Fy, 0I6MIQF

©lictrertes

TELECOM 87 - A REPORT

Switch, mode power supply
Wide band aerial booster & splitter
The rise & rise of the micro
Portrait of Sir Clive Sinclair

1

Front cover

A computer controlled test
system from SIRA measures the
optical transfer function and
the modulation transfer
function of lens systems to give
an Indication of the quality
and the ability to resolve fine
details in the image with
adequate contrast.

Photograph CROWN
COPYRIGHT RESERVED

STRATEGIC ELECTRONICS
INITIATIVE (SEI)

- THE PAPER WAR

The Department of Electronics has recently brought out two discussion
papers devoted to the national policy on electronics and excerpts from
the papers appear elsewhere in this issue.

The proposals are high sounding, often abstract. On occasions, it states
the obvious but there are some valid observations which need to be
taken seriously. The proposals have already received some flak and they
are dubbed “Old wine in new bottle”.

We cannot be expected to adopt the proposals wholesale. As has been
hinted, the papers are meant for public discussion. Hopefully, such
discussion will bring out a national consensus, and prune the paper into
a workable scheme.

The basic question whether liberalised policy helps In the growth of the
industry crops up again. The most widely held and perhaps, a valid view,
has been that the growth of India's electronics Industry during the last few
years is mainly due to the liberal import policies. Yet, the growth has not
kept pace with the targets. The new proposals seem to imply invisible
bureaucratic controls and here begins the controversy.

A welcome suggestion has been the criterion for foreign collaboration.
The gross exports do not reflect the reality. The net foreign exchange
earned is, indeed, the real indicator of the export capability of a unit.
Inter-ministerial wrangles proved to be the bane of our policies and they
have often resulted in wasteful and avoidable foreign exchange outlay,
not to speak of the damage done to indigenous technology
development. For mending fences, the bureaucrats need no policy
paper and they have to simply realise their collective responsibility.

The discussion papers rightly point out many of the ills afflicting the
electronics scene in the country and they have not been diagnosed for
the first time. Solutions to these ailments are to be found. That remains to
be done by all those concerned with electronics.

■ ■988 2.07

ELECTRONICS NEWS • ELECTRONICS N

Ites Jubilee

The Indian Technical Education Soci-
ety, an autonomous body, with over 300
affiliated technical institutes, celebrated
its silver jubilee, towards the close of
1987, Apart from conducting examina-
tions in various trades, for which about
10,000 students appear every year, the
ITES has set up the Institute of Techni-
cians for aspiring and trained techni-
cians.

Delivering the key note address at a
function to mark the silver jubilee, the
chairman of Electronics Commission,
Mr P.S. Deodhar, pointed out that
largely technical institutes lacked in
adequate staff which resulted in their
poor management. Students in private
institutes needed better technical educa-
tion, he added. Referring to the speech
of Mr S.Y. Side, president, ITES, Mr
Deodhar offered to supply kits to the
ITES. So far about 350,000 kits have
been sold to 100 small scale industries.
The department of electronics has de-
cided to begin new courses on manufac-
turing engineering, mechanical man-
ufacturing process and circuits and de-
signs on chips, according to Mr
Deodhar.

Dr P.K. Patwardhan of Bhabha Atomic
Research Centre, called for ruralisation
of technical education and stressed the
need for more privatisation of institutes,
resulting in less dependence on the gov-
ernment. More polytechnics and en-
gineering colleges without proper equip-
ment and infrastructure cannot deliver
the goods as IITs and engineering col-
leges produced only trainable boys but
not trained ones. Practical training in the
industry and in the field filled this gap.
The ITES, by its reputed and profes-
sional contributions, turned out as a
strong technical Alma Mater, according
to Mrs Chitnis, vice-president of ITES,
with its office at 173, Vidyanagari Marg,
Kalina, Bombay-400 098.

Meltron in Brief

Maharashtra Electronics Corporation
Ltd. recorded a turn over of Rs. 17.70
crores in 1986-87, its ninth year of opera-
tion. During the year, the first EPABX
system based on the C-DOT technology
rolled out of the Meltron’s ultramodern
telematics division at Aurangabad. An
outdoor broadcasting van was developed
indigenously at the company's au-
diovisual division at Bombay for the All-
India Radio. The company's communi-
cation factory at Nagpur supplied to AIR
a studio transmitter link.

Meltron’s audiovisual division will
shortly introduce multi voice logging re-
corders for use in the defence services
and professional mixing consoles and
audio racks for studio applications. The
radio communication factory at Nagpur
proposes to add synthesised portable
transreceivers and rugged versions of its
existing products for the army and para-
military forces. The new factory at Au-
rangabad will take up the production of
pulse code modulation equipment and
rural automatic exchanges based on indi-
genous C-DOT technology.

WISITEX 88

The fifth World Instrumentation Sym-
posium and Industrial Electronics
Exhibition (WISITEX) will be held from
February 4 to 10,1988 in Bombay. The
symposium will cover industrial process
automation and safety instrumentation,
laboratory automation and quality con-
trol, medical and bio-techpology, new
trends in communcication and informa-
tics, advanced manufacturing
technologies, advances in materials
technology, digital control for power
plants and high temperature supercon-
ductors and their industrial applications.
Buyer-seller meets and release of an
electronics . and instrumentation direc-
tory are the additional features of
WISITEX 88.

BEL TRANSMITTER

The end of 1987 saw the beginning of a
new era for the Bharat Electronics Ltd.
which handed over its first 50 kilowatt
short wave transmitter to the govern-
ment of India. Simultaneously, BEL ef-
fected the formal delivery of a one
kilowatt medium wave transmitter.

BEL has to deliver 20 SW transmitters of
50 KW range to the AIR. These trans-
mitters are being manufactured under an
agreement with Brown Boveri of Swit-
zerland. An automatic turning sys'tem in
the transmitter enables it to be tuned
within three minutes to any frequency in
the short wave band.

The one kilowatt medium wave trans-
mitter, is a fully solid state equipment
developed as a successor to earlier de-
signs which used valves. BEL has an
order for nine of these transmitters.

Derailing Computers

The computerisation plan of the Indian
Railways is facing rough weather- as the
planning commission has suggested dras-
tic pruning of the proposed outlay of

Rs.1200 crores. the commission had set
up an expert panel to examine the
technology upgradation of the railways.
The commission has opined that the pro-
ject should be taken up in phases with
the computerisation of dense areas like
workshop, maintenance of the wagon
fleet and the railway depots. The expert
panel has suggested that railway network
computerisation of such an ambitious
scale has not been attempted anywhere
in the world. In Britain and Canada, lo-
cation of the wagon fleet was vital since
the wagons in these countries are owned
by several private parties. In India the
entire fleet is owned by the railways.

The panel has asked the basic question if
such a computerisation would justify the
high costs and whether the railways
would be able to find the huge resources,
totally estimated at Rs.2200 crores for
the project. Selective computerisation of
operation intensive areas has been re-
commended as a solution.

Domestic Snags

The Centre for Development of Telema-
tics, C-DOT, which won acclaim by de-
veloping an indigenous electronic ex-
change in a record time at a given cost,
seem to be facing some teething troubles
in commercial production. The 512-line
exchange was scheduled to be ready by
early 1988 but problems relating to the
development of necessary software have
come in the way. It took giant companies
with massive R & D infrastructure nearly
a decade to develop the software for digi-
tal electronic exchange.

C-DOT is currently testing its 512-line
exchange at Delhi Cantonment. Though
the calls are going through, some
parameters are yet to be fulfilled.

As an exception, the department of tele-
communications is believed to have al-
lowed the C-DOT to hook its system to
the existing regular telephone network
so that the on line problems of the new
exchange can be quickly identified. The
department always allowed only proved
and tested systems to be hooked to its
network.

Meanwhile, C-DOT has sought an addi-
tional outlay of Rs.52 crores to meet the
cost overruns of its 512-line project and
to undertake futuristic research in the
area of Integrated Services Digital Net-
work. The Planning Commission has ap-
proved Rs. 18 crores against this request.
The department of electronics and de-
partment of telecommunication are sup-
porting their venture, C-DOT, despite

ELECTRONICS NEWS • ELECTRONICS N

the delay in commissioning the projects
as the main objective here is indigenisa-
tion. The electronic switching system-11
depends on the success of the expanded
512-line project. But, the government
would wait till the indigenous technology
is perfected rather than buying it from
abroad.

C-DOT's entry into ISDN, however, is
viewed with scepticism by the Planning
Commission. Before entering this high
technology area, C-DOT would do well
to master its 512-line max project, it is
felt. But the DOT feels the time is ripe
for venturing into ISDN.

Reports indicate that C-DOT is going
ahead with its work on the one-lakh line
project of ESS-II at ITI Bangalore, by
choosing about 140 vendors from all over
the country. An estimated 12 million
pieces of metal film resistors and 80,000
units of wire wound resistors would be
required for the project for which about
40 vendors have been identified. Simi-
larly, vendors have been chosen for the
supply of 52.47 lakh ceramic capacitors,
4.06 lakh electrolytic'capacitors, 48,000
polysterene capacitors and some tan-
talum capacitors. Similar transistors,
diodes, semiconductor devices, hybrid
micro circuits, transformers and IC soc-
kets would be produced.

Strategy And Initiative

A new electronic policy has been prop-
osed by the department of electronics in
two discussion papers entitled "Strategic
Initiatives for electronics" and “New in-
itiative for R & D in electronics”.

The papers puts forth nine objectives for
developing indigenous electronics in the
country with self reliance as the thrust
area in all sectors & the objectives are:
To improve efficiency of agricultural, in-
dustrial and commercial infrastructure
by ensuring efficient, low cost and high
availability voice and data communica-
tion; penetrating into remote locations
of the country.

To ensure rapid dissemination of pro-
ductive and educational information in
the country through creation of efficient
radio, TV and video network;

To harness electronics technology as a
tool to bridge the gap between rural and
urban India, by emphasising rural de-
velopment and nation-building ac-
tivities.

To enable the rural poor to be informa-
tion rich and to speedily educate coun-
try’s illiterate and semi literate popula-
tion through audiovisual media, enabl-
ing them to be more productive;

To maximise export growth that will
more than balance the foreign exchange
outflow needed for essential imports, by
creating profitable opportunities to the
manufacturers of export and by helping
them to be on an equal footing with their
world wide competitors.

To strengthen country's defence through
strategic development of our own war-
fare and relevant implements;

To tone up intelligence network and sec-
urity systems through better hardware
maintenance.

To make India a well recognised and
sought after source of selected electronic
merchandise and services, by capitalis-
ing on our strengths such as in computer
software, ferrites, labour intensive
equipment and components; and to
achieve better planning, measurement
and mid-course correction through
quicker data acquisition, processing and
wider access.

Quoting the seventh plan document, the
paper estimates the gross output in elec-
tronics by 2000 AD. to be Rs. 50,000
crores from the present level of Rs. 3,850
crores. The output target for 1989-90 is
Rs. 10,800 crores.

The procedure for processing industry
will be streamlined to the effect that the
application processing time will be re-
duced to 22 weeks from the present 72
weeks, the paper says.

Delicensing will be extended to all sec-
tors of electronics industry which are
based on indigenous technology. De-
licensing will also be extended to all pro-
jects where the design infrastructure
capability exists within the country.
Systems in engineering houses and inde-
pendent commercial R & D houses will
be registered as DOE units and will be
treated on a par with manufacturers and
actual users.

Wherever import of technology is in-
volved, care will be taken to ensure that
foreign collaborators do not benefit at
the cost of the country. The government
will indicate a few norms such as
minimum plant capacity and maximum
price for plant and fee for technology.
The need to import technology is often
out of compulsion due to the inability or
lack of planning by various user agen-
cies. This will be avoided by appropriate
coordination with the user agencies.
Decentralisation of electronics industry
approval process by transferring small
scale industry approvals to state directo-
rate of industries will enable state level
implementation of small projects, ac-
cording to the paper.

All industries, irrespective of their size,
will be asked to submit data about their
production and imports to a national
data base. No governmental assistance
will be available unless all required data
in standardised format is submitted
along with any application for govern-
ment assistance.

Thrust will be applied to increase per
capita availability of telephone in rural
areas.

To encourage production based on indi-
genous knowhow, special incentives
have been proposed. They are :

Reduced duty on capital goods, raw
material and components tor production
based on indigenous R & D; extra excise
duty on products made from SKD and
CKD kits; total delicensing for indigen-
ous knowhow-based entrepreneurs, with
no capacity restriction; those who failed
to fulfil the promised indigenisation to
be denied further foreign collaboration;
expenditure on R & D with the qualifica-
tion and experience of personnel to be a
criterion for deciding indigenisation; im-
ported test equipment, strictly for in
house R & D of Indian companies, will
be duty free.

On foreign collaborations, the policy
paper says that emphasis should be on
net foreign exchange earned rather than
on the gross value of the export.
Commending the virtue of rationalisa-
tion of fiscal policies, the paper says that
the long-term import duty structure for
materials, components and finished
equipment will be further streamlined.
The usual criticism that poor economy of
scale of production hindered the elec-
tronics industry in India can no longer be
true since there is no limit to capacity and
foreign collaboration proposals have
been approved in. large number. How-
ever. the home market is not growing
and products designed in India and
based on Indian components are fewer.

To help expand the market for compo-
nents, the suggested policy options are;
Promoting products which are designed
in India; allowing realistic duty draw-
back and cash assistance to ensure that
the cost of materials, machinery and
power accessible to the Indian exporters
is the same as incurred by his foreign
counterparts by refunding all such extra
costs; and giving diriect export aid like
that of Electronic Sales Support Organi-
sation programme which includes com-
prehensive assistance from warehousing
to office facilities.

If necessary, an Electronics components

ELECTRONICS NEWS • ELECTRONICS N

development fund may be set up to fi-
nance the component industry.

All basic materials needed by the elec-
tronics component industry like copper,
aluminium, gold-nickel alloys, plastics
and adhesives are very expensive in
India compared to the world market
prices, it is proposed to give a capital
base to such industries in public sector as
an outright grant and then asking them
to compete with imported prices without
further assistance. Or else, such basic
materials may be made available at con-
cessional duty to the component industry
through agencies like ET & T.

The lead paper has called for the forma-
tion of a holding company under DOE
for the strategically critical areas of elec-
tronics where private investment is not
forthcoming. The holding company will
have the task of investing in major pro-
jects such as selected semiconductors
and high technology professional equip-
ment.

A centre for excellence in advanced
maintainance techniques and maintaina-
bility research for computers, communi-
cations and control systems will be set up
by the DOE. This centre will give techni-
cal and managerial guidance in establish-
ing national resource centres for mainte-
nance in different states.

Real-time solids modeller

The Real-time Solids Modeller from Sili-
con Vision provides a sophisticated
design tool for creating 3-D solid or
wireframe objects of any complexity.

Unlike other packages, this system can
perform full colour hidden-surface
removal for any solid object at high
speed. Colour hard copy of the designs
can then be produced on a range of

popular pen plotters or printers for pro-
fessional results. Further information
from Silicon Vision Ltd • 47 Dudley
Cardens • HARROW HA2 ODQ •
Telephone 01-422 2274.

CMOS version of 80186

Intel has recently introduced a
CHMOS* version of its 80186, a highly
integrated microprocessor for embedded
control applications in industrial auto-
mation, communications, and office
automation. It is type-coded 80C186 and
is currently available for 12.5 MHz oper-
ation; a 16 MHz version will become
available shortly.

The 80086 in compatible mode is pin-
for-pin compatible with HMOS* 80186
systems, but it has been enhanced to
support Intel’s next-generation numerics
co-processors.

It is claimed that, under normal
operating conditions, the 80086 needs
only 20% of the power the HMOS 80186
requires.

*CHMOS and HMOS are patented pro-
cesses of Intel Corp.

World's fastest 8-inch
Winchester disc drive

A new family of 8-inch Winchester disc
drives that combine the world’s fastest
seek performance with a very high level
of reliability has been announced by Ver-
mont Research.

The first product in the new Ascutney
family, the model 7030, is a 600 Mbyte
unit with extremely fast seek times: a
mere 22 ms for full stroke seek, which is
nearly 40% faster than most 8-inch disc
drives. A rigorous testing programme
has established a predicted MTBF of
over 30,000 hours.

• Cteeve Road » LEATHERHEAD
KT22 7NB • Telephone (0372) 376221.

Thermoprobe replaces
voltmeter

A new, low-cost electronic test instru-
ment called Thermoprobe quickly ident-
ifies dead active components on PCBs
without direct contact.

The solid-state device consists of a ther-
mistor probe connected to a modified
Wheatstone bridge circuit. It is designed
to measure minute temperature changes
of 0.02 °C. Since defunct resistors,
transformers, diodes, or ICs, do not
emit heat they can be quickly identified
on the unit’s built-in S-meter as the
probe is moved in close proximity to

Its small size makes the Thermoprobe
particularly useful in field service appli-
cations for computers, electronic instru-
mentation, video, and hi-fi equipment.
It operates from a 9-volt battery.

The Thermoprobe is available at $21.95
from Melrifast • 51 South Denton
Avenue • New Hyde Park • New York
11040 • USA.

Revolutionary electron beam
tester

An innovative new system for testing
VLSI chips has been introduced by
Schlumbergcr Technologies Automatic
Test Equipment. Combining scanning
electron microscope (SEM) technology
with CAD/CAE tools within a familiar
workstation environment, the
company’s Integrated Diagnostic System
(IDS) 5000 Workstation brings un-
precedented efficiency and accuracy to
the field of VLSI diagnosis and
characterization, yet remains easy for
engineers to use. The system was
launched at the recent Productronica
trade fair in Munich.

Full details from Schlumbergcr
Technologies, Automatic Test Equip-
ment • Ferndown Industrial Estate •
WIMBORNE BH21 7PP • Telephone
(0202) 893535.

TELECOMMUNICATION NEWS • TELECO

Indian Maritime Communica-
tion

India will shortly join the select band of
nations in providing a coastal earth sta-
tion link to the Inmarsat space segment,
enabling direct maritime communica-
tions between ships and shore based in-
stallations.

The Videsh Sanchar Nigam Ltd., a gov-
ernment of India undertaking, will set up
the CES at a cost of Rs. 16 crores at Arvi
near Pune in Maharashtra by mid-1989.
The International Maritime Satellite Or-
ganisation (INMARSAT) is an inter-
governmental, international body com-
prising 53 members. India is one of its
founder members. Set up in 1979, In-
marsat provides reliable commercial
communication links like telephone,
telex, data transmission and so on via
satellite to all the ships and off-shore

In the Inmarsat systems, the INMAR-
SAT organisation provides the satellites
or space segment. Presently, nine satel-
litess have been leased, three each in In-
dian Ocean, the Pacific and the Atlantic
Ocean regions, from the International
Satellite Organisation (Intelsat satel-
litee), the European Space Agency
(Marecs satelliotes) and Comsat Organi-
sation of USA (Marisat satellites).

A satellite relays signals received from a
ship to distant shore station and from dis-
tant shore station to ships. Ship-satellite-
ship links use L brand (1.5 GHz /
1.6GHz) frequencies while shore-satel-
lite-shore links use C band (6 GHz /
4GHz) frequecies. A ship desirous of
having Inmarsat facilities should be
equipped with a ship terminal called ship
earth station (SES) and it is the responsi-
bility of the ship owner to have one such
approved terminal fitted on board.

The present standard A, SES terminal
costs US $ 25,000. Inmarsat will soon in-
troduce very cheap and simple ship ter-
minals called standard C terminals.
These terminals will cost $ 5,000 each.
The Inmarsat system was put into opera-
tion on February 1, 1982 with about six
CESs, serving 1000 SESs. By 1987, it
grew to serve 6200 ship terminals, in-
cluding 40 belonging to the Indian
maritime fleet, through 20 coastal earth
stations. In the Indian Ocean region.
Seven CESs operating at present are
Yamaguchi (Japan), Eik (Norway),
Odessa (USSR), Thermopylae
(Greece), Nakhoda (USSR), Jeddah
(Saudi Arabia) and Psary (Poland).

Inmarsat is now poised to serve flying
aircraft as well as land mobile services
like trucks and railways. It has plans to
provide communication links to flying
aeroplanes to carry air traffic control sig-
nals, traffic data and speech signals. In-
marsat is procuring new satellites of high
capacity to meet the growing demands
for ship services and to meet the new air
and land services. The second genera-
tion satellites to be procured by mid-
1988 will have 125 channels for tele-
phone as against the 40 in the existing
satellites.

The Indian earth station at Arvi will
serve both standard A and standard C
ship terminals and flying aircrafts in fu-
ture. It may be possible to ring up a per-
son flying on board an aircraft from
home or office. Operation of this station
will result in savings in foreign exchange
which is now being spent by the Indian
ships for routing their calls through a
foreign earth station. Indian ships will
spend much less amount in Indian cur-
rency by using the Arvi station. Also, the
competitive tariff will attract more
foreign ships to use our station.

The Arvi station will be linked to the na-
tional and international gateway ex-
changes at Bombay. In addition to con-
ventional telephone, telex and slow
speed data services, the station will pro-
vide instant communication to Rescue
and Coordination Centres from a ship in
distress. The station will handle SOS
messages from Emergency Position Indi-
cating Radio Beacon system. EPIRBs,
being developed by Inmarsat, will get
detached from a ship in disaster, float in
water surface and contact with water will
activate the beacons to transmit signals.
The Arvi station will pick these signals
and pass on to the rescue centres on a
priority basis.

Telecom Commission

The government of India is considering a
proposal to set up a "Telecom Commis-
sion” on the lines of the Atomic Energy
Commission. The commission is ex-
pected to oversee telecom operations,
evolve policies, formulate the guidelines
and determine frequencies of various
channels. The day to day operations and
implementation of projects will be given
to state level telecom bodies.

The idea of Telecom Commission is said
to have emanated from Mr Sam Pitroda
technology adviser to the Prime Minister
and the founder of C-DOT.

With the formation of Telecom Commis-
sion, department of telecommunication

may be wound up. The department has
been functioning both as a policy making
body and implementing agency.

Dot Lab

The Department of Telecommunication
will set up a Laboratory for research in
optic fibre technology at an estimated
cost of Rs. 2 crores. The proposed
laboratory will have links with the Inter-
national Telecommunication Union,
based in Sweden.

Link With Vienna

A scientist at Bhabha Atomic Research
Centre, Bombay, can now directly get
information from the International
atomic Energy Agency, Vienna, through
computer terminals.

The Vienna database is now accessible
through International Subscriber Dial-
ling. As an experiment, BARC’s compu-
ter division connected a computer termi-
nal through a modem to a telephone line
with ISD facility. BARC library and in-
formation services division formulated a
search query in the format that is accept-
able to the package that retrieved infor-
mation from the database.

Command can be given to get the re-
trieved information file printed at Vie-
nna and the output can be obtained
through mail. The retrieved items, either
in toto, or any selected portion such as
title or author or abstract, can be read by
giving appropriate commands. The file
containing the retrieved items can also
be seen in part based on stipulated condi-
tions. Since data transmission is at the
rate of 300 bauds, browsing retrieved
items is likely to take a good deal time
and consequently it will cost a lot of
money.

SWITCH MODE POWER SUPPLY

Following the general introductions into the operation and application of switch mode power supplies in
reference "» and a \ this article discusses a practical design of a versatile, compact and highly efficient
SMPSU rated at 2.5 A.

The circuit described here is based
around the integrated switch mode
power supply controller Type L4960
from SGS. Briefly recapitulating the
general introduction in reference this
IC has the following features:

■ Input voltage range: 9—50 VDC

■ Output voltage adjustable between 5
and 40 V.

■ Maximum output current: 2.5 A.

■ Maximum output power: 100 W.

■ Built-in soft start circuitry.

■ Stability of internal reference: ±4%.

■ Requires very few external compo-
nents.

■ Duty factor: 0—1.

■ High efficiency: >j up to 90%.

■ Built-in thermal overload protection.

■ Built-in current limiter for short-
circuit protection.

The pinning of the 3 regulators in SGS’s
L496X series of switch mode regulators
is shown in Fig. 1. The Type L4964 is
housed in a special 15-pin enclosure, and
can supply up to 4 A. The Type L4962
is the least powerful version (1.5 A),
housed in a 16-pin DIL package.

The internal organization of the Type
1.4960 has been discussed with reference
to Fig. 1 in |J| . The operation of the on-
chip soft start circuitry, and the current
limiter, is illustrated by the waveform
diagrams in Fig. 2a and 2b, respectively.
The thermal shutdown circuit in the
L4960 is activated when the chip junc-
tion temperature exceeds 150 °C.

For the sake of safety, the proposed
SMPSU is a transformer based design.
The alternating input voltage to the
board is obtained from the secondary of
a mains transformer, which ensures that
the DC input to the regulator is at least
3 V higher than the required output
voltage at the maximum output current.
It goes without saying that the trans-
former is preferably a toroidal type.

Circuit description

Figures 3a and 3b show the circuit
diagram of the mains section, and the
DC power supply, respectively. The alter-
nating voltage from the secondary
winding is applied to the respective in-
puts on the supply board, while the
centre tap is connected to ground. The
unregulated input voltage, i/i, for the
L4960 is supplied by a full-wave rectifier

2.22 MHO

circuit composed of two 3 A diodes Type
1N5404, Di-D:, and an electrolytic
reservoir capacitor, Ci. Network Ri-Cj-
Cj defines the gain of the closed regu-
lation loop. A second network, C:-R:,
is dimensioned for an oscillator fre-
quency of about 100 kHz. The function
of capacitor O is twofold: it defines the
period of the soft start ramp (see Fig.
2a), as well as the average short-circuit
current. The feedback input of the regu-
lator is connected to the junction of out-
put voltage divider Rj-Rj.

The output voltage, (/«., of the L4960 is
calculated from

i/o = 5.1[(R.i + Rj)/Rj] assuming that
t/i-(/o£3 V.

Remember that the minimum value of
U\ is 9 V. A fixed output voltage of
5.1 V (±4%) is obtained when Rj is
omitted, and Rj replaced with a wire
link. When R> has the fixed value of
5K6, Rj alone determines the output
voltage:

Uo = 9 V: Rj = 4K28
Uo=\2 V: Rj = 7K58
Do =15 V: Rj= 10K87
Uo= 18 V: Rj= 14KI6
Do = 24 V: Rj = 20K75

These resistor values are, of course,
theoretical, and require ascertaining in
practice. The output voltage is readily

SGS Ates.

.

b ,*

..... -Iiriniirn

1-

Fig. 3. Circuit diagram of the mains section (a) and the switch mode power supply (b).

Resistors ( ± 5%):

Rl-ISK

R2=4K3

R3;R4 = dimension as required for Uo: see text
Capacitors:

Ci » lOOO/i: 40 V; radial (pitch 7.5 mm|

C 2 = 2n2
C3=33n
C4=390p

Ca = 2/i2; 40 V: radial (pitch 2.54 mml
Ca;C7 = 220/i; 40 V; radial (pitch 5 mm)
C8=100n

Semiconductors:

Di;D2 = 1N5404

D3=8R05* (SGS-Ates) or BYV28 (Mullard)

ICi = L4960* (SGS-Ates)

' SGS-Ates (UK) Limited • Planar House •
Walton Street • Aylesbury HP21 7QJ. UK

Inductor:

Li = 150. . . 300/iH toroid suppressor, e.fl.
Newport 1400/1 1/3 or Siemens B82500-B-
A10 (axial). Home made: approx. 60 turns of
01 mm. enamelled copper wire on a suitably
rated 0 1 5 - 20 mm ferrite core.

Miscellaneous:

T0220 style heat-sink.

T0220 style mica washer and bush.

tap; ratings as required (see textl. E.g. ILP
series: Jaytee Electronic Services • 143

Kent CT6 6PL. Telephone: (0227) 375254.
PCB Type 880001 (see Readers Services page).

4

made variable by fitting R- = 6K8 and
replacing Rj with a 25K potentiometer.
Power diode D> is included as a safety
measure. 'This fast rectifier limits the
negative potential at the input of choke
Li to a safe -0.6 to -1 V when the
output transistor in the regulator is
switched off. In the absence of Ds, the
voltage at pin 7 would increase
dangerously to several volts below the
ground potential. Choke Li is an essen-
tial part in the L-C network for noise

and ripple suppression at the output of
the supply.

Construction

The compact printed circuit board for
the SMPSU is shown in Fig. 4. The com-
pletion is extremely straight-forward.
Start off by selecting resistors Rj and Rj
as explained above. Fit the components
in the centre, Ri,..Rj inclusive and
C2 . . . Cs inclusive. Prior to soldering

Fig. 4. The printed circuit board for the compact SMPSU.

onto the board, regulator ICi and
power diode D; are bolted back to back
onto a common heat-sink as shown on
the component overlay. Do not forget to
keep the heat-sink electrically insulated
from the metal tab of D.i with the aid of
a thick" mica washer and a plastic bush.
It is possible to use the Type BYV28
diode in position D>. This device is
housed in an axial SOD64 enclosure,
however, and requires a different mount-
ing method than the TO220 style 8R05
from SGS. The BYV28 is mounted up-
right, but the anode lead is not bent
down until is has been soldered to a
cable eye. The eye is then insulated and
fitted onto the heat-sink as detailed
above. Whatever diode or mounting
system is used, check the insulation with
a continuity tester! Push the leads of
ICi and Dj into the respective holes un-

til the heat-sink rests securely onto the
board surface. Solder the leads and cut
off their excess lengths. Now fit the re-
maining components, Li, Ci, Cs, C-, Cs,
Di and D 2 . Be sure to observe correct
polarization of the diodes and the elec-
trolytic capacitors. Great care should be
taken to avoid any likelihood of a short-
circuit between the winding on the choke
core and the heat-sink for the regulator.
It is recommended to secure Li with the
aid of a central nylon bolt and nut.

Testing

Check the position, insulation and
orientation of all the parts on the board
before connecting this to the secondary
of the mains transformer. It should be
noted that the supply requires a load at

all times for proper operation. When the
SMPSU is fed with 30 VAC, and loaded
with 2 A at an output voltage of 5 V, the
temperature of the heafisink should not
exceed about 60 °C at room tempera-
ture. The efficiency of the supply under
these conditions is approximately 68%.
With a load of 2 A, the efficiency in-
creases from 80% at U° = 10 V, 85% at
Uo = \5 V, to 87% at at £/» = 25 V. Gb

References:

111 Switch mode power supplies, Elektor
India, November 1987.

1:1 High-current switching regulator IC
simplifies supply design, Elektor India,
November 1987.

DOUBLE TRACE EXTENSION FOR
VLF ADD-ON UNIT

by E Fano

A handful of components and some minor alterations to the VLF add-on unit for oscilloscopes enable
this popular unit to be used for two-channel measurements.

Thanks to its versatility, low cost, and
case of construction and use, the VLF
add-on unit for oscilloscopes published
in reference 10 has become one of the
most popular construction projects
featured in Elektor Electronics . The fol-
lowing description shows that the circuit
is readily modified to achieve two-
channel operation on a single-beam os-
cilloscope. The required modification
and the extension circuit are useful for
many applications involving the simul-
taneous analysis of 2 slowly varying
quantities.

Input multiplexing and 256
bytes more

In the original design of the VLF storage
unit, address lines A8 to A10 inch of
RAM IC 2 are kept permanently logic
low. This means that only the first 256 of
the available 2048 programmable lo-
cations in the RAM are used for storage
of converted data. The idea behind the
present extension is to drive address line
A8 with a signal that creates an ad-
ditional data block of 256 bytes. This
block is written to during every second
display — conversion cycle, and can thus
hold the digitized data for a second in-
2.24

put channel. The input signal for opamp
ICi is, of course, multiplexed according-
ly.

The modifications

The bold lines and the shaded area in
Fig. 2 give all the necessary details on the
modifications and the extension circuit.
It is recommended to cut the connection
between pins 23 and 22 of the socket that
holds RAM IC:. This modification is
performed at the track side of the board,
and effectively insulates pin 23— address
line A8 — of the RAM. Solder a wire to
the insulated area that connects to socket
pin 23, and run it to pin 3 of ICiob ( =
output 1QA). Construct the input
multiplexing circuit in the shaded area
on a small piece of prototyping board,
and connect it to the VLF storage unit as
indicated by the bold lines. The elec-
tronic switches in the 4066 -toggle on
each pulse from output 1QA. This ar-
rangement ensures that the correct data,
i.e., the measurement values for each
channel, are stored in the relevant 256
byte area in the RAM. Zener diodes D,\
and Db protect the inputs of the 4066
against overvoltages. To prevent distor-
tion of the displayed image, the signal

Fig. 1 Output waveform of the 2-channel
storage unit using incorrect (a) and correct
(b) trigger settings on the oscilloscope.

Fig. 2 Circuit diagram of the complete 2-channel unit. The modifications are shown in bold lines, the extension in the shaded

applied to input B may not go negative, parasitic capacitance blocking the References:

Figures la and lb illustrate the effect ob- decimal counter, and may be remedied

tained with the modified and extended by fitting a 33 k£2 series resistor on the in VLF add-on unit ' Ibr osctllloscopcs.

VLF storage unit. trigger output line to the oscilloscope. Elektor India, March 1987.

Finally, some constructors have reported This resistor also protects the output of

the absence of count pulses on pin 7 of FF2 against short-circuits on the trigger

IC10 on selecting display range b output. Th

(12.5 s). This effect is probably due to

SOFTWARE UPDATE FOR uV
CONTROLLED FREQUENCY METER

The software for the Elektor micropro-
cessor controlled frequency meter (see
reference 11 ’) can be updated as shown
here to enable selecting triggering on the
leading or trailing edge of pulses when
the meter is set to the event count mode.

The addition to the existing program en-
sures a stable readout when a sawtooth-
like input signal is applied to the meter,
because triggering can take place on the
fast rising edge.

As from now, the EPROM for the pro-
ject, ESS536N, contains the update.

Reference:

01 Microprocessor controlled frequency
meter. Elektor India, February 1985.

,98a 2.25

SYNCHROTRON X-RAYS
REVEAL HOW ICE FLOWS

by Dr Robert Whitworth, Department of Physics, Birmingham University

Ice covers a great deal of the Earth’s surface. Understanding the way it flows is important for many
practical purposes such as exploiting oil reserves and helping the engineer in an ice-bound environment.
An ice physics group at Birmingham University is employing a powerful research tool, the synchrotron
source of X-radiation at the UK Science and Engineering Research Council’s Daresbury Laboratory, to
reveal detail of the flow mechanism of ice in a way not previously possible. Ice has been found to be
an especially suitable material for study by this means and the group’s findings promise to yield new
insights into crystalline materials.

The process of slip by which plastic deformation occurs in crystals. The original block at the
left is deformed to the shape on the right by one horizontal layer of atoms slipping over
another hy the spacing between two atoms. This docs not lake place all at once: in the in-
termediate stale shown here, slip starling at A has reached the broken line BC This is called
a dislocation line and is a region of distortion in the crystal. The deformation of crystals oc-

Formation of images of a crystal by X-ray topography using the beam from the Synchrotron
Radiation Source. Because the beam is highly parallel, each point on the crystal diffracts the
X-rays to the corresponding point in each of the diffraction spots formed on the film.

Everyday experience tells us that pieces
of ice arc brittle. Yet glaciers, large
masses of ice, flow down the sides of
mountains. In the polar regions huge
sheets of ice flow out towards the
oceans. The significance of such flow is
brought home when something unusual
happens as, for example, when the Hub-
bard Glacier in Alaska advanced rapidly
during 1986 and blocked an inlet from
the sea.

Engineers who work in the ice-bound
regions of the world erecting buildings
or drilling for oil have a great interest in
the mechanical properties of ice, while
ice physicists seek to understand the fun-
damental principles of its behaviour.
With ice, or any other solid, there is still
a great deal to be done before we can
answer the question “Given the
crystalline structure of a material and
the properties of a single molecule, how
fast will the material deform when a
stress is applied to it?” At that point the
problems of engineering and of physics
interact and progress in either field can
be of benefit to the other.

We can tell from the regular forms of
snowflakes or the patterns formed by
frost that ice is crystalline. This means
that the molecules within it are arranged
in a regular pattern. Bulk ice consists of
many crystals or grains of different
orientation, all joined together to form a
solid mass. When such polycrystalline
ice is deformed, several processes may
occur. With each grain, layers of mol-
ecules may slide over one another; this is
known as slip. In addition, grains may
move relative to their neighbours. Dur-
ing deformation the structure suffers
damage, which can be annealed out by
the diffusion of molecules or by the ,
nucleation and growth of new grains.
These processes occur in all crystalline
materials, but their relative importance
depends on the material and on the tem-
perature. The basic process is slip; but
when ice is close to its melting point the .
other processes are important, too.

Dislocations

It is a little over 50 years since it was

realized that one layer of a crystal can-
not slip over another as a rigid whole. If
we imagine that slip over one atomic
spacing starts at a corner, for example
point A in the first diagram, and con-
tinues up to the line BC, this line marks

a region of distortion in the crystal. It is
called a dislocation line or, for short,
just a dislocation . Arrangements of the
atoms near B and C are then as il-
lustrated. The plastic deformation of a
single crystal takes place through the

movement and multiplication of dislocations during the plastic deformation of a single crystal
ve prints, so dislocations appear bright against a dark background. The width of the region

Three topographs from a sequence showing the
of ice at - 20 °G These illustrations are negali
shown is 4 mm on the crystal.

movement of these dislocation lines.
How much stress is required to move
them and the way in which they move
have been extensively studied in metals
and many other materials: the most
powerful technique for observing dislo-
cation lines is transmission electron
microscopy. A limitation of this tech-
nique, however, is that the samples have
to be very thin foils held in a vacuum,
and it has so far proved impossibly diffi-
cult to study the properties of dislo-
cations in ice in this way.

A different technique, which has in the
past been used successfully to observe
the dislocations in ice, is known as X-ray
topography. Great improvements to this
technique have recently been achieved
through the use of X-rays from a syn-
chrotron radiation source. By using the
source at the UK Science and Engineer-
ing Research Counsil’s Daresbury Lab-
oratory we have been able to investigate
the motion of dislocations in crystals of
ice several millimetres in size in a degree
of detail that has not previously been
possible.

The Synchrotron Radiation Source is a
large installation in which a beam of
electrons with an energy of 2 GeV (giga-
elcctronvolts) is bent by magnets to cir-
culate in an orbit of diameter 30 m. As
it does so it emits intense electromag- ]
netic radiation in the plane of the orbit,
with a spectral range from infra-red to |
X-rays. When the X-ray beam falls on a j
suitably oriented single crystal, as shown
in the second diagram, the crystal dif-
fracts the X-rays to produce a number of
beams in accordance with Bragg’s law of
diffraction. When the beams fall on a
screen behind the crystal they form
spots. In topography experiments at
Daresbury the X-ray beam travels 80 m
before it reaches the crystal, so it is virtu-
ally parallel. Because of this, every point
on each diffracted spot corresponds to a
particular position in the crystal. How-
ever, the diffraction from any point
where the crystal is distorted will be
slightly altered in such a way as to pro-
duce contrast at the corresponding point
in the diffraction spot. In our ex-

periments the diffraction spots are
recorded on photographic film, thereby
providing images of the crystal known as
topographs. Dislocations are clearly vis-
ible within them, usually as dark lines. A
sequence of such topographs taken while
the crystal is under stress reveals the mo-
tion of dislocations as the crystal
deforms. Compared with a normal X-
ray source the synchrotron radiation
beam is very intense and exposure times
are comparatively short (typically 15 s).
This makes proper dynamic experiments
a realistic proposition for the first time.

Multiplication

Crystals used in these experiments must
be of very high quality because the in-
itial density of dislocations has to be low
enough for them not to obscure one
another in the topographs. Large, single
crystals are specially grown in Birm-
ingham for the experiments. In the
course of deformation, the density of

dislocations increases, as seen in the
series of topographs making up the third
illustration. This increase is essential to
the process of deformation. The dislo-
cation line is the boundary on the slip
plane between the region which has
slipped and the region which has not;
the rate at which the boundary moves
corresponds to the rate of deformation,
but if all the dislocations that were pres-
ent to begin with merely moved across
their slip planes and out of the surface
the deformation would stop. New dislo-
cation lines are created by a process of
multiplication, the mechanism of which
was first suggested in 1950 by F.C. Frank
of Bristol University and W.T. Read of
Bell Telephone Laboratories. If a dislo-
cation line does not lie entirely on one
slip plane, but makes a step from one
plane to another, the dislocation lines on
(he separate planes form a spiral around
the step as shown in the fourth illus-
tration. This means that the dislocations
do not pass out of the crystal but con-

in the lower right of the sequence of topographs.

uafv 1988 2.27

tinuously grow longer. Near the centre of
the spiral, slip amounts to as many
atomic spacings as there are turns in the
spiral.

In the sequence of topographs, a pair of
opposite spirals can be seen developing
in the lower right-hand corner. The pos-
ition of the step from which they
originate is marked by the arrow in the
first topograph. Often, the multipli-
cation is more complicated as is shown
in the topograph on the front page.
When we recall that the dislocation line
is the boundary of the region which has
slipped, it is remarkable how convoluted
the boundary can become. These
topographs of ice are some of the
clearest images of dislocations moving in
crystals of this size ever obtained in any
material; the study promises to yield an
understanding of the processes of
dislocation multiplication in bulk
samples which will be significant in the
understanding of other materials, too.
The crystal structure of ice Has a hex-
agonal axis of symmetry which is easily
seen in the shapes of snow flakes, and
deformation of single crystals is know to
occur almost entirely by the slip of
planes, called the basal planes, lying

Diagram of the costal structure of ice. The circles represent water molecules, each of which
is linked to four others by hydrogen bonds. The hexagonal axis of symmetry is vertical and
the normal slip planes (the basal planes) are horizontal. There arc two sets of such planes,
the glide set and the shuffle set, between which slip may occur. Possible prismatic slip planes
arc shown as vertical. As in the case of basal slip, the slip direction on these prismatic planes
is horizontal, approximately into or out of the plane of the diagram.

This image shows part of a single crystal of ice after a small amount of plastic deformation. It is an X-ray topograph, formed by diffraction
of X-rays from the Synchrotron Radiation Source at the UK Science and Engineering Research Council's Daresbury Laboratory. Deformation
takes place through the movement of defects in the crystal known as dislocations, seen here as bright lines. Dense bands of them are spreading
from the left-hand side and are just beginning to do so from the right, generated by a process of dislocation multiplication which can be seen
happening more clearly in the formation of the tangled loops at the lower right. The spiral to upper right of centre was formed by diffusion
of defects prior to deformation. Taken by scientists from Birmingham University, the topograph is of an area of crystals one centimetre across.

perpendicular to this axis. The final
diagram of the structure shows that there
are two kinds of basal planes between
which slip might occur, in two sets called
the glide set and the shuffle set. it has
not yet been possible to find out by ex-
periment which set slips. By comparison
with other materials of similar structure,
and for theoretical reasons, the glide set
seems the more likely. From the diagram
it is far from obvious why ice should not
deform just as easily by slip on vertical
planes known as prismatic planes. But
experiment shows that it takes about 50
limes more stress to deform ice in that

To deform a polycrystalline mass of ice,
slip within the grains must take place in
more than one set of parallel planes if
the grains are not to come apart at the
boundaries. This is the main reason why
polycrystalline ice is stronger than a
single crystal. From our topographs we
have found that, in certain circum-

stances, dislocations on prismatic planes
move as fast as or faster than those on
basal planes. This comes as a surprise to
anyone who knows how difficult it is to
deform ice along the prismatic planes,
and it looks as though the preference for
basal slip arises not because it is difficult
to move dislocations along other planes
but because multiplication is easier in
the basal plane.

Another effect that is important poten-
tially is that the strength of ice is very
sensitive to the presence of minute traces
of certain impurities. Some years ago.
Dr S.G. Jones ans Dr J.W. Glen in our
laboratory found that one part in a
million of hydrogen fluoride reduces the
stress needed to deform a single crystal
of ice to as little as one-quarter of nor-
mal. It may be significant that small ad-
ditions of hydrogen fluoride also pro-
duce large changes in the electrical
properties. We hope to use X-ray top-
ography to investigate these effects, and

so learn more about the processes which
limit the rate of movement of the
dislocations.

A proper understanding of the physics
governing dislocation movement in a
material is necessary for a theoretical in-
terpretation of the strength of that
material. Metallurgists have made great
use of knowledge gained from electron
microscopy in developing new alloys for
use in aircraft, nuclear reactors and
elsewhere under other extreme con-
ditions. A similar understanding of the
physics of ice is needed to predict or
modify the behaviour of ice in the en-
vironment. It is even more necessary
when we wish to forecast how ice will
behave in less familiar conditions. For
example the recent discovery that some
of the moons of the planets Jupiter and
Saturn are composed mainly of ice has
led to questions about how ice would
deform on a geological time scale at high
pressures and at low temperatures.

Transistor letter symbols

lipolar

SIR CLIVE SINCLAIR: SUPER
ELECTRONICS ENTREPRENEUR

by Martin Ince

As both a leading developer of original
ideas in ihe electronics industry and a
businessman who pioneered mass mar-
keting of his products well below the
prices of his rivals. Sir Clive Sinclair has
firmly established himself as the best
known Britain in his field today.

A public opinion poll a few years ago in
the United Kingdom placed him in a list
of the top ten scientists of all time
alongside Leonardo da Vinci and Albert
Einstein, although he has no pro-
fessional academic background and did
not even go to university. After over 25
years in business, he is still producing
ideas for new electronic products and
launching the firms to manufacture
them.

To achieve this, he often drives research
teams to turn his ideas into products
ahead of the competition worldwide.
But his commercial origins lie in techni-
cal fields far removed from the wafer
scale chips and portable telephones that
now dominate his business plans.

Early commercial pursuits

It is noteworthy that his early days offer
little clue that he would become the one
individual associated in many people's
minds with the spread of high tech-
nology products into everyday life. He
left school at 17, with a less than glitter-
ing academic record, and before starting
his career as an independent business-
man, he was employed on a magazine
for amateur electronics enthusiasts.
Wireless World.

The publication is famous throughout
electronics and communications, and for
publishing Arthur C. Clarke’s 1945
futuristic article on geostationary com-
munications satellites — creations that
now dominate international telecom-
munications.

His first company, Sinclair Radionics,
opened its doors in 1962 and in many
ways marked the start of his career as an
electronics entrepreneur. Its business was
to supply kits to radio amateurs who
wanted to make their own equipment,
but had previously had to go to a variety
of sources for the necessary parts.

The idea was ingenious and worked
comparatively well, but lacked the other
key ingredient of Sinclair’s business
projects in recent decades— the possi-
bility of a mass market for the product.

Fig. 1. Sir Clive Sinclair, electronics genius.

There simply are not millions of people
wanting to build their own radios.
However, there was a potential mass
market for his next generation of
products— a range of cheap calculators
and digital watches designed (in another
Sinclair hallmark) to make sophisticated
equipment available at well below the
prices of rival suppliers.

Low cost computers

While some satisfied users of Sinclair
calculators are still to be found in the
United Kingdom, the increasing com-
petition from Far Eastern and other
major suppliers of cheap consumer elec-
tronics soon meant that these prod-
ucts — commercial successes when they
were launched — ceased to be major
money-spinners. But by now his atten-
tion was on another market— that for
computers — where his name would be
made on a wider scale.

At this stage, his tactic was to produce a

Fig. 2 . Sir Clive Sinclair and his miniature

computer for under £100 — vastly less
than the price of the Apple machines
then sweeping the world. The result of
Sinclair’s computer development pro-
gramme was a series of machines of in-
creasing power, starting with the ZX80
and proceeding through the ZX81, the
Spectrum (regarded by the experts as the
best Sinclair computer) and the QL (for
Quantum Leap).

Sir Clive’s vital role in British electronics
has been to open up markets for sophis-
ticated goods at low prices, sometimes in
a way that has benefited competitors
more than himself.

His own computers were not best-sellers
in the long term, and his interests in this
field have now been sold to Amstrad, the
firm that has popularized home comput-
ing and word processing in the United
Kingdom. Another Sinclair product, the
C5 electric car, which is no longer on the
market, was even less of a success, this
time because of technical problems with
the vehicle itself.

High technology telephones

But he shows no signs of bowing out of
the high technology business. He is a
director and part owner of a new firm,
Shaye Communications, which plans a
new, miniaturized form of portable tele-
phone for business and private users.
This time the aim is to produce a port-
able telephone at about half the size and
a fifth of the price of existing machines,
with a retail price of about £200.

The idea is to manufacture a telephone
that can be used via a network of public
access points, providing a route into the
general network and therefore offering
many advantages over cellular telephony.
Because Britain has an acute shortage of
radio spectrum space, the telephones
would economize on radio spectrum
space and would be less prone to inter-
ference and eavesdropping than existing
cellular phones.

Moreover, Sinclair recently launched his
latest computer, the Z88, a cheap laptop
machine costing about £230, weighing
less than 4 kg and under 3 cm thick. It
cannot be sold under his name, however,
because Amstrad now has the rights to
call a computer a Sinclair. Instead, it
is sold by Sir Clive’s new firm, the
Cambridge Computer Company. De-
spite missing out on university himself,
Sinclair has long found Cambridge a

congenial site for his high technology
companies.

He also has plans to tackle the highest
levels of computer technology via a new
firm called Anamartic, which aims to
develop a wafer scale electronic memory.
British banks have put several million
pounds sterling behind the company,
which proposes to produce a practical
memory on which the chips used are
connected on a single wafer, allowing
more speed and scope than existing
memory devices.

Same Sinclair thinking

Sinclair must hope that Anamartic will

live up to its name, which comes from
the Greek for “faultless”. His idea of
establishing a British presence in wafer
scale memory technology is undoubtedly
ambitious.. The technology is one that
other British manufacturers would prob-
ably choose to import from Japan or
elsewhere, but he did run Europe’s
largest calculator manufacturer at a time
when that area was thought to be beyond
the scope of Britain’s own electronics in-

Anamartic bears clear signs of tra-
ditional Sinclair thinking. If the tech-
nology developed at his Metalab devel-
opment centre in Cambridge works, it
will provide a breakthrough in cost and
simplicity, which will allow computers to

be yet cheaper, smaller and more
available.

The first market Anamartic wants to
tackle is business automation, including
manufacturers of office workstations.
But anyone who knows Sir Clive’s taste
for taking technology to consumer
markets cannot doubt that this will be
his next stop if it proves a success.

THE RISE AND
RISE OF THE MICRO

by C H Freeman

California’s ‘Silicon Valley’, that conglomerate of many semiconductor and electronic equipment
manufacturers, has its roots in ‘the Fairchild takeover’. Fairchild Semiconductor had been bought out in
1959 by Fairchild Camera and Instrument. With the takeover there came the implementation of new,
rigid management structures. The managers of the old style Fairchild found themselves the middle-
management of the ‘new’ company, and became dissatisfied and disillusioned. Resignations began.
Those who resigned began to set up their own semiconductor manufacturing companies, companies
such as the (now) familiar Signetics, Intersil et al emerged. In total, about 50 IC companies have their
roots in Fairchild.

Birth of the
microprocessor

It was in 1968 that Robert Noyce, general
manager at Fairchild, left the company
to co-found Intel. Noyce had been one
of the founding fathers of Fairchild
Semiconductor in 1957. He had been in-
timately involved with semiconductor
technology for most of his professional
life and had acquired a very considerable
expertise in that field. It was Noyce who
pioneered research and development of
the monolithic integrated circuit at Fair-
child Semiconductor, quite literally from
sketchbook to production line in the
early 1960s. Such was his reputation that
venture capital flooded into the newly-
founded Intel with Noyce’s announce-
ment that the company intended to
specialize in memory chips; very much a
growth sector in the IC market of the
day.

During the next two years, Intel’s repu-
tation grew, along with the sophisti-
cation of its products. By 1970 Intel had
introduced a 256 bit RAM, stimulating

in earnest the switch from magnetic core
to IC main memory. It was in the sum-
mer of 1969, however, that Intel were ap-
proached by Busicom, a now-defunct
Japanese calculator manufacturer, to de-
velop a set of chips for a new line of pro-
grammable electronic calculators. Intel
had recently announced the development
of new IC manufacturing techniques for
making 2,000 transistor chips and
Busicom hoped Intel would be able to
make even more sophisticated devices.

Fig. I. The IBM 7351 Computer .Photograph
courtesy of PPM Instrumentation.

The Busicom engineers had prepared
preliminary designs which called for 12
chips, performing logic and memory
functions, in each calculator. By varying
the 'program' held in the ROM chips, a
whole line of calculators with different
capabilities and functions could be pro-
duced. Intel assigned Marcian E. Hoff,
Jr to the project. Hoff studied the
Busicom designs carefully and con-
cluded that, whilst technically feasible,
they would be too complicated to pro-
duce in a cost-effective manner; there
were simply too many chips per device.
The more Hoff considered the problem,
the more he became convinced that one
particular route led to the solution of the
problem. Along that route lay the con-
cept of a general-purpose logic chip
which, like the central processor of a
computer, could perform any logical
task. Such a micro-sized processor — a
microprocessor — would be program-
mable, acting on instructions and data
held in RAM and ROM.

The advantages were clear. Busicom’s
calculators could now be re-designed

india februaiv *988 2.31

into cost-effective 4 chip (microproces-
sor, ROM, RAM and an I/O interface
chip) machines, rather than the original
12 chip uneconomic devices initially
mooted. As Hoff critically examined
these proposals, he began to see the
promises such a device held. A reduced
chip count meant fewer interconnections
between discrete IC’s, a simpler, more
flexible and powerful design specifi-
cation could be issued. The most ex-
citing prospect, however, lay in the
device’s programmability, and program-
mability meant versatility. This scheme
was accepted by Busicom and in late
1970 the first microprocessor, the Intel
4004, began rolling off the production
lines.

The 4004, as its title suggested, was a 4-
bit microprocessor. The 4-bit architec-
ture was, at that time, state-of-the-art
stuff. A wider data bus was not
technically possible at the time. This
mattered little, however, as the 4-bit ar-
chitecture. was perfectly suited to pro-
cessing single decimal digits. The other 3
chips were similarly of limited capa-
bility. The ROM, containing the
calculator program, had a capacity of
just 2k bits, whilst the RAM held 320
bits. The calculators were duly produced
by Busicom, using the 4 Intel chips and
the few months after the introduction of
the calculators onto the market saw the
chips prove themselves as reliable, flex-
ible, cost-effective pieces of hardware.
The 4004 had been developed exclusively
for Busicom and Intel had no marketing
rights to the device whatever. The
economics of the calculator business in
the summer of 1971 saw Busicom ask In-
tel for a price cut on its chips. In ex-
change for this cut, Intel were to receive
full marketing rights to the 4004 and its
associated support chips. With the
marketing rights to the chips firmly es-
tablished, Intel launched the chips onto
the public market in November 1971.
The advertising hype used to promote
the ’new’ MCS-4 family belies the uncer-
tainty behind the launch. Intel’s
marketing division really did have a
'fingers crossed’ attitude towards sales.
No one was really certain if the public
would buy the devices. The first advert
says much about the type of application
envisaged for the MCS-4 family. To
quote from the advertisement:

“MCS-4 systems provide complete com-
puting and control functions for test
systems, data terminals, measuring
systems, numeric control systems and
process control systems. . .MCS-4
systems interface easily with switches,
keyboards, displays, printers,
readers. . . ”

Orders were slow. The initial sales levels
attained by the 4004 were not high
enough to justify further microprocessor
development. It was most unlikely that
the next Intel microprocessor, the 8008,

2

Fig. 2. Transferring schematics to PCB
layouts by computer. Photograph courtesy of
Betronex Ltd.

would have reached research stage had it
not been for a CRT research project
undertaken by the Display Terminal Cor-
poration (as they were then known).
Texas Instruments and Intel had entered
into an agreement with DTC, the object
of which was to produce a monolithic
processor capable of controlling a CRT
terminal.

A few months after drafting the agree-
ment, Texas Instruments pulled out,
leaving Intel to continue development
alone. Intel came up with the goods:
almost. The chip worked well, but
slowly. Too slowly for its intended appli-
cation as a CRT controller and the Intel-
DTC contract was dissolved. This left In-
tel with an (apparently) unsaleable mi-
croprocessor on their hands; what were
they to do? Almost unbelievably, the
8008 microprocessor was released on
general sale in 1972 on the assumption
that it would assist sales of memory
chips! The Intel microprocessor research
team were duly disbanded and that, it
was thought, was that. However, it was

Fig. 3. Elektor Electronics’ Junior Computer
announced in May 1986 was built by
thousands of computer enthusiasts all over
the world.

far from being ’that’. To everyones sur-
prise, the 8008 was a big-seller and,
almost overnight, microprocessors
began to be perceived in a new light: as
a growth market with attendant impli-
cations of large capital returns. The
changes the 8008 imbued into the semi-
conductor manufacturers manifested
themselves in the explosion of ’me
too’ microprocessor manufacturers:
Rockwell, Signetics, Motorola etc. all
announced their intentions to market
their own microprocessors. Intel, realiz-
ing they were on to a good thing, hastily
re-assembled their design team and set
them to work on the next processor, the
8080. The commercial desirability of
software compatibility had long been
realized within the industry and the 8080
was designed to be software compatible
with the 8008. A year after the introduc-
tion of the 8008 the 8080 was launched
to an even more enthusiastic reception.
Competitors set to design and market
their own processor chips in direct com-
petition with the 8080. It is significant
that the next three years saw the in-
troduction of chips now' regarded as in-
dustry standards: the 6502, the 6800, the
Z80, the TMS1000 series etc.

By 1976 the microprocessor had estab-
lished itself in many applications,
proving itself a reliable, efficient, cost-
effective component. The semiconduc-
tor manufacturers had come to accept
them as a necessary and lucrative part of
a manufacturing operation. The micro-
processor was here to stay!

Enter the minicomputer

The late 1950s had seen the computer
evolve into a useful, if difficult and ex-
pensive, tool. The computer’s place in
the scheme of things seemed clear; scien-
tific research, defence depts., large
automated accounts depts., educational
establishments— they all found comfort-
able uses for their new-found computer
power. The advent of time-sharing
systems saw an even greater degree of
computer penetration. Aside from the
obvious advantages of allowing many
users within a single company or organ-
isation to simultaneously ’use’ the
organisations computer, a new use for
time-sharing systems was spotted. Com-
mercial time-sharing services began to
appear. Customers of the service would
use remote terminals to communicate
with the time-sharing computer via
’phone lines, buying time on the
machine on a minute-by-minute basis.
By 1960 time-sharing was very much a
growth sector of the computer market
and was being confidently hailed as the
’way of the future’.

The big machines required to run any
sort of computing operation were still
costly, however: they were costly in terms
of full-time operations staff needed,

costly in terms of physical space and
power required and, above all, costly in
terms of initial outlay. In the early 1960s
a new need for a breed of smaller and
relatively less complex (and hence rela-
tively inexpensive) machines was begin-
ning to be identified. Thus it was that in
1963 the American computer manufac-
turers Digital introduced the PDP-8. In
comparison with the mainframes of the
day it was a limited machine. The PDP-8
ran just one program at a time, pro-
cessed data in 12 bit words and had just
4K words of memory, but the advantages
the machine offered were obvious. The
PDP-8 was, physically, about the size of
a large domestic freezer, did not require
a legion of trained support staff and,
most importantly, cost just $18,000; a
fraction of the price of then available
main-frame computers. The PDP-8 sold
extremely well, penetrating both old and
new markets. Scientists, defence and
educational establishments, industrial
plants and financial' institutions all
welcomed the new minicomputer with
open arms. Every drop in the price of the
PDP-8 captured new customers, hitherto
unable to afford their very own com-
puter power.

By the end of the ’60s a firm dichotomy
had been established, with the main-
frame and the minicomputer occupying
two different and well-defined market
sectors. A measure of the impact of the
mini can be gained from the fact that by
1971 at least 70 American firms were
manufacturing them, with estimated
sales running into many thousands of
millions of dollars. The minicomputer
had arrived.

Birth of the microcomputer

The development of the first microcom-
puter is a story that is, economically
speaking, quite different from the devel-
opment of the minicomputer.

The first serious, documented research
into the possibility of producing a home
(or personal) computer occured at
Digital (again) in the early ’70s. It is
most probable that other companies in-
vestigated the personal computer idea
also, but the Digital investigation re-
mains the most documented experience.
David Ahl joined Digital in 1969 as a
market researcher. Ah engineer by trade,
Ahl became involved in marketing the
company’s minicomputers to schools,
colleges and other ’small’ institutions.
In addition to the ’regular’ type of
orders for machines from institutions
and groups, Ahl occasionally received
orders from individuals for a Digital
mini. This set a train of thoughts in mo-
tion in Ahl’s mind as he began to
wonder if there would be a market for a
simple, low cost personal computer. In
1973 Ahl moved to Digital’s research and
development wing, investigating (among

other things) new markets for certain
products. A separate engineering study
was being simultaneously pursued in the
R&D group, the object being to manu-
facture a small business computer. Ahl
became interested in the project and
began to eagerly investigate the
marketing side whilst the engineering
chaps worked on perfecting two proto-
types, which were eventually persuaded
into working order in early 1974. Details
of these prototypes remain sketchy, but
one was a ’modified computer terminal’,
the circuitry being constructed from
discrete logic and memory chips; no
microprocessors were involved in its
construction. The second prototype,
however, was quite a different prop-
osition. More radical and daring in its
design philosophy, this machine was a
portable self-contained computer, about
the size of a thick attache case, compris-
ing a monitor, keyboard, floppy disc
drive and the main processor itself.
Microprocessors were incorporated in
the design of this second machine. In
May 1974 a management committee
convened at Digital’s headquarters to
discuss the future of the project. The
technical half of the committee were,
understandably, enthusiastic. The
machines worked well (aside from
teething troubles with the floppy disc
drive) and could be manufactured econ-
omically. The sales department, making
up the other half of the committee, were
far less enthusiastic. Why, it was
reasoned, should educational depart-
ments buy the machines? A time-sharing
computer would be much more cost-
effective. Why should the ordinary
householder buy them? There were no
conceivable ’home' applications. In the
end the salesmen won the day and the
project was dropped. The computer in-
dustry, it seemed, was indifferent to the
prospect of a personal computer. It mat-
tered not that such machines were
technically feasible and, indeed, could
be manufactured and sold on a cost-
effective basis: no market for such
devices could be seen. The early develop-
ment of the microcomputer thus rested
squarely on the shoulders of the hob-

Fig. 4. A stimulating application of the
microcomputer: Mailbox Chess, a British
Telecom (Prestei) service.

byists — people whose primary motiva-
tion was technical rather than financial.
The introduction of the Intel 8008 mi-
croprocessor, the 8 bit successor to the
4004, stimulated the interest of one
Jonathan A. Titus, an electronics hob-
byist. Studying the specifications of the
8008, Titus realized that the chip was
powerful enough to run a microcom-
puter and set about designing his own
such machine. By Autumn 1973 Titus
had a working prototype and, wishing to
share the design with other members of
the hobbyist fraternity, contacted
various amateur electronics periodicals,
enquiring whether they would care to
publish his design. The offer was finally
taken up by the (American) ’Radio-
Electronics’ magazine — a publication
firmly aimed at the hobbyist. The July
1974 edition carried a constructional
article on the Mark-8 (as it had been
christened) together with a parts source
guide and approximate costing. The
Mark-8 was a strictly limited machine.
Its innards contained 256 bytes of RAM
expandable up to 16K and no ROM:
Titus would have had to pay significant
amounts to Intel to produce ROMs to
his specification. Input/output con-
sisted of lamp and switch technology.
Despite its limitations, interest in the
Mark-8, and subsequent sales of the
bare circuit boards, far exceeded expec-
tations. Yet, in spite of this, Titus did not
even consider forming a computer
company, regarding the Mark-8 as more
of an ’educational’ project than a com-
mercial proposition.

The interest stimulated by the introduc-
tion of the Mark-8, coupled with Intel’s
newly released 8080 microprocessor,
prompted a small electronics company
called M1TS to introduce the Altair
8800. The 8800 was designed for the
American hobbyist electronics publi-
cation ’Popular Electronics’. The pro-
ject was interned to be printed as a series
of constructional articles in the maga-
zine and, like the Mark-8, was aimed
firmly at the hobbyist market. The basic
machine was designed with expand-
ability very much in mind. The 256 byte
memory supplied with the basic kit
could be expanded right up to the maxi-
mum 64k bytes possible with the 8080
processor by means of slot-in memory
boards. Other peripherals, such as a
CRT terminal, printer, alphanumeric
keyboard and papertape reader were also
on the drawing board. Combining
peripherals, memory expansion boards
and the 8800 itself could produce, for
the first time, a really useful, relatively
low cost system. Such a system formed
the first — the very first — fully fledged
personal computer on the market. The
8800 made its first appearance in the
January 1975 edition of Popular Elec-
tronics and was an instant sensation.
The machine was offered to readers of
the magazine at a cost of $650 fully

assembled and $395 in kit form. Such
prices for a machine of such high speci-
fication were unheard of and orders
poured into MITS, who had great diffi-
culty in fulfilling customers orders.
Customers experienced delays of up to 6
months (sounds familiar!) before
delivery of their computers, and the
promised peripherals did not materialize
until early 1976: one year later (sounds
even more familiar!). The popularity of
the hardware stimulated an 8800 based
software market. A BASIC interpreter
was written for the machine and
marketed, with success, by MITS. The
8800 can, justifiably, be said to have
been the first ’real’ personal computer,
and its success helped fuel the belief that
a significant market for personal com-
puters might exist after all.

As interest in the Altair 8800 grew, the
first ’computer clubs’-, run by and for
amateur enthusiasts, were being formed
all over America. It was at one such club
in California’s Silicon Valley that
Stephen Wozniak, a young self-taught
computer engineer, became interested in
the possibility of fabricating a
’homebrew’ computer. Wozniak exam-
ined the specifications of available
microprocessors and concluded that the
comprehensive instruction set of the
newly-introduced 6502 would serve his
purposes best. Wozniak set to work,
writing a BASIC interpreter before going
on to design and construct the hardware.
The finished article was a single board
computer with 4K of onboard RAM and
circuitry to enable direct connection to a
monitor. Wozniaks friend, Stephen
Jobs, saw a market for the computer and
tried to persuade Wozniak to enter into
a business partnership with him. Woz-

niak, meanwhile, had demonstrated the
board to a gathering of other homebrew
enthusiasts and the reception had been
so enthusiastic that he had approached
Hewlett-Packard, his employer, and tried
to interest them in manufacturing his
computer. Hewlett-Packard refused,
however, doubting the existence of a suf-
ficiently large market. Jobs thought dif-
ferently. He approached several potential
buyers and eventually signed a contract
for 100 boards at $500 each. Jobs and
Wozniak went into partnership and the
Apple Computer Company was born.
The Apple I (as it has subsequently be-
come known) had no keyboard, ter-
minals, disc drives etc. and was clearly
targeted at the hobbyist ’tinkcrcrs’
market. In total, Apple sold around 175
boards at $500 apiece, leading Jobs and
Wozniak to consider the project a great
success. So much of a success, in fact,
that the pair went to work on a second
machine, Wozniak taking care of techni-
cal development and Jobs looking after
the business side of Apple. In October
1976 Jobs received significant help from
a venture capitalist who decided that the
newly-developed Apple II was just right
for the mass market. With a firm
business plan and significant financial
backing, the Apple II went on sale in
1977. Sales of the machine rapidly grew,
boosted by a hard advertising campaign,
and by the end of 1977 the Apple Com-
puter Company’s annual sales were
estimated to be in the region of
$775 000. Next year’s sales were even
better and Apple were named as the
fastest-growing company in American
history. The software base for the
machine grew rapidly, helping to fuel
sales to further heights. The personal

computer had arrived.

The establishment of a mass-market for
the personal computer shook other
manufacturers out of their complacency
and, in time-honoured fashion, began to
jump onto the bandwagon. From here
on, market forces begin to take control
and the story of further developments in
the field of personal computing becomes
less concerned with technical innovation
and more concerned with hard
economics. One further development,
however, merits a brief glance. In 1981
IBM entered the fray with its PC, the
machine becoming an instant success in
its target area (the so-called middle-
business market). With IBM sales of the
PC well established, the familiar
scenario of ’mc-too’ PC clone manufac-
turers, introducing machines in direct
competition with IBM appears again.
The resulting economic morass is a
startlingly familiar cocktail of market
forces.

Conclusion

With a well established trichotomy of
mainframc/mini/micro the future may
well see significant technical develop-
ments, but in a capitalistic economy,
where market forces reign supreme, it
will be the businessmen, and Jiot the
scientist, who will decide what sees the
light of day. With the benefit of hind-
sight it is difficult to imagine that in the
future technical constraints, rather than
economic ones, will form the sword of
Damocles over continued development.
Time alone will tell.

The level of water in a water tank water detector

can be measured in various ways,
some of which can of course be more
complicated than others. The circuit
published here lights a LED when-
ever the level of water drops below
the electrodes. With a high water
level,, the FET hardly conducts or
not at all, because the gate is then
connected to earth and there is no
voltage difference between gate and
source. When the water level drops,
the gate/source connection is inter-
rupted. The gate is then at a positive
potential by means of the 820 k
resistor, thereby causing the FET to
conduct. The LED will now light. If
the reverse operation is needed, that
is the LED lights when the electrodes
are short circuited by the water, just
connect the 'earthy' electrode in the
circuit to the positive and wire R2
between gate and source.

M

2.34,

TEST & MEASURING EQUIPMENT

Julian Nolan’s discourse on dual-trace oscilloscopes is continued this month with reviews of the Gould
OS300 and the Grundig MO20.

Part 1: dual-trace oscilloscopes (B)

Gould OS300

Gould have a high reputation for the
quality and robustness of their
oscilloscopes and the OS300 follows in
this trend, in spite of it being the
cheapest scope manufactured by Gould.
The Gould range of oscilloscopes com-
prises real-time and digital storage units
at prices well over £18,000 for a complete
digital acquisition system. Gould also
manufacture a wide range of other
specialized electronic products such as
digital chart recorders, logic analysers
and emulators, etc. Gould also have
available a wide range of accesories for
its oscilloscopes, such as protective car-
rying cases, rack mounting kits and of,
course, probes.

The OS300 is delivered in an easily ident-
ifiable blue/white box and is well pro-
tected by the standard polystyrene or
similar cutouts which surround it. The
scope is supplied with a comprehensive
manual and, I believe, a mains lead (the
review model was not). Since the mains
connection is made via a standard IEC
style socket, there is no problem should
the lead supplied be lost or too short.
The scope itself is quite small as far as
the height and the width are concerned,
although the depth is a relatively long
460 mm. It weighs only 5.8 kg. The
OS300 is equipped with a multiposition
aluminium stand (with plastic handgrip
at centre). However, a price is paid for
the multiple positions of the handle in
that the sides have to be pulled out to
facilitate movement. This is not as easy
as it sounds when the stand needs to.be
rotated by 90° or more, owing to the in-
flexible nature of the aluminium.
Voltage selection can be made for a
100 V, 120 V, 220 V or 240 V line by 2
slider switches provided on the back
panel, on which Z modulation and,
unusually, ramp-out sockets as well as
ground are provided.

The front panel is clearly laid out. All
trigger functions are controlled by seven
push-button switches, though the fact
that two switches have to be latched sim-
ultaneously to enable Ext or TV trigger-
ing could be initially inconvenient.
Alternate or chopped modes and TV line
or frame triggering is automatically
selected by the timebase switch, being
dependent upon the sweep speed selec-
ted. Although bringing about a slight
decrease in versatility when the scope is
used in these modes, this arrangement is,
however, justified by the resulting in-

Fig. 7. Gould OS300 oscilloscope

crease in ease of use of the scope.
Curiously, the on/off switch is com-
bined with the intensity control, which is
slightly inconvenient as the intensity
control is reset during each on/off oper-
ation. Due to the quick heat cathode of
the CRT, the intensity control can be set
any time after the 10 seconds or so
warm-up time of the CRT, which limits
the inconvenience. The CRT bezel is not
equipped with camera mounting facility,
but Gould point out in the manual that
this problem can be solved by holding a
suitable camera against the tube face.
As can be seen from the specification,
the maximum Y amplifier sensitivity is
2 mV/div, which is effective across the
20 MHz bandwidth ( - 3 dB). This ex-
tends upto 10 V/div calibrated or
25 V/div uncalibrated. The wide at-
tenuation range allows a good range of
input signals to be accurately measured.
The good sensitivity permits a xlO
probe to be used with low voltage high
impedance sources for the minimum
minimum loading. On the whole, the Y
amplifiers performed well, having a
good response up their specified band-
width and risetime but it was noticeable
that the response did dip sharply after
25 MHz. As on most modern scopes in
this price range, the input stage of the
pre-preamplifiers is formed by FETs,
here in a source follower configuration.
The use of FETs obviously helps provide
the high input impedance which is a vital
quality of any oscilloscope. In common
with those on some other oscilloscopes
in this range, the Y amplifiers exhibit a
small amount of drift (approximately
Zi cm), during the ’warm-up’ period
thus making readjustment of the
horizontal trace position necessary, after
the initial 10 minutes of operation. An
accuarate 1 V peak to peak (± 2%),

1 kHz square wave for x 10 probe cali-

bration is provided.

The trigger performance of the OS300 is
good, with triggering possible on a very
large range of signals without major ad-
justment of the trigger threshold. Com-
plex signal patterns are also handled
without too many problems, although a
digitized sine wave with its third MSB bit
missing did require some fine adjust-
ment of the triggering level to obtain a
stable trace. High frequency triggering
was also good with triggering on wave-
forms with frequencies very considerably
higher then the 20 MHz Y amplifier
bandwidth. The triggering selectors con-
sist of a bank of 6 push-button switches,
some of which have dual functions, e.g.,
two being depressed for one function.
This leads to a slight decrease in flexi-
bilty in some modes such as external
triggering or TV sync, although it has to
be said that this is minimal. Line or
frame sync, is selected in the TV mode
by the timebase switch and, although in
theory this may appear again to be
somewhat inflexible, in practice this is
not so, as the automatic timebase selec-
tion is quicker and easier than manual
switching. This also applies to chopped
or alternate switching, which is also
selected by the timebase speed. The
chopping frequency is a notable
500 kHz, enabling relatively high fre-
quency signals to be measured more ac-
curately in this mode rather than in the
alternate mode where, because of its
nature, phase comparisons of waveforms
can be inaccurate.

Sensitivity is good across the entire
bandwidth ranging from about 1.5 mm
at 2 MHz to 4 mm at 20 MHz. The
OS300 still maintained a reasonable sen-
sitivity at frequencies above 20 MHz.
Alternate channel triggering of the
timebase is not available and because of
this, synchronization of the two input
waveforms is necessary in dual trace
mode in order to obtain a stable trace.
Trigger coupling is also limited by the
absence of low frequency or high fre-
quency filters which are usually provided
in scopes in this price range. Where these
would normally be necessary, typically
where the required trigger signal has a
very high content of LF for an HF wave
or HF for an LF wave modulation, the
problem can usually be got around by
critical adjustment of the triggering
threshold. This is, however, not always
the case and consequently necessitates
the use of the external trigger facility.

The remaining trigger functions are
fairly standard ones such as AC or DC
coupling etc.

The timebase speed selection switch is to
the right of the scope and has a total of
18 speeds ranging from 0.2 s/div to
0.5 /is/div. This can be expanded to
50 ns/div by a x 10 switch, making
viewing of fast risetime waveforms easy.
My main criticism of the OS300 is its
below average focusing, which is evident
at most time base speeds, and results in
a lack of sharpness and definition that
some other scopes with 2 kV tubes
possess. This is naturally most
noticeable on the x 10 timebase speeds.
The OS300’s lack of automatic focusing
docs not help in the matter, so that a
slight readjustment of the focusing con-
trol is necessary when changing over a
wide range of timebase speeds. The
brightness on the scope is admitedly very
good for a 2 kV accelerating potential
tube, but even at low intensity levels the
trace still insists on appearing slightly
defocused— both the focus and
astigmatism controls level and certainly
the brightness of the Mullard quick-
heat-cathode CRT is very good. This
quick heat cathode, as used in many
modern TV’s, enables the scope to be-
come operational very much quicker
(under 10 secs) than other instruments in
its class (typically 30 secs).

The Gould OS300 is also fitted with Z
modulation and, unusually, a ramp out-
put. This output, although being only
3.5 V in amplitude (peak) can easily be
amplified if necessary and used to drive
such things as, for example, a VCO
which in conjunction with a swept LO
detector could be used to provide a spec-
trum analyser type display. This perhaps
slightly ambitious application is one of
many for which the ramp output can be
used, as it ensures the synchronization
of any external device with the scope
thus producing a locked display. It is sur-
prising that this helpful feature is not in-
cluded in more scopes.

Internal construction is centred around a
single large epoxy glass PCB which
houses the vast majority of the circuitry,
while a smaller vertical PCB contains the
Y amplifiers and some timebase compo-
nents. The Y amplifiers themselves are
fully screened. Parts of these amplifiers
are constructed from IC’s which are
housed in TO88 metal packages for
further noise immunization. Overall in-
ternal construction is very neat: there are
very few interconnecting wires, and,
despite the fact that both PCBs are
single-sided, also very few wire links.
The PCBs and other large components
such as the tube and the high-quality
mains transformer are all mounted on a
robust steel subframe onto which the
sheet metal outer housing is fitted. A
large part of the CRT is shrouded to
guard against X-ray radiation etc. Most
of the components used in the OS300 are

2.36 1988

Fig. 8. Close-up of OS300 controls

of a high standard, but I have my doubts
about the potentiometers used for some
of the continuously variable front panel
controls, which have a somewhat uneven
movement, and upon inspection prove to
be little more than presets. There is,
however, no reason why these should
prove to be unreliable. On the subject of
presets, all are clearly labelled on the silk
screened PCB and all are easily access-
ible. Components used in the scope
come from a wide range of manufac-
turers ranging from National Semicon-
ductor for some of the ICs to Magnetic
Shields Ltd for the CRT shield.

The manual for the Gould is particularly
good, totalling 26 pages plus the A3 size
circuit diagrams. It covers the OS300 in
some detail including sections on oper-
ation, maintenance and calibration.
There is also a detailed circuit descrip-
tion so fault finding, etc., should present
no problems. In addition, there are a
large number of diagrams, both circuit
and mechanical, which should be of help
in many circumstances.

Conclusion

The Gould OS300 has its good and bad
points. The good points include the ex-
cellent quality of construction, its rug-
gedness, the quick-heat cathode of the
CRT, ease of operation, and the several
time saving features. It is also worth
bearing in mind that the instrument
undergoes a large number of quality
control procedures during and after con-
struction such as the 36-hour soak test
which is carried out on the completed in-
strument. While these arc in no way
unique to Gould, they do appear to be

Table 5. Specification

ELECTRICAL CHARACTERISTICS
Designed for IEC 348 Cat. 1 .

Line voltage: - 100,120,220,240 VAC
± 10%: externally adjustable. Power
50 Watts. Line frequency 45 . . . 440 Hz.

MECHANICAL CONSTRUCTION
Dimensions: - W 305 mm. H 140 mm,

D 460 mm

Housing: - Sheet steel
Weight: — approx. 5.8 kg

Y AMPLIFIER ETC.

Coupling AC, Gnd or DC on both

Operating modes: -
CH 1 alone
CH 2 alone or inverted
CH 2 + CH 1

Alternate or chopped (500 kHz)

CH 1/CH 2. switching automatic.
Frequency range (full deflection)

. 0 ... 20 MHz ( - 3 dBl (DC coupled)
Risetime < 17.5 nsec.

Deflection factor 1 2 steps;

2 mWdiv. . . 10 V/div ± 3% (Min
25 V/div, fine control fully anti cw).

Input coupling AC, DC or Gnd.

Input impedance 1 MQ/28 pF; Max input
voltage 400 V DC or pk AC. Input
protected.

Input impedance 1 MQ/28 ;,f : Max input
voltage 400 V DC or pk AC. Input
protected.

X-Y MODE

CH 1 X-axis, CH 2 Y-axis. Bandwidth DC-
1 MHz (-3 dB). Less than 3° phase shift
at 50 kHz.

TIMEBASE

Deflection factor 18 stops;

0.5 «sec. . .0.2 sec/div ± 3% in
1/2/5 sequence. Min speed 0.5 sec/div
(fine control fully anti cw).

Expansion x 10, extends max. timebase
speed to 50 nsec/div. Expansion error
s ± 2% extra.

TRIGGERING

Trigger modes: - Auto (bright line).
Normal, TV (active line or frame sync,
automatically selected from timebase

Trigger coupling: — AC, DC.

Trigger sources: - CH 1, CH 2, Ext.
Triggering slope: - positive or negative,
switchable.

Trigger sensitivity: — Internal (DC) 5 mm
i to 20 MHz. External (DC) 400 mV to
j 20 MHz; All modes.

MISCELLANEOUS
CRT -make Mullard, measuri
100 x 80 mm, accelerating
2 kV, quick-heat cathode. F

Compensation signal for divider probe,
amplitude approx. 1 V ± 2%; frequency

Z modulation— input by 4 mm socket; 2 V
visible mod. + 40 V complete blanking.
Ramp out. . . + 3.5 V peak from 5 kQ.
Covered by 2 year warranty.

Fig. 9. Internal

a of OS300

particularly stringent in this case, being
backed up by the instrument’s two year
guarantee. As to the bad points, my
main criticisms are the focusing of the
CRT and the lack of alternate triggering.
Summing up.'the Gould OS300 oscillo-
scope is probably best suited to environ-
ments where its high grade of construc-
tion will be valued, such as educational
establishments or servicing departments,
and its few shortcomings will not be no-

The Gould OS300 was supplied by
Gould Electronics Ltd., Instrument
Systems, Roebuck Road, Hainault, Il-
ford Essex IG6 3UE. It retails at £342 +
VAT.

No other oscilloscopes under £1,000 are
manufactured by Gould.

Grundig MO20

The German company of Grundig has
long been renowned for its high quality
consumer products, and, perhaps not
surprisingly, the Grundig MO20 largely
keeps to this mould. The MO20 is one of
3 oscilloscopes manufactured by Grun-
dig aimed at the ’under £1000 market’:
the others being the M022 and the
M053. The MO20 is the cheapest in this
range retailing at £365 +VAT. This
price, in common with many other
foreign products, has been adversly af-
fected by the exchange rates: it was £66
pounds lower at £299 in August of 1986.
The Grundig MO20 arrived well packed
in a cardboard box, substantially larger
than the instrument itself, which was en-
cased in a number of polystyrene
cutouts. On removal, the Grundig
turned out to be fitted with a two pin
continental plug, which entails either
using an .adaptor or, more preferably,
refitting with a 13 amp plug. The lead
feeds straight into the scope and is not
fitted with a socket. This is unfortunate
because it is of only average length and
in some situations a longer lead may be
required. The stand is moulded in a
tough plastic and is well up to sup-
porting the heavier than average
(8 '/z kg) MO20, but it has only one pos-
ition. The front panel is clearly laid out,
although the vertical mode switches are
initially confusing. However, the method
of using two switches for the vertical
mode pays dividends when in the X-Y
mode, which due to the design is ex-
tremely versatile. Upon inspection of the
manual, it would be all too easy to de-
scribe the Grundig as a run-of-the-mill
oscilloscope, with perhaps the notable
feature in this price range of having
automatic peak value triggering, but this
is not the case. A summary of the
specifications is provided in Table 7.

As can be seen, the maximum calibrated

10

Fig. 10. The Grundig MO20 oscilloscope

Y amplifier sensitivity is 5 mV /cm
although, unusually, the fine adjustment
control calibrated to give a maximum
gain of x 2.5 actually increases the sensi-
tivity, resulting in a sensitivity of 2 mV.
I found this to be accurately calibrated,
although the system of using the fine
control to increase sensitivity was at first
off-putting as on most scopes it
decreases sensitivity. The Y amps per-
formed well handling an almost full
height (6 cm) sine wave with little
degradation in height over the full
20 MHz bandwidth of the instrument.
When fed with a fast rise time pulse, the

Y amps responded within the specified
limit of 17.5 ns. It was pleasing to note
that the range extends to 20 V/cm,
which permits the measurement of rela-
tively high voltages in the dual trace
mode without any overlapping of traces,

Fig. 11. Close-up of MO20 controls

while using a x 10 probe. Both channels
are invertable, making for easier subtrac-
ting, as no swapping of connectors is re-
quired for changing the subtraction
waveform. One minor but annoying
point was that both Y amplifiers on the
review model exhibited a small amount
of drift (about Zi cm) over the fist 10
minutes of operation, necessitating an
adjustment in their trace position. This
does not, however, affect normal
measurements outside the ’warm up’
time during which I found both channels
very stable. A 1 V peak-to-peak 1 kHz
probe calibration waveform is provided
having a rise time of 5 /js.

The MO20 boasts an automatic peak
value trigger, which enables the scope to
lock on to virtually any repetitive wave-
form of above 7 mm in amplitude, with-
out adjustment of the triggering
threshold. This has advantages in that
alteration of the manual trigger value is
not necessary for variations in waveform
amplitude etc. thus making operation
easier. Triggering on the MO20 is com-
prehensive: auto (peak triggering not
bright line); normal, i.e. no bright line;
or auto trigger and TV frame and line
modes. On top of this, it also features
AC, DC, LF (fs^8 kHz) and HF
(for^.10 kHz) coupling. The source for
these functions can be selected from
either CHI, CH2, LINE or an external
source. A notable exception here is that
in common with most other
oscilloscopes in this price range, alter-
nate triggering from CH.l and CH.2 is
not available. This means that when
both channels are being used both the
input wavforms must be synchronized to
a single source to enable both trees to be

Selection of the trigger mode is by a
number of lever operated switched.
These are not only easy and convenient
to operate, but also provide a clear indi-

cation of which function has been selec-
ted; they are also used for vertical mode
selection. I found the triggering facilities
of the Grundig very good, although a
slightly more sensitive auto trigger func-
tion would have saved adjustment of the
Y amps in a few cases. The scope trig-
gered successfully on a large variety of
waveforms ranging from NRZ pulses to
a heavily modulated sine wave at
20 MHz. In some cases, it was necessary
to switch into normal mode to obtain a
fully stable trace, but this was relatively
rare. The auto trigger function does
however take a few tenths of a second to
lock onto a signal, and during this time
the display is blanked, producing a
rather disconcerting flickering if the Y
amplifiers range is changed or the trig-
ger signal falls below the trigger
threshold. Despite this, the auto trigger
function performed very well, its
usefulness extended by the inclusion of a
LED indicator to show the trigger state
of the scope.

Time base speed selection is by means of
an 18-position switch and ranges from
0.2 scc/div to 0.5 //sec/div, a x 10 switch
increasing this to 50 ns/div. A fine ad-
justment control is capable of slowing
the trace down by a factor of up to 2Zi.
The focusing and brightness of the
VALVO CRT were very good even at
maximum sweep speeds, although there
was a small amount of dcfocusing when
the x 10 control was brought into oper-
ation, the intensity control being
suitably advanced to compensate for the
loss in brightness. Defocusing was also
observed when the intensity control was
at maximum. These problems are really
to be expected with a 2 kV tube as fitted
to virtually every scope under £500. The
maximum sweep speed ( x 10) proved to
be fast enough for observing signals at
the full 20 MHz bandwidth, the over-
shoot on a 10 MHz square wave being
clearly visible. The CRT is Clearly
marked with the appropriate gradua-
tions for rise time measurement, which
is made easy by the comprehensive trig-
gering facilities. There is no filter in
front of the CRT so that the colour of
the CRT appears true: greyish white with
a red graticule. Although Z modulation
is not fitted as standard I was surprised
to read in the manual that it can easily
be fitted by the addition of one BNC
socket connected to a clearly marked
point on the PCB. The bright level is
specified as <0.8 V and the dark level as
>2.0 V.

Overall the construction of the Grundig
MO20 appears to be excellent for an os-
cilloscope in its price range. The case is
solidly constructed from relatively heavy
gauge steel and, since the instrument
only consumes 35 Watts of power, no
ventilation slots are necessary. The front
and rear surrounds are cast metal: most
of the front panel is plastic, and the rear
is of sheet steel. The use of steel does, of

Fig. 12. Internal view of MO20

course, have its drawbacks in terms of
weight over the more frequently used
aluminium, but in my view the extra
strenght obtained compensates for this.
On removal of the top cover the internal
layout was found to be very neat, with a
minimum of interconnecting wires. The
majority of the components are
mounted on two fibreglass PCBs: one
large board houses the power supplies, Y
amplifiers, blanking circuit etc, and a
smaller board houses the ramp gener-
ator, triggering filters etc. Both boards
are single sided, the larger of the two
having a large number of wire links.
Connections between the boards arc
made by a thick form of ribbon cable.
Both boards are silk screened, and are
fitted with numerous test pins, all of
which are clearly labelled as to their
function and, where appropriate, ex-
pected voltage. The majority of the inte-
grated circuits are manufactured by
Texas Instruments, although several, in-
cluding the input amplifiers, are from
other manufacturers. The input'
amplifiers are soundly constructed and
fully shrouded with sheet metal, as is the
CRT. The attenuation range switches ap-
pear to be of a high quality construc-
tion, and have a positive action. One
minor point is that both the attenuation
switches and the timebase switch have no
end stops, so that it is all too easy to ac-
cidentally switch from the maximum
range to the minimum or vice versa. Per-
sonally, I didn't like this because of the
reason outlined above, but this a purely
personal view. The timebase switch,
which is based around a small vertically
mounted PCB, did not appear to be of
such quite good quality as the attenuator
switches, but was still acceptable. All
other components appeared to be of a
very high quality.

Servicing or calibration of the scope
should be no problem, as most of the
components are fairly standard,
although there are several thick film re-
sistor networks. The large number of
marked test points greatly eases fault-
finding and calibration. Unfortunately,
the manual only gives a brief list of the
internal presets and suitable calibration
procedures. None the less, these are
quite helpful and should be sufficient
for most users requirements. The

Table 7. Specification

ELECTRICAL CHARACTERISTICS: -
Protection class 1 .

Line voltage: - 1 10,220,240 VAC
± 10%: internally adjustable. Power
35 Watts. Line frequency 45 . . . 65 Hz.

MECHANICAL CONSTRUCTION
Dimensions: - W 375 mm, H 160 mm,

D 430 mm

Housing: — sheet steel
Weight: — approx. 8.5 kg

Y AMPLIFIER ETC.

Operating modes: —

CH 1 alone or inverted
CH 2 alone or inverted
CH 1 + CH 2

alternate or chopped (250 kHz) CH 1/CH 2
Frequency range 16 cm deflection)

0. . ,20 MHz (-3 dB).

Risetime < 1 7.5 nsec.

Deflection factor 12 steps:

5 mV/div ... 20 V/div ± 3%

(Max. 2 mV; fine control full cw).

Input coupling AC, DC or Gnd.

Input impedance 1 MQ/25 pF, Max input
voltage 400 V (peak including DC
voltage).

X-Y MODE

CH 1 X-axis, CH 2 Y-axis. Bandwidth
DC-1 MHz (— 3 dB). Less than 3° phase
shift at 50 kHz.

TIMEBASE

Deflection factor 0.5 nsec/
div. . .0.2 sec/div + 3% with
1/2/5 divisions.

Expansion x 10; extends max. timebase
speed to 50 nsec/div; expansion error
s ± 2% extra.

TRIGGERING

Trigger modes: - Auto with peak value
triggering from f> 10 Hz; Normal; Active

Trigger coupling: - AC, DC,

LF (fo=.8 kHz), HF (f„~10 kHz).

Trigger sources: - CH 1, CH 2, Line, Ext.
Triggering slope: — positive or negative,
switchable.

Triggering sensitivity: — Internal si cm
at 20 MHz, External s 500 mV at
20 MHz, Normal mode.

Trigger signal indicator: — green LED.

MISCELLANEOUS

CRT-make VALVO. measuring screen
1 00 x 80 mm; accelerating voltage 2 kV;
beam rotation by front panel adjustment.
Compensation signal for divider probe,
amplitude aprox. 1 Vpp, frequency 1 kHz.
Z modulation-possible by addition of BNC
socket, bright level S0.8 V, dark level
a: 2.0 V.

Covered by 1 year warranty.

Table 8-

CATEGORY

Unsatis-

factory

Good

Very

Good

Excellent

TRIGGER FACILITIES

x

TRIGGER PERFORMANCE

CRT BRIGHTNESS

CRT FOCUSING

Y AMP PERFORMANCE

INTERNAL CONSTRUCTION

EXTERNAL CONSTRUCTION

X

OVERALL SPECIFICATION

EASE OF USE

MANUAL

X/Y PERFORMANCE

X

manual itself is quite good, covering a
total of 16 pages (plus block and circuit
diagrams), of which »8 pages is in
German, the remainder being the
English translation. Overall, the A4 size
manual provided a good, if short, sum-
mary of the functions, set-up pro-
cedures, and applications of the MO20.
In addition, details on fitting the Z-
modulation socket, test characteristics
and detailed specifications were also in-
cluded. The circuit diagrams themselves
are very clearly laid out and easy to
follow, as is the block diagram. These
are provided on fairly large pieces of
paper which makes reference easy.

Conclusion

The Grundig MO20 is certainly worth
considering, its particular strengths ly-
ing in its construction and advanced
triggering facilities. The construction is
very good for an oscilloscope in this
price range and the two-year guarantee
offered with the scope reinforces this.
The triggering facilities are good, with
the particular bonus of automatic peak
value triggering, as well as comprehen-
sive filtering. It is a pity, however, that
alternate triggering is not available. The
CRT and drive circuitry are also very

good, producing a clear and bright trace,
even at the maximum deflection speed.
To sum up, the Grundig MO20 is cer-
tainly worth buying if you require a well-
built oscilloscope with a good range of
facilities.

The Grundig MO20 oscilloscope was
supplied by Electronic Brokers and
retails at £365 plus VAT. Electronic
Brokers arc at 140-146 Camden Street,
London NW1 OPB; telephone 01-267
7070.

Other Grundig scopes under
£1000

M022— As the MO20 plus automatic
timebase selection, triggerablc second
timebase, hold off control, Z mod. Cur-
rent price £499 + VAT.

M053— As the M022 plus alternating
second timebase, digital timebase dis-
play, 50 MHz bandwidth, delay time
multiplier. Current price £750 + VAT.

When developing photographs a
light meter can be a useful gadget.
Of course, the more luxurious the
better, but as the diagram here
shows, a simple circuit can also be
very effective. The circuit is con-
structed around a photodiode. In a
camera the diode generates enough
voltage to control a meter directly.
Often, however, there is not enough
light uqder the enlarger. That is why
a transistor has been added. Unfortu-
nately, this also puts an end to the
original design's simplicity, for now a
battery will also be necessary. One
solution is to connect the light meter
to an ordinary multimeter switched
to the resistance range. This is
indicated inside the dotted line.
Remember that the positive test lead
of most meters is the negative pole
for the resistance range!

Operation is as follows. The test
diode is connected between the
collector and base of the transistor.
The more light that falls on the
diode, the more the latter will con-
duct, thus providing the transistor
with more base drive current. The
collector current then increases
(almost) linearly and this is indicated
on the meter's dial.

M

cheap light meter

,2.39

Telecom 87: a preliminary report

Telecom is the quadrennial communi-
cations exposition organized by the In-
ternational Telecommunication Union
(ITU). It is the world’s largest, most
authoritative and comprehensive tele-
communications event, attracting exhibi-
tors and visitors from all over the world.
In the 4 years since Telecom 83, great
strides have been made by the telecom-
munication industry. This could not
have been reflected better than by the
main theme of Telecom 87: "The Com-
munications age: networks and services
lor a world of nations". The ITU, as
part of the United Nations, has a co-
ordinating, advisory and regulating
function as regards worldwide telecom-
munication. But "a world of nations" is
indicative of incompatibility, or, quite
literally, a world of difference, between
nations’ standards and preferences for
expanding existing telecommunications
networks. In spite of the modern means
available for global communication,
technology in each of the ITU’s 162
member countries is developing at its
own pace, often dictated by political and
economical factors. The ITU organizes
Telecom to enable national PTTs and
telecommunication companies from all
over the world to see the latest technical
achievements. This, hopefully, leads to
discussions on the ways in which these
can be integrated into existing telecom-
munication structures. Again, there is
the risk of incompatibility between exist-
ing and new systems, and this is exactly
what the ITU strives to eradicate or pre-
vent with the aid of a comprehensive ex-
position. In this respect, it is interesting
to quote the Secretary General of the
ITU, R. E. Butler: "The successive
Telecom exhibitions owe their success to
the fact that the ITU was in a position to
offer, within a neutral framework im-
mune to external influence, facilities for
meeting and interchange which were
soon recognized as a valuable service
rendered to our member countries as
well as to the operating organizations
and the telecommunications industry. ”

Telecom 87 was held in Geneva, the
headquarters of the ITU, from 19 to 27
October 1987. The venue was the Palais
des Expositions (Palexpo), which is
within walking distance of the Geneva
airport. The total number of exhibitors
was close to 900, and some 50,000
visitors were received representing
200,000 entries. The total net exhibition
surface was 53,500 m ! inside, and 9,000
nr’ outside Palexpo. These figures mark
the growth and importance of the tele-
communications industry when com-
pared to the net surface of 10,000 m'
used for the first Telecom exhibition in
1971. Asociated events during Telecom
87 were:

2.40 >888

■ 5 Forums, in which press represen-
tatives, experts and officials, and

other Telecom 87 attendants from the
ITU member countries were given a
chance to discuss openly various themes
related to telecommunication;

■ the 3rd Book and Audiovisual Fair on
telecommunication, electronics and

associated disciplines;

■ the 5th International Film Festival
culminating in the award of

the”Golden Antenna”;

■ the 5th Youth in the Electronic Age
competition.

The first day, 19 October, was certainly
the most interesting for the numerous
press representatives, since Telecom was
not yet opened for the general public.
Although it was announced that ac-
credited press members would have a
unique opportunity of arranging
meetings and interviews with VIPs and
ITU officials, the day (and much of the
evening) saw installation personnel and
many engineers working frantically to
get their company’s booth, display, or
stand ready in time. All this business of-
fered a unique chance to study brand
new equipment and connection tech-
niques from close by before it was care-
fully tucked away behind attractive look-
ing panels with letters and lighting in
various colours. Painting, opening
crates, carpet laying as well as connect-
ing terminals to mainframes and digital
exhanges was carried out as ’’last mi-
nute” work. Outside the Palexpo
building, on a specially reserved site, 30
or so dish aerials of widely varying size
were being brought into service to enable
permanent and ’’live” demonstations of
satellite communication for the coming
exhibition days. The signals to and from
the transportable dish stations were car-
ried to the various stands via cables, or
short-range microwave links via the roof
of the Palexpo building.

Telecom is not an exposition where one
has to press through masses of visitors to
get a glimpse of the material on display.
The various commercial and technical
representatives of exhibiting companies
were eager to demonstrate new equip-
ment to any interested visitor, although
some of the questions asked were so
deeply involved with technical details as
to require noting separately and for-
warding to an expert colleague in the rel-
evant department ”at home”. Press re-
leases were not scarce either, making it
very difficult, if not impossible, to leave
Palexpo without bags of documentation
and souvenirs for the occasion. The
press room was hectic, and cluttered with

paper. A number of companies used the
press notice board to criticize com-
petitors’ products by means of hurriedly
issued information bulletins. Cocktails,
lunches and important speeches were
scheduled and again postponed for
maximum impact on the press represen-
tatives.

The main theme of Telecom 87 was
excellently reflected by the Hall of
Nations, in which individual countries
had set up ’’national pavilions” accom-
modating the stands of many smaller,
and often highly specialized, companies.
The Italian, French, Northern
American, Scandinavian and West
German pavilions were remarkable for
the impressive number of companies ac-
commodated. Many companies were
present with their corporate stand, and
in addition as part of a national pav-
ilion. Examples were Siemens, Motorola
and Philips/AT&T.

Companies and institutions with related
products and services were located close
to one another in the halls. Inmarsat’s
’’Great Maritime & Land Mobile
Show”, and the 1:4 scaled model of
Ariane 4 were successful eye-catchers
calling visitors in the direction of a
number of important representatives in
the communications satellite field:
EuroSatellite, ESA, Inmarsat, Hughes
Aircraft, Intelsat, ArianeSpace, General
Dynamics and EutelSat.

Telecom 87 focused heavily on integrated
digital networks (ISDN) and satellite
communications. NTT (Nippon Tele-
phone & Telegraph Company), Mat-
sushita, NEC, GEC/Plessey and
AT&T/Philips attracted many visitors
with live demonstrations of ISDN. In an
ISDN system, telephone calls, telex, fac-
simile, slow-scan images, and LAN-like
digital services are all ’’packed together”
for high-speed transmission via fibre op-
tic links or geosynchronous communi-
cation satellites. System X from
GEC/Plessey is generally expected to
govern the new technical outlook of all-
digital, and ISDN compatible, data ex-
change systems. In fact, System X equip-
ment has been in use for a number of
years in all British Telecom trunk ex-
changes, but the latest enhancements to
the system as regards the transmission
speed achievable over satellite links has
aroused the interest of many national
PTTs looking for ways to extend existing
telephone networks. ISDN and satellite
communication are closely related, and
the relevant products and services of
many leading telecommunication
companies will be discussed in next
month’s issue of Elektor Electronics.

STEREO LIMITER

A quality limiter for use in tape recorders, transmitters, public address systems, and discotheques.

A limiter is an electronic volume adjust-
ment circuit in which AF signals are
amplified up to a predefined level of the
input amplitude. When this level is
reached, the gain of the amplifier is re-
duced to ensure that a fixed, maximum,
output level is not exceeded. In other
words, the output amplitude remains
constant irrespective of fluctuations of
the input signal above the limiting
threshold. Limiting is, therefore, often
referred to as dynamic range compres-
sion. Figure 1 shows the dynamic
response— U« as a function of U, — of
the proposed limiter.

The design described here is based on a
pair of standard gain controlled
amplifiers which ensure a dynamic range
compression of about 46 dB. The
limiting threshold is reached at an input
voltage of about 50 mV: the output
voltage is then about 670 mV.

Circuit description

With reference to the circuit diagram of
the stereo limiter in Fig. 2, opamp Ai
sums the signals applied to the L and R
inputs, and provides the gain control
signal for the limiter chip Type NE572 in
position ICj. Although it is economical
to provide a gain control signal common
to both channels, the result is, of course,
the likelihood of mutual and inap-
propriate gain reduction on the stereo
outputs. Fortunately, this effect does not
raise problems for programme material
played at average to loud levels, and the
differences in output volume on the
channels are certainly tolerable at less
than 5 dB.

Both channels in the Type NE572 dual
programmable analogue compander
(compressor-expander) from Valvo/
Mullard comprise a full-wave rectifier, a

Fig. I. Dynamic response of the stereo
limiter.

buffer and a linearized, temperature
compensated gain ceil. All these operate
independently from the corresponding
section in the other channel. The recti-
fier translates the AF signal from Ai
into a direct control current for the
buffer, which in turn controls the output
current provided by the associated gain
cell, marked AG in the circuit diagram.
The attack and recovery constants of the
gain controlled buffers are determined
with the aid of external electrolytic
capacitors Cs-Ct (L) and Cu-Cu (R).
The outputs of the current controlled
gain cells AG are connected to the feed-
back resistors of opamps A5 (L; R 5) and
A6 (R; Ru). Hence, the output current

provided by the gain cells controls the at-
tenuation introduced by As and At,. In
the present application, the operation of
the gain cells is, therefore, comparable to
that of a current controlled electronic
potentiometer. Output opamps A; (L)
and A-i (R) are dimensioned for an am-
plification of about 4.7. The oscillo-
grams of Fig. 3 show the dynamic
response of the limiter.

It is evident that the technical character-
istics of the proposed limiter are a com-
promise between what is useful on the
one hand, and practical for most appli-
cations on the other. This means that the
input threshold, the output level, the
dynamic range and the tracking(gain dis-

tribution) of the channels are dimen-
sioned such that the circuit is suitable for
a wide variety of applications. In some
cases, the technical characteristics may
need altering, however.

Resistor R.< (R12) sets the maximum am-
plification for an optimum signal to
noise ratio in the absence of an input
signal. The maximum usable resistance
is about 680K. The gain cells operate at
a bias potential of about -5 V, while
the + input of the associated oper-
ational amplifier (pin 5; 3) is connected
to ground. This means that the maxi-
mum drive for As (As) is about 1.4 V™,.
Both gain controlled opamps function
as an alternating voltage amplifier, and
do not, therefore, need a coupling ca-
pacitor to the associated output driver.
The attack constant is determined with
C12; Cs (R; L), the release constant with
Cu; G, (R; L).

The main point in the dimensioning of
the control circuit concerns the selection
of the control voltage for the gain cells.
In practice,- it was found that the drive
margin can not be set much higher than
- 25 dB, corresponding to the already
stated 50 mV (0 dB-a-i m w in 600 Q).
The input voltage should, therefore, not
exceed 130 rnVnm to avoid overdriving
the limiter, since this would then operate
linearly again, amplifying the input
signal. To avoid any risk of this happen-

Fig. 3. Automatic level control obtained with the limiter. Small signal response (b) and large
signal response (a). Upper channel: output; lower channel: input.

ing, it is recommended to fit presets with
a value of, say, 100 kQ at both limiter in-
puts.

The tracking (gain balance) of the chan-
nels is optimized with the aid of Pi. The
correct adjustment is reached after
checking, noting and comparing the
dynamic response curves of the L and R
channel with the aid of a calibrated sine-
wave generator, an oscilloscope and a
true-rms meter.

In the absence of these instruments, ac-
ceptable results are obtained when Pi is
set to the centre of its travel.

Construction follows the usual pattern
of fitting the components as per the
parts list and the white overlay on the
PCB. Fit the ICs in sockets, and do not
forget the 2 short wire links between C20
and C21. The capacitors in the corners
of the PCB are bipolar ( non-polarized )
types.

Attention: pin 5 of IC2 is erroneously
left unconnected on the PCB. This is
readily amended by running a short
length of light insulated wire from pin 5
to the ground connection of C20.

Construction and use

The ready-made printed circuit board
for the stereo limiter is shown in Fig. 4.

The supply voltages for the limiter can
be obtained by stepping down ± 10, ± 12
or ± 15 V rails available in the equip-
ment to incorporate the stereo limiter.

2.42

Zener diodes and discrete regulators are
equally suitable for providing the
regulated ±7 V supply voltage.

The limiter is best connected perma-
nently between the outputs of a line
driver or mixer, and the inputs of the
power amplifier. After establishing the
drive margin of the system, the output
and input level presets (if used) are
sealed to avoid overdriving the power
amplifier and the limiter, respectively.Gb

SCIENCE MOBILISES TO BEAT
MURDER IN THE AIR

by Bill Pressdee, BSc, CEng, MIEE

The problems confronting airport
security are basically the same as those
involved in the custody of any major in-
dustrial complex of national import-
ance. These include theft, ranging from
petty larcency to bullion robbery, illicit
incursion, ranging from unauthorized
entry to military takeover, and specifi-
cally terrorist attack, either in the air-
port or in the air.

Combatting these problems calls for ad-
equate surveillance by man or machine
to discover illegal acts, and the appro-
priate detection of illicit devices or
materials in sufficient time to apply
remedial action.

Incidents at airports — such as bullion
robbery, the smuggling of drugs, or the
discovery of explosives in hand lug-
gage— are generally reported in iso-
lation. However, to be effective, an air-
port system must be comprehensive to
cover every aspect of security. For each
airport, the security system must first
take account of the particular site prob-
lems.

There are many security devices available
for use around the airfield, and these
should be deployed to monitor various
zones of increasing risk, starting at the
perimeter fence: microphonic cable to
detect break in; closed circuit television
(CCTV) cameras to scan various sectors
of the airfield (possibly connected to

motion detectors); and infrared, micro-
wave and underground pressure detec-
tors to discover intruders and vehicles in
unauthorized areas.

Searching the public

Chubb Alarms is a major British secur-
ity company marketing a comprehensive
range of such devices, and has con-
siderable expertise to advise how an area
can best be protected.

Of more critical concern at present,
however, are the airport areas to which
the public have access— in particular the
interfaces between the public areas and
those restricted to passengers and air-
port staff. Through these pass the

Luggage inspected for explosives using the
A.I. Security Type 97.

passengers, their baggage and their hand
luggage, each piece of which may be
concealing weapons or explosives, or the
means for making or assembling them.
It is at these interfaces that, as the
criminal and would-be terrorist become
more ingenious, the detectors deployed
against them need to become more effec-
tive.

CCTV is an important general means of
monitoring the concourse of the airport
and noting irregularities. The recent de-
velopment of charge coupled device
(CCD) cameras using solid state image
sensors is a major step towards improved
CCTV surveillance. The units are ex-
tremely small, allowing covert oper-
ation; they have a long life, are robust,
need negligible maintenance and work at
low voltage with power consumption of
just a few watts.

Finding metal

Coupled via a fibre optic taper to a
microchannel plate image intensifier,
they are capable of operation over all
ranges of ambient illumination from
bright sunlight to starlight. The English
Electric Valve Company Ltd has recently
announced a comprehensive range of
CCD cameras and sensors manufactured
in its new factory at Chelmsford, eastern

>2.43

England, the most advanced CCD fa-
cility in Europe.

Firearms and most other weapons will
incorporate a substantial amount of
metal, which may be detected by X-ray
machines or metal detectors. Passengers,
on entering the airport’s departure area,
and possibly also before embarking,
may be required to pass through an arch-
way incorporating a metal detector (and
perhaps including an explosives detector
as well, as in the A. I. Security Entry
Scan Type 85).

The threshold sensitivity of the detector
will be set to discriminate between, say,
a small pistol and loose coins in the
passengers’ pockets: warnings of signifi-
cant metal detection may then prompt
further investigations by means of a
physical body search or hand held metal
detector such as the GM2 made by
Graseby Dynamics. .

Baggage and hand luggage is normally
inspected via an X-ray machine, tended
by an operator trained to identify sus-
picious opaque profiles. Recent products
from Astrophysics Research, which has
supplied over 2000 such machines world-
wide, include a combined check-in desk
and X-ray screening system and mobile
systems for spot checks.

Explosives threat

Detection of illegal objects depends at
present on the operator’s alertness and
experience, but the advent of micro-
processors operating at several million
instructions per second heralds the de-
velopment of expert detection systems in
which the X-ray responses will be com-
pared automatically with a multitude of
those from known weapon types.
Explosives represent the most deadly of
the armaments available to the terrorist
and also, even when made into impro-
vised bombs, the most difficult to detect.
The minimal amount of metal in the
detonator or triggering device is unlikely
to reach the alarm thresholds for arch-

way or hand held metal detectors.
Accordingly, over the last two decades
several types of explosive detector have
been developed. Explosive compounds,
their additives and decomposition prod-
ucts emit minute quantities of a charac-
teristic vapour, which is possible to
sample and identify— although with
some modern military explosives this is
far from easy.

The simplest sniffer devices available
rely on direct ionization of the vapour
from the explosives in air. These include
typically the Graseby Dynamics PD4C
and the A. I. Security Model 35, which
are light, compact and easy to use.
Although they perform a useful func-
tion, their sensitivity is limited, their
discrimination medial, and they are
unlikely to respond to certain military
explosives.

Finer sampling

A more sensitive and selective range of
instruments is available, based on gas
chromatography. It includes the A. I.
Security Model 97. These devices rely on
a constant and very pure supply of inert
gas, and the means for introduction of
the atmospheric sample into the gas
stream. The penalties for increased
sophistication, however, appear in terms
of increased size, weight, warm-up time,
response time, and cost.

The only other detectors in being are the
very complex and accurate instruments
mostly confined by their size and lack of
portability to laboratories, except for a
recent product of Graseby Dynamics,
the Ion Mobility Spectroscope (IMS),
Model PD5 which shows considerable
promise.

The IMS operates by first drawing an air
sample through a probe and over a mem-
brane, which excludes dust and moisture,
but permits the diffusion of the vapour
molecules. The molecules are then ion-
ized by a weak Nickel-63 beta emitter
and subjected to a 1000 V DC field, con-

trolled by a gating grid which allows the
passage of the ions in discrete samples.
The drifting ions become ranged
spatially in order of their mobilities,
and, on reaching the collector electrode,
present a current waveform character-
istic of the ions in the sample. The
internal air stream is circulated by a
pump and dried. Then certain dopant
chemicals are added in minute quantities
to enhance the sensitivity.

The waveform received by the collector is
digitized and fed to a microprocessor, in
which its characteristics are assessed
against patterns for various explosives,
while vapours derived from other
substances are disregarded. The PD5 has
a hand-held unit with digital readout
connected by an umbilical to a briefcase,
making it ideally portable.

In many aspects the means for ensuring
good airport security are becoming bet-
ter and more sophisticated. It remains
only for airport authorities to develop
security systems that use these means ef-
fectively.

A. I. Security (Division of Analytical In-
struments Ltd), Pampisford, Cambridge
CB2 4EF.

Graseby Dynamics Ltd, 459 Park
Avenue, Bushey, Watford WD2 2BW.

EEV Solid State Devices, (English Elec-
tric Valve Company Ltd), Waterhouse
Lane, Chelmsford, Essex CM1 2QU.

Chubb Alarms Ltd, 42-50 Hersham
Road, Walton-on-Thames, Surrey
KTI2 IRY.

Astrophysics Research Ltd, 100 Vale
Road, Windsor, Berkshire SL4 5JP.

LIGHT POWERED
THERMOMETER

An accurate, automatically operating electronic thermometer that indicates temperature on a digital
readout without the need for batteries or a mains supply.

The thermometer described here is pow-
ered by an amorphous solar cell. In con-
trast to other types of solar cell, this uses
a non-crystalline silicon layer. Amorph-
ous solar cells arc produced by con-
trolled deposition of silicon on a glass
surface, which forms the top of the cell.
The production method is relatively
simple and cost-effective, but has the
disadvantage of yielding cells with a
relatively low efficiency. The basic struc-
ture of an amorphous solar cell is shown
in the diagram of Fig. 1. The element is
composed of 3 series connected cells
secured on a glass plate. Each photon
that enters a cell causes the release of an
electron from a silicon atom. The release
generates electric energy, which can be
used for powering the thermometer cir-
cuit — provided, of course, there is suf-
ficient incident light on the solar cell.

Circuit description

With reference to the circuit diagram of
Fig. 2, the temperature sensor is formed
by IC2. This is the well-known precision
centigrade temperature sensor Type
LM35CZ from National Semiconductor.
Housed in a plastic TO92 enclosure, this
device gives a linear output voltage of

+ 10.0 mV/°C. It does not require any
external calibration or trimming, and yet
provides an accuracy of ±0.25 °C at
room temperature, and ±0.75 °C over
the full -40 to +110 °C temperature
range. The 1C is internally calibrated
such that 0 °C corresponds to an output
voltage of 0 V. The remaining functions
of the thermometer circuit are a volt-
meter, a read-out, and a decimal point
shifter. All these are combined in a single
integrated circuit Type ICL7136 (ICi),
and a VA digit liquid crystal display
(LCD). The oscillator internal to ICi is
operated at the lowest possible clock
speed to ensure minimum power con-

sumption of the circuit whilst avoiding
display flicker. The thermometer read-
out is calibrated with the aid of Pi.
Components Di and Rn enable the
sensor to provide a negative output
potential when the temperature falls
below 0 °C. LEDs Di and D2 do not
function as light sources, but as
reasonably stable 1.6 V references that
require a forward current of only a few
micro-amps. Standard zener diodes give
better regulation, but are not suitable
here in view of the relatively high for-
ward current required for the stabiliz-
ation effect.

The circuit around IC3 is a voltage

Fig. 3. Track layout and component mounting plan. The completed printed circuit board can
be fitted in a transparent Hcddic enclosure.

monitor that switches the thermometer
off via T 2 when the potential supplied
by the solar cell falls below 7.0 V. This
protective measure effectively prevents
erroneous read-outs: for accurate oper-
ation sensor IC 2 requires a minimum
supply voltage of 5.5 V, while the refer-
ence source internal to ICi should be
fed with 7.0 V or more. Schmitt-trigger
IC 3 in the voltage monitor switches T 2
on again at an input voltage of 8 V, i.e.,
the circuit is dimensioned for a hysteresis
of 1 V. The switch-on threshold is set at
7.0 V with the aid of preset P 2 . The cur-
rent consumption of the thermometer in
the de-activated and activated state is
about 10 and 200 |/A, respectively.
When the circuit is in the de-activated
state, and there is moderate incident
light, the solar cell can only supply
about 100 /iA, so that Ct is charged to
8 V. The thermometer is switched on,
and draws more current than can be sup-
plied by the solar cell. This means that
C6 is discharged: the supply voltage
drops below 7.0 V, and the thermometer
is switched off again after a few seconds.
This automatic on-off arrangement
enables taking temperature readings even
in less favourable weather conditions.
The hysteresis of IC3 can be increased
by reducing the value of R~, and re-
adjusting P 2 . It is possible to use a
smaller capacitor for C<>, so that the
thermometer is switched on rapidly
when the light intensity increases. Fi-
nally, the function of D 7 is to limit the

supply voltage to 12 V when there is in-
tense sunlight on the solar cell.

Construction and setting up

The printed circuit board for the light
powered thermometer is shown in Fig. 3.
The completion of the board should not
present difficulty, but care should be
taken handling and mounting the fragile
LC display. Do not overlook the 2 wire
links on the board.

Do not yet fit IC 2 , and apply + 1.000 V
to the points intended for the Von. and
GND terminals of the sensor. Adjust Pi
for a display reading of 100 °C. Remove
the voltage source connections, and fit
IC 2 . The completed PCB and the solar
cell are made to fit in a transparent Hed-
dic enclosure. The space is quite tight,
and the drop of sealing resin on the LCD
may have to be flattened by careful fil-
ing. One side of the enclosure of ICi is
treated likewise. Use wire-wrap terminal
strips so that the face of the LCD is
pressed against the inside of the lid. Drill
a few holes in the enclosure to prevent
heat building up inside. The response of
the sensor to rapid temperature changes
can be improved somewhat by glueing a
small piece of thin metal sheet onto the
flat side of the TO92 enclosure.

Some spare room is available in the
enclosure for an optional 9 V (PP3) bat-
tery. A switch can be fitted to select be-
tween the battery or the solar cell as the
power supply for the thermometer. TW

Parts list

Resistors (±5%):

Ri;R2:Ra=1M0
R4 = 220K
R5=270K
R6=1M8
R7 = 10M
Rs=1M5
R9 = 470K
RlO = 680K
Rn =33K

Pi = 200K or 220K multiturn preset
P2 = 1M0 preset H

Capacitors:

Ci;C2;C3=100n
C4 = 82n
C6 = 47p
Ce = 470«; 16 V
semiconductors: ’

Di;D2-LED:3mm;red
D3 = AA1 19

D4= zenerdiode 12 V; 0.4 W
Tl*BS170‘

T2 = BC547B

ICi =ICL7136CPL* *

IC2 = LM35CZ (Maplin order no. UF51FI
IC3=TLC271 +

* available from Cricklewood Electronics

available from Universal Semiconductor
Devices Limited.

Miscellaneous:

LCD = the following types may be used:
LTD221-C01 (Mullard/Videlec; for distributors
refer to InfoCard 507 in the April 1987 issue
of £fl:

43D5R03 (Data Modul/LXD; 2D Electronics •
Wellington House • 2 Kentwood Hill •
Reading. Tel. (07341 4204401:

3901 or 3902 (Hamlin; Hamlin Electronic Europe
Umited • Park Road • Diss • Norfolk IP22
3AYI.

Solems Type J0887JB01"

Enclosure Heddic Type 222-G".

PCB Type 87188 (refer to the Reeders Services
page).

PP3 battery (optional).

" Heiland Electronic Design & Development •
Hermann Loans Strasse 11 • D-4410
Warendorf 3 • West Germany. Telephone:

+ 49 125821 7550.

Availability in the UK: Chartland Electronics
Limited • Chartland House • Twinoaks •
Cobham • Surrey KT11 2QW. Telephone:

(037 284) 2553.

For further details on distribution: Barnes
Newman International Limited • Office Suite 1
• The Square, Forest Row • East Sussex
RH18 5ES. Telephone: (034 282) 2708. Telex:
95637.

mometer.

WIDEBAND AERIAL BOOSTER
AND SPLITTER

A single transistor amplifier and a matched RF signal distributor ensure that the signal provided by the
aerial or cable network can be fed to several radio or TV sets without loss of quality.

Many people buy a second TV scl for
use in a location other than the living
room. After experiencing the disap-
pointing reception obtained by the use
of the built-in or set-top aerial, it is often
decided to connect the new TV to the
same antenna input as the main set.
However, the assumption that the signal
strength is high enough to feed 2 TV sets
is immediately proved false by ghost ef-
fects and considerably increased noise
on. . .both sets!

Aerials and coax

A good quality directional aerial is the
best RF amplifier. It is often frequency
selective, consumes no power, introduces
no noise, and gives considerable amplifi-
cation. A typical multi-element Yagi
aerial for UHF TV reception has a half
power opening angle of about 15°, and
a power gain of 12 to 16 dBi. Since the
aerial is usually mounted in the highest
possible location on the roof, its output
signal needs to be fed down to the re-
ceiver via a cable that ensures minimum
signal loss, freedom of induced inter-
ference, and correct matching at both
ends. The loss introduced by a cable of
any type (coax, twin-feed) is directly
related to its length, and the frequency
of the signal it carries. For coax, the loss
is generally lower with increasing cable
diameter. At 100 MHz, for instance,
Type RG213/U cable has an attenuation
of about 5.7 dB per 100 m, while
RG58/U is specified at 14 dB. RG213/U
(Z<i = 50 Q) and RG58/U (Zo=53.5Q)
have an outside diameter of 10.3 and
5 mm, respectively. The commonly
used, general purpose, white coax for
use in the home is certainly not better
than RG58/U.

Although it is assumed here that con-
structors are familiar with the general
rule that an RF amplifier should be fit-
ted as close as possible to the aerial, the
following example may prove helpful to
illustrate the practical consequences if
this is not the case.

Any electronic amplifier produces noise.
Assuming that the circuit in question
receives an input signal of 10 //V, and
amplifies this, say, 10 times while adding
1 ,uV noise, it is readily seen that its out-
put signal is composed of 100 /jV

Fig. t. Circuit diagram of the wideband
aerial booster.

’’signal” and 1 /j\ noise, i.e., the signal
to noise (S/N) ratio is 100:1. When the
downlead coax has an attenuation of 4,
the TV set receives 25 /tV ’’signal” and
0.25 /j\ noise. Thus, the S/N ratio is not
affected by the coax cable.

When the aerial amplifier is fitted at the
low end of the downlead coax, i.e., close
to the TV set, it receives a signal of
2.5 ,uV, which it amplifies to 25 fiV. The
noise level at the output is, again, 1 ^V,
however, so that the resultant S/N ratio
is only 25:1. The signal amplitude is still
25 fiV, but the noise level is quadrupled
from 0.25 //V to 1 //V. The conclusion is
obvious: the amplifier should be fitted
as close as possible to the aerial, and
the connection between them should be
made in high quality (low loss) cable. In
general, the amplification of the aerial
booster ensures that the available S/N
ratio is not affected by the downlead
cable, even if this has considerable at-
tenuation.

Wideband aerial booster

The amplifier described here is a wide-
band design with a frequency range of
about 80—800 MHz. Its advantages are
mainly the ease of construction, and the
absence of tuned circuits. Its inherent
disadvantages, are, however, equally im-
portant to note. The absence of any
form of selective filters in the circuit may
give rise to cross-modulation and
blocking in the vicinity of powerful
transmitters (mobile radios, TV
transmitters, cellular radio repeaters,
etc.).

With reference to the circuit diagram of
Fig. 1, the amplifier is based on low
noise RF transistor Type BFG65 from
Mullard. Components P1-R1-L3-G. are
fitted close to a mains adaptor that pro-
vides a regulated output of 12 V. The
amplifier is fed via the core of the
downlead coax cable. Choke L.< prevents
the RF signals being short-circuited in
the supply, while G keeps the ampli-
fier’s supply voltage away from the TV
input. The RF signal provided by Ti is,
therefore, superimposed on the supply
voltage.

Zenerdiode Di keeps the base of Ti at
4.7 V below the collector potential. In-
ductors Li and L2 prevent RF feedback

Fig. 3. The completed amplifier and supply hoards connected to the coax cables.

effective grounding, and the shortest
possible connection of the centre core.
The amplifier is fitted in a waterproof
ABS enclosure for fixing onto the aerial
mast. Drill 2 or 3 small holes in the
underside of the enclosure to prevent
water gathering inside. The supply sec-
tion is fitted in a small enclosure, located
on the attic, or behind the TV set,
together with the mains adaptor.

Turn the wiper of Pi for maximum re-
sistance. Measure the current drain of
the completed amplifier by connecting
an ammeter between the adaptor output
and the +12 V input on the supply
board. Pi is carefully advanced until a
weak station is sufficently amplified,
without running into cross-modulation
caused by stronger signals. Do not ex-
ceed 25 mA on penalty of damaging Ti.
When the amplifier is used for
distributing signals on the cable network
as discussed above, it is recommended to
replace Ti by a Type BFG96, operated
at a collector current of 75 mA (this may
require adapting Ri).

between the collector and the base whilst
passing current through Di. The combi-
nation of electrolytic and SMD
capacitors (Cs-C 2 and C+Cj) ensures
optimum decoupling for the entire fre-
quency range of the amplifier.

The aerial booster plus supply section is
constructed on the PC board shown in
Fig. 2. All parts are fitted at the copper
side. The PCB is cut to separate the
amplifier and the supply section. Drill a
5 mm hole to receive Ti, whose leads
are cut to size, and soldered flat onto the
relevant copper areas. Ascertain the pin-
ning before fitting the BFG65! SMD
capacitors Ci, C 2 and C> are carefully

glued into position before soldering. In-
ductors Li and L: are wound as 6 turns
of 00.2 mm (36SWG) enamelled copper
wire through miniature (3 mm) ferrite
beads. The enamel coating is carefully
scratched off the connecting wires, these
are tinned, pushed through the respec-
tive holes, soldered at the copper side,
and cut off at the other side of the
board. The fitting of the zenerdiode and
the 2 electrolytic capacitors should not
present difficulties. Mount a small tin or
brass screen across the transistor as
shown on the component overlay. Figure
3 shows the completed aerial booster
plus supply. Note that the coax cables
are clamped onto the boards to ensure

Signal divider

The previously described amplifier has
sufficient gain to enable dividing its out-
put signal between a number of TV sets
in the home. All the signal dividers to be
described should obtain their input
signal from C<., i.e., they must not be
fitted between the amplifier output and
the supply.

The signal dividers are assumed to be
terminated in 75 2. Figure 4 shows the
most elementary set-up of a coax signal
divider. Although the input signal is cor-

(25 + 75)/2 = 50 Q

Fig. 4. This signal divider is loo simple to
give optimum results.

rectly terminated in 75 £2, both TV sets
at the outputs see a source impedance of

75 + (75//75) = 112.5 Q.

A better circuit is shown in Fig. 5a. In
this, there are three 25 Q resistors and 3
termination resistors for the input and
the 2 outputs. Each signal path has a
25 Q resistor and a parallel combination
with an equivalent resistance of

so that the characteristic impedance is
50 + 25 = 75 Q. In the circuit of Fig. 5b,
the signal amplitude on each of the 3
outputs is one third of that at the input.
For an n- way divider, the value of R is
calculated from

8 *JTT x75IS1 '

Fig. 5. Signal dividers that maintain the correct termination and source impedance in 75 Q
cable networks.

Fig. 6. An alternative 4-way signal divider.

The resistor values obtained from this
equation can be approximated with the
nearest E12 or E96 value; for the 3-way
and 4-way dividers of Figs. 5b. and 5c,
the respective values of 39 8 and 47 Q
can be used with impunity.

It is possible to further subdivide the
signal with the aid of the circuit shown
in Fig. 6. This forms an alternative to
the circuit in Fig. 5c, and enables
economizing on coax cable, since the
distribution point need no be central to
all 4 connections. It is important to note
that all outputs on the dividers described
here require termination in 75 Q, either
by a TV or radio set, or, if this is not
connected, by a 75 2 resistor. Finally,
Fig. 7 shows a practical version of a 2-
way divider fitted in a metal enclosure to
prevent stray radiation. B;D

Fig. 7. Suggested construction of a 2-way
signal divider in a metal enclosure.

,2.49

A circuit is only as good as its
power supply. Even the best
designed circuits can be upset by
supply ripple, poor regulation,
mains transients and power supply
instability. 1C voltage regulators
have taken much of the donkey
work out of power supply design
for many applications, but the
problems do not always end there.

Power supply circuits are usually de-
signed almost as an afterthought. The
circuit requiring power has probably
been developed on the bench using the
ubiquitous lab power supply. In general,
a power supply for an electronic circuit
should approximate an ideal voltage
generator as well as possible. This means
that the output voltage should remain
constant over the norma! output current
operating range: in other words, the
supply should have zero output im-
pedance.

The supply voltage should not only
remain constant under static conditions,
but should also still remain constant
when dynamic changes of current are
occurring. As mentioned in the intro-
duction, 1C voltage regulators make it
easy to achieve these objectives at the
output terminals of the supply, but at
points in the circuit remote from the
supply the situation may be very differ-

Dcpending on the type of circuit, the
inductance of supply wiring and p.c.
board tracks, though small, may have
a dramatic effect. The worst offenders
in this respect are probably TTL logic
circuits. When the output state of a
TTL gate changes this results in a
I change in the current drawn by the
| circuit of around 2. m.A. A static
current difference of 2 mA would
probably result in a supply voltage
variation of a few microvolts. The
; dynamic situation can be very different,

The fall (or rise) time of .. i ll gate
from logic I to logic 0 level s of
the otder of 10 nanoseconds. The rate

of change of current ^ is therefore
dt

2 mA in 10ns or 200,000 amps/sec. The
voltage V across an inductor L equals
L— , so a wiring inductance of the order
of I pH between the supply and the
point where the current change oc-
curred would result in a voltage drop of
200 mV.

This is, of course, a gross oversimplifi-
cation, but it docs demonstrate the sort
of volLage transients that can occur on
supply lines in TTI, circuits. When such
transients arc of sufficient magnitude
they can cause gates to change stale and
can cause spurious triggering of flip-
flops, counters apd monostables.

The problem can be attacked in several
ways. The simplest approach is to pro-
vide local decoupling of the supply at
various points in the circuit by means of
small capacitors, typically 10 to lOOn.
These should preferably be ceramic
types which have low self- inductance
(figure la). Nine times out of ten this
will effect a cure, but in some cases the
cure can be worse than the disease, for
it is possible that Ihe capacitor will form
a resonant circuit with the inductance
of the supply lines. When this is excited
by the switching transients of the TTL
the problem can be magnified rather

than reduced.

The solution here is to damp the res-
onant circuit by including a resistor in
series with each capacitor. 1 J2 carbon
composition resistors should be used
for this purpose (film types often have
a high self-inductance). This is shown in
figure 1 b. An alternative to this is to use
electrolytic capacitors of say 10 p/6.3 V.
Modern electrolytics have fairly low
self-inductance, but the internal resist-
ance is often sufficiently high to pro-
vided adequate damping.

Figure 2a shows the use of RC combi-
nations to decouple several IC’s fed
from the same supply rail. Rather than
using a single supply rail running from
one IC to another a better solution,
whenever possible, is to run separate
supply rails from the power supply to
each IC or group of two or three IC’s.
In this way interaction between the
various IC’s is reduced, since the self-
inductance of each supply line actually
helps to isolate the IC’s from one
another (figure 2b).

Another useful tactic is to make the
supply lines fairly thin, as this increases
their resistance and thus reduces the Q
of the self-inductance. While on the
subject of separate supply lines it is
worth considering ‘on-card stabilising'.
In complex TTL systems, where there
are several p.c. boards, a separate IC
voltage regulator is often used on each
board. This has several advantages:

1. Interaction between boards is vir-
tually eliminated.

2. Heat dissipation is spread over several
IC regulators rather than being con-
centrated in one central power
supply, so cooling problems are re-
duced.

3. Each board can be fed from a re-
motely located unregulated supply
without worrying about voltage drops
along the supply lines (provided the
voltage reaching the boards is suf-
ficient for the regulators to operate).

Diode decoupling

Of course, not every circuit will be used
with a stabilised supply, there are cases
where it is simply not necessary or is
uneconomic. In such cases the simple
transformer, bridge rectifier and
smoothing capacitor is generally used
(figure 3). Off load this gives a reason-
ably constant d.c. voltage, but as the
load current increases the supply will
exhibit an increasing ripple voltage (see
figure 5).

Some times a circuit requiring a ripple-
free supply at a relatively low current,
must be operated from the same
unstabilised supply as a circuit requiring
a higher current. A typical example of
this is an audio amplifier where the
preamp is fed from the same supply as
the power amplifier. Generally, the
power amplifier can tolerate higher
ripple voltages than the preamp.
Clearly some way must be found of
isolating the preamp from the ripple
voltage caused by the power amplifier.
This can be achieved by RC filtering, as
in figure 3. The high current supply is

, 2.51

first i.f. amplifier, mixer, BFO, second
i.f. amplifier, demodulator and a.f.
stages. Each of the stages would have
its supply connection via a parallel
resonant circuit tuned to the centre
frequency handled by that stage (e.g.
10.7 MHz in the case of an f.m. i.f.
amplifier). This will act as a wavetrap
to prevent RF getting back onto the
supply lines (figure 7a).

Where the circuit handles a wide range
of frequencies it may be necessary to
use a filter with a wider stopband by
connecting two or more resonant cir-
cuits (of different frequencies) in series,
as in figure 7b.

At very high frequencies the use of a
series resonant circuit to shunt unwanted

RF. signals down to ground is also
recommended. The combination of
parallel resonant and series resonant cir-
cuits is shown in figure 7c. For circuits
working in the hundreds of megahertz
the required inductance for the series
circuit can be obtained simply by
trimming the capacitor leads to an
appropriate length. H

Figure 3. Where a ripple free supply is taken
from a non-ripple free supply it may be

Figure 4. Substitution of a diode for R1 gives
a great improvement. Once C2 has charged it
is completely isolated from Cl and the ripple
depends solely on the current drawn by R|_.

Figure 5. Ripple voltage on Cl for increasing

Figure 6. The ripple voltage on C2 remains
ripple voltage on Cl.

Figure 7. Three methods of supply decoupling
in RF circuits, a. Narrowband filter, b. Wide
stopband filter, c. Combination of parallel and
series resonant circuits for VHF use.

selex-3i

MINI - SYNTHESIZER

The sound of electronic organ
and synthesizer has become
quite common everywhere.
However, if you want to buy an
instrument from the market it
would be quite costly.

What we have presented here is
a simple low cost circuit for a
Mini-Synthesizer. It will also help
you in learning how the electronic
synthesizer functions. The
description is not only of the
electronics involved, but also
includes references to 'tones'
and 'octaves'. What voltages and
frequencies mean to the
electronics expert, tones and
octaves mean to the musician.

To fully understand the mini
synthesizer one must have at
least a working knowledge of

frequencies, as well as the tones

Block Diagram

Figure 1 shows the simplified
block diagram of our synthesizer.
On the left side is the 'Tone
Register' or the part which
decides the tone that will be
produced. Then comes the heart
of the synthesizer - the VCO, or
the Voltage Controlled Oscillator.
The frequency of the VCO is
dependant on the DC voltage
present at its input. In case of a
professional electronic organ,
there are twelve individual
oscillators, which form the basic
tone generators. Other octaves
are desired from these main
oscillators. In case of a
synthesizer, there is one VCO (or
sometimes more) which
generates the frequencies,
depending on the DC voltage at

In case of the synthesizer circuit
given here, we have thirteen
different voltage inputs and one
VCO. These thirteen voltage
inputs are available in the Tone
Register, which is nothing but a

bank of trimpots connected
across a reference voltage. The
sliding contacts of the trimpots
are set at different positions so
that they give different voltages.
These voltages are used as the
control voltages for the VCO. The
connection between the tone
register and the VCO is through a
flexible wire, with a probe tip

end being permanently soldered
to the VCO input. Thus by
touching the various outputs
from the tone register with the
probe tip, different frequencies
can be generated by the VCO.
The output of the VCO goes to a
mixer. This delivers the highest

frequency. Secondly, this output
also goes to a frequency divider
which divides the VCO frequency
by 2, 4, 8 and 16 and gives; four
different output frequencies.
These four outputs of the
frequency divider are also fed to

the mixer pass through switches,
and can be connected or cut off

.. 988 2.53

from the mixer as desired. The
divider thus forms the sub-octave
generator. These sub-octaves

frequency by closing the
switches. By opening the switch
in the main frequency input and
only closing the switches in the
i-octave inputs, it is possible
only to feed the sub-octaves and
le main frequency. The
mixer also consists of
potentiometers to decide how
percentage of which
frequency passes through to the
amplifier.

The Circuit

The complete circuit (except thf
mplifier, of course.) of the
nini-synthesizer is given in
igure 2. The circuit is exactly
similar to the block diagram we
studied in figure 1 . You can easil
recognise the tone register, the

frequency divider and the mixer.
The VCO consists of A1 and A2,
whereas A3 and A4 just serve as
buffers to avoid overloading the

The tone register is formed by
the trimpots PI to P14. The VCO
consists of A1, A2, R1. R2, R3,

PI 5, Cl and D1. The frequency
divider is formed by IC2. The
mixer consists of potentiometers
PI 6 to P20, resistors R6 to R11
and switches S2 to S6.

A3 forms the buffer which
generates the voltage reference
for the tone register and virtual
ground for all other Op amps. R4
and R5 are part of a voltage
divider which give a voltage of
4.5 V at the non inverting input of
A3. The inverting input of A3 is
shorted to its output, so that A3
behaves as a voltage follower,
and the output is stabilised at 4.5
V. This output is used as the
reference for the tone generator.

It also serves as the virtual
ground for the other Op amps,
thus eliminating the need for a
dual power supply. Effectively it
generates a supply with + 4.5V,
— 4.5 V and the virtual ground.
The current consumption for the
circuit is about a few mA, and the
9V battery can last for a long time.
The trimpots PI to PI 3 are
connected parallely across the
4.5 V reference supply through a
common trimpot P14. The total
resistance of the parallel bank of
resistors comes to about 7700
and P14 is also adjusted to about
the same value. The voltages at
the sliding contacts of PI
lie between 2.25 and 4.5 V with
reference to the virtual ground.
These voltages are brought out
an a piano type key pattern mad
of aluminium or copper foil.

PI 3

The detailed circuit of the
synthesizer.

There are four main sections -
The Tone Register, VCO,
Frequency divider and the Mix.

2.54,

become positive. Output of A2
now jumps to + 4.5 V again. Tf
charging discharging cycle thus
repeats, and what we get at the
output of A1 is a triangular wav

Figure 3 shows the VCO section
of the circuit again, along with
the voltage wavefarmes present
at different points in the VCO.
The output of the VCO at pin 14 is
fed to the mixer through the
buffer A4.

Inside the mixer circuit, it also
passes through R6, to the
potentiometer PI 6. Pin 14 also
feeds the frequency divider 1C
4024, which gives four outputs
Q 0 , Qi, Q 2 and Q 3 . These are V4,
'A, VS and V le of the input
frequency. These signals pass
over resistors R7, R8, R9 and RIO
to the potentiometers P17, P18,
PI 9, and P20 of the mixer circuit.
The mixed signal appears at R1 1
and goes out to the amplifier
over capacitor C3.

Switches S2 to S6 are the octave
selector switches, which decide
the signals to be mixed to
produce the output of the
synthesizer. The settings of PI 6
fo P20 decide the percentage in
which these octaves are mixed in
the output signal. R1 1, C2 and C3
combination serves as a low pass
filter and improves the quality of
the sound generated by the
synthesizer. The output can be

component layout of the circuit
in two parts. Figure 4 shows the
VCO and the frequency divider,
whereas figure 5 shows the tone

Connection of the VCO input
must be done through a shielded
wire, and the shield should be
connected to ground. The
components of the mixer circuit
should be wired directly in the
sythesizer enclosure. The
keyboard should be constructed
on the top of the enclosure, using
aluminium foil or copper foil. If
you have a facility for etching
PCBs, it can be etched on a PCB
and mounted on top of the
enclosure. Your craftsmanship
will be the limit on how good a
keyboard you can construct. The
minimum requirement being
that it should be able to make an
electrical contact! Figure 6 shows
the keyboard pattern. You can
use a combination of aluminium
and copper foil to give an

After the electronics is finished,
the musical alignment must be
done. The so called tuning of the
synthesizer. For this you will
need a properly tuned reference
instrument, a piano, a flute or a
guitar. The tuning procedure is
as follows:

2. Bring PI 3 approximately tc

3. Adjust PI almost near the

4. Close SI. and switch on thi
audio amplifier.

5. Align P14 in such a way thi
between the highest and
lowest tone lies
approximately 1 octave.
Adjust P13 again if require!

The VCO Section reproduced for
better understanding, with
important voltage waveforms at
various points inside the VCO.

Component layout of the VCO
and the frequency divider, on
SELEX PCB- size 1.

. 2.55

a-2.73

R N No. 39881/83

Oiam\7\ati Z:l££tram£s Vv\ Lt/L

S-17, M.l.D.C. Bhosari, Pune 411 026. India. (0212) 82682. 83791 Cable : CHAMPION

J. 0145-505 MCCI IN.