UK 4 — elektor april 1979

elektor

Volume 5

48 decoder

Elektor Publishers Ltd., Elektor House,

10 Longport, Canterbury CT1 1PE, Kent, U.K.

Tel.: Canterbury (0227) 54430. Telex: 965504.

Office hours: 8.30 - 12.45 and 13.30 - 16.45.

Bank: 1. Midland Bank Ltd., Canterbury, A/C no. 11014587
Sorting code 40-16-11, Giro no. 3154524.

2. U.S.A. only: Bank of America, c/o World Way
Postal Center, P.O. Box 80689, Los Angeles,

CA 90080, A/C no. 12350-04207.

3. Canada only: The Royal Bank of Canada,

c/o Lockbox 1969, Postal Station A, Toronto,

Ontario, M5W 1W9. A/C no. 160-269-7.

Please make all cheques payable to Elektor Publishers Ltd. at the
above address.

Elektor is published monthly.

Number 51/52 (July/August) is a double issue.

SUBSCRIPTIONS: Mrs. S. Barber
Subscription 1979, January to December incl.:

U.K. U.S.A./Can. other countries

surface mail airmail surface mail airmail

£8.50 $21.00 $31.00 £8.50 £14.00

Subscriptions normally run to December incl. Subscriptions from
May issue: U.S.A./Can. other countries

U.K. surface mail airmail surface mail airmail

£5.50 $14.00 $ 20.00 £5.50 £9.00

Back issues are available at original cover price.

Change of address: Please allow at least six weeks for change of address.
Include your old address, enclosing, if possible, an address label from a
recent issue.

ASSISTANT MANAGER: R.G. Knapp

ADVERTISING MANAGER: N.M. Willis

National advertising rates for the English-language edition of Elektor
and international rates for advertising in the Dutch, French and German
issues are available on request.

EDITOR U.K. EDITORIAL STAFF

W. van der Horst L Meiklejohn

J. Barendrecht
G.H.K. Dam
P. Holmes
E. Krempelsauer
G. Nachbar

TECHNICAL EDITORIAL STAFF
A. Nachtmann

J. Oudelaar
A.C. Pauptit

K. S.M. Wal raven
P. de Winter

Technical telephone query service, Mondays only, 13.30 - 16.45.
For written queries, letters should be addressed to dept. TO.
Please enclose a stamped, addressed anvelope or a self-addressed
envelope plus an IRC.

ART EDITOR: F. v. Rooij

Letters should be addressed to the department concerned:

TQ = Technical Queries ADV = Advertisements

ED = Editorial (articles sub- ADM = Administration

mitted for publications etc.) EPS = Elektor printed circuit
SUB = Subscriptions board service

The circuits published are for domestic use only. The submission of
designs or articles to Elektor implies permission to the publishers to
alter and translate the text and design, and to use the contents in other
Elektor publications and activities. The publishers cannot guarantee to
return any material submitted to them. All drawings, photographs,
printed circuit boards and articles published in Elektor are copyright
and may not be reproduced or imitated in whole or part without prior
written permission of the publishers.

Patent protection may exist in respect of circuits, devices, components
etc. described in this magazine.

The publishers do not accept responsibility for failing to identify such
patent or other protection.

Dutch edition: Elektuur B.V., Postbus 75, 6190 AB Beak (L),
the Netherlands.

German edition: Elektor Verlag GmbH. 5133 Gangelt, W-Germany
French edition: Elektor Sari, Le Doulieu, 59940 Estaires, France.

Distribution in U. K. :

Seymour Press Ltd.. 334 Brixton Road, London SW9 7 AG.
Distribution in CANADA: Fordon and Gotch (Can.) Ltd.,
55 York Street, Toronto, Ontario M5J 1S4.

Copyright ©1979 Elektor publishers Ltd. - Canterbury.

Printed in the UK.

contents

elektor april 1979 - UK 5

In early TV games, the
'objects' on the screen
consisted of little more
than two vertical bars
('bats') and a small
square ('ball'). Nowadays,
complete cowboys,
battletships or jet aircraft
are more in demand. The
pP TV games computer
can cater for all tastes,
p. 4-06 and p. 4-26

contents

selektor

4-01

pP TV games 4-06

An introduction to the microprocessor-controlled TV games
computer.

For mobile public
address systems, in
particular, an amplifier
that will produce well
over 100 watts while
running off a standard
car battery can be
extremely useful. The
Stentor was designed to
provide a truly
'stentorian' mouthpiece
in this type of
application.

p. 4-14

By fully utilising the
available output power in
a sound installation, it is
often possible to produce
clearly intelligible speech
in even the noisiest of
environments. The
Assistentor uses a two-
pronged approach:
dynamic compression to
boost low-level signals
and frequency response
tailoring to put all the
power into the most
important part of the
speech spectrum.

p. 4-20

A few of the possibilities
of the TV games
computer are illustrated
on this month's cover:
racing, cards, and even
simple arithmetic.

Photographs courtesy of
Philips Gloeilampen-
fabrieken NV, the
Netherlands.

car light reminder <m. Penrose) 4-13

An extremely simple circuit that will remind the motorist to
switch off his headlights before leaving the car. No
components are connected in series with the existing wiring,
so that it is easy to install and will not affect the operation of
the car electrical system.

stentor 4-14

In Homer's Iliad, Stentor is the name of one of the Greeks at
Troy, who it is said had a voice louder than that of 50 men.

Stentor thus seemed a suitable name to call an amplifier
which, to continue the analogy, provides the power of
50 transistor radios!

The distinguishing feature of this amplifier is that it is port-
able (!) since it can be powered from a normal 12 V car bat-
tery and is therefore ideally suited for open air P.A. work.

TAP Switch (U. Sussbauer) 4-19

This circuit for a touch activated switch uses only 4 V* com-
ponents!

assistentor 4-20

Everyone must at some time have had the maddening experi-
ence of standing on a railway platform and straining anxiously
to make out what the station announcer has just said — a
situation which provides a good example of the problems
facing P.A. systems used in environments with high levels of
background noise.

One solution is to employ a dynamic range compressor to
boost the amplitude of the softer portions of speech above
the threshold level of the background noise, without altering
the level of the louder speech passages. When used in con-
junction with the 'Stentor' power amplifier the result should
be a P.A. system which will make itself heard in even the

noisiest of crowds.

zero pF screen 4-23

quiz master 4-24

In many quiz games it is important to be able to determine
who is the first to reply to a question. To avoid disputes and
unseemly family arguments the simplest approach is to use an
impartial electronic 'quizmaster'.

building the TV games computer 4-26

The introductory article on page 4-06 describes what the
TV games computer looks like and what it can do. Without
further ado, this article gives the constructional details.

market 4.35

advertisers index UK 22

Supplement:

BASIC (part (2), an introduction to a
simple computer language.

selektor

alektor april 1979 - 4-0 1

—duastiL;

Salting away spare power

Meeting peak demands for electricity
means idle machinery at other times.

One way to generate more economically
is to run the main stations at a more
even rate and store their spare, off-peak
power in the form of compressed air in
underground caverns. It can then be
called back to augment the supply at
peak hours. Storing some of this energy
through heat exchange promises to
make such systems even more efficient.
In most countries, and certainly in the
industrialized world, electricity gener-
ating authorities are faced with the same
problems: how to supply all the
demands of their consumers and how to
make the best use of their generating
equipment. Every authority also faces
the fact that the main demand comes in
the working day, when equipment runs
flat out to meet the need. The load is
only slightly relieved as darkness falls
because the domestic consumer then
starts to make big demands on the
supply. But late at night the demand
from all users drops markedly, and the
pressure on turbines and generators is at
last reduced.

One could be forgiven for thinking that
the ease-up is no bad thing. But any
machinery that is subject to peaks and
troughs in usage becomes more and
more susceptible to wear and,
subsequently, more likely to break
down. The family car never runs better
than at cruising speed on a good road,
when it consumes less fuel per mile
than at any other time. In the power
generating field this cruise condition is
highly desirable, for it would allow
turbines to be run at a steady rate
through the 24 hours. Peak demands
would have to be met by stored energy,
and it is the store problem that prevents
the best use of power output potential.
The only storage available to electrical
authorities is hydrostorage, in which
excess output is made to pump water
up into a convenient reservoir. Then,
when peak-period power is needed, the
water is released to run down through
channels to drive turbines. But not all
countries are blessed with the geological
formations that lend themselves to dams
and reservoirs, and in some countries
water itself may be scarce. So what is
the answer? The battery system to store
vast amounts of power for domestic and
industrial use has yet to be invented;
but there is hope from another source,
and that is to store the surplus power
below ground.

At Huntorf, in West Germany, a
unique power storage system is now in
use. Though the principle has been
discussed for some years, this system
employs the first commercial prototype
generating unit using compressed air as
a storage medium. During the off-peak
period the turbines in the power station

drive a compressor unit and compressed
air is channelled into nearby salt
caverns. Then, when demand reaches
its peak, the stored air is released under
careful control to drive a turbo-
generator, which adds its weight
to the station’s output; but,
unfortunately, the heat generated by
the compression process has long since
been dissipated. To develop enough
expansion to drive the turbine
efficiently, fuel has to be added to the
air and the mixture ignited.

So, futuristic as it is, the Huntorf
system still wastes quite a lot of energy.
First, fuel has to be provided for com-
bustion; second, in the compression
stage a good deal of heat is lost by being
dissipated. Every schoolboy knows that
when you operate a bicycle pump it
produces heat. The same is true with
the Huntorf compressors, and the heat
there goes into the salt cavern. There is
the danger that it might harm the
storage chamber by making the walls
become semi-plastic and thereby reduce
the available storage space. If the
storage were not the ideal salt caverns,
with their guaranteed sealing qualities,
but rock caverns, or caverns with water
all round them, the heat would create
even greater problems. It would make
‘dry’ rock even more dry, giving rise to
crumbling and subsequent leaks, and if
a seal were provided by placing the
compressed air inside an underground
water ‘bubble’, quite clearly the water
would soon be driven off by the high
temperatures involved. It is this problem
that has attracted the attention of
Britain’s Central Electricity Board.

At their Marchwood research station,
near Southampton, they have been
working for some time on ways to make
use of the heat generated in the com-
pression stage of storage. What they
propose is that once the turbines are
switched to the compression stage,
instead of the air being directed straight
to the storage cavern it should be passed
through heat-retaining units. These
would be giant ‘silos’ containing heat-
retaining materials such as dense fireclay
pebbles. The air subsequently in store
would be cool, so it would not cause
the storage chamber to deteriorate. But
that is only a relatively small advantage.
It is further proposed that when the
trapped air is released to drive a power
turbine at peak periods, it should pass
back through the silos, where it would
pick up enough heat to start the turbine
without having to add fuel. In effect
you would have a turbine that did not
burn any fuel at all. Ambitious this may
well be, but small-scale tests have shown
it to be feasible and many American
authorities are looking at the ‘no-fuel’
method very seriously indeed.

In the opinion of all concerned, salt
caverns form the best store, not only

because of their sealing qualities but
because they can be easily created. All
that is needed is to tap a hole into the
ground in a known salt area and then
pump water into it at high pressure.
Unwanted deposits are washed out by a
pumping system. Proponents are quite
convinced that natural rock caverns
could be utilised, too, with perhaps a
sealant applied to the rock face. And
the idea of storage in a bubble becomes
feasible if the compressed air has been
cooled.

It is understood that the Huntorf unit
reaps back about 70 per cent of the
energy expended in storing its output.

A figure of 80 to 85 per cent or even
more is possible by first removing and
then replacing the heat.

Theoretical storage volumes are in
figures that stretch the mind, but
various authorities in the UK and else-
where are seriously evaluating this sort
of storage. And who knows, perhaps we
are witnessing the first steps along the
road to the power station gas turbine,
as we know it, becoming obsolete.
Spectrum no. 158

(451 S)

Computerised camping

Wigwam International Ltd. of Norwich
are using a computer system to keep
track of holiday bookings for 220 tents
and 37 caravans on seven sites in
France, Spain and Italy.

At the end of each day charts of each
site with the latest state of bookings and
vacancies for the whole summer are
produced in a few minutes at
160 characters per second by a Tally
1612 printer.

Since campers may stay one night or
ten, vacancies occur in random patterns.
Re-arranging bookings to make
maximum use of tents and caravans
used to be difficult and time-consuming,
and producing the up-dated charts by
hand was tedious and open to error.

Now when a booking is entered the
computer searches for a vacancy and,
if it finds one, allocates it. Then the
printer produces a letter of confir-
matio to the camper on pre-printed
stationery. When the peripheral
bookings — on ferries or airlines, for
example — are confirmed, the printer
runs off the invoice. It is also used to
produce statistics and analysis reports.
Tally Limited, Tally House
7 Cremyll Road,

Reading RG1 8NQ,

England

(453 S)

I

I

Looking into engines with lasers
(Dr. B. W. Dale)

Many people are surprised when they
find that research is going on into the
way internal combustion engines work.
What more needs to be known about an
engine that has been around since the
last century, is based on a simple
principle, and has been the subject of
countless engineering studies?

Because internal combustion engines
can be made to perform quite well
without our needing to know precisely
what goes on inside the cylinders, there
has been little incentive to study in
detail the way they work. But several
factors have combined recently to
create problems for engine designers.
Growing awareness of environmental
problems has led to pressure for quieter,
less polluting engines. Realization that
the world’s oil resources are being used
up has produced two more pressures:
engines will be expected to become
progressively more efficient in their use
of fuel and, more important, the
practice of refining motor fuel to suit
engines may have to cease. It looks as
though we shall have to design our
engines to accept, as time goes by, fuels
that fall more and more short of ideal,
especially if the oil companies are
forced, in the end, to synthesize liquid
fuel from coal.

Engine designers have a fairly clear idea
of the sort of fuel-air mixture that is
needed for desirable combustion
characteristics. The problem lies in
creating suitable mixtures and in
transporting them, without deteriora-
tion, inside a practical engine. These
points can best be illustrated by
example. In a fourstroke diesel engine,
air is drawn into the cylinder during the
induction stroke by a downward motion
of the piston. The inlet valve closes and
the air is compressed by an upward
motion of the piston, becoming very
hot in the process. Fuel is injected just
before the end of this compression
stroke. Combustion starts within about
a millisecond, so during this short delay
the fuel must become thoroughly mixed

with the air for combustion to be
satisfactory.

Rapid mixing is commonly achieved by
ensuring that the air in the cylinder is
swirling and turbulent at the moment of
injection, and by injecting the fuel in
such a way that it immediately breaks
up into very small droplets. Note,
however, that the swirling and turbulent
motion must be imparted to the air
during the induction stroke, half a cycle
before combustion takes place. In
practice it is found that, for no apparent
reason, some air-inlet ports are very
much better than others in achieving the
desired result. An understanding of this
problem would remove a large element
of uncertainty from diesel engine design.
The problem is mainly how to measure,
and then to try to acount for, the way
the fuel droplets interact with gas-flows
inside engines.

Complex Problems

In a typical petrol engine the fuel and
air are mixed in the carburettor. The
fuel-laden air then has to flow past the
throttle plate and the bends and
junctions in the manifold without
depositing the fuel droplets onto the
solid surfaces. In the manifold there is
the further problem of ensuring that all
cylinders receive the same amount and
quality of fuel-air mixture, in spite of
the fact that cylinders 1 and 4 are
further away from the source of the
mixture than cylinders 2 and 3. Such
problems are often solved by putting
weirs or baffles inside the manifold, but
these are usually more effective at some
engine speeds than others, which means
that the manifold functions well only
under certain engine conditions. Here,
too, the problems are those of
understanding gasflows and their inter-
action with liquid droplets. They are
very complex problems, bearing in mind
that at a typical engine speed the flow
in the manifold changes direction some
100 times a second because of the cyclic
demands of the four cylinders.

Many ingenious techniques have been
developed for studying fuel-air motion
I inside engines. High-speed photography

has achieved considerable success, but is
essentially qualitative. The depth of the
volume in which the fuel droplets and
other particles are in focus (that is, the
depth of field) is never certain; the
technique is also off-line, which means
the data does not become available
during the experiment. Most other
techniques involve inserting probes into
the engine cavities; they may be vanes,
heated wires, tubes, and so on. None of
these give unambiguous results. They
clearly interfere with the normal
functions of the engine and they have
never been used with complete success
on a running engine.

Over the last decade, a number of
optical techniques have been developed
to measure various physical phenomena
in engineering. One of the most highly
developed is laser anemometry, for
measuring fluid flow. There are several
ways of constructing a laser anemo-
meter. The most straightforward is the
so-called real-fringe system.

Crossing Beams

Here, we must first consider some of the
wave-like properties of light. A light-
wave is a propagation through space of a
small, rapidly oscillating, electric
potential, in the same way that a water-
wave is a propagation across a surface of
an oscillating depth, or potential, of
water. If we represent fluctuations in
the electric potential by alternate light
and dark shading, laser beams can be
depicted as shown in figure 3, which
represents the crossing of two laser
beams. It is the peculiar and very useful
property of laser beams that the wave-
fronts are spaced absolutely evenly and
are parallel. In the cross-over region of
the two beams the total oscillating
electric potential is the sum of the
components from the two beams, and
here a very useful phenomenon occurs.
Along the line A A', for example, one
beam always reinforces the effect of the
other (in our representation the shading
is either particularly heavy or
completely absent). Remembering that
all the wave-fronts are moving in the
direction of propagation at the speed of
light, any point along the line A -A' will
experience a particularly strong
oscillating electric potential, and will
therefore be intensely illuminated. But
along the lines B-B', orC— C', one laser
beam exactly cancels out the effect of
the other (in our representation the
intensity of shading does not change).
The electric potential does not oscillate
at any point along lines B— B' or C— C',
so the illumination here is zero. It will
be seen that the cross-over region of the
two beams consists of alternate light
and dark ‘fringes’ parallel to A-A'. In
laser anemometry the cross-over region
is referred to as the control volume, and
typically occupies 0.01 cubic millimetres.

elektor april 1979-4-03

asHsan

A small liquid droplet or particle of dust
happening to pass through the control
volume meets the alternately light and
dark fringes and, if it is observed from a
distance, appears to flash on and off as
it passes through the fringe system. If
the repetition frequency of the flashes
is measured, the rate at which the
particle is passing through the control
volume can be found. This is the
principle of real-fringe laser anemo-
metry. Unless they are specially filtered,
all normal fluids contain an abundance
of minute particles. If the fluid is
flowing the small particles follow the
flow very closely, so measuring the
movement of the particles is tantamount
to measuring the movement of the fluid.
The laser anemometer just described, in
common with other types of anemo-
meter, suffers from three intrinsic and

serious drawbacks. First, although the
rate at which particles move is
measured, the direction in which they
move is not discovered. Second, the
anemometer gives wrong information in
cases where the turbulent fluctuations
of the flow (which can be thought of as
the random movements of tiny eddies
of fluid super-imposed on the bulk
flow) are about the same size as the
average flow. In such instances small
pockets of fluid move, momentarily, in
a direction opposite to the bulk flow.
The anemometer, unable to distinguish
between pockets of fluid flowing in
opposite directions, treats all pockets as
travelling in the same direction and gives
an incorrect value for the average flow.
Third, the instrument is embarrassed by
stagnation regions, such as occur at the
centre of vortices, where no particles

Figure 1. Four stages in the engine cycle of a
four-stroke diesel engine. During the
induction stroke (a) air is drawn past the inlet
valve into the cylinder. The inlet valve is
closed during the compression stroke (b) and
the air is compressed, becoming hot enough
to ignite the fuel when it is injected. Injection
takes place when the piston reaches 'top dead
centre' (c). After about a millisecond,
combustion begins, in this case starting in the
combustion bowl hollowed out of the piston
crown. During the working stroke (d) the
piston is forced down by the hot, expanding
gases produced by ignition of the fuel.

Figure 2. In the four-cylinder petrol engine,
the fuel-air mixture is prepared in the
carburettor. Fuel-laden air then has to travel
past the throttle valve, which controls the
total flow of fuel-air charge, and through the
manifold and inlet valves to the cylinder. The
inlet valves open one at a time (usually in the
order 1, 3. 4, 2) to allow fuel-air mixture to
be drawn into the cylinders.

Figure 3. Wave-theory representation of two
laser beams crossing. Light and shaded bars
show the wave-fronts of alternating electric
potential. In the cross- over region the two
beams reinforce one another along A-A' but
cancel out along B-B' and C-C'. The cross-
over region therefore consists of a pattern of
light and dark fringes parallel to A-A'.

Figure 4. Frequency-shifting gives laser
anemometry a considerable advantage when
dealing with highly turbulent flows. In (a)
flow velocities V are plotted against the
fractions of fluid travelling at those velocities.
The plot shows some of the fluid to have a
negative velocity; that is, it moves against the
main flow. An anemometer unable to
distinguish between positive and negative
velocities would count negative values as
positive and give a wrong average value. By
shifting the frequency, as in (b), all velocities
are made to appear positive and the measured
profile is the same as the true one. Bias
introduced by this technique is subtracted
later to give the true velocities of the fluid.

Figure 5. (a) Optical system for measuring the
steady-flow pattern set up when air is forced
through a diesel-engine inlet port, past the
valve and into a transparent cylinder. A laser
beam is split in two and the beams directed to
meet at a point. A particle moving through
the fringe system formed in the control
volume scatters light in a stream of pulses at a
rate which is a measure of the particle
velocity. The scattered light is collected by a
lens and detected by a photo-multiplier, and
the pulse rate is worked out by an electronic
signal processor, (b) Optical measurement of
gas velocity from back-scatter uses the same
principle as that for forward scatter.

cross the fringes and no signals are
recorded.

All three difficulties can be overcome
by causing the fringe system itself to
move in a direction at right angles to the
fringes. Suppose, then, that the fringes
are made to move faster than any
pockets of fluid that are present.
Particles moving in the same direction as
the fringes will appear to be moving
very slowly, and particles moving in the
opposite direction will appear to be
moving very fast. If the bias introduced
by the fringe motion is then allowed
for, the true velocities of the particles
can be found. This resolves the
ambiguity in the direction of motion,
and permits valid measurements to be
made in highly turbulent flows.

Stagnant regions are no longer a
problem because fringes are swept past
the particles; if there are particles, there
will be signals. The fringes can be moved
within the control volume by making
the two incoming laser beams have
slightly different optical frequencies, a
technique known as frequency-shifting.
It can be done either by inserting an
electro-optical device, such as a Ken-
cell, into the path of one of the beams,
or by using a rotating diffraction grating
as the device for originally splitting one
laser beam into two.

Substantial Advantages

The device of shifting the frequency,
coupled with an inherent lack of
sensitivity to heat released within the

selektor

alektor april 1979 - 4-05

system under study, gives the laser
anemometer substantial advantages over
other techniques in the study of
complex and/or combusting flows. The
laser technique interferes to only a very
slight extent in the normal working of
the system under study, although it
does require optical access; inserting
windows is often a serious problem. It is
sometimes necessary to ‘seed’ the flow
with extra scattering particles to speed
up the measurements.

Typical optical arrangements for making
laser anemometry measurements are
shown in figure 5 . Small particles scatter
light in all directions, but mostly
forward. The easiest way of conducting
an experiment is therefore as shown in
the first diagram (5a), which shows the
measurement of the steady flow-pattern
produced when air is forced through a
diesel engine inlet port, past the valve
and into a transparent cylinder. But
such an arrangement is often impracti-
cable in engineering work, for it needs
two transparent windows.

The so-called back-scatter technique is
shown in figure 5 b. The scattered light
is collected through the same window as
is used for the incoming beams. Optical
access problems are greatly reduced, but
because the light scattered in the
backward direction is 1 000 times less
than that scattered in the forward
direction, the measurements themselves
are a lot more difficult.

At Harwell, the Research Laboratory of
the United Kingdom Atomic Energy
Authority, we have had a great deal of
experience, over a long time, in applying
laser anemometry to the study of gas
and liquid flows in difficult environ-
ments met with in engineering. Towards
the end of 1976 we were given the task
of demonstrating that laser anemometry
could be applied to the study of internal
combustion engines. The work was
funded entirely by the UK Department
of Industry, and we have been advised
and guided by a panel of senior
engineers drawn from the UK motor
industry. Imperial College, London, has
collaborated in several aspects of the
work.

So far, our internal combustion engine
project has been highly successful.
Demonstration experiments have
included the measurement of gas flows
in the manifold of a motored petrol
engine, in the cylinder of a research
petrol engine (motoring and running),
and in the cylinder of a motored
production diesel engine. (By motored,

I mean that the engine is turned over at
normal working speeds by an electric
motor, without combustion in the
cylinders.)

The problems

The first problem to be considered is
that we need to know what is happening

at specific parts of the engine cycle, a
substantial but tractable problem for
modem electronics. Each particle
passing through the control volume
produces a train of light pulses, which
are collected in a photo-multiplier and
converted into a train of electrical
pulses. The pulse repetition frequency
is measured and the information is fed
into an electronic store. Simultaneously,
a device attached to the crank-shaft of
the engine is made to produce either a
stream of timing pulses, or a digital code
providing instantaneous identification
of the point reached in the engine cycle.
The timing information can be handled
in several ways. The pulses can be used
to direct the signals from the photo-
multiplier to any one of a number of
separate electronic stores, each
corresponding to a different part of the
engine cycle, or the timing device can be
used to ‘label’ each signal from the
photo-multiplier with a code identifying
the point in the engine cycle at which
the signal originated.

Many optical techniques of studying
engines suffer from the fact that viewing
windows become sooted up during
running, and tend to collect droplets of
fuel, smears of oil, and so on. Laser
anemometry manages to surmount the
problem of dirty windows because it
depends upon measuring the repetition
frequency of light signals rather than
their intensity.

With modem techniques, repetition fre-
quencies can be determined from very
weak signals. The problems of fuel and
oil films or droplets on viewing
windows, and of stray reflections from
internal surfaces, are solved largely by
careful focusing of the image of the
control volume onto the collection
aperture of the photo-multiplier. This
ensures that light originating from
anywhere other than the control volume
is strongly defocused. Light emitted
during the combustion process can be
removed by optical filtering.

Engine vibration is not a problem
because the dimensions of the control
volume are such that the time particles
take to cross it is short compared with
the time taken for the engine to move
significantly through vibration. The
main effect of the vibration is to
cause the whole of the control volume
to oscillate slightly in relation to the
engine. In practice this means that the
results obtained are averaged over a
volume of several cubic millimetres,
which is acceptable.

Other Optical Techniques
Air-flows represent only a small part of
the picture. Clearly, to understand the
fuel-air mixing process, for example, it
would be necessary to be able to
measure the sizes of fuel droplets and
the way they evolve. Although

commercial instruments are available for
measuring droplets, they are mostly
unsuitable for engine work. We are
developing an instrument for measuring,
simultaneously, the size and the velocity
of fuel droplets, as functions of crank-
angle, inside engines. If successful, such
an instrument would be able to examine
how well fuel droplets of various sizes
follow the gas flow, a matter that is
fundamentally important.

Knowing how fuel droplets of various
sizes move in an engine would provide
much of what needs to be known about
the mixing of fuel and air; but the
picture would still be incomplete, for
little would be known about the
vaporized fuel. We are hoping to
develop laser-Raman scattering to the
point where it can be used to discover
the concentrations of fuel, oxygen,
nitrogen, carbon dioxide, carbon
monoxide and so on in the vapours
inside engines. The technique, which
depends upon the propensity of gas
molecules to extract small, discrete
amounts of energy from passing beams
of light, has been demonstrated
successfully upon simple flames, but the
application to real engine problems may
be several years away.

(Spectrum no. 159.)

(448 S)

Booze from STC?

A huge underground reservoir beneath
one of Standard Telephones and Cables'
factories in South Wales supplies water
of such quality that the company is
supplying it to blenders of whisky, rum,
gin and brandy.

The first tanker load of water has been
shipped to Avery’s of Bristol where it
has been used in the reducing process of
750 cases of malt scotch whisky
destined for the United States.

STC’s telephone cable factory in
Newport, Gwent, uses over 40,000
gallons of water a day in its manufac-
turing processes. When the 1 976
drought threatened production, STC
had to look for an alternative water
supply. The company found the reser-
voir under its own factory and the
Welsh Water Authority granted STC
a licence to abstract up to 18 million
gallons of water a year.

Subsequent tests have shown that the
natural untreated water is completely
free of bacteria. The water is filtered
only to prevent any possibility of dis-
colouration from the sandstone through
which it passes.

(452 S)

4-06 - elektor april 1979

MP TV games

jiP TV games

Doing battle with chips.

'Chips', in the electronic sense,
used to be transistors. Then came
the Integrated Circuit, quickly fol-
lowed by 'MSI ICs' (Medium Scale
Integration Integrated Circuits) —
quad opamps, for instance. Inevi-
tably, these were followed by 'LSI
ICs' (Large Scale Integration):
complete alarm clocks, TV games
and microprocessors. It's amazing
how much circuitry manufacturers
can squeeze onto a 'chip' now-
adays!

However, microprocessors have
posed an unexpected problem.
We've got them, but very few ama-
teur electronics enthusiasts seem
to know what to do with them!
The best approach would appear
to be to simply use them as 'com-
plicated integrated circuits', build
something interesting around them,
and quickly proceed to forget what
particular component makes the
unit tick. That particular approach
is what we have in mind with the
circuit described here. Once com-
pleted, it is a box with a keyboard
and joystick controls; after play-
ing some 'space sounds' off a tape
or disc into it, it becomes a sophis-
ticated TV games machine — with
full colour, all sorts of 'object
shapes', score-on-screen and sound.
At a later date, it should also
create an irresistible urge to
'program' ones own games. This
will prove to be relatively easy: the
unit just happens to contain what
microprocessor 'freaks' call
'comfortable monitor software' . .

The TV games computer contains a
microprocessor - /iP for short. So be it.
There’s no reason to be scared of /zPs -
they don’t bite! Furthermore, abso-
lutely no knowledge of microprocessors
is required to have fun building and
using this unit. At the same time, it of-
fers a unique possibility of getting to
grips with fiPs at a later date, simply by
exploiting its capabilities to the full.

This article is intended as a basic intro-
duction to the unit.* The descriptions
of what it does — and how — are suffi-
cient for those readers who want to
build and use it with the TV games pro-
grams available via the Elektor Software
Service (ESS). A much fuller descrip-
tion, for those who want to develop and
program their own games, will be sup-
plied with the p.c. board (and is also
available separately — for those who
have access to facilities for etching
double-sided boards with plated-through
holes ...).

For many of our readers, however, we
expect that this information will be of
little interest at present, so we do not
intend to publish it in Elektor. Let’s
concentrate on what it does.

It’s great fun! For some time now,
we’ve had a working prototype set up in
our offices and our editorial staff regu-
larly crowd around as two colleagues en-
gage in a heated battle of wits. We’ve
also got a solo game - man against
machine - which is basically an extend-
ed version of noughts and crosses. It’s
called ‘four in a row’. The ‘machine’ has
proved to be a formidable opponent ...
What, exactly, can it do? Well ...

What it does

The unit is intended for TV games. This
means that it must create a (colour)
picture on a standard (PAL) TV set, cor-
responding to some game - man against
man or man against machine. The effect
— and the fun — can be enhanced by
including sound effects. Games can be
won or lost, so a ‘score’ display is also
required.

* The actual constructional details are given
in a companion article elsewhere in this issue.

The first requirement for a ‘game’ dis-
play are ‘objects’ on the screen. In early
TV games, these consisted of two verti-
cal bars (bats) and a small square (ball).
Nowadays, complete cowboys, battle-
ships or jet aircraft are more in demand.
The TV games computer can cater for
all tastes.

A single ‘object’ can consist of up to
80 squares in an 8 x 10 rectangle (see
table 2). By filling in the necessary
squares, quite detailed objects can be
displayed: as an example, in figure 1 a
locomotive fills the centre of the screen.
The same object can be repeated in dif-
ferent parts of the screen; furthermore,
several sizes are available: the little
steam engine at the bottom of the
screen in figure 1 is a scaled-down ver-
sion of its big brother in the centre. To
be more precise, any object can be dis-
played in four different sizes: 1:1, 2:1,
4:1 and 8:1.

The shape of the object can be re-pro-
grammed quite rapidly, so that the same
‘section of memory’ can produce sev-
eral different objects (or object sizes) in
the same picture. However, this is a
complicated system for microprocessor
specialists only; for most users it is more
interesting to know that four com-
pletely different objects can be ‘stored’
simultaneously. A battleship, sub-
marine, jet aircraft and ‘missile’, for in-
stance; or, for a more peaceful game,
two footballers, a goal (displayed twice)
and a ball.

The colour of the objects can also be
determined according to personal taste:
a choice of eight colours is available.

Table 1.

The TV games computer in a nutshell

- both joystick and keyboard control;

- video output suitable for PAL colour TV
receivers (UHF and VHF input);

- sound effects via built-in loudspeaker;

- microprocessor controlled, and therefore
programmable for a wide variety of games;

- cassette interface included, for easy load-
ing and storing of programs (e.g. ready-
made games supplied through the Elektor
Software Service);

liP TV games

elektor april 1979-4-07

. d

Each of the primary colours (red, green
and blue) can be selected indepen-
dently, so that the various mixtures are
also available; for instance, red plus
green equals yellow. All in all, red,
green, blue, three mixtures (orangy-yel-
low, greeny-blue and purple), white and
black can be produced.

For most games, objects are not
enough: a ‘background’ is also required.
This may consist of anything from a
partial boundary to a complete cross-
hatch pattern. Building up the desired
background with the TV games compu-
ter is similar, in many ways, to building
up an ‘object’. Basically, the back-
ground consists of 160 squares: 10 rows
of 16 squares each. Each square is de-
fined by its upper and left-hand sides:
the right-hand side and bottom ‘belong’
to the adjoining squares. Any of these
sides can be displayed, as required; fur-
thermore, by ‘widening’ the left-hand
edge it is possible to fill in the entire
square.

This system can be used to create a wide
range of different background patterns.
A few examples are shown in figure 1:
all the squares along the bottom row
have been ‘filled in’ to produce a solid
edge. The checker-board pattern in the
lower left- and right-hand corners is
achieved by filling in alternate squares;
the cross-hatch along the top consists of
the sides of the corresponding squares
only. Various other horizontal and verti-
cal lines, bars and dots are all variations
on the same theme.

The colour of the general background
‘between the lines’ and that of the lines

themselves can both be chosen inde-
pendently. As before, a choice of eight
colours are available. The only restric-
tion is that if the same colour is chosen
for both background and ‘object’, the
latter will not be visible!

The final point, as far as the picture is
concerned, is the possibility of display-
ing the score. The TV games computer
offers four alternative display modes,
each consisting of four digits. These
digits can be displayed either at the top
or at the bottom of the picture; further-
more, they can be used either as a single
four-digit number (e.g. 2650) or as two
two-digit numbers (26 50).

In figure 1, the latter option is displayed
at the top of the screen.

The colour of the ‘score display’ is al-
ways chosen to contrast with the colour
of the background lines.

So much for the picture. However, this
by no means exhausts the capabilities of
the games computer. As mentioned ear-
lier, sound effects can also be included:
the unit contains a ‘programmable
squarewave generator’ that can be used
to produce all kinds of squawks and
whistles via a small built-in loudspeaker.
In most games, the score depends to a
large extent on ‘hits’ and ‘misses’. In
this unit, collisions between each object
and any other object or the background
are all detected and stored individually.
This information can be used for a va-
riety of different effects, quite apart
from up-dating the score. Suitable
sound effects can be triggered; an object
can be made to change shape, disinte-
grate, or simply disappear; the back-

Figure 1. This photo illustrates the various
display possibilities of the TV games com-

ground can be altered; and so on. For
instance, in a particularly war-like game
like a shoot-out between two cowboys,
each time one of them is ‘hit’ he could
fall down and then proceed to float
slowly up to and off the top of the
screen - to the accompaniment of
‘heavenly’ music!

The facilities discussed so far are the
basic ‘pieces’ out of which a game can
be constructed. For an exciting game,
the picture and sound must be modified
continuously - as the players manipu-
late their joy-stick controls and/or key-
board. This is where the ‘brain in the
box’ comes in: once the microprocessor
has been programmed for a particular
game it will set up the necessary objects
and background, monitor the signals
from the players’ controls and detect
any ‘collisions’, and proceed to modify
the display, add sound effects and up-
date the score accordingly. The beauty
of this system is that merely running a
new program from tape into the ma-
chine is sufficient to obtain an entirely
different game — with the number of
possible games being virtually unlim-
ited!

Furthermore, the microprocessor is suf-
ficiently ‘intelligent’ to be a formidable
opponent in solo games (man against
machine), provided the rules of the
game are not too complicated: the
memory in the basic unit is insufficient
for games like chess.

Having discussed, in general terms, what
the ‘games computer’ can do, the next
step is to find out how. In other words,
what makes it tick?

4-08 — elektor april 1979

pP TV games

2

I

III?

4 analog

13-bits ADDRESS BUS

oro

How it works

At this stage, we only intend to give a
very broad outline of the operating prin-
ciples. Some more detailed explanations
will be given later; a truly complete de-
scription is outside the scope of this ar-
ticle — it is available separately, as men-
tioned earlier.

A functional block diagram is shown in
figure 2. In the first block, ‘CPU' stands
for Central Processing Unit. This is the
actual microprocessor chip — the
‘brain', in other words. It controls all
the other units, calling them up as re-
quired via the ‘address bus'. Although
microprocessor specialists seem to take
particular delight in describing this type
of process in (mistifying) jargon, it is
basically a fairly straightforward circuit.
Code locks and code switches are fairly
well-known, and several circuits have
been published over the years.

Imagine, now, that each of the other
blocks contains one or more of these
code switches. When the correct code
appears at the input of one of these
switches (i.e. power on the correct lines
and no power on the wrong ones), the
switch operates and the corresponding
part of the block is turned on. Provided
each individual circuit in the whole unit
is operated by its own code switch with
its own unique code, the CPU can turn
on any one of them by simply putting
the correct code on the address bus.

A ‘brain’, on its own, is fairly useless.
A ‘memory’ is also required, and two

different memories are actually available
in this unit. The first is a ‘ROM’ (Read
Only Memory); it contains all sorts of
‘standard reference’ information — how
to read programs off a tape and how to
store them on tape, for instance. The
second block of memory is labelled
‘RAM’, which stands for ‘Random Ac-
cess Memory’. Information can be
stored here and recalled as required; its
primary use is for storing the program
for a particular game, as well as informa-
tion that changes in the course of each
game - such as the scores.

CPU, ROM and RAM could be called
the ‘internal organs’ of the unit. On the
outside, there is a TV set, a keyboard,
joysticks, a loudspeaker and a tape re-
corder. In computer language, these are
all called ‘peripherals’. None of these, of
themselves, can be controlled by the
CPU - a very unsatisfactory state of af-
fairs. This failing is rectified by adding
suitable conversion and switching units
between the peripherals and the compu-
ter proper. Once again in computer lan-
guage, this type of ‘buffer unit’ is called
an ‘interface’.

Going back now to the block diagram,
the remaining sections require little fur-
ther explanation. The fourth block is
labelled ‘PVI’ (Programmable Video In-
terface); it is the ‘interface’ which pro-
vides the outputs for the TV set and the
loudspeaker. It also takes care of the
analog inputs from the joystick con-
trols. The next block, the ‘keyboard in-

terface’, is exactly what its name im-
plies. Finally, an unnamed block takes
care of the remaining in- and outputs —
the tape or cassette recorder in particu-
lar. The other 7 pairs of in- and output
lines from this block can be disregarded
for the present.

The text ‘8-bit bidirectional databus’ in
the link along the bottom of the blocks
is another piece of computer jargon. It
signifies that output signals from the
various sections can be passed along to
other units - the arrows illustrate the
possibilities. The CPU, for instance, can
receive information from any of the
other blocks, or it can itself provide in-
formation. By contrast, the ROM and
the keyboard can only provide informa-
tion (on request, when called up via the
address bus).

The way in which the various units
work together (under central control of
the CPU) is best illustrated by an exam-
ple.

Painting by numbers

Take the steam engine in figure 1. How
do you go about getting it on the
screen?

The picture on the screen is determined
by information stored in the Program-
mable Video Interface (PVI). However,
this unit will not accept information
from the keyboard, or even from the
keyboard interface, and so for the pur-
pose of this explanation we will have to

»iP TV games

dektor april 1979 - 4-09

resort to a trick: the RAM (the section
of memory in which we can store any
desired information) will be used as a
go-between. First, however, let us see
what information the PVI requires.
Without going into too much detail ...
Obviously, the shape of the ‘object’ is
the first thing to look at. As stated ear-
lier, any object is built up in an 8 x 10
rectangle: 8 horizontal and 10 vertical
divisions. At this point, computer jargon
becomes useful: it can be used as short-
hand. A ‘bit’ is a single unit of informa-
tion: voltage present or absent on a par-
ticular wire; for digital systems, ‘1’ or ‘0’
respectively. A certain number of bits
(the number of bits depending on the
system in question) is called a ‘byte’.

For the ‘games computer’ a byte con-
sists of 8 bits; in other words, all infor-
mation along the data bus is passed in
groups of 8 bits, each being 0 or 1 .

Each row in the 8 x 10 rectangle for one
object consists of 8 squares. Each of
these squares corresponds to one ‘bit’,
so a complete row is one ‘byte’. There-
fore, 10 bytes are required for one ob-
ject. If we take a closer look at the loco-
motive (figure 3), the top row consists
of four squares, followed by a three-
square gap and one final square. The bit
corresponding to each filled-in square is
‘1’, so the first byte is 11110001.

This ‘binary number’ can also be written
in shorthand. Reading from right to left,
the first bit counts for 1 , the second for
2, the third for 4 and the fourth for 8.
These four bits, added, can therefore
correspond to any number between 0
and IS. So-called hexadecimal numbers
also run from 0 to 15: the numbers
from 0 to 9 are followed by the first

letters of the alphabet - A is 10, B is
1 1, and so on. Reading the first byte as
two groups of four bits, we can there-
fore use hexadecimal numbers as short-
hand:

1111 = F and 0001 = 1, so the first
byte is FI.

Using this system, the complete picture
can be described as ‘FI, 51, 55, etc.’, as
shown in figure 3. So far so good: we
can now describe any object by means
of ten (hexadecimal) numbers. How-
ever, these numbers must be stored in a
section of the PVI that will be ‘address-
ed’ by the CPU when it decides that an
object should be displayed. The ‘ad-
dresses’ within the PVI (and anywhere
else, for that matter) are also given in
hexadecimal numbers. The ten bytes
corresponding to the shape of the first
object are stored in the PVI at addresses
1F00 through 1F09. The following
four addresses (1F0A through 1F0D)
are used to store the horizontal and ver-
tical position coordinates of the object
and any desired duplicate(s); two fur-
ther addresses (1F0E and 1F0F) are
used as ‘scratch-pad memory’ (for the
moment, they are best forgotten) and
the shape of the second object is stored
at addresses IF 10 through 1 FI 9. And
so on.

The position of the object has been
mentioned briefly. This is determined,
quite simply, by the distance from the
top left-hand comer of the screen to the
top left-hand comer of the 8 x 10 rec-
tangle. The picture is divided into 227
horizontal and 252 vertical units; the
horizontal and vertical displacement of
the object are determined by the num-
bers stored in bytes 1F0A and 1F0C

Figure 2. Functional block diagram of the

Figure 3. Painting by numbers!

respectively (for the first object) — the
horizontal coordinate and the vertical
offset (with respect to the main object)
of a duplicate are stored in bytes 1F0B
and 1F0D.

The first two bits in byte 1FC0 deter-
mine the size of the first object (the
third and fourth bit — from the left -
are for the second object, and so on).
‘00’ is the smallest size, ‘01’ is twice as
large, ‘10’ is x 4 and ‘11’ is x 8.

Finally, the colour of the objects is
stored in bytes 1FC1 and 1FC2. Read-
ing from right to left, the first three bits
determine the presence or absence of
each of the primary colours in object 2 ;
the next three bits determine the colour
in the first object; the two left-hand bits
are unused.

Now, let’s cut a long story short. We
know what information the PVI needs;
let’s put it there and see what happens.
Utilising the ‘monitor software’, pres-
ent in the ROM, the necessary informa-
tion can be stored in the RAM - after
which the CPU can be instructed to
transfer this information to the PVI, to
produce the required picture on the
screen.

The information is typed in via the key-
board, as shown in table 3. Basically,
the complete section of memory be-
tween addresses 0A00 and 0ACA is first
erased, after which the correct informa-
tion for the desired object is stored in
the Random Access Memory (starting at
address 0A00). Finally, a brief pro-
gram is started that causes the CPU to
transfer the requisite information from
the RAM to the PVI — and the picture
appears on the screen.

The explanations given in the table are

4-10 — elektor april 1979

nP TV games

This program can be used t
keyboard as follows:
keystroke(s) display on

0900 05

0901 CA

0902 06

produce an 'object' on the screen. I

start 'monitor' program
user: 'I want to store a prog
Computer: What is the first
The first address is 0900.
Roger. What data? Cxx' is tl
already stored here).

The first 'data byte' is 05.
The second is CA.

n be used ti

C;B

C; A

0901 CB ;

0901 CA
0; 6 0902 06

The complete program is as follows:
address data

0900 05CA )

0902 06CA 1

0O4A00 I
CD7F00 |

FA78 J
0C1E88 1
4410 i

0907
090A
090C
090 F

0911

0921

0923

0926

0929
092B
092E

0930
0933

0400 1

05CA L
06CA f
CD4A00
FA7B J
04F£ I
CC 0AC8 |
CC BAC9 J
0409

CC 0AC6 /
0402 \

rs 0A00 through 0ACA.

V in all RAM addresses
00 through 0ACA (i.e.
s section of memory).

score' display (by
impossible' numbers — FF —
ore bytes 0AC8 and 0AC9).

IF 0900

09 in byte 0AC6).
Determine the sizeo
(02 = half-size).

Having loaded tl
'erase' program s
keystroke

PC = 0000
PC = 0916
(blue)

PC - 0916

address

0A00

0A01

0A02

0A03

0A04

0A05

0A06

0A07

0A08

0A09

0A0A

0A0B

0A0C

0A0D

run a program . . .
gat address 0916'.
sm is running.
ie 'monitor' program.

ta describes the shape
ngine (see figure 3).

Horizontal position of object.

Horizontal position of duplicate(s) (off screen).
Vertical position of object.

Vertical offset of duplicate(s).
in the screen, by using the 'load PVP program

After operating the and 'MEM' keys, up to
three more objects can be stored (starting at
addresses 0A10, 0A20 and 0A40); a back-
ground can be added (from address 0A80);
the sizes and colours of the objects can be
varied (0AC0 . . . 0AC2); background and
screen colours can be chosen (address 0AC6
- note that the background will only appear

if the second data 'number' is J
nally, data in address 0AC7 deti
sound. Note that it is possible to
fro through the memory withoi
data by operating the '+’ and
complete chaos results, the 'eras
can be run — from address 0916. h

intended as a challenge: by playing
around with the shape, colour, position
and size information all sorts of other
objects can be created. Have fun!

Background

The background is programmed in much
the same way as an object. It was men-
tioned earlier that the background con-
sists of 160 squares: 10 rows of 16
squares each. Each square is defined by
its top and left-hand edges, so that there
are 320 edges in all; each of these can be
selected independently, by setting a cor-
responding bit in the PVI to ‘1’. The 16
edges in one row correspond to 1 6 bits —
two ‘bytes’, in other words. For one
row of squares, two bytes are therefore
required for the upper edges and two
further bytes for the left-hand sides; the
10 rows thus require 40 bytes in all.

As illustrated in figure 4, the addresses
of these bytes run from 1 F8<9 and 1F81
for the first row of ‘upper edges’ to
1FA6 and 1FA7 for the last row of
‘left-hand sides’. As when ‘program-
ming’ for an object, storing 1111 1111
(= FF) in both the first and second
bytes (1F80 and 1F81) causes the
whole first row of ‘upper edges’ to ap-
pear on the screen. Similarly, ‘FF’ and
‘2A’ in the third and fourth byte,
respectively, will produce the pattern of
left-hand edges shown.

Unlike the situation when displaying ob-
jects, the size and position of the back-
ground are, of course, fixed. The size of
the individual lines can, however, be
varied. Starting from the top of the pic-
ture, one bit is used to determine
whether the first row of ‘top edges’ ap-
pears as full-width lines or whether each
‘edge’ is shortened to a dot. A second
and third bit can similarly widen the up-
per and lower halves of the left-hand
edges to fill the adjoining square. The
next row of squares is dealt with in the
same way, using three more bits. Fi-
nally, two further bits can be used to
widen all the lines (top edges and left-
hand sides) in the first two rows, to
either j or \ the width of a square.

To sum it up: 8 bits - one byte - deter-
mine the width of all the lines in the
first two rows. Four further bytes take
care of the other four pairs of rows; the
addresses of these five bytes run from
1FA8 to 1 FAC.

The various possibilities are shown in
figure 4. The information given in hexa-
decimal numbers to the left of the draw-
ing determines which lines appear; it is
stored under the 40 corresponding ad-
dresses. To the right of the picture, the
addresses and contents of the five ‘hori-
zontal extension bytes’ are listed.

One final point remains to be discussed,
before the background can be conjured
up on the screen: the colour. One byte
in the PVI takes care of this (address:
1FC6). From right to left, the first three
bits determine the presence or absence
of the three primary colours in the gen-
eral background between the lines

Figure 4. The background can be constructed
from a wide variety of horizontal and vertical
lines, squares and dots.

Figure 5. The score display: two groups of
two digits or one group of four.

(‘screen’); the fourth bit turns the back-
ground lines on and off (‘background
enable’) — note that, unless this bit is 1,
no background at all will appear!

The fifth through seventh bits deter-
mine the colour of the lines in the back-
ground; the last bit is unused.

Using the same principles outlined earlier
(table 3, in particular) it is now possible
to produce any desired background pat-
tern, by entering the necessary informa-
tion into the PVI (again using the RAM
as go-between).

What's the score?

Displaying the score on the screen is
simplicity itself. As mentioned earlier,
the score consists of four digits. The
first two are stored in the PVI under
address 1FC8 (four bits for each digit);
the second pair in byte 1FC9.

The only other points to determine are
the position and type of display re-
quired. In byte 1FC3, the extreme right-
hand bit determines the position: 0 for
the top of the screen and 1 for the bot-

4-12 - elektor april 1979

nP TV games

tom. The second bit in the same byte
determines the score format. As shown
in figure 5, a 0 causes the score to be
displayed as two two-digit numbers; a
single four-digit number is obtained by
storing a 1 at this bit.

No other possibilities for position or
size are available, and the colour is iden-
tical to that of the background lines. In
most cases, these limitations do not
pose too much of a problem. However,
if more flexibility is required, it is of
course quite possible to use one or two
of the ‘objects’ to create a score display
of any size, shape or colour and in any
position!

Collisions

The score and sound effects will depend
to a large extent on collisions. When an
object hits the background, this is de-
tected and a bit corresponding to that
object is set to 1. For the four objects,
the four left-hand bits in byte 1FCA are
used for this purpose. Similarly, the six
right-hand bits in byte 1FCB are used to
signal inter-object collisions.

The CPU can ‘read’ these two bytes as
required, and use the information to up-
date the score, change the direction of
travel of an object, add sound effects,
and so on.

Sound effects

The only sound produced is a square-
wave. The frequency is determined by
an 8-bit number, stored in byte 1FC7 in
the PVI. If the number is zero (00), no
sound is produced; otherwise, the out-
put frequency is determined by the
value stored. To be precise, the fre-
quency equals

f _ 7874

6

where n is the value stored in the ‘sound
byte’. For instance, if n = 01, the out-
put frequency will be approximately
4kHz; as higher numbers are stored the
frequency decreases, until for the high-
est value (FF) the output becomes ap-
proximately 30Hz.

At a later date, other sound effects can
be added if required. Although the
games computer itself cannot produce
them, it has other outputs that are un-
used at present. These can be used to
trigger external sound effect generators.

Instructions for use
Having seen, in general terms, what the
games computer can do and how it does
it, the next step is to get it on the job.
Assuming that the complete unit has
been built, adjusted and connected up
to the TV set as described in the com-
panion article, it is now time to start
typing instructions on the keyboard
(figure 6).

For the moment, the most important
‘procedure’ to know is how to transfer
programs from an ESS record (or tape)
to the memory in the computer. After
switching on, the keys ‘reset’ and ‘start’
are pressed, whereupon the letters ‘I1II’
should appear at the lower lefthand
corner of the screen. The machine is
now ready for programming.

Each program on tape or disc is preced-
ed by a so-called ‘file number’. This is a
single hexadecimal digit (not 0), and is
intended as an aid when looking for one
particular program out of several on the
same tape. Having operated the reset
and start keys, the rest of the procedure
is therefore as follows:

- press the ‘RCAS’ key (‘Read Cas-
sette’). The computer will respond
by asking for the file number of the
program: ‘FIL = ’.

- type in the desired file number;

^ O 0*0 +

□

®

I 3

,5

Hffi) c f

\p 1/2^ ^REG^] 8 f

prfpffff

®

®

mmm

®

mmm

— operate the *+’ key;

- start the tape at some point before
the start of the desired program.

The file number will now appear at the
top of the screen (‘FIL + 3’, for in-
stance) and the machine will start its
search for the correct program. If it
finds other file numbers first, it will dis-
play these on the screen; meanwhile,
two dots under the V sign will flash
rapidly. As soon as the correct file num-
ber is located it will proceed to store the
following program; this is indicated by
the fact that the two dots now flash
slowly. When the complete program has
been stored, the text ‘PC = 0916’ (for
instance) will appear; the four-digit
number is the so-called ‘start address’
for the program, as specified on the
tape. As soon as this text appears,

- the V key is pressed. This starts the
program and, with it, the game!

Should an error occur while the pro-
gram is being read from the tape, the
transfer is stopped and the text ‘Ad =
0D00’ (for instance) will appear. The
number signifies the first ‘address’ in the
group of instructions (‘block’) where
the error occurred. In this case, the tape
must be wound back and the whole pro-
cedure repeated.

None of the other keys will normally be
required until one starts writing pro-
grams. At that point, the more extensive
information supplied with the p.c.
board will be required anyway, and it
contains full details on how to use the
‘monitor’ program. However, to satisfy
initial curiosity, we can run through
them quickly:

— WCAS, which stands for ‘Write Cas-
sette’. This key is used when storing
programs on tape. The machine will
first want to know the first address
in the program (BEG =); then the
last address (End =); then the so-

Figure 6. The keyboard.

car light reminder

elektor april 1979 — 4-13

car light reminder

called start address (Sad =); and
finally the file number (FIL =).

- BPI/2: this key is used for inserting
so-called break points in a program,
as an aid to ‘debugging’ it.

- REG: this is used for checking or al-
tering the contents of seven 8-bit
‘registers’ and a 16-bit ‘status regis-
ter’ in the CPU.

- PC, for ‘program counter’. This key
can be used for entering the ‘start ad-
dress’ of a program ; when a program
has been copied from tape, this will
normally have been taken care of
automatically.

- MEM: the ‘memory’ key. This is used
when storing a ‘home-brew’ program
in the memory by means of the key-
board. It was also used in the pro-
gramming example given earlier (ta-
ble 3).

So much for the keyboard. For that
matter, so much for this introductory
article! All that remains is for us to ex-
press our hope that everybody will get
as much fun out of this ‘playful compu-
ter’ as we have! H

M. Penrose

There are many different designs for
alarm circuits which remind the motor-
ist to switch off his lights before leaving
the car. The advantage of the circuit
presented here is that no extra com-
ponents have to be connected in series
with the existing wiring, so that it will
not effect the operation of the car
electrical system. Furthermore the cir-
cuit is extremely simple, consisting
solely of a DC buzzer, a double-pole
double-throw switch, and a handful
of diodes (depending upon the num-
ber of functions to be monitored).

The accompanying circuit diagram
shows how the car headlamps, side-
lights, fog lamps, and heated rear
windscreen can be monitored. Note
however that the circuit will not indi-
cate if the above functions are actually
working!

With SI in the position shown in the
diagram, the buzzer will be activated
the moment the engine is switched
off and one of the monitored functions
is left on. Switching off the function

concerned also cuts the buzzer.

If one wishes to leave a particular
function on, e.g. the parking lights,
then SI should be switched to its
alternative position, whereupon the
buzzer is disabled until the engine
is started again. The alarm can then
be re-armed by switching SI back to
its original position.

Normally there is a resistance (Rl) in
parallel with the ignition system (as
a result of the various dashboard indi-
cator lights, fuel guage etc.) which is
sufficiently small to ensure that the
buzzer will be activated in the event
of the contact breaker coming to rest
in the open position when the car is
stopped. If, however this resistance
is too large, a 100...200J2 resistor
(2 W) can be connected in parallel
with Rj_. It may be preferable to use
a small lamp - roughly 0.1 W/12 V -
since due to its positive temperature
coefficient, the more power it dissi-
pates, the greater its resistance becomes.

H

4-14 -elektor april 1979

In Homer's Iliad, Stentor is the
name of one of the Greeks at
Troy, who it is said had a voice
louder than that of 50 men.
Stentor thus seemed a suitable
name to call the amplifier
described here, which to continue
the analogy, provides the power of
50 transistor radios!

The distinguishing feature of this
amplifier is that it is portable (!)
since it can be powered from a
normal 12 V car battery and is
therefore ideally suited for open-
air P.A. work.

There are a number of situations in
which high output power amplifiers are
required, but where connection to a
mains outlet is either difficult or down-
right impossible. One need only think
of e.g. political meetings and demon-
strations, fStes, school sports days,
processions etc. The Stentor has there-
fore been designed specifically to meet
the requirements of ‘bags of Watts’ and
portability. Excluding the use of a
mains supply results in some degra-
dation of amplifier output. However
the aspects of the amplifier’s perform-
ance which are worst affected, namely .
an increase in harmonic distortion, are
of secondary importance in the type of
application for which the Stentor is
intended (harmonic distortion is obvi-
ously less significant than, e.g. cross-
over distortion or clipping).

Starting with a car battery as the source
of supply voltage, what kind of output

power can be achieved using conven-
tional amplifier output stages? If one
takes the example of a conventional
output stage the power delivered into
the load can be calculated with the
following simple equation. Assuming
that we neglect the voltage losses in the
output transistors, resistors and capaci-
tor - and bearing in mind that the
calculation is only valid for a sinewave

power P is given by: P = g“r^£> where

U = supply voltage

RL = loudspeaker impedance.

Thus in the case of a 12 V supply and
a 4 ft loudspeaker the output power
would equal 4.5 W. Naturally this is a
ridiculously small figure for a public
address system. However, there is
another type of output stage which is
commonly used in power amplifiers,
and that is the bridge configuration.

In a bridge amplifier, two output stages tapped choke (e.g. a mains transformer
are used, and the input signal is applied with two identical secondaries, the
in antiphase to each of their inputs. The primary not being used). The centre tap
loudspeaker is connected between the is connected to the positive supply rail,
two outputs. The result of employing The change in the current drawn by one
this configuration is that the amplitude transistor from the supply line via the
of the output signal is doubled, and centre tap induces a voltage in the other
the power delivered into the load winding which can be as large as the
(i.e. the loudspeaker) is quadrupled, supply voltage. Thus, per half period of
Thus assuming our 1 2 V supply and the waveform, a signal whose ampli-
4 S2 loudspeaker, an output power tude is twice the supply voltage can
of 18 W is theoretically possible. appear across the loudspeaker, so that

There still remains one further trick the peak to peak amplitude of the
which we can use, and that is to connect loudspeaker voltage can be as much as
a transformer in parallel with the four times the supply voltage (see
loudspeaker. The transformer should, as figure 1). The power delivered into the
far as possible, have two identical loudspeaker is therefore not 18 W, but
windings, and is connected as a centre- 72 W (for a 4 (1 load), and always

Figure 1. This figure illustrates the voltage-
doubling effect of coupling the output stage
to a transformer.

Figure 2. Block diagram of the Stentor.

Figure 3. Complete circuit diagram of the
Stentor. LEDs D2 and D3 set the base bias
voltages of T1 and T2, and should light up
only very faintly or not at all. The same is
true of D1, which is used to set the quiescent
current level of the output stages. The adjust-
ment procedure for the quiescent current is
quite straightforward: an ammeter is connec-
ted in series with the supply line (make sure
that the wiper of P2 is initially turned fully
clockwise and that of P3 fully anticlockwise
- i.e. both wipers are at supply common). P2
is now adjusted until a quiescent current of
0.5 A is obtained. P3 is then adjusted until
the meter reads 1 A.

elektor

ril 1979 -4-17

assuming that the supply was capable of
providing sufficient current, the theor-
etical figure of 144 W could be achieved
with a 2 S2 load. One could even im-
prove on this figure, since a supply
voltage of up to 14 V is not inconceiv-
able.

Block diagram

The block diagram of the Stentor is
shown in figure 2. As can be seen, the
circuit employs the principle of adding
a transformer to increase the output
power. The input signal is first fed to
a summing amplifier, thus permitting
several signals to be mixed together
before being amplified (see figure 5).
The actual amplifier contains two
identical output stages, one of which is
driven directly from the output of the
summing amplifier, the other being
driven via a phase inverter. Since the
two output transistors are biased to
virtually their cut-off points, each
output stage will conduct during alter-
nate half cycles of the input waveform.
The pre-driver stages are formed by
differential amplifiers, whilst the driver
and output transistors are connected as
current amplifiers. To reduce crossover
distortion to a minimum, the output
stages are in fact biased to slightly above
the cut-off point, so that a quiescent
current flows even in the absence of an
input signal. The bias levels can be
varied between 0 and 1 .4 V by the two
potentiometers (P2 and P3 in the circuit
diagram). As explained, the output
transistors are connected to a trans-
former with a centre tap which is
connected to the positive supply line.
Initially, under quiescent conditions,
both ends of the transformer are at
positive supply potential. During one
half cycle of a signal waveform one of
the output transistors, e.g. T7 is turned
on, current will be drawn through the
upper coil causing the collector voltage
of T7 to fall. However the change in
current flowing through the upper coil
will cause an induced voltage in the
lower coil of the opposite polarity to
that across the upper coil, i.e. as the
latter falls from +14 V to zero, the
former will swing up to +28 V (see
figure 1). On the second half cycle of
the input waveform, the process is
reversed. T8 turns on (and T7 turns off)
pulling down the lower coil, and in
doing so induces an opposite voltage in
the upper coil. Thus the total voltage
difference across the loudspeaker on
each half cycle is 2 x supply, and
although the voltage never actually falls
below 0 V, since the current reverses on
alternate half cycles, the loudspeaker
effectively ‘sees’ an AC voltage that
swings between +28 V and -28 V.

Circuit diagram

The complete diagram of the Stentor is
shown in figure 3.

A1 forms the input summing amplifier,
the gain of which is determined by the
ratio R4/R3. The voltage divider net-

Figure 4. Track pattern and component
layout of the p.c.b. for the Stentor (EPS
790701. The output transistors and coupling
transformer are not mounted on the board.
T7 and T8 should be provided with suitably
large heat sinks (taking care to ensure that
they are insulated from the heat sink by
means of mica washers etc.).

Figure 5. This figure illustrates how several
input signals can be summed at the input
of the Stentor. If only one input signal is
used PI can function as a volume control.

Figure 6. Wiring diagram for the Stentor. A
great deal of care should be taken when
wiring up the circuit. The connecting wire
should be as heavy and stiff as possible. It
is strongly recommended that the diagram
shown here is strictly adhered to.

s

— 2*

max. 10 x

79070 -5

work R1/R2 provides a reference
voltage of half supply voltage. This is
applied to the non-inverting input of
the op-amp, so that, in the absence of
an input signal, the output of A1 will
also equal half supply. Any AC voltage
applied to the input of A1 will thus be
inverted and summed with half supply
voltage. A2 is also connected as an
inverting amplifier, so that the signals
fed to the two output stages are in
antiphase.

The output stages, which are identical,
consist of a differential amplifier (T9
and T10 respectively), which is biased
via the current source provided by T1
and the red LED D2 (T2 and D3) and
driver and output transistors T3, T5,
T7 (T4, T6 and T8). Note that it is
important that a red LED must be used
to set the bias voltage of the differential
amplifier, since LEDs of other colours
have different forward voltage charac-
teristics. The collector current of T1
flows through both halves of T9 in
relative proportions which are deter-
mined by the gate voltages of the two
FETs. The input signal is fed via R9 to
the left-hand gate of T9, and this gate
voltage determines how hard T3, T5
and T7 are turned on. However the
current through T7 also flows through
R30, so that the voltage dropped across
this resistor also determines the gate
voltage of the other half of T9, i.e.
negative feedback occurs. The collector
current of T7 is proportional to the
input voltage.

As already mentioned, R13 and D1

4-18 — elektor april 1979

provide a DC voltage of 1.4 V. By
means of P2 and P3 a portion of this
voltage can be used to bias the output
stages such that a small quiescent
current will flow even in the absence of
an input signal. This has the effect of
reducing cross-over distortion.

Construction

Track pattern and component layout of
the printed circuit board for the Stentor
are shown in figure 4. Construction of
the circuit should not present any
special problems, however the value of
certain components will depend upon |
the impedance of the loudspeaker being
used. Thus with a 2 S2 loudspeaker the
2N3055’s (T7/T8) should be replaced
by the slightly more expensive 40411.

The transformer must be able to supply
at least 4 A in the case of a 4 loud-
speaker, and 8 A into a 2 speaker.

It is extremely important that the
terminals of the transformer primary
are well insulated as the turns ratio
between primary and secondaries causes
dangerously high voltages to be induced
on the primary.

A wiring diagram for the Stentor is
shown in figure 6.

Loudspeakers

As was emphasised at the start of the
article, the Stentor cannot lay claim to
hi-fidelity performance. The total distor-
tion of the amplifier is in the region of
10%, although for the most part this is
harmonic distortion which has very
little effect upon the intelligibility of
the spoken word.

Having described how the Stentor
manages to pump large numbers of
Watts into a loudspeaker, it is worth
spending just a little time examining
what type of loudspeakers should be
used. Note the word loudspeakers, for
it would be a waste of time and money
trying to use a single speaker to handle
the output power of the Stentor. A
less expensive and more efficient
approach is to connect a number of
smaller speakers in series/parallel con-
figurations, as shown in figure 7. Most I
loudspeaker manufacturers market a
small diameter speaker costing under
£ 5 and with a power rating of anything
between 5 to 15 W. By mounting the
speakers close together on a panel, or
above one another in a column, a
further improvement in efficiency can
be obtained due to reinforcement
effects.

Care should of course be taken to
ensure that all the speakers are connec-
ted in phase. This can be checked by
briefly touching a battery across the
terminals of the loudspeaker combi-
nation. The speaker cones can be seen
(and felt) to move in or out and the
movement should be in the same
direction for all of them. If one is found
to be incorrect the connections to its
terminals should be reversed. M

TAP switch elektor april 1979 - 4-19

TAP switch

U. Sussbauer

I

I

The advantage of this circuit for a touch
activated switch is that it requires only
one set of contacts and uses only two
inverters, two resistors and a pair of
capacitors. The circuit functions as
follows: At switch on, the input of N1
is low, since C 1 is discharged. Since the
input of N 1 is low, the input of N2
must be high and the output of N2 low,
which of course holds the input of N 1
low — thus the circuit is latched in a
stable state.

j In the meantime capacitor C2 charges
J up, via R2, to logic ‘1’. If the touch
J contacts are now bridged, the logic ‘1’
on C2 is applied to the input of N1
(C2 > C 1 ), taking the output low (and
the output of N2 high). The state of the
Q and Q outputs is thus inverted.
Bridging the contacts again causes C 1 to
discharge into C2 so that the outputs
revert back to their original state. If the
contacts are bridged for longer than the
time constant R2 • C2, then the outputs
will change state again. If the contacts
are permanently bridged, the circuit will
in fact oscillate at a frequency deter-
mined by the above time constant.

With the component values shown, the
contacts should not be bridged for
longer than approx. 1 second. This can
be extended by increasing the value of
C2. H

■ elektor april 1979

assistentor

Everyone must at some time have
had the maddening experience of
standing on a railway platform
and straining anxiously to make
out what the station announcer
has just said — a situation which
provides a good example of the
problems facing P.A. systems used
in environments with high levels
of background noise.

One solution is to employ a
dynamic range compressor to
boost the amplitude of the softer
portions of speech above the
threshold level of the background
noise, without altering the level of
the louder speech passages. When
used in conjunction with the
'Stentor' power amplifier the
result should be a P.A. system
which will make itself heard in
even the noisiest of crowds.

Environments with considerable back-
ground noise (open-air gatherings, factor-
ies etc.) present a particular problem to
P.A. systems. The signal level of speech
varies considerably, so that whilst the
louder portions may be readily audible,
the quieter passages will be swamped.
Unfortunately, simply turning up the
volume of the P.A. system is not an
answer, since increasing the signal level
to the point where the softest portion
of speech is above the threshold level of
the background noise means that the
louder passages will be deafening (and
more than likely distorted by the
amplifier clipping).

The dynamic range of speech (i.e. the
difference between the loudest shout
and the quietest whisper) is around
35 dB. However it is fact that this can
be reduced to approximately 5 dB
without adversely affecting the intelligi-
bility of the speech. Thus it is possible
to amplify the low level speech signals
above the threshold of the background
noise, whilst leaving the high level
signals more or less untouched (see
figure 1). This process is known as
dynamic compression and is the func-
tion of the circuit described here. Since
it can be used in conjunction with the
Stentor power amplifier contained else-
where in this issue, the circuit has been
dubbed the ‘Assistentor’.

However the Assistentor is not just a
dynamic range compressor, it utilises
another interesting feature of human
speech to further reduce the volume
required to render a speech signal
intelligible. Most of the power in a
speech signal is concentrated in the
lower frequencies. However these fre-
quencies are of little importance for the
intelligibility of the signal. For example,
passing a speech signal through a 6 dB
per octave highpass filter with a cut-off
frequency of 1 kHz would reduce the
power of the signal by 77%. However
the intelligibility of the signal remains
virtually unaffected (92% of the orig-

The Assistentor imposes the frequency
response of figure 2 upon the speech
signal. In addition to rolling off the low

frequency signal components, a small
amount of treble cut is also applied.
This has the effect of preventing sibilant
hiss in the speech signal.

Dynamic range compression

The basic principle of a compressor is of
controlled attenuation. The input signal
is fed to the attenuator and then to an
amplifier, the output of which is recti-
fied, smoothed, and used to control the
attenuator. Thus as the input signal
increases, so will the control voltage
applied to the attenuator and hence the
degree of attenuation will be relative to
the input-signal level. The result is that
the overall gain of the circuit becomes
smaller as the input signal level increases.
The rectifier output is smoothed in
order to give a control voltage which
will follow the envelope of the signal
waveform.

A number of different non-linear devices
could be used to control the attenuator,
e.g. FETs, voltage dependent or light
dependent resistors, etc. However a
simple and effective approach is to
employ silicon or germanium diodes. As
can be seen from figure 3, the dynamic
resistance of a diode decreases as the
current through the diode increases.
This fact was utilised in the design of
the Assistentor as shown in the block
diagram of figure 4.

The input signal voltage dropped across
the diode attenuator is determined by
the size of the control current I c .
However this control current is itself
derived from the output of the attenu-
ator (which thus functions as a current-
controlled resistor). As the input voltage
falls, so does the control current through
the diode, with the result that the
dynamic resistance of the diode
increases and more of the signal voltage
is dropped across it. Thus more of the
Input signal is amplified and appears at
the output of the circuit.

The dynamic range of the input signal is
reduced from around 35 dB to approxi-
mately 6 dB, i.e. the quietest passages
of speech are amplified by roughly

assistentor

elektor april 1979 -4-21

Figure la. In an environment with a high
level of background noise the quieter portions
of the speech signal are swamped.

Figure 1b. By compressing the dynamic range
of the speech signal, the quieter passages are
lifted above the threshold of the background
noise, without affecting the amplitude of the
louder passages.

Figure 2. By rolling off the low frequency
response of the speech signal it is possible to
drastically reduce the power of the signal
without impairing intelligibility.

Figure 3. A dynamic compressor can be
obtained by exploiting the non-linear transfer
characteristic of a diode.

Figure 4. Block diagram of the Assistentor.

30 dB, without the louder passages
being affected.

Circuit diagram

The circuit diagram of the Assistentor is
shown in figure 5. As is apparent, not
one, but four diodes, arranged in a
bridge network, are used to form the
current controlled attenuator. The
reason why four diodes are employed is
that the control current also causes a
voltage drop across the diode, which is
superimposed upon the signal voltage.
Variations in this control voltage can
give rise to spurious ‘clicks’. With the
arrangement shown in the circuit
diagram the signal is applied differen-
tially, and is amplified by a differential
amplifier at the output of the attenu-
ator. The voltage produced by the
control current, however, appears in
common mode at the amplifier inputs
and is rejected.

To ensure that the operation of the
attenuator is as symmetrical as possible
and to optimise control signal rejection,
the diodes should ideally be matched
pairs. The best solution is to use a
monolithic bridge rectifier, of the type
employed in power supply circuits.
Before the input signal is fed to the
diode attenuator, it is amplified by T1
and Al. Several RC networks around
A1 ensure that the circuit has the fre-
quency response shown in figure 2. The
gain of this stage can be varied by means
of PI.

A2 functions as an inverting amplifier so
that antiphase versions of the amplified
input signal are fed to the controlled
attenuator. A similar configuration is
present at the output of the attenuator:
A3 amplifies the compressed signal,
whilst A4 is again connected as a phase
inverter. Antiphase versions of the com-
pressed signal are therefore available,

assistentor

elektor april 1979 - 4-23

Figure 5. Complete circuit diagram. The con-
trolled attenuator is formed by a diode bridge
network. The output signal is amplified
differentially in order to reject the common
mode control voltage.

Figure 6. Track pattern and component
layout of the printed circuit board for the
Assistentor (EPS 790711. A number of
precautions are necessary to reduce r.f. pick-
up — see text.

Figure 7. In a number of applications it is
useful to provide the Assistentor with a by-
pass switch. This can be connected as shown.

Figure 8. The transfer characteristic of the
Assistentor with PI set for maximum sensi-
tivity. As can be seen, full dynamic com-
pression occurs at an input voltage of only

permitting full-wave rectification via T2
and T3. The output of these transistors
is used to charge CIS, the smoothing
capacitor. The voltage on this capacitor
determines the control current at the
emitter of T4. The discharge time of
this capacitor, and hence the decay time
constant of the compressor, can be
varied by means of P2.

The Assistentor has two outputs. Out-
put A provides a signal level suitable for
use with power amplifiers such as the
Stentor, whilst output B provides a low
level signal. The amplitude of both out-
put signals can be adjusted by means of
potentiometers P3 and P4 respectively.
Since the circuit has a considerable gain
(up to approximately 86 dB), it is
susceptible to r.f. pick-up. For this
reason the input, outputs and supply
line are decoupled with the aid of ferrite
beads LI . . . L4 and a sprinkling of 1 n
capacitors. This precautionary measure
is particularly important if the circuit is
to be used in the vicinity of an r.f.
receiver or transmitter.

7

Q Assistentor ^

V

\o*

Construction and use

The printed circuit board for the
Assistentor is shown in figure 6. The
ferrite beads LI . . . L4 are printed as
coils on the component overlay. The
desired self-inductance can be obtained
by making a wire link using insulated
wire and winding the link through the
ferrite bead three or four times.

Any type of bridge rectifier will suffice
for Bl. However it is best not to use
versions with too high a current rating,
i.e. a 100 mA type should be preferred
to one of several amps. In the circuit
diagram and parts list the B30C100 is
mentioned, however this is a nominal
value, which only serves' as a rough
guideline.

The sensitivity of the circuit can be
adjusted by means of preset P 1 . If the
range provided by this potentiometer
proves insufficient, then the value of
R 1 5 can be altered. Increasing the value
of this resistor increases the sensitiviy
(and also the level of both output
signals).

The Assistentor is not restricted to use
with the Stentor or other power ampli-
fiers of that nature. The provision of a
low level output means that it can be
employed with dictaphones, intercoms,
office- or school loudspeaker systems,
or in amateur radio installations to
achieve the maximum modulation depth
without overmodulating, and thus
obtain maximum range. Figure 7 shows
how the Assistentor can be provided
with a switch, in order to allow the
circuit to be switched in and out of
operation. This facility is indispensable
if using the Assistentor in conjunction
with an amateur radio transmitter.
Finally, figure 8 shows the transfer
characteristic of the Assistentor with P 1
set for maximum sensitivity. The volt-
age level at output B is given for various
values of the input voltage. As can be
seen full-scale compression occurs at as
low an input voltage as 1 mV.

Lit.:

Selektor, Elektor May 1978,

P- 5-02. H

zero j>F
screen

In a large number of applications it is
necessary to screen signal leads from
external interference (mains hum, r.f.
pick-up etc.). Unfortunately, between
the signal-carrying conductors and the
screening metal sheath there is inevitably
a certain capacitance. The longer the
leads the greater this capacitance will
be.

It is possible that the capacitance of
cables can adversely affect their high
frequency transfer characteristic, one
example being the influence that the
screened lead between pick-up and
preamp has on the frequency response
of a moving coil cartridge. Although the
capacitance cannot be eliminated,

fortunately there is a way to overcome
its effect. If we ensure that the load on
the capacitance remains constant (e.g. is
always zero) or at any rate does not
depend upon the AC signal voltage, then
the disadvantage of screening can be
removed.

How is this done? The screen is con-
nected to the signal conductor via a
voltage follower, the output impedance
of which is considerably lower than the
impedance of the cable. The connection
can be made at either end of the signal
lead (the other end of the screen is left
floating). The result is that the screen
no longer has any capacitance effect,
since the signal lead and screen are both

carrying the same AC voltage. The cable
remains effectively screened, however,
since the opamp has an extremely low
output impedance.

If the screen was also used as an earth
return, then an alternative means of
realising the earth lead must be sought.
There are several possibilities: one can
use an extra conductor inside, or an
extra screen outside the neutralised
screen, or both. In the latter case the
screen is connected to case earth both
at the ‘transmitter’ and ‘receiver’ end,
and thus no longer functions as a return
lead. K

4-24 - elektor april 1979

quiz mai
ster quiz
master c
iuiz mas
ter quiz
master g
[uiz mas

In many quiz games it is important
to be able to determine who is the
first to reply to a question. To
avoid disputes and unseemly
family arguments the simplest
approach is to use an impartial
electronic 'quizmaster'.

Figure 1. Circuit diagram of the quiz-master.

Figure 2. Using this simple arrangement of
gates it is possible to construct a six-input
(excluding the reset input) NAND.

Figure 3. Track pattern and component
layout for the quiz-master printed circuit
board (EPS 79033).

quizmaster

elektor april 1979 -4-25

The diagram of a suitable circuit is
shown in figure 1. The arrangement
shown is for 4 players, however the
design can be extended for any number
of competitors. The operation of the
circuit is extremely simple.

Each of the modules shown in figure 1
consists of a flip-flop which is set by
means of a pushbutton switch. The
Q output of each of the flip-flops is
connected via a NAND gate to the reset
input of each of the other flip-flops. As
soon as one flip-flop is set, the remaining
flip-flops will therefore be inhibited,
since their reset inputs are held high.
The Q output of the flip-flop which has
been set goes high, turning on the as-
sociated transistor and lighting a LED,
thus providing a visual indication of
which pushbutton was pressed first. The
circuit is reset for the next round by
pressing S5.

The RC network at the input of each
module is simply a pulse former which
prevents undefined logic states occurring
in the event of both the set and reset
inputs of the flip-flop being high at the
same time.

It is a simple matter to extend the
circuit to accommodate more than four
competitors. For each additional person
an extra flip-flop circuit is needed and
the number of inputs to the NAND gates
must be increased accordingly.

The printed circuit board was designed
to use a CD 4068 which is an 8-input
NAND gate. However, it may be more
convenient to use a 4-input NAND
(MC 14012, CD 4012) and a simple
method of extending the number of
inputs is shown in figure 2. This con-
figuration is equivalent to a 6-input
NAND, excluding the reset input, and
would therefore be suitable for seven
competitors. It must be remembered
that any unused inputs should be con-
nected to the positive supply rail.
Construction of the circuit should not
present any problems. If desired, the
pushbutton switches can be connected
to the board via long lengths of wire.
The relatively low impedance of the
pulse former networks ensure that hum
and other types of inductive inter-
ference are sufficiently suppressed. If
need be, the resistor values of these
networks can be reduced to 1 k. H

4-26 - elektor april 1979

lilding the TV-games computer

building
the TV-games
computer

Elsewhere in this issue, an
introductory article about the
TV games computer describes
what the finished unit looks like
and what it can do. In this article,
our main interest is how to build
the unit. After a brief description
of the circuit, we will concentrate
on the constructional details and
calibration procedure.

One glance at the main circuit diagram
is probably sufficient to scare off all but
the hardiest electronics enthusiasts.
However, the block diagram is not
nearly as bad (figure 1) and, once it has
been understood, it will serve as a guide
through the main circuit.

As explained in the introductory article,
the ‘brain’ of the TV games computer is
the microprocessor chip, or ‘CPU’. By
passing ‘switching signals’ along the
‘address bus’ ( 1 3 wires), it can operate
all other sections of the unit as required.
Information is passed from one unit to
another along the ‘data bus’ (8 wires);
finally, several special-purpose control
signals are connected direct from the
CPU to the unit(s) involved.

Without a memory, a brain is fairly
helpless. Three distinct types of mem-
ory are available in this unit: The Read
Only Memory (ROM), containing the
pre-programmed ‘monitor software’; the
Random Access Memory (RAM), that is
used for storing the actual ‘game’
program; and, finally, a cassette or tape
recorder for long-term storage of as
many different ‘game’ programs as may
be desired. The type of memory that is
to be used at any particular moment is
selected (under CPU control) by the
‘address decoder’; the exact part of that
memory from which data is to be
retrieved (or in which it must be stored)
is selected by the CPU itself, via the
address bus.

Since most tape and cassette recorders
are designed for audio work, using them
to store digital signals requires some
careful signal handling. The digital
output from the computer to the tape
must be AC-coupled and filtered to
remove the extreme high-frequency
components; the input from the tape
to the computer must be boosted in
level and ‘cleaned up’ to produce a
recognisable digital signal. These oper-
ations are performed by the section
labelled ‘cassette interface’ in figure 1 .
The sections described so far are com-
mon to virtually any computer system.
‘Intelligence’ - the CPU - and ‘mem-
ory’. It may be noted, in passing, that
precisely this part of the system is
shown on the left-hand page of the main
circuit diagram. However, the TV games
computer has still to be connected to
the ‘player’ controls (joy-sticks and
keyboard), the TV set and a loud-

speaker.

First, the controls. The joy-sticks, being
potentiometers, are basically analog
devices. To adapt them to the otherwise
digital system, some kind of analog-to-
digital conversion is required: a ‘joy-
stick interface’, which is actually part
of the PVI. However, to save space
only one ‘interface’ is actually available
and so only one joy-stick can be dealt
with at a time. For this reason, a ‘joy-
stick selector’ is included, to switch to
and fro between the two controls. This
unit is, of course, under CPU control
(via a separate connection); it is also
connected to the ‘keyboard interface’.
The latter unit feeds the data from the
keyboard onto the data bus — again
under CPU control (via the address bus).
So much for the inputs. The outputs —
to TV and loudspeaker — are rather
more complicated. Happily, most of the
work is done by a single IC: the PVI,
or Programmable Video Interface. This
unit is comparable to a ‘slave’ micro-
computer: under CPU control it stores
and supplies data as required; it detects
certain situations (e.g. inter-object colli-
sions); based on the data stored, it
‘creates’ the corresponding picture and
sound signals. In fact, the ‘sound’ in
particular is simplicity itself. A single
buffer stage, connected to the corre-
sponding PVI output, can drive a small
loudspeaker.

To create the ‘picture’ the PVI must
enlist the aid of a few ancillary circuits.
A crystal-controlled oscillator produces
the basic timing signals. These are fed,
through frequency dividers, to another
useful IC: the ‘Universal Sync Gener-
ator’ (USG). This unit takes care of all
the sync signals required for a modern
colour TV set, and produces some
additional synchronising signals for
other parts of the circuit almost as a
by-product.

One set of outputs from the USG is
fed to the PVI, to tell it what part of
the picture is actually being ‘written’
at any particular moment. Based in part
on this information, the PVI produces
a group of output signals that determine
what colour must be displayed at that
point in the picture, in order to repro-
duce the required display of objects,
background and score. These outputs
are fed, via a ‘gating’ circuit controlled
by the USG, to the final section: the

building the TV-gar

elektor april 1979 - 4-27

‘digital video summer’. This unit does
exactly what its name implies: it sums
the outputs from the crystal oscillator,
the USG and the PVI to produce the
total video output. ‘Summing’, in this
case, should be taken in the broadest
possible sense: it includes frequency
division, gating and level matching.
Having a video output is all very well,
but most TV sets only have UHF or
VHF inputs. This problem is quickly
solved, by adding a UHF/VHF modu-
lator.

The circuit

Having discussed the block diagram, we
can now risk a quick look at the main
circuit diagram (figure 2). However,
there is little point in going into great
detail. The main point to note is that
the layout corresponds, by and large, to
that of the block diagram: the various
sections are drawn in the same relative
positions.

The CPU (IC1) is at the left; the address
and data busses run along the top. The
address decoder for the memory and
input/output selectors (IC6 and IC7),
the ROM (1C2) and the RAM (IC13 . . .
IC28) require little explanation. The
only point to watch is that 2 1 1 2’s used
in the RAM must be 450 ns versions (or
faster). The input- and output selectors
(IC8 and IC9, respectively) offer eight
‘serial’ inputs and eight outputs. How-
ever, only one of each is used in the
basic unit; the other seven in- and out-
puts are available for other external
devices if required. The ‘output filter’
in the cassette interface consists of three
resistors and two capacitors; the ‘input
buffer’ uses an opamp to boost the
signal to TTL level.

So much for the first half of the circuit!
This ‘thumbnail’ description should
suffice to give a general idea of ‘what
does what’. A fuller and more detailed
explanation is given in the supplemen-
tary information supplied with the

printed circuit board.

The second half of the circuit is rather
more complicated at first sight (well,
even at second and third sight . . .).
However, it is not too difficult to locate
a few important sections, by referring
to the block diagram. The ‘heart’ of
this part of the circuit is the PVI (IC3):
as mentioned earlier, it is almost a
‘slave’ microcomputer in its own right.
The fact that it operates in close collab-
oration with the CPU is readily appar-
ent: it is the only subsection that is
connected directly to virtually all the
address and data lines.

The PVI is flanked (quite litterally, in
the circuit) by the joystick selector
(1C 10), the ‘loudspeaker interface’ (Tl)
and the keyboard with its interface
(IC 1 1 and IC12). None of these merit
any detailed discussion at this point.
The keyboard (or keyboards, depending
on how you look at it - them?) will
be explained later, from the user’s
point of view.

Now for the remainder of the circuit.
To be honest, this type of circuit should
either be described in detail or not at
all . . . However, we will attempt to give
a very rough outline, without Qs, Qs,
logic ones and zeroes.

Immediately below the PVI, a crystal
oscillator (IC31) is used to generate the
main timing signals. One of these goes
down and around, passes through a
divider stage consisting of IC32, IC35
and IC36, and finishes up as the ‘clock’
input signal for IC4, the ‘USG’. This
‘Universal Sync Generator’ is more
important than its relative size in the
circuit might lead one to expect: it
produces the complicated synchronising
signals required for a modem (PAL)
colour TV set. Furthermore, it produces
reference signals that are used by both
the PVI and the CPU; finally, it ‘gates’
the video outputs from the PVI (i.e.
turns them on and off, as required) via
an intriguing selection of NANDs,

inverters and EXORs (IC29, IC30 and
IC40).

What remains in the circuit was all
lumped together in the block diagram
as ‘Digital video summer’. This section
(consisting of IC33, IC34, IC38, IC39
and assorted NANDs and inverters)
combines signals from the crystal
oscillator, the gating network at the
PVI outputs and the USG to produce
the final video output. The correct
relative levels of the various signals are
determined by the resistor network in
the ‘summer’ (R54 . . . R62).

The complete circuit shown in figure 2
is contained on a single printed circuit
board. More on this later. However,
two further units are required: a power
supply and (in most cases) a UHF/VHF
modulator.

Power supply

Any stabilised supply, capable of
delivering 5V at 2A, is suitable. A
simple circuit is shown in figure 3.
Although this configuration may seem
rather peculiar, especially where Tl and
T2 are concerned, the principle is quite
straightforward. If the load current
increases, the integrated voltage regu-
lator (IC1) will attempt to supply this
current itself. However, in so doing it
will increase the voltage drop across R2,
thereby turning on T2 - which then
proceeds to deliver the bulk of the
current. In the event of a short circuit
occurring, Tl limits the current through
T2 to a safe value and internal protec-
tion circuits maintain the dissipation in
the IC within its limits.

A suitable printed circuit board is
shown in figure 4.

UHF/VHF modulator

Or rather: ‘VHF/UHF TV-modulator,
see Elektor 42, October 1978’. Why
design a new one if an existing circuit
is good enough?

The circuit is shown in figure 5, and the

printed circuit board in figure 6. Re- Figure 2. Complete
peating the circuit description would be circuit,
rather a waste of space, so we can
quickly proceed to the

Construction details.

The TV games computer is built up
from four basic units: the main circuit,
keyboard (figure 9), power supply and
UHF/VHF modulator. The wiring be-
tween these units (and various external

diagram of the main odds and ends: joy-sticks, loudspeaker)
is shown in figure 7.

The p.c. board for the main circuit
(figure 8) requires some comment. It
is a double-sided p.c.b. with plated-
through holes. A prime example of
modem technology with, regrettably,
the associated ‘teething problems’.
Technology has not yet reached the
point where plated-through holes are
100% reliable (at a reasonable price,

building the TV-games computer

elektor april 1979 -4-29

that is) and, as an interim solution,
manufacturers often simply mount
all components on the board and reject
any circuits that don’t work. This is
rather unsatisfactory for the home
constructor, and he is more inclined
to ‘trouble-shoot’. For complicated
circuits like the TV games computer,
this can be extremely time-consuming.
For this reason, it is advisable to check
the board before mounting the com-

ponents. A first, visual, check is poss-
ible by holding the board up to the
light and looking through the holes:
the plating should be clearly visible.
To make assurance doubly sure, each
hole can be tested individually with
a multimeter: with one probe on each
side of the board, the resistance should
be zero.

When mounting the components, it is
strongly advised to use a suitable

miniature soldering iron and first-
class IC sockets. In a unit like this,
tracing dud contacts afterwards can be
a harrowing experience.

The main board has a large number of
in- and output connections, most of
which however remain unused in the
basic tv games computer. The connec-
tions between main board and keyboard
(see figure 7) are clearly numbered on
both boards. Note that the two wire

Figure 3. A suitable 5V power supply.

Figure 4. Printed circuit board and com-
ponent layout for the power supply (EPS
79073-1).

Figure 5. The UHF/VHF modulator, orig-
inally described in Elektor October 1978.
Note that IC1 is not required in this appli-
cation.

Figure 6. Printed circuit board and com-
ponent layout for the UHF/VHF modulator
(EPS 9967).

links shown as dotted lines on the
keyboard p.c.b. should not be mounted
in this case.

The key numbers shown on the com-
ponent layout for the keyboard corre-
spond to the actual addresses of the
keys. However, for normal use the
indications shown in figure 10 are prefer-
able, since they correspond to normal
use in the monitor program. If the
keyboard was only to be used in this
application, it would have been more
logical to run all the keys together into
one 7x4 block; however, the keyboard
will often be used as two separate small

keyboards for two players. For this
reason, the suggested layout is designed
for easy ‘cutting along the dotted line’.
The only constructional comment
regarding the power supply is that all
due care should be taken! If, due to a
fault occurring at a later date, the
supply voltage suddenly jumps to way
over 5V, several expensive ICs may die
a sudden death. For the same reason,
the supply voltage must be adjusted to
5V before connecting it to the rest of
the unit. A drop of laquer on the
potentiometer will not only keep it
from sliding off position, it will also

building the TV-games computer

elektor april 1979 - 4 31

serve as warning not to touch it at a
later date.

The UHF/VHF modulator must be
adequately screened. Of course. Modu-
lators should always be mounted in
metal boxes. This unit can be powered
from the main +5V supply, so the
regulator (IC1) can be omitted and the
holes in the board for its two outer pins
can be bridged with a wire link.

Calibration procedure

With everything neatly built and all
interconnections made, and after a
final visual check of the wiring, it is
now time to switch on. Note that, as
stated earlier, the power supply must
already be tested and adjusted to 5 V.
The calibration procedure is simplicity
itself. In fact, the modulator contains
exactly the same number of adjust-

ment points as the rest of the circuit —
two, to be precise!

UHF/VHF modulator
Set PI to its mid-position and tune the
TV set to one of the harmonics of the
carrier. When the carrier is picked up,
the snow-storm effect on the screen
of the TV set will disappear.

Turn P2 up to maximum.

This completes the initial adjustment.
Some ‘touching up’ will be dealt with
later.

Main circuit

Operate the reset and start keys. After
correct adjustment, this should cause
a blue screen to appear with four yellow
letters at the lower left-hand comer.

The only adjustment points in the
main circuit are PI and C9 in the crystal

5

12 ...15 V

D1 = 1N4148

oscillator. These can be ‘calibrated’
by looking at the picture:

- if PI is incorrectly adjusted, the
oscillator will not run at all, in which
case no picture will appear. The
simplest adjustment procedure is to
turn PI a little bit further than
necessary to obtain a picture.

- C9 determines the oscillator fre-
quency, and incorrect adjustment
will lead to poor colour or even no
colour at all.

Final touches

Having obtained a picture, it becomes
a simple matter to adjust for maximum
picture quality:

The TV set is tuned to the sideband that
gives the best picture; if tuned to the
wrong sideband the picture will tend to
appear negative.

If the picture lacks vertical synchronis-

Parts list for
UHF/VHF modulator

Resistors:

R1 = 33 k
R2 = 22 k
R3.R9 = 470 fl
R4 = 1 k
R5 = 220 «

R6 = 270 n
R7 = 150 n
R8 = 6k8
R10.R11 = 100 n
R12 = 1k5
R13 = 68 St
PI = 2k5 (2k2)

preset potentiometer
P2 = 1 k preset potentiometer

Capacitors:

Cl ,C7 = 33 p
C2 = 120 p
C3.C4.C5 = 8p2
C6 - 22 p

C8.C9 = 1 <x/16 V tantalum
Semiconductors:

T1 ,T2 = BF 194, BF 195, BF 254,
BF 255, BF 494, BF 495
T3 - BFY 90
D1 = 1N4148

IC1 - not required (see text)

Miscellaneous:

LI - 1 pH
XI = crystal,

27 MHz approximately.

Figure 7. Interconnection between the various

Figure 8. Printed circuit board and com-
ponent layout for the main circuit (EPS
79073). The board is double-sided with
plated-through holes. Note that it is repro-
duced here at 70% of actual size.

Semiconductors:

IC1 - 2650A ISignetics)

IC2 = 2616 (Signetics)

IC3 = 2636 (Signetics)

IC4 = 2621 (Signetics)

IC5 - LM339 (National Semiconductor)

IC6 = 74LS1 39

IC7 ■ 74LS138

IC8.IC38.IC39 = 74LS251

IC9 = CD 4099

IC10 = CD 4053

IC11 = 74LS156

IC1 2 = 74LS258

IC13 . . . IC28 = MM2112-4

(450 ns access time)

IC29 » 74LS08

IC30 = 74LS05

IC31 = 74LS04

IC32 = 74LS86

IC33.IC35.IC36 = 74LS1 13

IC34 = 74LS109

IC37 = 74LS00
IC40 = 74LS136
IC41 - 74LS10
T1 -BC517

Resistors:

R1 . . . R21.R25.R33.R35.R55- 10 k
R22,R23,R26,R27,R30,R31 ,R39 . . . R42
R46 . . . R53.R57 = 4k7
R24.R71 = 15 k
R28 = 470 k

R29.R44.R45.R56.R67 = 2k2
R32.R34.R70 = 47 k
R36,R37,R38,R62,R64,R65.R66,R69
= 1 k

R43 = 33 k
R54 = 1 k5
R58.R59 = 820 «

R60 = 680 12
R61 = 470 n
R63 = 100 k
R68 = 220 SI

Capacitors:

Cl .Cl 5.C16.C1 7.C18.C20 = 1 50 n MKH

C2.C4.C1 4- 1 n

C3 = 3n3

C5 = 3n9

C6.C7 - 47 n

C8 = 470 p

C9 = 0 . . . 22 p trimmer
C10.C11 - 68 p (cer.)

Cl 2= 56 p (cer.)

Cl 3 = 220 p/6 V
C19 = 100 m/6 V

Sundries:

Xtal: 8.867 MHz
Loudspeaker: 100 n/500 mW
2 Joysticks (P3 . . . P6): potentiometer
values 680 k.

28 'digitast' switches (for the keyboard)

9 0999999 99 9 0 99990 9 90 0 990 9

ation (i.e. rolls), or if some local broad- suitable in most cases. However, for program was custom-designed (by Phi-

cast transmitter interferes with the the odd recorder the output signal lips) for this application. However,

picture, PI can be readjusted slightly level may be too high or too low, in initial contacts with several major

and the TV set retuned accordingly. which case the value of R70 can be component suppliers have shown that

P2 on the modulator board can be used altered accordingly. In some cases there is considerable interest in supply -

to adjust the contrast. (excuse the pun) problems may also ing complete kits, which will include the

C9 on the main board influences the occur if the cassette recorder is placed factory-programmed ROM. H

colour above all else. Both C9 and PI too close to the TV set. The solution

should be adjusted for optimum picture is, of course, simple: move the recorder

quality. further away.

The components used in the circuit are
Final notes admittedly anything but standard. In

The cassette interface should prove particular, the ROM with the correct

etektor april 1979 — 4-35

lililiiil:

Spectrum analyzer with
10 Hz resolution to
22 GHz

Using microprocessor control, the
Model 8S66A Spectrum Analyzer
from Hewlett-Packard gives the
user the ability to more
accurately resolve closely spaced
signals over a wider range of
amplitudes than previously
possible. Using the Hz resolution
bandwidth, the analyzer’s
sensitivity is -137 dBm to 1 GHz,
-134 dBm to 5.8 GHz, -15 dBm
at 22 GHz.

This high sensitivity includes the
presence of built-in pre-selection
from 2 to 22 GHz. Related
attributes are 80 dB dynamic
range and full frequency range
amplitude of t 2.2 dB. This
performance permits direct
measurement on microwave
signals of line-related sidebands
that are 50 dB down. In addition,
frequency accuracy virtually that
of the internal frequency
reference error ( 1 x 10"’ per day)
can be achieved. This performance
is complemented by features that
make the 8566A extremely easy
to operate, and easy to set up for
automatic operation with HP-IB
(IEEE-488).

The power and convenience of
the 8566A’s microcomputer-
based controls and CRT readout
simplify and speed use in so many
ways that previously-impractical
analyses now become routine.

The fact that harmonic mixing is
employed to extend frequency
coverage to 22 GHz has been
made transparent to the operator.
Functions, such as center or
start/stop frequencies, frequency
span, and amplitude level may be
keyboard-set, varied with ‘analog’
knob (actually a rotary pulse-
generator), or incrementally
keystepped. Other operating
parameters such as resolution
bandwidth, sweep time, video
filter and rf attenuation are
automatically coupled to the
entered data for proper signal
display.

However these parameters can
also be manually entered at
different values depending upon
the application.

A tunable marker in the 8566A
makes basic measurements more
accurately and more easily by
directly measuring a signal, or by
speeding the process of
magnifying the portion of the
spectrum to be analyzed. With
the marker set to the signal peak,
signal amplitude and frequency
are numerically displayed on the
CRT. A second marker makes
relative measurements instantly
available, with numerical display
of the difference in amplitude

and frequency between the two
markers. This is useful for
modulation and distortion
measurements.

Hewlett-Packard Ltd.,

King Street Lane,

Winnersh, Wokingham,

Berkshire RGU SAR, England.

(1096 M)

Electronic baseball

An electronic baseball game
which simulates all the functions
of the real game in a self-
contained, battery-operated unit
has been introduced by Micro
Electronics Ltd. The micro-
processor-controlled game, which
uses light-emitting-diode displays
to simulate the movements of
players and ball and to keep
score, allows players to use their
own offensive and defensive skills
and strategy to influence the

In operation, a player chooses the
desired speed of pitch (slow,
normal or fast) and then presses
the ‘pitch’ button, which starts a
scoreboard display cycling
through all possible events at the
desired speed. When the ‘bat’
button is pressed, the cycling of
events stops, and a crowd cheer
(for a ‘hit’) or a buzz (for an ‘out’)
is heard.

In addition, players can introduce
offensive or defensive options in
the form of ‘pinch hitters’ or
‘relief pitchers’, respectively. The
control circuitry is designed so

that the three pinch hitters have a
higher ‘batting average’

(400 against 270 for the normal
team players), and the two relief
pitchers, although more
‘powerful’ to start with, are
designed to ‘tyre’ to the original
pitcher’s effectiveness after five
batters. There are several ways in
which players can use their skill:
the pitcher has three choices of
speeds, giving him the
opportunity to analyse the batter
to see whether he hits fast or slow
balls, and the batter can time his
pressing of the ‘bat’ button (with
practice) to obtain more and
better hits.

An alphanumeric display
scoreboard provides a continual
display of home and visitor score,
number of relief pitchers and
pinch hitters remaining, event
counter (showing score of present
batter), inning counter, ‘at bat’
indicator (showing which side is
in), number of ‘outs’, and
indicators to show when pinch
hitters or relief pitchers are
introduced.

The game is mounted in a sturdy
plastic case with a representation
of a baseball diamond, with
individual lights showing the
location of the base runners. An
input is provided for an external
9 V DC supply.

Micro Electronics Limited,
Consumer Products Division,

766 Finchley Road,

London, NW 11 7TH England.

market

4-36 — elektor april 1979

Automatic Bridge

The Wayne Kerr B605 Automatic
Component Bridge is a new
microprocessor-based, fully
automatic L, C, R, Q and D
measuring instrument featuring
AUTORANGE and AUTOTRIM
with 0.1% accuracy over a wide
dynamic range. It will provide fast
measurements with guaranteed
reliable performance and easy-to-
use operational characteristics and
yet it is still in the medium price
range for laboratory-standard test
instruments.

The instrument is compact with
simple frontpanel controls. Once
an ‘unkown’ component is
connected to the test terminals,
the LCD readout performs
interactively with the user to
guide him/her through a very
short and simple procedure to
arrive at a correct measurement.
The display offers readings with
resolution of 0.001 pF on the
C-range. Consistency is achieved
by using a design requiring only
one 'standard' component and all
the circuitry is referenced to this
single resistor before each
measurement sequence. The basic
accuracy of the B605 is 0.1%
which is guaranteed across the
whole impedance range of the
bridge and is achieved within the
300 ms taken for a measurement
cycle.

The measurement range is broad
enough to suit all conceivable
component measurements, and
there is a choice of three front-
panel selectable measurement
frequencies; 100 Hz being
available for electrolytic
capacitors and large inductors;

1 kHz as a general-purpose test
frequency ; and 10 kHz for low
value components such as very
small capacitors. The display
instructs the user when ‘AUTO
TRIM’ or ‘AUTO RANGE’
adjustments are required, and
clearly states all results in the
most practical units.

AUTOTRIM is carried out under
MPU control, fully automatically.
Once the frontpanel control has
been pressed, the machine takes
four readings on each of the three
operating frequencies - stores the
results - and then averages them,
thus ensuring that residual trim
errors do not affect the results.
Pressing the AUTORANGE
button automatically selects the
correct impedance range for the

component being measured. This
range is then held even after the
component is removed, and thus
a batch of similar components can
be quickly measured without
further range changes.

The B605 terminals are suitable
for 2, 3 or 4-terminal measure-
ments, but a small matching
adaptor, the JU4, provides for
single-handed loading and removal
of components with axial, radial
or preformed leads. On the top of
the JU4, two inline slots give
positive connection for parallel
wires, and on the front of the
unit, spring-loaded jaws (opened
by depressing the platform)
provide firm connection for axial
wired components.

Wilmot Breeden Electronics
Limited,

Durban Road Bognor Regis West
Sussex P022 9RL, England.

(1121 M)

Transistor and
diode tester

It is claimed that the use of multi-
meters for the in-circuit testing of
transistors and diodes is likely to
become outdated with the devel-
opment of a new tester by
Ampmace.

The swamping effect of parallel
resistors which has often led to
ambiguous readings in the past
has been overcome in this tester
by the feeding of a sufficiently
large, but finely controlled,
current into the junction under
test.

Audio signals replace the
conventional visual indications of
the performance of the junction
under test, making the tester
simple enough for use by
unskilled personnel. A steady
tone, warbled tone or silence
indicate normal junction, short-
circuit junction or open circuit
junction respectively on diodes
and transistors in or out of circuit.
This new piece of pocket-sized

equipment with its automatic
on/off circuit for long battery
life, and built-in protection up to
1,000 V AC or DC, has added
versatility in that it may also be
used as a continuity tester.
Ampmace Limited,

Unit 96, Somers Road,

Rugby, Warwicks, England.

(1122 Ml

Verorak cabinet joining
kit

Joining together two or more
Verorak (R) 19" racks or cabinets
of the same size has now been
simplified with the introduction
of joining kit AB067 from Vero
Electronics. The kit is suitable for
all 1 2 types of Verorak currently
available and joining time is less
than ten minutes with a single
spanner being the only tool
required. Once joined the racks
have an attractive appearance and
all round sealing to prevent the
ingress of dust or moisture can be
achieved with the sealing tape
included in the kit.

Veto Electronics Ltd.,

Chandler's Ford Industrial Estate,
Eastleigh,

Hampshire, S05 3ZR, England.

(1 127 M)

Teletext decoder

It is comforting to know that
when television ‘addicts’ become
bored with the daily problems
experienced by people living in
dreary streets and cardboard
motels, or when the ‘feature’ film
is dated 1930 (!), an interesting
and informative alternative is
available from the television
media - in the form of news,
weather reports, stock-market
information etc. presented as
magazine-type pages - Teletext.
Although becoming less of a
rarity, television sets with built-in
Teletext decoders are nevertheless
rather expensive luxury items.

However, an ‘add-on’ unit,
making the system less costly,
has been produced by Radofin
Electronics Ltd. This unit can be
plugged directly into most
domestic television receivers and
once set up the decoder will
switch instantly from normal
programmes to Teletext at the
touch of a button. With features
such as double-height character
facility, black and white and true
PAL colour, it meets the latest
broadcast specifications by the
BBC and IBA. The decoder has
remotely controlled push-button
page change and is able to receive
both CEEFAX and ORACLE
transmissions.

Radofin Electronics (U.K.) Ltd.,
London W6, England.

(1128 M)

On/off dimmer

The use of micro-electronics in
the traditional dimmer switch has
resulted in a big jump forward in
the concept and design of these
popular products.

This dimmer has no moving parts
and to dim or raise the light level
the user merely places a finger on
the stainless steel centrcplatc.
When the finger is taken away the
light remains at the existing level.
The light can be switched on or
off by a rapid touch of the same

The Davi Litetouch also
incorporates a simple to replace
fuse which provides protection
for the electronics while RF inter-
ference has been minimised to
conform with pending legislation.
AvaOabe in a choice of two
colours (black or white) the
dimmer has a pleasant appearance
and will fit most U.K. and
continental type wall boxes. The
forward projection from the wall

Davi Marketing Ltd.,

46a High Street, Stamford,
Lincolnshire, England.

(1123 M)

advertisement

Elektor April 1979 - UK 18

WHO and WHOM

A directory of electronic component suppliers to

IF THERE IS A COMPONENT SHOP IN YOUR AREA NOT LISTED BELOW^^AS^^^J^NOY^

Spectron Electronics
Manchester Ltd

7. Oldfield Road. Salford.

Greater Manchester
Tel 061 834 4583

Elektor readers

AREA 1

ACE Mailtromx Ltd.,

Tootal Street. Wakefield.

W. Yorks. WF1 5JR.

Tel: 0924-250375

A. Marshall (London) Ltd

85 West Regent St..

Glasgow G2 2QD
Tel: 041-332-4133

Aitken Bros. & Co.

35 High Bridge.
Newcastle upon Tyne.
NE1 1 EW
Tel: 0632-26729

The Amateur Radio Shop

4 Cross Church Street,
Huddersfield, HD1 3PT.

Tel: 20774

Bell's Television Services

190. Kings Road.

Harrogate. Yorkshire.

Tel. (0423) 55885

Derwent Radio

5 Columbos Ravine,
Scarborough, N.Yorkshire.
Tel: (0723)63982
Electrovalue Ltd.

680 Burnage Lane,

Manchester M19 1NA
Tel: 061-432-4945

Gamma Enterprises

373 Blackness Road.

Dundee. DD2 1TL
Phone Dundee 643061
Greenbank Electronics
94 New Chester Road,

New Ferry, Wirral,

Merseyside L62 5 AG.

Tel . 051-645 3391

Harrogate Electronic
Services,

25. Regency Parade,
Harrogate.

Shudehill Supply Co. Ltd.

53 Shudehill.

Manchester M4 4AW,

Tel: 061 8341449

Tandy,

Flottergate,

Riverhead Centre, Grimsby.
Tel: 57780

M & B Components Ltd

86 Bishop Gate Street.

Leeds 1 LSI 4B8.

Tel 0532 35649

AREA 3

Progressive Radio,

93 Oale Street.

Liverpool 2 25D
Tel (051) 2360982

Electronic Assembly
Services,

Bright Street Works,

Bury, Lancs. BL9 6AQ

AREA 2

Strutt Electrical &

Mechanical Engineering Ltd.

3c Barley Market Street.
Tavistock, Devon.

Tel: Tavistock 5439 Telex 45263

A. Marshall (London) Ltd

1 Straits Parade.

Fishponds Rd.

Bristol BS16 2 LX
Tel: 0272 654201

Crystal Electronics

40 Magdalene Road,

Torquay.

Devon.

Tel: 22699

G.F. Milward

369 Alum Rock Road.
Birmingham B8 3DR.

Tel. 021-327 2339

L.F. Hanney

77. Lower Bristol Road,

Bath BA2 3BS. Avon
Tel: 0225-24811

The Radio Shop

16 Cherry Lane.

Bristol BS1 3NG.

Tel Bristol 421196.

S.T.D. Code 0272

Durrant Radio,
(Component Service)

9. St. Marys Street.
Shrewsbury, Shropshire.
Tel: 61239

Swift Electrical,

17-19, Harborough Road,
Northampton. NN2 7AX
Tel: 0604 712999

Marco Trading,

The Old School, Edstaston,

Nr. Wem. Shropshire. SY4 5RJ.
Tel: 094872 464/465

Marcson Electronics,

155 High Street.

Chasetown,

Walsall, Staffs.

Tel: (05436) 4632

Steve's Electronics,

15/17 The Balcony,

Castle Arcade.Cardiff.

Tel: (0222)41905

A. Marshall (London) Ltd

40-42 Cricklewood Broadway.
London NW2 3ET
Tel: 01-452-0161
325 Edgeware Road.

London, W.2
Tel. 01 723-4242

Ambit International

25 High Street.

Brentwood. Essex.

Tel. (0277) 216029.

Telex 995194

Anco Distributors,

50 Rainsford Road.

Chelmsford. Essex,

Tel : (0245) 58605
Arrow Electronics
Coptfold Road.

Brentwood. Essex.

Tel: 0277-219435

Audio Electronics

301 Edgeware Road,

London W2 1 BN
Tel: 01 724 3564

Basic Electronics Ltd

18 Epsom Road.

Guilford, Surrey

B. Bamber Electronics

5 Station Road,

Littleport, Cambs. CB6 1QE.
Tel. ELY (0353) 860185

Baydis Ltd.,

54. Mortimer Street.

Herne Bay, Kent.

Tel: Herne Bay 64586

Bl PAK,

18 Baldock Street.

Ware. Herts
Tel: (0920) 61593

Bywood Electronics

68 Ebberns Road.

Hem el Hempstead.

Herts HP3 9QRC
Tel: 0442-62757

C. N. Stevenson

236 High Street.

Bromley.

Kent. BR1 1 PQ
Tel: 01 464 2951/5770

Cavern Electronics

94 Stratford Road.

Wolverton. Milton Keynes.

Tel. Milt. K 314925
Charles T own
89 Carrington Street.
Nottingham,

Tel Nottm. 868933 and 55489

C.P. Developments

16 Hughenden Road.

High Wycombe.

Bucks HP13 5DT
Tel. 0494 30043

C.T.S. Ltd

20 Chatham St..

Ramsgate. KentCTII 7PP.
Tel. Thanet 54072

Custom Electronic
Controls

45. Picardy Road.

Belvedere. Kent DAI 7 5QH.
Tel. Erith 34476

Dawes Electronics

121 Dawes Road,

London SW6.

Tel: 01 381 3975

Direct Electronics

627 Romford Road,

Manor Park,

London El 2 5AD.

Tel. 01-553 1174

D.P. Hobbs

11 King Street.

Luton, Beds.

Tel: (0582) 20907

Eagle electronics

10. Eagle Street.

Ipswich.

Suffolk, 1P4 1 JB.

Tel: Ipswich 58075.

Electrovalue Ltd.

28 St. Judes Road.

Englefield Green,

Egham,

Surrey TW20 0HB
Tel: Egham 3603
Eley Electronics
100/104 Beatrice Road.
Leicester,

Tel. Leicester 871522

Frank Mozer Ltd

5 Angel Corner Parade,
Edmonton, London N18.

Tel. 01-807 2784

Frazer-Manning Ltd

26 Hervey Street.

Ipswich, IP4 2ES.

Tel. 50975

Foreway Services

19 Old High Street.

Headington.

Tel: 0865 68472-3

GB Garland Bros. Ltd

Chesham House,

Deptford Broadway,

London SE8 4QN,

Tel . 01 -692 4412
Greenway Electronic
Components
(East Grinstead) Ltd.,

62 Maypole Road,

Ashurst Wood,

East Grinstead.

Sussex.

Tel: 034 282 3712

Harrington Colorvision

9 Queen Street.

Colchester. Essex,

Tel. Colchester 47503

HB Electronics

22. Newland Street,

Kettering Northants.

Tel. (0536) 83922

Henrys Radio

404 Edgeware Road,

Lodon W2.

Tel: 01 723 5095

H.G. Rapkin,

11. Kettering Road.
Abington Square,
Northants.

Hi Fi Care,

245 Tottenham Court Rd..
London W1

advertisement

Eiektor April 1979 - UK 19

ILP Electronics Ltd

Crossland House.

Nackington.

Canterbury, Kent.

Tel: 0227 63218

J.T. Filmer,

82 Dartford Road.

Kent DAI 3ER
Tel: (0322) 24057

Kays Electronics,

195 Sheffield Road.

Chesterfield,

Derbyshire
Tel: (0246) 31696
Lloyd Electronics
126. High Street.

West Wickham, Kant,

Tel: 01 777 3247

MTRS

20 White Hart Street.

Mansfield, Notts.,

Tel. Mansfield 23107
Maydale Electronic
Services

2 Wellesley Parade.

Godstone Road, Whyteleafe.
Surrey CR3 0BL,

Tel. Upper Warlingham 5169

Mays of Church Gate

12/14 Church Gate
Leicester LEI 4AJ.

Tel: 58662

Orchard Electronics

Orchard House, St. Martins St.
Wallingford, Oxon 0X10 ODE,
Tel. Wallingford (0491) 35529

Electronic Centre

45 Lower Hill Gate,

Stockport, Cheshire SK1 1 1JQ,

Tel: 061 480 9791

Phonosonics

22 High Street

Sidcup,

Kent

RB Electrical &
Electronics,

24 Springfielfl Park,

Hollyport,

Maidenhead,

Berks.

Tel: 0628 39798

R.S.M. Television

38 Carrington Street,

Nottingham,

Tel. Nottm. 868933

S & A Enterprises,

25, Common View,

Letchworth, SG6 1BZ.

Herts. Tel: 01 467 0092

Noble Electronics,

26. Lloyd Street.

Altringham, Cheshire,

WA14 2DE

Tel.: 0619414510

Norman Inskip,

16 New Road,

Chatham, Kent.

Tel: (0634)811119.

Smiths of Edgware Road

287-289 Edgware Road,

London W2 1 BE,

Tel. 01-723 5891

Swan ley Electronic*

P.O. Box 68
Swanley, Kent.

Tel. Swanley 64851

C & L Electronics

81 Rosemary Road,

Nr. Jackson Corner,

Clacton on sea, Essex.

Tel: 0255 23266

S. R. Services

21b, Park Road, (entrance
Queens Road) Chislehurst,
Kent. Tel: 01 4670092

Suttons

50 Blue Boar Row.

Tel'. 27171

Tandy,

1 Emmanuel Street.
Cambridge CB1 1 NE
Tel: (0223) 68155

Technomatic Ltd.,

17 Burnley Road,

London. NW10
Tel: 01-452 1500
Telex: 922800

Telecraft

53 Warwick Road,

New Barnet.

Herts EN5 5EQ
Tel: 01-440-7033

Teleradio

325-7 Fore Street,
Edmonton,

London N9 OBE
Tel. 01 807 3719

Brian J. Reed,

161 St. Johns Hill,

Battersea.

London SW11.

Tel: 01 223 5016

QC Trading

1 St. Michaels Terrace.

Wood Green.

London N22 4S,

Tel. 01-889 7593

Vero Electronics Ltd.

Retail Dept..

Industrial Estate,

Chandlers Ford.

Hants S05 3ZR

Tel: Chandlers Ford 2956

Videocraft

Assets House.

Elverton St..

London SW1P2QR
Tel: 01-828-2731
Watford Electronics
33 35 Cardiff Road.

Watford
Herts WD1 80S
Tel. 0923 37774

T. Powell,

306 St. Pauls Road.

Highbury Corner.

London N1 .

Tel: 01 226 1489

Zartronix

1 1 5 Lion Lane,

Haslemere. Surrey.

Tel: 0428 52445

Rush moor Electronics

43, Queens Road,

Farnbo rough,

Hants GU14 6JP

Tel. Farnbourough 515373

AREA 4

The Electronic Centre
16 Collage Square East
Belfast 1
N. Ireland
Tel. Belfast 27357

international

Vadelec Electronics

24-26 Avenue de I'Heliport,
1000 Bruxelles.

Tel: 02 218 26 40
Telex: 26061

CANADA

Kitronic

26236 - 26th Ave..
RR 5. Aldergrove,
British Columbia

DENMARK

Dansk MINI Radio

Nr. Farimagsgade.

57-59.

1364 Copenhagen K.

Aage Nielsen Eftf.

1, Sortedam Dosseringen
2200 Copenhagen N.

Hobby Electronics,

37, Nedergade
5000 Odense

Frederikshave Hobby
Elektronik

9, Havnegade
900 Frederikjhavn

Elektroden,

5, Tordenskjoldsgade
8200 Arhus C.

IB's Radioservice,

Hovdrupvej 22,

9400 Norresundby

WK Electronic,

6, Skoletorvet,

8600 Silkeborg.

Kolding Elektronik,

40, Agtrupvej 40,

6000 Kilding

Ulla Electronic,

Christiansgade 58,

9000 Alborg

Radiolyttemes

Indkobscentral,

18, Borgergade,

1300 Copenhagen K.

Holte Elektronik,

Holte Midtpunkt,

2840 Holte

Lillie Elektronik,

89, Sondergade,

6500 Vojens

Gettermann Electronic,

Radhuspassagen,

7100 Vejle

EL-Star Hobby,

143 Fmsensvej,

2000 Kobenhavn F.

WM. B. Peat &
Company Ltd

25/26 Parnel Street
Dublin 1 , Ireland
Tel. 749973/4

Rotec,

16, Jernbanegade.

4800 Nykobing Falster

BERIC,

43 Rue Victor Hugo.
Malakoff 92. Paris
PENTASONIp

5 Rue Maurice Bourdet
75016 Paris
Selectronic,

14, Boulevard Carnot,
59800 Lille, France
Tel: (20) 55.98.98

INDIA

PRECIOUS ELECTR
CORPORATION

3 Chunam Lana.

Dadasaheb Bhadkamker Marg,
BOMBAY -400 007

INDONESIA

Inel Co.,

Office: Jl. Naripan 98,
Bandung,

Tel: 59544

ISRAEL

K.D.M. ELECTRONICS

Tel Aviv, LTD.,

21 Tchernihovsky St.,

P.O.B- 4770

ITALY

Vulpetti Guiseppe,
i2VTT Via G.Cesare,
4-22063 Cantu.

ELCOM

34170 Gorizia,

Via Angiolina 23

SOUTH AFRICA

PHILTRON (PTY) LTD.,

P.O. Box 2749
Pretoria 0001.

SWEDEN

COILTRONIC

Box 5007,

16305 Spaanga

KITEL Distribution

Box 21038

S-100 31 STOCKHOLM

TURKEY

ELEKTRONIK DERGI
VE KITAP

Yayinevi P.K. 1126,

Istanbul,:

NEW ZEALAND

Orbit Electronics,

161 Hobson St.,
Auckland 1,

New Zealand.

Elektor April 1979 -UK 20

advertisement

TkAINjAM

TRANSAM COMPONENTS LTD., 12 CHAPEL STREET, LONDON, N.W.I. TEL: 402 8137

The exciting new

TRITON

Personal Computer

Basic in Rom: a powerful 2k Tiny
basic resident on board,* makes Triton
unique, easy to use and versatile.
Graphics: 64 Graphic characters as
well as full alpha numerics.
Single Board: Holds up to 8k of
memory, 4k RAM and 4k ROM, sup-
plied with 3k ROM and 2k RAM.
Memory Mapping: 2 mode VDU, I/O
or memory mapped for animated
graphics.

Cassette Interface: crystal controlled
Modem tape I/O with auto start/stop
+ “named” file search.
U.H.F. TV Interface: On board uhf
modulator, plugs into TV aerial socket

COMES COMPLETE with keyboard,
case, full power supply, quality
through hole plated P.C.B. Full
(118 page! instruction manual. A

E owerful lk Monitor and 2k Tiny
asic in Eprom all I.C. Sockets.
All components can be bought
separately, so you can start con-
struction on a low budget. Full details
of prices and discounts are shown in
our new 1979 catalogue.

EXPANSION BOARDS
MOTHERBOARD -8 slot
A new 8 slot Triton Motherboard is
now available based on Eurobus. It
allows easy expansion and has its own
meaty power supply.

8k STATIC RAM

Eurocard size (160 x 100mm) 8k
Static Ram fully buffered, on board
regulation and decoding uses 4k
(lk x 4) static rams. Just plugs into
Motherboard for memory expansion.
8k EPROM BOARD
Designed to take 8 x 2708 Eproms on
the Triton bus. Don’t forget our
Programming Service

TRITON KIT £286

MOTHERBOARD KIT £50

8k RAM CARD £97

8k EPROM CARD T.B.A.end Feb.

Prices exclude VAT 8%.

118 Page Manual inc. P8iP £5.70
Full details in Catalogue 30p. + SAE.

TC itronic

AUDIO, DIGITAL AND ELECTRONIC
COMPONENTS LTD.

26236 26TH AVE R R 5 ALDERGROVE B.C. VOX 1A0

856-

CANADIAN DISTRIBUTORS to ■■■ HU 1W1 magazines, binders and P.C.
specialty shops of : boards for elector projects.

WESTERN CANADIAN DISTRIBUTORS OF “7“'°',
PERSONAL COMPUTING, 73 MAGAZINE R.O.M.
INTERFACE AGE, BYTE, KILOBAUD,

SYBEX BOOKS

THE ONLY WESTERN CANADIAN
STEREO MAIL ORDER HOUSE

CURRENT and BACK ISSUES
are available by mail order for
the above magazines and books.
Send a self addressed stamped
envelope for a complete list of
magazines and books.

Send one dollar for catalog and
current price lists. Components,
Kits, speakers, amplifiers, rec-
eivers etc.

DEALER ENQUIRIES WELCOME

SEND ALL MAIL 1Z .. . AUDIO, DIGITAL AND ELECTRONIC

ORDERS TO' I V ltruilic COMPONENTS LTD.

26236 - 26TH AVE., R.R. 5, ALDERGROVE, B.C. VOX 1 A0