
of these books are available nection and packaging tech- shove” we had to make some
from the EPE Online Store at nologies like printed circuit stern decisions. And when one
www.epemag.com boards (PCBs), flexible circuit peels the outer layers away, it be-
boards, hybrids, and multichip comes obvious that the three ab-
solutely core electronics develop-
ments of the 20th Century are the
vacuum tube, transistor, and inte-
grated circuit, as discussed
above. Of course, if you disagree
(or if you simply crave more once
this series is finished), please feel
free to vent your feelings by inun-
dating the Editor with your letters
and emails.
PART 5 – Waveforms, Frequency,
and Time
by John Becker
What we are doing during this Teach-In In the first four parts we discussed (and
2000 series (of at least 10 parts) is to lead you gave you practical experience of) several types
through the fascinating maze of what of commonly used electronic components.
electronics is all about! We are assuming that Last month we also showed you how to
you know nothing about the subject, and are interface your experimental circuits to a PC-
taking individual components and concepts in compatible computer, allowing pulse
simple steps and showing you, with lots of waveforms to be displayed on the screen. We
examples, what you can achieve, and without it now discuss other waveforms and how to
taxing your brain too much! observe them on your PC, adding just one
Greetings again! We expect Before that, though, we screen. Connect the computer’s
you’ve got your breadboard cou- shall discuss a little more fully parallel port to your breadboard,
pled to the computer, all not just the above two pro- without the oscillator connected
powered-up and ready to go. But, grams, but also the preceding to any of the five input points
wait, is the oscillator on your one, Parallel Port Data Display/ (IN0 to IN4).
breadboard the same as that de- Set. In doing so we shall also
scribed in Fig.4.1 and Fig.4.3 of introduce you to the first of sev-
eral concepts in Digital Logic, INPUT REGISTER
Teach-In Part 4 last month? If
not, make it so, then we’re ready that of the AND function. Now look at the Direct Input
to start… From the main menu, call Byte box on the screen (Photo
up the Parallel Port Data Dis- 4.8 last month). What you
play/Set program to your should see is the Original Byte
MAKING WAVES
Last month, we left you with
the suggestion that you should
explore two of our simulation pro-
grams: Computer as Frequency
Counter and Pulse Input Wave-
form Display. (Remember that
this software is available for free
download from the EPE Online
Library at www.epemag.com,
Ed.)
In the second of these pro-
grams, at some stage of your ex-
perimentation, hopefully you pro-
duced a screen display that looks
something like that in Photo 5.1.
It is waveforms that are the main
subject of our discussion this Photo 5.1. Typical pulse waveforms input to the
month.
computer via the simple interface board.
TEACH-IN 2000
Bit Power Decimal Binary lows the correct logic on all five ANDed answer = 00010000
breadboard inputs to be seen as (decimal 16)
7 27 128 10000000 such in box 2, as the New Byte. therefore bit 4 is high. Con-
6 26 64 01000000 This type of manipulation versely:
5 25 32 00100000 can be extremely useful when
4 24 16 00010000 using the computer as an item Register value = 01001000
3 23 8 00001000 of test gear to monitor a digital (decimal 72)
value shown as 10000xxx bi- circuit. AND value = 00010000
nary (where x can be either 1 or (decimal 16)
ANDed answer = 00000000
0). This indicates that the com- FREQUENCY COUNTER (decimal 0)
puter’s input port register (a
special form of digital storage It is also possible to read therefore bit 4 is low.
device) believes that its eight the status of each register bit by
bits are at the logic states ANDing it with particular binary Thus all that is needed to iso-
shown. values (a Binary Logic subject late the status of a bit is to AND it
coming in Part 6), and this is with the required value. If the bit
Your breadboard interface what is done in the Computer as is low an answer of zero will re-
circuit only uses five bits, but the Frequency Counter pro- sult; if it’s high an answer greater
which five of the eight bits ap- gram. Call it up from the menu, than zero will result. So you sim-
parently available? Probably and couple your oscillator IC1a ply check whether or not the an-
from your experiments you al- pin 2) to IN0 as you have done swer is zero and act accordingly.
ready know the answer, the left- before. It is best if diodes D2 The AND command/facility is a
hand five, bit 7 to bit 3 are those and D3 are omitted and a 1kW powerful tool in computing and
used in connection with bread- electronics.
resister used in position D3
board inputs IN4 to IN0.
(also as you’ve done before). Note that in the computer
As far as the computer is program that actually controls
You will have discovered
concerned, bits 2 to 0 are not to what you are now seeing on
that the program only responds
do with data input (what their screen, binary values cannot be
to frequency input signals from
logic values are depends on recognized as such, consequently
one interface input path at a
your computer system – the au- the decimal equivalent is that
time, and that for this path to be
thor’s various computers return used in the Frequency Counter
seen as active by the program,
different values). program.
the Active Bit in the screen box
What might also be puz- has to be appropriately set from This program has been writ-
zling you (although you’ve pos- the keyboard (key <B>). Each ten to isolate the bit as selected
sible just shrugged your shoul- active bit has an associated in the screen box and to detect
ders over it and thought that’s AND value stated – each value each time it changes state during
the way it is), is that bit 7 is al- is one of the powers of decimal a period of one second, dividing
ways set at the opposite logic to 2: the answer by two to find out how
what you expect through con- many times the bit has been high
When ANDing one binary
necting signals to IN4 on the in that period. This result is dis-
number with another, individual
board. Well, you’re right, that is played on screen as the
bits in the answer will only be at
the way it is, bit 7 is inverted by (approximate) frequency at which
logic 1 if the same bit in both
the computer. starting values is also at logic 1. your oscillator is running.
Underneath the upper If either or both bits are at logic The internationally agreed
boxes is the correction formula 0, so the same bit in the answer unit of frequency is the Hertz, ab-
used to change the value di- will also be at logic 0. For ex- breviated to Hz, as shown on the
rectly received by the port regis- ample, to test for bit 4 being screen. (Strictly speaking, Hertz
ter (box 1) to that in box 2, Cor- high in a register that probably in this context should be written
rected Input Byte. The meaning has other bits set as well (bits 3 with a lower-case initial – hertz.)
will become clear after Binary and 6 in this example): In the present example we have a
Logic has been discussed in frequency of 1Hz (one cycle per
Part 6. Basically, all that is be- Register value = 01011000
second. Back in history, fre-
ing done is to re-invert bit 7 and (decimal 88)
quency was actually defined in
shift all the bits along to the AND value = 00010000 cycles per second – CPS or C/S).
right by three places. This al- (decimal 16)
TEACH-IN 2000
t
PULSE DISPLAY (A) TRIANGLE
HARMONICS
t
In the Pulse Input Wave- t
form Display program (call it up We would now like to illus- (B)
POSITIVE-GOING
RAMP
on screen), AND is again used trate a problem that can occur

to isolate the selected bit. In this
t
when monitoring a waveform, (C)
NEGATIVE-GOING
case, though, the result causes and one which can also affect
RAMP
the drawing of a horizontal line the correct response when digi- t
to represent the bit’s status, tal electronic circuits are used in

(D) SINEWAVE
high or low. The “vertical” line is real-life situations – that of rela-

t
t
simply drawn by the program at tive rates of response between (E) SQUARE WAVE
the detected transition between one circuit and another. The ef- t
logic levels. fectiveness of the illustration on t
REGULAR
Having completed one screen, though, rests on how (F) PULSE
crossing of the screen, the fast your computer is running. If

t
waveform re-commences on the it runs too slowly, it may be diffi- IRREGULAR

(G)
next allocated path, erasing any cult for the required effect to be PULSE
previous waveform. The rate of produced. Nonetheless, let’s try

transition is selectable via the it. (H)
COMPLEX
Display Step option (note that First, set your oscillator to a
this does not actually change slow speed so that the individ- Fig.5.1. Examples of wave-
the rate at which the register ual pulses (as square waves – form types. Frequency mea-
input is examined – sampled – it equal length highs and lows) are surements can be assessed
just changes the distance the clearly seen and with wide spac-
line travels for each sample). It for periods arrowed for (A) to
ing. It is probably best to select (F), but cannot be meaning-
is not possible to assess fre- a Display Step value of 10 (use
quency from this display (a true fully measured for irregular
the <+> key).
oscilloscope would allow this to
be done, however). Slowly increase the oscilla-
tor’s rate and watch the pulses guish between them.
The facilities offered by this close up to each other (you may Continue to increase the
display and the previous pulse need to change capacitor C1 to frequency – what you should
counting program can be of a smaller value). Eventually, the see next (your computer’s sam-
great use when testing other cir- pulses will be so close together pling speed permitting) is that
cuits in the future. that you can probably not distin- as you increase the frequency,
the waveform spaces start to
widen, probably with uneven
spacing (see Photo 5.2). Why?
The answer is not that your
oscillator has decided to run er-
ratically or more slowly all of a
sudden. What is happening is
that the computer now does not
have time to respond to each
pulse. So what you are seeing
on screen is a sub-harmonic of
the actual frequency being mon-
itored. The unevenness of the
spacing is due to the computer’s
sampling rate and the speed of
the oscillator not being synchro-
nized.
Photo 5.2. Example of “under-sampled” pulse If you try this with the Fre-
quency Counter program, you
waveforms into the computer via the simple
will see the apparent frequency
interface board.
TEACH-IN 2000
rate change downwards as the
transition to sub-harmonic sam-
pling occurs.
It is even possible that as
the oscillator’s frequency is in-
creased still further, sampling
may occur on every third pulse,
or every fourth pulse, and so on,
almost indefinitely. It is unlikely,
though, that our demonstration
program will actually allow you
to go beyond one-in-three.
This situation is one that
you have to consider very care-
fully when using a frequency
counter or oscilloscope (whether
as true items of test equipment,
or as computer-based sampling
programs). It is also something
to be considered when you start Fig.5.3. Relationships between upper and lower sections for a
designing digital logic circuits – uniformly alternating waveform.
can all the digital integrated cir-
cuits keep pace with the rat Examples of seven different Press the up-arrow key so
e at which others expect the types of waveform will be seen that the orange background is
m to? Highly unpredictable re- on screen (see Photo 5.3 and on the square wave.
sults can occur if they can’t! Fig.5.1) as follows: Note the point at which any
Not only does the problem o) Triangle one of the pulses first becomes
reveal itself in digital electron- high and the point at which it
o) +VE Ramp (positive-going
ics, but you can also experience ends being low (and the next
ramp – upwards)
allied situations when dealing pulse commences). You will see
with analog signals – a signal at o) –VE Ramp (negative-going that the square wave has a
one frequency adversely (or de- ramp – downwards) white box around it to indicate
sirably in some cases) affecting o) Sine this period. In Fig.5.1 it is repre-
a signal at another. o) Square sented by the “t” arrows.
In a future part we shall o) R-Pulse (regular pulse) This period is known as a
show you how two analog sig- o) Complex cycle. On screen you will see
nals respond to each other. that it takes just one second.
In fact an eighth waveform When a waveform consists of
is available via the screen – many equal length cycles its fre-
WAVEFORM TYPES press the down arrow key three quency is defined as the num-
times. Where R-Pulse was will
DISPLAY now become I-Pulse (Irregular
ber of cycles that occur in one
second, which, as we said a few
Put your breadboard as- Pulse) and the pulses will be paragraphs back, is expressed
sembly to one side for the mo- seen to be very irregular. in Hertz.
ment (you’ll need it again for the
Experimental article, though). In the right-hand box you
Let’s show you examples of SCALING AND CAL- will see that the time (t) for one
some different waveform types. CULATIONS pulse and its consequent fre-
We shall also explain how you quency (f) is stated:
can do some simple calcula- The screen display is
scaled, at present to represent a t = 1 secs, f = 1Hz
tions in respect of frequency
and time. From the main menu period of 10 seconds from the
start of any waveform to its end. Press key <X> to change
select Frequency and Time the screen’s scale to 1 second.
Vertical dotted lines indicate the
(well, what else?!). Each vertically dotted interval is
1-second interval points.
TEACH-IN 2000
now 0××1 seconds. The frequency be at different amplitudes. To crosses any point on the wave-
represented is therefore 10 cycles separate each frequency and its forms. Timing measurements
per second – 10Hz. The right- amplitude out from the main can then be taken between any
hand box confirms this. waveform requires very com- two or more points where the
You/we have been able to do plex equipment (such as a fre- horizontal and waveform lines
the conversion from time to frequency analyzer – expensive!). cross.
quency in our heads, almost in-
stinctively. As shown in the box, WELL-DEFINED SELF-TEST
though, there is a formula for it:
It should also be obvious For a bit of timely entertain-
f = 1/t that even with a well-defined ment, we’ve added a Self-Test
waveform, you need to be able option to test your use of time
Conversely, if you want to to time the complete period of a and frequency calculations –
find out the cycle period (time) of cycle. It does not matter, press <S> to enter or exit it.
a waveform whose frequency you though, at which point you actu-
know, the formula becomes: ally start the timing.
With the square wave you
CYCLE POWER
t = 1/f
can start timing as the pulse In Part 3 we explained that
Press <F> and this formula rises (at its leading edge) and the amount of current con-
will be shown, together with the end at the start of the leading sumed by a circuit can be de-
calculated answer. edge of the second pulse. Alter- fined according to the voltage
Press any of the <+ – * /> natively, you can start as the applied across it and the resis-
keys to expand or contract the pulse falls (at its trailing edge) tance through which it flows, I =
waveform, and variously swap and end as the next pulse’s V/R. When the voltage is at a
between the two formulae using trailing edge begins to fall. The nice steady DC level and the
<F>. The units in which f and t white outline box on the wave- resistance is constant, the cur-
are expressed are changeable by form illustrates this point when rent drawn is the same at all in-
using <U>. you set it to lower frequencies stants of measurement.
These formulae can be ap- (also see Fig.5.1). If we want to calculate cur-
plied to any regular waveform, When looking at a wave- rent (or any of the factors de-
including the first six on screen form on a screen (whether it is a fined in Ohm’s Law and its
(and in Fig.5.1). With irregular computer screen as in this sim- derivatives – Part 3) in relation
waveforms, such as those for I- ulation, or an oscilloscope to a changing voltage (e.g.
Pulse and Complex, periodic time screen), the commencement of waveform) rather than steady
and frequency for the waveform a cycle does not necessarily oc- conditions, the situation be-
as whole cannot be ascertained. cur where you first see the comes more complex.
All that can be assessed from waveform appear; the cycle For any particular instant,
the irregular pulse, for example, may have started long before we can of course say that condi-
is the period of any individual the screen responded to it. Al- tions are constant at that in-
pulse. In terms of overall fre- ways take a measurement when stant, and calculate accordingly.
quency of the waveform, this an- it is obvious where the exact However, we often need to
swer is meaningless. start of the cycle occurs. Our know the current drawn over a
simulation program, you will no- period of time, rather than in-
It is possible, of course, to tice, takes this into account. stantaneously.
use a frequency counter with this
irregular pulse chain and deter- Examine the position of the With a square wave or reg-
mine the number of pulses that white box for all the waveforms ular pulse waveform, average
occur in any 1 second, but the that can display it and observe current can be calculated ac-
answer for each 1-second sam- at which points of a cycle mea- cording to the period for which
pling period is likely to be differ- surements may be taken. the pulse is high (full current
ent. It is worth noting that most flow) relative to the period for
oscilloscopes have a facility for which it is low (minimum current
With the complex waveform,
placing a horizontal line across flow).
there are in fact a great many dif-
ferent frequencies involved in its the screen. This can be moved When the pulse is changing
make up, but they are all likely to up or down and placed so that it between 0V and a known volt-
TEACH-IN 2000
D C
age value, it’s just a matter of current per unit of time in re-
ratios of the on to off (high to (A) spect of a triangular waveform
low) periods of the pulse. is the same: multiply the height
A B (voltage or current) by the
For example, if a 0V to +5V AREA OF RECTANGLE ABCD = AB x BC OR DC x AD length (time) and divide by two,
pulse is high for 0××5 seconds AREA OF TRIANGLE ABC = AREA OF TRIANGLE ADC
= (AB x BC)/2 OR (AD x DC)/2
e.g.:
and low for 0××5 seconds (i.e. a C
square wave), we can say that Iav = Imax x Tcycle/2, or:

the average current drawn in (B)
Iav = (Imax/2) per cycle time
one second is half that which
would be drawn if the maximum A
B
E
Suppose a peak current of
current flowed for 1××0 seconds. AREA OF TRIANGLE AEC = ((AB x BC)/2) + ((BE x BC)/2)
OR (AE x BC)/2
100mA is drawn when power is
supplied by a triangular wave-
If the on-off periods are not
VE form alternating between 0V
equal, the formula used is just:
(C) VPK and +5V and the cycle period is
Average Current = On Period/ 1 second. The minimum current
Cycle Period x Maximum Cur- T
0V
drawn is zero. Therefore the av-
rent. AVERAGE VOLTAGE FOR PERIOD T = VPK/2 erage current drawn is simply
SIMILARLY AVERAGE CURRENT FOR PERIOD T = IPK/2
100mA/2 = 50mA per second.
For the above example, and Photo 5.2. Example of
If the waveform is alternat-
assuming a continuous maxi- “under-sampled” pulse wave- ing between two other levels, a
mum current of 2 amps, this re- forms into the computer via similar principle applies, but re-
lationship becomes: the simple interface board. quires just a bit of extra calcula-
Iaverage = Ton/Tcycle x Imax = 0×5 sec/ tion. Suppose, for example, that
amount of time, no matter how the waveform causes the cur-
1 sec x 2 amps = 1 amp per
small, see Part 6 next month.) rent to alternate between
second.
The maximum peak current is 100mA maximum and 20mA
a predictable or measurable minimum:
TRIANGULATION value (that when the voltage is at First take the difference be-
For waveforms such as tri- its maximum), so too is the mini- tween the two extremes: 100mA
angle and ramp, the calcula- mum current (that when the volt- – 20mA = 80mA. The average
tions are just as simple, irre- age is at 0V, i.e. a current of zero of the current difference for this
spective of their relative rise amps). waveform is 80mA/2 = 40mA.
and fall times. Consequently, when a trian- But there is the “standing” cur-
By definition, the voltage of gular waveform is alternating be- rent of 20mA, which has to be
triangle waveforms rises at one tween 0V and a given maximum added to the average differ-
constant rate, and falls at an- voltage, the average current ence. Thus the overall average
other constant rate. This is true drawn during one cycle is half current drawn = 40mA + 20mA
irrespective of the relative rates that of the peak current. = 60mA.
of rise and fall for the wave- Do you remember from What happens, though, if
form. school days how you found the the waveform is evenly alternat-
In the case of a uniformly- area of a triangle having sides of ing between –6V and +6V, for
shaped triangle (isosceles – different lengths? (See Fig.5.2.) example? Instinct might just
having two equal sides), rise The principle for finding average suggest that if 100mA is drawn
and fall periods are equal in VE
at +6V, then –100mA is drawn
length. Note that a ramp wave- VPK (1) at –6V, thus the average current
form is a special case of trian- 0V drawn is zero. But, you might
gle, in which the vertical edge
VPK (2)
wonder, doesn’t that mean the
VE
occurs instantaneously, making T battery would never run down?
the ramp duration the same as VPK (1) = VPK (2)
Sadly, not so…
AVERAGE FOR PERIOD T = VPK (1)/2
the cycle period. (In fact an in- It’s certainly true to say that
stantaneous change in level
Fig.5.3. Relationships be-
the average current or voltage
never occurs in electronics – tween upper and lower sec- value of a waveform which al-
every change takes some tions for a uniformly alternat- ternates symmetrically above
TEACH-IN 2000
and below a zero value will be CONST pi = 3.141592653589# / VE
zero if measured over a long 180 PK x 0 707

PK x 0 636 VPK VPK PK
enough period, although this volts = 1
VAV VRMS
0V TIME
does not mean that the power totalvolts = 0
consumed is zero.
FOR angle = 0 TO 180
When it comes to express- totalvolts = totalvolts +
VE
ing an average in relation to a (SIN(angle * pi)* volts) Fig.5.4. Definition of the four
triangular waveform uniformly value relationships for a pure
NEXT angle
alternating above and below a sine wave.
zero point, it is usually taken as averagevolts = totalvolts / 180
the average over just half of a PRINT averagevolts relating to just one half
cycle, not over the full cycle. of the cycle. Since a sine
The answer now will be
The half cycle can be either on wave is symmetrical
0××6366036V (approximately),
the negative or positive side of above and below 0V,
say 0××636V. Vpk is the same what-
the zero midway level. The ef-
fective answer is same. (See ever half (positive or
Fig.5.3.) SINE WAVES negative) of the wave-
However, what we can say form is measured.
Since sine waves are obvi-
with respect to both sides of the ously subject to a different set Vav – as said in the previous
waveform is that it has a peak- of rules to triangle waves, let’s section, a sine wave’s
to-peak value twice that of its examine them a bit more average voltage is that
peak value. The peak value in closely. From the main menu taken over one complete
the previous example is 100mA, select Sine Wave Value Rela- half-cycle. We showed
therefore its peak-to-peak value that it has a value 0××636
tionships. Photo 5.4 shows what
is 200mA. you should see displayed. times that of the peak
voltage.
The relationships we now
WAVING PROOF discuss are specifically those Vpk-pk – the peak-to-peak volt-
with regard to pure sine waves. age is obviously twice
If your computer has Quick- that of the peak voltage
BASIC or QBasic installed, you They do not apply to any other
type of waveform. Furthermore, (Vpk).
can prove for yourself that for a
triangle wave the average volt- if you see any waveform rela- Vrms – well, first we had better
age or current for one side of its tionships referred to anywhere explain the concept of
ramp is half the peak value, us- and the shape of the waveform RMS:
ing the following BASIC routine: is not actually stated, then the
relationships are assumed to
ramplength = 10: ’ (seconds) refer to sinusoidal conditions. RMS VALUES
totalvolts = 0 Looking at your screen dis- You have discovered in ear-
FOR volts = 0 TO ramplength play (and Fig.5.4), you will see lier parts that when a current
STEP 1 the representation of a pure flows through a resistance,
sine wave swinging symmetri- power is dissipated. We went on
totalvolts = totalvolts + volts
cally above and below a zero say that heat is generated by
NEXT volts that dissipation. For a fixed re-
level and plotted horizontally
averagevolts = totalvolts / ram- sistance, and steady (DC) volt-
with respect to time. Four verti-
plength age or current, the heat gener-
cal arrows indicate four major
PRINT averagevolts relationships exhibited by a si- ated is calculable – often ex-
nusoidal voltage waveform: pressed in watts (see Part 1). In
You will get an answer of the Ohm’s Law section (Part 3)
5V. o) Vpk peak voltage you saw that power can be cal-
Now do the same for a half o) Vav average voltage culated in several ways, such as
2 2
a sine wave (having an angular o) Vpk-pk peak to peak P = I x R, P = V /R, and P = I x
change of 0° to 180°) and voltage V.
whose peak voltage is 1V. Re- o) Vrms r.m.s. voltage An RMS (root of the mean
member that the sine of 0° is 0 square) value states the effec-
and that the sine of 90° is 1. Vpk – the peak voltage is that tive value of alternating current
TEACH-IN 2000
or voltage in terms of the heat AV = Pk x 0××636 and Pk = RMS x 1××414). The scaling of the values is
that will be generated through a changeable using <X>.
resistance. By definition, the
You will find this facility to be of enormous value when you de-
value is that which will produce
sign your own circuits in the future (or wish to analyze those of other
the same amount of heat in the
designers).
resistance as would a direct
voltage or current of the
same magnitude. SELF-TEST
Any waveform shape can Finally for this month’s Tutorial – through the Self-Test option the
be related to an RMS value, but computer randomly selects sine wave related questions. You are ex-
the value is really only mean- pected to use your calculator, and then compare your answer to the
ingful when referred to a known computer’s, which will be displayed when you press <A>. You can
shape. For a sinusoidal wave- cheat a bit if you want by changing the values offered by the com-
form the RMS value is 1××414 puter!
times the peak value.
And thinking of changing values, do please read and digest our
“measured observations” discussion in Panel 5.1 following the Exper-
SINE WAVE RELA- imental section.
TIONSHIP
The four sine wave units of EXPERIMENTAL
measurement listed in the previ- In this month’s Experimental section we show you how to add an-
ous section have their relation- other circuit to the interface board. This allows you to actually view
ships tabulated at the bottom of on screen the various waveforms you’ve been generating with the
the main screen box displayed. variable mark-space oscillator of Fig.4.3 last month,
The screen display relation- These include triangle, rising ramp, falling ramp, square wave,
ships are interactive (although and regular pulse. Sine waves, irregular pulses and complex wave-
the waveform itself is not). Each forms we shall illustrate via your breadboard later in the
formula can be selected using Teach-In series.
the four keyboard arrows. Vari-
ous values can also be set
through the smaller left-hand NEXT MONTH
box. Thus you can have the In Part 6 next month, we return to Binary Logic (briefly discussed
computer calculate relationship when the Parallel Port Data Display/Set program was described).
answers in respect of values of The discussion will include Digital Logic Gates – OR, NOR, XOR,
your choosing. XNOR, AND, NAND. As usual, the screen displays will be interactive.
On entry to the screen you
will see a highlight on Pk x
0××636, allowing you to find the
average value (AV = ) when the
peak value is known. Referring
TEACH-IN 2000
to the values box, Pk is shown
as 1V. At the top of the main
EXPERIMENTAL 5
box is shown the answer to the ANALOG-TO-DIGITAL CONVERTER
calculation, in this case the
same answer that we illustrated
We are now going to describe a very simple circuit that you can
earlier with the BASIC sine
assemble on your breadboard, and which will allow you to use your
wave calculation example.
computer to view various analog waveforms, such as the triangle
We suggest you experiment and ramps discussed in this month’s Tutorial. It is capable of display-
with the different formulae and ing other waveforms as well.
values of your choice (the op-
However, we won’t try to deceive you – the circuit is highly use-
tions vary depending on the for-
ful, but it’s also very limited!
mula). Try to memorize some of
the formula (two useful ones are Commercial equipment such as oscilloscopes, or commercial
TEACH-IN 2000
software for computer-based
virtual ’scopes will do far more
than our simple demo software.
The author’s Virtual Scope pro-
ject of EPE Jan-Feb ’98 is also
an extremely sophisticated item
of computer-based test gear,
but it has to be said that the
construction of its complex
printed circuit boards is not
suited to novices.
The intention of the Teach-
In analog interface is basically
to let you view the various
waveforms that you create on
your breadboard. However, it
can be used not only with the
demo circuits we describe
throughout this Teach-In series,
but it will be of use with some of Photo 5.6 Typical analogue waveform as demonstrated
your own future designs.
using the complete interface circuit
Should we have space later
ANALOGUE TO DIGITAL CONVERTER
in the series, we’ll describe the PIN COMPUTER
6V BLOCK CONNECTOR
sort of facilities you should ex-
TB1 SK1
pect from a fully-fledged oscillo- 1 8
scope (and which need not be REF VE
SIGNAL 2 IC2 7 1 OUT 0 D0
very expensive). INPUT VIN TLC549 CLK 1
3 6 12 IN 1 SELECT
REF D OUT 13
4 5 2 OUT 1 D1
GND CS 2
ANALOG INTERFACE
A very simple analog-to- 0V
14 GND GND
16
digital converter (ADC) inte- (BATTERY VE)
grated circuit (IC) is the only ac-
tive component in this part of Fig.5.5 Circuit diagram for the serial analog-to-digital con-
your interface assembly. The verter interface.
circuit diagram for its connec-
tions is shown in Fig.5.5.
We shall discuss the nature
of the ADC (IC2) later on in this
article. In the meantime, we
want you to assemble the circuit
on the breadboard and have a
look at some waveforms on
your screen.
The breadboard layout for
the ADC (and the computer in-
terface from last month) is
shown in Fig.5.6. Assemble the
components into the bread-
board, following the latter’s
numbered holes.
The oscillator should be
Photo 5.5. Detail of the interface board with the ADC included.
same as that referred to at the
start of this month’s Tutorial (i.e.
TEACH-IN 2000
with diodes D2 and D3 in- log sampling in a moment. converts the voltages to equiva-
cluded). Use a 100uF capacitor Also at the top right of the lent binary numbers. You saw in
for C1 and adjust preset VR1 to screen is the single word DI- Part 4 the representation of a
a midway position. RECT. This means that data is number in both binary and deci-
Connect the resistor/capaci- being sampled in real-time mal (Parallel Port Data Display/
tor junction on the oscillator (here and now). Set).
(IC1a pin 1) to the ADC at its By pressing <C> the word The ADC is capable of
input pin 2 (Signal Input as CAPTURE appears instead. In “reading” the voltage level at
shown in the breadboard lay- this mode, the computer sam- many thousand times per sec-
out). Also connect the ADC’s ples the data as fast as it can ond and is controlled by logic
data output pin 6 (D OUT as and temporarily stores it in level signals from the computer
shown in Fig.5.5) to IN1 on the memory. Once a full set of sam- (or other device in other appli-
computer interface part of the ples (640) has been received, cations). The ADC used in this
board. Crocodile-clipped links the computer then plots them as Teach-In demo is a serial ADC,
will do in both cases. a screened waveform. which means that its output data
Power up the board and run is read one bit at a time.
This technique allows
the Analog Input Waveform Dis- higher speed waveforms to be Parallel ADCs also exist, in
play. On entry to the display, a sampled than does the Direct which the 8-bit data is read as a
yellow line should be seen method. The drawback is that single 8-bit byte. Parallel ADCs
traversing the screen from left there is a pause between each are much faster to read than
to right, its vertical position full screen change of data serial types, but require more
moving up and down. Adjust (computer speed dependent). computer control and data lines
VR1 (or change the value of than we have (readily) available
C1) until the moving line begins Experiment with different for your Teach-In breadboard.
to show several cycles of a oscillator frequency rates and
waveforms, using preset poten- To start each voltage level
triangular-like waveform, as you sampling and digital conversion,
saw demonstrated in the simu- tiometer VR1, and different val-
ues for capacitor C1. Also find the computer sets the ADC’s CS
lation program Frequency and input (chip select) high via out-
Time, see Photo 5.6. out how Direct/Capture and Dis-
play Rate settings have their put data line D1. This action
Should the yellow line just benefits. Note that each time causes the ADC to “read” the
be sitting towards the bottom of the Display Rate is changed, voltage present on its signal in-
the screen, check that the Port the waveform line restarts from put (Vin) at that moment.
register selected is still the cor- the left of the screen. The ADC has an internal
rect one (as discussed in Part high-speed oscillator that then
4). Also double-check that your Also see whether you can
replicate some of the stranger controls the data conversion
ADC’s breadboard assembly process. (Incidentally, chip is a
corresponds to the connections waveforms you have seen
through the ADC-Demo pro- term frequently encountered in
shown in Fig.5.6, and that the electronics and is colloquially
oscillator is still correctly assem- gram. (It has to be said that
those of you with higher-speed used to mean any integrated
bled. circuit device.)
computers will fair more easily
Towards the top right of the since the waveforms will be The result of the conversion
screen is shown the Display traced faster on the screen.) is a binary number between
Rate, 10 at the moment. Using 00000000 and 11111111
Just for interest, try con-
the <+> and <–> keys, this can (decimal 0 to 255). A conver-
necting the oscillator’s digital
be set to any value between 1 sion value of zero results from
output (F OUT, IC1a pin 2) to
and 10 and controls the rate at an input voltage of zero. The
the ADC instead of the analog
which the waveform line maximum conversion value of
waveform.
crosses the screen. The slower 255 occurs (in this breadboard
the rate, so more waveform cy- assembly) when the input is at
cles are shown. SERIAL ADC DEVICE the same voltage level as the
Note that changing this rate As the Analog-to-Digital ADC’s power supply. This con-
also changes the rate at which Converter’s name states, it al- version range is determined by
data is acquired (sampled) from lows an analog signal (e.g. volt- the voltage levels to which the
the ADC. We shall discuss ana- age waveform) to be input, and ADC’s +REF and –REF pins are
TEACH-IN 2000
connected (6V and 0V in this
case).
In other applications, the
pins may be connected to other
reference voltages to provide a
different conversion range. The
reference voltages must lie at or
between the power supply volt-
ages. (Note that the maximum
voltage at which this ADC can
be powered safely is +6××5V.)
READING BITS
Once conversion is com-
plete, the binary data can be
read bit-by-bit by the computer.
It is read in order of bit 7 to bit 0
of the binary conversion value. Photo 5.7. Analog-digital-analog demonstration screen. You
(In theory, about 40,000 conver- can observe each bit being “assembled” to a binary number,
sion and data-read cycles per and change the sampling step rate.
second can take place – but not
does whatever it has been told ADC DEMO
with this Teach-In demo!)
to do with the data, in our case
To read each bit, the com- it either stores it or draws a Run the Analog-Digital-
puter takes the ADC’s CLK screen line in relation to it. After Analog Demo program from the
(clock) pin high via output line which the next sample can be main menu to see the animated
D0 (OUT0). The data is then taken. principle of how the ADC con-
read from the ADC’s Dout pin version is output from your
As an example of the com-
via the breadboard’s IN1 path – breadboard to the computer
puter’s data acquisition routine,
computer parallel port register (see Photo 5.7).
take the conversion value of
bit 4. The left-hand box (Input)
binary 10010111 (decimal 151).
The data on register bit 4 The computer first sets its data shows a single sine wave, cy-
will either be at logic 1 or logic storage value to 0 (00000000). cling between 0V and +5V. It is
0. As discussed when we de- The reading routine results in shown feeding into the serial
scribed the data input process in the following binary values of Analog-to-Digital converter,
Part 4, the value of the bit is the chip and the assembled where the present voltage level
isolated from the other register data storage: is given, together with the ADC
bits (using an AND command) conversion value in both deci-
The computer’s assembled
and set into the rightmost (bit 0) mal and binary. For the sake of
value at step 8 is the same as
position of the 8-bit binary value this demo, the ADC is assumed
that of the original ADC conver-
being assembled by the com- to be referenced so as to gener-
sion value.
puter. Between each bit, the ate an output of 0 for 0V and
value is multiplied by two to 255 for +5V.
shift all the bits left by one ADC Assembled The ADC is shown con-
Step
place. Value Value nected to the computer. At pre-
The computer then sets the 0 10010111 00000000 sent, the immediate ADC binary
ADC’s CLK line low, causing it 1 00101110 00000001 value is repeated as the com-
to set the next bit of its binary 2 01011100 00000010 puter’s received value. The
conversion onto the output pin. 3 10111000 00000100 computer then feeds this value
Taking the CLK line high again 4 01110000 00001001 into a Digital-to-Analog Con-
allows the computer to now read 5 11100000 00010010 verter (DAC). DAC devices will
this bit, and so on for all eight be discussed another time – this
6 11000000 00100101
bits. one converts the 8-bit binary
7 10000000 01001011
At this point, the computer into an equivalent output volt-
8 00000000 10010111
TEACH-IN 2000
Photos 5.9 and 5.10. Exam-

ples of the screen demo set
for different waveform and
sample rates.
Photo 5.8. Another example of the ADA demo screen, this
input waveform, the output
time showing (right) the result of using fewer sampling waveform makes the same
step, i.e. the sampling rate is
age, in this example having the tended to represent the behav-
just right in order to replicate the
same scale as the ADC. ior of any particular serial ADC
original.
Also at present, the DAC device, it is a very generalized
interpretation. However, the output wave-
shows the same values as the
form will become an imperfect
ADC, and the resulting wave- To terminate the bit-shifting
copy of the input if its sampling
form is displayed in the right- process, press <B> again.
rate is reduced – press the <*>
hand box (Output).
key once, to set a sampling rate
To see how the computer SAMPLING RATES to half that of what it was
reads in the ADC’s binary con- (Sample 2). Now observe the
version value bit, press <B>. While discussing pulse fre- extra jaggedness of the output
quency counting in Part 4, we shape.
Note how the left-hand ADC
referred to the problems of sam-
bit (bit 7) drops down and Press <*> a few more times
pling data at rates slower than
moves right to insert itself into and observe the increasing dis-
ideal. We can now illustrate one
bit 0 of the computer’s storage tortion. What is happening is
way in which the problem mani-
value (which starts off at zero that the output waveform is be-
fests itself.
for each cycle of eight bits). ing “traced” horizontally across
With binary step sampling the screen at a given rate. The
Note also how the ADC and
off, first note the shape of the movement represents the Time
storage values shift left by one
two sine waves (Waveform axis we discussed in the
binary place during the copying
Rate 1, Sample 1, as stated in Resistor-Capacitor Charging
process, with the previous bit 7
the bottom text line). Graph of Part 2, and in the
being “lost”. A zero value enters
the ADC value at bit 0 at each Although you will notice that waveform period timings discus-
shift left. the waveforms have slightly sion in this month’s Tutorial.
rough edges, they are as close The sampling rate, though,
When all eight bits have
to sine waves as we can get is not fast enough to keep pace
been input to the computer, the
with the program that creates with the trace, which continues
final byte value is copied to the
the display you are watching. to show the vertical value last
DAC, and the process starts
Just for background interest – sampled until the next one is
again by the ADC taking an-
the waveforms are plotted as taken. The steps become even
other sample, with the Input
180 steps (pixels) across the more pronounced the more you
waveform having shifted slightly
screen, and 140 steps vertically press <*>, and the less recog-
to the right.
between top and bottom wave- nizable the output waveform
Please be aware that the form extremes.
process illustrated is not in- becomes.
For every step made by the
TEACH-IN 2000
pling, in this case in the com-
PANEL 5.1 – MEASURED OBSERVATIONS puter simulation program, where
You might think, perhaps, that in order to know the true state of the sine wave is calculated ac-
affairs regarding component values, timings, voltages and currents cording to an angle count being
etc., all you need to do is take some measurements. True, measure- incremented – small increments
ments of anything that happens in electronics can be taken, though for slow waveforms, greater for
for some of them extremely sophisticated equipment is needed. higher rates. Nonetheless, het-
But, and this is a big “But”, no measurement can be taken instan- erodyning is a very real effect;
taneously, it’s spread over a “window in time”. Consequently, the undesirable in some cases
measurement does not show the actual state of the condition being (audio sampling for example),
measured at a specific point in time, it simply shows an averaging- but highly beneficial in others
out of what might be numerous values occurring within the period of (e.g. radio reception). The word
measurement. superhet is derived from the
term, although the full term in
Furthermore, the value reading shown on the measuring instru- the radio reception context is
ment only reveals the value to within so many decimal places, or actually supersonic heterodyne
within an estimated fractional distance from a marker on a scale. – a superhet receiver mixes the
Fortunately, in many electronic circumstances, only a close approxi- incoming radio frequency with a
mation to the actual value is needed, although there are instances local oscillator frequency, to
when greater precision would be desirable. produce a specific (fixed) inter-
There is a further important factor which affects the accuracy of mediate frequency (IF) from the
measuring some aspects of electronic circuits – the measuring instru- which the desired signal can be
ment itself can affect the characteristics of the circuit point being more easily amplified and ex-
measured; significantly so in some instances, and sometimes it is tracted.
necessary to take ingenious steps to try to circumvent the problem,
but even then a degree of interference will still exist. Electronics is
riddled with practical examples of the Principle of Uncertainty! DIVERSION END
After that diversion, and
– with Sample = 1), you will no- coming back to the ADC sub-
RATEABLE LESSON tice that another waveform ap- ject, you should now have a
pears to be superimposed on greater understanding of what it
The lesson we hope you will the top and bottom of the over- can achieve and what its fail-
take in from this is that when all input screen display. This ings can be. In addition to the
sampling an electronic signal, illustrates another by-product of subjects listed at the end of the
whether it’s digital or analog, the how two frequencies can react Tutorial, next month we shall
sampling rate should be care- with (modulate) each other. examine Sampling in greater
fully chosen to obtain the opti- detail. Before then, see what
mum results. Too slow a rate is The result is two new fre-
quencies, one that is the sum of you can discover for yourself
obviously undesirable. A further about the subject with the aid of
example of sampling is dis- the originals and one that is the
difference between them. The your digital and analog wave-
cussed next month when we il- form sampling displays.
lustrate Logic Gates (program effect is known as heterodyning
Digital Sampling and Logic (from the Greek heteros, mean- Till next month, the author
Demo is the one we shall dis- ing “other”, and dynamis, mean- waves goodbye and exits
cuss in this context – see the ing “strength”). screen-right!
main menu). Suppose, for example the
Play around with other input two base frequencies are 2kHz CORRECTION
Waveform Rates (<*> and </>) and 3kHz, the sum of the fre-
quencies is 5kHz, and the differ- Part 4, Feb. ‘00, Fig.4.6.
and Sample values (<+> or <– SK1 pin 23 should read pin 32.
>) and see what you observe. ence is 1kHz. These two new
frequencies replace the original SK1 pin 16 should read pin 23.
You’ll come across some quite The PCB is correct.
remarkable output waveforms. base frequencies. It should be
pointed out that the reason for
In some cases, especially at the Input display also showing
higher input waveform rates heterodyning is because it too is
(25, 26, 44 and 59 for example actually created through sam-
SURFING THE INTERNET
By Alan Winstanley
Radio Bygones Message server and generally see what web in search of links, which
Board the magazine is about. Ameri- would then be cataloged and
can and Canadian readers are added to the enormous Al-
Welcome to our monthly welcome to have a look at taVista database back home.
column related to Internet is- www.radiobygones.com and You could also add your own
sues. On the web site for the an on-line version will hopefully URL manually just to be sure,
print edition of EPE at be produced in the very near and indeed the task of register-
www.epemag.wimborne.co.u future. ing one’s own URL into all the
k you will find indexes listing the relevant search engines – there
projects contained in issues dat- are some 1,500 or more of them
ing back to 1996, and also fur- Search and you will – is now an important element
ther details, including color pho- find... an advert when creating new web sites.
tos, of the constructional pro- If you type in the URL (Apparently only I see the joke
jects which have appeared in www.altavista.digital.com you in AltaVista’s “Add URL” confir-
EPE within the last 18 months will of course be redirected to mation message stating that
or so. Some readers will also the popular search engine of “this URL was retrieved in 4.997
know from the EPE Chat Zone AltaVista, now at seconds and will be added in a
that a full redesign is now being www.altavista.com. Apart from day or two.”)
considered, as the site has a web site refresh, the develop- In January 1998, Compaq
evolved over the past four years ment of this interesting and im- (www.compaq.com) acquired
and is getting ready for a total portant search engine has DEC along with its Alpha micro-
rebuild. A shopping cart system largely gone unnoticed by many processor technology and the
is under trial, as well. We wel- UK users. For the benefit of AltaVista engine, and nearly two
come your suggestions and newcomers, AltaVista was origi- years later in November 1999
feedback by email regarding our nally created by the computer Compaq announced a new line-
Internet presence to webmas- manufacturer Digital Equipment up of “supercomputer” Alpha-
ter@epemag.demon.co.uk Corp. (DEC) as a showcase for based servers and workstations
We have also just opened a their Unix and NT mainframes for 3D, CAD and Internet server
message board system for Ra- and servers. Digital Equipment applications. Compaq hadn’t
dio Bygones readers – check in saw their powerful search en- been idle with AltaVista though,
at (deep breath) gine as a good advert, to and the trusty old search engine
www.epemag.wimborne.co.u demonstrate the power of their was destined for greater things.
k/radiobygones/wwwboard to Alpha processors and servers to Very early last year Compaq
leave messages or to follow up search their enormous database announced the development of
regarding antique radio sets or of the world wide web, and to AltaVista as a separate com-
just share your nostalgic memo- sally forth with the closest pany – and at about the same
ries. Advertisements from the matches to a search enquiry. time it announced that it had
antique radio sector are also Indeed AltaVista has always purchased
welcomed – and they’re free! been a personal favourite, partly www.shopping.com, a very
because it offers Boolean com- popular online retailer in the US.
Still on the subject of Radio mand search options that come
Bygones, we can announce that Then in the middle of last
as second nature to many elec-
we have now obtained the do- year Compaq announced that it
tronics enthusiasts (especially if
main of had sold a majority stakeholding
you followed our series Teach-
www.radiobygones.co.uk and in AltaVista to the Internet busi-
In 98: An Introduction to Digital
a web site related to our sister ness development and manage-
Electronics).
magazine will be open in the ment group CMGI
near future. You will be able to AltaVista’s powerful “spider” (www.cmgi.com), formerly
subscribe online using a secure (a network search and retrieval College Marketing Group Infor-
program) would traverse the mation Services. CMGI will de-
Net Work
velop AltaVista further into what page. 95 included a direct and un-
they hope will become the wanted desktop link to MSN, its
world’s largest portal. fledgling mail service and Inter-
Disappearing URLs net Content Provider. This was
Coupled with the fact that
One potential problem immediately branded a prime
Compaq Internet-ready desktop
seems to be surfacing with Al- example of Microsoft’s own
PCs were to include a ready-
taVista: users are complaining (failed) attempts at Internet
made link to AltaVista, it be-
that their own web site seems to empire-building.
came clear how Compaq was
have disappeared from its list-
starting to embrace the com- Users and manufacturers
ings, and this can be attributed
mercial forces of the Internet complained even more when
to the rebuild at the end of last
and steer business the way of Microsoft’s Internet Explorer
year. Webmasters should do a
AltaVista. Compaq said that browser was being foisted on
quick search for themselves
they would meld their consumer them, to the alleged detriment
(literally) on AltaVista and re-
Presario Internet PCs with of Netscape, yet they seem
submit their URL. This option is
CMGI’s Internet services, by happy for consumers to be
buried in the Advanced Search
providing keyboard and web steered from their home desk-
page (Add/Remove a URL).
browser access to AltaVista and tops towards a portal site which,
Some users have quoted a pe-
other CMGI web offerings. with a bit of luck, will ultimately
riod of up to six weeks before
An updated AltaVista web entice them into breaking out
their URL appears again, which
site was mooted in June ’99. their credit cards in a “seamless
obviously means that they will
The old logo and layout would information and shopping expe-
lose traffic or business opportu-
be replaced by a fresh new rience” as Compaq calls it.
nities during that time.
number in cheerful yellow along Nevertheless, AltaVista re-
For those of you wishing to
with all the usual portal offerings mains a firm favourite as a
ensure that your web pages re-
of news, travel, shopping, jobs search engine, although on my
ceive a higher scoring in search
and so on. The shopping.com own recently redesigned web
engine listings, you should have
site would also be restructured site (http://home-
a look at
and enhanced. pages.tcp.co.uk/~alanwin)
www.searchenginewatch.com.
AltaVista has already This provides details on most of there is a Google search engine
grown into a key portal site, the commonest engines and installed on my “Links” page.
which was rated at the ninth other tips. Google is tremendously fast and
largest domain on the entire In- slick, with none of the portal
It is interesting that there padding of AltaVista.
ternet in early 1999, and conve-
seems to be no objection to
niently accessible from Compaq You can contact me, as al-
Compaq’s eagerness to provide
desktop PCs. A quick look at ways, by email at
a direct link to AltaVista, which
the Compaq Presario web alan@epemag.demon.co.uk
by Compaq’s own admission is
pages (www.compaq.com/
also intended, in turn, to route
mypresario/internetservices/)
consumer Internet traffic to e-
on “How to Search” takes you
commerce sites (to the tune of
directly (surprise) to My Al-
several million hits over Christ-
taVista, where users are encour-
mas 1998 alone). Yet many PC
aged to configure a start-up
users rebelled when Windows
of this circuit – it is the key to its
by ALAN WINSTANLEY and IAN BELL operation. The symmetry is so
important that in order for this
Our surgery writers continue their exploration of circuit to work well the two
operational amplifiers by delving into the innards of transistors must have exactly
typical devices to explain basic opamp principles of the same characteristics, i.e.
operation. they must be matched.
These characteristics must
A few months ago now stages, often with frequency
remain matched all the time –
readers Mohab Refaat and compensation, and finally an
something that, given the high
Tony Soueid inquired about output buffer providing power temperature sensitivity of
the use of one of the most gain to drive external loads, semiconductor
fundamental building VOLTAGE devices, can only
GAIN
blocks of electronics, really be achieved
the operational amplifier + DIFFERENTIAL
OUTPUT OUT if the two
BUFFER
or opamp. Mohab asked – AMPLIFIER (POWER AMP.) transistors are
about choosing opamps physically close
and we explained all the Fig.1. Typical opamp block diagram. together on the
major opamp same piece of
characteristics in the Dec ’99 silicon. Also, integrated circuit
and Jan ’00 issues. but with no voltage gain. All of designers use special layout
these stages are direct- techniques to make sure that
We now go on to address coupled, which means they
Tony’s main point: “I don’t know transistors that should be
are connected without matched do indeed have the
what is inside that ‘black box’… coupling capacitors. Direct
it’s based on a differential pair of same characteristics, despite
coupling means that opamps temperature variations and any
transistors, but it’s far from are able to amplify DC and
being that simple. Can you imperfections in the
very low frequency signals. semiconductor manufacturing
please supply me with some
information?” The circuit diagram for the process.
basic differential pair is shown This would seem to make
This is a big topic, one that in Fig.2a. Notice the symmetry
can and does fill whole life difficult for the hobbyist or
textbooks, but we will try to give + VCC student who is interested in
a brief overview of some experimenting with these
important points. circuits using individual
RC RC components, however it is
possible to purchase matched
IDENTICAL TWINS VO1 VO2
Tony is right to say that TR1 c c TR2 1 8
opamps are based on the Vi1
b b
Vi2
“differential pair”, which we’ll
e e 2 7
look at in detail in a moment,
but first look at Fig.1 which
shows a general block diagram 3 6
of an opamp.
LM394
The circuitry of an IE N.C. 4 TOP VIEW 5 N.C.
operational amplifier often
comprises: a differential input
Fig.2b. Pinout details of
stage, with voltage gain
followed by one or more further, the LM394 supermatch
–VEE
single-ended voltage gain pair (National Semiconduc-
Fig.2a. Basic differential pair.
Circuit Surgery
subsequent
transistors (such as the National current). stages, so having a differential
Semiconductor LM394 If the two input voltages pair as the first stage is a good
“Supermatch pair” in Fig.2b) change together (this is known way of reducing drift.
and some transistor arrays also as a common-mode input
contain differential pairs (e.g. If we change the (still equal)
signal) then the symmetry will input voltages by a large
the CA3086 “npn array”) for just not be disturbed and IE will still
this kind of role. enough amount then the circuit
split equally between the two will cease to function as just
transistors. You may think that described. For example, if we
SINGLE-MINDED changing the input voltage must take the input voltages down to
change the collector and emitter near VEE then the current source
The basic differential pair currents, but it does not have to, may no longer function properly.
(Fig.2a) has two inputs Vi1 and because the emitter voltage is This would determine the
Vi2 and two outputs VO1 and VO2. free to change whereas IE is opamp’s common mode input
Each transistor has a collector fixed by virtue of the constant range. Any lack of matching
resistor RC as a load. Small current source. between the transistors would
differences in the input voltages
The ability of the matched- probably result in some shift in
cause relatively large changes
transistor circuitry to reject output voltage as the inputs
in the output voltages, also in a
signals which are the same on varied together, which would
differential manner (i.e. as one
both inputs (common mode) is manifest itself as a non-ideal
output voltage increases, the
not only important because it common-mode rejection ratio
other will decrease by the same
gives us the function of a (CMRR) for the opamp.
amount). We can, however,
choose to use only one output differential amplifier, but also
because it makes the design of
(referred to as taking a “single-
high-performance, high-gain DC INVERTED VIEW
ended output”).
amplifiers possible. For The high sensitivity of the
Key points to understanding example, if the temperature of a transistor’s collector and emitter
the circuit’s operation are firstly single transistor amplifier currents to base voltage comes
that a transistor’s collector and changes then the bias currents in to play when we make the
emitter current are very change too, and so therefore do voltages at the two inputs
sensitive to changes in its base the circuit voltages. slightly different. This breaks
voltage, and secondly that the the symmetry and causes a
emitters are connected to a In capacitively coupled (i.e.
AC) circuits this does not matter larger proportion of IE to flow in
constant current source. The one transistor than the other.
following discussion assumes because the temperature
that both transistors are changes are slow and are below For example, if we increase
switched on – that is, their base- the cut-off frequency due to the Vi1 slightly and decrease Vi2 by
emitter voltage is greater than coupling capacitor. However, if the same amount, then more of
about 0××6V. there is no capacitive coupling IE will flow in transistor TR1 than
(as in an opamp), any changes in TR2. This will cause TR1’s
The constant current source in voltages due to temperature collector to fall lower than
means that the sum of the two (or other forms of “drift”) are TR2’s, so output VO1 will be
emitter currents must always be effectively indistinguishable lower than VO2. Thus, if we take
equal to IE. If the two base from the required low frequency a single-ended output from the
voltages are equal, and the signals and will be amplified by collector of TR2, Vi1 will act as
transistors are identical, then it subsequent stages. the non-inverting input (+) and
follows that IE will split equally Vi2 will act as the inverting (–)
However, if the temperature
between the two transistors, input. This denotes what effect
of a properly matched
they will draw the same base a signal on either input has on
differential pair changes, both
current, and their collector the polarity of the output:
transistors are affected equally
currents will be equal. As the increase the non-inverting input
and there is no change in the
two collector resistors are equal, and the output effectively
output (the drift is a common
the voltages dropped across increases too. Increasing the
mode signal). The worst place
them will also be equal inverting input at Vi2 will cause
to get drift is in the first stage as
(assuming there is no output the output VO2 to fall (invert).
the error is amplifier by all
Circuit Surgery
PROPORTION OF IE +VCC
IN EACH TRANSISTOR
1.0 R1 R2 R3
OUTPUT VO2
TR2 TR1
OUTPUT VO1
TR1 c c TR2
b b
INPUT Vi1
e e IREF
0.5
INPUT Vi2
IE
c TR3 TR4 c
b b
e e
V i1 – Vi2/mV
–100 –75 –50 –25 0 25 50 75 100 –VEE
Fig.3. Typical characteristics of a basic Fig.4. Basic differential pairs – simple current
differential pair. mirror biasing.
Note that differences over a MIRROR CURRENT optimal performance (transistor
few tens of millivolts (mV) result gain etc. varies with bias
in most of IE flowing in one or A basic differential pair with current). The supply current
other of the two transistors. current mirror biasing is shown required may also be a
Over a large input difference in Fig.4, which will hopefully be consideration when choosing IE.
range the response of the circuit familiar to regular readers. The Although a device such as the
is exponential (see Fig.3), but emitter current can be set using: LM394 Supermatch pair has a
for just a few millivolts IE = (VCC – VEE – VBE(TR4)) / R3,
maximum collector current
difference between the inputs rating of 20mA, National
the change in the output voltage where VBE will be typically guarantees most parameters
difference is near-enough 0××6V to 0××7V. To choose the RC over a range of 1uA to 1mA.
directly proportional to the input collector resistors (R1 and R2)
difference (the central part of for maximum swing, set the
Fig.3). So we have the linear quiescent (“idle”) output voltage
OPAMP SELECTOR
differential amplifier that we to half the positive supply. Thus Table 1 shows a
require for an opamp input the quiescent voltage across the comparison between a number
stage. collector resistors is VCC / 2. of popular opamps. It is by no
To turn Fig.2a into a means comprehensive, as
Since we set IE above and
practical circuit we need a thousands of opamps are
the collector current is
current source, this can also be available, but it will at least
approximately IE / 2, then RC for
achieved using a couple of enable you to compare the
each transistor in the pair can
matched transistors such as the specifications of many well-
be calculated using:
Supermatch pair, although there established types. You can use
are also more sophisticated RC = (VCC / 2) / (IE / 2) the information we have
current sources employing more = VCC / IE. provided in previous issues to
transistors. For a more detailed decipher the meanings of the
discussion of current sources So if the supplies are ±9V data: expressions such as
please refer to Circuit Surgery and we chose a bias current of “Open Loop Gain” and “Slew
May and June 1999 in which we about IE = 1mA then we get R3 Rate” should now be readily
discussed these types of circuit = 18K and R1 = R2 = 9K. understood (we hope).
in depth. For any given transistor IE / The manufacturers’ data
2 should be chosen to give must be consulted for more
Circuit Surgery
design information as needed, can stress a device over time, electrode. It all made sense.
as our figures may often only ultimately leading to some lasting The trouble is, in modern
apply under certain conditions damage. Simply bolt it to a semiconductor electronics we
(supply voltage, temperature generously-sized heatsink and it talk in terms of “conventional
etc.). The World Wide Web will perform fine. Proper thermal current flow”. We all do this
offers for the first time the resistance calculations are the without thinking, but it’s
possibility of readers fetching best way of determining what size extremely bizarre to anyone
data directly from the heatsink to use. ARW. coming into electronics from
manufacturer. It’s usually in other sciences (notably physics
Adobe Acrobat PDF format, and chemistry). Under this
which needs the free Acrobat CONVENTIONAL
convention, electric current is
reader from www.adobe.com CURRENT deemed to flow from positive to
Next month we will look at Do you know why the plate of negative, although in real life it
how gain can be improved by a valve (vacuum tube) is called an flows in the other direction.
using transistors instead of anode whilst the plate on a More than one Physics
resistors as loads, and consider semiconductor diode is called a teacher has torn a strip off me
the problem of getting all the cathode? Most confusing. E.J. for apparently not knowing
bias voltages right when you Bibby. which way current flows in a
cannot isolate stages using circuit: my apologies to Physics
coupling capacitors. IMB. When I first started reading
up on electronics in the early teachers everywhere but
1970’s, my very first text book unfortunately the convention is
HOT REGULATOR started with valves (but hey, I’m now so well entrenched around
not that old!). Their operation was the world that it will never
I built a 1A power supply change. (Can you imagine the
described in terms of real
with a 317-type variable chaos if it did? Which way
electron flow, i.e. what actually
regulator. The data says that it round would you connect your
happened in terms of the physics
is a 2A regulator but when I multimeter?)
of the electron. The simplest
draw 1A, the regulator gets very vacuum tube is the diode, In a semiconductor diode,
hot and the voltage slowly consisting of a cathode (which is conventional current flows in the
drops. Why? a piece of metal warmed up by a direction of the arrowhead
So asked a reader in the heating element) together with an symbol – from anode to
EPE Chat Zone on our web site anode “plate”. cathode. In a silicon diode, the
(www.epemag.wimborne.co.u Electrons boil off the hot anode (a) must be 0××7V more
k) recently. cathode and, being negatively positive than the cathode (k) for
Regulators are usually charged, are attracted towards a “forward current” to flow from
protected against excess the anode, which when positively anode to cathode. However, the
current and thermal overload. It biased will “accept” these anode (electron) current in a
sounds as though you haven’t negative electrons. The current of vacuum tube flows from
heatsinked the device properly electrons which flows through the cathode to anode.
(if at all). It’s imperative that the valve in this way is called the I’m afraid that we have
regulator is allowed to dissipate anode current. history to blame for this
any power efficiently, to prevent By placing an electrode conundrum, but in practice
the chip from overheating. between the cathode and anode everything works fine. After all,
Fortunately, the LM317 – and applying a grid bias voltage we know what we mean, don’t
like many other three terminal to it, the anode current can be we? ARW.
regulators – will suffer no controlled – thus a triode valve is
immediate damage from created which can be used as an
inadequate heatsinking, amplifier. This, together with my
because it will simply current- scant knowledge of Nuffield
limit and gradually shut itself Physics (as my schoolteacher of
down. As you saw, the output the time would testify), meant that
voltage slowly falls during this I started out in electronics
shut-down process. However, knowing that electric current
any repeated cycling like this flowed towards the most positive
V/uS
Input
uVoC
Output
Supply
(GBW)
current
current
ratio dB
rejection
Slew rate
Input bias
Maximum
Maximum
CMRR dB
Open loop
resistance
gain V/mV
Bandwidth
Differential
Max power
diss. (mW)
Input offset
Input offset
voltage mV
voltage drift
Output short
input voltage
input voltage
circuit current
voltage swing
Supply voltage
supply voltage
TL071 +/-18V +/-30V +/-15V 680 1.4mA +/-13.5V -- 200 65pA 1012ohm 18.0 3.0 100 13.0 3MHz 100 Fast slew, jFET input, low noise
TL081 +/-18V +/-30V +/-15V 680 1.4mA +/-13.5V -- 200 30pA 1012ohm 18.0 3.0 86 13.0 3MHz 86 Low power, jFET input
-0.3 to
TLC27M2C 18V +Vdd 725 285uA -- -- 275 0.7pA -- 2.1 1.1 94 0.62 650kHz 93 Dual, low voltage, precision
+Vdd
-0.3 to

TLC27M7 18V +Vdd 725 285uA -- -- 275 0.7pA -- 2.1 0.19 94 0.62 635kHz 93 Low offset, low power
+Vdd
Vcc- -5
LT1013AC +/-22V +/-30 -- 0.7mA +/-14V 28mA 2500 -12nA 400M 2.5 0.04 117 0.4 800kHz 120 Dual, single rail, high gain
to Vcc+
OP07C +/-22V +/-30V +/-22V 500 -- +/-13V -- 400 +/-1.8nA 33M 0.5 60.0 120 0.3 600Hz 103 Low noise, bipolar input
Copyright © 2000 Wimborne Publishing Ltd and

Low noise, low offset, low drift,
OPA27GP +/-22V +/- 0.7 +Vcc 500 3.3mA +/-13.8V -- 1500 15nA 2G 0.4 0.025 122 1.9 8MHz 120
precision instrumentation
OPA177GP +/-22V +/-30V +/-Vs -- 1.3mA +/-14V -- 12000 0.5nA 45M 0.7 0.02 115 0.3 600kHz 115 Precision, bipolar, instruments
V+ +0.7
OPA544T 70V -- -- 12mA -- 4.0A -- 15pA 1012ohm -- 1.0 106 8.0 1.4MHz -- High voltage, high current, TO220
to V- -0.7
AD711JN +/-18V Vs +/-18V 500 2.5mA -- 25mA 400 20pA 3x1012ohm 7.0 0.3 88 20.0 4MHz 95 Precision, high speed, low offset
AD744JN +/-18V Vs +/-18V 500 3.5mA -- 25mA 400 30pA 3x1012ohm 5.0 0.3 88 75.0 13MHz 95 Precision, FET input
741 +/-18V +/-30V +/-15V 500 1.7mA +/-14V 25mA 200 80nA 2M 15 1.0 90 0.5 1.5MHz 96 Obsolete general-purpose bipolar
LM10 45V +/-40 -- -- 270uA -- -- 400 10nA 500k 2.0 0.3 102 -- -- 96 Low voltage, reference output
Low voltage, battery operation,

LM308 +/-18V -- +/-15V 500 150uA +/-14V -- 300 10nA 40M 6.0 10.0 100 -- -- 96
precision
Table 1: Opamp Selector.

LF411 +/-18V +/-30V +/-15V 670 1.8mA +/-13.5V -- 200 50pA 1012ohm 7.0 0.8 100 15 4MHz 100 Low offset, jFET input
LF441 +/-18V +/-30V +/-15V 670 150uA +/-13V -- 100 10pA 1012ohm 10.0 1.0 95 1.0 1MHz 90 Low power, jFET input
LMC6001- -0.3 to
+/-Vs -- -- 750uA 14.6V +/-30mA 1400 25fA >1T 10.0 0.35 75 0.8 1.3MHz 80 Ultra-low input, instrumentation
AIN +16V
LMC6081 15V Vs -- -- 450uA 14.5V -- 1400 10fA >10T 1.0 0.35 85 -- 1.3MHz 94 Precision, low offset CMOS
EL2044 +/-18V +/-10V Vs -- 5.2mA +/-13.6V 75mA 1.5kV/V 2.8uA 15M 10 0.5 90 325 60MHz 80 Low power, low voltage
Circuit Surgery
EL2001 +/-18V -- +/-15V -- 1.3mA +/-11V +/-100mA 998 1.0uA 8M -- 2.0 -- 2000 70MHz 75 High slew rate, high speed buffer
NE5534 +/-22V 0.5 +Vs 1150 4.0mA <+/-16V 38mA 100 400nA 100k 5.0 0.5 100 13.0 10MHz -- Low noise, audio, instrumation
+V +8 to
CA3130 16V 8.0 -- 300uA <15V 20mA 100 5pA 1.5T 10.0 8.0 90 <30.0 4MHz 74 MOSFET input, rail-to-rail output
-V -0.5
+V +8 to
CA3140 36V 8.0 -- 1.6mA 13V +40mA 100 2pA 1T 6.0 5.0 90 7.0 3.7MHz 80 MOSFET input, bipolar output
-V -0.5
+V +8 to
CA3420 22V 15.0 9.0 150uA -- 2.6mA 100 0.05pA 150T 4.0 5.0 80 0.5 0.5MHz 90 Low voltage, portable instruments
-V -0.5
EPE Online, March 2000 - www.epemag.com - 227

A ROUNDUP OF THE LATEST EVERYDAY NEWS
FROM THE WORLD OF ELECTRONICS
DIGITAL FINGERPRINTING
Digital cameras can now not only assess vehicle speed limit violations, but also
detect fingerprints on banknotes - Barry Fox reports.
Criminals beware. An ad- hangs on the quality and in- and approved on 1 April 1999
venturous local police force in tegrity of the photos. for private data networks. Cam-
the West of England is pioneer- High intensity white light is eras use infrared flash to log the
ing the use of electronic imag- focussed on the print area by time taken for a car to travel a
ing to lift faint fingerprints from fiber optics, and the image cap- mile. They were catching 4000
banknotes and check. The sys- tured on disc by a high resolu- a day but no summonses were
tem relies on watermarking tion digital camera with 3648 x issued. Now that the cameras
technology developed for elec- 4623 image sensor; 12-bit are Type Approved the police
tronic speed trap cameras but monochrome coding captures and local councils can install
spurned by the Home Office. subtle grays in a 13 Megabyte them – but they are expensive.
The mark stops lawyers discred- file. Says Signum’s Marketing
iting fingerprint evidence by ar- Because the image is in Manager Alan Bartlett, “The UK
guing that digital images can be digital code, the lab can use authorities are working on the
easily doctored. Fourier Transform analysis to principle that most motorists will
Three years ago Esther separate the regular pattern of not go to the expense of legally
Neate, senior fingerprinting de- banknote printing from the irreg- challenging a speeding fine. In
velopment officer at the Wilt- ular fingerprint. Labs have tried the US people just shoot the
shire Constabulary, persuaded to do this with film and color fil- speed cameras to pieces any-
the force to go out on a limb ters, but electronic analysis pro- way. But 30 year jail sentences
and replace the ageing film duces much clearer pictures. can hang on fingerprint evi-
equipment in her lab with an all- dence which will be challenged.”
An identifying watermark,
digital system. with date, time and place, is The Wiltshire Constabulary
“The FBI and a few forces embedded in the image using will not identify individual cases
round the world are mixing film VeriData iDem software from that have relied on digitally
and digital technology, but we British company Signum Tech- recorded evidence. “But for the
are first to replace film com- nologies of Witney. The mark is last 18 months all cases this lab
pletely,” says Neate. invisible to the eye but can be has handled have used the pro-
detected by analysis of the im- cess,” says Esther Neate.
age file. If even one pixel in the Wiltshire’s camera can work
Chemical Integrity image is altered, it shows. with a laptop PC running Win-
Prints from a sweaty finger dows NT, to capture fingerprints
are 95 percent water and five or footprints at the crime scene.
per cent salts, amino acids and
Speed Trapping
Police forces in Singapore,
fats. Suspect banknotes and Signum developed VeriData Turkey, Australia, New Zealand,
checks are serially treated with for electronic speed traps, but in Belgium and the US are now
14 different chemicals such as April ’99 the Home Office ap- asking the fingerprinting labora-
ninhydrin, and a photograph proved the use of SVDD (Speed tory in Devizes to help them set
taken at each stage because Violation Detection Deterrent) up similar systems.
each chemical reacts with a dif- digital cameras as long as the
ferent sweat component and images are securely encrypted
destroys the previous reaction. or carried by the closed data
By the end of the treatment, the networks used on motorways.
paper is stained, toxic, and use- SVDD was tested on the M1
less as evidence. So any case and M20, in Leicester and Kent
NEWS......
to 2Mbits/s, but will not be ready
until 2002.
Orange recently launched a
new service called High Speed
Circuit Switched Data, which
hikes the speed to 28××8Kbps.
One-2-One will wait a year until
different technology, called
General Packet Radio Service,
is ready. Both systems are au-
thorized by ETSI, the European
Telecommunications Standards
Institute, but are not fully com-
patible.
HSCSD uses less error cor-
CORDLESS SOLDERING IRON rection to increase the basic
BS Manufacturing tell us that their latest soldering iron, the P100, GSM data rate to 14××4Kbps, and
sets a completely new standard by offering precision heating power then gangs channels together to
in the form of a tiny pen-style tool. The P100 is 19cm long and give 28××8Kbps or higher.
weighs just 57g, delivering up to 120W output – double that of some
GPRS works on the as-
competitive irons, allowing it to be used for heavy duty electrical
sumption that most users do not
work and silver soldering, in addition to precision electronics and
need constant data speed. A
model making applications.
pool of capacity serves several
The fuel used is liquid butane/propane gas, stored in the users at the same time, with bits
translucent handle. The tank can be refilled from a standard gas allocated as and when they are
canister (cigarette lighter type). Typically each refill provides around needed. One-2-One plans a
45 minutes of continuous use. Gas is ignited by means of a spark GPRS service for September
from a flint inside the tool’s cap, with its flow regulated using a slider, 2000, with speeds up to
allowing fine adjustment down to 20W. 56Kbps, rising to 112Kbps by
A wide range of attachments for the P100 are offered, to 2001.
configure it for soldering, hot knife cutting, slicing, heating, igniting, “HSCSD is a technology
shrink wrapping, melting, shaping and other uses. A rich choice of cul-de-sac” says Craig Tillotson,
tips for soldering/desoldering is available, from 4××8mm wedges to One-2-One’s Director of Strat-
chisels and angles as small as 1mm. egy. “GPRS hardware will cope
The iron costs around 18 UK Pounds and may be purchased with HSCSD, but HSCSD hard-
online (via the web site below), or from distributors. ware will not handle GPRS. And
if Orange pass on the true cost
For more information contact BS Manufacturing Ltd., Strawhall
to subscribers, HSCSD access
Industrial Estate, Carlow, Ireland.
will cost at least ten times as
Tel: +353 (0) 503-41340 much as GPRS.”
Fax: +353 (0) 503-40363 “Not so”, says Stuart Scott,
Email: sales@vulkangt.com Orange’s Manager of Internet
Products. “We have not yet set
Web: www.vulkangt.com
war that will replicate round the

CELLPHONE world. Most countries now use Filter Software
Europe’s digital GSM system
DATA WAR but email and Internet access, – Free!
at 9××6Kbps, is painfully slow.
By Barry Fox Microchip, those ingenious
The completely new Univer- PIC manufacturers, have told us
Rival British cellphone net-
sal Mobile Telecommunications that you can now download
works Orange and One-2-One
System promises data rates up some filter design software from
have started a GSM data speed
NEWS......
their web site, free! FilterLab is
a software design tool that sim-
plifies the design of active filter PORTABLE POWER
systems using opamps as ana-
log filters. It provides full
schematic diagrams of filter cir-
cuits with component values
and display of the frequency re-
sponse.
FilterLab supports the de-
sign of low-pass filters up to 8th
order, with Chebyshev, Bessel,
or Butterworth responses, from
frequencies of 0××1Hz to 10MHz.
Once the filter response has
been identified, FilterLab gener-
ates Bode plots and the circuit
diagram. It also generates a
Spice model for time domain
analysis, streamlining the de-
sign process.
To download Filterlab, ac-
cess Microchip’s site at:
www.microchip.com
Power Lines “Run virtually anything in your car!” exclaims a Press Release
from Merlin Equipment. Merlin’s Cherokee unit simply plugs into a
and Health car’s cigarette lighter and converts low voltage battery power to stan-
The National Radiological dard 230V AC mains power. A normal UK 13A socket on the unit al-
Protection Board (NRPB) is to lows direct connection to appliances.
investigate recent claims that a The Cherokee 150 is capable of supplying up to 150 watts of
causal link between power lines power continuously. For appliances that require a surge of power
and human health can be estab- (TVs for example), it can provide 300 watts instantaneously. The
lished. It states that these converter is overload, overheat and short-circuit protected. In the
claims need to be compared event of the input battery voltage dropping below 10××8V, the unit will
with the findings of the first pa- cut out – ensuring that you can re-start your car’s engine!
per from the UK Childhood Can-
cer Study (UKCCS) which Merlin have a large range of other products designed for in-car,
shows no increased risk of caravan or boat use.
childhood cancer associated For more details, contact Merlin Equipment, Dept EPE, Unit 1,
with magnetic fields from the Hithercroft Court, Lupton Road, Wallingford, Oxon OX10 9BT, UK.
electricity supply. This definitive Tel: +44 (0)1491-824333
study, looking at actual cases of
childhood cancer and controls, Fax: +44 (0) 1491-824466
is the largest of its type in the Email: sales@the-merlin-group.com
world. Web: www.the-merlin-group.com
For more information, con-
tact NRPB, Chilton, Didcot,
Oxon OX11 0RQ, UK.
Electronic Purse
One of the many documents that have come to us in connection
Tel: +44 (0) 1235-822744
with the Smart Card 2000 exhibition and conference (8-10 Feb 2000,
Fax: +44 (0) 1235-822746 Olympia, London), highlights the question “Have Europe’s banks in-
Web: www.nrpb.org.uk vested millions developing a product no-one wants?”. The product
referred to is the “electronic purse”.
NEWS......
Electronic purses have the consumer to accept. If the
So will we be jingling the
been the cherished goal of public is not ecstatic about the
coins in our pockets in 10 years
banks and other global organi- concept, neither are many of
time? As we go to press, such
zations for around 10 years. Ap- Europe’s bankers who are fac-
matters are due to be discussed
parently, though, in many mar- ing losses of up to two euros per
by delegates at the conference.
kets consumer feedback indi- purse card per annum. In some
cates they are not a viable countries the costs of persuad-
proposition. ing the customer to load and
The concept of a cash-less use the card exceed incomes by
society is proving difficult for several times.
Robert Penfold looks at the Techniques of Actually Doing it!
In a previous Practically omega symbol up to 999 ohms). than about one millimeter in di-
Speaking article we considered Large values are expressed in ameter.
the subject of capacitors, and either kilohms (kW or just k) or Any lettering on a compo-
this month we move on to that megohms (MW or just M). A nent this small would have to be
other humble component – the kilohm is equal to 1,000 ohms, minute, and would also be eas-
resistor. and a megohm is 1,000,000 ily obliterated. Color codes are
ohms. relatively easy to read, and
At a guess, in most projects
about half the components are even if they become damaged it
should still be possible to read
resistors, so beginners have to COLOUR BAR the values of components cor-
get to grips with resistors right
from the start. One immediate problem rectly.
facing the beginner is that most The normal resistor color
The basic unit of resistance resistors are not marked with
is the Ohm, but this is a small code has four bands, with three
values using normal text char- bands grouped together. It is
unit of measurement. Hence the acters. Instead a system of
resistors used in electronic cir- these three that indicate the
“color coding” is used, and there value of the component while
cuits often have values of thou- are four or five colored bands
sands or even millions of ohms. the other one shows the toler-
marked around the body of ance rating of the component.
The usual abbreviation for each component. The tolerance is simply the
ohm is the Greek letter omega This may seem to be an un- maximum deviation from the
(W), but these days an upper necessarily awkward way of marked value given as a per-
case letter “R” is sometimes handling things, but you have to centage. Thus, if a 100 ohm re-
used instead. (In EPE and EPE bear in mind that the average sistor has a tolerance rating of
Online we tend to use the resistor is an extremely small five percent, its actual value
component. You would be between 95 and 105
will often be deal- ohms.
ing with resistors The group of three bands
that are no more indicates the first two digits of
Table 1: Resistor Color Code

Color Band 1 Band 2 Band 3 Band 4
Black 0 0 x1 --
Brown 1 1 x10 1%
Red 2 2 x100 2%
Orange 3 3 x1000 --
Yellow 4 4 x10000 --
Green 5 5 x100000 0.5% Fig.2. Two methods of five-band
Blue 6 6 x1000000 0.25% resistor color coding.
Violet 7 7 -- 0.1%
Gray 8 8 -- -- Table 2
White 9 9 -- -- 1.0 1.1 1.2 1.3 1.5 1.6
Gold -- -- x0.1 5% 1.8 2.0 2.2 2.4 2.7 3.0
Silver -- -- x0.01 10% 3.3 3.6 3.9 4.3 4.7 5.1
None -- -- -- 20% 5.6 6.2 6.8 7.5 8.2 9.1
Practically Speaking
the value and the multiplier. more than about five percent Nevertheless, this form of
Fig.1 shows the function of each away from a preferred value. five band coding does seem to
band. Table 1 shows the mean- Most resistors are available be used on the resistors sold to
ing of each color when it ap- in the full E24 series, but some amateur users. When applied to
pears in each band. are only available in the E12 normal E24 values the third
As an example, suppose series, which is every other digit is always zero, and the
that a resistor has red–violet– value in the E24 series (1××0, 1××2, third colored band is therefore
orange–gold as its color code. 1××5, etc.). Most electronic pro- black.
The first two bands indicate the jects only use resistors having To compensate for this ex-
first two digits of the value, and values from the E12 series. tra zero the multiplier value is
in this case red and violet re- reduced by a factor of ten. Tak-
spectively indicate that these ing our earlier 27k example, this
are two and seven. The third BUNCH OF FIVES would become red–violet–
band is orange, which means Rather unhelpfully, many of black–red–gold. This gives 270
that the first two digits must be the resistors now sold to ama- x 100 = 27,000 ohms.
multiplied by one thousand in teur users have five band
order to give the value in ohms. codes. These operate in the
manner shown in Fig.2. The first
COMPOSITION
This gives 27 x 1000 and an
answer of 27,000 ohms (27 kilo- of these is quite easy to use be- In component catalogs you
hms (27k)). The color of the cause the first four bands give will find resistors described as
fourth band is gold, and the re- the value and tolerance rating in “carbon film” and “metal film” or
sistor therefore has a tolerance the normal way. The additional “metal oxide”. The simplest re-
rating of five percent. fifth band shows the tempera- sistors are the carbon composi-
ture coefficient of the compo- tion type, which are basically
nent, which is not something just pieces of carbon with an
PREFERRED that is normally of any rele- electrode attached to each end.
VALUES vance. If the fifth band is ig- These have now been re-
nored, the other four give the placed by carbon film resistors,
Resistors are normally value and tolerance rating in the
available in what is called the which consist of a former made
usual fashion. from an insulating material hav-
“E24” series of values. The ba-
sic E24 series of values is listed The second form of five ing an electrode at each end. A
in Table 2, but values ten times band coding is probably the film of carbon is deposited on
higher, a hundred times higher, more common one, and is the former, and the resistance
etc. are also available, up to a slightly more difficult to deal value obtained depends on the
normal maximum of 10 with. Again, it is not that far re- thickness and the exact compo-
megohms (10M). Values one moved from the four-band sition of the film. Carbon film
tenth and one hundredth of the method. resistors are adequate for most
basic values are also available, applications, and are the type
The first two bands indicate
but are relatively difficult to ob- normally specified in compo-
the first two digits of the value,
tain. nents lists.
and the last two bands provide
This range of values might the multiplier and the tolerance Metal film resistors are the
look rather random at first rating. The difference is that an usual choice for more demand-
glance, but each value is additional middle band is used ing applications. They are simi-
roughly ten percent higher than to indicate the third digit of the lar in construction to carbon film
the previous value in the series. value. resistors, but the film is based
Together with the inclusion of on a metal oxide instead of car-
This method of coding can
the same values in various bon. Resistors of this type nor-
handle non-standard values
decades, this means that one of mally have close tolerances of
such as 26××7k, but as these are
these “preferred” values will al- two percent or better, and gen-
not used in amateur electronics erate less electrical noise than
ways be close to the required this is irrelevant to the electron-
value for a resistor. In fact the any form of carbon resistor.
ics enthusiast. The normal four- Their values are also affected
ideal value calculated by a cir- band method of coding is all
cuit designer should never be less by temperature changes
that is needed. and aging.
Metal oxide resistors are marked on the resistor in the
needed for some demanding same form that it appears on a
applications, such as in critical circuit diagram. In other words,
stages of test equipment and in the letter used to indicate the
low noise audio preamplifiers. unit of measurement is also
They can be used in place of used to denote the position of
carbon resistors for general use, the decimal point. A value of
but it makes sense to use 2××7k would be marked as “2k7” Fig.3. High power wirewound
cheaper carbon resistors in any and a value of 0××47 ohms would resistor in an aluminum heat
application where they will suf- be marked as “0W47’’ (or dissipater.
fice. “OR47”).
sistors, and this must be fol-
There will usually be other lowed “to the letter”.
HIGH POWER marks as well, some of which
Some resistors do actually might simply be the makers
name, a batch number or some- POTENTIOMETERS
have the values written on the
body using ordinary text charac- thing of this type. Of more use, The terms “potentiometer”
ters, but in recent years this is there will probably be a wattage and “variable resistor” tend to
something I have only encoun- rating and a letter to indicate the cause a certain amount of con-
tered on higher power resistors. tolerance rating of the compo- fusion. A potentiometer (“pot”)
Most of the resistors used in nent. Table 3 shows the corre- has three terminals, and be-
electronic circuits have to dissi- sponding tolerance rating for tween two of these there is a
pate very low power levels, and each of the code letters used. fixed resistance. There is a vari-
small resistors having ratings of High power resistors have able resistance between these
about 0××25 watts are perfectly various compositions, but the terminals and the third one.
adequate. wirewound variety is by far the A potentiometer consists of
most common. This consists of a track of carbon with a terminal
Some circuits have the odd
a coil of resistance wire wound at each end of the track. The
resistor or two that has to han-
around what is usually a ce- third terminal connects to a
dle higher power ratings. Com-
ramic former. wiper that can be moved along
ponent lists normally indicate a
suitable power rating for all the One slight problem with the track by means of a spindle.
resistors anyway, but a suitable wirewound resistors is that the There are also preset poten-
rating should certainly be given coil of wire also acts as an in-
for any high power types. ductor, although most compo- Table 4
nents of this type are con-
It is very unusual for resis-
structed in a fashion that mini- Potentiometer
tors having power ratings of Letter
mizes this problem. Even so, Type
more than about one watt (1W)
wirewound resistors are less
to have the value marked using A Linear
than ideal for some applica-
colored bands. The larger physi- B Logarithmic
tions, and if a different type is
cal size of these resistors C Anti-logarithmic
indicated in the components list
makes it possible to mark the
it is advisable to use the speci-
value using text characters of tiometers that have no spindle,
fied type.
reasonable size. but can be adjusted using a
Very high power resistors screwdriver. With an open con-
The value is invariably
have metal fins to help conduct struction preset potentiometer
heat from the component into the track, wiper, and terminals
Table 3 the surrounding air (Fig.3). are all clearly visible (see
Many of these resistors also Fig.4). The greater the amount
Letter Tolerance have to be mounted on a sub- of track between the wiper and
F 1% stantial piece of metal, which one of the fixed terminals, the
G 2% acts as a heatsink and provides greater the resistance between
J 5% further cooling. With resistors them as well.
such as this, the article should
K 10% The normal way of using a
give guidance on using the re-
M 20% potentiometer is with an input
voltage across the track, and a is due to the way we perceive
variable output voltage is then sound rather than a problem
available from the wiper with the potentiometer.
(moving contact) and one of the When used as a volume
track terminals. Strictly speak- control, a logarithmic poten-
ing, the components you buy tiometer gives a much better
are always potentiometers hav- control characteristic. Logarith-
ing three terminals, but in some mic potentiometers are used for
applications it is a variable re- little other than volume controls.
sistance that is required. It is Anti-logarithmic controls are dif-
then only necessary to use the ficult to obtain and are only
wiper terminal and one of the Fig.4. “Open Track” preset needed for a few specialist ap-
track connections. plications. If you use a poten-
potentiometer.
Potentiometers are avail- tiometer of the wrong type the
able in three types, which are tions. So does an anti- circuit will still work, but the con-
the linear (lin), logarithmic (log) logarithmic potentiometer, but trol will be awkward to use.
and anti-logarithmic varieties. A with the high and low values the
linear potentiometer gives ap- other way around.
proximately equal resistance AS EASY AS ABC
When used as a volume
between the wiper and the two control a linear potentiometer The values of potentiome-
track terminals when it is at a gives an odd control character- ters are marked using text char-
central setting, as one would istic, with the volume seeming acters, together with the type
expect. to jump to a high level when it is “log”, “lin” or “anti”). These days
A logarithmic potentiometer advanced slightly from zero. many potentiometers are
produces vastly different resis- Further advancement then marked with a code letter to in-
tances under the same condi- seems to have little effect. This
John Becker addresses some of the general points readers have raised. Have
you anything interesting to say? Email us at editor@epemag.com!
meaning of many words 2000 series. It is always useful

WIN A DIGITAL changes amongst the general to reinforce what you have read
MULTIMETER population. There are many with practice. I have subscribed
cases, too, where words have to the Online version of your
acquired different meanings magazine and am looking
The DMT-1010 is a 3 1/2 digit depending on the context in forward to every issue. I have a
pocket-sized LCD multi-meter which they are used. few suggestions which I would
which measures a.c. and d.c. It’s worth remembering (or like to see if possible.
voltage, d.c. current, and is it?!) what Alice’s Wonderland I enjoyed the PDF files
resistance. It can also test friend Humpty-Dumpty said (in a being separate files so that I
diodes and bipolar transistors. rather scornful tone), “When I could save the Teach-In 2000
use a word it means just what I from the Nov ’99 issue to a
Every month we will give a
choose it to mean – neither floppy disk to view it at work on
DMT-1010 Digital Multimeter breaks and between service
more nor less”!
to the author of the best calls. (Remember the
Readout letter. “paperless office” hype?). I am
TEACH-INS PLUS EPE not able to do that with the Dec
ONLINE ’99 issue. The only way to carry
the information with me now is
Dear EPE,
* LETTER OF THE to print those pages out as the
Thank you for not only the whole file is over 3MB. I hope
MONTH * new Teach-In 2000 series, but you have plans to make the
for the Teach-In 1998 series as complete series and the
A-LIVE-A-LIVE-OH! well. I credit knowledge gained previous Teach-In series
from that for helping to secure a available for purchase via the
Dear EPE, new job. Many companies are Internet as well. This would be
A friend recently drew my now giving a written skills test most helpful to readers like
attention to the current series you for technical positions and I myself in Macon, USA, whose
are running on Oscillators (since would not have been able to only source for EPE is one
July '99). Having obtained the pass the electronic and digital bookstore which stocks four
September issue, I sent for the portions without your magazine. copies. I missed some of the
previous issues. What interesting (My experience is in hydraulic PIC Tutorial because it was sold
reading they make! and pneumatic systems.) The out.
series was so well written, I
I have been pleasantly Thanks for a great
learned enough basic
surprised by the exchange of publication.
electronics as well as digital to
information with readers in Alan Craig
be hired. I have since
Readout, and the informative Macon, Georgia, USA
completed a well-known
level of Circuit Surgery.
correspondence course in basic
Arthur Lawrance electronics for which my
via the Net employer will reimburse me. I Alan E-mailed his
found your Teach-In series to be comments to our On-Line Editor
better written and more Alan Winstanley, who E-mailed
It is an interesting fact that, understandable than theirs. I back:
despite attempts at also wish they had been able to We’re really pleased to hear
standardization, English continues provide the interactive software that Teach-In 98 was of benefit
to evolve and, irrespective of I downloaded for the Teach-In to you, your compliments are
“official” definitions, the perceived
Readout
much appreciated. Writing that 2000 series is finished, we are Steve Gooch
series was hard work! We tried planning on offering the whole via the Net
to mix together the essential series on a single CD (actually a
theory along with some practical mini-CD that would fit into your
work. It was also difficult to co- wallet, so that you can easily I wondered if Steve had a
ordinate the material originating read the articles whilst on the power supply regulation
from four writers, to a very tight road). problem, the rectified output of
monthly deadline. The advent of REC1 not providing sufficient
Furthermore, we’re also voltage when the display has
digital cameras helped a lot. planning on offering the entire many digits active. This could
The development and set of EPE 1999 issues on affect synchronization and
production time for Teach-In 98 these mini-CDs for ease of indeed the correct operation of
was shorter than that of Teach- reference. Watch our web page the PIC. Making this suggestion
In 2000, which has been in at www.epemag.com for more to Steve, he later replied:
preparation for at least a year. details in the near future.
There has therefore been much
more time available to develop Thanks for your suggestion.
the Teach-In 2000 programs (It PERSEVERENCE The problem was with the PIC
hasn’t felt like it! JB), and its programming. Why, remains a
author John Becker is very Dear EPE, mystery, data transmission
skilled at producing QBasic I have had the PIC Electric speed? A preprogrammed chip
electronics demo or test and project (Feb-Mar ’96) hanging from Magenta cured all,
measurement software. around for some time, and however. Thanks for the
much to my irritation I have excellent magazine – and the
It is possible to buy
been unable to resolve a back-up support.
photostats of any out-of-print
problem with it: the LED display
EPE Back Issue articles direct
continually shows flashing PIC16F877 PROBLEM
from the UK. These can be
FFFF. The middle FF are fairly
ordered via our secure server, SOLVED
stable, but the outer two FF
accessible from the EPE web
segments are dimmer and flash Here’s another tale with a
site at
in a more pronounced fashion. happy ending. First the problem:
www.epemag.wimborne.co.u
k The circuitry has been
constructed on PCBs purchased
Dear EPE,
from you, using recommended
Alan W also forwarded Alan components obtained through Many thanks for the
C’s queries on EPE Online to its RS. I have checked connections PIC16F87x Mini Tutorial (Oct
Editors Max and Alvin in and component layout and can ’99). It was just what the doctor
Alabama, USA, from whence find no problem with these. I ordered, the inclusion of the
the electronic version is served have tried down-loading the Basic program was a
out. Max responded: program for the PIC a number masterstroke, my attempts to
of times, from a ‘486 66MHz calculate baud rates had
PC. Using both the on-board produced some improbable
Thank you for your kind components, Simple PIC results, most of which would not
comments – it’s always great to Programmer and also the have fitted into an 8-bit register.
receive positive feedback. Sorry PICtutor board, I always end However, I am experiencing
to hear that you would prefer to with the same results. difficulties re-programming a
receive the Online magazine as Adjustment via the calibration PIC16F877. I’ve built a board
multiple small PDFs. In fact, the buttons appears to do little. The along the lines of your Data
main reason we decided to +15V, –15V, +5V and TP7 Logger (Aug-Sept ’99), minus
move to a single large PDF is check out OK. The A-D provision for EEPROMs, and
that a lot of readers requested it reference voltage has been connected it to my EPE PIC
that way (to make it easier to adjusted as advised. Please Tutorial (Mar-May ’98) board
print out the entire issue in one could you offer some and all went well. I ran the input
go). However, you will be happy suggestions? port test program of Toolkit Mk2
to know that once the Teach-In (May-Jun ’99), all voltages
Readout
present and correct, plugged in reconfigured as I described in parallel port, all well and good.
my ’877 and compiled and Mini Tut. Next I decided to get
programmed TKTEST4 into the Then comes the following “EASYPic’n” to start learning.
PIC. It ran and the LEDs flashed. reply back from Derek a couple This is where it’s all got a bit
I then compiled (with no of weeks later: cloudy. EASYpic’n writes out
errors) my own program (well the code ready to be compiled
actually most of it was yours from by MPASM. Undeterred by this I
Mini Tut) intended to initialize the Have cured my problem! – tried to use TASM instead (it’s
USART with a baud rate of by accidentally re-compiling all I have!). Of course doing this
31.250kHz and loop 10101010 to TKTEST1 and loading it. It ran throws up lots of errors and it
give a nice square wave out of straight away. The problem with takes a while to sort them out.
the port. I re-configured the PIC my program was due to having At first this wasn’t too bad, but
and then programmed it with my already defined PAGE1 and as the progs move on its all
own program. Nothing appeared PAGE0 in the header. I then re- getting a bit too much.
to be working, so I assumed that equated them to suit the
Now, I did download
there was a problem with my SETBAUD routine in your Mini
MPASM from Microchip’s
program and re-loaded Tut. I assumed that it would be
website and thought that would
TKTEST4. Nothing happened, the OK to use either routine, to
be it. (Ha, as if!) Of course
PIC appeared dead. It seemed select pages. Apparently not.
MPASM converts my .ASM files
impossible to program it now. In After removing the equates for
to .HEX files. But when it comes
case the PIC had developed a RP0 and RP1 from the header,
to downloading to the chip,
fault I plugged in another. the program ran.
SEND.EXE needs to see .OBJ
TKTEST4 went straight in and Many thanks for your files. (Dare I ask what the
ran. I then attempted with my own concern. difference is?) which is fine with
program again and I have exactly TASM but no good with the hex
the same problem i.e. I cannot re- files of MPASM. So what do I
load TKTEST4 successfully. An interesting situation and
do with these hex files?
solution – from which all us PIC
I have thoroughly checked Next I downloaded MPLAB
programmers should learn a
my board, PIC Tutorial board, from Microchip. This is fine and
lesson!
etc. and can see no dry joints. I I could probably get used to
have checked the clock with an that. The thing is I’m not sure if
oscilloscope and the 4MHz clock TASM, MPASM MEANS that has something built in that
is running, I have re-checked all SPASM will do something with the hex
voltages with the input port test files and then download them.
program of Toolkit Mk2 again and Dear EPE,
Maybe its Picstart Plus.
all voltages are present and Recently I got hold of the
Any reference to
correct, the 12V programming Simple PIC Programmer as
programming the chip seems to
voltage switches between 12V featured in EPE Feb ’96. Once
refer to serial connection, and
and 5V as it should. built it has all worked fine and
guess what – my TASM thingy
has spurred me on to delve
Derek Johnson plugs into the parallel port. Is
deeper into PIC programming,
via the Net my kit a bit of a dinosaur? How
However, I would appreciate it if
do I get my MPASM generated
you could just clear up a couple
hex files to the chip? Do I now
That, then, is the outline of of things that are driving me
need a serial programmer
Derek’s PIC problem, and he had bonkers.
instead? (I bet this is where free
obviously taken the correct steps Firstly, the kit came with downloads end.)
in trying to determine the cause of TASM. I assume that TASM is a
Is there another EPE project
the problem. The only thing that forerunner of MPASM which I
that gets around these
occurred to me was that it might see and hear about wherever I
problems.
be a PIC configuration problem – go, and a specially written
the settings having become (guessing again) program called Mick Tinker
corrupted in some way. I SEND.EXE. The prog is via the Net
suggested such to Derek and compiled with TASM and
recommended that he downloaded to the chip via the
Readout
Such confusion Mick! First let to comment further?
me say that you seem not to have FLAWED PIRACY-
been a regular reader of EPE, PROOFING? PIC vs AVR ETC
otherwise your knowledge of PIC
programming requirements and Dear EPE, Dear EPE
techniques would have increased I found the Pirate-Proof PICs are definitely excellent
through reading the several CDs (News, Dec ’99) interesting microcontrollers, but it does not
articles that we have published on but flawed. Have the people at mean they are the only
the subject since the Simple PIC C-Dilla forgotten that a microcontrollers. There are
Programmer. computer comes with an Audio other good microcontrollers like
In particular you should read Line-in and some more Atmel AVR or Scenix SX. By
the PIC Tutorial series of Mar- expensive sound cards have using only PICs you are limiting
May ’98 (which discusses PIC digital inputs, and certain CD your magazine’s resources.
programming at some length), Players have digital outputs? So Some good projects with other
if a CD player disregarded the microcontrollers will definitely
PIC Toolkit Mk1 of July 98 (which
false error correction code open your magazine to a much
discusses not only programming
would the false code be sent wider audience.
but also the differences between
MPASM and TASM), PIC Toolkit through the digital outputs? I am not saying to throw
(Hmm, I wonder). away the PIC, it should be
Mk2 of May-June ’99 (which is a
much enhanced version of the But that aside, pirates would included, as it is very good for
Mk1 and has many of the just sample from a CD player’s first time programmers, and a
features you obviously have need audio outputs to a PC’s audio lot of your loyals also use PICs,
for, including the ability to inputs, and using a good PC but loyals like myself feel that
translate between MPASM and sampler, save on to a hard we should not be confined to a
TASM – it also allows the newer drive track by track. After doing single subject but rather explore
PIC16F87x series to be so they could (without the false all the possibilities.
programmed, as well as the ’84 code) be put on a CDR disk and Let me tell you what
series). then copied as many times as happened to me, I am a student
they liked. All in the time it and my professor gave an
We strongly recommend that takes to play a regular CD, so it
you, and other readers in a similar assignment, to design a data
seems the pirates will still be logger but to design it from any
position, should read these one step ahead, and that it just other microcontroller except
articles, which I am sure will clear merely slows them down. PICs. None of my colleagues
up a fair number of your
I think true anti-piracy will along with me were able to do
problems. I would comment,
come when CD media is old hat that. We had to hear a long
though, that as you have MPLAB
(I predict in about 10-15 years), lecture on how we have gotten
you should make an in-depth
and solid state memory sticks or used to spoon-feeding, and had
attempt to get to know it, and to confined all our attention on a
cards are the norm. Using
also obtain the other hardware single topic. After graduating
encryption and digital ID tags,
associated with it. As good as and getting a job we might be
the consumer when buying
Toolkit Mk2 is, it does not cover asked by our superiors to design
music from a store would have
the full range of PICs that are his ID put onto the card, this ID a project from Atmel AVR, then
manufactured, whereas would come from the what will we do?
Microchip’s hardware/software manufacturer of the card player
suites are designed to do so and would be unique, therefore Ziyad Saeed
(after all, Microchip are the not allowing it to be played on via the Net
manufacturers of the PICs and so any other player even if copied.
provide a full backup for their use
in industry). Rather than discuss it Darren Portsmouth Editor Mike Kenward replied
all further here, do read the via the Net directly to Ziyad:
above-mentioned articles, I can understand that as a
available as back issues (or student you will need to learn
photocopies in some cases) from An interesting point, Darren, about other microcontrollers, but
the EPE Editorial office. and one I am not qualified to you should realize that for the
comment on. Any readers care type of projects we publish the
Readout
PIC is usually the best and voltages different to 12V. But days. f you really feel in danger
easiest solution. what are these threshold of “running out”, keep a back-up
voltages for a 6V battery? I fuel supply available, in your
We have published projects
would assume the upper case keep another battery
for the Atmel AVR
threshold level to be 6V, but I handy. I compliment you,
microcontrollers, but few
have no idea what the lower though, on your initiative in this
readers were interested. Whilst
one should be. I would be most matter. Building the Voltage
we do publish a range of
grateful if you could recommend Monitor will not only prove to be
educational items we cannot
to me suitable threshold useful constructional
undertake to teach you about all
voltages, bearing in mind that experience, but using it will also
the subjects you will undertake
the project is to be used in help reinforce your concept of
– sometimes you will need to
conjunction with the power electronic power consumption.
find resources elsewhere.
supply for your Teach-In series. FTP PLUS TI2K
I would like to say that I find Dear EPE,
To which I will add my own the Teach-In series excellent. It
comment that I was personally explains concepts in a clear, In Readout it seems some
very disappointed that the AVRs thorough and practical way, and folks have a problem
did not receive the reader I have really enjoyed learning downloading from the FTP site,
response I had hoped for. I had through it. or they are getting corrupted
felt that we should actively code. I always use WS_FTP,
demonstrate that John Thornton which is a free shareware FTP
microcontrollers other than PICs Sunderland program. It’s very easy to use
were in commercial use and I for both down and up loading!
was quite prepared to learn The Teach-In circuits should I have downloaded the
about them for myself and on function even with voltage levels Teach-In 2000 software and
behalf of you all! But, as so few well below 5V, even as low as have to say I like it, it will prove
readers have expressed an 4V. When new, your 6V battery very useful. The pots screen is
interest in AVRs, I too shall stick useful and the cap-resistor time
will probably deliver about 6××5V.
with PICs, which I must add, are constants, can’t wait to see how
I would probably regard 5××5V as
not only “good for first time you convert the printer port to a
being the level at which I would frequency counter! What I
programmers”, but are widely
replace the battery, and so set would like to see added is a 555
used by professional designers
one threshold for a little above timer time-freq calculator and
in industry.
that, and another for about 6V basic opamp configuration with
as advance warning that “fuel” is gain calc, etc.
VOLTAGE MONITOR beginning to get a bit low. But in
many ways, it’s a somewhat Mel Saunders
Dear EPE,
arbitrary matter since the via the Net
I am writing in regard to the amount of current being
Voltage Monitor Starter Project consumed will determine what Hopefully, you should know
in the Feb ’00 issue. Since I am may be regarded as a by now details were given in
a beginner and am following the reasonable life expectancy for Feb ’00 issue, easy isn’t it?!
Teach-In 2000 series, I thought the power remaining in the Sorry to disappoint, though, I’m
that this would be an excellent battery. It’s like with car driving: not covering 555s in TI2K. The
project to build and use to my fuel light comes on when the aim of the series is to achieve a
ensure that the voltage level of
tank is down to a quarter full. It broader sweep without getting
the battery I am using when
is typically a 500 mile tank and into specific named devices.
doing the practical experiments
so I know that I probably still Besides which, there’s been
does not fall below a critical
have well over a hundred miles enough published on 555s to fill
level.
before having to walk! A the British Library twice over
You give very clear and hundred miles on the motorway (much of it originating in EPE) –
easy-to-follow instructions for is less than two hours of fuel I’ve no wish to add to the glut!
determining threshold voltages remaining. Driving locally to the
of the detectors when using the shops and back, the same fuel
device to monitor batteries of probably represents several
with DAVID BARRINGTON module, floppy disk, BT47 relay,
9-way PC serial lead (25-way
extra), and all other
Some Component RF Solutions (UK)
components. They have even
Suppliers for EPE Online Tel: +44 (0) 1273-488880
included a low-voltage stepper
Web: www.rfsolution.co.uk
Constructional Articles motor and UK mains adapter,
RS (Radio Spares) (UK) “plug” type, power supply.
Antex
Web: www.rswww.com All this for just 34.99 UK
Web: www.antex.co.uk
Speak & Co. Ltd. Pounds plus 3 UK Pounds post
Bull Electrical (UK) Tel: +44 (0) 1873-811281 and packing. For full details
Tel: +44 (0) 1273-203500 contact Magenta Electronics
Email: sales@bull-
electrical.com EPE ICEbreaker 2000.co.uk or E-mail:
Web: www.bullnet.co.uk sales@magenta2000.co.uk.
Apart from the specially
CPC Preston (UK) programmed PIC16F877 Finally, we understand that
microcontroller chip, most of the some overseas readers
Tel: +44 (0) 1772-654455
other components needed to (particularly in USA) may have
EPE Online Store and Library difficulty in obtaining the ZTX
construct the EPE ICEbreaker
Web: www.epemag.com project are fairly common items. transistor. The designer informs
Electromail (UK) The “firmware” program in the us that most general purpose
Tel: +44 (0) 1536-204555 chip is loaded and copy- npn types rated at 1A 60V
protected (in the upper half of should work (though not tried) in
ESR (UK) this set-up.
Tel: +44 (0) 191-2514363 the 8K program memory) and is
Fax: +44 (0) 191-2522296 not available in any other way –
Email: sales@esr.co.uk you must purchase the High Performance
preprogrammed PIC chip to
Web: www.esr.co.uk
build this project.
Regenerative Receiver
Farnell (UK) Some of the components
We have reached a special
Tel: +44 (0) 113-263-6311 needed for the High
agreement whereby we are able
Web: www.farnell.com Performance Regenerative
to offer a ready-programmed,
Gothic Crellon (UK) 20MHz version, PIC16F877 Receiver may be hard to track
Tel: +44 (0) 1743-788878 chip together with a floppy disk down. We have not included the
containing the ICEbreaker Jackson type tuning capacitors
Greenweld (UK) in our pricing for this project, as
Fax: +44 (0) 1992-613020 software and the printed circuit
board (code 7000257) -- see the it will depend on the condition
Email: greenweld@aol.com and “newness” of these
Web: EPE Online Store at
variables. We suggest you shop
www.greenweld.co.uk www.epemag.com
around for these items as they
Maplin (UK) Also, the ICEbreaker could add as much as 30 UK
Web: www.maplin.co.uk software, including the simple pounds, or nearly double the
test program and demo price, of this Receiver. Try Bull
Magenta Electronics (UK) programs, is available for free Electrical or J&N Factors (Tel:
Tel: +44 (0) 1283-565435 download from the EPE Online +44 (0) 1444-881965), who may
Email: Library at www.epemag.com be able to offer a good price, if
sales@magenta2000.co.uk
A special package has been they still stock them. You can
Web: also try Mainline Surplus Sales
www.magenta2000.co.uk put together by Magenta
Electronics, which contains the (Tel: +44 (0) 870-241-0810).
Microchip following: preprogrammed We found that some of the
Web: www.microchip.com 16F877 (20MHz version), type numbers quoted for the
Rapid Electronics (UK) printed circuit board, solderless TOKO tuning range coils did not
Tel: +44 (0) 1206-751166 breadboard, LCD display tally with our information and
Shop Talk
could have caused real they had discontinued stocking N. R. Bardwell (Tel +44 (0)
problems. However, thanks to them. To solve this problem 114 255-2886) – Digital
the designer’s, Raymond Haig, FML Electronics (Tel: +44 (0) Multimeter special offer.
efforts in double-checking with 1677-425840) purchased some
the TOKO suppliers, we now specially and, at the time of
have the correct type numbers. going to press, we understand
PLEASE TAKE NOTE:
they still have stocks. Scratch Blanker Jan '00
The TOKO coil numbers
and ranges used in the The rest of the components, We have been informed
Receiver have been set out in a including the ceramic resonator, that the MN3004 delay-line and
table (next month) and were should be readily available the MN3101 clock generator ICs
purchased from Bonex Ltd (Tel: items. Just one point, specify called for in the Scratch Blanker
+44 (0) 1753-549502). Type the L suffix when ordering the are no longer produced or
numbers and order codes are as BC184L general-purpose stocked by Maplin. However we
follows: CAN1A350EK, 380- transistor, because other types understand that Sky Electronics
350; RWO6A7752EK, 357-752; have differing pinouts to this (Tel +44 (0) 20-8450-0995)
RWR331208NO, 351-208; one. have some.
154FN8A6438EK, 356-438; The single-sided printed
KANK3426R, 363-426; circuit board is available from
KANK3337R, 363-337; the EPE Online Store (code
MKXNAK3428R, 363-767. 7000258) at www.epemag.com
The rest of the components
for this project should be widely
stocked. The three Receiver Automatic Train Signal
printed circuit boards are All parts, including the
available as a set and are LF351N IC, for the Automatic
obtainable from the EPE Online Train Signal, this month’s
Store at www.epemag.com “starter project”, should be
We could not close without stocked by our regular
saying that the author has components advertisers. The
produced a really “professional” choice of LEDs, 3mm or 5mm,
Regenerative Receiver – almost will depend on gauge and size
a piece of nostalgic art! of your model railway layout.
The LED current is not very
high, so “high brightness” types
Parking Warning System are preferable.
A few dedicated parts are
called up for the Parking Teach-In 2000
Warning System and may not
No additional components
be obtainable from your usual
are called for in this month's
local component stockist. The
installment of the Teach-In 2000
PIC260435 infrared sensor/
amplifier/demodulator came series. For details of special
from Farnell, code 139-877. packs readers should contact:
(This device has nothing to do
with PIC microcontrollers.) ESR Electronic
Components – Hardware/Tools
Turning to the HT12B or and Components Pack.
HT12A encoder and the HT12D
decoder ICs, these caused Magenta Electronics –
considerable sourcing problems Multimeter and components, Kit
last time they were used in a 879.
published design. At that time, FML Electronics (Tel +44
they appeared in a well-known (0) 1677-425840) – Basic
company’s catalog, but, in fact, component sets.

EPElectronics 2000 03

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

EPElectronics 2000 03

Uploaded by

Copyright:

Available Formats

Volume 2 Issue 3

HIGH PERFORMANCE REGENERATIVE RECEIVER - Part 1 173

AUTOMATIC TRAIN SIGNAL - by Robert Penfold 179

INGENUITY UNLIMITED - hosted by Alan Winstanley 194

EPE ICEBREAKER - by Mark Stuart 183

SERIES AND FEATURES

TECHNOLOGY TIMELINES - Part 2 Days of Later Yore, plus Fundamental 198

TEACH-IN 2000 - Part 5 - Waveforms, Frequency and Time 208

NET WORK - THE INTERNET PAGE surfed by Alan Winstanley 221

CIRCUIT SURGERY - by Alan Winstanley 223

PRACTICALLY SPEAKING - by Robert Penfold 228

REGULARS AND SERVICES

READOUT - John Becker addresses general points arising. 235

Copyright © 2000 Wimborne Publishing Ltd and EPE

Copyright © 2000 Wimborne Publishing Ltd and EPE

How to avoid having an unwanted rear entrance to your garage!

This is a device to aid you in INFRA-RED INFRA-RED BUZZER

the proximity of the vehicle by

positive supply, which should A7. Since pins A0 and A1 on

Fig.4 (below). Wiring from the circuit

Fig.5 (above). The LEDs mounted in

Completed remote LED

LEADOUTS FOR ALTERNATIVE TRANSISTORS

which is directly connected to

together with an erratic

Powerful local radio

transmitters can swamp

regenerative receivers (they

even cause problems with

superhets of advanced design).

inclusion of what is known as a

capacitor C1 form a parallel

tuned circuit, which presents a

high impedance at resonance,

see Fig.1. When the inductor/

capacitor combination is set to

the frequency of the offending

The problem is invariably

Waves, and suitable component

Fig.2. Alternative electronic tuning sys- Frequency (kHz) Frequency (kHz)

Fig.2. Complete circuit diagram for the Automatic Train Signal

Layout of components inside the small plastic box. Note, the

Zener Diode Tester - A tap on the knee

The Purpose of the circuit dia- 35V

gram illustrated in Fig.1 is to help FS1

unidentified Zener diode. It is cen- 10V

voltage regulator (IC1), which is 0V

mode. The current is set to under E

mula I = 1××25V/R, and this flows D6

nected in parallel with the Zener diodes D1 and D2 and capacitors D8

rectly. By adjusting control poten- much lower secondary voltage

grammed into a chip before it can

tems – such as the EPE PICtutor 2

details) and are adequate for the

but for larger changes and pro- 11

grams it can be tedious, and er- 32

duce as to eliminate. RE2

grams rely on more advanced 17

PC, reads input pins, and 0V (GND)

switches output pins to match the

RA2/AN2/VREF 4 37 RB4 MCLR/VPP/THV 1 28 RB7/PGD

RA3/AN3/VREF 5 36 RB3/PGM RA0/AN0 2 27 RB6/PGC

RA4/TOCK1 6 35 RB2 RA1/AN1 3 26 RB5