100123300 000000000

• ^ jw&m.

1 LL^L'MiXii

M \ui&0< 1 ’

contents

selektor

8K RAM + 4,8 or 16K EPROM on a single card

Many readers have requested that the 4K RAM card for the Elektor SC/MP
system be updated. The new card presented in this article contains a total
of 8K of RAM, up to 16K of EPROM and can be used with either of the
Elektor SC/MP systems or the Junior Computer.

precision power unit

electronic linear thermometer

Conventional mercury thermometers are, of necessity, rather fragile and
therefore break very easily, always at the most inoppotune time. Electronic
thermometers, on the other hand, are not as fragile and can enjoy a much
longer life. The 'readout' can be a great deal more accurate with infinitely
better legibility. The electronic thermometer featured here can be built
by anyone to fit almost anywhere which is certainly not true of the
conventional type.

I the Josephson computer

Large computers can consist of over a hundred thousand logic circuits
| and are able to execute more than a million instructions per second. Fast
: as this may seem, a computer 20 times faster than this is currently being
1 developed by IBM. It stems from a brand new branch in technology,
namely, the Josephson technique

VOX printed circuit board

UK EOITORIAL STAFF
T. Day E. Rogans

TECHNICAL EDITORIAL STAFF

ejektor: measuring multipath

G.H.K. Dar
P. Holmes
E. Krempel
G. Nachbar
A, Nachtmi
K.S.M. Wall

high speed readout for elekterminal

With minor a modification to the Elekterminal it is possible to
entire contents of the display (TV screen) on a cassette tape.

musical box

electrolytology

There is nothing new about the fact that electrolytic capacitors possess
inductance due to the method used in their construction. At high
frequencies their impedance will be largely determined by this parasitic
induction. An electrolytic capacitor even acts as a band pass filter and
thus also has a resonant frequency.

What fewer people will be aware of is that the capacitance is clearly
frequency dependent. Elektor takes a closer look 'inside' electrolytic
capacitors this month.

curve tracer

This article describes an economic curve tracer for transistors and diodes.
No really professional test instrument, of course, but an extremely useful
aid to quickly carry out a general test either to compare transistors or
select them.

( using the vocoder

I Several months ago (Elektor no. 56, December 1979), Elektor published
a 10 channel vocoder. When building a vocoder there are a few 'obstacles'
which ought to be taken into account. Readers who have already built one
. and are familiar with it will find that this article provides useful
I information on how to improve on the vocoder's technical qualities.

market

advertising index

eleklc

1980-9-01

Bright outlook for liquid crystals

Reliable, long-life liquid-crystal displays
are now familiar features of mass-
produced watches and pocket calcu-
lators. But the work of the university,
government and industrial research
teams that made them available so
cheaply is by no means at an end, for
the growing complexity of devices
means more stringent specifications of
materials. Liquid crystals of a newly-
developed type promise to hold the key
to further important innovations.

Digital electro-optic displays using liquid
crystals are now commonplace in
watches, pocket calculators and clocks;
and. thanks to advertising, the terms
liquid crystal and liquid-crystal display
are now familiar to all. The success of
such devices is entirely dependent upon
the quality of the thin, fluid film of
liquid-crystal material that is used to
present the figures and is shown in the
photo.

Progress in making satisfactory liquid-
crystal displays was held up for a long
time simply because liquid-crystal
materials with good, stable character-
istics were not available, but things
suddenly changed with the discovery of
the cyano-biphenyl class of liquid

Immediate demand

The potential of these liquid crystals
seemed to be excellent, and it was borne
out by an extensive test programme at
RSRE. There was an immediate demand
for the new materials. BDH Chemicals,
at Poole, in the South of England,
solved the problems of production and
made them in large quantities, pure
enough to guarantee reliable perform-
ance in display devices.

One interesting feature of this collabor-
ation between a university, a govern-
ment research laboratory and a chemical
company is that the research, in bringing
together two scientific disciplines,
chemistry and physics, has led to more
in the nature of mutual stimulus than to
problems of communication and under-
standing.

Liquid crystals

What are liquid crystals? They have
been known for many years, having
been discovered by the Austrian botatist
Reinitzer in 1888. He observed that the
organic chemical cholesteryl benzoate
melted sharply at 145°C but did not
form a clear liquid. Instead, it gave a
cloudy fluid which became clear only
when heated to 179°C. The inter-
mediate phase was eventually recog-
nized as a liquid crystal. Soon, other

CH,0-{3- n = N “O" 0CH J

O

CHjO -^^-CH = N-^^-OCO.CH,

CH,0 HC =(^)= CH OCH 3

Figure 1 . Soon after the discovery of the first
liquid crystals, it was found that other organic
compounds behaved in a similar way.

A common feature of the materials was that
the molecules were long and narrow, and

organic compounds were found to
behave in a similar way, and a common
feature of all the materials was that the
molecules were long and narrow. More-
over, they all contained ring systems
and double bonds which lent them
rigidity (see figure 1).

It is not unreasonable that compounds
made up in such a way should behave as
they did. Their crystal lattices consist of
a rigid, three-dimensional, ordered array
of the rodlike molecules; when the
crystals are heated, thermal vibrations
eventually overcome the intermolecular
interactions, the molecules become free
to move in any direction and the rigid
crystal ceases to exist. With less elon-
gated or more nearly spherical mol-
ecules, the material is a true liquid, with
total disorder of the molecular system.
But with rod-shaped molecules there is
a strong tendency for their long axes to
stay parallel, over considerable molecu-
lar distances, even after the crystal has
collapsed. This brings about a fluid but
nevertheless quite highly organized

Because of the molecular order which

2

(a)

(b)

Figure 2. Reversible transition from a nematic liquid crystal (a) to the disordered isotropic
liquid (b).

9-02-1

>r September 1980

persists, the material in this phase has
many of the optical properties of a
crystal, yet it can flow; that is why we
use the term liquid crystal. Only when
we heat the material to a higher tem-
perature is the truly disordered, liquid
state produced (see figure 2).

It is quite common for the molecules in
liquid crystals to retain a parallel
arrangement, in layers. When this hap-
pens the crystals are known as smectic
liquid crystals and are rather viscous;
they have not, so far, been of much
commercial use. A great deal more
important are the so-called nematic
liquid crystals, which are much more
fluid and possess a non-layered, parallel
molecular arrangement.

Another form of the nematic liquid
crystal is found when the molecules of
the compound are optically active, that
is, when they can take up either a right-
handed or left-handed structure, related
one to the other as an object is to its
mirror image. Through the asymmetry
of the intermolecular force fields in a
liquid-crystal phase made up entirely of
right-handed or of left-handed mol-
ecules, the molecules are no longer
arranged mainly in parallel in three
dimensions; instead, the molecular
arrangement may be thought of as in
the diagram in figure 3. Here, the mol-
ecules lie parallel to one another in
sheets, but there is no ordered arrange-
ment of their ends. Crystals with this
sort of structure are known as
cholesteric liquid crystals.

Passing upwards through a stack of such
sheets, we find that the long axes of the
molecules lie progressively in a single-
handed sense, forming a right-hand or
left-hand helical arrangement with a
well-defined pitch. Because of the
relationship X = Pn, where X is the wave-
length of the incident light, P is the
pitch and n is the refractive index
(usually about 1-5), such crystals have
the property of selectively reflecting
coloured light, when P is within the
range of wavelengths of visible light. For
this reason, suitably pitched cholesteric
liquid crystals are used in digital ther-
mometers and for various forms of sur-
face thermography; the pitch and the
colour of the reflected light change with
temperature.

Obviously, the existence of materials
giving fluid but ordered states of matter
provides a considerable challenge to
inventively minded people to find other
applications for them. The need to heat
a solid to form the nematic or choles-
teric phase was a severe drawback, but
in turn it challenged organic chemists to
provide materials that formed those
phases in the ambient temperature
range. The aim was eventually achieved,
but at first the materials had certain
drawbacks.

The first room- temperature phases w

3

Figure 3. Portrayal of the molecular arrange-
ment in a cholesteric liquid crystal. The arrow
represents the direction of the molecular long

CH,(CH,),-^-^-CN (b)

CH,(CH 2 )nO-^-^-CN (c)

ch,<ch,)„-£3“0“0 -cn (d)

Figure 4. The first room-temperature liquid-
crystal phases were formed by compounds, or
mixtures of compounds which all had the
general form shown in (a), where the linking
group -A-B- was of the type -N-CH-,

— N=NIOI— , -CO.O- and so on. Such linking
groups made the compounds unstable or
coloured, and this problem was eliminated
by linking the two rings directly in a biphenyl

or alkoxy groups were used at the ends of the
molecules to give 4-alkyl- and 4-alkoxy-
4'-cyano-biphenyls as in lb) and (cl. The

developed.

formed by compounds, or mixtures of
compounds, which all had the general
form shown in (a) in figure 4, where
— A— B— , the linking group, was of the
type — N=CH— , -N=N(0)-, — N=N— ,
—CO.O— and so on. But these linking
groups made the compounds chemically
or photochemically unstable, or
coloured them. The best group was the
ester linkage -CO.O-. Nevertheless, the
existence of these room-temperature
nematics allowed physicists to explore
the potential value of such materials
and. from the early 1960s, it seemed
certain that they would be exploited.

The disadvantage of the central group •
was recognized and it was eliminated
by linking the two rings directly in a
biphenyl structure. Shortening the mol-
ecules might seriously restrict the tem-
perature range of phases, so, from
previous experimence, a cyano group
was used as one of the end groups. The
group for the other end was chosen to
be a saturated alkyl group or an alkoxy
group. In this way, the idea of the
4-alkyl- and 4-alkoxy-4’-cyanobiphenyls
was born. Their structures are shown in
(b) and (c) respectively (figure 4).

One of the first materials synthesized
was that shown as (b) in the diagram,
with a chain of five carbons. It melted
at 21.5°C and was nematic until 35°C;
the nematic phase was maintained
indefinitely at room temperature and
down to 4 C in a supercooled state. The 1
nematic phase was colourless, the com-
pound highly stable chemically and
photochemically. and the material was
non-toxic and apparently devoid of
harmful properties. A wide range of
these compounds was made. Several
melted at low temperatures and the
compounds of the form (c) in the dia-
gram had the higher -temperature change
of phase from nematic liquid crystal to |
isotropic liquid.

Derivatives

From this range of compounds, mix-
tures could be produced that behaved
as a system melting at around 0°C or
just below and remaining nematic until
over 50°C. This temperature range was
judged to be too narrow and led to the
development of analogous p -ter phenyl
derivatives such as that shown as (d) in
the diagram.

The nematic phases of these materials
persist at temperatures up to well above
200°C, but by incorporating suitable
amounts of compounds such as (d)
with, say, n = 4, in mixtures of the
materials (b) and (c) new systems were
obtained with wider nematic ranges
from -10°C to 60.5°C, -12°C to 72 C
and so forth.

With their strongly dipolar cyano group
at one end of the molecule, these com-
pounds have a high positive dielectric
anisotropy, which means that the elec-
tric permittivity along the long axis is
greater than it is across the short axis;
so, in the nematic phase, the molecules
have a strong tendency to align parallel

to the direction of the field. This is just
what was needed, for the availability of
such wide-range nematic systems on a
commercial scale from BDH Chemicals
allowed electronics companies to make
electro-optic displays capable of excel-
lent performance and long life.

The watch or calculator display is
simply a thin sandwich of the nematic
liquid crystal between two glass plates
coated on their inner surfaces with a
transparent, conductive film of an elec-
trode material such as ln 2 0 3 or Sn0 2 .
By treating the electrode surfaces in a
suitable way, the molecules of the liquid
crystal are made to lie parallel to them
but their directions are arranged at a
right angle across the film. The inter-
vening molecules in the film, which is 6
to 12 pm thick, take up a quarter helical
arrangement, as shown in figure 5. If
light entering the sandwich is polarized
in a plane parallel to the long axes of
the molecules at the film's surface, it is
guided through a right angle as it passes
through the film and it emerges through
a second polarizer set at right angles to
the one through which it enters. So, in
its so-called off state the cell transmits
brightly and can be used to produce
bright reflections if a back-mirror is
used. But when a small voltage, of about
2,5 V in a 12-/jm cell is applied across
the film, the molecules turn rapidly to
align at a right angle to the electrodes.
Light is no longer guided as it passes
through the cell, which appears black.

It is obvious that if only parts of the cell
are activated electrically (for example,
groups of seven bar electrodes in a
figure-of-eight arrangement) we have a
way of presenting black information on
a bright background without having to
generate light energy within the device.
When we switch the field off, the mol-
ecules rapidly realign to form the
twisted state. This is the basis of the dis-
play in which the cyano-biphenyl liquid
crystals have been so successful. The
only movement is molecular, so the
response times are very fast, and, with
a power consumption in the microwatt
range, batteries have a long life.

Cholesteric devices

Biphenyl liquid crystals provide us with
a means of making stable cholesteric
materials. We need only introduce a
branched carbon chain for X instead
of CH 3 (CH 2 ) n - or CH 3 (CH 2 ) n O-,
and provided it makes the system
active optically, for example, with a
(+)-CH 3 CH 2 CH(CH 3 )CH 2 - group, the
material will be cholesteric.

Materials of this sort were made at Hull,

Figure 5. The change in molecular arrangement when a twisted nematic liquid-crystal is
switched electrically from the bright off state (left) to the dark on condition (right). When the
field is switched off, relaxation from the on to off state takes place rapidly through forces
starting at the surfaces.

and they went into commercial prod-
uction. Incorporating them in the
nematic hosts shown in (b) and (c) of
the diagram above produces cholesterics
with a range of pitch values that depend
upon concentration. Such compounds
are added in very small amounts to
nematic liquid crystals used in the
twisted nematic device to ensure that
when the device is switched off, the
quarter helix always regenerates with
the same sense, so avoiding patches
through reversed-twist areas. In greater
concentration, they are added to
nematics to form quite high-pitch
cholesteric phases for use in another
form of device which does not need
polarizers. In such devices, the twist of
a cholesteric of positive dielectric
anisotropy can be unwound electrically
to yield a nematic phase, which then
twists back to a cholesteric when the
electro-motive force is switched off.
There is an optical contrast between the
switched-on and switched-off states,
which can be made more pronounced
by dissolving dichroic dyes in the liquid
crystal, thereby making it possible to
produce a negative colour contrast in
these cholesteric-nematic phase change
displays, with numbers in white on a
coloured background.

The cyano-biphenyl and cyano-terphenyl
materials have provided the essential
means of producing the reliable liquid-
crystal display devices which are now
preferred to light-emitting diode dis-
plays, particularly for battery-operated
instruments. But devices are becoming
more and more sophisticated, which
means more stringent requirements of
performance and properties of the
liquid-crystal material, so there is still

a great deal to do.

For example, in multi-function calcu-
lators and watches, matrix addressing
of the display greatly reduces the num-
ber of individual electrical contacts
needed for the electrode elements.
Mixtures of the cyano-biphenyl liquid
crystals with other nematic materials
allow this to be achieved, but for ideal
performance we need materials with
operating-voltage thresholds that are
largely independent of temperature.
Mixtures of a new type of nematic
material developed at Hull have now
been found at RSRE to be less depen-
dent on temperature than any earlier
materials are, and they may well be the
key to further success.

Cholesteric materials of negative dielec-
tric anisotropy are the basis of positive-
contrast colour displays, with coloured
numbers on a colourless background,
making use of dye-cholesteric-nematic
phase-change. Such devices are at their
best if the materials are of the type with
a low birefringence, that is, with refrac-
tive indexes that have much the same
values in the directions of the long axes
and the short molecular axes. Work
going on at Hull, RSRE and BDH
Chemicals aims to combine these
desirable physical characteristics in
liquid crystals that are stable at room
temperature.

Prof. G.VZ. Gray, in Spectrum no. 167
(570 S)

9-04 - i

RAM + 4,8 or 16K EPROM

The extremely popular 2114 RAM 1C
has four times the memory capacity of
the type used for the original (4 k) RAM
card. This means that a card with twice
the memory capacity can be con-
structed with half as many ICs. It will
be apparent that this leaves a certain
amount of 'free space' on the actual
board. Why not fill it up with EPROM?
By doing this we can kill two birds with
one stone - there is no Elektor EPROM
card as such for any of the Elektor
computer systems.

The decoder (IC5) divides the entire
address range into 4 k ‘pages'. Each
memory section (including the RAM
area) can be placed anywhere within the
64 k address range. When 2708 s are
used for the EPROMs, they can be
positioned on any page by connecting
a single link from IC5 to both inputs of
N1. Two pages are required if 2716s
are used and this involves installing two
wire links between IC5 and N1. If,
however, 2732 s are used, one complete
page can be allocated to each EPROM.

8KRAM+4,8 or
16K EPROM

on a single card

Many readers have requested that
the 4 k RAM card for the Elektor
SC/MP system be updated. The
new card presented here contains
a total of 8 k of RAM, up to 16 k
of EPROM and can be used with
either of the Elektor SC/MP
systems or the Junior Computer.

The 27xx series of EPROM was chosen
as the 2708, 2716 and 2732 are all pin
compatible (Ik, 2k and 4k EPROMS
respectively). Obviously, to be 'univer-
sal' certain connections have to made
'programmable' (see circuit diagram in
figure 1). The address decoding and the
logic level on the chip select inputs
depend on the particular type of
EPROM used. These connections can be
altered by means of wire links on the
printed circuit board (figure 2).

input B address order

2708 A10 All

2716 A12 All

2716 A12 All

2732: IC7 deleted (see text).

IC25- 26-27-28

IC25-27-26-28 beginning at an even page
IC26-28-25-27 beginning at an odd page

Table 1 . This table shows the connections to the A and B inputs of IC7 for the different types
of EPROM.

Table 2

RAM

EPROM

1k0 - 1000
Ikl = 1400
1k2 = 1800

1 k5 - 2400 .
1 k6 = 2800 .
1 k7 = 2C00

of IC5 to inf
N2

13FF
1 7FF
1 BFF
1 FFF
23FF
27FF
2BFF
2FFF

IC25 =
IC26 =

IC28 *

: of

3000

3400

3800

2707

2716

. . . 33FF
. . . 37FF
. . . 3BFF
I...3FFF

3000 . . . 37FF

4000 . . 47FF

3800 . . . 3FFF
4800 . . 4FFF

3000

4000

6000

connect pin 14
of IC5to

connect pins 14 conn
and 3 of IC5 to 3. 1 1
inputs of N1 IC5 t

2732

. . . 3FFF
. . -4FFF
. . . 5FFF
. . . 6FFF

9 ... 1 2 of IC7
(IC7 is deleted!).

Table 2. An example of the possible address format when the RAM section is sequentially
followed by the EPROMs.

This will be explained in detail further

The next step in the address decoding
is to enable the individual memory ICs.
As far as the RAM is concerned this will
be in sections of 1 k (two ICs per sec-
tion). The EPROMs, on the other hand,
will be in sections of 1 k, 2 k or 4 k (for
2708s, 2716s and 2732s respectively).
The RAM section is taken care of by the
3 to 8 line decoder IC6. One half of a
similar 1C, IC7 (2 to 4 line decoder) is
used to select the EPROMs. Wire links
are included to pre-program the A and
B inputs of IC7 for the particular type
of EPROM to be used (see table 1 ) . The
order of addressing will be slightly
different when 2716 s are used, but this
should not cause any problems in
practice, provided the EPROMs are
programmed (and installed) in the
correct order. Table 2 gives an example
of the relevant addresses and connec-
tions for when the RAM section is
placed on pages 1 and 2 followed
sequentially by the EPROM sections.
The memory card iscompletely buffered
to keep the load on the bus system to
a minimum. The address bus is buffered
by IC1 and IC2. These are uni-direc-
tional buffers which have PNP inputs
requiring a very low input current. The
same is true of the data bus buffers
(IC3 and IC4). These are bi-directional,
the direction of data transfer being con-
trolled by the logic level on the com-
mon select line. When this line is low
the buffers enable the transfer of infor-
mation into the RAM section, and when
the select line is high the data con-
tained in the RAM/EPROM sections can
be read out.

While the memory card is not being
addressed the data bus buffers are held

in the write mode (via gates N3 and N5) section when a WRITE signal is present other microprocessor systems,
to ensure that the card is unable to (via N4). The two wire links shown at

interfere with the data bus. When the the inputs to N4 enable the memory Arranging the memory blocks

card is being addressed, the buffers are card to be used with the Elektor SC/MP The way in which the address decoding j

switched to the read mode. Data can system or the Junior Computer (both is done on this card makes for a large

then only be entered into the RAM inputs connected to 31a), or with most degree of flexibility — provided you I

know what you're doing! The first thing
to realize is that IC5 divides the address
area into 4kByte blocks, and that N1
(with inputs V and W) selects one or
more of these for the EPROMs, whereas
N2 (inputs X and Y) selects two 4kByte
blocks for the RAM. In general:

V or, via a wire link, to positive supply.
The 4kByte field is further subdivided
by IC7 (connected to address lines A10
and All), to select the EPROMs as
follows:

IC25 V0O0 . . . V3FF
IC26 V400 . . . V7FF

4kByte EPROM 2x 4kByte EPROM

Two 4kByte blocks are required,
one for IC9 . . . IC16 and one for
IC17 . . . IC24. One of these blocks
must be on an even-numbered page
(0, 2, 4, etc.) and the other on an odd
page. For example, X = 4 and Y = 5
would define a consecutive RAM area
from 4000 . . . 5FFF.

EPROMs type 2708

For four of these IkByte EPROMs, a
4kByte address field is required. This is
selected by connecting one of the
outputs from IC5 to N2 ('V'); the other
input to N2 ('W') is either connected to

EPROMs type 2716

An 8kByte address field is required in
this case (4 x 2kByte) . The same
principles apply as discussed above for
the RAM area: V must be connected to
an even-numbered output from IC5, and
W to an odd-numbered output. For
example, if V = 2 and W = 7, the four
EPROMs will correspond to the
following address fields:

IC25 2000 . . . 27FF

IC26 7000 ...77FF

IC27 2800... 2FFF

IC28 7800... 7FFF

Note that IC25 and IC27 form a 4kByte

pair, as do IC26 and IC28.

EPROMs type 2732

Each of these ICs corresponds to a
4kByte address field - in other words,
to one output from IC5! In this case,
no further subdivision of this field is
required, so that IC7 becomes redun-
dant! N1 is not required either, but its
two inputs (V and W) must be connected
to +5 V by means of wire links.

The four 4kByte blocks required can be
programmed by wire links direct from
the corresponding outputs of IC5 to the
holes intended for pins 9 ... 12 of IC7
(k ... n). Pin 9 (k) corresponds to IC25,
pin 10 (I) to IC26, etc. This means that
if, say, IC28 is to be located on the last
page, a wire link must be taken from
output F of IC5 to pin 12 (n) of the IC7
position.

Wire links and unused positions

An important point to note is that
unused inputs should not be left
floating. This was already mentioned
above, as regards N1 and N4. The same
obviously applies to N2, if the total
RAM area is not to be used as yet:
unused inputs must be connected either
to +5 V or to an unused output from
IC5.

Particular care should also be taken with
the wire links at the inputs to IC7 and
IC25 . . . IC27. These depend on the
type of EPROM used, as follows:

2708: P-Q, S-T, e-f, a-c.

2716: P-R, S-T, e-g, a-d.

2732: e-g, a-b.

Finally, it should be noted that the
supply common (0 V) connection to the
board must be applied via two sets of
connector pins: 4a/c+16a/c and
32 a/c. These two sets are not inter-
connected on the board! K

9-08 - ele

As we all know it is virtually impossible
to construct a precision voltage source
with standard components. Close
tolerance devices (both active and
passive) are very difficult to obtain. If
resistors are connected in series and
parallel to produce the required value, a
tolerance of 0.1% is out of the question.
The solution, therefore, is to find
integrated components with 'everything
included'. The majority of so called
precision voltage regulators suffer from
the disadvantage that they can only
provide one output voltage. Elektor's

The voltage across RsET is buffered by
an opamp before being fed to the
internal series pass transistor. The 1C
contains integrated 0.1% tolerance
resistors between various pins which can
be interconnected to provide a combi-
nation of values. Resistors with values
of 5 k, 10 k, 2 k and 6 k are located
between pins 9 and 7, 7 and 6, 6 and 5
and 8 and 4 (ground) respectively. The
output current is determined by. L|_im
(100/zA) and the current flowing
through RSENSE and can be calculated
from:

|H*< vision power imil

When calibrating voltmeters an
accurate reference voltage is, of
course, essential. As far as digital
voltmeters are concerned the
reference voltage will have to be
very accurate indeed for such
meters to be of any significant
value. If close tolerance precision
resistors are incorporated in the
DVM's input attenuator, the
reference voltage should meet
accuracy requirements outside the
basic measuring range. Thus, a
reference voltage is required with
an accuracy that corresponds at
least to the tolerance of the
divider resistors. The precision
power unit described here
produces several reference voltages
with an accuracy of 0.1%! To
make full use of the degree of
precision achieved, the unit has
been incorporated into a high
quality power supply.

design team, however, have discovered
a little known 1C from National
Semiconductor which is able to produce
several accurate voltages and has
excellent characteristics and can also be
incorporated into a power unit as a
'normal' regulator 1C. This device has
the part number LH 0075.

The block diagram of the precision
power unit is shown in figure 1 . As can
be seen, it is very similar to conventional
circuits of this kind. Pre-stabilisation has
been included to limit the input voltage
to the regulator 1C. This protective
measure is quite justified considering
the cost of the 1C. Both voltage and
current can be adjusted separately. By
including a pair of series pass transistors
an output current of up to 2A can be
produced. A description of the technical
specifications for the unit is given in
table 1.

The internal structure of the 1C is
shown in figure 2. A constant current
source is fed to a zener diode via a field
effect transistor. This produces a highly
accurate, temperature stable reference
voltage with a variation of 0.003%/°C!
This reference voltage is then used to
produce two further constant currents
(*SET a nd ■LIM)- Tbe output voltage is
determined by the (1 mA) current
flowing through resistor RsET and can
be calculated from the formula:

UQUT = >SET • RSET

. r lim ,

l0UT< ™ xl "BiI5Ii" lLIM

The 1C can also be used as a pro-
grammable current source when pin 9 is
connected to ground via a 25 k resistor.
The output current will then be deter-
mined by the values of Rum and
RSENSE- A potentiometer could be
used for R|_|M so that the output
current can be made adjustable.

Circuit diagram

The complete circuit diagram of the
precision power unit is shown in figure
3. The maximum secondary voltage of
the mains transformer is limited to 30 V
so as not to exceed the input require-
ments of IC2. The transformer voltage is
rectified by B1 and smoothed by Cl
before being fed to the pre-stabiliser IC1 .
LED D1 indicates that the circuit is
switched on.

By including a zener diode (D2) in series
with the ground lead of I Cl, its output
voltage is 'raised' to 30.2 V to provide
an adequate (and safe) input level for
IC2.

The output voltage of the circuit can be
adjusted by means of potentiometer P2
which is connected as shown in figure 4.
The output current limit is set by PI,
R2 and R6. Resistor R6 is included in

Table 1

Technical Data

Variable Output voltage : 0 V to + 25 V

Fixed Output voltage: + 1 .5 V, 2 V, 5 V. 6 V. 8 V. 10 V, 1 2 V, 15 V,

18 V

Voltage regulation:
Ripple suppression:
Presettable current limit:

type 0.008% A/
80 dB
Oto 2 A
type 0.075%

Table 1. Technical :
indeed precise.

eification of the precision power unit. As the figures show, the unit is

parallel with PI to reduce the maximum o
output current to 2A, while R2 acts as a
current sense resistor. The output
voltage is selected by a multi-position
switch (see figure 4) connected, as
mentioned before, to the internal
precision resistors of IC2. This switch
connects the various resistors in series or
parallel as required.

Transistors T1 and T2 increase the
current output capability of the supply
and the resistors in their emitter leads
(R7 and R8I ensure that the current is
divided equally between them.

Resistor R3 is included as a dummy
load for the unit and diodes D4 and D5
protect the circuit from negative
transients.

The output voltage and current can be
monitored by including a moving coil
meter and a double pole switch. If the
unit is to be used to power HF circuits
an extra 100 nF capacitor should be
connected directly across the output.

Figure 2. The ir

C3 = 2(i2/25

©»

o;

©

0 On

©

6

7

0 \

Figure 7. The new scale for the 1 mA moving

Construction and setting up

The printed circuit board and com-
ponent layout for the precision power
unit is shown in figure 5. The socket for
IC2 can be made from 'socket strip' by
cutting off four strips of three contacts.
A suggested front panel layout for the
unit is shown in figure 6. Once this has
been attached, and the meter scale
substituted for the one shown in
figure 7, the unit can be wired as shown
in figures 3 and 4.

All wiring should be carried out with a
great deal of care and attention to detail
as one mistake could burn a hole in
your pocket.

After the wiring has been checked
thoroughly (several times! ), the output
voltage of IC1 should be measured
without IC2 inserted. If this voltage is
any higher than 32 V there is something
wrong with the pre-stabilisation circuit
which could result in damage to IC2.

If the voltage is correct the unit can be
switched off and IC2 inserted. Again,
check several times that the 1C is
positioned correctly. With S2 in the
'voltage' position, S3 switched to one of
the preset ranges and a voltmeter
connected across the output, the unit
can be checked and P3 adjusted to give
the correct reading on the scale of Ml .
The current range can be adjusted with
the aid of a known load resistor. Switch
the unit off, turn PI fully anti-clockwise
and switch S3 to the 10 V position.
With a load resistor of 10S2/10W
connected across the output (or a
universal meter switched to the lA-range
-or higher) rotate P2 until the meter
needle stops moving. According to
Ohm's Law a current of 1A will then
flow through the load resistor. The
meter scale can be adjusted by means of
P4.

Once the above checks have been carried
out successfully the unit can be installed
in a suitable case and is ready for use. H

Figure 6. Proposed front i

9-1 2 - elektor September 1 980 electronic linear thermometer

electronic
lin ear

thermometer

a semiconductor used as a temperature sensor

The conventional mercury
thermometer has been with us for
a long time, mainly because it
serves its purpose very well. It
does however suffer from a
number of major disadvantages.
They are, of necessity, rather
fragile and therefore break very
easily, always at the most
inopportune time. A relatively
long period is required for them to
stabilise and they are not the
easiest thing in the world to read.
Electronic thermometers, on the
other hand, are not as fragile and
can enjoy a much longer life. The
'readout' can be a great deal more
accurate with infinitely better
legibility. Furthermore, they can
be built by anyone to fit almost
anywhere which is certainly not
true of the conventional type. The
actual sensor, in this case a
semiconductor diode, is very small
allowing it to be mounted in
previously impossible situations. A
further advantage is that, due to
its linear characteristics, expensive
equipment is not required to
calibrate the unit.

Various types of sensors are available
for the purpose of constructing a fully
electronic thermometer. Temperature
sensitive resistors are often used, with
either a positive temperature coefficient
(PTC) or a negative temperature (NTC).
A temperature coefficient that is
positive means that the resistance
increases with the temperature while
with a negative coefficient the resistance
will decrease with temperature.

The disadvantage of thermal sensitive
resistors, however, is that they are not
linear. The characteristic which
represents the curve of the resistance as
a function of the temperature is not a
straight line, but slightly curved. There-
fore, unless elaborate compensation
networks are included, a resistor can
only be used within a small temperature
range, for then the small part of the
curve used can be considered to be a
straight line. For greater temperature
ranges a different sensor will have to be
used. With respect to fairly high

temperatures of up to 1000° C thermo-
couples are needed. These, however,
demand a rather specialised technique
(cold welding compensation, compen-
sation of temperature influence as a
result of current passing through, etc.)
and are therefore not suitable for
domestic use.

Temperature sensors using semiconduc-
tor diodes or transistors do not suffer
from these drawbacks. They can be
applied within a wide temperature range,
are not complicated in structure and are ’
as compact as the other sensors.

The temperature sensitivity of the
semiconductor sensor is based on the
principle that the forward voltage will
change with temperature when the
forward current is maintained at a
constant level. An example of this is
shown in figure 1 , when the forward
voltage is a function of the temperature
in the BAX 13 diode. It will be seen
that it can exhibit PTC or NTC charac-
teristics depending on the value of the
forward current. At a current of 1 mA
the diode has a distinct negative
temperature coefficient which reduces
as the forward current is increased. At a
figure approaching 75 mA the forward
voltage is practically temperature
independent which can of course be
very useful. When the forward current is
increased beyond this point the diode
then behaves with a positive temperature |
coefficient. All very interesting but not
of great importance to us in this
instance.

What is significant, however, is that all
the lines in figure 1 are straight. In fact
this linearity continues at temperatures
below freezing point. Thus, the BAX 13
would make an ideal temperature sensor. 1
Furthermore, most of the common

J. Borgman

1980 - 9-13

Figure 2. The circuit diagram of the linear thermometer. Using the 1N4148 as a sensor, the temperature can be displayed with the aid of a
moving coil meter or a digital voltmeter with 'floating' inputs.

0- 30 0-300 pA - 30
0- 30 O-IOOpA - 30
0- 50 0-300 pA - 50
0- 50 0-500 pA - 50
0-100 0-1 mA -100

• 2 x 3k3 in

arallel.

- 30° C
► 50° C
f 50° C
-100°C

Table 1. A number of alternative values fo

types of diodes have similar character-
istics providing the forward current is
kept as constant as possible.

The circuit diagram

The circuit diagram shown in figure 2
was first published in the 1979 Summer
Circuits issue and many readers
requested a printed circuit board for it
which is now available. A suitable diode
to use as a temperature sensor can be
the common 1N4148. Its forward voltage
drops by about 2 mV per rise in C.

If a diode is to be used as the basis for
temperature sensing it is important that
two main conditions are met in the
circuit. Firstly, as previously mentioned,
the forward current through the diode
must remain as stable as possible.
Secondly, the circuit measuring the
forward voltage of the diode must have
a high impedance.

The sensor diode, D1, is included in a
bridge network which is supplied with
the reference voltage from a 723 1C. At
0°C the bridge must be completely
balanced, that is, there must be no
voltage difference between the two
inputs of IC2. The 741 will, in fact,
ensure this itself since R7 is connected
to the inverting input and constitutes a
feedback path. The 1C will maintain or
make the voltage across R7 equal to the
voltage at its non-inverting input.

When the temperature of the sensor
diode is 0°C, the voltage across R7 will
be equal to the voltage across R6 plus
part of the preset PI . The output of IC2
is then adjusted to zero by PI, effectively
balancing the bridge. The display of the
thermometer is a meter in the output of
IC2. The meter has been connected in a
bridge rectifier circuit and it will there-
fore read in one direction only. This
then makes the adjustment of PI a

simple matter.

Any variation in temperature will result
in a change in the forward voltage of the
sensor diode D1 . Since the voltage across
R7, D1 and P2 is the reference voltage
of the 723, and therefore constant, any
change in the forward voltage of D1 will
result in a voltage variation across R7.
This will be immediately detected by
IC2 which will react by passing a small
current, via the meter, to P2 to compen-
sate for the change. Any variation of the
current through D1, due to changes in
forward voltage, will therefore be
avoided by the reaction of IC2. Thus
the path through the meter bridge circuit
and R8 is a servo control loop to
maintain the current through D1 at a
constant level. The current level through
R8 (the current which counteracts the
change in voltage across R7) reflects the
temperature of the sensor diode. The
meter will, of course, indicate this and
can be provided with a scale graduated
in degrees.

As mentioned, the meter is connected in
a bridge rectifier circuit. This means
that the meter will always read in the
same direction, regardless of whether
the temperature of D1 is above or below
0°C. In other words, the meter will give
an identical reading for both +10°C and
-10°C. Some indication is needed to
show whether the temperature is above
or below zero.

So far, only the reference voltage section
of the 723 has been used. Since this 1C
also contains an opamp with a transistor
output, this could be used for the
indicator circuit, with the addition of a
few other components. The opamp is

ft

.

«

used as a comparator with the output
of IC2 and the reference voltage being
connected to the non-inverting and
inverting inputs respectively. Assuming
that the circuit is calibrated for zero
deflection of the meter at 0°C, a fall in
temperature will result in an increase in
the output level of IC2. This will take
the non-inverting input of the opamp
high and with it the output. Transistor
T1 will turn on, lighting the LED. When
the temperature rises above zero, the
reverse process will occur and the LED
will be extinguished.

If a digital voltmeter with a 'floating'
input is available, the meter described
above may be omitted. The DVM may
be used to measure the voltage across
R8 in the feedback loop. This will
correspond to the current passing
through it and therefore to the tempera-
ture of the sensor diode. If this option is
used the output of IC2 can be directly
connected to R8, since the meter, its
bridge rectifier and the temperature
polarity indication circuit (R 1 ... R4,
T1 and D6) need not be used. It will be
obvious that neither of the terminals of
the DVM must be earthed. The polarity
of the temperature (above or below 0°C)

will be indicated by the positive or
negative sign of the DVM display.

Construction and calibration

The construction of the Linear
Thermometer should not present any
difficulties if the printed circuit board is
used. The layout for this is shown in
figure 3. Room on the board has been
allowed for the small 12 V 100 mA
mains transformer required. The
completed printed circuit board may be
fitted in a type BOC430 plastic case
from West Hyde or the 65-2518 H from
Vero. The mounting holes on the board
have been drilled to fit either of these
boxes.

For the temperature sensor the 1N4148
diode is recommended. This may be
mounted at a reasonable distance from
the circuit board if desired. For air
temperature measurements, the diode
can be used without any form of cover,
provided it is protected from accidental
damage by some means. To measure the
temperature of electrically conductive
fluids, the diode will need to be electri-
cally insulated. The insulation should be
as thermal as possible for obvious

reasons.

Depending on which type of meter is
chosen, it may be necessary to adjust
the values of R8A and R8B as well as
the limits of the measuring range.
Table 1 gives some guidelines for this
purpose. The connection points for a
DVM are shown in figure 2.

Before calibration can begin, a quantity
of ice must be produced from distilled
or demineralised water (available at the
chemist's). The ice is crushed and placed
in a position where it can melt slowly.
The sensor diode is then put in the
melting ice and PI is adjusted for a zero
reading on the meter. This then is
freezing point calibrated, now for the
other end of the scale. With P2 in a
central position, the sensor diode can be
placed in boiling water (also distilled or
demineralised). The voltage across R8
can now be set to exactly 1 V by P2.
This should complete calibration but if
a reliable reference thermometer is to
hand, further checks can be made on a
comparison basis. M

Large computers can consist of over a hundred thousand logic circuits
and are able to execute more than a million instructions per second.
Fast as this may seem, a computer 20 times faster than this is currently
being developed by IBM. It stems from a brand new branch in
technology, namely, the Josephson technique. Its most striking aspect
is that it can only function at extremely low temperatures, at which
most life has come to a complete standstill, for it is then that electrons
move with an increased velocity.

superconductors supercede semiconductors

die Josephson
computer

1 . An electron microscopic photo of part of a Josephson chip. Similar t
semiconductor technology, circuits may be photolithographically mini.

as nothing to do with semiconducting. This is
oseconds (50.1(T 12 seconds). Even faster swi

There are two ways in which to expand
a computer's capacity: either by includ-
ing more logic circuits, or by enabling
them to work at a higher speed. The
Josephson computer, the 'super brain'
of the near future, draws its strength
from its rapidity.

The speed at which a computer carries
out its instructions is measured in the
cycle time or clock generator period.
The large computers in operation today
have a cycle time of around 3050 ns
(nanosecond = one millionth of a sec-
ond). The world's fastest computer,
which oddly enough is not an IBM
design, but a CRAY (small scale special-
ist industry), has a cycle time of 12 ns.
With the aid of the Josephson technique
it is hoped to reduce this to 1 ns. In
actual fact, the first prototypes will
probably have a cycle time of 2 ns, but
even this amounts to their being
20 times faster than the large present-
day computers. Apart from their speed,
the prototypes' performance will re-
semble that of the IBM 370/168, one
of the biggest existing computers.

Achieving such a
short cycle time
is not only a mat-
ter of searching
for high-speed
logic circuits. It
also involves solving the problem per-
taining to the transport of countless
electrical signals. In one nanosecond an
electrical signal can only travel about
15 cm, which means if that is to be the
cycle time, the dimensions of the entire
computer will have to be no more than
15 cm. For this reason, the Josephson
computer as designed by IBM will be
13.5 x 13.7 x 14cm.

The question is now: will the hundred
thousand logic circuits required by an
extensive computer be able to fit into
such a tiny space? Yes, by means of
large scale integration (LSI) which
modern technology has fortunately
already achieved. Several tens of thou-
sands of chips can be integrated. How-
ever, if they belonged to the semi-
conductor type, the circuit would be
doomed to disintegrate after a very
brief lifespan, as they would dissipate
several kilowatts.

Thus, what is needed is a technology
including similar miniaturisation possi-
bilities, but which at the same time
produces higher speeds and much less
dissipation. All this is achieved by the
Josephson technology. Figure 1 shows
the result: a Josephson chip.

Superconductors and electron
tunnels

In 1962 while still a university student,
the British physicist Brian D. Josephson
laid the theoretical foundation for the
Josephson effect named after him. It is
based upon two physical phenomena:
superconductivity and electron tunnel-
ling.

Superconductivity was discovered in

9-16 -

191 1 by a university professor at Leiden,
Heike Kamerlingh Onnes. He noted that
certain metals (superconductors) lose all
resistance to electric current flow when
they are cooled to below a certain tem-
perature (which is different for each
superconductor). The resistance literally
drops down to zero ohms. Kamerlingh
Onnes found that superconductivity will
only take place if the current is main-
tained at a certain level. If it rises above
that value, the metal will start to act
as an ordinary conductor, despite its
being sufficiently cooled. It also ap-
peared to be possible to disturb the
superconductivity with a magnetic
field.

It was only in 1957 that a satisfactory
explanation for this phenomenon could
be given. One of the people responsible
was John Bardeen, one of the three
inventors of the transistor. What it
comes down to is that in the super-
conducting state an electric current
must not be regarded as a stream of
'single' electrons, but of 'pairs' (Cooper
pairs, named after another founder of
the theory). The electrons belonging to
such a pair now move in step with each
other, so to speak, and no longer need
to 'cling' to the atom nuclei. With each
other's help they shoot between the
nuclei. Superconductivity stops when
the electron pairs become separated for
some reason. This may be due to an
increase in temperature or current, or
due to a magnetic field. Strictly
speaking, superconductivity is only valid
for direct currents; as for alternating
currents, they cause a slight deviation
from the 'ideal' superconductivity until
far into the high frequency range.
Whereas superconductivity was ex-
plained long after its discovery, with
electron tunnelling (the tunnel effect)
it was quite the opposite. The theory
has been in existence for some time
before the phenomenon could be
demonstrated during the sixties. It has
nothing to do with superconductivity
and in fact also occurs at everyday tem-
peratures. It is the tunnel diode, often
applied as a gigahertz amplifier or as a
fast switch, which makes use of the
effect.

Contrary to what might be expected,
a thin insulator between two conductors
will allow an electric current to pass.
Thus, despite the fact that its ohmic
resistance is infinite, current will flow.
This involves quantum mechanics and
so is rather complicated. Basically, it
means that the electron should not
only be regarded as a particle, but also
as a wave phenomenon. It 'rebounds'
as it were against the barrier formed
by the insulator, but being a wave it
penetrates it slightly, provided that the
insulator is not too thick.

Josephson: superconducting
tunnelling

Josephson combined the two physical
phenomena by applying the electron

tunnelling theory to electron pairs
responsible for superconductivity. This
is because an electron pair can also be
considered as a wave. Remarkably, the
thin insulator, which really should not
allow any current to pass at all, was now
found to act as a superconductor. This
occurred when the metals around it
were in a superconducting state.

This effect is called the Josephson
effect. A year later is was also observed
in the American Bell laboratory.

A thin insulator between two super-
conductors is called a Josephson junc-
tion. This is the principle behind the
Josephson computer. Since supercon-
ductivity only takes place at very low
temperatures, the entire computer is
cooled by submerging it into liquid
helium. Its boiling point is around
4.2 degrees Kelvin (— 269°C). Thus,
more than anything else, the Josephson
computer has to 'keep its cool'. There
are additional applications for the
Josephson effect outside the computer
field. It can, for instance, be used to
measure tiny magnetic fields and volt-
ages, and can also be applied in micro-
wave technology.

The Josephson junction used as
a switch

As we have just seen, a superconducting
material can be brought out of this state
in three different manners: by an in-
crease in temperature, an increase in
current and by the creation of a mag-
netic field. This is not only true of
superconducting metals, but also of the
Josephson junction - even more so, in
fact. That is why Josephson called it a
'weak super conductor'. When it is

brought out of its superconducting
state, it does not start acting as a normal
conductors, like metal, would, but as an
ordinary tunnel junction. In practice,
this means that the Josephson junction
then demonstrates a resistance of a few
hundred ohms, so that it is possible to
switch its resistance from zero to a few
hundred ohms. The Josephson com-
puter uses this phenomenon.

Switching from a superconducting to a
resistant state, and vice versa, happens
at a speed that few physical processes
can match. The switching time is around
6 picoseconds (a picosecond is one
billionth of a second), less than 1% of
the 1 ns cycle time IBM seeks to achieve
in the Josephson computer. By way of
comparison, the fastest semiconductor
switches take ten times longer.

This incredible speed is not the only
advantage the Josephson junction has
to offer. Its dissipation (heat develop-
ment) when in a superconducting state
is nil, even when a current of 0.1 mA is
flowing. For, after all, its resistance is
also nill In fact, dissipation will be very
low, even in the resistant state, as the
circuit's supply voltage will be approxi-
mately 10 mV.

A Josephson computer including a
16 Mbyte memory capacity is therefore
expected to dissipate a mere 7 watts of
electrical power. What a difference,
compared to the amount of kilowatts
produced by present-day computing
monstersl

This does not mean that a Josephson
computer will not lead to high elec-
tricity bills. On the contrary, cooling it
to 4.2 Kelvin will require about
15 kilowatts. Cooling techniques have
fortunately long since been developed.

2

Figure 2. The Josephson computer will have an unusual appearance. Most of it will be taken up
by the cooling system required to maintain 4.2 K (-269°CI. This requires a lot of energy:

15 KW, whereas the computer itself needs only 7 WO). A: compressor for the cooling;

B: cooling system; C: interface and supply operate at room temperature; D: in- and output
connections; E: the actual computer; F: liquid helium at 4.2 K.

ilektor September 1980 -9-17

3

f

Figure 3. The structure of a Josephson
junction (3a) and its voltage/current ratio
curve (3b). The Josephson junction consists

very thin insulator, the Josephson barrier,
between them.

The operation of a cooling installation
(cryostat) hardly differs from that of a
domestic fridge. Its design is such, that
it can be switched off for as long as a
hundred hours at a time without any
detrimental effect to the superconduc-
tivity. Figure 2 outlines the installation.
It consists of a cryostat able to hold
460 litres. A compressor takes care of
the cooling. The actual computer, the
block of less than 4 litres, is submerged
in the cryostat.

The U-l curve

The relationship between the current
passing through a component and the
voltage across it can be expressed in
the form of a graph: the U-l curve. The
U-l curve of a Josephson junction is
shown as a circuit diagram in figure 3a
and as a graph in figure 3b. It is rather
unusual, as it has two curves. It looks
as though at certain I current values the
voltage U can assume two different

What happens if the current through a
Josephson junction is allowed to rise
above zero? First we remain within the
lefthand plot of the curve. The current

4

Figure 4. The influence of a magnetic field
on the voltege/current ratio curve (4a). The
diagonal is the load line caused by including

increases, but the voltage is still 0 volts.
Its resistance is nil and it is in the
superconducting condition. This con-
tinues until the current rises above
I max. for as soon as this happens the
junction will cease to be in a super-
conducting state. Thus, we jump (liter-
ally) to the right branch of the curve
and there is a voltage across the junc-
tion. If the current is allowed to drop
down below l max we continue to
remain in the right branch of the curve,
for there is still a voltage across the
junction. The superconducting con-
dition will only occur again, therefore,
if the current is reduced to below l m j n ,
or if the voltage is brought down to
below U m j n , which comes down to the
same thing.

Thus, the Josephson junction can be
made to switch from a superconducting
state to a resistant state by increasing
the current passing through it very
briefly. Switching back to a super-
conducting state is achieved by
decreasing it very briefly. The Josephson
junction has a memory function: the
superconducting and resistant states
both being stable, they can be fixed.
This distinguishes it from a transistor.

for in the latter's case at least two
transistors are required to store a single
bit. In this manner, a memory com-
ponent is prevented from dissipating
power in either of its two conditions.

Magnetism

By varying the current passing through
a Josephson junction, we can switch it
from its superconducting state to a
resistant state and back again. But this
isn't always convenient, as in elec-
tronics we prefer to switch a current
with the aid of another independent
current or voltage. Ideally, the
Josephson junction should have a
base- or gate-electrode, or something
similar. Fortunately, this appears to be
possible and not even all that compli-
cated.

Use is made of magnetism. The values
'max and Umin < as als0 'min) are found
to depend on the size of the magnetic
field. This is shown in the curve in
figure 4a. The maximum super-
conducting current l max drops to
'max 0 when a magnetic field is applied.
The minimum voltage in the resistive
state U rn in w'" then drop to U m j n 0 .
Thus, the Josephson junction can be
preset at a fixed current 1 1 . If a mag-
netic field is applied, it will switch from
its superconducting state to normal
conductance. Similarly, it will switch
back to superconductivity when the
magnetic field is removed. For this to
happen, there must be a voltage U] of
between U m j n 0 and U m j n across the
junction. This is done by switching it in
series with a load resistor R|_, as drawn
in figure 4b. The diagonal Ij-'-'b ' n
figure 4a is a load line, like the ones
found in transistor graphs. It indicates
the various current/voltage combi-
nations which are possible after the load
resistor has been added. Switching
happens along this line.

Since the supply voltage Ub is very
low (a few millivolts), very little power
is dissipated in the resistor. In the
resistant state about 0.5 /jW is con-
sumed. Often self inductances are
employed for the load in the Josephson
technique.

How is a magnetic field generated?
Simply by emitting an electric current
along the Josephson junction, for each
current has its own magnetic field
around it. Since the junction is highly
sensitive to the magnetic field, a small
current is all that is necessary. Figure 5
gives an enlarged view of the way in
which a Josephson switch can be
introduced into a chip. Above the
Josephson junction there is a control
channel through which the control
current l c flows. The arrows indicate
the magnetic field created by the
current. Current I through the junction
is affected by the much smaller control
l c . The Josephson switch is a current
controlled current switch. In 1965 the
IBM technician, Juri Matisoo, succeeded
in making and testing such a switch.

9-18 — elektor September 1980

Standard component: the SQUID

It would be an advantage if the junc-
tion's sensitivity to magnetic fields were
at a maximum, for then the control
current could be low. This is achieved
by making the surface area of the
junction as large as possible. However,
this also has disadvantages: for not only
does a large junction naturally occupy
more chip space, but it will also switch
more slowly. The Josephson junction
has a capacitance, a slowing-down
factor, which increases with junction
surface area.

This dilemma has been solved by the
development of the SQUID, a kind of
Josephson 'standard component' with
two or more small Josephson junctions.
In the SQUID use is made of the coop-
eration of different Josephson currents.
This cooperation is rather complex and
may be compared with the interference
of wave forms (light, for instance). It is
connected with the fact that Josephson
currents tend to be unevenly distributed
over a Josephson junction in the pres-
ence of a magnetic field. Figure 6

5

Figure 5. Highly enlarged outline of a Josephson switch. The current I is switched under the
influence of a magnetic field 0 (indicated with arrows! inducted by the control current l c .

6

illustrates this. It shows what the
current distribution in increasingly
magnetic field looks like when the
current through it is equal to the
maximum superconducting current
•max- As the magnetic force augments,
•max will decrease and may even equal
0 at a magnetic flux of 0 2 . At that level
in magnetic force, the Josephson cannot
attain a superconducting state, however
small the current flow through it. If the
magnetic field is further increased
however, l^eix will rise again. Not
drawn is how l max will again equal
zero, if the magnetic field force is
further increased. This is why the mag-
netic field to maximum super-
conducting current ratio forms such a
peculiar, periodical curve (figure 7).

Now the two Josephson junctions may
be made so that at a certain magnetic
field force, one junction can easily
become superconducting and the other
not at all, while at another level the
situation will be exact reverse. The two
junctions are both controlled by the
same control current. Thus, the control
current sends the controlled current
into a certain direction: either via one
Josephson junction, or via the other.
This is what occurs in a SQUID, a
'Superconducting QUantum Inter-
ference Device'. It enables the sensi-
tivity of a large Josephson junction to
be combined with the speed of a small
one. SQUIDs can be made in all sorts
of versions and may include more than
two Josephson junctions.

Logic circuits

SQUIDS form the pillars of the

Figure 6. When a magnetic field 8 is created the current through a Josephson junction is not Josephson computer. They enable all

evenly distributed. At certain levels of magnetic force, the total current through the junction the known logic circuits in semiconduc-

may even become nil (82). tor technology to be made: inverters,

gates, flipflops. An AND gate, for

Blektor September 1980 — 9-19

instance, may be made by controlling a
SQUID with not one control current
but two. Then the two control channels
will be created above the junctions. The
SQUID will switch only if both the
control currents are large enough. Such
a type is also called a 'current injection
device’, as shown in figure 8. Similar
circuits make OR gates possible.

As for flipflops, these can be constructed
in various ways. One of the most
interesting makes use of induced
superconducting loop currents, for such
currents flow indefinitely!

Alternating voltage

A remarkable characteristic inherent to
the Josephson junction is its perfectly
symmetrical non-polarised structure.
This means it can be connected either
way around. What is more, a Josephson
circuit may be equally well fed with an
alternating voltage, which is the case in
the Josephseon computer. The great
advantage here is that the supply voltage
will then function as a clock signal. In
other words, its stomach is also its heart
and so quite a few electrical connections
may be omitted. This also helps to reset

The power supply of the first prototype
computer will consist of a 500 MHz
sine-wave oscillator producing 7 watts
power. It is 'on dry land' and so is not
cooled. On each of the more than ten
thousand chips included in the Josephson
computer, there will be a number of
voltage controllers to limit the voltage
to an upper threshold of 12 millivolts.
The sine-wave will then have become a
square wave. By voltage limiting in
many places interference between signal
paths is avoided.

The synchronisation of such a high
speed computer poses a serious problem.
During a two nanosecond clock period
an electrical signal will only travel
30 cm. Highly specialized techniques are
needed to ensure that processes occur
simultaneously in the way they should.

The material

Although it should not be taken lightly,
a Josephson computer is not difficult to
manufacture. This is because familiar
techniques may be used on a large scale,
such as the manufacture of semiconduc-
tor IC's. Although the materials used are
different, the procedure is very similar:
layers are evaporated, patterns are
applied photolithographically and
etched. Complicated semiconductor
processes like diffusion and implantation
are not even necessary for Josephson
chips. On the other hand, more layers
are applied (ten to fourteen, instead of
three to six) and the tunnel barrier (the
insulator between the superconductors)
is very hard to make, because it has to
be so thin.

The materials used in a Josephson
computer have to comply with two
obvious requirements; they must be

7

Figure 7. As a result of the uneven distri-
bution drawn in figure 6, this remarkable
relationship between current I through a
junction and the magnetic field 0 is
established. Such an effect is used in the
'Josephson standard component', the
SQUID.

capable of sustaining freezing tempera-
tures and great changes in temperature.
After all, a Josephson computer is built
and probably repaired at room tempera-
ture. Because of the fluctuations in
temperature, the materials used must
have similar expansion coefficients,
among other things, which reduces the
choice of possibilities considerably.

The latter aspect is quite a headache for
IBM technicians. It is true that enormous
progress has been made (the error factor
after 400 temperature cycles has already
been brought down from 99% to 0.1%),

but there are so many chips that the
sensitivity to temperature changes is still
far too great.

Josephson chips are based on silicon,
like the semiconductor type. Silicon was
chosen because its use in the semi-
conductor industry is well established.
Some experts believe that more is
known about silicon than about any
other material on earth.

Contrary to semiconductor chips, the
silicon here takes no part in the electrical
process. Thus, its semiconducting charac-
teristics are not at all involved, for the
Josephson computer silicon is merely an
insulator. The fact that it is also a good
heat conductor is an extra advantage.
Different insulating and protective layers
are made from another material used in
semiconductor technology: silicon
oxide. The superconducting layers
consist of the metal niobium or of a
lead alloy (with bismuth, or with
indium and gold).

The Josephson barriers made from lead
and indium oxides are subjected to very
tough requirements. They are no thicker
than 4 to 6 nanometres, about thirty
atom diameters (the other layers are
about 100 nm thick .) Furthermore, the
density of that layer is highly critical,
as the maximum superconducting
current depends on it exponentially.
What it comes down to is that the layer
must be made in such a way that the
average density may only vary from the
standard by less than an atomic di-
ameter. This is like covering an acres
with a layer of soil three centimetres
thick without it varying anywhere by

Figure 8. A current injection device, one of the methods to produce a logic AND gate by means
of the Josephson technique. The gate works with two Josephson junctions, shown as vague
circles in the dark horizontal rectangle. The left-hand junction has five times the surface area of
its right-hand counterpart. The smallest dimensions are around 2.5 pm, which is as tiny as an LSI
semiconductor chip (IBM photo).

9-20 - elektor September

the Josephson computer

more than a millimetre. This feat
required brand new evaporation
techniques.

Soldering with mercury

The chips couldn't be connected until
another new technique was developed,
for of course the Josephson chips
couldn't just be mounted onto printed
circuit boards. Apart from the undesir-
able cooling effects, the computer with
its more than ten thousand chips would
be far too big.

How it is put together is shown in
figures 9 and 10. Figure 9 displays a
module about 30 x 25 x 15 mm. The
chips here are packaged very closely
together. Not only do the substrates of
the chips consist of silicon, but so does
the rest of the module. Without any
other form of case, the chips are
mounted onto the small cards face
down, according to the 'bonding' process
(from the semiconductor technology).

block meas
(shaded arc
consist of mono crystalline silicon, aloi
affixed by photolithographical means.

elektor septen

1980-9-21

This makes the heat transfer to the
liquid helium highly efficient.

The chip cards are connected to the
larger card by means of minute
connectors. These have 'micropins'.
A micropin is 0.2 mm long and
0.075 mm in diameter. Larger pins
would cause the magnetic field to be
too large, which would slow down the
signal transfer and cause cross-talk. The
individual distance between the micro-
pins is half a millimetre, (an ordinary
DIL-IC's pins are 2 mm apart). The
micropins are connected, not with
solder, but with mercury, as this
solidifies at low temperatures (below
-40°C). At room temperature, the
liquid mercury is stored in drops with
a 0.4 mm diameter in specially made
cavities.

Four modules (which sometimes vary in
size) in figure 9 are combined to form a
"W module'. Twenty one such modules
constitute the entire computer shown in
figure 10. More than the thousand chips
have been collected in a block of less
than 14 x 14 x 14 cm. The CPU and the
fast 32 K byte scratch memory are both
included in one of the twenty one
W modules. The other twenty are
occupied by the large 16 M byte main
memory. Once it is submerged in liquid
helium, this block outshines all present-
day computers.

Why a Josephson computer?

It is almost certain that a Josephson

computer can be built. Already,
complete 16 K RAMs and CPU chips
have been made according to the
Josephson technique. Problems left to
be solved involve developing a sufficient
resistance to temperature changes, as
mentioned above. Furthermore, suitable
fast I/O systems still have to be created,
for without its 'hands and feet' even the
most cold blooded of brains will not be
up to much. At this stage, however,
these problems seem unsurmountable.
Whether a Josephson computer will ever
be a market product is hard to say. The
microprocessor is threatening to put an
end to the golden age of the large-size
computer. Fewer calculations are dealt
with 'centrally' in favour of the small,
specialised microcomputer systems. It
does not look as if the world is desper-
ate for even larger and faster computers.
Nonetheless, IBM must see some future
in its Josephson computer, as otherwise
it would not have put so much time and
effort (and money!) into research.
There are still a number of fields where
the present-day computing monsters
lack ability. Computer simulations of
physical or economic processes, for
instance, could be far more accurate and
take place on a larger scale. Computer
simulations are also important in
weather forecasts and in some fields of
purely scientific research (nuclear
physics). And of course then it could
have military applications.

Another field which would welcome a
huge computing capacity is pattern

recognition, in which the computer
interprets sound (speech) and video
signals (written text, video and radar). A
third possibility is provided by the great
data banks which must be accessible for
many users at once.

In any case, IBM is already speculating
upon fields which are still reserved for
the ordinary semiconductor micro-
computers and which one day, in the
distant future, could well be taken over
by Josephson technology . . .

Sources:

Spektrum der Wissenschaft, Juli 1980:
'Superleitende Computer',

Juri Matisoo;

IBM Research Highlights, June 1979:
'Experimental IBM circuits are the
world's fastest';

IEEE Spectrum, May 1979:

'Computing at 4 degrees kelvin',

W. Anacker (IBM). M

Temperature

A word on temperature, which is a
peculiar concept. It is quite different
from heat. Heat can but cause a change
in temperature, no more. In modern
physics temperature no longer has much
to do with cold or heat. Rather it is
considered as a gauge of the trembling
pertaining to atomic nuclei. Atomic
nuclei do not stay in a fixed position,
but move around a fixed point. Using a
little imagination, a particle of matter
may be seen as a swarm of mosquitoes.
The swarm remains stationary, whereas
the individual mosquitoes are highly
mobile.

The more atomic nuclei in movement,
the higher the temperature. To the
physicist, therefore, temperature is
inherent to matter- it is one of its
characteristics.

If the temperature of a piece of matter
is lowered, the atomic nuclei start
moving less wildly. This is true of all
material. If the temperature is low
enough, the atomic nuclei will become
immobile. Since temperature is a
measure for atomic mobility, it is not
strange that the temperature at which
the nuclei become immobile is the same
for all kinds of matter.

Standing still is as immobile as you can
get. This brings us to the conclusion
that a lower temperature will therefore
not be possible. For this reason, the
temperature at which atomic nuclei
stand absolutely still is called the
absolute zero. It is somewhere around
— 273.4°C. It can also be expressed as
OK (in the past: 0°K). K represents
Kelvin, the absolute temperature unit.

The tunnel effect

The tunnel effect is based on the fact
that a thin insulator applied between
two conductors will allow an electric
current to pass. The phenomenon is
explained in quantum mechanics. What
this boils down to is that particles do
not have a certain fixed mass, speed,
energy, etc., as was believed in classical
(newtonian) physics. A random distri-
bution is involved. You could say that
in classical physics a particle used to be
considered as a hard little globule
moving in perfect orbits at a well defined
speed, whereas in quantum mechanics
everything is much 'haxier'. Here a
particle looks more like a cloud, not
clearly circumscribed, but ending
'somewhere'. The averages of the various

chance distributions of mass, speed and
energy will however still be the same as
in classical physics.

Classical physics has no explanation for
the tunnel effect. According to it, the
particles - electrons — would all have
too little energy to be able to penetrate
the thin insulator barrier. Quantum
mechanics states, however, that although
the average energy of the various
particles would be deficient, nevertheless
it is possible for one particle to have
enough energy. Some particles must
therefore be almost immobile, whereas
others are highly mobile.

In other words, according to quantum
mechanics, the particles are no longer all
identical.

9-22 - elektor September 1980

VOX prir

UK pinted
circuit board

particular voice pattern, since both the
bandwidth and the centre frequency are
adjustable (with P2 and P3 respectively).
The two other potentiometers are to
preset the input sensitivity (PI) and the
length of the delay (P4). In practice, the
range of PI should be adequate since
the gain of the microphone preamp A1
can go up to 100 X.

Delay time however can be a matter of
personal preference. With the values of
P4, R20 and C7 as given in figure 1 the
delay time is adjustable between 0.5 and
2.5 seconds. The range can be altered by
changing the values of any or all of
these components.

The layout of the printed circuit board
for the VOX is shown in figure 2.
Everything in figure 1 is included on the
board with the exception of the two
stereo potentiometers, the relay and the
microphone.

For a more precise description of the
VOX circuit readers are referred to the
previous article published in Elektor 56
(December 1979). H

The VOX switch published in the
December 1979 issue of Elektor
attracted a lot more attention
than expected and it is for this
reason that a printed circuit board
has now been produced.

A brief recap of the original article will
be useful to those readers who are
unfamiliar with the purpose of a VOX
switch. Basically it is a voice operated
electronic switch, normally used to
operate a transmitter/receiver. It can
have other uses of course, but its main
purpose is to allow the hands to be free
when using the microphone. As soon as
a sound is picked up by the microphone
the VOX will switch the transmitter/
receiver to 'transmit'. At the end of the
speech passage the VOX will switch
back (after a short delay) to 'receive'.
The delay is presettable and is included
to cater for breaths and hesitations.

This VOX is fine but there are drawbacks
in practice, Sound picked up by the
microphone can include squeaky chairs,
doors closing or even beer cans popping
open, should this occur in the vicinity.
It is not of course desirable for the trans-
mitter to switch on in these instances.
The Elektor VOX switch avoids this
problem by the addition of a filter
designed to exclude all frequencies
other than those in the speech band.
The filter can be made 'active', to a
certain degree, to the characteristics of a

Resistors:

R1,R3,R4,R10,R13,R16 - 10 k

R2.R1 7 = 47 k

R5,R6,R7,R14,R19 = 22 k

R8.R11 = 3k9

R9.R12- 1k2

R15- 100 k

R18 = 4k7

R20 = 220 k

R21 ,R22 = 6k8

Capacitors:

Cl - 1 n (MKM)
C2.C3 = 22 n
C4.C5.C10 - 100 n
C6 - 2p2/16 V
C7 = 4p7/16 V
C8.C9-220 p/16 V
C11 = 100 p
Cl 2= 27 n

T1 ,T2,T3 = TUN
T4 = TUP
D1,D2,D3= DUS
IC1 = TL 084
IC2 = 4528

Miscellaneous:

PI ,P4 = 1 M preset
P2 = 1 M lin.

P3 = 10 k log.

LI - 5 turns 0.1 . . . 0.25 CuL
wire on ferrite bead

Britain's first com
computer kit.

The Sinclair ZX80.

£ 79?- 5

Please note: many kit makers quote VAT-exclusive pnces

Fewer chips, compact
design, volume production -
more power per pound!

The ZX80 owes its remarkable low price to
Its remarkable design the whole system is
packed on to (ewer, newer, more powerful
and advanced LSI chips A single SUPER ROM,
tor instance, contains the BASIC interpreter,
the character set, operating system, and
monitor And the ZX80s IK by te RAM Is
roughly equivalent to 4K bytes in a conven
tional computer - typically storing 100 lines of
BASIC (Key words occupy only a single by te )

The display shows 32 characters by 24 lines
And Benchmark tests show that the 2X80
is faster than all other personal computers
No other personal computer otters this
unique combination ol high capability and

1

• Ready moulded case

sir -

sirn=lE
zxso

Science of Cambridge Ltd

iplete

The Sinclair teach-yourself
BASIC manual.

If the soecifications of the Sinclair ZXSO
mean little to you -don't worry They're all
explained in the specially -written 128-page
book free with every kltl The book makes
learning easy, exciting and enjoyable, and
represents a complete course in BASIC
programming - from first principles to complex
programs (Available separately -purchase
price refunded if you buy a ZX 80 later i
A hardware manual is also included with

The Sinclair 2X80. Kit: £79.95.
Assembled: £99.95. Complete!

The ZX80 kit costs a mere £79 95 Can't
wait to have a ZX80 upandrunn!ng?No
problem! It's also available, ready assembled
and complete with mains adaptor, tor
only £99 95

Oemand tor the ZX80 is very high use the
coupon to order today tor the earliest possible
delivery All orders will be despatched in slnct
rotation We'll acknowledge each order by
return, and tell you exactly when yourZXSO
will be delivered If you choose not to wait, you
can cancel your order immediately, and your
money will be refunded at once Again, of
course, you may return your ZXSO as received
within 14 days for a full refund We want you to

no doubt that you will be

I nonco To: Science of Cambridge Ltd. 6 Kings Parade
criDsi Remember all prices shown include VAT. posta
rUHM Please send me

measuring mill

Multipath Distortion

Multipath distortion; an unpleasant
phenomenon, especially with FM stereo.
Often the only thing you can do to
combat it is to empirically rotate the
aerial. What's important is that the dis-
tortion produced by the receiver can
even be recognized as a multipath pro-
blem. For quality FM receivers, there-
fore, a multipath meter is highly desir-
able. The standard recipe for such an
indicator can be improved on as this ar-
ticle will show.

Figure 1 . Multipath distortion is caused when the transmitted signal reaches the receiver via two
(or more) paths of different lengths.

of multipath. The amount of interference
incurred depends on the kind of modulation
applied and on the wav in which demodula-
tion takes place in the receiver.

If during AM (Amplitude Modulation) a few
side band components are amplified or weak-
ened. this will merely cause the demodulated
signal to be 'coloured'. If, however, the mo-
dulation depth at peak amplitude is greater

to short wave radio will be familiar with this

phenomenon. Where distant radio stations
are concerned, short wave signals constantly
reproduce themselves along various paths
through the ionosphere.

Multipath indicator circuit

In figure 3 the block diagram of an FM recei-
ver has been drawn in which the ordinary S
meter is expanded with the additional facili-
ty of multipath indication. With the switch
in position 1, it operates as a multipath indi-

derived by detecting the outputs of all the IF
amplifiers (here A1, A2 and A3) and adding
them. Why not just the output signal of A3?
Because A3 will already be limited to average

9-28 - ele

1980

The heart of the extension circuit
consists of two multiplexers, IC1 and
IC2. The information at one or the
other set of inputs is passed to the
outputs of the multiplexers depending
on the (logic) state of the select input.
The data lines of the keyboard and
those of the memories are each connec-
ted to a separate 'group' of inputs.
When the select input is taken low the
keyboard data will be passed through to
the outputs, and when the select input
is high the data from the memories will
pass through. To be able to store the
memory contents on tape, therefore,
the select input will have to be taken
high. This is accomplished as follows:

address counter so that the data stored
in the next memory location becomes
available. When the WRITE pulse occurs
(immediately afterwards) the UART
will operate once more.

The entire cycle is repeated until the
complete page has been 'dumped'. The
'end-of-page' pulse (RP) inhibits both
the DAV pulse and the R/W pulse via
FF2, N6 and N7.

By pressing the reset button (S2 of
figure 1) FF1 and the RS flip-flop
(N1/N2) are both reset and the
Elekterminal can be operated normally
once more. The information stored on
cassette can be re-entered via the serial

high speed readout
Km* ek'khrniiual

With a minor modification to the
Elekterminal it is possible to 'store'
the entire contents of the display
(TV screen) on a cassette tape.

The majority of the connections
can be wired to the existing
expansion sockets. For the
remaining connections just three
of the copper tracks between the
UART and the CRTC on the main
board of the Elekterminal have to
be broken.

When the start button, SI of figure 1 , is
depressed, the 0 output of FF1 goes
low. As the Q output of FF 1 is high, the
memory WRITE signal is inhibited by
N4. When one of the keys on the ASCII
keyboard is depressed, preferably the
space key or a control key, a strobe
pulse (KS) will be generated. The
leading edge of the strobe pulse will
cause the data from the keyboard to be
entered into the UART in parallel. With
switch SI in figure 2 closed, this
information will be passed out of the
transmitter section of the UART and
back into the receiver section in series.
Once a complete character has been
transferred in this manner the Data
Available (DAV) output, pin 19, of the
UART will go high for a short period of
time. The trailing edge of this pulse will
set the RS flip-flop formed by N1 and
N2 which in turn will take the select
inputs of the multiplexers high. The
memory data, together with the R/W
signal, will now be present at the
multiplexer outputs. Shortly after the
DAV pulse, an R/W pulse is generated
which acts as a substitute for the strobe
pulse. This means that the data held in
memory will be shifted in and out of
the UART repeatedly. This information
passing out of the UART in series can
now be stored on cassette. Since the
recorder will have to be running before
a key is depressed, the character which
starts the cycle will also be recorded.
This is why it is advisable to use the
space key or a control key, as these will
have very little effect on the actual
display.

A DAV pulse is generated after each
complete byte has been shifted out
(and back in again). The leading edge of
this pulse also increments the memory

The modifications

1. Break the copper track between
pin 6 of IC19 (Nil) and pin 3 of

IC1 .

. IC6.

2. Break the copper track between
pin 16 of IC10 (CRTC) and pin 19
of IC8 (UART).

3. Break the copper track between
pin 3 of IC16 (N12) and pin 23 of
IC8 (UART).

4. Connect points A1, A2, B1, B2, Cl
and C2 of figure 1 to the correspond-
ing points in figure 2.

5. Connect pin 27 of IC10 (RP) in
figure 2 to the point marked RP in
figure 1.

6. Connect the following:

pin 9 3 of IC1
pin 6 of IC1
pin 10 of IC1
pin 13 of IC1
pin 3 of IC2
pin 6 of IC2

7. Disconnect poi

tween the keyboard and IC8 in fig-
ure 2 and re-connect as follows:

KB0 from keyboard to pin 2 of IC1
KB1 from keyboard to pin 5 of IC1
KB2 from keyboard to pin 11 of IC1
KB3 from keyboard to pin 14 of IC1
KB4 from keyboard to pin 2 of IC2
KB5 from keyboard to pin 5 of IC2
KB6 from keyboard to pin 1 1 of IC2

8. Finally, connect the outputs of the
two multiplexers in figure 1 to the
UART (IC8) in figure 2 as follows:
pin 4 of IC1 to pin 26 of IC8

pin 7 of IC1 to pin 27 of IC8
pin 9 of IC1 to pin 28 of IC8
pin 12 of IC1 to pin 29 of IC8
pin 4 of IC2 to pin 30 of IC8
pin 7 of IC2 to pin 31 of IC8
pin 9 of IC2 to pin 32 of IC8 H

figure 2:
point MO (IC6)
point Ml (IC5)
point M2 (IC4)
point M3 (IC3)
point M4 (IC2)
point M6 (IC1)
KB0 . . . KB6 be-

9-30 -

ar 1980

musical box

Readers who collect musical boxes will probably think that an
'electronic musical box' sounds as crazy as a gas telephone or a steam
radio. After all, what made the musical box so enjoyable was winding
it up and listening to its familiar tune. The circuit presented here shows
that electronics can be used to replace the wear-prone internal workings
of a musical box. In fact, an advantage over its old-fashioned
counterpart is that this circuit is able to play no less than 27 tunes.
Applications can also include toys, video games and doorbells.

mii.sk*;il Ink

As can be seen from figure 1 , the actual
melody generator is a single 1C (IC4). It
is the AY-3-1350 from General
Instrument Microelectronics, a company
with an excellent name for solid state
musical devices. The circuitry around
IC4 generates the clock signal, selects
the melody required and amplifies the
output level.

To select a particular tune, one of the
connections marked A . . . E will have
to be grounded and pin 15 of the
melodic chip must be connected to one
of the points marked 1 ... 4. There are
several ways in which the desired code

can be presented to the 1C. One method
is to use wire links, another is to incor-
porate switches and a combination of
the two is also possible. The printed
circuit board has been designed to
accomodate either of the two methods
shown in figure 2.

If the circuit is constructed exactly as
shown in figure 1 and wire links are
placed between points K . . . N and
R...V (see figure 2a), the following
procedure will take place.

When one of the pushbuttons, S A • • . Sg
is pressed, one of the points marked
A . . . E will be connected to ground via

figure 2a figure 2b melody

H ij S2 S3

— s A

— S B

— Sc

— s D

— S E
KR S A
KS S B
KT Sc
KU S D
KV Sg
LR S A
LS S B
LT Sc
LU S D
LV S E
MR S A
MS S B
MT S C
MU S D
MV S E
NR S A
NS S B
NT S C
NU S D
NV S E

S F

SG

0

0

0

0

A Toreador
B William Tell
C Halleluiah Chorus
D Star Spangled Banner
E Yankee Doodle
A John Brown's Body
B Clementine
C God Save The Queen
D Colonel Bogey
E Marseillaise

B Deutschland Lied I

C Wedding March
D Beethoven's 5th
E Augustine
A A Sole Mio
B Santa Lucia
C The End
D Blue Danube
E Brahm's Lullaby
A Hell's Bells (specially composed)

B Jingle Bells
C La Vie en Rose
D Star Wars
E Beethoven's 9th

Descending Octave Chime
Westminster Chime

R7 “ 100 k zener

R8.R17 = 2k7 D18 * 5V6/400 mW zener

R10.R12.R16- 3k3 Capacitors: T1 " TUP

R11 = 27fi Cl . . . C5 * 10 n T2 = BC517

R13.R14.R18 = 33 k C6.C8.C1 1 - 100 n T3 ” TUN

R 1 5 * 560 k C7 = 220 p IC1 - 4049

R19 ■ 47 k C9 * 220 n IC2.IC3 = 4066

R20= 100H C10.C12- 10 m/16 V IC4 = AY -3-1350

one of the diodes D1 . . . D5. Each can be seen from figure 2b, points few microamps. Transistor T1 and the

pushbutton has a total of five melodies A . . . E can be grounded by means of a zener diode D18 are included to drop

at its disposal. The choice can be cut six position wafer switch, S3. Switch S2 the voltage down to 5 V for those parts

down to one by means of a wire link, connects one of the points K . . . N to of the circuit requiring a lower voltage.

Thus, one of five predetermined point P. The melody will be initiated The nominal loudspeaker impedance is

melodies can be selected per switch and, upon depressing Sp. Resistor R6 and 8 J2, but if one with a higher impedance

in addition, two well-known chimes the electronic switch ES5 are not is to be used, the value of R20 can be

may be 'played' by depressing Sp or Sq. necessary for this latter option. They reduced accordingly. Switch SI is still

Table one shows the melodies which are are shown outside the dotted line as to be mentioned. Its function is to

available and the combination of ES5 is contained in a separate 1C to select between a 'piano' sound with

connections required to select each one. ESI . . . ES4. slow decay (position (a)) and a constant

The code numbers and letters corre- The oscillator is formed by C7, R8 and volume 'organ' sound (position (b)). It

spond to those given in the circuit and PI together with part of IC4. The pitch should keep the children amused for

in the component layout shown in of the melody being played can be hours! M

figure 3. adjusted by PI, the length of each note

The second method is to use a pair of can be adjusted by P2, leaving P3 to

multi-way switches, in which case the regulate the volume,

area inside the dotted line in figure 1 Two 4.5 V batteries are all that is

may be omitted. This will enable any required to power the circuit as the

one of 25 melodies to be selected. As quiescent current consumption is only a

52 ■ 5 position wafer switch

53 * 6 position wafer switch
LS ■ 8 Si/0.5 W loudspeaker

(see text)

electrolytology

ar 1980 - 9-33

Capacitors are a vital part of electronics,
so it is important to realise exactly how
they work. In its simplest form the
capacitor consists of two flat metal
plates which are separated by an electri-
cally insulating substance called
dielectric (see figure 1). When a voltage
is applied to the plates (figure 2), the
following happens. The electrons
(negatively charged particles) which
originate from the negative pole of the
voltage source, will repel the electrons
on plate (b) when they reach plate (a)
(since similar charges repel each other).
The electrons on plate (b) will be
attracted to the positive pole of the
voltage source. Electrons are therefore

eleeirolytology

an inside look at capacitors

There is nothing new about the
fact that electrolytic capacitors
possess inductance due to the
method used in their construction.
At high frequencies their
impedance will be largely
determined by this parasitic
induction. An electrolytic
capacitor even acts as a band pass
filter and thus also has a resonant
frequency.

What fewer people will be aware
of is that the capacitance is clearly
frequency dependent. This has to
do with moving ions in the
electrolyte, which will be
discussed later.

being moved — in other words there is
an electric current flowing. Since plate

(a) is being charged and electrons are
disappearing from plate (b), a potential
difference occurs across the plates.
Whenever this voltage is equal to that at
the source, the electron flow will stop.
The capacitor will then be fully charged.
It should be mentioned that there is no
current flowing from plate (a) to plate

(b) , for both plates are separated by
insulating material.

The current is of a temporary nature
(until the capacitor is fully charged). By
continuously reversing the polarity of
the applied voltage the current can be
maintained. That is why a capacitor will
only pass alternating current.

The amount of current depends on the
amount of charge which is displaced
inside the capacitor, which in turn
depends on the applied voltage and on
the value of the capacitor. The relation-
ship between voltage and charge is
expressed in capacitance. The greater
the capacitance, the more charge is
displaced for a given voltage and more

1

AC is allowed to pass through.

There are various methods of increasing
the value of capacitance. In the first
place by having a larger plate surface
area, secondly by using a thinner
dielectric and thirdly by using an
improved dielectric. In order to obtain
the greatest possible capacitance from
the smallest possible size, manufacturers
have examined various techniques.
Capacitors are normally made from very
thin sheets of metal foil separated by a
thin dielectric. The thinner the dielec-
tric, the greater the capacitance, but at
the same time the maximum voltage
that can be applied has to be reduced to
avoid breakdown. To increase the

capacitance even further, several layers
of metal foil and dielectric can be piled
on top of each other (see figure 3). This
is called a layered capacitor.

The dielectric may consist of paper,
plastic or a type of ceramic material.
Thus, there are paper, polyester,
polycarbonate and ceramic capacitors.
Each type of dielectric gives rise to its
own special characteristics and makes
the capacitor suitable for certain
purposes.

In addition to the layered construction,
there is the more common wound
method where the metal foil and the
dielectric are rolled up (see figure 3b). A
capacitor of this kind will have a higher
parasitic induction than the layered
type.

Up to now these have all been foliated
capacitors, being made up from thin
strips of metal and insulating material.
In spite of the extremely thin metal foil
and dielectric used, dimensions increase
at an alarming rate at high capacitance
values and working voltages. For this
reason, the maximum value of foliated

2

Figure 1 . In its simplest form a capacitor
consists of two flat metal plates which are
separated by an electrically insulating material

Figure 2. By connecting a voltage to the
plates, a charge can be produced due to the

Figure 3. To obtain large capacitances in small dimensions, several layers o* metal foil and dielectr
Another method is to roll up the metal foil and dielectric (3b).

capacitors is restricted to a few micro-
farads. For larger values electrolytic
capacitors must be used.

The electrolytic capacitor

The plates of the electrolytic capacitor
also consist of very thin metal foil. The
material will be either aluminium or
tantalum. Taking the aluminium type as
an example, it will be found that
basically the electrolytic capacitor has
the same structure as an ordinary
capacitor: two plates and an insulator.
Since the electrolytic capacitor is
polarised, it has an anode plate (positive)
and a cathode plate (negative). The
cathode contains not only the metal foil,
but also an electrolyte (electrically
conductive fluid). In figure 4 the
simplified structure of the electrolytic
capacitor is shown. The cathode only
serves to pass current to the electrolyte
by way of its large surface area.

The dielectric consitsts of aluminium
oxide, a good insulator with a high
breakdown voltage (800 million volts
per metrel). This means the dielectric
can be very thin enabling large capaci-
tances (even up to 1 farad) to be
reached in relatively small dimensions.
The layer of aluminium oxide is obtained
by anodising the aluminium foil.
Anodising is an electrochemical process,
whereby the aluminium is dipped into
an electrolytic bath (figure 5). A voltage
is applied (the activating voltage)
between the bath and the aluminium
which acts as the anode (positive). The
oxygen ions (negatively charged), in the
solution, combine with the aluminium
and the density of the layer of
aluminium oxide created depends on
the value of the activating voltage. The
density of the dielectric can therefore
be controlled very accurately. The
oxidised aluminium foil is then ready
for use as the anode of the electrolytic
capacitor.

Nowadays, all electrolytic capacitors are
wound. The foils of the anode and the
cathode are separated by a layer of
paper for two reasons. Firstly, to
prevent a short circuit between the two
aluminium foil layers and secondly, to
act as a holder for the electrolyte
(sponge effect).

4

Figure 4. In the cate of the electrolytic capacito
metal foil, but alto includet an electrolyte lelect
of aluminium oxide obtained through anoditatic

To increase the capacitance of electro-
lytic capacitors the anode plate is
etched before oxidisation to provide a
greater surface area (see figure 6). As
the cathode is made up from a fluid, it
will adapt itself to the rough surface
area of the anode. Modern production
methods for electrolytic capacitors
almost always follow this construction
method. The electrolyte need not
always be a fluid, often a form of 'paste'
is used. Hence the terms 'wet' capacitors.
As mentioned before, the electrolytic
capacitor is polarity conscious, with the
anode always being positive with respect
to the cathode. The voltage across the
capacitor must never exceed the
oxidising voltage, as this would cause
the anodisation process to continue and
the electrolytic capacitor to explode
due to the heat produced. If the electro-
lytic capacitor is connected incorrectly
(with the anode negative in relation to
the cathode) the aluminium foil of the
cathode plate will be subject to
anodisation and again the capacitor will
come to a bad end. For AC purposes
special bipolar electrolytic capacitors
(which are not polarity conscious) have
been developed .

The electrolytic capacitor and its
impedance

We have already seen that the winding
method produces an unwanted side
effect — that of induction. At higher
frequencies especially, the parasitic
induction contributes greatly to the im-

r, the cathode plate is not only made up of
rically conductive fluid). The dielectric consists

Figure 5. Anodisation is an electrochemical
process, whereby the aluminium is dipped in
an electrolytic bath and 'plated' with a layer
of oxide by means of an electric current. Here
the aluminium acts as the anode.

Blectrotytology

er 1980 - 9-3

pedance (AC resistance) of the electro-
lytic capacitor. With the exception of
bipolar types, capacitors are only suit-
able for DC anyway, so is there a prob-
lem? The answer is yes, for when an AC
signal is superimposed on a DC level this
will create considerable problems. Con-
sider for instance, what happens in the
smoothing circuit for mains supplies,
when circuit voltages are disconnected
or when two amplifier stages are AC
coupled. In these cases it is impossible
to merely rely on the given value of the
capacitor.

In addition, the electrolytic capacitor
has a resistance produced by the electro-
lyte. This resistance is highly dependent
on temperature. The frequency
dependence of the impedance is clearly
shown in the equivalent circuit of the
electrolytic capacitor (figure 7). Basi-
cally, the electrolytic consists of a
capacitor, a resistor and an inductor
connected in series. By way of illus-
tration, the impedance curve of an
electrolytic capacitor of 100 jrF/63 V
has been plotted at different tempera-
tures in figure 8. At frequencies of up to
60 ... 80 kHz (at 20°C) the impedance
is mainly determined by the R and C in
the equivalent circuit, and at higher
frequencies by the R and L. The curve
also shows that the electrolytic capacitor
has a resonant frequency, where its
impedance will be at a minimum. In
other words, it acts as a band pass filter
for high frequencies (LCR series loop).

AC and DC capacitance

Figure 7. The equivalent circuit of the As stated, the cathode of an electrolytic

electrolytic capacitor. capacitor is made up of an electrolytic

80087-8

(Siemens Data Book 1980/811. V 1 P

Measured capacit

* in pF at:

Type

47 pF, 350 V
6800 pF, 25 V
680 pF, 25 V
100 pF, 25 V
4,7 pF, 25 V

0 Hz 50 Hz
54,1 49,2

112% 103%

8760 7370

120% 100%

829 759

111% 101%

133 122

110% 101%

4,27 4,04

109% 103%

100 Hz 1000 Hz

47.9 43,2

100% 90%

7330 6670

100% 90%

749 699

100% 93%

121 110

100% 90%

3,92 3.47

100% 88%

fluid (or paste) Current conduction in
a fluid takes place rather differently
than in a solid. In solids only electrons
move about, whereas in fluids ions also
take part. Because of their small size
and mass, electrons are very mobile and
can keep up with the speed of voltage
variation. This is not the case with the
much larger and heavier ions. These are
much slower, especially at low
temperatures. If the temperature
becomes low enough for the electrolyte
to solidify, the ions will be frozen, as it
were, and will not take part in the
conduction. Only the electrons will then
be able to displace the charge (a charac-
teristic of solids). The result is a greatly
reduced capacitance.

Since ions are less mobile, they will have
difficulty in penetrating the deepest
pores of the etched anode, as they do
not have enough time. For this reason
the deepest pores will not be effective in
the operation of a capacitor under a
superimposed AC, which means a smaller
anode surface area to operate on. Thus,
the effective value of an electrolytic
capacitor under DC conditions will be
greater than with AC. In other words,
the capacitance is frequency dependent.
Electrolytics therefore have a DC and an
AC capacitance. The AC capacitance is
measured according to DIN standards
with a 50 Hz signal of <0.5 V (low
enough to prevent destruction) and at a
temperature of 20°C. The I EC standard
prescribes a measuring frequency of 100
or 120 Hz. The DC capacitance is
determined by timing a single discharge
from an electrolytic capacitor charged
to a nominal voltage.

The DC capacitance is usually 1 .1 to 1 .5
times greater than the AC value. The
greatest differences are found with
electrolytic capacitors having a low
maximum working voltage. The
dielectric in these is very thin and so the
dimples in the rough anode are rela-
tively deeper after anodisation than in
the case of capacitors with a high
maximum working voltage.

After the 'digifarad' article was publ ished
(Elektor 54, October 1979), several
readers drew our attention to the fact
that the values of electrolytic capacitors
measured with this instrument have to
be interpreted with care. This is because
the digifarad measures capacitance
according to a method which is very
similar to that used to determine the
DC value. Since the AC value is indicated
on most electrolytic capacitors, the
digifarad will more often than not
overestimate the value. This is not
necessarily incorrect, but it will have to
be taken into account when the
capacitor is used.

By way of illustration, several values of
electrolytic capacitors at different
frequencies are given in the table. M

Literature:

Siemens Data Book 1980/81,
'Aluminium and Tantalum Electrolytic
Capacitors'.

aer 1980

The design

Figure 1 gives the layout of a design
which escaped notice in last year's
Summer Circuits '79 (no. 6). This is a
cheaply constructed curve tracer for
transistors and diodes. No really pro-
fessional test instrument, of course, but
an extremely useful aid to quickly carry
out a general test either to compare
transistors or select them. Naturally,
hobbyists will have to have an oscillo-
scope (with separate x and y inputs),
because the curves will be displayed on
the oscilloscope screen.

Since it is impossible to tell which
transistor characteristic is more import-
ant than another, there is no such thing
as the 'most important curve'. Transistor
handbooks speak of the most read curve.

to the Y input and the ground connec-
tion of the oscilloscope 'hangs' resistor
R7. This is the TUT's collector resistor
and the voltage across it is naturally
proportional to the collector current of
the transistor tested. In this way, an 'lc'
will appear on the vertical axis of the
oscilloscope. The TUT's emitter is
connected to the X input so that the
collector/emitter voltage (Uce) can be
read horizontally on the screen.

What causes the curves to appear on the
screen? Two signals are fed to the TUT.
A 5 step position staircase waveform is
fed to the base and during each step a
sawtooth is fed to the collector. This
means the collector voltage changes
continually at a certain base drive
current. This occurs at quite a speed so

transistor oirve tracer

Ic/UcE characteristics directly onto the screen.

There are never enough simple
circuits which provide useful and
low-cost additions to the 'home
lab'. This particular design
possesses all the advantages to
make it a glorious example of its
kind. It offers oscilloscope owners
a neat, additional measurement
facility. It is easy to build,
contains common parts and is
inexpensive. Reason enough to
design a printed circuit board for
it.

This involves the Ic/UCE characteristics
where the collector current is plotted as
function of the collector/emitter
voltage at different drive currents.
Figure 2 gives an example of such a
characteristic. At the same time it
(roughly) indicates the drive currents
the curve tracer uses. The current
amplification may be directly derived
from the Ic/Uqe characteristics and,
after a few calculations, so may the
transistor's output impedance. The
latter is affected by the curve's slope.
Generally speaking, the more horizontal
and straight it is, the higher the collector/
emitter impedance.

Back to the schematic. The transistor
under test is indicated as TUT' as usual.

that the oscilloscope screen simul-
taneously shows 5 characteristics for
5 different base drive currents. The
staircase signal and the sawtooth wave-
form are controlled by means of an
astable multivibrator. The AMV consists
of T1 and T2 and generates a square
wave with a frequency of approximately
1 kHz.

The sawtooth is obtained very easily by
integrating the square wave via R5 and
C5. Creating the staircase voltage is a
little more complicated. During the
positive half-cycle of the square wave
produced by the AMV, C3 is charged to
a maximum which is equal to the supply
voltage. During the negative half-cycle,
C3 will turn on transistor T3 and thus

Figure 1. The circuit diagram of the curve tracer.

B. Darnton

2

R7 = 330 n
R8 = 270 k

Capacitors:

Semiconductors:

C1,C2.C4 = 100 n
C3= 22 n

C6 = IOOm/IOV

T1 . . . T4.T6 = TUN
T5 = TUP
D1 = DUG

Figure 3. This is how the curves ep

to the TUT's base via R8) will become
a little lower. By loading C4 intermit-
tently, each successive negative half-
cycle will reduce the emitter voltage of
T4 in steps until T4 starts to conduct
turning on T5. C4 is soon discharged
and a new cycle starts.

The number of stages which make up a
single cycle is determined by the ratio
of C3 to C4 and is 5 here. By adjusting
the value of C4 the number of stages
(and thus the number of curves indicated
on the screen) can be changed as
required.

In practice

The photo in figure 3 shows how the
curves appear on the oscilloscope screen.
The circuit's only flaw now comes to
light - the characteristics are traced
from right to left, instead of the other
way around. Unfortunately, nothing can
be done about this. In practice it does
not present a problem. What is serious
however, is that the tracer is only
suitable for NPN transistors. NPN types
cannot be tested with it. If this is
considered to be a drawback, however,
there is a cheap solution: two printed
circuit boards may be built instead of
one. The circuit requires few com-
ponents, so why not? The second circuit
will then be a PNP version. ForTI . . .T4
and T6 use TUPs, T5 will be a TUN, C6,
D1 and the supply leads will be
switched around. Furthermore, such a
PNP version will trace the curves from
left to right, only now the Y axis will be
negative so that they will appear upside
down on the screen. A little strange
perhaps, but you'll soon get used to

As mentioned above, diodes may also be
tested. These are connected with the
anode to R7 (1) and the cathode to the
supply zero (X). The l/U characteristics
of the diode in question will now appear
on the screen. Figure 4 shows the
printed circuit board. It is highly
compact and can be built in less than no

Last word. Since the circuit only requires
a few mA, the supply will not have to
be very 'heavily' tested. However, the
supply voltage must be well regulated
for it to work properly. M

9-38 -

1980

)r Vocoder

using die
EMotor Vocoder

Several months ago (Elektor
no. 56, December 1979), Elektor 1
published a 10 channel vocoder.

When building a vocoder there
are a few 'obstacles' which ought
to be taken into account. Readers
who have already built one and
are familiar with it will find that
this article provides useful
information on how to improve
on the vocoder's technical
qualities. To start with it is a good
idea to check the initial
adjustment.

F. Visser

Each channel in the vocoder contains
three presets. Two of these are intended
to eliminate leakage of the Voice and
Carrier signals to the vocoder's output;
the third sets the dynamic range of the
voltage control circuit (in the analyser
section, where the audio signals are split
up into small bands and are converted
into DC control voltages). This is im-
portant if the vocoder is to respond to a
wide range of input signal levels and
reproduce the speech sounds as accu-
rately as possible. In passing, it should
be noted that this high 'responsiveness'
may cause a disturbing side effect when
the vocoder is used during live perform-
ances, where there is usually a high level
of interference. In such cases the
vocoder will analyse and synthesize the
entire complex sound, producing an
undesirable cacophony. Further on in
this article, methods will be suggested
to suppress these side-effects. For the
moment, however, let us concentrate
upon setting up the vocoder properly.
The best way to start is to adjust
potentiometers PI, P5 and P9 in the
band pass, high pass and low pass filters
respectively. These presets compensate
the output offsets of the filters that
follow the rectifiers in the analyser
section. To a large extent, this deter-

mines the vocoder's dynamic range.

The offset should not be more than
5 mV. If this cannot be achieved it may
be advisable to modify the offset
compensation slightly, as shown in
figure 1. In the original design HA 4741
type opamps were used, as these have
a smaller offset than the TL series.

Unfortunately, they are also more
difficult to obtain and more expensive.
If all the U 0 ut buses are now connected
to the Uj n buses, there is no danger of
undesirable offset voltages turning on
the OTAs in the synthesizer section (or
cutting them off — if the offset is
negative).

Elektor Vocoder

iber 1980 - 9-39

The vocoder's dynamic behaviour is
further determined by the following
adjustment: the cutoff point of the
OTAs. This can best be done with the
aid of an oscillator and an oscilloscope
or an AC millivoltmeter. The (sine
wave) oscillator is connected to the
carrier input and is tuned to each
successive filter frequency in the syn-
thesizer section. The signal voltage is
set to about lOVp-p, measured at
pin 7 of A4, A14 and A24. The Uj„
potentiometer on the front panel is
turned up fully and now the oscillo-
scope or millivoltmeter is used to check
the output of A10, A20 or A30. The
preset potentiometers P4, P8 and PI 2
are adjusted to the point where the
output signal just stops decreasing (see
figure 2).

Finally, the leakage from control input
to audio output of the OTAs must be
reduced to a minimum. Usually, it will
not be possible to eliminate this entirely
- but it is worth while trying (even
replacing the OTAs, if necessary), since
break-through of the speech signal to
the vocoder output seriously affects the
overall performance. Figure 3 shows the
measurement set-up; P2, P6 and P10 are
adjusted for minimum break-through.
Best results will be obtained when the
leakage of the single phase rectified sine
wave signal, applied to the speech
inputs, is not greater than 5 mV p-p at
the vocoder output. In practice, this will
not be easy to achieve. It has been
found that only 200 out of every 1,000
OTAs manage it!

If an oscilloscope and an oscillator are
available, it is a good idea to check the
pass-band and gain of all the filters.
Obviously, any deviation with respect to
those particular aspects can lead to an
undesirable colouring. If, however, good
components are used (and mounted in
the correct positions!), any error should
be so small as to be negligible.

How to use the vocoder

Having set up the vocoder properly, the
next question is what to do with it. Its
most common application is as a 'voice
processor'. A recent 'hit' in the charts
is 'Funky Town' by Lipps Inc, in which
the voices of two members of the group
are transferred to the sound of a syn-
thesizer. The introductory lyrics are
difficult to understand (even for
Americans!). One reason for this could
be that the key chosen for the melody is
rather high and, as our previous article
on the vocoder stated, it is important
that the frequency spectrum of the
carrier signals overlap that of the speech
input. If the carrier consists almost
exclusively of high frequency com-
ponents and the modulation signal (in
this case the voice) is in a lower fre-
quency range, only the higher har-
monics of the voice will be super-
imposed on the carrier signal, as shown
in figure 4. Furthermore, a woman's

9-40 - elektor September 1980

Elektor Vocode

voice appears to be used as the modu-
lation signal on this recording, with a
formant range that is less suitable for
the classical vocoder with a relatively
small number of channels. Later on in
'Funky Town' the melody is played in
a lower key and a male voice sings the
lyrics. The improved intelligibility is
very noticeable!

The Elektor vocoder has the advantage
that it can offer a reasonable solution
to the problem of non-overlapping fre-
quency spectra. By connecting the
voltage control outputs of the analyser
to channels one or two places higher up
in the spectrum instead of to the con-
trol input of the corresponding syn-
thesizer channel, the significant spectral
information is moved up, as it were, to
a range that encompasses the higher
carrier frequencies. This technique,
known as 'formant shift', will be dealt
with in depth later on in this article.

In addition to the vocoder's use as a
voice processor there are many ways in
which sounds can be superimposed on
different kinds of carrier signals. The
best way to get to know the vocoder is
to systematically carry out experiments,
using a microphone and a simple saw-
tooth or pulse generator.

The microphone

As far as the microphone is concerned,
a high quality type is best: if the modu-
lation spectrum is free from coloration,
the end product will also be good. Not
everyone will be able to afford a high-
priced microphone, of course, so a few
suggestions on how to obtain good
results with a reasonable quality micro-
phone may prove useful.

In the first place, it may prove useful
to give the microphone pre-emphasis
- in other words, emphasize certain
frequencies, where necessary, or atten-
uate them. This is done by means of
tone controls or with separate filters.
One of the most important corrections
to be made is to attenuate the low fre-
quency range. It is difficult to give
precise figures for this, as it of course
depends on the type of microphone
used and also on the distance between
the mouth and the microphone. The
closer the microphone, the more low
frequency components will reach the
analyser, not to mention the sound of
breathing and explosive consonants
(p, k, etc.).

Sometimes, depending on the high fre-
quency spectrum of the carrier signals,
it may be advisable to boost or atten-
uate the treble range. As a rule, a
standard Baxandall tone control with a
turnover frequency around 1 kHz is

The carrier

Many sound sources may be used as
carrier material, but a simple function

elektor September 1980 -

OClBQBtlDaDB

p [i n r

•’U’LIU'U'U'U’U-U'LT

generator with a control range between
about 20 Hz and 1 kHz would be ideal
for the first experiments. The most
suitable wave forms to experiment with
are triangle, square wave, sawtooth and
pulseforms. Should such a generator not
be available, you can always build one
based on one of the many Elektor cir-
cuit designs.

Monitoring the results

The best way to judge the results is to
use headphones. The system can also be
used to drive a conventional audio
system with loudspeakers, but head-
phones are preferable as they avoid
acoustic feedback problems.

A few simple examples

When the microphone, generator and
headphones are connected (figure 5)
and everything is switched on, the first
experiments may be carried out. If you
don't want to fall back on sentences like
'Testing . . . one . . . two . . . three . . .'
it is perhaps useful to have a text in
front of you. Experience has taught us
that not everyone possesses the 'gift of
the gab' at such moments!

The frequency of the generator is set at
about 50-60 Hz, using a pulse wave-
form. The result will be a resonant,
clear, synthesized voice. If the fre-
quency remains unchanged, the result
sounds like the 'Cylon effect'. Cylons
are robot-like creatures from the
American TV series and film: 'Battlestar
Galactica'. A vocoder was in fact used

to produce their robot voices.

By raising the carrier frequency while
continuing to speak, the synthesized
voice can be made to change in pitch. It
will become less intelligible once the
frequency is above 500-600 Hz; this
effect was mentioned earlier, when dis-
cussing the Funky Town recording.

It should be clear that the pitch of the
synthesized vocoder product depends
exclusively on the carrier’s pitch. The
next test to be described will demon-
strate this.

The frequency is set to a low value, tor
instance 100 Hz, and now the pitch of
the voice is changed by singing instead
of speaking, or by producing other
sound varying in pitch. You will notice
that the resulting timbre will change,
as if a band-pass filter were being used,
but that the fundamental frequency will
remain the same. This is because the
generator is still set at a fixed fre-
quency. Nevertheless, this is a source of
regular misunderstandings. Witness the
fact that the vocoder is often compared
to a harmonizer or to a pitch shifter
— equipment used to shift the funda-
mental frequency and the spectrum of
speech or music.

If the same good intelligibility is re-
quired at higher frequencies, 'formant
shift' can be used. The Elektor vocoder
is one of the few vocoders on the
professional market that offers this
interesting facility. Formant shift liter-
ally means shifting the intelligibility
information to a higher or lower fre-
quency range. By coupling the output
voltages of the analyser to the control
inputs of synthesizer filters which do

not have the same F 0 , the measured
formants are transposed to another
place in the spectrum. If, for example,
the voice at the speech input is much
lower than the fundamental frequency
of the carrier signal, the result can be
made more intelligible by shifting the
formants to a higher carrier spectrum.
The synthesized 'voice' will become
clearer and at the same time assume
an entirely different character. This
phenomenon can be used with great
success to produce 'funny' voices.

The higher the analyser spectrum Is
moved up, the more the voice will
sound like Donald Duck. If the analyser
spectrum is transposed down, the
speaker will sound as if he suffers from
the proverbial hot potato. Quite a
different way to manipulate the form-
ants is 'formant inversion'. To obtain
this effect the analyser and synthesizer
channels are cross-coupled. Not sur-
prisingly, the result will be practically
unintelligible. All transient sounds, such
as K, P, T and hissing sounds will be
superimposed on the low end of the
carrier spectrum, whereas the low fre-
quency information in the speech signal
will control the high end of the carrier
spectrum. Furthermore, of course, the
formants will be thoroughly mixed.
A good example of this is the '0' sound
which comes out as a 'U'. In spite of
the fact that the result is virtually
unintelligible, this effect can be useful
when making (complex) musical sounds.
This is illustrated in figure 6.

The results obtained so far through
speech synthesis will all sound robot-
like. In the first place, this is due to

the pulse signal used as a carrier: it con-
tains a lot of higher harmonics, creating
a slightly grating, 'mechanical' sound.
If a sawtooth is used instead of a pulse
shaped signal as a carrier, the result will
be softer. This illustrates that the
carrier's complexity affects the timbre.
To attenuate the robot sound further
there are all sorts of other tricks.

By modulating the carrier signal, for
instance with a low frequency sinewave
or triangular signal, a much more life-
like 'human' sound is produced. Other
modulation effects may involve a low
frequency random signal or, even better,
a control signal that is derived from the
fundamental frequency of the original
speech. This can be simulated by tuning
the generator to the voice pitch and
then adjusting it by hand to follow the
inflections. When an accurate frequency/
voltage converter ('pitch extractor') is
used a very natural sounding voice can
be synthesised, which shows that the
intonation of the voice is a very essen-
tial part of human speech. A few
suggestions to obtain carrier modulation
are given in figure 7.

Unvoiced consonants
Up to now, the unvoiced consonants
(S, SH, SK, SY, K, T, P, F, etc.) have been
neglected. These cannot be successfully
reproduced by only using a sawtooth
or pulse as a carrier. To synthesize
unvoiced consonants, a detection system
is required with the aid of which noise
can be added to the carrier signal at
the right moment. Since the Elektor
vocoder does not (yet) possess that
Voiced/Unvoiced detector, another
trick will have to be used for the
moment.

A very clever expedient was developed
by Harald Bode, vocoder manufacturer,
and he has now taken out a patent for
it. Bode constructed a sort of 'bypass'
circuit for high frequencies derived from
the analyser section. In the case of the
Elektor vocoder this has been provided
by means of potentiometer PI 7 on the
highpass filter. This contains the high
frequency range of the speech spectrum
where most unvoiced sounds are pro-
duced. By adding this signal directly
to the output, a reasonably complete
'speech' signal may be obtained.
Nevertheless, it is worthwhile to listen
to the unvoiced sounds as they are
reproduced when pulse or sawtooth
waves form the carrier signal. By pro-
ducing hissing and 'plop' noises in the
microphone while switching the gener-
ator from triangle to squarewave to
sawtooth to pulse waveshapes, you can
hear how important it is to have a wide
carrier spectrum for unvoiced sounds.
Using a triangular wave, which only has
even' harmonics, the result will be very
poor, whereas the pulse which contains
all the harmonics will produce some-
thing remotely like an S or an F.
Whistling into the microphone with a

fixed pulse frequency as a carrier will Depending on the effect that they wish
also show how much high frequency to achieve, it may be advisable to con-
energy it possesses. nect an effects box between their

instrument and the vocoder carrier
input, with which additional high fre- .

The vocoder for musicians quency components may be added to

The experiments just carried out may the original sound. Examples of such
seem a little too simple, but they devices are phasers, flangers, boosters,
emphasize the basic operation of the distorters, fuzzers, frequency doublers,
vocoder. Once the user really feels he etc.

understands exactly what is happening. It may also be interesting to connect
the variety of applications will only be the guitar to the speech input of the
limited by his imagination. When vocoder, while using an organ, string
used for musical applications, the quartet or synthesizer as the carrier
vocoder will be restricted to keyboard signal. This of course requires strict
and string instruments. After all, a saxo- coordination between the various
phone player can hardly be expected to players. Chords or a melody will be
blow and talk and sing all at the same played on the keyboard instrument,
time! whereas the guitar is used to play a

Guitar and bass guitar players will melody or a rhythmic pattern — prefer-
discover that more often than not the ably monophonic, so no chords. The
dynamic range of their instrument will newly generated sounds will have the
not be sufficiently wide to produce envelope shape and some of the spectral
intelligible or clearly articulated sounds, characteristics of the guitar. Many other

Elektor Vocoder

aptember 1980 — 9-43

The vocoder at live performances

When performing with the vocoder on
stage during a concert, a few aspects
need to be treated with care.

There are basically two characteristics
in the vocoder, which could turn the
performance into an absolute catas-
trophe.

In the first place its sensitivity or
'responsiveness' which was mentioned
earlier. Like so many devices, the Great
Compromise will have to be sought.
Providing the vocoder with a wide
dynamic range may create chaos in
noisy surroundings. This is because the
vocoder makes no distinction between
what it hears and what it is supposed to
hear. ('Not in front of the vocoder!')
Everything that enters the analyzer is
processed in the usual fashion and
appears synthesized at the output of
the equipment and those of you who
have experienced the result know what
a terrible din that can be!

The only suitable methods to suppress
such sensitivity to undesirable noises is
to use a highly directional microphone
which is spoken into from as short a
distance as possible or to use two micro-
phones in antiphase. The latter method
is illustrated in figure 8.

When two (identical) microphones are
used in this way it is important to speak
or sing in front of one of them at as
short a distance as possible. A plop cap
and a bass roll-off filter are indispens-
able. Another advantage of this method
is that acoustic feedback may be notice-
ably reduced. Feedback sensitivity
happens to be another drawback of the
vocoder, as a result of the phase shifts in
ranges where the syntheser filters over-

The vocoder in the studio

The above-mentioned precaution to
curb nasty side effects are of course less
important in recording studios and may

musical instruments may of course be q
equally well combined.

For electronic pianos the same applies
as for the guitar. Here too, the use of
some kind of effects device is rec-
ommended.

Organists and synthesizer players have a
much easier time. A nice effect which
can be produced on most keyboard
instruments is the bass effect, by
making explosive noises with the mouth
! in the microphone and letting them
) decay. Wind instruments like the tuba,
trombone, etc. can be imitated with a
t little practice. Electronic synthesizers,

> like the Elektor Formant, offer an
* extremely wide range of possibilities.

. Apart from generating carrier sounds,

! the synthesizer can also be used to
produce signals to control the vocoder
■ synthesizer inputs directly, and the

- analyser outputs of the vocoder can be

I used to control numerous units in the

' modular synthesizer.

even be totally unnecessary. The that the material may not be spectrally technicians seem to understand that

vocoder is an instrument which is highly wholly suitable and that the synchronis- the vocoder needs to be played, like any

suitable for use in the studio, provided ation between the Voice and Carrier other instrument, and that learning to

that a few details are taken into account signals may not be sufficient. play may take some time.

- particularly when dealing with The problem in the sound studio is Finally, figure 9 provides a few examples

existing recordings. The vocoder is not a often that 'time is money' and so a in which the vocoder can play an

miracle machine with a 'talent button' producer will sometimes get a little interesting part, especially if more
or a 'success filter', but an instrument impatient if the vocoder does not voltage control equipment is available,

which one must learn to use, preferably obtain astounding results at first bat. Figure 10 gives a few suggestions for

in the initial stages of a musical pro- Vocoding is then postponed until the peripheral devices to make the vocoder

duction, where required. If 'vocoding' final mix-down stage, where it is often more versatile. The voiced/unvoiced

is postponed until all the material is much more difficult to obtain the detector, in particular, is scheduled for

recorded on the various tracks of a desired effect. publication in the near future,

multi track recorder, there is a chance Fortunately, more

and more sound

7 segment display i
Limited. The displays

Limited. The displays can be clearly read up
to 35 metres away. Units in the 'S’ Series
operate using simple low-voltage signals to
create local magnetic fields behind the
segments. Each segment is simply a rotating

range of guides, edge connectors, and other acce
mpectron Wes , Hyde Developments L td..
y read up Unit g p ark St , nd Esl
■S' Series Aylesbury. Bucks HP20 1ST
signals to Telephone: Aylesbure 102961 2044 1

A la, 9 e 11 digit planar display prot
' 1b/q mi readability while the basic 1 Hz resol
be reduced to increase measurem
This is a microprocessor based ir
with the MRU controlling the met
and sequence as well as the display ar
T ape-stereo-rad io-clock d isplay functions. Remote control of mode. I
airchild's Optoelectronics and reset are provided through an
ie F LB 4010 4-digit LCD. IEEE 488 interface from which re

This 42-pin tape-stereo-clock display features also be output. The
four digits 10 mm high plus decimal points, can also be used for
Other symbols available include a stereo mode number from 0 to 9!
indicator, a tape mode indicator, tape direction l O monitoring,
indicators, and AM-FM radio mode indicators. Elex Systems Ltd.,
Transf lector and reflector versions are Crossmay House,
available. With transflective LCD operation Bracknell,
the display has the ability to allow light to Berkshire.

pass through the cell from the rear as well as Tel. 103441 52929.

Because moving parts are kept to a minimum,
and there are no filament lamps, operational
life is extremely long. An additional feature is
the use of semi-hard magnetic cores for the
electromagnetic drives, which allows high
velocity character changes driven from rapid

Two sizes of unit are available initially, with
character sizes of 40 x 32 mm (Type SO) and
80 x 40 mm (Type SI). Each unit includes a
built-in driver circuit which controls the
segments using 16 V dc signals. Maximum
power consumption occurs during character
change and units are rated at 0.07 Watts.
Impectron Limited.

Foundry Lane,

Horsham,

W. Sussex. RH 13 5PX.

Tel: 0403-501 1 1

A case with a difference

West Hyde Developments have recently
enlarged their Mod-1 range. This now amounts
to some 62 different enclosures and 37 chassis

?B.B;B.Bs

reflect light incident on the front. The glass-

biphenyl liquid crystal material very stable. A
typical connection system would consist of a
PC board, a pair of conductive elastomer
connectors on which the display is placed. A
bezel would then hold the assembly together.
Operating voltage is typically 3 Vrms in the
form of a square wave at 32 Hz. Contrast
ratio is typically 20 : 1 .

Fairchild Camera & Instrument (UK) Limited.
230 High Street.

Potters Bar,

Herts EN6 5BU,

Telephone 107071 51111

Photodiode for optical sensing
operation

The Symot 320/1 range of photovoltaic
photodiodes now includes the SP - 10N type

active area 91.16 mm 1 . The photo-voltage at
25° C is 0.43 V. and the photo-current is
590 mA (test conditions 1000 Lux). Dark
current is typically 1.5 mA (reverse voltage
2 V) and the capacitance is 7,000 pf. Peak
sensitivity is at 830 nM wavelength.

The package is equipped with flying leads,
covered in vinyl tubing.

This photodiode is suitable fo
of optical sensing applicatio
shape is advantageous, includin'
light reflection, smoke dete>

The RPC series offers almost 200 user options,

moulded grey or black plastic, and is available

or silver plated brass or phosphor bronze. It
offers the choice of panel or p.c.b, mounting,
and is supplied with a choice of 'solder bucket'

from 200 n to 20 MSI. Range and function
selection si by two rotary switches on the
clearly coded front panel, which make the
instrument very easy to use.

The multimeter is powered by a 9 V carbon-
zinc or alkaline battery (PP3 or equivalent),
the latter giving a typical life of 200 hours.
Battery-low indication is provided by the
multimeter's display, which shows 'BAT'
when less than 10% of useful battery life
remains. Automatic decimal-point, polarity
and overrange indication is also provided.

The case is of high-impact ABS plastic, and
the display is shock-mounted behind a tough
polycarbonate plastic window. The battery
and the protection fuse are easily accessible,
and a single calibration control is provided.
Estimated mean time between failures is in
excess of 20 000 hours.

The Gould Alpha V measures only 1 78 mm
(7 in) x 76 mm (3.07 in) x 38 mm (1.5 in)
and weighs 282 g. Accessories supplied with

weighs only 9 kg, measures 284 x 138 x 400
mm, and costs just £ 1,440 with a two year

House of Instruments,

34/36 High Street,

Seffron Walden,

Essex, CBI0 1EP.

Tel: (0799) 22612.

(1632 M)

Miniature Dustproof Presets

The HO 621 A ia a low cost miniature cermet

adjustment cap for applications throughout
all types of electronic circuitry. The 6 mm
diameter makes the trimmer small enough to

Swindon,

Wiltshire SN2 6BN,

Tel: Swindon (0793) 693681-7.

(1622 M)

Low-cost hand-held digital
multimeter

The new Alpha V, the latest and smallest in
timeters from Gould Instruments 9 Division,

versatility and ruggedness. The Alpha V has a
3’/i-digit liquid-crystal display, and the
25 measuring ranges cover the five basic
functions of DC voltage. AC voltage, DC cur-
rent, AC current, and resistance. Costing only
£85.00 (plus V.A.T.). the Alpha V is being

accessories available are a soft protective
carrying case, high-voltage probe, r.f. detec-
tor, and a special set of test leads rated at
2 kV RMS and 20 A.

Gould Instruments Division,

Roebuck Road,

Essex.

Telephone: 01 500 1000

(1637 M)

Low cost digital storage
oscilloscope

Designated the Model MS-1650, this versatile
instrument combines a 10 MHz real time
oscilloscope with a digital storage system
employing a 1024 x 8 bit memory.

Digital storage offers several advantages over

for less demanding situations.

in ?n

200 North Service Road,
Brentwood - Essex. CM 14 4SG.
tel. (0277) 230909

116291

Tool range Ltd..

Upton Road.

Reading,

RG325A

Telephone: 1 0734 1 29446

11557 Ml

Tuneable inductors

Toko’s range of tuneable inductors now
includes the 12VX series of high inductance
tuneable coils - available with primary/tap
and secondary up to 68 mH nominal,
although tuning 30% of this centre value
by slug core adjustment.

1cm grid

Switch operation is by a 12-volt relay and the
unit has a loss of less than 1 dB over the
frequency range 0-500 MHz.

Telecommunications Accessories Limited.
Thame Industrial Estate, Bandet Way,

Thame. Oxon 0X9 3SS.

Great Britain.

Tel: Thame 3621/2/3.

(1567 M)

Math processor chips boost micro-
computer performance

Intel has introduced two new math processor
chips, the 8231 (fixed point) and 8232
(floating point), which increase the perform-
ance of a microcomputer system by a factor
of up to 100 times ktfien carrying out
mathematical operations. Both chips act
as dedicated peripherals interfacing directly
to Intel's 8080. 8085 and 8088 microcom-

puters in addition to all other general-purpose
processors with 8-bit data buses.

The 8232 will perform 64-bit double-pre-
point addition, subtraction, multiplication
and division. Double precision operation is

ments such as chromatographs and specto-

carried out over a wide dynamic range with

racy. The processing time depends on the
data, however, a typical single-precision
multiply takes approximately 100 micro-
seconds. The 8232's floating point algorithm
is a subset of the proposed IEEE floating
point standard which is the same as used

software and the ISBC 310 arithmetic pro V -
cessor board. In fact, this standard is used
in all Intel hardware and software to ensure
that programs written in different languages
and run on different systems will always
yield the same result.

The 8231 is intended for use in process and
industrial control applications requiring a real
time mathematics capability over smaller

mode, performing 16 and 32-bit addition,
subtraction multiplication and division. Other
functions which can be performed by the
8231 include the calculation of sine, cosine,
tangent, cosecant, secant, cotangent, square
root, logarithm, natural logarithm, exponen-
tials and powers.

Both the 8232 and 8231 math processor chips

grammed algorithm controller, an 8 by
16-bit operand stack, a 10-level working
register stack, command and control registers
and a read only memory containing all the
control software. All transfers between the
host processor (including operand, results,
status and command information) take place
over an 8-bit-bidirectional data bus.

Both chips are manufactured using Intel’s
HMOS technology and are available in 24-pin
dual-in-line packages. Each requires +12V
and +5 V power supplies.

INTEL Corporation IUK I Ltd.,

4 Between Towns Road, Cowley,

Oxford OX4 3NB.

Tel: Oxford 108651 771431

(1562 M)

Liqhtweight polypropylene tool
box

The new 1 51 Cantilever tool box from Racco
is made from rugged polypropylene and has
overall dimensions of 500 x 220 x 220 mm.
Four tool trays each 480 x 75 x 35 mm fold

are provided with movable dividers. A fold
flat lid and two safety catches complete an
extremely lightweight package. Finished in

The Racco 151 Cantilever Tool Box costs
£ 1 2.50 + VAT
Toolrange L TO.,

Upton Road,

Reading RG3 4JA,

Tel: Reading 10734) 29446 or 22245

(1573 M)

Illuminated push button switches
offer RFI shielding

The capital RF 335 series of illuminated push
button switches is now available from Symot

The normal rating is 2 amps, 250 VAC.

Steel housings, finished with a black oxide
bezel are standard and stainless steel clips

The panel cutout dimensions are 0.928 inches
x 0.972 inches.

Momentary and alternate action switches with

able. Lens modules can be full-screen or hori-

zontally or vertically split and a simple adap-

available in all standard colours. The lens mo-

dule design ensures that the TH4: flange base
lamps do not move during operation, thus re-
ducing lamp failure.

Lamp replacement is easily achievable from

Symot Limited,

22a Reading Road.

Henley -on- Thames,

Ox on. RGP 1AG.

Tel: 1049 12) 2663.

Flatter flat pack relay

Miniature p.c.b. type LZN relays from IMO
Omron are now flatter. These highly reliable
ultra low profile (11.5 mm) flat pack type
miniature relays, available in two or four pole,
employ the international grid terminal ar-
rangement. The LZN has triple gold flashed
silver contacts rated 3A @ 24 VAC. these twin
bi-furcated contacts plus the unique card life
off system for the contact drive ensures re-
liable switching of low voltages. Operating
voltages are 6-48 V DC. These competitively
priced flat packs have extremely long life, in
excess of 100 million operations minimum,
low coil resistance plus reliable low voltage
switching makes the LZN ideal for the alarm

349 Edgware Road,

London W2 IBS,

Tel: 01 723-2231/4 and 01 402-7333/6.

1575 M

DIL switches sealed

Dual in line switches, type 338. are now
offered by Symot Limited. The principal

advantage claimed for these switches is the

terminal sealing technique which is designed

to prevent contamination from flux and

These DIL switches embody a single point
contact system, with self-cleaning wiping
action which results in a low and stable
contact resistance. The initial figure of less
than 50 milliohms at 2 V DC, 10 mA, is
guaranteed for over 10,000 switching oper-

Temperature coefficient is fully docu-
mented over the range — 25°C to +60°C.

Brendwood ■ Essex. CM 14 4SG
tel. (0277) 230909

(1560 M)

UHF cavuty filters

Toko’s 232MT series of helical resonators

for communications quality UHF filters.
Originally developed for applications ranging
from UHF RF filters to the first IF of SHF
(satellite broadcasting) systems, the series
are available in the range 380 - 480 MHz
with either two or four chambers.

The insertion loss is only 1.8 dB max for four
pole units, with a shape factor (6/60 dB) of

width of : 25 MHz at 60 dB. an ultimate
stopband of some - 70 dB.

plastics have been chosen to increase the
reliability of these switches, which are rated
at 50 V DC, 100 mA, non switching and
5 V DC. 100 mA in the switching mode.

The terminal pitch is 2.54 mm x 7.62 mm.
Symot Limited,

22a Reading Road,

Henley-on-Thames,

Oxon. RG9 I AG
Tel: (049 12) 2663

(1572 M)

ar 1980 -UK 19

Build the Practical Electronics 1
handheld DMM. This superb product
offers professional precision with
extended battery life. Five function
operation (AC and DC VOLTS, AC
and DC CURRENT, RESISTANCE)
*n check diodes. 0.5” LCD

with ability t

display with Battery Low' warning.
Auto-polarity, Auto-zero. Full protection
against transients and overloads with
ability to withstand mains on any range.
0.5% basic DC accuracy and 15
different ranges. It measures AC/DC
voltages from 0. 1 mV to 500V. AC/DC

\ TOTAa

LCD DISPLAY^

FULLY ASSEMBLED

[ DMM (INC. LEADS)!

^^^^^To'Lascar Electronics. Unit 1, Thomasin Road, Basildon, Essex.

Please send me Dale □ FULLY ASSEMBLED DMM (INC. LEADS) £47 09 □

PE— DMM KIT £39 04G ICL 7106 £10.87 Q LCD DISPLAY £9.72 Q PCBE6 27Q

FEATURED

ASA

PROJECT IN

PRACTICAL

ELECTRONICS

sc/mputer w

sc/mputer w

This first book of a series describes how to build
and operate a simple microprocessor system
based on the National Semiconductor Micro-
processor (INS 8060). The system may be
extended to meet various requirements - these
will all be discussed in the SC/MPUTER books.

PRICE/UK £3.70 inc P&P

OVERSEAS SURFACE MAIL . . £3.90 inc P&P