9ROXPH  ,VVXH 

'HFHPEHU 
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, Febuary 1999 - www.epemag.com - XXX
Maxfield & Montrose Interactive Inc
 Copyright  1999, Wimborne Publishing Ltd
and Maxfield & Montrose Interactive Inc.,
PO Box 857, Madison, Alabama 35758, USA
All rights reserved.

:$51,1*
The materials and works contained within EPE Online — which are made available
by Wimborne Publishing Ltd and Maxfield & Montrose Interactive Inc — are
copyrighted. You are permitted to download locally these materials and works and to
make one (1) hard copy of such materials and works for your personal use. International
copyright laws, however, prohibit any further copying or reproduction of such materials
and works, or any republication of any kind.

Maxfield & Montrose Interactive Inc and Wimborne Publishing Ltd have used their
best efforts in preparing these materials and works. However, Maxfield & Montrose
Interactive Inc and Wimborne Publishing Ltd make no warranties of any kind, expressed
or implied, with regard to the documentation or data contained herein, and specifically
disclaim, without limitation, any implied warranties of merchantability and fitness for a
particular purpose. Because of possible variances in the quality and condition of
materials and workmanship used by readers, EPE Online, its publishers and agents
disclaim any responsibility for the safe and proper functioning of reader-constructed
projects based on or from information published in these materials and works.

In no event shall Maxfield & Montrose Interactive Inc or Wimborne Publishing Ltd be
responsible or liable for any loss of profit or any other commercial damages, including
but not limited to special, incidental, consequential, or any other damages in connection
with or arising out of furnishing, performance, or use of these materials and works.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1082
Maxfield & Montrose Interactive Inc
 PROJECTS AND CIRCUITS
PIC MICRO-PROBE - by Jim Main 1088
A hardware tool to help debug your PIC software.

MAGNETIC FIELD DETECTOR - by Robert Penfold 1093

Where’s the attraction? Find out with this Starter Project!

INGENUITY UNLIMITED - SPECIAL - hosted by Alan Winstanley 1099

Serial Port Splitter; Elderly Person Monitor; Rechargeable PP9 Battery; Class-D 30W Audio
Amplifier; National Lottery Predictor; Tumble Dryer Alarm; Narrow Band Vision;
Squash/Badminton Scorer; Time-Lapse Unit for Camcorder; Audio Limiter.

LOFT GUARD - Terry de Vaux-Balbirnie 1110

A lofty way to keep your power bills low!

GINORMOUS STOPWATCH - Part 2 - by Ned Stojadinovic 1116

Adding giant digits to your Stopwatch display

SERIES AND FEATURES

CIRCUIT SURGERY - by Alan Winstanley & Ian Bell 1124
Opamps 101; Fusible resistors

INTERFACE - by Robert Penfold

A serial approach to PC add-ons.
1128
NEW TECHNOLOGY UPDATE - by Ian Poole 1131
Curly nanotubes could determine the nature of future transistors.

PRACTICAL OSCILLATOR DESIGN - Part 6 1133

Resistor/capacitor oscillators by Raymond Haigh
Worked examples and circuit info for hands-on constructors.

TEACH-IN 2000 - Part 2. Capacitors and RC Timing - by John Becker 1142

Essential info for the electronics novice, with breadboard experiments
and interactive computer simulators.

The Joy of EL! - Part 1 - by Mike Ceranski 1157

Everything you needed to know electroluminescent cable but were too afraid to ask!

NET WORK - THE INTERNET PAGE surfed by Alan Winstanley 1160

Jet Scream; Helps Us to Help You; Brow-beaten; Help-line

REGULARS AND SERVICES

EDITORIAL 1084
NEWS - Barry Fox highlights technology’s leading edge. Plus everyday 1162
news from the world of electronics.

READOUT - John Becker addresses general points arising. 1167

SHOPTALK - with David Barrington 1172
The essential guide to component buying for EPE Online projects.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1083
Maxfield & Montrose Interactive Inc
 NEW LISTBOT SERVER
As you know, every month we send each of our subscribers an email message to
announces the availability of the new magazine and provide a summary of the articles
featured in this issue. Over the past few months we have seen a significant increase in
the number of our subscribers, and we have therefore been obliged to streamline our
email notification system.
In order to help us maintain our notification system, we will be moving our email
distribution list to a ListBot server, and we will then use this server to provide you with
your notification email each month. The ListBot server is self-maintaining, in that it
automatically removes any duplicate email paths or paths that have become invalid. It
also allows you to change your email path and remove yourself from the subscription list
if you require, which all goes to help reduce the overhead of running and maintaining
our email lists.
To kick this new system off you will receive a notification from the EPE Online ListBot
server stating that your name has been added to the EPE Online Newsletter
subscription. In order to continue receiving email notifications from us, you will need to
confirm your acceptance of joining the new ListBot server. This is a one-time action that
you will need to take in order to “sign up” for the notification service – all it takes is to
simply “Reply” to the original email from the ListBot server.

EXTRA SERVICES
As you can imagine, the editorial department receives a great deal of email each
month from subscribers (and sometimes from would-be subscribers) requesting us to
provide a design service for “a special project that I am working on” (usually qualified by
“and I’m on a very tight deadline, so please respond ASAP”).
While we strive to provide you with a magazine that is both informative and
interesting each month, we simply do not have the resources to provide a consultative
design service to our readers. (If only, there were 48 hours in each day).
However, we do try to respond to all emails dealing with articles in the magazine and
any queries regarding subscriptions within one business day.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1084
Maxfield & Montrose Interactive Inc
SCRATCH BLANKER
It seems that reports of the death of vinyl discs have been somewhat exaggerated. While it is true that
new vinyl records are not made in significant numbers any more, there is a thriving second-hand market. In
fact many types of record are now hotly collected, including some that were manufactured quite recently. In-
terest in vinyl records may still be quite strong, but the drawbacks that resulted in compact discs taking over
have not gone away.
Noise caused by dust getting into the grooves is one problem, but with proper care and handling this can
be minimized. Physical damage to this very vulnerable form of recording is probably the main problem, and
there is no easy solution to this one. Most new vinyl recordings were supplied complete with a few “clicks” and
“pops”, and even when handled with due care they tend to gain some more over the years.
This stereo circuit provides a delay of less than one millisecond to the audio signal so that “clicks” can be
detected and removed before the listener hears them. Make listening to your old vinyl a pleasure again.

FLASHING SNOWMAN
If you wish to make an electronic project popular you give it some flashing LEDs, or you do if you believe
the in-joke that was popular in the electronic magazine publishing business some years ago. This joke came
about because one of the magazines now incorporated into EPE published a project that was basically just a
soap dish fitted with some LEDs that flashed. Apart from looking pretty, it did not actually do anything, but that
did not stop it from being by far the most popular project ever published by that magazine!
This project is very much in the flashing soap dish tradition, it is just a polystyrene ceiling tile fitted with
some LEDs that flash. It is a simple but amusing Christmas decoration that should raise a smile or two.
The tile is fashioned and painted to look like a snowman (or snowperson?), and it has the LEDs to form
the eyes, nose and mouth. The idea is to arrange the LEDs so that the snowman's expression alternates be-
tween an internet style smile and frown. This is just a suggestion, and there is plenty of scope for doing your
own thing. You could obviously use a different character such as Father Christmas as the basis of the project,

FROST ALERT
Vehicle manufacturers strive to make our driving experience ever more comfortable. Such items as ride
control, power steering, electronic engine management, and air-conditioning now appear as standard fit on
many vehicles and as a result we become more and more detached from the outside environment. One of the
most important considerations would seem to be the quality of the road surface, a factor that directly affects
our style of driving.
We generally notice if the road is wet, muddy, or covered in leaves and take appropriate precautions, but
we can't see black ice. For this we need an external temperature sensor, warning us when the likely conditions
for ice on the road are developing. A cheap solution is the LCD strip attached to the driver's wing mirror, but
despite looking very pretty (especially if you have heated wing mirrors …), it tells us nothing about conditions
down at road level. The Frost Box gets us right there, with a simple three-color LED indicator to keep you in-
formed.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1085
Maxfield & Montrose Interactive Inc
 $ VHOHFWLRQ RI (/(&7521,& %RRNV
IRU +REE\LVW DQG 6WXGHQWV
The Great Logo Adventure
A cartoon-illustrated, family activity book for
0RUH GHWDLOV DUH
exploring animation, graphics, math, geometry,
... Ideal for teaching programming concepts to DYDLODEOH RQ WKH
young people of all ages.
FREE CD-ROM (for PCs and Macs) contains (3( 2QOLQH ZHE VLWH
Logo software plus lots of other stuff.

Controlling the Forrest Mims

World With Your PC Engineer's
Connect your MS-DOS/Win- Notebook
dows PC to the real world with This revised edition
this best-selling book! includes hundreds of
Comes with all software useful circuits de-
(executable files plus C, Basic, signed and built by
and Pascal Forrest using com-
monly available inte-
grated circuits and
other components.

Bebop to the Bebop BYTES Back

Boolean Boogie (An Unconventional
(An Unconventional Guide to Computers)
Guide to Electronics)
This in-depth, highly readable, This follow-on to Bebop to the Boolean
up to the minute guide shows Boogie is a multimedia extravaganza of
you how electronic devices information about how computers work.
work and how they're made -- FREE CD ROM contains the
the only electronics book Beboputer Computer Simulator, along
where you can learn about mu- with over 200 megabytes of mega-cool
sical socks and the best time multimedia.
of the day to eat smoked fish!

9LVLW WKH (3( 2QOLQH 6WRUH DW

ZZZHSHPDJFRP DQG SODFH \RXU RUGHU WRGD\

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1086
Maxfield & Montrose Interactive Inc
 $ VHOHFWLRQ RI (/(&7521,& %RRNV
IRU +REE\LVW DQG 6WXGHQWV
Programming Micro- Designus Maximus
controllers in C More Unleashed
Info or Buy Now!
This book opens with a quick (Banned in Alabama)
review of the essentials of C
programming and then exam- This unabridged and unexpur-
ines in depth the issues faced gated tome contains the defini-
when writing C code for micro- tive collection of Clive "Max"
controllers. Maxfield's wildly popular arti-
cles published in leading elec-
tronics magazines. FREE CD-
ROM contains a logic synthe-
3D Graphics Super- sis tool, a digital logic design
computing on Win- system, & ...
dows NT
Explains the concepts behind
2D and 3D computer graphics.
In addition to easy to under- Simple, Low-Cost
stand text, the book is packed
with superb color graphics that Electronics Projects
fully illustrate the points behind More Info or Buy Now!
each topic. ($9.99 Special Of- Whether you're a beginner to
fer! -- While Stocks Last!) electronics or an old hand, don't
miss this new book of do-it-
yourself electronics projects.

Integrated Circuit
Hobbyist Handbook
More Info or Buy Now! 0RUH GHWDLOV DUH
DYDLODEOH RQ WKH
Work with CMOS or TTL digi-
tal ICs? Use op amps and

(3( 2QOLQH ZHE VLWH

other linear devices? Need an
application circuit using popu-
lar, readily available ICs?
Then this handy reference is a
"must have"!

9LVLW WKH (3( 2QOLQH 6WRUH DW

ZZZHSHPDJFRP DQG SODFH \RXU RUGHU WRGD\

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1087
Maxfield & Montrose Interactive Inc
 3,& 0,&52352%(
E\ -,0 0$,1
A neat little tool to help debug your PIC o) The code is held-up out-
microcontroller code! putting the Debug word for
a minimum amount of
If there’s one thing that irks and the author always try to in- time.
most when developing PIC micro- corporate a bank of them on a o) It may be possible to use a
controller software, it’s that sink- printed circuit board wherever pin that is currently being
ing feeling when you apply power possible. This gives you an on- used for other output du-
for the first time and the thing just board debugger straight away – ties.
sits there smugly doing but only if you have the port
pins to spare. This is chip- The second point needs
(apparently) nothing.
dependant, and remember some further explanation. Ide-
In fact, your PIC chip is prob- some of the smaller 8-pin PICs ally, this output pin will be
ably whizzing away inside some only have something like a processor sta-
loop or other, or resetting itself tus LED pin – or at least a pin
several thousand times a second. where a short duration word isn’t
You’ll be none the wiser. going to upset whatever is con-
nected to it (a relay, for in-
BUGS AND OPTIONS stance, should ignore a word of
less than 100 microseconds or
There are various solutions so). In this way, the
to debugging your code and debug word is trans-
indicating what’s going on parently output on the
inside your chip. You could pin.
be using a software simu-
The Micro-Probe de-
lator, such as MPSIM.
scribed here is connected
You might have some
to the target output pin
LEDs attached to a spare
and “listens” for any valid
port and be lighting them up
debug words coming from
at various stages of your code to
that pin. There are eight possi-
see how far you’re getting. You
ble Flag codes. It does not mat-
might, perhaps, have a serial port
ter whether the output pin is
in your project and be sending the
held high or low before the word
odd character to indicate position
is output – the probe will pick up
– you might even be fortunate one or two input/output pins to the pattern in each case.
enough to have an in circuit emu- begin with.
lator (you wish …).
All of these debugging meth- CIRCUIT
SINGLE PIN
ods are well and good, but suffer DESCRIPTION
from disadvantages. MPSIM re- SOLUTION
quires time and effort to set up, The Micro-Probe has a very
An answer to this situation simple circuit and uses the
and if you’re running on-chip pe- is to use an existing output pin
ripherals such as serial ports, ever-popular PIC16F84 in its
to output a very short duration 10MHz version. It is such a sim-
PWM, I2C, SPI etc., then its use Debug word (of around 64ms).
becomes very limited. ple circuit, in fact, that it does
There are two advantages to not really warrant a printed cir-
Using LEDs attached to a this: cuit board. Consequently, it has
spare port is an excellent idea,
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1088
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
4!
!>$ ! B[ B[
%F Q [
!$ 4$
4& 4%
F5
B!
$': $
=3<B B2
&
3# B#
3!
 !%`
!& $':
'
! P ?C3! B2!
7[ 7[
H!
3" ! =8j (
B2"
! ^
3$ 93! B$
!%` @93!&6($! $':
!% )
?C3" B2#
!'
>3 B1
!( !
>3 B1! B2$
B% B[ B[
>3
!
B1" $':
" !! 4& 4'
4&
4" >3 B1# B2%
!>$ ! #
C97>1< Q [
B1$
!"
B2&
B&
Q $':
B" !#
4# B2' 7[ 7[
$['
@?G5B [
7>4
%
7>4
Fig.1. Complete circuit diagram for the PIC Micro-Probe.

been designed onto a piece of circuit comes in via diode D2.

stripboard. This drops the incoming signal SOFTWARE
With reference to the circuit voltage level by the same
amount as D1 (so that you are The software for the Micro-
diagram in Fig.1, the PIC, IC1,
not over-volting the PIC with Probe is split into two parts – the
is set up in a crystal-timed con-
respect to the supply). R2 is a code run by the Micro-Probe it-
figuration with a 10MHz crystal,
pull-down resistor for the PIC’s self, and the Target code you
X1, and two 15pF capacitors,
input pin RA4. All other Port A have to add to the target applica-
C3 and C4. Note that because
pins are configured as outputs tion to output the debug words.
the PIC is to be operated at
10MHz, it must be configured in the software and can be left It should be noted that the
for the HS crystal option prior to unconnected. Target program has been written
it being programmed. Four bi-colour LEDs are specifically for use within pro-
connected to Port B via ballast grams that are intended to be as-
The 5V power supply for the
resistors R3 to R6. These de- sembled through MPASM
PIC is connected to the circuit
vices are actually two LEDs in (Microchip’s own assembler soft-
by means of two flying leads
one package, connected back- ware).
with test clips on the ends. The
clips allow the power to be ob- to-back across each other. De- The Target code cannot be
tained from the supply of the pending on the direction of the used with programs written in
circuit under test. Since it would current, the device illuminates TASM (the Shareware assembler
be very easy to mix up the po- either red or green, so eight sig- language used in many EPE and
larity when connecting into a nal indications can be displayed EPE Online projects). Nor can the
target circuit, diode D1 is in- using four of these little rascals. EPE PIC Toolkit (both Mk1 and
cluded to protect the PIC. If you prefer, you can re- Mk2) interpret the Include instruc-
place the bi-colour LEDs with tion embodied in the Target pro-
A power-on LED, D3, is
eight individual LEDs, remem- gram. Experienced programmers,
taken across the supply in se-
bering to add a ballast resistor however, should have no diffi-
ries with ballast resistor R1. You
to each. culty in re-writing the small
can then tell straight away if one
amount of code involved to suit
of the power clips has fallen off! When a valid word is re-
the TASM/Toolkit structures.
The signal from the target ceived by the Micro-Probe, one
or more of the LEDs will light.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1089
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
TARGET MACROS MACRO ROUTINES through a debug word, destroy-
ing its timing. Re-enable it as
Looking at the Target code There are three distinct required afterwards (see your-
first, this allows you to add Macros: PIN X, SYNCWORD and prog.asm).
macro-routines to your program. DEBUG X.
Macros are very powerful and When you want to reset and
At the lowest level is the turn off all the LEDs, just re-
flexible batch-type commands,
macro PIN X. This takes an argu- move one of the power leads
which consist of instructions to
ment of 0 or 1 and sets or clears temporarily (or fit a Reset switch
the compiler to generate code
the selected output pin accord- if you like).
at compilation time. An example
ingly. It then loops for a number
of using the macros to generate
of times according to the clock
the debug words is shown in INCLUDE FILES
frequency (defined by CLK) to
program file yourprog.asm.
time the length of the pulse. Keep things tidy by putting
First of all, it is important to the body of all the aforemen-
The macro SYNCWORD
be able to generate the correct tioned macros into an Include
starts with a 0 for the start bit, fol-
duration of pulse for a number file. To do this, open up a blank
lowed by binary 101 to uniquely
of target clock frequencies. The page within MPLAB, type in the
identify that this is a debug word.
macros generate the correct du- Macros, and save as: de-
It does this by calling PIN X four
ration of pulse for an integer bug.inc into c:\program
times (e.g. PIN 0, PIN 1, PIN 0,
number of megahertz frequency
PIN 1). files\mplab
(1, 2, 3 etc.). It is necessary to
point out that your target pro- Macro DEBUG X is the one When you want to add de-
cessor should be crystal or ce- you call from the body of your bug code to an application, just
ramic resonator clocked – RC code with the relevant argument, put the command:
(resistor-capacitor) clock gener- where you want to signal that the
code has reached that particular INCLUDE DEBUG.INC
ation is not really accurate or
stable enough for the Micro- point. DEBUG first calls SYNC- below your processor-
Probe. WORD, and then adds the 3-bit specific Include line, and then
In yourprog.asm it will be code for the relevant Flag point. add the CLK and DEBUGPIN
seen that the clock speed (CLK) Finally, a stop bit 0 is added to definitions.
is defined for 16MHz: the end.
When you want to take out
or disable the debug code gen-
#DEFINE CLK .16 ;SPEED IN
MHz
USING MICRO-PROBE eration, then just “comment-
out”, with a semicolon as usual,
Where you want to signal a
Note that a decimal point is the Include line (as well as com-
point having been reached in
placed in front of the 16, which menting out the various invoca-
your code (say entering a subrou-
signifies to the compiler tions of DEBUG in your code).
tine), then add the line:
(MPASM or compatible), that The CLK and DEBUGPIN
the value is in decimal. The ap- DEBUG X (where X = 1 to 8) definitions can safely stay in
propriate value for the speed of your application.
your target circuit should be For example:
substituted in place of the 16. DEBUG 1 lights the first LED
The pin of the target circuit green
BIT-BANGING
which the Micro-Probe is to DEBUG 2 lights the first LED red The Micro-Probe works by
monitor is defined in your- what is known as “Bit-Banging”
DEBUG 7 lights the fourth LED
prog.asm as Port C pin 7: – that is, it constantly samples
green
an input pin (RA4) and looks for
#DEFINE DEBUGPIN If your target application changes in its logic states. To
PORTC,7 makes use of interrupts, then do this, you have to time the
make sure that you disable the instructions carefully so that you
Any Port and any pin can be
global interrupt enable bit are always sampling in the cor-
substituted in place of PORTC,7
(INTCON,GIE) before calling the rect part of the incoming bit
as required.
Macro. This is to ensure that an (interrupts are of no use here
interrupt does not happen halfway because of the short duration of

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1090
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
? > = < ; : 9 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 : ; < = > ?
the incoming pulse train).
! ! 4' 4%
7[ 7[
When the level is unchang-
B[ B[
ing, then a sample loop occurs
% % B
every ten instruction cycles (1 7[ 4& 7[ 4$ $ 4&
B
cycle = 0 4ms at 10MHz), or ev-
¬ & B B[
B
B[ %
ery 4ms. It compares this level #

with the previous sampled level ! !

by XORing them together. If the
result = 1, then a change in 93! [
H!
level has occurred. !% !% 3 B" C97>1<
$ 3 Q 9>@ED
4"
Assuming that the output #
3
started off low, then the first " 3!
sample will occur somewhere " " 
B
inside the first “1” of “101” (the ! 4!
Q 4# [ 7>4
Syncword). The pin is re-
sampled six cycles later (2 4ms) ¬
%F
"% "%
to make sure that the sample
point is not too near the leading Fig.2. Component layout and stripboard track details for the
edge of the first pulse. There-
after, the pin is sampled every PIC Micro-Probe.
8ms to sample each pulse in the
same place. CONSTRUCTION
COMPONENTS If the Syncword is wrong, Stripboard is used for the
Resistors then the process is abandoned Micro-Probe construction. The
R1, R3 to R6 470 ohms (5 off) and the sampling process starts
R2 4k7 type used in the prototype is
All 0.25W 5% carbon film or better
from the top. that specifically designed for
Capacitors
Once the three bits of data mounting integrated circuits and
C1 100u radial electrolytic, 16V have been obtained, we have a which has a break running up
C2 100n ceramic number between 0 and 7. This the middle. If using ordinary
C3, C4 15p ceramic (2 off) is multiplied by four (by per- stripboard, cut the tracks appro-
Semiconductors forming the RLF instruction priately to keep the two sides of
D1, D2 1N4001 1A 50V rectifier twice) before being added to the the PIC isolated from each
diodes (2 off) program counter (a computed other.
D3 red LED, 3mm
D4 to D7 bi-color (red/green) GOTO). The component layout and
LEDs (4 off) Using as an example the underside track view are shown
IC1 PIC16F84-10P (10MHz
situation where the data is 0, 4 x in Fig.2.
version), preprogrammed
0 = 0, which is then added to It is likely that the stripboard
Miscellaneous the program counter (PCL). The
X1 10MHz crystal will be larger than you need. If
program counter always points so, use a sharp knife to score
Stripboard, with central separating to the next instruction that is to the stripboard where you want
strip, 25 strips x 15 holes (see text); be performed, so adding 0 to it to cut it (on the copper side). It
plastic case, 75mm x 56mm x just results in the next instruc-
25mm; 18-pin DIL socket; nuts should crack cleanly over the
and bolts, M3 x 20mm, countersunk
tion being performed as normal. score when you bend it with a
(2 off); spacers, M3 x 6mm (2 off); In this case that means that bit pair of pliers. File down the
probe clips, miniature, one each of 0 of LED_REG is set, and bit 1 rough edges.
red, black, yellow; extra-flexible is cleared before jumping for-
wire for probes; connecting wire, ward to the Port writing section. Drill two 3mm mounting
solder, etc. holes in the positions indicated,
In the Port writing section and make the various breaks in
See also the (LED_EXIT), the shadow regis-
SHOP TALK Page! the copper using the same drill
ter LED_REG is written to Port bit.
B and lights the relevant LED
Approx. Cost Use an 18-pin DIL socket
Guidance Only $30 before looping back to the top.
(turned-pin is best) for the PIC.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1091
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
It’s preferably to use insulat- through the case and tighten the
ing sleeve on all the wire links grommet to clamp them.
to prevent shorts and then sol- The lid of the case is drilled
der in all the components, ob- with 3mm holes to line up with
serving polarity for the diodes the LEDs coming up from un-
and the electrolytic capacitor. derneath. If you gauge the
The bi-colour LEDs have the height of the LEDs correctly,
red anode denoted by the then with the lid on the box,
longer lead, so make sure the they will protrude slightly above
short lead goes to the position the surface. Make a paper tem-
marked as Rk on the layout dia- plate with the positions of the
gram. LEDs on it and tape this to the
You can’t damage these lid prior to drilling.
LEDs by getting them the wrong Put both bolts through the
way round, but your colors will bottom of the box and slip on
be reversed. the 5mm spacers, followed by
At first, only solder in one the stripboard. Thread 3mm
When you run the target
lead of each LED, so that you nuts onto the bolts and tighten.
processor, the first LED (D4)
can adjust the height to fit the Put the lid on the box, guiding
should light green on the Micro-
box before soldering in the the LEDs through the holes and
Probe. Check operation for the
other one. fasten using the screws supplied
other seven Debug states.
with the box.
Finally, label the front panel
Solder spring loaded test
ENCLOSURE clips onto the ends of the wires
and your Micro-Probe is ready
Use a small plastic case for for action!
now protruding from the box
housing the Micro-Probe. Drill and you have yourself a com-
two holes in the bottom of the pleted Micro-Probe! RESOURCES
case for the stripboard’s mount-
ing screws. Use a countersink Software for the Micro-
tool so that the countersunk TESTING Probe is available free from the
bolts will sit flush with the sur- Power the unit from the tar- EPE Online Library at
face. get board using the power clips www.epemag.com
Drill a 2mm hole in one end and attach the signal probe to
for the signal wire, and a 5mm the required pin. The power-on
hole in the other end for the LED should be illuminated, if
power leads to pass through via not check your connections and
a clamping grommet. (When circuit.
you mount the stripboard, you With the Debug Include file
may need to file its top end to in your default directory, put the
clear the grommet.) clock and pin definitions into
Solder the leads directly your code, as discussed earlier.
into the board at the positions Enter a Debug command (e.g.
indicated. Pass the leads DEBUG 1) into your code.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1092
Maxfield & Montrose Interactive Inc
 0$*1(7,& ),(/' '(7(&725
E\ 52%(57 3(1)2/'
You will find the attraction with this novel, low-cost top, and a certain portion of the
starter project. supply voltage at intermediate
points. The two electrodes are
This very simple project can HALL EFFECT half way up the slice, and con-
detect fixed magnetic fields or sequently there is half the sup-
Detecting varying magnetic ply voltage at each one. This
fields that are varying at an audio fields is quite easy, and requires
frequency. Fixed or slowly chang- gives zero output voltage
nothing more than an inductor across the two electrodes.
ing field strengths are registered to act as the sensor. Unfortu-
on a center-zero meter, which in- nately, static fields do not pro- Applying a suitable mag-
dicates the polarity in addition to duce any output from an induc- netic field to the device “skews”
the relative field strength. Audio tor and require a totally different the current flow and the poten-
frequency fields, such as those approach. tial gradient, producing an im-
produced around mains and au- balance in the output potentials.
dio transformers, are detected via The only common form of The stronger the magnetic field,
a crystal earphone that can be magnetic sensor that “fits the the greater the difference in the
used to monitor the output signal. bill” is a linear Hall effect de- output voltages.
vice. A Hall effect sensor is a
The unit is not intended to form of semiconductor, and is Applying a magnetic field of
provide accurate measurement of actually a very simple type of the opposite polarity skews the
magnetic field strength, and is component. Fig.1 helps to ex- current flow in the opposite di-
aimed at those who like to experi- plain the way in which a Hall rection, giving an output signal
ment with something a bit differ- of the opposite polarity. The
3EBB5>D 6<?G 3EBB5>D 6<?G output signal therefore indicates
the strength of the magnetic
field and its polarity.
5<53DB?45 5<53DB?45
It is important to realize that
a Hall effect sensor only works
?ED@ED ?ED@ED
F?<D175 F?<D175 if the magnetic field is applied
to one side or the other of the
silicon slice. Applying the field
to the front, back, top, or bottom
1 2 of the sensor does not affect the
Fig.1. A Hall effect sensor is little more than a current flow in a manner that
slice of silicon. (a) normal and (b) with magnetic will produce any imbalance at
field influence). the electrodes. Consequently
it will not produce any
ent. Although quite simple, the effect device works. output voltage.
unit is reasonably sensitive. A The sensor is just a slice of
small and not very powerful bar
magnet can be detected by the -
silicon having electrodes on op- SENSOR
posite surfaces. A current is
prototype at about 100mm from Practical Hall effect sensors
passed through the silicon, and
the sensor, and drives the read- are more than just the sensing
this produces a potential gradi-
ing to full scale at a range of element itself, and they are in-
ent in the silicon. There is zero
about 30mm. variably in the form of and inte-
volts at the bottom of the slice,
grated circuit containing the
the full supply potential at the
sensor plus some additional cir-
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1093
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
B54E354 9>3B51C54
?ED@ED ?ED@ED signal from the sensing element
and produces a single output
=17>5D =17>5D
that is at about half the supply
C > potential under standby condi-
tions.
Placing a north pole of a
magnet close to the surface of
the sensor that carries the type
number produces a reduction in
> C the output voltage, and placing
DI@5 a south pole close to this sur-
>E=25B face gives an increase in the
output potential (Fig.2). The fre-
C5>C?B C5>C?B
quency response of the device
Fig.2. A Hall sensor indicates is flat from DC to 23kHz, which
the polarity of the field as well means that it encompasses the
as its strength. full audio range.
Magnetic Field Detector front
cuitry. Some sensors provide a
switching action, and others pro- CIRCUIT OPERATION panel layout. The Hall effect
vide an output voltage that is The full circuit diagram for sensor is mounted externally
proportional to the applied field the Magnetic Field Detector ap- in a probe arrangement, such
strength. pears in Fig.3. IC1 is the Hall as an old pen case, and con-
In this application, it is only effect sensor and IC2, a preci- nected to the circuit board via
devices in the second category sion opamp, is used to provide the screened cable.
that are of any use, and the de- some additional amplification.
vice chosen for this design is The amplifier is an operational
typical operational amplifier.
the UGM3503U. This is an inex- amplifier inverting mode circuit,
which has resistors R1 and R4 Using negative feedback re-
pensive device but it has a very
as the negative feedback net- duces the voltage gain of the cir-
useful level of performance and
work. cuit as a whole to a more usable
is very easy to use.
figure, and this “closed loop” gain
It has just three terminals, The innate voltage gain of
is equal to resistor R4 divided by
which are the supply and output IC2, or the “open loop” gain as it
R1. This works out at a little over
terminals. An internal differen- is termed, is extremely high at
300 in this case. Higher voltage
tial amplifier boosts the output DC and low frequencies. In fact,
gain would obviously give better
it is over 100,000 times for a
sensitivity, but it would also give
problems with noise and drift.
C!

B" ?>?66
Opamp IC2 amplifies the
! [ voltage difference between the
21<1>35
B%
##[
input voltage to resistor R1 and
@B?25
FB! g the voltage at its non-inverting
![ =5!
input (pin 3). This second voltage

'

F
# ! P1 ! P1
B!
![%
93" &  2! can be adjusted via potentiometer
93! ?ED " ?@''7@ &F
$ h 11
VR1, and in practice it is adjusted
E7>#% #E

B$ $
B&
to produce a voltage that
7>4
$' [
##[ matches the normal output poten-
3"
tial from IC1. This produces half
 ! ^
the supply potential at the output
3!
51B@8?>5
! P of IC2.
B#
! [ C;!

The potential divider formed

by resistors R5 and R6 also pro-
duces an output of half the supply
potential. Meter ME1 is con-
Fig.3. Complete circuit diagram for the Magnetic Field Detector.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1094
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
COMPONENTS STRENGTH OF CHANGE the meter under standby condi-
tions.
Resistors In both cases the greater the
change in the output voltage from The opamp also needs to
R1 1k5
R2, R3 10k (2 off) IC1, the higher the reading from be able to work properly with a
R4 470k the meter. The meter therefore supply potential of just 6V. The
R5, R6 33k (2 off)
indicates the relative field OP077GP is reasonably priced
All 0.25W 5% carbon film
strength and the polarity of the and gives good DC perfor-
Potentiometer magnetic field. mance in this circuit. On the
VR1 1k rotary carbon, linear other hand, its open loop band-
Applying a “North Pole” close width of 600kHz equates to a
Capacitors
C1 100u radial electrolytic, 10V to the surface of the sensor that closed loop bandwidth of only
C2 100n polyester, 5mm lead carries the type number produces about 2kHz in this design.
spacing a positive reading, and applying a
Semiconductors “South Pole” to it generates a If audio rather than DC per-
IC1 UGN3503U Hall effect sensor negative reading. This may seem formance is of most importance
IC2 OP77GP precision opamp
to be at odds with Fig.2, but bear it would be advisable to use a
(see text) TL071CP for IC2. This will give
in mind that IC2 inverts the sig-
Miscellaneous nal. quite good DC performance
S1 s.p.s.t. miniature toggle switch plus a more respectable audio
B1 6V (4xAA cells) battery pack, The value used for resistors bandwidth of around 10kHz. To
with holder and PP3-type R5 and R6 controls the sensitivity
battery connector. compensate for a lack of sym-
SK1 3.5mm jack socket
of the meter circuit. The specified metry in the TL071CP’s output
ME1 100uA - 0 - 100uA moving values permit ME1 to be driven to stage resistor R6 should be re-
coil panel meter full scale in both directions pro- duced from 33 kilohm to 27
vided the battery is reasonably kilohm.
Medium size plastic or metal box; fresh, but their value is high
0.1 inch matrix stripboard, size 19
holes by 20 copper strips; 8-pin
enough to prevent the meter from
DIL socket; control knob; crystal suffering anything more than very CONSTRUCTION
earphone, with lead and plug; twin- minor overloads. The stripboard layout for the
screened cable, about 0.5 meters;
multistrand connecting wire,
Capacitor C2 couples the out- Magnetic Field Detector is
solder pins; solder, etc. put of IC2 to earphone socket based on a piece that measures
SK1. This enables the output sig- 19 holes by 20 copper strips.
See also the nal to be monitored using a crys- The component layout and in-
SHOP TALK Page! tal earphone, but satisfactory re- terwiring, together with the posi-
sults are unlikely to be obtained tions of the breaks in the copper
Approx. Cost using any other type of earphone strips, are shown in Fig.4.
Guidance Only $26 or with headphones. A board of the required size
(Excl. earphone, case, & batts)
A 6V battery supplies power must be cut down from one of
to the circuit, and the current con- the standard sizes in which it is
nected between the output of sumption is only about 9mA. Do sold. The holes are very close
IC2 and this potential divider, together so use a hacksaw to
not use a 9V battery as this
and it therefore responds to the cut along rows of holes rather
would result in the maximum
voltage difference between the than trying to cut between them.
supply voltage rating of IC1 be-
two. Under standby conditions This inevitably produces quite
both points will be at the same ing exceeded.
rough edges but they are easily
potential, giving zero voltage filed to a neat finish. Next, drill
across the meter. An increase in GOOD PERFOR- the two 3mm diameter mount-
the output voltage from IC1 pro- ing holes and make the four
duces a decrease in the output MANCE
breaks in the copper strips. A
from IC2, and a negative de- In order to produce good re- special tool for cutting the strips
flection on the meter. A de- sults in this circuit it is necessary is available, but a handheld
crease in the output potential for the operational amplifier to twist drill bit of about 5mm in
from IC1 has the opposite ef- have good DC performance. Oth- diameter does the job just as
fect, producing a positive indi- erwise there could be major prob- well. Make sure that the strips
cation from th emeter. lems with drift, and DC offsets are cut across their full width.
could make it impossible to zero
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1095
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
! % ! !% !)
The circuit board is now 21<1>35 1
2 2!
ready for the components and 3
B54
link-wires to be added. With a 4
small board such as this the or- 5 
6 C!
der in which the components 7
8 B
are fitted is not really important, " 2<13;
9
3! B
but it is best to work methodi- :
;
$ ?>?66

FB!
cally across the board so that =
<
93"
nothing is overlooked. > 3 B
? %
B "
Neither the OP077GP or


@ !
TL071CP is static-sensitive, but A B
B
F # =5!

it is a good idea to use a holder D? 93!

?ED
C B&
@B?25 D
for any DIL integrated circuit. 7>4

Be careful to fit IC2 and elec- C3B55> D9@ C;!

trolytic capacitor C1 the right 51B@8?>5

way round. B9>7

Fit single-sided solder pins

at the points where connections ! % ! !% !)
will be made to potentiometer D
VR1, meter ME1, etc. It is one- C
B
millimeter diameter pins that A
@
are required for stripboard. “Tin” ?
the pins with plenty of solder so >
=
that it is easy to make reliable <
connections to them. ;
:
9
8
CASING-UP 7
6
5
Virtually any medium size 4
plastic or metal case should be 3
2
able to accommodate this pro- 1
ject. However, be careful to
choose one that has sufficient Fig.4. Stripboard component layout interwiring and details for
depth to take the meter and the breaks required in the underside copper tracks.
battery pack. The latter consists
of four AA size cells in a plastic and these are available from are easily located as they are at
holder. Connections to the many DIY superstores. the corners of a 32 millimeter
holder are made using an ordi- square having the same center
nary PP3 style battery clip. Al- Alternatively, it can be cut
using a fretsaw, coping saw, or as the main cutout.
though the circuit has a fairly
high voltage gain the layout is miniature round file such as an
not critical, and it is just a mat- “Abrafile”. Another method INTERWIRING
ter of designing a layout that is would be to mark out the cutout,
drill a series of small holes just The hard wiring is reason-
easy to use. ably straightforward. SK1 is a
inside this mark and then “join-
One slightly awkward as- up” the holes to form the re- 3 5mm jack socket, and most
¬

pect of construction is fitting the quired cutout. With any of these sockets of this type have a built-
meter onto the case, because methods it is advisable to cut in switch that is not required in
this requires a large cutout to be just inside the perimeter of the this application. Accordingly,
made in the case. Most moving required cutout, and then en- one tag of SK1 is left uncon-
coil meters require a 38mm large it to precisely the required nected.
round mounting hole and the size using a large round file. The Hall sensor (IC1) is
easiest way of making this is to mounted externally and con-
use an adjustable hole cutter Four smaller (3mm diame-
ter) mounting holes are also re- nected to the main unit by way
(also known as a “tank” cutter), of a piece of twin-screened ca-
quired. The positions of these
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1096
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW

Completed detector showing

the earphone socket on one Layout of components inside the two halves of the case. Note
side panel. the space for the battery pack.
the sensor to the screened lead As explained previously,
and use insulation tape or applying the pole of a magnet to
sleeving to ensure that the sol- one of the four smaller surfaces
@?C9D9F5
dered joints cannot short-circuit of the sensor will not produce a
CE@@<I together. The sensor will be significant output signal. In use
7B?E>4  F neater if it is built into a probe, the orientation of the sensor
?ED@ED based on an old pen for exam- should therefore be adjusted to
Fig.5. Connection details for ple, but this is not essential. maximize the meter reading.
the UGN3503U Hall effect Placing the probe against
sensor. TESTING the power cable of virtually any
mains powered device that is
When the unit is first switched on should produce a
ble about 0 5 meters or so in
¬
switched on it is likely that the
length. An entrance hole for the meter will be driven fully posi-
cable must be drilled at a strate- tive or negative. With careful
gic point in the case, and if a adjustment of Balance control
metal case is used the hole VR1 it should be possible to
should be fitted with a grommet zero the meter, and placing the
to protect the cable. The screen probe near any magnetized ob-
is used to carry the ground (0V) ject should then produce a suit-
connection. able response from the meter.
Rather confusingly, the The meter movement itself
plastic encapsulation of the contains a permanent magnet,
UGN3503U Hall effect sensor and placing the probe near this
chip seems to be completely should produce full-scale de-
symmetrical. The only way of flection of the meter. Placing
identifying the three leads is to the opposite face of the probe
use the type number on the near the meter should then pro- Completed circuit board show-
body of the device as a refer- duce full-scale deflection in the ing the four link wires and the
ence point, see Fig.5. Connect opposite direction. opamp C2 mounted in its
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1097
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
50Hz “hum” from the earphone. ter the average field strength. The circuit is reasonably
Alternating fields will not pro- This will normally be zero due to stable, but occasional readjust-
duce an indication from the me- the opposite poles in the signal ment of VR1 will be required.
ter because the meter will regis- canceling out one another.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1098
Maxfield & Montrose Interactive Inc
 ROLL-UP, ROLL-UP! Win a Pico PC-Based Oscilloscope
Ingenuity is our regular round-up of readers' own • 50MSPS Dual Channel Storage
circuits. We pay between $16 and $80 for all material Oscilloscope
published, depending on length and technical merit. • 25MHz Spectrum Analyzer
We're looking for novel applications and circuit tips, not
simply mechanical or electrical ideas. Ideas must be the • Multimeter
reader's own work and must not have been submitted • Frequency Meter
for publication elsewhere. The circuits shown have • Signal Generator
NOT been proven by us. Ingenuity Unlimited is open to
ALL abilities, but items for consideration in this column If you have a novel circuit idea which
should preferably be typed or word-processed, with a would be of use to other readers, then a Pico
brief circuit description (between 100 and 500 words Technology PC based oscilloscope could be
maximum) and full circuit diagram showing all relevant yours.
component values. Please draw all circuit schematics Every six months, Pico Technology will be
as clearly as possible. awarding an ADC200-50 digital storage oscil-
loscope for the best IU submission. In addi-
Send your circuit ideas to: Alan Winstanley,
tion, two single channel ADC-40s will be pre-
Ingenuity Unlimited, Wimborne Publishing Ltd., Allen
sented to the runners up.
House, East Borough, Wimborne, Dorset BH21 1PF.
They could earn you some real cash and a prize!

The power for the charging (220 ohms in parallel with 27

Rechargeable PP9 Battery circuit is provided by an ohms will do).
– Energy Saver external 12V to 15V DC power Six 1 2V NiCad cells are
¬

If you discard exhausted PP9 supply capable of providing

placed in series and wired
layer-type batteries, this can 50mA or so. This is hooked up
across the PP9 battery
become an expensive process as via a DC power socket SK1
terminals and they will be
these batteries cost about three which is also fitted into the
charged by the constant current
pounds each. It was decided to battery housing.
of IC1. The on-load voltage of a
provide an alternative using In this circuit IC1 is fully charged set of six cells was
Nickel Cadmium cells together configured as a constant-current measured at just under 10V with
with an extremely simple (not voltage) regulator, and the an average current of 25mA
charging circuit, which is built current flowing is limited by the being drawn.
within the housing of an series resistor R1. The current I
exhausted PP9 battery. The A steel-cased PP9 should
is 1 25/R1, hence for a 50mA
¬
be prized apart and its contents
circuit diagram is shown in Fig. 3. current R1 is about 24 ohms carefully disposed of as
chemical waste, then the circuit
built inside and the case folded
B! 9 - % ]1 )F @@)
C;! 9>
93! ?ED
5=@DI
back together again. My present
<=#!'D 31> ?><I rechargeable PP9 has
14: 2!
undergone about 40 charging
43 9> & h !f" cycles during its existence and I
>Y3T
anticipate many more – what a
B! "$ : 9 -
! "%F
B saving!
D. Allen
Fig.3. Rechargeable PP9 circuit. Note all components are Cheltenham
housed inside the discarded PP9 casing.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1099
Maxfield & Montrose Interactive Inc
 ,QJHQXLW\ 8QOLPLWHG

serial port of the PC (and also disconnected from the PC.

Serial Port Splitter – Line from devices X and Y) to TTL/ When Select is at logic 1,
Sharing CMOS levels for manipulating signals on port B are routed to
When you run out of spare by IC3, a data selector/multi- the output port instead. In this
serial ports on your PC, the cir- plexer. Signals on the two sets case, Tx is connected to Rx2,
cuit shown in Fig. 1 may be of inputs (A0 to A3 and B0 to Rx to Tx2 and Rx1 is held at
used to effectively add another B3) are selected and routed to logic 1 so the PC communicates
port. The idea is to share the the output (YO to Y3) by the Se- with device Y.
PC serial port between two ex- lect input (pin 1 of IC3).
Switching between device X
ternal RS232 devices (device X When Select is at logic 0, and device Y is controlled by
and device Y in Fig. 1) and the signals on IC3 port A are routed the RTS signal from the PC se-
PC communicates with them to the output port (Y0-Y3). In rial port. RTS can be toggled by
one at a time. this case, the Tx of the PC se- a piece of simple software,
In the circuit diagram shown rial port is connected to Rx1; Rx which configures the control
in Fig.1, IC1 and IC2 are the to Tx1 and Rx2 is held logic 1 registers of the UART chip in
familiar MAX232 voltage level (idle condition). The PC there- the PC.
translators which convert the fore communicates with device
RS232 signal levels from the X, and device Y is effectively W. Ip, Belfast

 9

 
 

& &
CC CC
µ µ
9 9

 

   
9 9

   

& 9 & 9

& &

 µ 
µ 
 9  9
9² 9²
 

,& ,&
& ² & ²

 
& &


 µ 
 µ
& &
9  9
0$; 0$;
& &

 µ  µ
9 9 6.
 

& ² & ² 6(5,$/ 3257

2) '(9,&( ;


 
77/
56
56  77 /
&026 287
& 026 2 87
6(/ 9 CC  ,1  
  ,1   7[
576 $

$


$
77/ 56
56  77 /
  &026 ,1 87  

  ,1 & 026 2 87 
$ < 5[
7[

,&
 

< * 1'


+&$
< 1&

 

% <
77/  77/ 56


&026 ,1   &026 ,1 87  

1 & % 5[


%



%


*1' * 1' (


7[
 

77/  * 1'
56

 &026 ,1
 28 7 

5[ 77/
56

&026 287 6. 
 ,1 

6(5,$/ 3257

2) '(9,&( <
6.

3& 6(5,$/
*1' *1'
3257

'% & 211(& 725  

9

Fig.1. Circuit diagram for the Serial Port Splitter. Note that pin 6 of IC1 and IC2 is at – 10V
with respect to the 0V line.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1100
Maxfield & Montrose Interactive Inc
 ,QJHQXLW\ 8QOLPLWHG

 D85C5 3?=@?>5>DC
1B5 38?C5> D? 79F5
254B??=
Class-D 30W Audio Am-
!8j CAE1B5 G1F5
1D @9> # ?6 93!
$ (
plifier – Power Play
C9DD9>7
B"  B# BCD F
B??=
"['

!&
Audio amplifiers are typi-
'
@B5CCEB5 @14
C8?G> B!
49C
F
2!
&F cally class-AB in operation, and
#!
?@5B1D54

C!
&[(
 B$ 93! ?ED 3<; whilst these produce good qual-
%%%

&
A&'
"$
ity amplification they are also
D8B

"
DB97
93" quite inefficient at 50 to 60 per-
=93B?CG9D38
?@5B1D54 2I F
'$83D$ $
cent or so. A class-D amplifier is
3?==?45 ?B
21D8B??= DB! ! !!
BCD much more efficient, with effi-
4??B 23%$(
F
C"
"G9B5
312<5 R
S
(
G4!
ciencies of between 90 percent
U
3! @95J? to almost 100 percent being
C?E>45B
possible as it is essentially a
switching circuit.
Fig.2. Circuit diagram for an Elderly Person Monitor. A suggested circuit diagram
for a 30W Class-D Audio Ampli-
the counter is enabled and begins fier is shown in Fig.4. The in-
Elderly Person Monitor – counting. If the time taken to get coming audio signal is amplified
Take Care from the bed to the commode or by the inverting operational am-
bathroom (where a seat or door- plifier IC1, with adjustable vol-
An elderly relative who re- ume controlled by potentiometer
sides with us occasionally falls activated microswitch, S2, auto-
matically resets the counter again) VR1. A PWM (pulse width mod-
accidentally, and has laid there ulation) signal is produced by
for some time in a distressed is long enough for counter output
Q6 (or Q7 perhaps) to go high, the comparing the audio signal with
state without being able to sum- a 100kHz triangle wave. This is
mon help. Consequently, a sim- alarm WD1 sounds in a neighbor-
ing room so that one can investi- achieved using the comparator
ple independent alarm was de- IC6. Resistor R13 is used to
signed and the resulting circuit gate and check that the person is
all right. provide positive feedback and
is shown in Fig.2. C6 is a speed-up capacitor,
Unless a “reset” operation is A delay of between one and which improves comparator re-
applied before a certain time two minutes was selected to allow sponse time. The comparator
period has elapsed the alarm the elderly person sufficient time output swings between †7 5V.
¬

will automatically sound. The and also because in practice the

The pull-up resistor R12 pro-
principle of operation can be microswitch S2 wasn’t always oper-
vides +7 5V whilst –7 5V is pro-
¬ ¬

adapted as required and may ated. In our case the switch opens
when the person leaves the com- vided by the open emitter input
inspire other ideas. of the comparator (pin 1 of IC6).
mode, and so IC2 begins counting.
While the person is in bed a When this signal swings
pressure pad (S1) under the If the time taken for the person
returning to bed (which resets the positive, transistor TR1 acts as
mattress is held in the closed a current sink, which increases
circuit condition. This maintains counter) is again long enough for
the alarm to sound, then that per- the voltage drop across resistor
the 4040 12-Stage Ripple R16; this voltage drop is enough
Counter IC2 in its reset state via son is standing up, or returning to
bed or has fallen. Since an elderly to turn MOSFET TR3 on. When
transistor TR1 and so the piezo the signal swings negative, TR2
sounder WD1 is disabled. person is unlikely to remain stand-
ing for more than (say) two minutes acts as a current source causing
Clock pulses of approxi- the voltage drop across R17 to
and is also unlikely to take more
mately 1Hz frequency are fed increase sufficiently to turn TR4
than two minutes to return to bed, it
continually from the 555 timer on. Essentially, MOSFETs TR3
is probable that the person has
IC1 (pin 3) to the counter input and TR4 are activated alter-
fallen.
of IC2 at pin 10 (CLK), but have nately, producing a PWM sig-
no effect until the person gets The prototype operated from a nal, which swings between plus
out of bed (in our case, to use a safe 6V battery, which could be and minus 15V.
commode but it could be rechargeable.
It is now necessary to re-
adapted to be reset by, say, a store this amplified PWM signal
C. Embleton
bathroom door) at which point back into a reproduction of the
Northallerton, N. Yorks
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1101
Maxfield & Montrose Interactive Inc
 ,QJHQXLW\ 8QOLPLWHG

incoming audio signal. This is

?ED@ED
achieved by averaging out the

# G
:
<C

!%F
(

F
 !%F
PWM signal using a 3rd order
Butterworh low-pass filter with
its cut-off frequency (25kHz)

8
<"

P "%



much lower than the triangle

P
3!"
3!!

!
wave frequency, ensuring large

3!#
P !



'& 8
P
3)

<!

P
3!
!
attenuation at 100kHz. The re-

!
sulting output is an amplified

^
^

3(
3'
reproduction of the input audio

!
!
T

T
c

c
signal. The triangle wave gener-

DB$
9B6%"
DB#
9B6)%#

W
ator is based around IC2 and

:
B!(

("
U
S

S
IC5, whereby IC2 is effectively

:
:

23#"'
DB"

B!'
B!&

!"
!"
a square wave generator with

DB!
23##'

R
positive feedback provided by
R7 and R11. Diodes D1 to D5

B !%

FB#
! [
![
acts as a bi-directional clamp

B!$
![

1=@<9DE45
DB91>7<5
[
B!#
(D3 being a Zener diode),

`
3&

G1F5
!
!
B!"
![

'
clamping the voltage to about

B!!

(["
!
†6V. An ideal integrator is
<=#!!
93&
(

$


&
## `
3%
#

"
formed by preset VR2, capaci-

<6#%%>
93%
tor C5 and IC5, which converts
' %F


'

$
""[
B!


"

#
a square wave into a triangle
&

wave. Preset control VR2 allows

93$

6B5 AE5>3I
'$!

DB91>7<5

FB"
'

!>$!$(
!>$!$(


G1F5

4$

4%
!
the frequency to be altered.
#

"

Q


[
The output of IC5 (pin 6)
3$
P !

Q
%F!
4#
provides feedback to IC2, and

[
! [

! [
B(

B)

' %F

resistor R14 and preset VR3

P

Q
[

[
!"[
B'
3#

!>$!$(

!>$!$(
!
&

4!

4"
form an adjustable attenuator

:
B&

!%F
 !%F

F
&(
93#
'$!

allowing the magnitude of the

'

$
 #

"

triangle wave to be adjusted.

B%
![

'
!%F
!%F

After construction, VR2 and

F?<E=5

!
<=#!!
93"
[
B$

$
VR3 should be adjusted in order
FB!


!
! [

&

"

#
! [
B#

to provide the best quality out-

14'$$
93!
'

put. A pair of ordinary 741


"

op.amps (IC4 and IC3) are used

:
B"

as unity gain buffers to provide

&'

^
3"

the plus and minus 7 5V sup-

^
3!

¬
!

B!
![

plies. Capacitors C3, C4, C11,

1E49?
9>@ED

and C12 act as charge reser-

C; !

voirs, and the remaining capaci-

tors are for decoupling. The cir-
cuit requires a plus and minus Fig.4. Complete circuit diagram for the Class-D 30W Audio
15V supply rail, and it will drive Amplifier.
a 30W 8 ohm loudspeaker from
circuit consists of two CMOS
the LC network at capacitor C13 National Lottery Predictor 4017 decade counters each
and inductor L2. Note that small – It Could Be Us driven by a 555-based clock.
heatsinks maybe required for
MOSFET transistors TR3 and A simple form of random Counter IC2 will display
TR4. counter is illustrated in Fig.5, tens (0-4) whilst IC4 will display
which may help with the mentally- units. Therefore, a number be-
Lee Matthews exhausting process of selecting tween 0 and 49 will be displayed
Kirkby-in-Ashfield, Notts six entirely random numbers for on a series of light-emitting
the weekly National Lottery. The diodes upon the operation of

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1102
Maxfield & Montrose Interactive Inc
 ,QJHQXLW\ 8QOLPLWHG

push-switch S2, which enables the only way of knowing that it With the values shown, the de-
both counters. Separate is nearing the end of its cycle is lay is about six minutes, which
switches for tens and units when one of the neon indicators can be varied by adjusting the
could be used instead. on the machine extinguishes. values of capacitor and/or resis-
This indicates the start of a tor R4. A suitably powerful
Note that sometimes, num-
short “crease care” cycle after sounder would be the Maplin,
bers may repeat and zero may
which the machine stops. Some order code FK84F, or the
also be displayed.
kind of optically-isolated switch Squires, code 80-015 (takes
Edward Bibby followed by a delay seemed to more current – 35mA), which
Woolston, Warrington be the answer. can be heard in all parts of the
In the circuit diagram of house to warn that the cycle has
Fig.6, when the machine neon nearly finished. My wife has cer-
Tumble Dryer Alarm – It indicator goes out, the ORP12 tainly found it useful!
Won’t Crease You light-dependent resistor, R1, en-
Glyn Shaw
The need for the simple sures that the voltage on pin 12
(reset) of IC1, a 4060 oscillator/ Staines, Middx
Tumble Dryer Alarm circuit of
Fig.6 arose because our new counter, goes low which starts
tumble drier did not have a the counter. Output 14 at pin 3,
buzzer to indicate that it had which goes high at the end of
finished. My wife needed a solu- the delay period, is fed along
tion but vetoed absolutely any with the output of pin 7 into one
idea of digging into the back of of the AND gates of the 4081.
the machine and “fiddling with This provides a pulsed input to
the mains”! transistor TR1, which activates
the sounder WD1. Pin 5 of IC1
As the machine works by flashes the LED D1 when the
sensing how dry the clothes are, crease care cycle has started.

F5
!&
B#
F5 ! [
?>?66
B!
![ $ (
!$
3<; 93" 5>
!# C!

$ !'

 BCD F5

3! ' ! " # $ % BCD 7>4

! P 49C
" $ ' ! ! !% (
93! Q Q Q Q
%%%
4! D? 4$
FB!
& # [ [ [ [
! [ 2!
D8B ?ED@ED
)F
"
DB97
D5>C
 7>4 B"
3" ![
!P !

F5
!&

F5

B$
!$ 93$ !#
![ $ ( 3<; 5>
$ !'

 BCD F5

3# '   ! " # $ % & ' ( ) 7>4 BCD
! P 49C
# " $ ' ! ! % & ) !! ( !%
93# Q Q Q Q Q Q Q Q Q Q C"
%%%

FB"
! [ & # [ [ [ [ [ [ [ [ [ [ CD1BD
D8B ?ED@ED

"
4% D? 4!$
DB97

 7>4 E>9DC B%
3$ ![
!P !

Fig.5. National Lottery Predictor “random number” generator circuit diagram

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1103
Maxfield & Montrose Interactive Inc
 ,QJHQXLW\ 8QOLPLWHG

6
 9
when a positive clock pulse ap-
5
212))
pears on the clock input.
253

:'
The first flip-flop (IC3a) is
 
updated with every clock pulse
5
N 26& ,1  9DD
,& 75
  5 %&&
F
from the 555 timer and it re-
5

&/.
,&

23
 



N
E
members who won the last rally.
N &
Q
  
H The second flip-flop (IC3b) is
26& 287 23


567 & %
updated once all the buttons
Q 9
have been released. It copies
23 9SS

5  
what is stored in the preceding
N
flip-flop, and its output feeds the
5
N clock inhibit inputs (pin 2) of the
counters.
D

'

N
For example, if player B is
9
serving, the clock inhibit input
(pin 2) for counter B (IC6) will
be low and for counter A (IC5)
Fig.6. Circuit diagram for a simple Tumble Dryer Alarm. high. If player A wins the rally a
clock pulse goes to counter A,
the D-type flip-flops IC2a and but, because its clock inhibit is
Squash/Badminton IC2b.
Scorer – Final Call high the counter does not incre-
The output from IC2a is the ment. The first flip-flop now
The circuit diagram in Fig.8 debounced output from button “remembers’’ that player A
will keep the score in both bad- A, and the output from IC2b is should be serving next.
minton and squash games and that from button B. These de- Once button A is released
should end all those arguments bounced signals feed a JK flip- the second flip-flop is updated.
about what the score is or flop IC3 as well as the clock in- The circuit is then ready – if
whose turn it is to serve! puts to two decade counters player A wins the next point his/
The two pushbutton (IC4 and 1C6). her score will increase. If player
switches S1 and S2 are for The counters keep track of B wins the next shot however,
Player A and Player B. The um- the points that each player has the scores will not change but
pire simply presses the button scored, and their outputs will the serve will go back to player
corresponding to the player who drive 7-segment common cath- B.
won the rally. The circuit then ode displays directly. The other The scores for both players
calculates the new score and two counters IC5 and IC7 are are displayed on dual 7-
who should be serving next. for the tens of points for each segment displays. Note that the
When a typical switch but- player. person who is serving is indi-
ton is pressed or released, its In both squash and bad- cated by the decimal point of
contacts do not make a clean minton a player may only gain a their display being illuminated.
connection, instead they might point if he/she was serving. If Pressing both buttons at the
open and close (switch bounce) they were not serving but win a same time resets the unit.
several times before stabilizing. rally, they then serve for the
A typical period of time before a next point. In this circuit when a David Liddament
switch becomes steady (bounce player’s button is pressed the Caversham, Reading
time) is 5ms, which in this case corresponding counter for that
might add 2 or 3 points to a player receives a clock pulse:
player’s score! the counter will only increment if
One solution to get around the clock inhibit input is low (i.e.
this problem is to check the the player was serving).
state of the switches say every The JK flip-flops will latch to
50ms. Hence the 555 timer IC1 “remember” who was serving. In
is an astable multivibrator which this circuit one can imagine a
produces a square wave of ap- JK flip-flop as a simple Set-
proximately 20Hz. This clocks Reset bistable which is updated
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1104
Maxfield & Montrose Interactive Inc
 ,QJHQXLW\ 8QOLPLWHG

 9
The circuit diagram shown in
6

3/ $<(5 $
,& 3,1 
,& 3,1 
,& 3,1 
,& 3,1   
Fig.9 uses a 556 timer, (twin 555
,& 3,1 
,& 3,1  
,& 3,1 
,& 3,1 
D

 
D
7(1 6
timers in a 14-pin package), whereby
E E

6   each timer is configured as a one-

,&
F F
3/ $<(5 %

,&D


G






G shot monostable. The output from

  
5 (6 (7
H

 
H
each timer is used to trigger the input
 I I


J
 
J
of the other timer via an RC network.
;
&.


' 4
 
This arrangement is commonly
'28%/ ( ',*,7

,&D known as a cascade timer. The result


&20021&$7+2'(
&$55< 287 ',63 /$< LQ
 
  D D

4
 
&. E
 
E
81,76 is a dual timer with varying on/off
&.
5
&281 7(5 $ F
 
F times and a brief negative-going
,&
N 


G






G
pulse at either one of the trigger in-
H H


',6

5 (6 (7 I
 
I
puts (pins 6 and 8) every time each
5 &/2&. J
 
J monostable times out.
N ,1+,%,7


7+5(6
 '3

 
N


N
During each cycle, pin 6 and pin

75,* 8 are held at logic 1 by pull-up resis-
,&     5 5
-

,&D
4 - 4  Ω  Ω tors R2 and R3. A pair of back-to-
,&E


287




&.
 
back LEDs, D1 and D2, indicate

. 4
 
.
&.
4

5
 Ω
5
 Ω
whether the circuit is paused or film-

ing. When the output from one timer
,&E
&   
µ

 Y
 
 '3 N N
is high, the other will be low.

&/2&. 81,76
  








,1+,%,7 D
 
D

E
The trigger inputs A and B are
5 (6 (7 E


'
&.
4
 &281 7(5 %  
F
connected to pin 1 and pin 2 of a dual
,&
F

,&E NAND peripheral driver 40107 (IC3).

 
G G

 
 

 
H

 
H

I
The output is then taken from pin 3
4 &. I

  and connected to pins 6 and 7 of the

;
J J

&$55< 287

 '28%/ ( ',*,7
second driver. This produces a strong

&20021&$7+2'(
',63 /$< LQ negative pulse whenever either of the
&.
D
 
D 7(16
monostables changes state.
 
E E

F
 
F The resulting pulse can be used
,&
5 5
N  N


G






G
to power the LED emitter in a pre-

5 (6 (7
H

I
 
H

I
ferred opto-isolator (not shown) or a
,& 3,1 
,& 3,1  
,& 3,1  
,& 3,1 
J
 
J solid-state relay. The use of an opto-
,& 3,1  
,& 3,1 
,& 3,1  
,& 3,1  isolator ensures that no voltage or
9
current from the timer unit can inter-
fere with the camcorder circuitry.
Fig.8. Circuit diagram for the Squash/Badminton Scorer. The whole circuit can be powered
from a 9V battery. A 6V regulator IC1
Time-lapse Unit for Camcorder – In The Frame ensures that set times do not drift due
to decreasing battery voltage.
Many camcorder owners would like to produce more cre-
ative videos, such as time lapse films, which condense slow- Timing components VR1 with C3,
moving sequences into a short period. Unfortunately, time-lapse and VR2 with C6 should give a maxi-
facilities are only found on more expensive video cameras. mum time of 270 seconds. There is
no point in increasing this time pe-
All camcorders, however, have a REM (remote) socket, for riod, as camcorders automatically
use with a manual stop/start lead. The REM socket on cam- shut down if left in pause mode for
corders however is not the same as the REM socket on a cas- more than 5 minutes. The on-time
sette recorder, which is basically a simple n.o./n.c. (normal can be very short, i.e., enough to
open/normal closed) switch. capture two or three frames.
Manual control of a camcorder via the REM socket requires
“pulse operation”, i.e. a short pulse to start and a second short Philip Male
pulse to stop. This overrides the “pause control”, and places the Drake’s Broughton
camera in its pause mode during a break in filming. Pershore

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1105
Maxfield & Montrose Interactive Inc
 ,QJHQXLW\ 8QOLPLWHG

,1
,&
 
287

& 20

5 5 5 5 5
 D 

/ ('
N N N  N  N


& &    2372

Q Q  ,6 2/$725

5(6(7 9 5(6 (7
$

²
N


&
6
75,*  6(( 7 (;7
$

Q 

 ;

,&
23


,&
 ;

23
 5
 
95



&
95
 Ω
0 75,* 0
%
 
Q

  %

',6 7+5(6+
%

 
9
7+5(6+ ',6 5

 Ω

;

&9 *1' & 9

  
'

'

& &
& &

 µ Q  Q  µ

' $ 1' ' & $1 (,7+(5 %( 21( 5(' / ('

$ 1' 21( *5((1 /(' 25 %,',5(& 7,2 1$/

/( ' 72 6+2: : +(1 7,0(5 ,6 ,1 7+(

21 6 7$7(
$1'
2)) 67$7(


Fig.9. Circuit diagram for the Time-lapse Unit for Camcorder.

 9

$8',2
,1

5
5
N
 N

& &

 Q  µ 5
/,0,7


N
 '
 ' 5
5 /('

&D 1   N
N
  5

&E
N


7/ 
   

  
7/ 
&F 
&G 

5 5
 
7/   7/ 

N  N
&
 µ / ,0,7 $8',2
/ (9(/ / (9(/
& 5
Q N
95 95
N N
5
N

9

&
Q
² 9

&
 S & $8',2
 
 µ 2 87

&
 

 
0& 3

 
1 & 1&

Fig.10. Circuit diagram for an Audio Limiter.

plied to an audio power ampli- which together with diode D1

Audio Limiter – Just the fier, protecting those valuable acts as a precision rectifier. The
Limit tweeters! voltage across capacitor C4 sits
An audio limiter circuit was The design uses the at the wiper level of Limit con-
required which would accept a MC3340P electronic attenuator trol VR1 until the audio level
wide input voltage range without chip (IC1). Resistors R1 and R2 exceeds this value.
introducing too much distortion attenuate the input signal to a At this point diode D1 be-
when limiting. The circuit of Fig. level suitable for the MC3340. gins to conduct so the voltage
10 achieves this as well as al- The maximum level applied to across C4 rises. The voltage on
lowing a variable limit level and the device should be 500mV the output of IC2b (pin 7) then
output level. The circuit could RMS. falls. This voltage is inverted
be used in many areas, particu- and attenuated by IC2c and its
The input signal is also ap-
larly in limiting the signal ap- associated resistors. Audio
plied to IC2a via capacitor C2,
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1106
Maxfield & Montrose Interactive Inc
 ,QJHQXLW\ 8QOLPLWHG

Level control VR2 adds an off- C31> FB!

set to the output of IC2c, which B1D5 ! : "G

configures the attenuator in its >9@;?G 49C;

B51B F95G
linear region. The output of IC2c !"F 43
=?D?B
C!Q

2!
is then applied to the attenua- =!


! %F
4 35<<
tion control pin of the MC3340.
The opamp IC2d forms a simple
B!
comparator, which drives LED ?2:53D
<5>C "["

D2 during limiting. S

To set up the Audio Limiter, 3!
! P C!R

adjust VR2 so that the output of U

DB! C;!
IC2c is at least 4V to set the at- C31>>54
9=175
@8?D?

tenuator in its linear region. A DB1>C=9DD5B DB1>C9CD?B

D? =93 9>@ED 2"
C?3;5D ?6 )F

higher level can be applied to B149?

DB1>C=9DD5B

vary the relative output level.

Next, apply the maximum level
of audio and adjust VR1 until Fig.7a. Circuit/mechanical arrangement of the Nipkow 25-line
the LED illuminates. Back off Transmitter section.
VR1 until the LED just extin-
guishes. Any increase in the au- wish to experiment with other components. The LED D1 is a
dio level will now be limited to transmission media (e.g. wire- high-brightness green device
the level selected. based audio, intercoms etc.). placed in a flashlight reflector,
The transmitter section, and a piece of greaseproof or
Duncan Boyd tracing paper is placed over it to
which also shows the relative
Blackburn, Scotland obtain a more uniform spread of
placement of the mechanical
parts, is shown in Fig.7a. A Nip- light.
kow disc may be made from With this placed underneath
Narrow Band Vision – stiff card, using a plate to draw the “receiver” disc, a reasonably
Nipkow Disc-overy a circle 180mm diameter or so. uniform raster is obtained. Note
The object to be pictured that the picture requires the
The system shown in Fig.7
must be brightly lit, and it is room to be in near darkness if it
illustrates a simple but fascinat-
captured through a lens and is to be discernible by looking
ing electromechanical technique
converted into a narrow-band through the spinning disc.
for transmitting a small video
image over amateur radio vision waveform by TR1, a pho- The Receiver disc is rotated
bands. It consists of a simple totransistor placed in a plastic slightly faster than the Transmit-
modulator based on a Nipkow box behind the scanning disc. ter disc and the image will then
disc, a mechanical scanning de- The phototransistor (e.g. a be visible, although it may be
vice used in early television PN202, but other types may “rolling” or swirling. By applying
systems. The Nipkow disc has a work equally well) requires a 9V very light pressure to the re-
single-revolution spiral of small supply. A good-quality DC mo- ceiver disc, it can be synchro-
holes (25 in this case) which if tor (say, 12V DC) is powered nized to the point where you
rotated can be used to provide from a single D-cell and poten- can get a reasonably stable im-
raster scanning of an object. tiometer VR1 (rated at 2W) con- age.
trols its speed. The signal is de- A flywheel, formed from an
With the circuit shown, a
coupled by capacitor C1 and old loudspeaker magnet, was
basic 25-line monochrome
applied to the microphone input placed on top of the prototype
video image may be sent using
socket SK1 of the radio trans- receiver disc to add some mo-
amateur radio equipment over a
mitter. mentum and help with synchro-
good quality clear voice chan-
nel. This resolution is high nization. None of the parts are
enough for facial recognition of Receiver critical and substitutes may be
a person in close-up. It should made.
In Fig.7b, the loudspeaker/
not be compared to a slow-scan headphone output is fed to a An experimental but worth-
system, which can only send single transistor stage consist- while modification to the re-
still images. Readers may also ing of TR2 and surrounding ceiver is shown in Fig.7c, which
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1107
Maxfield & Montrose Interactive Inc
 ,QJHQXLW\ 8QOLPLWHG
>9@;?G 49C; 1@@<I 69>75BD9@ @B5CCEB5
D? CI>38B?>9C5 9=175
4! Q
892B9
7B55>
B" 9=175 1@@51BC 85B5
C"Q [
<54
%& [
6B?= C@51;5B ?ED@ED 7B51C5@B??6
?6 B149? B5395F5B  @1@5B

1
C;"
DB" ="

3" 23##' !"F 43

 ! P S =?D?B
FB" R
"[" D?B38
C"R
U B56<53D?B
FB#
2#
)F
3?>DB1CD
2$
! : "G
<54
4!
 ! %F
3# B#
! P %& : 4 35<< C=1<<
6<IG855<
C31>
B5359F5B
C@554
2

Fig.7b. Circuit/mechanical set-up for Receiver section.

DB$
offers a form of synchronization 23! (
U
"% 8?<5C @E>3854
S D8B?E78 49C;
control. This provides some C9J5 ?6

pulse advancement on the re- DB# C31>>54

9=175
23! (
C31> R !%  " ]]

ceiver disc’s rotation, which is C"R

8?<4 S U

now controlled by a transistor FB#

Darlington pair (TR3, TR4). ! [

R 
2$

It is important that good #F

" h 4
4" !
B$
:
="
!"F 43 =
35<<C =?D?B
quality smooth DC motors are ?1)!
7U

used, and the two motors 1

Q [
>9@;?G 49C;
!( ]] 491=5D5B
should have reasonably CD966 31B42?1B4

matched characteristics. Al- 2 " ?66

though the circuit is not perfect, Fig.7c. Improved Receiver add-on with Scan/Sync control
it is well worth the extra effort. (VR3).
Michael Robertson
Chasetown, Staffs

using VR1 (the timing capacitor many applications.

Pulse Modulated Inverter C2 should be 1nF minimum –
– Mains Motor Controller Both power MOSFETs must
ARW). be mounted on heatsinks and
A single-pulse modulated The NAND gates of IC3 the main On/Off switch S1
inverter circuit diagram is shown (4011B) are used to separate should also be capable of carry-
in Fig.11a, which can be used to positive cycle signals for the ing the full load current. The
operate a series-wound motor power MOS transistor TR1 and winding details of the trans-
up to 1hp in variable speed the negative cycle signal for former are also given. (I was
mode, from a 12V lead-acid car TR2. The two Zener diodes D1 unable to trace the power MOS-
battery. The series motor may and D2 provide protection for FETs used by the writer and a
be an electric drill or the drive the transistors whilst diode D3 substitute may be needed, e.g.
motor of a small electric vehicle and capacitor C3 help provide the IRFPG50 or similar, offered
or buggy for example. The cir- isolation between the driver and as a suggestion only. ARW)
cuit waveforms of various out- the output stage.
puts are shown in Fig.11b. M.T. Iqbal
Transformer T1 steps up
In Fig.11a IC1 (a 4047B) is Rawalpindi, Pakistan
the input voltage to a maximum
working as a 100Hz astable 200V AC. The potentiometer
which triggers an adjustable VR1 can be used for varying the
monostable (IC2). The period of output voltage in the range of
the monostable can be varied 50V to 220V AC, suitable for
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1108
Maxfield & Montrose Interactive Inc
 ,QJHQXLW\ 8QOLPLWHG
'

6
1

&

µ

&$5


   


%$77(5<

9 DD 9 DD

& '
$ 67$%/( $ 67$%/(


9
%

   
9 DD  9

26& 287 75,* 4

 &   

,&
D
4 $ 67$%/(
!Q

     

,&
& (;7 ²75,* G
4
9
%

&      

,&
$ 67$ %/( 5& &20 5(6(7
Q

95   
 5

& (;7 5(75,* E

²75,* N
N
%

    

5& &20 5 (;7 F

5(6(7

*1'
5   

9 SS


5(75,*
N


5 (;7 75,*

$
*1'

9

%
75
) (
7

5)*1
G

'
V

9

D
&7
9

6(5,(6

02725
0
'

9 V

N
75 75 02817(' 21 +($76,1.6 J

75
75$16)250(5 ,521 & 25( $   LQ 2 G

35,0$5< 6: *     78516

6(&21'$ 5< 6: *   78516

5)*1

9

Fig.11a. Circuit diagram for a Single-Pulse Modulated Inverter.

$
W

PICO PRIZE WINNERS %

W

It's time to decide the winners of superb PICO Tech- &

nology PC-based oscilloscopes, once again generously W

donated by PICO (www.picotech.com) for three lucky '

entrants whom in our judgment submitted the best ideas W

published in the past six months. As always, every entry

(
was judged on a number of criteria including the extent W

of “lateral thinking” or novelty, technical merit, resource-

fulness, appropriateness, and overall completeness. Pre- )
sentation was used as a tie-breaker.The final choice was W

difficult and, after careful consideration, EPE Editor Mike $&

287387

Kenward and Ingenuity Unlimited host Alan Winstanley W

jointly selected the following winners from the June-

November issues:WINNER: receives an impressive Fig.11b. Output waveforms at various
PICO ADC200-50 Digital Storage Oscilloscope, worth
over 450 UK Pounds!Lee Archer – TV Test Pattern Gen-
stages of the circuit.
erator (September 1999) – illustrating the adaptation of a
teletext timing chip, this circuit was considered to be thoroughly de-
veloped and complete.RUNNERS-UP: Both are lucky recipients of
PICO ADC-40 single channel PC-based oscilloscopes.Rev. Thos
Scarborough – Loop Aerial MW Radio (August 1999). This was a
novel radio receiver design using some traditional techniques, and
we are also happy to acknowledge the contributions made by our
most ingenious Reverend.Z. Kaparnik – One Volt LED. (November
1999). A number of intriguing and professionally presented microp-
ower circuits optimized to operate a LED from a single cell.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1109
Maxfield & Montrose Interactive Inc
 /2)7 *8$5' E\ 7(55< GH 9$8;%$/%,51,(

Has the light been left switched on? possible to increase the
operating time and details for
doing this are given later.
Having a permanently-wired which is left in some suitable Similarly, you could shorten it if
mains light in the roof space is position inside the loft. It required.
handy, especially if you keep a lot protects against leaving the light
of useful material up there. switched on by sounding a loud
warning after 8 minutes or some CHECKOUT
Unfortunately, it is all too easy to
leave it switched on as any user other preset time. This can be Before beginning
will testify. heard through the ceiling even construction work, check that
with the loft hatch closed and the loft space is reasonably dark
Once the hatch is closed, alerts the next person passing when the light is switched off.
there is no external sign that the by underneath it. Make sure you will be able to
light is on. It could then remain
In the prototype model, the site the unit where light from the
like that until the next visit –
specified operating time was lamp will reach it and, at the
possibly several weeks or even
found to be sufficient. If you same time, above some place
months later. In the meantime, a
happen to be working for a long where the sound will readily
significant amount of electricity
time in the loft, a Reset attract attention – for example,
would have been wasted and,
pushbutton switch on top of the near the top of the stairs.
more importantly, the life of the
bulb has been significantly unit may be operated every so Of course, the unit could be
reduced. often to reset the circuit and used in other similar situations.
hold the sounder off for a For example, to guard against a
further set time interval. This cupboard light being left
SELF-CONTAINED switch may also be used after it switched on inadvertently. You
The Loft Guard is built as a has begun to sound to stop it. could even site the buzzer
small, battery-powered unit, If you habitually spend long remotely if required.
periods up there, it would be The standby current

5
5 95 5 5 0

75
0 0 0 0 6((
7(;7
5 '
N 1 036$

D N
H
5

 &&
N 
 E
9
& 5 '
 

,&
Q N 1
F
   D N

 
,&/

75,*

5(6(7
287387

7+5(6

%
9


,&

',6
  [ $$


5
&
6((
*1' 6(( 7(;7
5(6(7
7(;7

5 & 5 :'
6
0 Q 0 9 72 9

 µ µ
  [ 0
6(( 7(;7

Fig.1. Complete circuit diagram for the Loft Guard.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1110
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
COMPONENTS
Resistors
R1, R2 10M
R3, R8, R9 470k (3 off)
R4 miniature light-dependent
resistor, 5M dark resistance,
5mm diameter (see text)
R5, R6 1M (2 off)
R7 100M cermet film
(or 3 off 33M -- see text)
R10 10M (2 off -- see text)
All 0.25W 5% carbon film, except
R4 and R7
Potentiometer
VR1 1M sub-miniature preset, vert.
Capacitors
C1, C2 47n miniature metallized Fig.2. Printed circuit board component layout and
polyester, 5mm pin (approximately) full size copper foil master pattern.
spacing (2 off)
C3 2u2 miniature metallized may be expected. voltage (nominally 4 5V) due to
¬

polyester, 5mm pin spacing

(2 off or as required -- see text) However, this will depend the potential divider action of
Test capacitor 100n metallized on how many times and for how equal-value resistors R1 and R2.
polyester, 5mm pin spacing long the buzzer sounds. While The non-inverting input (pin 3)
actually operating, the current has a voltage applied to it
Semiconductors
D1, D2 1N4148 signal diodes (2 off) rises to some 10mA. You could dependent on the values of the
TR1 MPSA65 pnp Darlington use a PP3 battery but the life resistors in another potential
transistor would be correspondingly divider.
IC1 ICL7611 micropower opamp
IC2 7555IPA low-power timer
shorter. In this case, its top arm
consists of preset potentiometer,
Miscellaneous
VR1, connected in series with
S1 miniature pushbutton switch, CIRCUIT fixed resistor R3 and the lower
push-to-make
WD1 Audible warning device, DESCRIPTION one, light-dependent resistor
103dB output at 1m minimum,
The Loft Guard circuit works (LDR) R4.
10mA DC operation maximum
B1 9V battery pack (6xAA cells), by sensing the change in As the illumination of the LDR
with holder illumination as the loft light is sensitive “window” is reduced, the
operated. Switching it on resistance of the device
Printed circuit board available triggers a timer, which holds the increases. In total darkness the
from the EPE Online store , code
sounder off for the preset delay specified LDR will have a
7000249 (www.epemag.com);
plastic box, size 138mm x 76mm x period. If the light is switched off resistance in excess of 5M . :
38mm internal; 8-pin DIL socket during that time, the circuit will Even when there is a small
(2 off); plastic stand-off insulators automatically reset ready for the amount of light, it will exceed
(3 off); PP3-type battery connector;
small fixings; multistrand connecting
next time. :
1M .
wire, solder, etc. The complete circuit In tests on the prototype in
diagram for the Loft Guard is the author’s loft, the “light”
See also the shown in Fig.1. It will be seen resistance was found to be some
SHOP TALK Page!
that operation depends on the
action of two integrated circuits.
:
100k . Of course, in any
Approx. Cost particular situation this value will
The first of these, IC1, is an
Guidance Only $30 operational amplifier (opamp)
depend on the relative positions
of the unit and loft light, plus also
(Excluding Batteries) responsible for the light-sensing the power rating of the bulb and
aspect while the other, IC2, other factors. The point is that
requirement of the prototype
carries out the timing. there is a wide difference
unit is less than 100mA. Using
the specified 9V battery pack, Looking at IC1 first, the between the LDR “dark” and
consisting of six AA alkaline inverting input (pin 2) is “light” resistance.
cells, a life of at least one year maintained at one-half of supply

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1111
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
MORE OR LESS out. The operating period would never time out since it
depends on the value of would remain triggered. While
Suppose preset VR1 is set
:
capacitor(s) C3 and resistor, R7. on standby, resistor R5
to a value of 300k . This is The higher the value of either or maintains the trigger input in a
added to resistor R3 to give the both of these components, the high condition and this prevents
resistance of the top arm of the greater the timing will be in possible false operation.
potential divider – that is, 770 proportion.
kilohms.
Under standby (“dark”)
KEEP IT UP
conditions. the resistance of the
HIGH VALUES The reset pin of IC2 (pin 4)
LDR will exceed this value. This Resistor R7 has a very high needs to be kept high to enable
will result in a voltage greater resistance (100 megohms) and operation of the monostable,
than 4 5V appearing across it
¬
the specified component may not and this is the purpose of
and hence at IC1 pin 3. When be available to all readers. It resistor R6. However, to allow
the loft light is on, the resistance could be made up from lower the circuit to settle down when
of the LDR will be less than 770 values connected in series and switched on and to prevent
kilohms and the voltage at pin 3 more will be said about this later. possible false triggering, it is
will fall below 4 5V.
¬ Capacitor C3 will probably held low for a short time using
consist of two separate capacitor C2.
When the voltage at the
components connected in parallel During this time the
opamp non-inverting input (IC1
(as shown in the Fig.1.) to monostable is disabled and
pin 3) exceeds that at the
provide the required capacitance. nothing can happen. The
inverting one (that is, under
The suggested value (2 2uF) will
¬ capacitor soon charges through
“dark” conditions), the opamp
give a combined effect of 4 4uF. resistor R6 and allows pin 4 to
output, pin 6, will be high. When ¬

go high.
it is less (“light” conditions), it Of course, you could use a
will be low. At the end of single 4 7uF, two 4 7uF or even
¬ ¬
Pushbutton (Reset) switch
construction, preset VR1 will be one or two 10uF capacitors S1 may be operated
adjusted so that this happens providing they were small enough momentarily at any time to
under the actual conditions to fit the circuit board layout. begin a new timing cycle and so
prevailing in the loft. Such an arrangement would give hold the warning buzzer off.
This works by taking the trigger
Note that both opamp a correspondingly longer time
input low for an instant.
inverting and non-inverting period.
voltages are derived from Using the values shown in the While IC2 output is high
potential dividers connected circuit diagram, the timing will be (that is, during the course of
across the power supply. As the about 8 minutes. It could be timing), the base (b) of
battery ages and the available reduced by using a single Darlington transistor TR1 will
voltage falls, the relative state capacitor having a lower value if also be made high (close to
of the inputs will remain required. positive supply voltage) via
unchanged. The circuit will resistor R9 and diode D2. Under
therefore still work correctly. Of When the LDR is dark – that standby conditions, the LDR R4
course, the battery pack will is, under standby conditions, the will be in near-darkness and IC1
eventually develop insufficient opamp output at pin 6 will be high pin 6 will be high. This also
terminal voltage to operate the and there will be no effect on IC2. provides a high state at TR1
buzzer satisfactorily and it will However, when the output goes base through resistor R8 and
then need to be replaced. low (i.e. when the light is switched diode D1.
on), a low pulse is transferred, via
Now look at IC2. This is an capacitor C1, to IC2 trigger input Since TR1 is a pnp
IC timer configured as a (pin 2). The monostable then transistor rather than the more
monostable. It may be activated begins a timing cycle. usual npn type, such a high
by a low pulse applied to the state will maintain the base at
trigger input (pin 2) – while high The purpose of capacitor C1 near emitter voltage and so hold
there is no effect. is to allow only a short pulse to it off. No current will flow in the
pass. This is because if IC2 pin 2 collector circuit and buzzer,
Once triggered, the output was maintained in a low state
(pin 3) goes high and remains WD1, will remain silent.
continuously, the monostable
like that until the circuit times Suppose some light reaches
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1112
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
%  :' 
the LDR R4, IC2 will be
triggered and a timing cycle will
begin. Opamp IC1 pin 6 will go
5 5
low but this will have no effect 95
H
D N
on transistor TR1 because this 5 '
E

5 D N
state is blocked by diode D1 5 5 '
5
5
5 F

;
which is now reverse-biased.
However, TR1 base will be kept 5
&
;
6(( 7(;7

5

:'²

high by the high condition of IC2

pin 3 and the buzzer will remain 5
&
&

&
9

%²

off.
When the monostable has

6(( 7(;7

timed out, IC2 pin 3 will go low

and this state will be blocked by
diode D2. Assuming light is still 72 6

falling on the LDR, TR1 base

will no longer be made high by
either path R8/D1 or R9/D2.
This allows it to go low via
resistor R10 and the device is ;

turned on (remember, it is pnp

transistor!). Collector current ;

then flows and the buzzer

operates.
The fact that TR1 is a
Darlington transistor results in it Fig.2. Printed circuit board component layout and
having an exceptionally high (approximately) full size copper foil master pattern.
current gain. Only a very small
base current (a fraction of a
microamp) is therefore
sufficient to operate the buzzer
hence the very high value of
resistor R10. Remember, the
flow of current is in the opposite
sense for a pnp transistor
compared with npn.

KEEPING IT DOWN
It is essential that the Components mounted on the completed circuit board.
continuous current requirement
of the circuit is kept very low to
Note that a single cermet-film resistor has been
minimize battery drain. This is used for R7 (see text).
achieved by choosing very low
To be effective, the buzzer circuit board (PCB) and the
power integrated circuits. Also,
must be of a very loud type yet topside component layout and
the resistors in the potential
have a current requirement of underside track master details
divider chains are made very
10mA maximum. The specified are shown in Fig. 2. This board
high. If the loft is reasonably
unit (103dB at 1meter) was is available from the EPE
dark under standby conditions,
found to work very well. Online Store (code 7000249) at
the resistance of the LDR will
www.epemag.com. All
also be high and this reduces
components are mounted on
still further the current flowing CONSTRUCTION this except the battery holder,
through the series arrangement
The Loft Guard circuit is buzzer, and pushbutton reset
of VR1, R3 and R4.
constructed on a small printed switch.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1113
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
Commence board amount of adjustment to its ready to work in darkness).
construction by drilling the three position. Adjust preset VR1 to
mounting holes in the positions approximately mid-track
indicated. Follow by soldering position and connect the
the IC sockets in position (do
POLARITIES batteries. Keep the switch wires
not insert the ICs at this stage) Now solder the polarity- separated so that the bared
then all other components sensitive components in place. ends cannot touch.
except capacitor(s) C3, light- These are the two diodes and Working on an insulating
dependent resistor R4, the Darlington transistor TR1. When surface (such as wood or
diodes and transistor. On no soldering the diodes, note that plastic) to prevent short circuits
account solder the ICs direct the cathode (k) end has a black at the PCB tracks, place the AA
to the board, because it band. When mounting the cells in the holder and connect it
would be very easy to transistor, take care to place it up. Peel back some of the PVC
damage them. as shown in the photographs tape to allow some light to
Note, resistor R10 (20M ) : with the flat face to the left. reach the LDR – the buzzer
consists of two individual 10M : Solder the battery connector
wires to the PCB. If the battery
may give a momentary “chirp”
which may be ignored.
devices connected in series
using the pads indicated (both holder has tag connections After about ten seconds or
positions are labeled R10). If instead of being the more usual thereabouts (remember, the
:
the 100M cermet film type
PP3 type, use short pieces of
stranded wire instead. Connect
timing has been reduced) the
buzzer should sound. If you re-
resistor specified for R7 is not
available, connect three 33M : pieces of light-duty stranded
connecting wire for the Reset
cover the LDR, it should stop
immediately. Similarly, if you
resistors in series instead using switch S1 and solder the buzzer touch together the switch wires,
the pads provided on the PCB – leads to the WD1 pads – the red it should stop.
the three positions are labeled one is the positive lead.
R7. If you have problems
Insert the ICs in their making it work, make sure the
The photographs show the holders, with the correct LDR window really is covered to
single specified resistor being orientation. These are both exclude almost all light – some
used. This is soldered directly CMOS devices and could types of black tape are far from
between the pads connecting possibly be damaged by static opaque. If necessary, carry out
IC2 pins 6, 7 and 8 – they are charge, which may exist on the the test in a dark cupboard. It is
labeled “x” in Fig.2. If you can body. To avoid possible not satisfactory to cover the
find no other way of doing it,
:
problems, touch something LDR window with a finger!
you can connect ten 10M which is earthed (such as a
resistors in series, zig-zag If all is well, disconnect the
metal water tap) before
fashion, and connect the ends battery holder and remove the
unpacking them and handling
of the “chain” to the “x” pads. ICs, again observing the anti-
their pins.
static precautions mentioned
Connect a 100nF “test” earlier. De-solder the buzzer
capacitor to either C3 position.
This will provide an operating
TESTING wires and test capacitor C3.
period of around ten seconds, Most readers will wish to With the required timing in
which will be more convenient carry out a basic test before mind, decide on the value of the
for testing purposes than the full mounting the circuit board in its capacitor, or capacitors needed
operating time. box. This will allow any errors to for C3 and solder them in place.
be corrected more easily. It Note that an electrolytic
Solder the LDR in position capacitor would not be
would be a good idea to tape
using the full length of its end satisfactory here due to its
over the hole in the buzzer for
leads for the moment. If the inherent high leakage current.
the moment to reduce the
specified miniature type of LDR
sound output because it is very
is not available, then the larger
loud!
ORP12 type could be used. BOXING UP
However, it would take up more Cover the LDR with a piece
of black PVC tape to simulate You are now ready to mount
space and would need a certain
placing it in darkness (or be the circuit board in its box. This

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1114
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
must be large enough to
accommodate the PCB, battery
pack, buzzer, and pushbutton
switch. You could use a more
compact case if you used a
smaller type of battery, but
remember that this will give a
shorter life.
Arrange the internal
components on the bottom of
the box and mark through the
PCB and sounder mounting
holes. Remove everything
again and drill these holes. Drill
a further hole rather larger than
that in the center of the buzzer
itself for the sound to pass
through. Note that the buzzer Positioning of components and circuit board inside the proto-
will be mounted so that the type case. Note the LDR “window” hole and Reset switch posi-
sound is directed downwards tion in the lid. The space to the right of the PCB is
(see photograph). This will allow for the battery holder.
the maximum amount of sound
to pass through the ceiling. Shortening the buzzer wires as as you can (as viewed from the
necessary, solder them back to top edge of the PCB) consistent
Mount the PCB temporarily
the PCB pads. Insert the ICs with correct operation. When
on plastic stand-off insulators.
again, taking precautions satisfied with the performance,
You may wish to mark the
against static charge build-up. secure the lid.
position of preset VR1 on the
side of the box so that a hole Mount the PCB and attach Check that the sound can
may be drilled to allow it to be the buzzer using a pair of long, be heard below the unit when
adjusted more easily. thin bolts. Do not forget to the loft hatch is closed. You
remove any tape from the could remove a small amount of
Measure the position of the
buzzer that was used to reduce roof insulation from around the
LDR “window” (top surface) and
the sound output during testing. case to allow the sound to pass
mark this on the lid of the box.
Insert the AA cells and secure through more efficiently but this
Drill a clearance hole for it. With
the battery holder to the base of was not found necessary with
the lid in place, and the LDR
the box using a small bracket if the prototype.
protruding, measure how much
necessary. It is suggested that the unit
the end leads need to be
shortened so that the window Place the lid temporarily in be allowed to sound every now
will be level with the face of the position but do not secure it yet. and again to check the
box. Adjust the LDR end leads as efficiency. When the buzzer can
necessary so that the window is no longer be heard as it should,
Remove the PCB and
level with the top face of the the batteries should be
adjust the LDR soldered joints
box (see photograph). Take replaced.
to give the correct clearance. It
care that they cannot touch one
would be a good idea to leave
another and cause a short-
the leads a little on the long side
circuit.
because they can be bent
slightly at the end to make small
adjustments to the height. FINAL CHECKS
Drill a hole in the lid for the Test the circuit under real
Reset switch and attach it. conditions. Try the unit in
Solder the switch wires leading different positions in the loft to
from the PCB to its terminals. find the best one. Leave preset
Drill the hole for VR1 VR1 adjusted as far clockwise
adjustment if this is needed.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1115
Maxfield & Montrose Interactive Inc
 *,1250286 6723:$7&+
E\ 1(' 672-$',129,&
Now you’re “up and running”, why not add some Decimal Binary
Giant Displays to your events Stopwatch? 0 0000
1 0001
This Large Digit Display unit SOFTWARE 2 0010
was originally designed for use 3 0011
In keeping with the author’s
with the Ginormous Stopwatch 4 0100
stated objective of designing
module presented last month. It 5 0101
without designing, he used two
has 178mm (7-inch) characters 6 0110
pieces of software from the Par-
and can use high brightness 7 0111
allax web site at
LEDs for dazzling daylight perfor- 8 1000
www.parallaxinc.com. These
mance. 9 1001
were from application notes
It can also be driven from a concerning receiving serial data Actually, four bits will allow
standard computer serial port with and utilizing a jump table to dis- you to count from 0 to 15
the optional adapter, allowing it to play digits on a 7-segment dis- (binary 1111), but we only need
be used as a scoreboard, bingo play. Readers are referred to to count up to 9. Let’s call these
number display, clock, etc. these notes. bits “n”, as in “nnnn”. Similarly,
four bits will allow us to have
It is interesting to note that
modules numbered from 0 to
CIRCUIT OVERVIEW it was easiest to choose the
15, call these bits “d” as in
same crystal frequency as the
The heart of the circuit is a Stopwatch module “dddd”.
PIC16C54 microcontroller and (3 2768MHz). This allowed the
¬

this has two relatively simple

author to play with the soft-
tasks. The first is to receive serial
ware’s “bit_k” constant without
data from the Stopwatch module
worrying about serial link com-
or computer serial port. The data
patibility between the Stopwatch
reaches the micro via an opto-
and Large Digit modules.
isolator (IC4), as discussed in
Part 1, and the individual digit Of course, large display
modules can be daisy chained modules that are to be driven by
together up to a maximum of 16 a computer must comply with
modules. the standard computer baud
rates and everything has been
The Stopwatch module only
standardized at 9600 bits/sec.
uses seven addresses. However,
when driven from a computer us- It was necessary, though, to
ing the Serial Port Converter, the come up with a protocol to ad-
software in the Large Digit Dis- dress the correct module and
play units will respond to all 16 tell that module what number to
addresses. display. This turned out to be
quite easy, and it can be done
The second task is to switch in one byte.
on the various segments on the
display to form the digits 0 to 9. First, consider the number
to be displayed. In binary you Completed “7-segment” Gi-
need four bits to display the dig- ant Display module. The fig-
its 0 to 9, like this: ures measure approximately
178mm by 100mm.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1116
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
6(5,$/ ,17(5)$&( /$5*( ',*,7
Computers and PIC micros  9 25  9

:,5,1* $ '% 3/8*

like to deal in bytes, which are 

eight bits, so the software

5

makes the “nnnn” and “dddd” 

5

6(( 7 (;7

bits into artificial bytes:

3,16

'
D

,&
1
 '75
N
'$7$ 287 
 '65  

 &'

dddd becomes dddd0000,  576

D
5 5,& 3,1 

which is one byte. 


&76

'$7$ 72 &219(57(5
'$7$ 287 
N

 

nnnn becomes 0000nnnn,  *5281' 72 &219(57(5

which is another byte. 75

%'
*1'

5
&20387(5 6(5,$ / F

The two bytes are ORed

N
3257
'$7$
,1387
E

together (inclusive-OR) bit by N

bit to form a single byte which

'

1

D
6(5,$/ 3257

looks like ddddnnnn. This sin- *1'

gle byte contains both the mod-

Fig.2. Circuit diagram for a simple Serial Port Converter Inter-
ule number and the digit to be
displayed. face add-on. The values of resistors R18 and R19 should be
330 ohms for 9V and 560 ohms for 12V
For example, to make mod-

ule 1 display the number 1, the

 ,&  D D D D

output byte would be 00010001.

 9 ,1 287
9
' '  ' '

 / 
N N N N

&20 To make module 2 display the

&

  µ
&

 µ
&

 Q
&

 Q
D D D D
number 1 it would be 00100001.
'  '  ' '

N N N N

'  72 ' 

CIRCUIT DIAGRAM
9

D D D D D

'  '  ' '

N N N N N
Referring to the circuit dia-
gram in Fig.1, data is received
75 via the opto-coupler IC4. The
D D D D

5

Ω
' '  ' '

 

H
% &
5

 N
N N N N

driving device (e.g. the Stop-

E

5  5 
watch) switches a LED inside the
Ω Ω
D D

opto-coupler on and off and the

   

F ' '

N N

light from its LED shines onto an

opto-transistor, switching it on
5

Ω
& &
 

;
 S  S

287
  0+ ]
5 5
and off in unison.
5 72 5

Ω
 

Resistor R1 holds the output


6(5,$/           

7+5R

0&/ 5
9( 26&  26&

5% 
 

,1
287  287  28 7 287 
of IC4, pin 5, at 5V until the tran-
287 28 7 287 

sistor switches on and shorts pin

,&
9
  

72& ., 5%  ,1

 

5 to ground. Pin 5 is connected

,&
 1 
5%  ,1

,& directly to the PIC microcontroller

    

5% 5%  ,1

8/1  

 

IC2 at its pin RB7, which is set up

3,& & 
D
5%  ,1

6 (5,$/ ,1
  
 6( ( 7 (;7 


N


*1 ' 5% 

5% 
 
,1

,1
as an input pin.
*1 ' &2 00

5$ 5$  5 $ 5$
 
When output pin 5 of IC4 is at
   

1&
0V, it switches on transistor TR1
5

N
5

 N
and, via current limiting resistor
R3, causes current to flow
5 5
through opto-coupler IC4 of the
next digit module. In this way the
N  N

modules are daisy-chained one to

6 6 6 6
the next.
Dual-in-line switch S1 to S4 is
used to set the digit’s module ad-
Fig.1. Circuit diagram for the Giant Digital Display module. dress number by placing the rele-

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1117
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW

5 

N ' ' ' '  '  '  '  ' '  ' 
D N N N N N N N N N N

'

' 

D
N N

D D D D D D D D D D
N ' 
N
' 

D D
' 

'

D N D D N

'  ' 

N D D N

'  '

N D D N

' 
' 

N D D N

' 
' 

N D D N

'
'

N D D N

' 
' 

N D D N

' 

'

N D D N

' 
' 

N D D N

' 

' 
'  ' ' ' ' ' ' ' ' ' 
N N N N N N N N N

D D

D D D D D D D D
N N

D
' 

' 

N D D N

'  ' 

N D D N

'  ' 


N D D N

' 
'

N D D N

'
' 

 
N D D N

' 
' 

N D D N

' 
' 

N D D N

' 

' 
5 

N D D N

' 
'

'  D N

'

' 
N D D N

5
' 
D
'  '   '  ' '  ' '  '  '  '  N
D D D D D D D D D
  D N

'

N N N N N N N N N N
N
'  D

5 

5

5 

5 

29

5  5 

6 (5,$/

5  &  &


7+52
287
5

H F

;
& 

75 E
5
21



&  &  &  5

,&
  9
,1 2 87

9

&2 0

 

6 ( 5 ,$ / ' $7$ ,1 3 8 7

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1118
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
COMPONENTS Module No. Switch Settings Display
Digit Module
Resistors
0 off off off off #
R1 470 ohms 1 off off off on hundredth seconds
R2, R4 to R7 10k (5 off)
R3, R15, R16 220 ohms (3 off)
2 off off on off tenth seconds
R8 to R14 100 ohms (7 off) 3 off off on on seconds
All 0.25W 5% carbon film 4 off on off off ten seconds
Capacitors 5 off on off on minutes
C1, C2 15pF ceramic
C3, C6 100n ceramic 6 off on on off ten minutes
C4 470u radial electrolytic, 16V 7 off on on on hours
C5 47u radial electrolytic, 10V
8 on off off off ten hours
Semiconductors
D1 to D78 red LED, 5mm, normal 9 on off off on #
or high brightness 10 on off on off #
TR1 BC558 pnp transistor
IC1 78L05 +5V 100mA 11 on off on on #
voltage regulator 12 on on off off #
IC2 PIC16C54 microcontroller,
preprogrammed 13 on on off on #
IC3 ULN2003 7 x Darlington driver, 14 on on on off #
common emitter
IC4 4N25 or 4N28 opto-isolator 15 on on on on #
Miscellaneous # Used in the computer version with the Serial Port Connector
S1 to S4 4-way DIL on/off switch
X1 3.2768MHz crystal (see text) Table 1: Module Selection Switches
PCB available from the EPE Online vant code on the PIC’s RA0 to IC3 acts as an intermediary
Store (www.epemag.com) code RA3 data pins. Pins RA0 and buffer.
7000247; 6-pin DIL socket; 16-pin
DIL socket, 18-pin DIL socket;
RA1 are normally held at 0V via This device is a rugged little
connecting wire, solder, etc. resistors R4 and R5; pins RA2 chip intended as a solenoid
and RA3 are normally held at driver and can handle almost
Serial Port Connector 5V via resistors R6 and R7. This 50V and 500mA, and is nice
Resistors method of biasing was done and cheap as well. It is essen-
R17 1k2 simply to make the board de- tially seven open-collector Dar-
R18, R19 330 ohm for 9V, sign easier and the software
560 ohm for 12V lington transistors that can be
takes it into account. When the turned on and off by the 5V and
Semiconductors appropriate switch is closed, the
TR2 BD681 (or equivalent, e.g. 0V logic level voltages from the
logic level seen by pins RA0 to PIC.
TIP141 or TIP142) npn
Darlington transistors RA3 is inverted
The LEDs are arranged in
D79 red LED, 5mm The status of the switches is
D80 1N4148 signal diode pairs in a series/parallel ar-
read whenever a serial data rangement, meaning that one
Miscellaneous byte is received by the PIC via
PCB available from the EPE Online
pair is connected in series with
its RB7 input. The 4-bit status the next pair. There is a voltage
Store code 7000248; connector to
suit serial port lead used
code forms the “dddd” bits re- drop of nearly 2V across each
ferred to earlier. LED or pair of LEDs in a parallel
See also the DISPLAY arrangement and the five pairs
SHOP TALK Page!
Pins RB0 to RB6 of the PIC are arranged in series.
Approx. Cost are used as the 7-bit output to Thus the five pairs will drop
Guidance Only $56 the seven sets of 10 LEDs that the 12V supply by 5 x 2V, or
about 10V, leaving the ballast
(Standard LEDs) make up the seven segments of
the display. The PIC16C54 can- resistor with 2V (12V – 10V) to
not by itself handle the current reduce to zero. The LEDs run
required by the LEDs and so well at about 20mA, so a simple

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1119
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
application of E = IR gives a LED on board (D79) to indicate rather tedious. It is suggested
value of 100 ohms for the bal- serial port activity and is a great that you test each segment as it
last resistors. help for trouble shooting. is finished.
The value of the ballast re- Start assembly of the Large
sistor is not critical and the CONSTRUCTION Display board (Fig.3) with the
LEDs will put out good light top right segment. Insert all the
from about 10mA to some The printed circuit boards LEDs and make sure that they
30mA, which is the maximum for the Large Digit Display and are all the correct way around,
for most LEDs. If you need to optional computer Serial Port noting that some high bright-
save power, try putting in 220 Converter Interface board are ness LEDs have different orien-
ohms ballast resistors and see available from the EPE Online tations to those of ordinary
how the light output looks. Store, codes 7000247 and LEDs. If in doubt, you can
7000248, respectively at check by temporarily connecting
The decimal point and colon
LEDs are done the same way,
www.epemag.com the LED in series with a 1k :
except that the LEDs are all in The component assembly resistor across a 12V power
series as there are not as many and track layout details for the supply.
of them. These LEDs are not boards are shown in Fig.3 and Flip the board over and sol-
controlled in any way and are Fig.4. der only one lead of each LED.
simply connected across the There is nothing difficult When you have done that, go
12V power supply, via limit re- about the construction but the back and grasp both leads of
sistors R15 and R16, constantly LEDs are, as may be expected, each LED and re-melt the sol-
remaining on while the power is
on.

SERIAL PORT CON-

VERTER
The digit modules can also
be driven from a computer se-
rial port with the aid of a con-
verter module interface (see
Fig.2). This is simply a Darling-
ton transistor switch (TR2)
†
which converts the 15V signals Completed control and power supply area of the display PCB.
from the serial port to voltages
of the correct polarity to drive '$7$
287 
'$7$
287  der while gently pulling up-
the opto-couplers. wards on the leads. This will
The transistor also provides seat each LED onto the cir-
the reasonably heavy current 75 *1'
cuit board and generally
required by opto-couplers con- N N
'
D make sure it is pointing
nected in “star” configuration '
5 '$7$
straight out from the board.
5
(see the last section of this arti- 
This is important, as high

D
5
cle). 9 brightness LEDs only appear
The converter has its own bright when you look directly
power supply because it has to onto them, if they are tilted
provide power to the internal they look dull and this makes
LEDs of the opto-couplers. The the display look patchy.
248

battery used can be 9V or 12V Go back and solder each

merely by changing resistors second lead and give the
R18 and R19. The values first soldered lead a touch up
:
should be 330 for 9V and with fresh solder if neces-
:
560 for 12V. Fig.4. Printed circuit board details
sary. Now solder in all of the
ballast resistors (R8 to R16)
The converter also has a for the Serial Port converter.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1120
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
',*,7  ',*,7 
and some power leads for the
12V supply.

DISPLAY TEST D
6723:$7&+
25 6(5,$/ 3257
&219(57(5
To test the segment, con-
nect the 12V supply and con-
nect a flying lead to ground
(0V). Touch the flying lead to
the end of resistor R13 that is
nearest to the bottom of the
',*,7  ',*,7 
board. The segment should light
up nice and bright.
If it does not, look for LEDs
the wrong way around, broken
tracks, or the wrong ballast re- 6723:$7&+
25 6(5,$/ 3257
sistor value, in that order. &219(57(5

If all is well, continue insert-

ing LEDs, testing, inserting, E

testing...
If any LEDs are a tight fit at
their skirts, gently file down their Fig.5. Suggested method of connecting the Giant Display
sides until there is room for
them to sit without colliding with
modules to the Stopwatch (Part 1) or Serial Port Converter. (a)
their neighbors. In “daisy chain” fashion or (b) “star” configuration.
Because the colon and dec- turn to pins 1 to 7 of IC3’s Now power down and care-
imal point LEDs are intended to socket and you should see each fully put the PIC
be permanently turned on, they of the segments light accord- (preprogrammed, of course)
(and/or their ballast resistor) ingly. into its socket, being very care-
should be omitted if those func- ful about orientation. Remem-
tions are not required on any of If you have connected the
colon or decimal point LEDs, ber that it is a CMOS chip and
the boards. so be sure to briefly ground
they should have turned on
Put in all the other compo- when you applied the power. yourself to discharge static elec-
nents and sockets for IC2 to
IC4, but do not install the ICs
just yet.
TESTING
Power up the board and at
the IC2 socket test for 5V and
0V at pins 5 and 14. This will
test the power supply regulator
IC1, and will also show up any
solder splashes or broken tracks
to these pins.
Switch off the power and
insert IC3, the LED driver de-
vice. To now test the operation
of the various segments, take a
flying lead and connect one end
to 5V, say to the link wire imme-
diately below IC1. Touch the
other end of the flying lead in One Display module being driven by last month's Stopwatch.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1121
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
tricity before handling it. Also STAR CONNECTION If you only want to drive the
insert IC4. modules from a computer, a
The digit modules are de- slightly different source code for
Turning on the power signed to be hooked up in “daisy the PIC has been included
should now give you a nice big chain” configuration, see Fig.5a, (called serin4.src) which re-
figure “0” and if not, immedi- and this should work well in quires the use of a 4MHz crystal
ately power down and start look- most cases. It is possible, espe-
ing for causes. The Stopwatch instead of the 3 2768MHz one,
¬

cially when many modules are and operates at 2400 baud. The
article last month has some tips used, for the signal to get a bit
on troubleshooting this type of slower baud rate is unnoticeable
lost in its trip down the chain; to our slow human senses and
circuit. remember the design allows up results in a design that is forgiv-
If you are using the Stop- to 16 digit modules to be used. ing of long serial cables and bit
watch module, connect it to one In this case, use the “star” rate errors in the computer or
digit board via a handy length configuration in Fig.5b where micro.
pair of leads, being careful to the driver transistor in the Stop-
connect signal and ground wires watch or Serial Port Converter
the correct way around. Select switches all of the opto-couplers SOFTWARE
the module address number via directly. Note that this will put The software for the Large
the DIL switch (S1 to S4) as per quite a strain on the battery of Digit module, including the
Table 1. Note that the software the Serial Port Converter or QBASIC demo program, is
“knows” that switches S3 and Stopwatch module, as it now available for Free download
S4 are connected in order of has to power all of the opto- from the EPE Online Library at
RA3 and RA2 (instead of RA2 couplers at the same time. www.epemag.com
and RA3 as might be expected).
To select a battery size, as- Preprogrammed PICs for
Power up both boards and sume that each module uses this module are available as dis-
start the Stopwatch. This should about 15mA when running and cussed on the Shoptalk pages.
immediately start the digit board plan accordingly. For example,
displaying the selected time Note that since publication
10 modules times 15mA is
unit. If it just sits on “0’’, use a of Part 1, the software has been
150mA and so a battery of
logic probe or similar to test for revised by the author. The new
1 2Ah (amp hour) capacity will
¬

a fast changing signal on pin 5 version is available from the

drive the display for eight hours. EPE Online Library.
of the opto-coupler, IC4.

PORT INTERFACE COMPUTER SERIAL

If using the Serial Port Con-
PORTS
verter, connect up the digit While developing this pro-
board and power as above. Now ject the author came across a
run the QBASIC demo program, strange fact: not all computer
making sure that the module serial ports operate at quite the
DIL switches are all off. Put in a same speed and the modules
different switch setting from the will consequently malfunction
list each time you run the pro- on some computers.
gram and the module should For those programming
immediately display the correct their own PIC and wanting to
number. drive the modules from a com-
You will know if the con- puter port, try varying the value
verter is working by observing of “bit_k” in the software for the
its LED. Whenever serial data is PIC. The comments section in
being transmitted it will flash the source code tells you how to
quite noticeably. do it.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1122
Maxfield & Montrose Interactive Inc
 $ VHOHFWLRQ RI (/(&7521,& &'5206

IRU +REE\LVW DQG 6WXGHQWV

Computer Desktop Encyclopedia

The award-winning source of computer terms, concepts, important products and
interesting stuff!! An invaluable reference and an unmatched resource for anyone
interested in computers. ($14.99 Special Offer! -- While Stocks Last!)

The PhizzyB Computer Simulator

The PhizzyB Simulator provides an accurate representation of the real PhizzyB,
including the ability to step and run through programs. Comes equipped with an
assembler and other tools, plus user manuals (as PDF files on the CD ROM).

PICtutor by John Becker

This CD-ROM, together with the PICtutor development board (as described when
you click on "more info"), will teach you how to use PIC microcontrollers. The
board also acts as a development test bed and programmer for future projects.

Electronic Circuits & Components

This CD provides an introduction to the principles and application of the most com-
mon types of electronic components, and shows how they are used to form cir-
cuits. The CD also includes the "Parts Gallery".

Digital Electronics (Mike Tooley)

This CD builds on the knowledge of logic gates covered on the Electronic Circuits
and Components CD (described above), and takes users through the subject of
digital electronics up to the operation and architecture of microprocessors

Analog Electronics (Mike Tooley)

This is a complete learning resource for the most difficult branch of electronics.
The CD includes a host of virtual laboratories, animations, diagrams, photos, and
text, as well as a SPICE electronic circuit simulator with over

Modular Circuit Design

This "web on a CD"-based educational software package contains essential infor-
mation for anyone undertaking high school-level electronics or technology courses,
and for hobbyists who want to get to grips with project design..

9LVLW WKH (3( 2QOLQH 6WRUH DW

ZZZHSHPDJFRP DQG SODFH \RXU RUGHU WRGD\

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1123
Maxfield & Montrose Interactive Inc
 by ALAN WINSTANLEY & IAN BELL VOLUME CONTROL
Both methods of volume
This month our team of surgeons commences an control obviously work, in that
opamp extravaganza, lifting the lid off these they both provide control over
indispensable amplifying circuits. Also, fusible the loudness of the sound
produced by the amplifier. I
resistors come under their beady eyes too! cannot say that I have seen a
Lebanon writes: formal rule for which method to
Welcome to the very last
use. However, we can make a
Circuit Surgery column of the Almost every design distinction between the two
1990’s, and we hope there is involving analog electronics approaches in that the input
something of interest to contains an opamp. I know resistance approach is an
everybody in our monthly how an opamp behaves and attenuator whereas the
round-up of readers’ queries the equations that rule its feedback resistance is a gain
and questions. behavior but what I don’t know control.
is what’s inside that black box. Both controls can be used
OPAMPS 101 All that we have been together, in some applications. I
We have had a couple of taught is that it is based on a therefore suggest that the “input
questions about opamps and differential pair of transistors, resistance control” is suitable
think that many readers will find but it’s far from being that when the maximum input signal
a discussion of this subject simple. Can you please supply is at a known reasonably fixed
useful. Mohab Refaat writes by me with some information? level. The amplifier can then be
email: designed to give full volume for
this signal level, and the input is
EPE publishes many It’s best to start with the attenuated for lower volumes.
circuits that involve audio second part of the first
question, because it deals with When the range of possible
effects or amplification. Some input levels is very large
use “low noise” opamps, such something very specific that
follows on neatly from our though, it will be necessary to
as the LF351 or the TL071. My be able to change the gain of
first question is, how can you recent discussion on the 22-
position volume control the amplifier to a level
select an opamp for a particular appropriate for the input being
application out of a large (Circuit Surgery, Oct ’99) and
volume power outputs from used at any one time. Ideally,
number of candidates? the Gain control would be set to
amplifiers (Sept ’99). We will
I found the use of a then move on to a more give maximum undistorted
“Volume” control in simple general discussion on opamps (non-overloaded) output with
amplifier circuits to be another over the next month or so. the maximum input in the
aspect I found a bit baffling.
Sometimes it is achieved using 719>
a pot. (potentiometer) as the
input resistance to the opamp,
B!
sometimes it is used in the
feedback network to control the F?<E=5
9>@ED ?@1!
gain. Are there any rules related
to the use of either method?
 ?ED@ED
Thanks for opening up the world 93!
of electronic circuit design to we
non-electronic engineers in a 7>4
simple way!
Also Tony Soueid from
Fig.1. Circuit to illustrate the two types of volume control.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1124
Maxfield & Montrose Interactive Inc
 &LUFXLW 6XUJHU\
9>F5BD9>7
9>@ED
everyone will
current situation and then left ?ED@ED aim for hi-fi quality – if a very
alone, with a separate control F!
high sound quality is not really
for volume. 
1F F?ED
needed then why pay more for a
F"
However, as gain also special opamp?
affects volume, the gain can be >?>9>F5BD9>7

set to give the desired volume 9>@ED

at any instance and this, of Fig.2. Op.amp circuit symbol INPERFECTIONS

course, reduces the number of Having given the
?ED@ED
controls needed. The F?ED 1H5C >?D D? C31<5 impression that opamp choice is
representative circuit in Fig.1 sometimes somewhat arbitrary,
F?`CQd  !%F
illustrates both types of volume it is worth pointing out that in
control, the input signal shown FY^CQd 719> 1-F?ED F"F! some cases it can make the
 !% PF
on the diagram may be from an difference between a circuit
FY^CQd  49665B5>D91<
external source or an earlier 
 !% FP 9>@ED F"F! functioning or not. I remember
stage in a larger circuit. F?`CQd
working on a partly developed
 !%F
prototype power control system
Fig.3. Graph showing the rela- for a CO2 laser, the existing
MAKING A CHOICE output circuit used a general-
To select an opamp you
tionship between opamp dif- purpose opamp, which was
need to know what the circuit – ferential input voltage and its simply not up to the job.
and hence the opamp – needs output voltage. Saturation oc-
The power measurement
to achieve, this will give you a curs when any increases in worked fine some of the time,
minimum specification for the the magnitude of the differen- but on other occasions would
device. Then purchase the tial input do not result in fur- not do anything. The problem
cheapest opamp which meets ther increases in output volt- was due to the high offset
all the specs! voltages, and more specifically
age. The values shown are an
It may not always be all that example for an opamp with a the drift in offset with time and
simple to calculate an opamp temperature. The circuit was
gain of 100,000 and a maxi-
specification in great detail, but replaced with one using special
you can use a bit of common mum output voltage of †15V. high precision peak
sense too. If your application is The gain of the opamp is measurement chips, which did a
an audio amplifier, it would be equal to the slope of the graph great job.
sensible to use a low noise between the saturation points. The above example
opamp and pay a bit more for a illustrates a couple of points.
better spec. On the other hand, constructional projects in EPE First, it is the imperfections in
if you are using it as a and EPE Online probably does “real” opamps (as opposed to
comparator to, say, switch on a not necessarily follow an “ideal” ones) that cause
heater when the output of a obvious process of selection! problems, so understanding
temperature sensor falls below The projects are designed by these and their impact will help
a certain point, then an ultra low various authors, many projects you avoid devices that are
noise “audio spec” device is not will be feasible with a wide unsuitable. Understanding
really needed. range of general purpose opamp imperfections will also
The range of circuits one opamps, and some authors will help you understand the internal
can design using opamps is so have their favorite ones that circuitry (with reference to the
vast that we cannot give they always tend to use. second reader’s question)
specific recommendations. The In other hobby projects, the because much of the design
best thing you can do is choice of opamp may well effort arises in reducing these
understand what all the opamp depend on what was available imperfections.
specs and ratings mean so that at the time! (e.g. see John Second, there are
you can make an informed Becker’s comment on his choice occasions where specialist chips
choice when necessary. of amplifier in the Musical other than opamps are the best
Having said this, the choice Sundial in the June '99 issue). option. The above case was one
of opamps in all the For audio projects, not example, another good one
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1125
Maxfield & Montrose Interactive Inc
 &LUFXLW 6XUJHU\
would be a sample and hold catalogs usually indicate o) Low-power/micropower –
circuit – you can build one using whether an opamp is intended suitable for use in systems
an opamp, but you will get for single or dualsupply such as mobile equipment,
better performance from a operation, otherwise check the where power consumption
sample and hold chip, and, of data sheet. The power supply is critical
course, some extra bits of connections are not always o) Wideband/high speed –
circuitry are already included. shown on schematics. for applications such as
Comparators are another case – The output voltage of an pulse circuits and video
all opamps can be used as opamp is given by VOUT = AV x where accurate
comparators, but it is often (V2 – V1) where AV is the open reproduction of complex
better to use chips that have loop voltage gain, V2 is the non- high frequency signals is
been optimized for this purpose. inverting input voltage and V1 is required
the inverting input voltage. This o) High-power/high current
ON SPEC “open loop” gain refers to the – opamps with high current
gain of the opamp itself without output stages capable of
It is worth looking at some any feedback circuitry. Opamps
of the specifications found on driving low impedance
are almost always used with loads
opamp data sheets and in some form of feedback though,
suppliers’ catalogs and which results in a gain for the o) Low drift/high precision
discussing how these may circuit that is different from that – amplifiers with minimal
effect your choice of opamp for of the opamp itself. offset voltage, and where
particular applications. We will accuracy is preserved over
also refer to some of these Note that the opamp a wide temperature range
specifications when we move on amplifies the difference in
voltage between its two inputs. o) Low bias/high
to look at the internal circuitry. impedance – FET input
But, before we start, we need to It is a differential amplifier.
The equation VOUT = AV x (V2 – opamps with very low input
define some basic things about bias currents for use in
the opamp, so let’s explore V1) always holds for totally ideal
device, but in reality is only buffer circuits or with large
them in greater detail. external resistors.
valid for a small range of (V2 –
The opamp is a high-gain, V1) and there are limits on the
direct-coupled amplifier, its Some opamps may
individual values of V2 and V1 arguably belong to more than
symbol is shown in Fig.2. The too. The opamp’s input-output
term “direct-coupled” means one of these categories. The
relationship is illustrated in Fig. specifications given on opamp
that the inputs and internal 3.
stages are connected directly, data sheets can be divided into:
not via coupling capacitors. This Some manufacturers group electrical ratings (maximum
enables the opamp to amplify their opamps into types suited voltages etc.); signal handling
DC and very low frequency to different kinds of application. (noise etc.) and offsets (which
signals. Typical descriptions may particularly effect DC accuracy).
include: We’ll discuss these and other
The opamp has two inputs – practical matters in the next
the inverting (–) and non- o) General-purpose – Surgery. IMB.
inverting (+) inputs – and an suitable for a wide range of
output. The inputs and outputs applications requiring
are usually referenced (applied moderate amplifier FUSIBLE RESISTORS
or measured with respect to) performance
Mark Lee asks: “I would
ground or 0V. o) Low noise – guaranteed appreciate an explanation of
Opamps usually have two very low noise for fusible resistors and how to use
power supplies, one at a applications such as them. They seem to be mainly
positive voltage with respect to sensitive measurement low resistance and low power
ground and the other at the and signal processing ratings. How do I use them in a
same magnitude negative with where noise from the circuit?”
respect to ground; however opamp must be within
many “single supply” opamps known bounds
are also available. Suppliers’ Fusible resistors are
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1126
Maxfield & Montrose Interactive Inc
 &LUFXLW 6XUJHU\
inserted into a circuit as an This means that they will The main thing is that they are a
ordinary resistor would be, disconnect the circuit, rather than fault-tolerant, fail-safe fireproof
except that they have the burning out or setting fire to the device.
special property that if they are board. They are only produced in Paradoxically there is even
overloaded for any reason (a a limited range of values (low a zero ohm resistor available!
circuit fault elsewhere ohms to a few kilohms) and would These are used by
downstream), then instead of be used in e.g. power supply or manufacturers using automated
burning out they are guaranteed monitoring circuits, where a PCB equipment, to apply a link
to go open-circuit within a combination of resistance and between two pads – it means
certain range of conditions. overload protection is required. that a machine which handles
resistors is, therefore, also able
to insert the equivalent of a wire
link instead. ARW.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1127
Maxfield & Montrose Interactive Inc
ROBERT PENFOLD fault caption should be erased.
A serial approach to PC add-ons This gives the program some-
where to display the data read
from the serial port. When the
In the previous Interface MOUSE EXPERIMENT program is run, the data dis-
(Oct ‘99) article we considered If your PC has a serial port played on the label should
the subject of serial port inter- mouse it is easy to experiment change as the mouse is moved
facing, and using a standard with serial port reading, and the around and the buttons are op-
RS232C serial port to send data raw data from the mouse can be erated.
to a user add-on. In this month’s read from the appropriate
Interface we will look at using a mouse port. Using Delphi 1, this
serial port to receive serial data. SYNCHRONIZATION
code could be applied to a timer
Much of the background in- component set with an interval For many applications it is
formation provided in the previ- of about 50 milliseconds: perfectly all right to take this
ous article also applies to using simplistic approach, and simply
a serial port to receive data. Re- Reading := Port[760]; read the port periodically to ob-
fer to the earlier article if you Str(Reading, S); tain the latest data available.
require information on the Label1.caption := S; For example, suppose that a
UART registers, setting the Thermometer is connected to
word format and baud rate, etc. The value of “760” in the the port. By reading the port the
first line is correct if the mouse latest temperature will always
With things reduced to the is on serial port two, but must be read and displayed. The fact
simplest level it is not difficult to be changed to “1016” if it is on that each new piece of data
read data from a serial port. Us- port one. The two variables may be read several times or
ing the methods outlined in the must be declared in the appro- the odd reading may be missed
previous Interface article it is priate part of the program by here and there will be of no
possible to set the required adding the following two lines in practical consequence.
baud rate and word format, and the section headed “var”:
is then just a matter of reading This is not the case in all
data from the base address of Reading : Byte; applications though, and in
the port. The base addresses some cases it may be neces-
S : String;
for serial ports one and two are sary to operate on the basis of
respectively &H3F8 (1016 deci- A label component must be sending a trigger signal to the
mal) and &H2F8 (760 decimal). added to the form, and its de- interface, and then reading in

Fig.1. Circuit diagram for the Simple Serial Interface add-on. It operates at 9600 baud.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1128
Maxfield & Montrose Interactive Inc
 ,QWHU)$&(
“x” number of bytes. It then be- Label1.caption := S; next line. This line repeatedly
comes essential to properly syn- Counter := Counter + 1; reads the Line Status register,
chronize the sending device and Str(Counter, S); and bitwise ANDs the result with
the program reading the data. one. This effectively strips off
Label2.caption := S;
Otherwise there is a risk of (say) bits 1 to 7, and reads only bit 0.
reading four bytes of data The port addresses used The program loops until the
twenty-five times each instead here are for port one. For serial returned value is one, which
of reading 100 bytes of data port two use these addresses: means that there is a fresh byte
once each. of data to be read. The port is
760 instead of 1016
There is no need for an ex- then read and the result is dis-
ternal handshake line to control 761 instead of 1017 played on Label1. Then the
the flow of data, and a status bit 763 instead of 1019 Counter variable is incremented
of the line Status register can 765 instead of 1021 by one and the new value is dis-
be used instead. Bit 0 of this played on Label2.
register is set to one when a In addition to applying this
complete byte has been re- program to a timer component
ceived and transferred to the these three lines must be used HARDWARE
receiver register. Writing a zero to declare the variables. The Simple Serial Interface
to this bit will reset it, but this is Reading : Byte; of Fig.1 can be used to test this
not normally necessary as it is program. The 6402 UART has
automatically cleared when the Counter : Byte;
been covered in previous arti-
data in the receiver register is S : String; cles and will not be discussed in
read. A further line must be detail here.
To ensure that each byte of added to the listing for the form, Transistor TR1 generates a
data is read only once it is just a and this sets the Counter vari- 2 4576MHz clock signal that is
¬

matter of using a software loop able at an initial value of zero. divided by 16 through IC1.
to monitor the received data bit, UART IC2 requires a clock sig-
and provide a hold-off until it is Counter := 0;
nal at 16 times the required
set to one. This prevents the baud rate, and this gives an out-
receiver register from being
read until a new byte of data is
OPERATION put signal at 9600 baud.

ready. In the original test program The control inputs at pins 34

we relied on the operating sys- to 39 of IC2 are hard-wired to
The Delphi 1 program de- produce the required word for-
tem to set up the serial port cor-
scribed previously is easily mat of 8 data bits, one stop bit,
rectly, but in real world applica-
modified to provide this hold- with no parity checking. Transis-
tions the program must do this.
off. In addition to the hold-off, tor TR2 acts as a simple line
The first four lines of the listing
this listing also implements a driver and inverter, but it does
set the port for 9600 baud oper-
counter that shows the number not provide proper RS232C out-
ation with a word format of eight
of readings that have been put levels.
data bits, one stop bit, and no
taken. A second label compo-
parity checking. Good results should still be
nent must be added to the form
to accommodate the counter. Using the control registers obtained provided the cable
to set the baud rate and word used to connect the interface to
Port[1019] := 128; format was covered in the previ- the computer is no more than a
Port[1016] := 12; ous Interface article, and will not few meters long. The output
be discussed again here. The (SK1) connects to the Ground
Port[1017] := 0;
fifth line switches off interrupts, and Receiver Data input of the
Port[1019] := 3; RS232C interface. These are at
and should ensure that the op-
Port[1017] := 0; pins 7 and 3 respectively for a
erating system does not upset
Repeat until things by reading bytes of re- 25-pin port, or 5 and 2 for a 9-
(Port[1021] and 1) = 1; ceived data. pin port (see Fig.2).
Reading := Port[1016];
The hold-off is provided by
Str(Reading, S); the Repeat...Until loop in the
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1129
Maxfield & Montrose Interactive Inc
 ,QWHU)$&(
IN CONTROL the maximum achievable trans- suggest that certain handshake
fer rate would probably be inputs could be read at the Mo-
A serial interface requires slightly less than this. dem Status register.
some form of control logic cir-
cuit to trigger the UART at the Although the routines pro- However, initial attempts at
appropriate times and send a vided here are written in Delphi writing to and reading from
stream of data. At its most basic 1, using the methods described handshake lines failed totally.
the control logic can consist of in previous Interface articles it Possibly these lines are imple-
nothing more than anoscillator, should be possible to use other mented via some other means,
which is all that is used in this versions of Delphi, Visual BA- but using them directly seems
case. SIC 6, or even GW BASIC. It is to be something less than
just a matter of outputting the straightforward.
A low power 555 timer, IC3, correct values to the serial port
is used in the standard oscillator registers, and then reading the
configuration. The values of base address. If a software
timing components resistors R8, hold-off is needed it might be
R9 and capacitor C5 set a low necessary to use a different
operating frequency of roughly loop structure with some lan-
1Hz. Therefore, about once per guages, but it should not be too
second the output of IC3 (pin 3) difficult to apply the same bit-
goes through a high to low tran- wise ANDing and looping tech-
sition and causes IC2 to send nique to provide the hold-off.
the eight-bit value on its inputs. Fig.3. The counter of the test
Although the test program is program should increment
assigned to a timer component EXTRA OUTPUTS each time fresh data is re-
that tries to take a reading every One or two readers have ceived.
50 milliseconds, which works asked whether it is possible to
out at 20 readings per second, it use some of a serial port’s
will only take about one reading handshake lines as general-
per second. This is due to the purpose outputs. The UART
software hold-off looping the data sheet would seem to sug-
program for about a second un- gest that the Data Terminal
til a new byte of data has been Ready (DTR) and Request To
received. If everything is work- Send (RTS) handshake outputs
ing properly, the counter should are respectively controlled by
therefore increment at about bits 0 and 1 of the modem con-
one and not 20 per second. trol register. This is at address
A slightly beautified version 764 for port two and 1020 for
of the program after 30 seconds port one. It would also seem to
of taking readings is shown in
Fig.3, the count has reached 31.
Of course, data can be trans-
ferred at a greater rate by in-
creasing the operating fre-
quency of IC3 and reducing the
time interval of the timer com-
ponent (or simply having a rou-
tine that continuously tests the
serial port).
However, bear in mind that
there are ten bits per serial byte
and with a baud rate of 9600
this works out at an absolute Fig.2. Interface connection de-
maximum transfer rate of 960 tails for 25-pin and 9-pin se-
bytes per second. In practice rial ports
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1130
Maxfield & Montrose Interactive Inc
TRANSISTOR TECHNOLOGY IS POISED FOR SOME MAJOR ADVANCES IN THE
NEAR FUTURE - IAN POOLE REPORTS.

It is just over fifty years could be further from the truth. The way in which they curl
since the first transistor was Many new ideas are surfacing, determines their electrical
made at Bell Labs. Since then and these will enable transistor characteristics.
many advances have been technology to surge forward and Fortunately, it is possible to
made, enabling the meet the demands to control the way in which this
performance to be improved tomorrow’s technology, both in curling takes place. By rolling it
beyond all recognition. performance and size. in a way that gives a straight
Bipolar technology has molecular alignment, the
improved from the early NANO-CURLS nanotube behaves like an
transistors that had cut-off ordinary conductor. However if
The idea of nanotubes has the curl is arranged so that
frequencies of only a few been covered previously within
kilohertz and low gains, to the molecular structure is twisted
this column (December 1998 then the nanotube behaves like
state where radio frequency EPE Online), but only in the
(RF) transistors are available a semiconductor.
application for producing very
that can operate to frequencies low resistance and high current A considerable amount of
of many Gigahertz and with carrying capacity conductors. experimentation was required to
much higher levels of gain than The nanotubes used for the enable the right properties to be
were previously possible. Not transistors that are being obtained. It was necessary to
only this, but field-effect developed are subtly different, have the right amount of curl. In
transistors (FETs) are now forming a semiconductor rather fact, the early nanotubes
widely available. than an ordinary conductor. consisted of multiple concentric
It is interesting to note that Although the concept has been layers. However, the nanotubes
the development work to devise known for several years, the that are used now consist of just
a semiconductor amplifying technology is revolutionary and a single wall comprising of a
device was initially focussed until recently it has not been single atomic thickness.
towards the development of a possible to realize it in a
field-effect device. However, physical form. TRANSISTORS
they were unable to make the Nanotubes used for
effect work, and they changed Having developed the basic
transistors have carbon walls semiconductor, the next major
the line of the investigations, made up from hexagonal
which resulted in the hurdle was to develop a useful
shaped matrices. Essentially device. Surprisingly, two
development of the bipolar they are vaporized carbon that
transistor. organizations announced they
has been condensed into a had succeeded. In 1998 the IBM
It took a few years before series of hexagons. To give a Thomas Watson Research
the field effect transistor was better view of what they are, Laboratories and the Delft
widely available. FETs also had they can be considered as a University in Holland both
a major impact on integrated very thin strip cut out of a claimed they had managed to
circuit technology, enabling the graphite carbon plane which has fabricate a transistor using this
degree of integration to be been rolled up and sealed at revolutionary new technology.
considerably increased. either end. The dimensions are
naturally very small, and the The device consisted of a
With transistor technology single nanotube having a
now very mature, it might be dimensions are measured in
atomic proportions. thickness of one atom. Once
thought that the rate of rolled the tube was about one
development would slacken as The carbon hexagons that nanometer in diameter. This
fewer developments were are used to make the tubes was connected between two
possible. However, nothing have a natural tendency to curl. electrodes that were about 400

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1131
Maxfield & Montrose Interactive Inc
 1HZ WHFKQRORJ\ 8SGDWHV
nanometers apart, and the particularly apparent. The very development work is still
whole structure was mounted on thin gate insulation has to be progressing. It is quite possible
a silicon substrate onto which a completely free from impurities that developments in other
layer of silicon dioxide had been as a result of its extremely small areas of semiconductor
set down to act as insulation. dimensions. technology may enable the
The nanotube then acted as the Atom leakage is also a requirement to be met by the
channel whose conductivity problem and interconnection time the development of the
could be controlled in the resistances also have to be nanotube transistor technology
normal way. investigated. The experimental has reached a sufficiently
devices produced so far have advanced stage.
Although the channel length
of the early development model had problems arising from the In order to introduce this
was relatively large, it could be very high resistance between new technology onto the market
made very much smaller. In the nanotube used for the new fabrication techniques are
fact, some working lab models channel and the contacts. required. This results from the
have been made with lengths of In current experimental fact that the extremely small
around 40nm and it is estimated devices the resistance has been sizes mean that “quantum well
that in future channel lengths of of the order of one megohm. effects” become an issue.
only 20nm should be Clearly there will be many To overcome this, new
achievable. advantages to be gained from materials are needed and in
As the speed of a reducing this value. By turn this leads to the fact that
transistor’s operation is primarily comparison the discreet FETs new processes and lines will be
controlled by the length of the that are widely used in today’s required. However with some
channel, this will result in a circuits have channel existing technologies nearing
considerable speed increase. resistances of only a few the end of their roadmaps, the
This means that considerable hundred ohms. The higher need for new technologies like
improvements will be possible values currently being obtained these nanotube transistors will
over the latest production FETs in the new devices will reduce be required to ensure that
fabricated using the latest 0 18 ¬
the high frequency performance semiconductor technology can
micron process, which have of the whole circuit in which keep up with developments in
channel lengths of around they are used. other areas and possibly stay
120nm. one step ahead.
WAFER THIN
FUTURE Another area that is being
This technology is very new investigated is that of producing
and still very much in its suitable wafers. Those that can
experimental stages, and much be produced at the moment, on
basic work is being undertaken an industrial scale, do not have
to ensure that the process can a sufficiently fine surface to
be reliably introduced into enable the minute nanometer
production apart from sizes required for the new
developing the basic transistors to be fabricated
technology. As a result it is sufficiently accurately. A rapid
likely to be several years before thermal oxidation process is
nanotube transistors are being developed but even when
available. Nevertheless, work is this has been perfected it is not
progressing apace. One of the expected that it will support the
problems results from the sizes below about 50nm for
minute dimensions used in commercial production, and this
these devices. It makes them will mean that the full capability
less robust and more open to of the new technology will not
problems arising from be realized.
impurities. The gate insulation Whilst no obvious solution
area is one where this is is even on the horizon,

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1132
Maxfield & Montrose Interactive Inc
 35$&7,&$/ 26&,//$725 '(6,*16
E\ 5$<021' +$,*+
Most text books deal with oscillators in a theoretical way. This series, prepared
with the electronics enthusiast and experimenter very much in mind, is intensely
practical. Tried and tested circuits are fleshed out with component values, and
their vices and virtues are exposed.
PART SIX – RESISTOR/CAPACITOR OSCILLATORS

Thus far we have covered so. shifted through 180 degrees.

oscillators which rely on quartz Resistors and capacitors fix This can be achieved by
crystals or inductors and capaci- the frequency of oscillation by inserting a network of resistors
tors to determine the operating controlling the phase of feed- and capacitors in the feedback
frequency. In this final part of the back, or by timing the action of path, the component values de-
series, circuits in which resistors switching circuits. termining the frequency at
and capacitors perform this func- which the desired phase shift
tion will be considered. takes place. In this way, the R/C
PHASE SHIFTING network fixes the frequency of
Resistor/capacitor (R/C) oscil-
lators are widely used for the gen- The signal at the base oscillation.
eration of specific waveforms (input) of a common emitter If care is taken with the as-
(e.g., sine, square, sawtooth) transistor stage is 180 degrees sociated circuitry, phase shifting
over the 5Hz to 50KHz range. out of phase with the amplified R/C oscillators can generate
Circuits of this kind will oscillate signal at the collector (output). sinewaves of high purity. The
from well below 1Hz to above For oscillation to take place, Wien bridge oscillator is the
2MHz, but a high degree of fre- feedback from collector to base classic example of circuits of
quency stability and waveform must be in phase, and the out- this kind. Here, the R/C network
purity becomes increasingly diffi- put signal must, therefore, be
cult to achieve above 100kHz or 75$16, 6725 / ($'28 76
9$/8(6 2)

&  &
$3352;

)5(48(1&<

PHASE SHIFTERS
Q +]

Q N+]
H E F
Q N+ =

%&  S N+ ]

A simple phase shifting os- %& 

5
%& 

cillator suitable for generating N

spot frequencies is given in

Fig.1. The usual formula relat-  

 9 72

ing frequency to resistance and  9

capacitance in circuits of this 5

& & &

kind is:
75
 N –
&

Q
%&

f = 65,000 / RC E
F

5 5

 N N 

when f is in Hertz, R is in
6,*1$/ 28 7

H . . .
9 5 0 6 :,7+

ohms, and C is in uF.

$ 9 6833/<

With this particular arrange-

ment, the frequency of oscilla-
tion is usually about 20 percent
lower than the figure derived by
Fig.1. Circuit for a simple spot frequency sinewave
calculation. oscillator
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1133
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
9$/8(6 2) $3352;

75$16,6725 /($'2876

&  & )5(48(1&<

is configured to give zero phase Q +]

shift at the frequency of oscilla-

75 75

Q N+]

H E F G J V Q N+=

tion. %&

%&
1
S N+]

%&

RELAXATION 5

N

OSCILLATORS 75
 
1


G
&

Capacitors take time to be-

9 72

Q

J 9

come charged when a DC volt- & & &

V

&

µ
age is applied across them via a 5

– –

75
N

resistor. The larger the values 5 5

&

%& 0 N
µ
of resistance and capacitance in

F

E
5 5

the series circuit, the longer the N N

charging time.
H

6,*1$/ 287

95 95 5 95

. . .
The rising voltage across Ω
9 5 0 6 :,7+

 N N N

$ 9 6833/<

the capacitor, as it is being

charged, can be used to trigger
a change of state in a transistor
switching stage. If this also re-
Fig.2. Circuit diagram for an adjustable spot frequency
sults in the capacitor being dis- sinewave oscillator with an output buffer stage.
charged, the cycle will start Accordingly, three R/C ele- tion will raise it.
again, and we have a circuit ments, each shifting the phase
which oscillates at a frequency The circuit is essentially a
by 60 degrees, are cascaded to spot frequency signal generator
determined by the amount of produce the required 180 de-
resistance and capacitance. that can operate from below 50Hz
grees phase inversion. Signal up to more than 50kHz. Its output
Arrangements of this kind attenuation is reduced to ac- waveform is of tolerable quality,
are known as relaxation oscilla- ceptable limits, but the amplifier but the impedance of the accept-
tors. They produce saw tooth or must still provide a gain of at ing circuit must be high or oscilla-
square waveforms which are least 29 times for oscillation to tion may be inhibited. An
rich in harmonics. Unijunction be maintained. impedance of 47kilohms, which
transistor and multivibrator os- In Fig.1, the combinations halves the signal output, should
cillators operate in this way. of R3/C1, R4/C2, and the input be regarded as the acceptable
resistance of TR1 (in parallel lower limit for reliable oscillation.
PHASE SHIFT with R1) combined with capaci-
tor C3, form the three-stage
OSCILLATOR phase shifting network. It should IMPROVED PHASE
A simple oscillator in which be noted that the capacitors and SHIFT OSCILLATOR
a network of resistors and ca- resistors in the network have
With the addition of two pre-
pacitors are used to shift the the same value. Increasing the
set resistors (potentiometers) and
phase of the feedback is shown amount of resistance and/or ca-
an output buffer stage, TR2, the
in Fig.1. Here, transistor TR1 is pacitance will lower the fre-
performance of the circuit is con-
configured as a common emit- quency of oscillation: a reduc-
siderably improved. The modified
ter amplifier with the output de-
veloped across the collector (c)
load resistor R2. Bias is applied
via resistor R1. COMPONENTS
In theory, a single resistor Capacitors used for phase shifting or timing should be
and capacitor combination can polystyrene, polyester, or Mylar film types. When identical capacitors
shift the phase of a signal are required (see Figs.1, 2 and 3) they should be of 10 percent toler-
through 90 degrees. This capa- ance or better.
bility cannot be utilized in prac- Note that this only applies to the circuits given here. Some phase
tice, however, because the sig- shift oscillators require 1 percent tolerance components before they
nal is excessively attenuated. will operate reliably.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1134
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
R/C OSCILLATORS version of the circuit becomes
an adjustable spot frequency WIDE RANGE AF
Combinations of resistors oscillator and is shown in Fig.2. GENERATOR
and capacitors can be used to The upper frequency limit is
fix the frequency of an oscilla- The frequency selective
around 60kHz, and the amplifier
tor. They do this in two ways: network at the heart of most au-
must have a gain of at least 29
dio signal generators was de-
o) By determining the phase times in order to maintain oscil-
vised by Wilhelm Wien, a Ger-
of signals in a positive lation.
man physicist, about a century
feedback loop. Circuits of Negative feedback devel- ago. Originally used as a mea-
this kind can generate high oped across the unbypassed suring bridge, the combination
quality sinewaves. emitter “resistor”, preset VR1, of series and parallel R/C ele-
o) By timing the switching of reduces the gain of transistor ments produces a network
the maintaining devices TR1. Setting this resistor so that which imparts zero phase shift
between on and off states. the circuit will only just oscillate at one frequency.
Arrangements of this kind results in the generation of a
Because there is no phase
are known as relaxation sinewave of high quality.
shifting within the R/C network
oscillators. They generate Replacing part of one of the at the frequency of oscillation,
square, sawtooth or pulse resistors (R3) in the phase shift- maintaining amplifiers for Wien
waveforms. ing network with pre-set VR2 bridge oscillators must have two
enables the frequency of oscil- stages. (Each stage imparts a
tary switch S1a and S1b. The lation to be adjusted slightly. (At shift of 180 degrees, and this
specified values more than 10kHz it can be shifted by plus results in the output being back
cover the entire audio frequency or minus 1 5kHz).
¬ in phase with the input). Pro-
spectrum. vided the gain of the amplifier is
The FET (field effect tran-
Ganged potentiometers, sistor) source follower stage three times or greater, oscilla-
VR1a and VR1b, form the resis- TR2 presents a very high tion will be maintained. With
tive arms of the bridge and set impedance to the oscillator and such a modest gain requirement
the frequency. Range limiter a suitably low impedance to the it is not difficult to apply heavy
resistors R1 and R2 ensure con- accepting circuit. Gate resistor negative feedback in order to
sistent operation over the full R5 is connected to a tapping on stabilize signal amplitude and
sweep of the potentiometers. the source resistor formed by improve waveform quality.
R6 and R7, rather than to the Wien bridge oscillators vary
negative rail. in complexity, and a simple, in-
AMPLITUDE expensive, yet very effective
By this means, correct gate
CONTROL biasing can be maintained with version of the classic circuit is
In order to obtain a high TR2 source (s) held at about given in Fig.3; a low distortion
quality sinewave, signal ampli- 4V, and this greatly improves audio frequency (AF) Signal
tude must be kept below the the signal handling capability of Generator. Here, the Wien net-
level at which the maintaining the stage. Moreover, the gate work is placed in a positive
amplifier begins to overload. resistor R5 is partially boot- feedback loop around a 741 op-
(Overload causes clipping or strapped and this increases in- erational amplifier IC (The feed-
flattening of the waveform put impedance to almost 10 back must be in phase, so the
peaks). megohms. non-inverting input at pin 3 is
used.)
Automatic control of signal Decoupling capacitor C5
amplitude in Wien bridge oscil- will not be needed in all cases. A low current filament lamp
lators is usually effected by an Variable potentiometer VR3, LP1 shunts a negative feedback
R51 type thermistor connected to the source of TR2 path (between output pin 6, and
(temperature dependant resis- by DC blocking capacitor C6, inverting input pin 2) in order to
tor). These devices are sensi- enables the output level to be stabilize the amplitude of oscil-
tive but expensive, and here an adjusted. lation. Bridge capacitors, C1 to
ordinary low-current filament C8, are selected by ganged ro-

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1135
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
$/7(51$7,9( 75$16,6725 /($'2876
',2'(

RELAXATION OSCIL- back and hold the sig- D

1
N

nal amplitude con-

LATORS stant.
6(( 7(;7 H E F

%&
%&
E H F

%)
%)
%&

Charging a capacitor, via a In practice, preset

resistor, is the most common VR2 is adjusted to
means of fixing the frequency of give the highest possi- 5
N
5
N
5
N 72
5
N 72
5
N

relaxation oscillators. The larger ble output consistent

N N

the capacitance and/or resis- with a perfect sinu- &

S 72 &
µ Q 

tance in the series circuit, the soidal waveform. If an 6(( 7(;7

&
S 72

longer the charging time and oscilloscope is not  Q

9

the lower the frequency of oscil- available to display a

lation. 75 75
trace, good results can – %)
F
%)
F

A widely used circuit of this be ensured by setting E E

kind is the astable multivibrator, VR2 so that oscillation H H

and a version which permits is only just main- 5

N
5
N
&
Q

some adjustment of the operat- tained. A 47 ohm pre-

ing frequency is given in Fig.4. set should be substi- 95
6,*1$/ 287
N

tuted for VR2 and R3 . . .

$3352;

The frequency determining & 9 5 0 6

if a supplier can be Q

networks comprise R3/C2 and

found. There is some
R5/C1. For an equal mark/ Fig.4. Circuit for an astable multivibrator,
amplitude “bounce”
space ratio (off pulses and on
when the frequency is with frequency shifting arrangement.
pulses of equal duration), R3
changed rapidly, and
and C2 must be identical to R5 FETs as voltage-variable con-
this is a feature of all Wien os-
and C1. trol resistors, or diodes as am-
cillators which incorporate a
A very approximate formula temperature dependant resistor plitude limiters, have been de-
relating frequency to resistance as a control element (the resis- vised to overcome this
and capacitance is: tance heats and cools compara- “bouncing”. However, unless the
tively slowly). Circuits using design is complex, they usually
f = 700,000 / RC exhibit higher distortion.
when f is in Hertz, R is in &
 75$16,6725
µ
ohms, and C is in uF. 723 9,(: /($'287


75
 
The frequency of oscillation H E F 
&
Q

is very dependant upon supply %&

3


  &

voltage and, to a lesser extent, 6E

 Q

transistor types, and the formula 5$1*(6 

& 95

is inevitably approximate. The

. +]  +]
5 Q N
 
Ω

9 72
. +]  N+] 
9
output is a square wave with a .

.
N+]  N+]

N+]  N+]

rounded leading edge. 95D

N
5

75
0

lamp is used in its place. 

  & %&
F

,&
Q

The resistance of a lamp 

 E
&
 µ


filament rises dramatically when 3 95E

N

H

current flows through it and 

6D

95
 Ω
5 5
raises its temperature. If the 

 Ω N

output at pin 6 increases, more &

 µ
&
Q

current flows and its resistance &

Q
&
Q /3
6,*1$/ 287
. . .
9 5 0 6
rises. Lamp LP1 is connected ),/$0(17 /$03
9 P$
&
5  µ
as the lower arm of a potential  Ω 25
9 P$

divider, VR2/R3 forming the up- 9

per section. An increase in the

resistance of the lamp will, ²9 72 ²9

therefore, increase the amount ²

of gain-reducing negative feed- Fig.3. Circuit diagram for a low distortion AF signal generator.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1136
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
WIEN BRIDGE to the required level and im- thousandth). A typical circuit
proves linearity. In theory, the arrangement, with the addition
A network of resistors and gain of this stage is approxi- of frequency adjusting refine-
capacitors, known as a Wien mately VR3 divided by R5 (i.e., ments, is given in Fig.4.
bridge, is used to determine fre- four times), but, in practice, it is Two common emitter tran-
quency in most professional au- rather less than this. Base bias sistor stages, TR1 and TR2, act
dio oscillators. With this net- is provided by resistor R4, C10 as switches, and their bases
work, phase shift is zero at one is a decoupling capacitor, and and collectors are cross coupled
particular frequency. A typical C11 blocks the flow of DC into by capacitors C1 and C2. Base
circuit is given in Fig.3. the accepting circuit. biasing is supplied by R3 and
The resistors and capacitors R5. These resistor and capaci-
in each arm of the bridge tor combinations, R3/C2 and
(VR1a/VR1b and C1/C5, C2/C6,
PERFORMANCE
R5/C1, act as the timing net-
etc.,) are of equal value, and Although simple and inex- works which determine the fre-
the standard formula relating pensive, the AF Signal Genera- quency of oscillation.
frequency to resistance and ca- tor circuit performs well when
preset VR2 has been correctly The coupling capacitors al-
pacitance is:
adjusted. Distortion figures as ternately charge, via the bias
f = 160,000 / RC resistors, and discharge, via the
low as 0 1 percent are claimed
¬
transistors, and the rising and
when f is in Hertz, R is in for circuits of this kind, and a falling voltages on the capaci-
ohms, and C is in uF.The actual check with an oscilloscope will tors switch the transistors on
frequency of oscillation is reveal that the sinewave is of and off, thereby maintaining the
around 10 percent lower than high quality. circuit action. The frequency at
the figure indicated by calcula- Output level remains con- which the switching, or oscilla-
tion, and the ranges quoted in stant over fairly wide shifts in tion, takes place is, of course,
Fig.3 are based on actual mea- supply voltage, and across the determined by the time con-
surements. switched ranges. Oscillation is stants of the R/C combinations.
The amplifier need only maintained up to 70kHz, but Collector loads are formed
have a gain of three times for performance begins to fall off a by resistors R2 and R7. Capaci-
oscillation to be maintained. little after 30kHz or so. tor C5 decouples the circuit
This modest requirement per- Constructors would have to from the supply line and C4
mits the use of heavy, ampli- commit themselves to consider- blocks the flow of DC into the
tude controlling negative feed- ably more expense and effort in accepting circuit.
back, and the quality of the gen- order to realize any significant
erated sinewave can be ex- improvement on this circuit.
tremely high. Note that the oscillator will not ADJUSTING THE
function correctly if a lamp with FREQUENCY
a higher wattage rating, or a
lower working voltage than 6V, The operating frequency of
OUTPUT LEVELS is fitted. simple astable multivibrators is
The simple control circuitry very dependant upon supply
places a rather low resistance voltage. Their frequency can
across the amplifier output, and RELAXATION also be shifted by applying a
variable bias to the base (b) of
the signal voltage available be- OSCILLATORS the transistors in order to modify
fore the onset of distortion is
limited to around 1V RMS. A The most common form of the triggering action.
larger output is often desirable, relaxation oscillator is the
Potentiometer VR1, con-
and the buffer stage transistor astable (i.e. non-stable) variant
nected across the supply via
TR1, in Fig.3, provides a mod- of H. Abraham and E. Bloch’s
range limiting resistor R1,
est amount of signal amplifica- multivibrator. Conceived by the
varies the voltage on the bases
tion. two Frenchmen in 1918, the
of the transistors. Resistors R4
name “multivibrator” was given
Emitter (e) resistor R5 is and R6 isolate the signal paths
to this type of circuit because
unbypassed, and the resulting and capacitor C3 decouples the
the output is rich in harmonics
negative feedback reduces gain bias supply. This arrangement
(they can extend beyond the
permits a fairly wide adjustment
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1137
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW

723 9,(:
of the nominal operating fre- )5(48(1&< 5$1*(6
  . +]  +]
quency, typically plus or minus . +]  N+]

20 percent. . N+]  N+]

5   . N+]  N+]

If a basic multivibrator is all N

that is required, omit VR1, R1,

R4, R6 and C3. 95
N



9 72

OPERATING 

5(6(7 


9 
9

,&
&9 1&

FREQUENCY 5
N

',6 
  –
The timing (bias) resistors 
7+5(6 287387

75,*
R3 and R5 can range in value 5$1*( 3
&
µ

from 47k to 470k, and the ca- 

6
*1'


pacitors, C1 and C2, from 47pF 


95
N
to several microfarads. This & &
µ
gives an operating range ex- &
Q
& 6,*1$/ 287
$3352; 9 33
tending from sub-audio frequen- Q Q

cies to 2MHz.
Small signal AF transistors Fig.6. Using the renowned 555 timer IC to
can be used up to 100kHz, but produce a 50Hz to 200kHz pulse generator.
r.f. devices will ensure reliable
oscillation at higher frequencies. CMOS SQUARE WAVE between gates IC1a and IC1b.
Suitable transistor types are The third gate, IC1c, is used as
also included in the circuit of GENERATOR a buffer stage.
Fig.4. A CMOS (complimentary Capacitors C1 to C6, se-
metal oxide semiconductor) digi- lected by rotary switch S1, en-
tal IC can be used as an excellent able the unit to cover from 10Hz
OUTPUT square wave generator. A typical to above 250kHz. Potentiome-
The output waveform is circuit is given in Fig.5, where the ter, VR1, acts as the frequency
rectangular with a rounded lead- inputs to three of the NOR gates control by varying the charging
ing edge. This rounding can be in a 4001B IC are wired together and discharging time of the ca-
eliminated by connecting to form inverting amplifiers. A re- pacitors. Range limiting resistor
1N4148 diodes between the sistor/capacitor timing network is R2 ensures consistent perfor-
transistor collectors and the connected in the feedback path mance over its full sweep.
coupling capacitors, C1 and C2 )5(48(1&< 5$1*(6  

(cathode (k) to collector (c)). . +]  +]

,& %
. +] +]
Additional one kilohm resistors . +]  N+]
723 9,(:

must be connected between the . N+]  N+]

 
. N+]  N+]
diode anodes and the positive . N+]  N+=

supply rail to maintain the circuit ,&D ,&E ,&F

action. 
%

 %

%

  

If the timing networks,   


R3/C2 and R5/C1, are identical, 5
/(9(/
&


5$1*( 95
 µ
the mark/space ratio of the out- 5
0
N
 3 N
9

put waveform will be equal. )5(4

&21752/  6


They do not, of course, have to 




–
be the same, and by tailoring 95
& & &
N
the component values, pulses &
Q
&
Q
&
S

of short duration separated by Q Q S

6,*1$/ 287
$3352;
comparatively long time inter- 9 33

vals can be generated.

Fig.5. Circuit diagram for a wide range, square wave genera-

tor using a 4001B quad 2-input NOR gate IC.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1138
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
CMOS SQUARE range, but performance be- resistor and a 100nF capacitor
comes erratic. in the timing network should
WAVE GENERATOR Most inverting CMOS gates make the circuit oscillate at
CMOS digital ICs can be should work well in this oscilla- around 1kHz.
configured as relaxation oscilla- tor, and the 4011B (quad two- Provided the supply voltage
tors in order to generate square input NAND gate) has the same is held between 8V and 12V,
waves of excellent quality. A pinout connections as the variations have a minimal effect
typical circuit is given in Fig.5, 4001B. on the frequency of oscillation.
where R2 and VR1, together Wider excursions cause signifi-
with a capacitor, C1 to C6, de- cant shifts.
termine the frequency of oscilla- SIMPLE PULSE
tion. GENERATOR SIMPLE SAWTOOTH
The usual formula relating In many cases the nature of
frequency to resistance and ca- the waveform is not important: GENERATOR
pacitance for this circuit is: all that is required is a signal to A device known as a uni-
f = 450,000 / RC test or trouble-shoot a piece of junction transistor can form the
equipment, or to generate an basis of a simple sawtooth gen-
when f is in Hertz, R is in audible tone. erator. Used almost exclusively
ohms, and C is in uF. A very simple and inexpen- in relaxation oscillator circuits, it
sive oscillator circuit, suitable comprises a tiny strip of n-type
The formula gives tolerably
for tasks of this kind, is shown in silicon material with non-
accurate results at low frequen-
Fig.6. Here a 555 timer, con- rectifying junctions (base 1 and
cies but, above 1kHz or so, the
nected as an astable multivibra- base 2) located at either end. A
frequency of oscillation is lower
tor, generates a pulsed wave- rectifying junction (emitter) is
than the figure given by calcula-
form. Various ranges are se- formed in a region of p-type ma-
tion. The ranges quoted in Fig.5
lected by switch S1 and poten- terial along its length.
are based on actual measure-
ments. tiometer VR1 sets the frequency The device acts as a volt-
of oscillation. age triggered switch. A typical
The circuit delivers a square
The timing capacitors, C1 to sawtooth generator circuit is
wave of excellent quality with
C4, are charged via R1, VR1 given in Fig.7, where resistor
an equal mark/space ratio.
and R2, but they discharge R1 and capacitor C1 determine
more rapidly through resistor the frequency of oscillation and
R1. The output at IC1 pin 3 is, R2 and R3 stabilize the transis-
OUTPUT therefore, a chain of pulses, and tor against temperature varia-
The loading effect of the adjustment of VR1 will alter tions.
output control VR2 reduces the both the frequency and the
available signal level, which is mark/space ratio of the out-
equal to the supply voltage put. Increasing the value of 5 5
N  Ω
when the oscillator is fed into a VR1 to one megohm will
high impedance. maximize the frequency
75
 
H
Frequency is affected by sweep with a single capaci- 1
9
E
changes in supply voltage, but tor. A sawtooth waveform is E E
H
to a much lesser extent than the available, at high
multivibrator circuit given in impedance, across the tim- 1 E
–

Fig.4. The mark/space ratio is ing capacitor.

6,*1$/ 287
almost exactly equal, and the If the simplest possible
&
Q
$3352;
Ω
5

9 33
square wave is of excellent spot-frequency signal gener-
quality. Output is constant over ator is required, VR1 and R2
the entire operating range. can be replaced by a single
Reducing the timing resistor fixed value resistor. A capac- Fig.7. Simple sawtooth generator.
R2 below 10k pushes the oper- itor can be permanently With the values specified for
ating frequency up to 2MHz and wired between IC1 pin 2 and R1 and C1 the circuit will
more on the highest frequency the negative supply rail, and oscillate at 1kH
VR2 can be deleted. A 100k
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1139
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
SIMPLE PULSE GEN-
ERATOR 5
N
5
N
5
Ω 75
1

The ubiquitous 555 timer 75

&
Q
J
G
 9

%&
H
IC, when connected as an E V

75
95

astable multivibrator, forms a N

1 E
–
F
very simple pulse generator. A H
5
0
5
N

typical circuit is given in Fig.6. E

&
µ
3
An approximate formula for  1&


the calculation of frequency,  6
5 & 5 95
 Ω  µ N
with this particular circuit, is: 5


N

N
& &
6,*1$/ 287
f = 2,800,000 / (R + 2000) x &
Q
&
Q
&
$3352;
9 33
µ S
C Q

when f is in Hertz, R is the

)5(48(1&< 5$1*(6
total value of VR1 and R2 in . +]  +] H

ohms, and C is in uF. . +] +]

E E
. +]  N+]
H E F G J V

The formula is reasonably . N+]  N+] %&

1
1
. N+]  N+] %&
accurate up to 5kHz or so, then . N+]  0+]
75$16,6725 /($'2876
the frequency of oscillation is
lower than the figure indicated Fig.8. Circuit for a linear sawtooth generator.
by calculation. Again, the A positive going pulse is The value of resistor R1 can
ranges quoted in Fig.6 are available at base 1, a negative range from 10 kilohm to one
based on measurement, not cal- going pulse at base 2, and a megohm (1M), and capacitor
culation. sawtooth (strictly speaking a C1 from 1uF or more down to
When a very simple and “shark’s fin”) waveform at the 100pF. Connecting a one
inexpensive means of trouble emitter. The impedance of any megohm potentiometer in the
shooting audio equipment is re- accepting circuit presented to R1 position will provide a wide
quired, this circuit is hard to the emitter must be high or the frequency coverage with a sin-
beat. The upper frequency limit unijunction action will be im- gle capacitor. The peak-to-peak
extends a little beyond 200kHz. paired. signal output at the emitter is

Emitter (e) impedance is

high when the device is off (not SAWTOOTH GENERATORS
conducting) and low when it is A unijunction transistor can form the basis of a very simple relax-
on. When the supply is first con- ation oscillator, and a typical circuit is given in Fig.7.
nected, capacitor C1 is dis-
charged, the emitter is at zero The following formula, which relates frequency to resistance and
potential and presents a high capacitance in the timing circuit (R1 and C1), produces tolerably ac-
impedance to the capacitor, en- curate results:
abling it to be charged via resis- f = 800,000 / RC
tor R1.
When a critical voltage when f is in Hertz, R is in ohms, and C is in uF. A sawtooth
(known as the “peak” point) has waveform with a peak-to-peak value equal to half the supply volts is
been developed across the ca- developed across the timing capacitor.
pacitor, the unijunction triggers The output of this simple, single transistor oscillator is non-linear
to the on state and the capacitor and at a high impedance, and an improved version is given in Fig.8.
discharges through the now low This more complicated circuit generates an extremely linear saw-
impedance emitter circuit. The tooth wave and has a low impedance output.
voltage falls to zero, the pro- Because of the way the timing capacitor is charged, it is not pos-
cess is repeated, and oscillation sible to quote a simple formula for the calculation of frequency. The
is maintained. measured ranges quoted in Fig.8 should, however, form a useful
guide to component values for spot-frequency versions of the circuit.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1140
Maxfield & Montrose Interactive Inc
 &RQVWUXFWLRQDO 3URMHFW
approximately equal to half the stant current generator stage, In Fig.8, current flow
supply voltage. transistor TR1. A FET source through transistor TR1 to capac-
follower buffer stage, TR3, pre- itors C1 to C5 (via switch S1) is
sents a high impedance to the controlled solely by the setting
LINEAR SAWTOOTH unijunction’s emitter and a suit- of VR1, and the charging rate of
GENERATOR ably low impedance to the ac- the timing capacitor is, there-
cepting circuit. By these means, fore, constant. This results in a
Whilst the sheer simplicity the limitations of the basic cir- linear voltage rise and a more
of the circuit arrangement cuit are overcome. perfect sawtooth waveform.
shown in Fig.7 makes it attrac-
tive for some applications, the When a capacitor is The buffer stage, TR3, is
high output impedance and non- charged via a resistor, the initial identical to the one adopted for
linear waveform limit its voltage rise is rapid, gradually the sine wave generator shown
usefulness. tailing off as it approaches a in Fig.2, and its operation has
fully charged state. already been described. Fre-
In the circuit diagram shown quency of oscillation is particu-
in Fig.8, the timing capacitor Because of this, the wave-
form developed across the ca- larly dependant upon supply
(C1 to C5) is charged via a con- voltage, and a well regulated
pacitor is not linear.
power supply is essential for the
correct operation of this circuit.
Stray capacitance acts as the
timing capacitor on the highest
frequency range.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1141
Maxfield & Montrose Interactive Inc
 3$57 7ZR  &DSDFLWRUV
E\ -2+1 %(&.(5
What we are doing during this 10-part Through these simple steps we hope to prove
Teach-In 2000 series is to lead you through the to you that using electronic components need
fascinating maze of what electronics is all not be a complex task and that, providing you
about! We are assuming that you know nothing think about each stage of what you are trying
about the subject, and are taking individual to create, you can actually design and build
components and concepts in simple steps and something that works!
showing you, with lots of examples, what you Last month we introduced color codes and
can achieve, and without it taxing your brain resistors. We now look at capacitors and show
too much! you some of the things they can achieve when
used with resistors.

In that fascinating bag of tive (–ve) wire, often shown with A CURVATURE
parts that you purchased last a “–“ (minus) symbol. The other
month, you will see a number of is called the positive (+ve) wire, IN TIME
blue (or black) tube-like compo- for which a “+” (plus) symbol Referring to Fig.2.2, plug
nents with two wires sticking out may be used. This is illustrated C1 into the breadboard with its
of one end (see Photo 2.1). They schematically in Fig.2.1a. This leads orientated as indicated by
are some rather remarkable com- type of construction is called the + symbol in the diagram
ponents called electrolytic capaci- radial. (the – symbol, although marked
tors. Find the one whose value A variant of electrolytic ca- on the capacitor, is often not
says 100uF. For this experiment pacitor case style is also manu- shown in layout diagrams). Now
consider the capacitor to be factured, as in Fig.2.1b, which is :
fit a 100k resistor (call it R1)
called C1. called axial construction and as shown. Insert terminal pins
One wire of C1 is usually whose +ve and –ve connections for the power supply connec-
longer than the other and has a are at either end as illustrated. tions (see Photo 2.2).
large arrow pointing at it from the Clip one power supply lead
case. This wire is called the nega- to the battery’s negative (–) ter-
minal and to the board as
 shown in Fig.2.2. Clip the other
power supply lead to the board
as well, but don’t connect the
other end to the battery’s posi-
tive (+) terminal yet.
It is conventional to use a
red lead for the positive power
supply connection. For the neg-
ative power supply connection it
 is the author’s preference to use
a green lead, although the use
of black is also common.
1 B1491< 2 1H91<
Note that in many circuits
Photo 2.1. A selection of Photo 2.1 A selection of (including those discussed in
electrolytic capacitors (radial electrolytic capacitors (radial this Teach-In), the battery’s neg-
construction). construction) ative connection is taken as the
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1142
Maxfield & Montrose Interactive Inc
 7($&+,1 
form as well sider your results.
the mascu- What timings have you got
line, he/she, for the 2V, 4V and 6V marks?
his/hers and Hopefully, about 4 seconds for
F9> so on. We the 2V, 10 seconds for 4V, and,
know that fe- oh – about 40 to 50 seconds for
males of the the 6V reading. And what do you
species are make of these timings? The volt-
F?<DC also inter- age steps have been at 2V inter-
F?ED B ested in elec- vals, yet the timings have be-
B54 ! tronics!)
F come progressively slower for
3?= Watch each step.
2<;
your meter Let’s now do things back-
3! and yell
 “NOW” as
wards. Watch the meter and tell
your friend to note the seconds
soon as it hand again as he unclips the red
shows a reading of 2V. At which lead from battery positive and
point your friend should tell you clips it (do it quickly) to the bat-
the number of seconds that tery negative. You now yell out
 have passed since clipping the
21DD5BI when the meter shows readings
&F red lead to the battery. of 4V, 2V and 0V.
Continue watching and tim- This time you should have
Fig.2.2 and Photo 2.2, ing and yell again when the me- found that 4V is reached after
breadboard layout for the ter reaches 4V and again at 6V about 4 seconds, 2V at 10 sec-
first resistor/capacitor timing (assuming you haven’t run down onds and (almost) 0V at about 40
your battery since last month’s to 50 seconds. You will see, too,
experiment. Teach-In – if you have, get a that timings have again become
new one!). Leave the battery progressively slower for each 2V
common reference point against fully connected while you con- step.
which voltage readings are
taken. As such, it is regarded as
being at zero volts (0V). PANEL 2.1 – AC AND DC VOLTAGES
Consequently, throughout Alternating (AC) voltages are those that repeatedly change their
this series (unless you are told magnitude above and below a midway reference voltage level (often
specifically otherwise) the me- taken as 0V, as in mains electricity supplies, but may be other volt-
ter’s COM lead should always ages). Direct (DC) voltages are those that remain at any fixed volt-
be connected to the battery age level, either above 0V or below it.
negative connection when tak- Strictly speaking, AC and DC actually mean alternating current
ing voltage readings. and direct current and, as such, to use the terms AC voltage and DC
Clip your multimeter’s leads voltage is incorrect. However, for some unknown reason, the terms
as shown and set the meter to AV and DV (which would be more appropriate when referring to alter-
the first volts DC range above nating and direct voltages) do not seem to exist.
6V. Note the reading, 0V at this It can be argued, however, that it is not the voltage that flows, but
moment. While you watch the the current. Indeed the term voltage merely represents a concept
meter, get a friend to clip the rather than something that actually flows. As an ancient (1962) copy
red wire to the battery’s positive of the Penguin Dictionary of Electronics puts it: “Voltage. Strictly, a
(+) terminal, noting the position difference of electric potential expressed in volts. However, the term
of the seconds hand on his is used more generally as a synonym for electrical potential”.
watch as he does so.
Voltage is certainly a more convenient term and ties in with the
(Girls, ladies, please forgive fact that the unit of measurement for potential.
the use of “he” and “his” forms
of reference – it would be te-
dious to this author’s typing fin-
ger to keep giving the feminine

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1143
Maxfield & Montrose Interactive Inc
 7($&+,1 
31@139D?B
3?>>53D9?>C combination has the most pro- charged state. The discharging
found effect upon the whole rate is governed by the same
realm of electronics technology. factors as controlled the charg-
But, before we examine the ing rate.
results of your timed yelling, Given enough time, all of
let’s first have a look at the ba- the electrical charge stored
sic nature of a capacitor. across the plates will reduce to
zero. But, you may ask, what
happens to the electrical charge
WHAT IS A CAPACI- itself? Principally, it is converted
TOR? into heat in the discharging con-
9>CE<1D9>7 ductor and capacitor’s internal
495<53DB93 A capacitor is a component
resistance, although in extreme
which has the capacity to store
circumstances some could be
electrical energy (hence its
3?>4E3D9F5 converted into light or radio en-
@<1D5C name). In this sense, capacitors
ergy. In normal use, you won’t
can be thought of as a type of
Fig.2.3. Basic construction of battery, but, unlike a battery,
notice any temperature change
a simple capacitor. in the capacitor or the conduc-
they do not depend on a chemi-
tor.
cal reaction for this function to
Don’t worry if the actual tim- occur. Rather, they take advan- A capacitor’s ability to be
ings you get are somewhat dif- tage of a convenient fact of na- charged by a voltage and to
ferent to ours, for a start you are ture that prevails when two hold the charge (almost) indefi-
not using an accurate timer, metal plates are placed close to nitely allows it to be used in
secondly, the actual values of a each other, but not touching, electrical and electronic circuits
capacitor compared to its stated and a voltage source is con- in a variety of ways:
value can differ by even as nected across them (see
much as 50% (more about this o) To simply store a voltage
Fig.2.3).
in due course). until it is needed
At the moment that the volt- o) To smooth out fluctuations
Additionally, your battery is age is applied, an electrical in voltage levels
probably not supplying exactly charge is transferred to the
6V. Furthermore, the resistance o) In conjunction with other
plates at a rate depending on
of your meter (which you estab- components, such as resis-
the voltage level applied, the
lished last month) is forming a tors for example, to deter-
material from which the plates
potential divider with resistor mine the rate at which volt-
are made, their total area, their
R1, so fractionally “pulling age changes occur at a
distance apart, the substance
down” the voltage at the junc- particular point in a circuit
which lies between them (called
tion of C1 and R1. the dielectric), and the amount o) To shorten or extend pulse
of resistance existing in the con- lengths
What you have just demon-
strated is that when a voltage is nection path (including the ca- o) To transfer changing differ-
applied to a capacitor via a re- pacitor’s internal resistance). If ences in voltage levels be-
sistor, the capacitor starts to the voltage is applied for suffi- tween one side of the ca-
charge up with the voltage, cient length of time, eventually pacitor and the other, in
commencing at quite a fast rate, there will be virtually the same other words, to allow alter-
but then more slowly as its voltage across the plates as nating (AC) voltages to be
charge increases, until near the available from the source. transferred whilst prevent-
end of when the voltage on the ing direct (DC) voltages
When the voltage source is
capacitor is close to that of the from flowing from one part
removed, the plates will retain
battery, the charging rate be- of a circuit to another (see
their charge until a conductor of
comes almost imperceptible. also Panel 2.1)
some sort is connected across
You have also shown that the them. As soon as there is a con-
opposite is true as well (i.e. dur- ducting path between the two CAPACITANCE
ing discharge). plates, the charge begins to flow
The charge/discharge prop- from one to the other, trying to VALUE
erty of a capacitance-resistance return to the previously un- The amount of electrical

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1144
Maxfield & Montrose Interactive Inc
 7($&+,1 
 
As with the resistor display
examples last month, you can
change the values associated
69H54 69H54 F1B912<5 DB9==5B with this demo. At the top right
>?>@?<1B9C54 @?<1B9C54 >?>@?<1B9C54 >?>@?<1B9C54
5<53DB?<ID93
you should see C highlighted
and its value as 1uF. Press the
Fig.2.4. Commonly used symbols for capacitors. <*> key (multiply) twice. On
charge that a capacitor can hold your breadboard is shown in each press the graph will redraw
is known as its capacitance Fig.2.5. (We didn’t comment on to suit the changed value. You
value (surprise, surprise!), and the battery symbol in Part 1, should see C = 100uF after the
the unit that is used to define it make a mental note of it now!) second <*> press.
is the “Farad”. This is named We’ll say more about ca- Press the down arrow key
after an electrical pioneer in the pacitors later, so back to your once to select factor R. Press </
nineteenth century called timing results… > (divide) once to set R = 100
Michael Faraday. He was a Lon- kilohms. Press the down arrow
doner, born 22-9-1791, died 25- once to select V, then press <–>
8-1867. DISPLAY GRAPH (minus) four times to set V = 6
More intimate information We can illustrate a capaci- volts. Again press the down ar-
about Farad values is given in tor’s rate of charge and dis- row to select T, and press <*>
Panel 2.2. charge using another of our once to set T = 10 seconds.
There are several symbols software demos. From the main
that can be used to represent a menu select Resistor-Capacitor TIME CONSTANT
capacitor, as shown in Fig.2.4. Charging Graph.
In front of you now is the
Some represent the type of ca- On entry to this display you graph that illustrates how an
pacitor, but there are also differ- will see a rising graph on a grid, ideal (and “empty” – uncharged)
ences of international standard similar to that in Photo 2.3. This 100uF capacitor charges via a
used in some cases. Those
used in EPE Online are the
particular graph is that for a 1uF
:
capacitor and 1M resistor, a
:
100k resistor when a voltage
of 6V is instantaneously applied
ones to the left of each pair. combination whose timings to it. The vertical axis of the
The circuit diagram for what would be far to fast for you and graph represents volts, and the
you have just been doing on friend to keep pace with! horizontal axis shows elapsed
time in seconds. The time be-
tween each horizontal step is
the value shown times 10, be-
cause T has been set to 10 sec-
onds per division.
Look closely at the graph.
Where it reaches the 2V grid
line, you can just about estimate
that the time taken so far is
about 4 seconds. It’s clear to
see that 4V is reached at about
10 seconds, and that it has just
about reached 6V at around 50
seconds. You can select a
“magnified” view of the 2V mark
by pressing <–> to make T = 1
second.
Why the blue horizontal line
just below 4V?, you must be
wondering. By convention, the
Photo 2.3. Capacitor charging graph displayed on the interac- line represents the 63% level of
tive computer screen the power supply voltage across

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1145
Maxfield & Montrose Interactive Inc
 7($&+,1 
F9>

experiment. graphs earlier.)

B! The blue line has changed Just for a bit of idle illustra-
21DD5BI
its position though. The reason tion, the circuit diagram for the
is simple, again by convention, R-C series is shown at bottom
F?ED
it now represents the 63% level right of the screen. Note how

below the starting voltage or the arrow changes direction and
3!
37% above the termination volt- value depending on the charge/
F age, in this case 6V and 0V re- discharge mode. The capacitor
Fig.2.5. Circuit diagram for spectively. symbol shown is that for a non-
The rate of change is said polarized type (see later, plus
the capacitor charge/dis- Fig.2.4), but in reality the sym-
charge experiment. to be exponential, and in its cal-
culation you ideally need a sci- bol should more reasonably be
entific calculator (and the knowl- that for an electrolytic when
the resistor/capacitor series. high values of capacitance are
The rate at which the capacitor edge of how to use it), because
the formula is a bit complex: used.
charges to that 63% voltage is
termed its time-constant, which We suggest you experiment
Vc = Vs x (1 – EXP(–t / CR)) with different range values on
is the value obtained when the
capacitance (C) and resistance for the charging rate, and: the screen, and if you think you
(R) values are multiplied. It is can actually time some of the
generally referred to as the CR Vc = Vs x (EXP(–t / CR)) graphs using your breadboard
(or RC) value. assembly, set R1 and C1 to the
for the discharging rate, same values as displayed. (Be
It is important to note that where: sure to read Panel 2.3, how-
the units for C and R must be ever!)
expressed with the correct or- Vc = voltage across the ca-
ders of magnitude. In the exam- pacitor In the accompanying Exper-
ple shown, C (100uF) is ex- Vs = voltage across the ca- imental article, we shall tell you
pressed in microfarads (100) pacitor/resistor series about how to combine two or
:
and R (100k ) is expressed in t = elapsed time
more capacitors to achieve dif-
ferent values. In this Tutorial
megohms (0 1), resulting in a
¬ CR = time constant section, though, it’s time to dis-
CR value of 10 seconds (100 x EXP = exponent cuss some more facts about ca-
0 1), as shown to the right of the
¬
pacitors – first have a read of
display. You will see these formulae
Panels 2.4 and 2.5, and then
shown, as appropriate, at the
With a 6V supply, the 63% read on from here.
top of the graph display.
voltage is 3 78V (6 x 0 63),
¬ ¬

shown alongside the CR value. We are not going to ask you

to memorize the formulae or CAPACITOR
test your knowledge of how to SELECTION
DISCHARGE GRAPH use them. Since you now have
a computer program that does it Some concepts referred to
So, we have illustrated the in this section are likely to be
for you, let it do the brain-
charging up of your R-C combi- alien to you. Where they are not
teasers! The answers for any
nation. The discharge illustra- further discussed here, they will
values not included in the C-R-
tion is similar, but in reverse. be covered in later parts of
V-T ranges provided can be es-
Return the time scale value to T Teach-In. We have to mention
timated from the nearest se-
= 10 secs, then Press <C>. them now as they are relevant
lected values.
The curve now starts high, to this section – you should re-
(What you can do more read it once you have read the
at 6V and smoothly descends to
simply, however, is calculate future parts. Should we not
0V. It crosses 4V at about 4
the RC time constant, by multi- cover something that you are
seconds, 2V at close to 10 sec-
plying the values of C and R. puzzled by, you can always ask
onds, and reaches 0V round
Do note that the values must be us to clarify it through Circuit
about 50 seconds – just as we
expressed in units of the correct Surgery or Readout pages.
predicted earlier, and you were
magnitude to achieve a valid
probably close to it with your There are several factors to
answer, as we said a few para-

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1146
Maxfield & Montrose Interactive Inc
 7($&+,1 
be considered when selecting a and at high temperatures (well
PANEL 2.2 – CAPACI- capacitor for a particular appli- above “normal” room tempera-
TANCE UNITS cation, which include: tures) all capacitors should be
A capacitance value of one significantly derated (assumed
o) Capacitance value to have a lower working voltage
Farad is a unit of charge, which, in
o) Working voltage than that stated). In normal ev-
practical terms, is far too large to
be useful in everyday electrical and o) Tolerance eryday applications, however,
electronic circuits. For conve- o) Leakage current this factor is usually irrelevant.
nience, the unit is usually divided o) Temperature coefficient Capacitors should always
and expressed in sub-units, such o) Stability be operated at well below their
as: nominal maximum working volt-
Unless you are involved ages. If a circuit is designed for
o) Microfarads, being one mil-
with a particularly demanding operation at 9V, for example, a
lionth of one Farad, and usu-
P
ally written as F (Greek “mu”
design, it is principally the first
two that will concern you, but
capacitor rated at a working
voltage of 9V or 10V should not
followed by a capital F), al- you should be aware of the fol- be used, rather, one rated at
though it is common for it to lowing: 16V or greater should be cho-
be written as “uF” or “mF”,
When substituting capaci- sen. Even one rated at 63V, for
since many keyboards do not
tors, either because they have instance, would be acceptable,
have the Greek symbol read-
failed in an existing circuit, or provided that its size (which is
ily available. (The use of “mF”
because the precise type speci- likely to be greater with in-
is to be deplored because it
fied in the components list of a creased voltage ratings) is suit-
really means millifarad rather
constructional project is not able for the circuit board on
than microfarad.) It is also
readily available from your nor- which it may need to be
common, where the meaning
mal supplier, it is important to mounted.
of the term is implied, for it to
P
be written simply as “ ”, in ensure that the replacement
performs to a specification
As a rule of thumb, the
quoted working voltage rating
component lists for instance.
which is at least as good as that should be at least 50% greater
Verbally, these abbreviations
of the specified component. than the voltage at which the
are often pronounced as
“mew” or “muff”. For example, However, it is quite permis- component is required to work
P
a 10 F capacitor might be re- sible to replace a capacitor that
has a working voltage rating of
in the circuit, although there are
occasions, such as in power
ferred as having a value of
15V by one rated at 25V, for in- supply circuits, where a much
“ten-mew” or “ten-muff”.
stance. The working voltage rat- greater margin should be al-
o) Nanofarads, being 1000- ing simply states the maximum lowed, possibly even as much
millionth of a Farad and usu- voltage at which a component as four times the nominal sup-
ally written as “nF”, although should be operated in normal ply voltage.
the “F” may be dropped where service. Generally speaking, a Where an AC voltage rating
it is implied in the context. higher working voltage rating is is specified, this is normally for
Verbally, the abbreviation nearly always acceptable elec- sinusoidal operation (sine
might be pronounced “en-eff” tronically (physical size permit- waves) at either 50Hz or 60Hz
or just “en”, i.e. a value of ting, of course). Similarly, a ca- (Hz, or Hertz, is a unit of fre-
10nF might be pronounced as pacitor with a tolerance of 20% quency in cycles per second).
“ten-en”. The use of the term can always be replaced by a Performance will not usually be
“nuff” is unlikely. similar one having a tolerance significantly affected at low fre-
o) Picofarads, being one million- of 10%. A better tolerance rat- quencies (up to 100kHz, or so),
millionth of a Farad and usu- ing is always acceptable electri- but above this, or when non-
ally written as “pF”, though cally. sinusoidal (e.g. pulsed) wave-
again the “F” might be forms are involved, the capaci-
dropped when it is implied. tor must be derated in order to
Pronunciation is usually “puff”, WORKING TO RULE minimize losses in its dielectric
although it might sometimes It is also important to note material which can produce in-
be heard as “pee”, i.e. “ten- that working voltages are re- ternal heating and lack of stabil-
pee” for 10pF. lated to operating temperatures ity.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1147
Maxfield & Montrose Interactive Inc
 7($&+,1 
Table 2.1. Capacitor varieties and their typical characteristics.
Metal Poly- Poly- Poly-
Capacitor Ceramic Electrolytic Mica Polyester Tantalum
Film carbonate styrene propylene
Capacitance 2.2p to 100n to 1u to 2.2p to 1n to 10n to 10p to 100n to 100p to
Range (F) 100n 47000u 16u 10n 10u 10u 10n 100u 470n
Typical +/-2 to -10 to +/-5 to +/-1, +/-2.5, +/-5 to
+/-20 +/-1 +/-20 +/-20
tolerance (%) +/-80 +50 +/-20 and +/-5 +/-20
Typical voltage 50V to 6.6V to 250V to 350V 63V to 63V to 50V to 6.3V to 100V to
rating (DC) 15kV 450V 600V (typical) 400V 630V 630V 35V 1.5kV
Temp. coefficient +100 to +1000 +100 to +35 to -150 to +100 to -200
-200 +60
(ppm/ o C) -4700 (typical) +200 +70 +80 +1000 (typical)

Stability Fair Poor Fair Excellent Fair Good Good Fair Fair/Good

Ambient temp. -35 to -40 to -25 to -40 to -40 to -55 to -40 to -40 to -55 to
range ( o C) +85 +85 +85 +80 +100 +100 +70 +85 +100

You should also be aware PANEL 2.3 – BEWARE THE FORCE!

that a sinusoidal waveform nor-
mally has its voltage quoted as Do be warned that you should NEVER insert or remove compo-
an RMS (root of the mean nents from a circuit board when the power is switched on.
square) value, whereas in fact Whilst this (arguably) is not so necessary to observe with passive
its peak value is nearly 50% components such as resistors and capacitors in a low voltage circuit,
higher (x 1 41), thus the chosen
¬ active components (to be met later) such as integrated circuits (ICs)
capacitor’s voltage rating must and transistors can die in such circumstances.
take this into account. It also important to note that capacitors can hold their charge for
a while even after the power is switched off. Ideally, you should allow
RIPPLE FACTOR a few seconds for them to discharge before handling them. With a
low voltage supply, such as 6V, this is perhaps not critical. However,
Capacitors used for smooth- with higher voltages, of greater than 30V for example, it is ESSEN-
ing and reservoir (substantial TIAL that you should allow for the discharge time. To really ensure
storage) applications in DC
power supplies must have an
that a capacitor is fully discharged, CAREFULLY touch a 10k resis- :
tor across its +ve and –ve connections for a few seconds – taking
adequate ripple current rating. great care that YOU do not touch the wires.
This rating refers to the AC
characteristic of the current (at We also have to caution you (not as the “Old Bill” but as friendly
the ripple frequency, e.g. 50Hz voices across the page!) – DO NOT use a metal tool (e.g. screw-
for UK mains operated power driver) to short out capacitor terminals for instantaneous discharge. It
supplies) which remains after can be damaging to both the capacitor and the screwdriver.
the principal alternating (AC)
itors required for power supply round. The positive side of the
voltage has been rectified to a
use. The chosen ripple current capacitor must always be con-
DC voltage. Without a capacitor
rating should always be greater nected to the side of the circuit
following the rectifier, the ripple
than the ripple current ex- that has, or is likely to have, the
voltage will be approximately
pected. highest voltage.
half that of the original AC
peak-to-peak voltage. It is the Across power supply lines,
job of the following capacitor to WHICH WAY this orientation of polarity will
smooth out that ripple, a task always be obvious – the posi-
that is complicated when large ROUND? tive side of the capacitor goes
currents are demanded by the A most important considera- to the positive supply line. It is
ensuing circuit. Component data tion when using polarized ca- not always so instinctively obvi-
sheets and catalogs will usually pacitors (e.g. electrolytic and ous when the capacitor is being
quote the typical ripple current Tantalum), is that they should used to couple AC signals be-
rating for the large value capac- be connected the correct way tween different parts of a circuit.
If in doubt, think about what DC

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1148
Maxfield & Montrose Interactive Inc
 7($&+,1 



PANEL 2.4 – CAPACITOR TYPES
Capacitors are manufactured as having two very basic character-
1H91< D1>D1<E=
istics, they are either: 5<53DB?<ID93 2514


o) Polarized, or
o) Non-polarized
the latter being manufactured as fixed and variable capacitance B1491<
5<53DB?<ID93 =931
types.
In circuit diagrams and constructional charts, a fixed capacitor’s
numerical identity is usually prefixed by “C”, e.g. C21. A variable ca-
pacitor may have its number also prefixed by “C”, although it is more
likely to be prefixed by “VC” (Variable Capacitor), or perhaps “CV”
(Capacitor Variable). 35B1=93 F1B912<5

Polarized capacitors, as their name implies, are very particular Fig.2.6. Examples of capaci-
about which side of them is connected to a (relatively) positive volt- tor body styles. Other styles
age. Connecting them the wrong way round can have dire results, a exist.
matter which is discussed in the main text. It is polarized capacitors
that you have been using so far, sub-type “electrolytic” – this is why levels are likely to exist if the AC
we stressed earlier that you should only connect their +ve and –ve signal ceases, and face the ca-
leads as shown. pacitor accordingly.
Non-polarized capacitors can normally be connected into a circuit There are instances, though,
either way round, although there are some circumstances where the when the polarity of the voltage
relative position of the output electrode foil is placed in relation to across an electrolytic might keep
other parts of a circuit. The colored ends of some polystyrene capaci- reversing (as in some types of
tors, for example, can indicate this type of polarity, although it is not oscillator, for example), adversely
a true polarity as referred to with regard to polarized electrolytic or affecting both the capacitor and
tantalum capacitors. (The author has never had occasion be con- the correct operation of the cir-
cerned about this detail, over several decades of doing electronics.) cuit. In this case, two equal value
electrolytic capacitors can be
Capacitors are also manufactured in a seemingly-bewildering ar- used in series, both negative
ray of sub-types, basically named in respect of the nature of the di- ends connected together, both
electric material used between the plates: positive ends facing outwards.
o) Electrolytic (polarized) The value for each capacitor
o) Tantalum (polarized) should be twice the total capaci-
tance required.
o) Polypropylene (non-polarized)
o) Polycarbonate (non-polarized) If a polarized capacitor is
connected the wrong way round,
o) Polyester (non-polarized) in extreme circumstances it can
o) Polystyrene (non-polarized) over-heat, causing damage to it-
o) Metallized film (non-polarized) self and other components, and in
o) Ceramic (non-polarized) a really severe case the capacitor
o) Mica (non-polarized) – sometimes called silver-mica may even explode. At the very
o) Trimmers – variable capacitors (non-polarized) least, the circuit may not operate
as intended.
o) Air-spaced – variable capacitors (non-polarized)
o) Paper – now rare (non-polarized) POLARITY MARKINGS
o) Oil-filled – now rare (non-polarized) Polarity is usually clearly
marked, but there are several
There are also sub-types of the sub-types! Have a look at a ma- ways in which it might be done.
jor component supplier’s catalog and prepare to be astonished… For- The ends from which the connect-
tunately, until you are much more into the depths of serious electron- ing wires come out may be
ics design, the subtle differences between some types need be of marked with “+” or “–“ signs, or
little concern. there might be a large arrow
pointing to the negative end or to
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1149
Maxfield & Montrose Interactive Inc
 7($&+,1 
PANEL 2.5 – IDENTITY CODING obvious, the negative lead will be
seen to be attached to the outer
The majority of capacitors now have their values printed on metal casing of the body. (The
them, although color-coded varieties are still to be found. Examples “opposite” term to axial construc-
of the color codes which might be encountered are shown in Table tion is radial, in which both capac-
2.2 plus Fig.2.7. As with resistors, the colors allocated to each nu- itor wires come out from the
meral from 0 to 9 conform to the standard color code system. same end – shown earlier in
Where capacitors have their values printed on them, the infor- Fig.2.1.)
mation may well be abbreviated or allocated a letter coding. Ceramic Non-polarized capacitors can
capacitors, for example, may have their tolerance and voltage rat- generally be connected either
ings coded as in Table 2.3. way round, although there are
A 3-digit coding is commonly used to mark some ceramic capac- specialized situations where the
itors. The first two digits correspond to the first two digits of the orientation in relation to the ca-
value, whilst the third digit is a multiplier which gives the number of pacitor’s outer foil may be signifi-
zeroes to be added to give the value in pF, e.g. 103 = 10000pF = cant (as we comment about
0 01uF.
¬
polystyrene capacitors in Panel
2.4).
Which brings us to the sometimes misunderstood use of pF, nF,
and uF. An nF value is 1000 times greater than pF, and 1000 times Be aware that with very small
less than uF. Therefore, the following typical conversions apply to polystyrene capacitors, an occa-
values seen on some capacitors: sional fault can be experienced in
that the leads can become de-
1nF (or 1n) = 1000pF tached internally. It is very un-
10nF (or 10n) = 10000pF = 0 01uF ¬
usual, but it can cause the capac-
itor to develop an open circuit, or
100nF (or 100n) = 100000pF = 0 1uF ¬
a short circuit.
However, despite all this possible coding, with many modern ca-
pacitors, their values are normally obvious from the uncoded infor- LIGHTLY CHARGED
mation printed on them (although you may need a magnifying glass
in order to read them). We suggest you now move
on to the Experimental portion of
this article and just generally play
Table 2.2. Tantalum capacitor color coding. around as suggested there. You
can even “lighten” up the experi-
Reading from the top, Bands 1 and 2 = Capacity, Spot = Multiplier, Band 3 = voltage ence as well.
Color Figure Multiplier Voltage Next month we look at com-
Black 0 1 uF 10V ponents whose values are not
Brown 1 10 uF -- rigidly fixed – variable resistors
Red 2 100 uF --
(variable capacitors will be dis-
cussed in a later part), and sensor
Orange 3 -- -- --
resistors. Herr Georg Ohm and
Yellow 4 -- -- 6.3V his famous Law also come under
Green 5 -- -- 16V scrutiny.
Blue 6 -- -- 20V
Violet 7 -- -- --
In this Tutorial portion of Part
Gray 8 0.01 uF 25V 2, while using different R1 and C1
White 9 0.1 uF 30V values on your breadboard to
Pink -- -- -- 35V mimic the screen display, you
might have come up against a bit
a particular wire (as we dis- crimp around the casing and the of a snag! The screen has speci-
cussed at the beginning of this circular face at that end is likely fied a C-R combination for which
Tutorial). With electrolytic ca- to be a plastic material, often you don’t have the component
pacitors having a wire at each black. values. Well, actually, you know
end (axial construction), the Also, where the lead con- you can make up the resistor
positive end is likely to have a nections to the capacitor are value using serial or parallel com-

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1150
Maxfield & Montrose Interactive Inc
 7($&+,1 

!CD 21>4
Table 2.3. Ceramic capacitor letter coding. C@?D
C@?D !CD 21>4
Reading from the top, Bands 1 and 2 = Capacity, Spot = Multiplier, Band 3 = voltage ">4 21>4
">4 21>4
Color Figure Multiplier Voltage
#B4 21>4
#B4 21>4
Black 0 1 uF 10V  
Brown 1 10 uF --
D1>D1<E= DI@5C
Red 2 100 uF --
Orange 3 -- -- --
Fig.2.7. Tantalum capacitor
Yellow 4 -- -- 6.3V
color coding.
Green 5 -- -- 16V
Blue 6 -- -- 20V
Violet 7 -- -- --
Gray 8 0.01 uF 25V
White 9 0.1 uF 30V
Pink -- -- -- 35V

TEACH-IN 2000 – EXPERIMENTAL 2

MEASURING AND CALCULATING CAPACITANCE
binations, as discussed in Part (C1) in your Tutorial Part 2
3!
1. It’s the capacitor values that charge/discharge experiment
are the problem. (Fig.2.2 and Fig.2.5). Resistor
Fret not! Capacitors too can 3" 3! 3" 3# R1 is given a value of 100k . :
be combined in series or paral- Do the charge/discharge
lel to achieve other capacitance 3# experiment, noting the time at
values. The rules are as simple 1 2
the 63% and 37% voltage lev-
as those for resistors, except els, i.e. 4V and 2V respectively.
that they are the opposite way Fig.2.8. Capacitor in series (You will find it easier to do this
round. (A), and parallel (B). experiment if you make up and
pacitors in Series and Parallel, use another short lead with two
crocodile clips on it.)
CAPACITOR accessible from the main menu,
allows you to set the values for Did you achieve timings of
COMBINATIONS two and three capacitors and about 15 seconds at the voltage
When capacitors are in se- have the computer calculate the points? That’s the time constant
ries, as are the three shown in resulting total series and parallel associated with R1 = 100k :
Fig.2.8a, the total capacitance values (see Photo 2.4). There is and C1 = 150uF, the latter be-
value (CT) is calculated as: also a Self-test option allowing ing very close to the answer of
you to check your understand- 149 2uF for C1, C2 and C3 in
¬
CT = 1 / ((1 / C1) + (1 / C2) + (1 ing of the two formulae in- parallel.
/ C3) + (etc)) volved.
which is, of course, identical
to the resistors in parallel for-
SERIES TEST
PARALLEL TEST
mula, except for the letter Return to your screen graph
change. Set up your breadboard as and set C and R to 150uF and
For capacitors in parallel (as
shown in Fig.2.9 (and Photo
2.5), in which three capacitors
:
100k , where you can see the
for the three in Fig.2.8b) the for- 15 seconds timing when the
are shown in parallel (as in graph slope crosses the per-
mula is simply: Fig.2.8b), where C1 = 100uF, centage line.
CT = C1 + C2 + C3 + (etc). C2 = 47uF and C3 = 2 2uF. This ¬

combination is being used in Now calculate the total ca-

The computer program Ca- place of the single capacitor pacitance if the same three ca-

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1151
Maxfield & Montrose Interactive Inc
 7($&+,1 
achieve that should be (ideally) 100 times
value when C the above 0 2 seconds, i.e. 20
¬

= 2 2uF. We
¬ seconds.
trust you’ll find
it to be 91k .: This now brings us to an
interesting point: how do you set
an exact time constant without
using multiple values of capaci-
SLOWER tors and resistors? The answer’s
TEST simple, and there are two ready-
On your made components that help in
breadboard, this, the variable capacitor, and
now use just the variable resistor (more com-
one capacitor, monly known as the poten-
with a value of tiometer). The latter we shall
Photo 2.4. Interactive computer screen for cal- 220uF (100 investigate next month.
culating serial and parallel capacitor times the
combinations. value of the LIGHTING UP TIME
above 2 2uF), ¬

Photo 2.5. and with the breadboard as- We are again going ask you
Breadboard sembly of Fig.2.2 modified to to use a light emitting diode
(LED), as we did in Part 1. We
layout for ex- suit. Testing your knowledge are also asking you to use an
amining ca- of resistor combinations, re- inverting logic gate (also known
place R1 with a made up
pacitors in
parallel.
value of 91k . Two resistors : as a NOT gate). You’ll be told
more about both devices on an-
will do it (within 100 ohms) –
what are they and how are other occasion, but you don’t
they connected? (Refer back to need to fully understand them if
Part 1 if in doubt.) you use them as we now tell
you.
Now do your time check
routine – the time constant The LED as you discovered
in Part 1, is a neat little device
pacitors are connected in series, that glows when a voltage is
as shown in Fig.2.8a. If you connected across it in a specific
don’t get an answer of approxi- direction via a suitable resistor.
mately 2 06uF, try again.
¬
It is important that the resis-
F9>
You won’t be able to do the tor should be used, since the
breadboard check with this LED cannot survive if more
value, the time-constant is too than about 2V is connected
fast in this instance, but you can F?<DC
across it. You are about to use it
use the display to show the F?ED B
with a 6V supply, and the resis-
graph for the nearest available B54 ! tor has to drop the voltage to an
F
value of 2 2uF, i.e. 0 22 seconds 3?= acceptable level. In this in-
:
¬ ¬

3" stance we want you to use a

with R1 at 100k . (Later on,
you could set up your own ex-
2<;
3!


3#
:
470 resistor, as we did previ-
periment using three capacitors  ously.
in series for which a time con- What we want to do is use
stant significantly longer than the LED (call it D1) to indicate
0 22 seconds is expected.)
¬

when a certain voltage has
The time constant for 21DD5BI been reached on a charging or
&F discharging capacitor. The prob-
2 06uF (call it 2uF) and R =
:
¬
lem is, though, that the time
100k is actually 0 2 secs. Use
¬

your graph display to find out

Fig.2.9. Breadboard layout :
constant when a 470 resistor
for capacitors in parallel ex- is used is too short for the ca-
what value of R is needed to periment. pacitance values you can realis-
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1152
Maxfield & Montrose Interactive Inc
 7($&+,1 
tically select. power supply is switched off shown in Fig.2.10, where the
We need, therefore, to use before inserting or removing symbols within the outline are
a technique that allows a rea- them. those for inverter gates.
sonably long time constant to be One further cautionary note: An inverter gate, as you will
set, and still to provide enough You will be aware that you can be told when we discuss digital
power to drive the LED via a sometimes generate sparks electronics in a later part, has an
:
470 resistor (call it R2). when combing your hair or tak- F5

This is where the logic gate ing off a sweater. This is caused !$ !# !" !! ! ) (

(call it IC1a) is used – as a type by the discharge of static elec-

of amplifier. Amongst your bag tricity, which can build up on
of parts you’ll find some black some substances, including
“caterpillars” with 14 legs, your body and that of animals,
seven-a-side. Find one that’s frequently by the action of fric- ! " # $ % & '

marked 74HC04. There are tion in a dry atmosphere. Such 7>4

likely to be lots of other forms of discharges, if they occur when
marking as well, but somewhere you touch some semiconductors
you should be able to discern can kill the devices – the level
the 74HC04 identity. of voltage discharge being
greater than the device is de-
The 74HC04 and the LED signed to handle. '
!$
are examples of components
that belong to the general class To avoid this happening, it
known as active devices (as op- is advisable to touch an earthed !

posed to the general class bare metal object immediately Fig.2.10. Pinouts and typical
called passives, of which resis- prior to handling integrated cir- case style for 74HC04 hex
tors and capacitors are exam- cuits. A water pipe is a suitable inverter gate. Note the inver-
ples). Like the LED, the 74HC04 object, as is the exposed bare
sion gate symbols within the
is another member of that enor- metal work of an item of
earthed mains powered equip- pinout drawing.
mous family of components re-
ferred to as semiconductors. It ment. When ICs have been
supplied in a black plastic foam, output that is at a level called
also belongs to a sub-group of Logic High when its input is at a
that family, generally known as or bag marked as being “static
sensitive”, leave the devices level called Logic Low, and vice
integrated circuits (often abbre- versa.
viated to ICs). More particularly, where they are until needed.
it is a digital logic IC. Then keep the handling of their So what’s Logic High and
legs to a minimum. Logic Low? Well, in this in-
The author reassures you, stance, High refers to +6V (the
SEMICONDUCTOR however, that for all the years power supply voltage level) and
HANDLING he has been handling ICs, he Low is simply 0V. The two
cannot remember killing one terms are respectively also
As with electrolytic capaci- with static electricity. They are known as Logic 1 and Logic 0.
tors, by far the vast majority of very robust, especially those
semiconductors can only be The logic gate, though,
manufactured over the last does not have to have exactly
connected to a power supply in decade or so.
one direction. Many can die if 0V or +6V on its input for the
connected the wrong way round. We shall discuss static elec- output to respond. There is a
Even if they don’t die, they will tricity further in a future part of range of voltage levels below
not work correctly. This is Teach-In. which the gate thinks it’s being
equally true for a 74HC04. provided with Logic 0, and
there’s range of voltage levels
Always connect semicon- INVERTER GATE above which the gate thinks it’s
ductors and other active de- being provided with Logic 1. In
The 74HC04 device is
vices into a circuit in the a region somewhere between
known as a hex (six) inverter
manner specified in circuit those two levels, the gate tends
gate – in other words it has six
diagrams, constructional lay- inverter gates within it, all us- to get a bit confused and may
outs or data sheets. Always able separately. Its pinouts are keep changing its mind about
ensure that the circuit’s
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1153
Maxfield & Montrose Interactive Inc
 7($&+,1 

F9> 1

F?<DC
FB3
B
B54
93!
!
F

3?=

2<;
3! B"

 4!
[ Q


21DD5BI
&F

2 F?ED

C55 D5HD

Fig.2.11 and Photo 2.6, breadboard layout for the first timing experiment using an inverter gate.

what logic level it’s being of-

F9>
fered. 1 C55 D5HD
6<1D
Although this dithering 
B!
would be a problem in a digital &F
21DD5BI
93!Q
:
B"
circuit, it’s of no great impor- Q [ !$
'$83 $
$'
FB3 " F?ED
tance for what we are going to !

do here, which is to connect the  ' Q

gate’s input to the resistor- 2 

3! 4!

[
capacitor series you have been F
charging and discharging.
Fig.2.12. Circuit diagram for the experiment in Fig.2.11, plus
(left) pinouts for a typical LED (light emitting diode).

INITIAL ASSEMBLY
F9> 1 Connect up your breadboard as shown in
Fig.2.11 (see also Photo 2.6). Note two things in
4" Q
4$
[ Q
4&
[
particular: the position of the flat side on the LED,
F?<DC
[ Q and the position of the “notch” (or dot/dimple, on
FB3
B54
B B B B some devices) of the 74HC04. (See also last
93!
! # % '
F

3?= month’s Practically Speaking article – Nov '99.)

2<;
3! B"
The circuit diagram for this component configura-
 4! B$
tion is shown in Fig.2.12.
[ Q 4#

Q [
Now perform some more capacitor charge/
B

[
& discharge experiments. You will see that the LED
 4% is on when the capacitor voltage is fairly low, and
21DD5BI
Q
&F
off when the voltage is fairly high. You may find
that the LED blinks a bit between the two levels –
2 F?ED! F?ED"
C55 D5HD
this is due to IC1 not being sure of its input logic
level. The effect is more likely to be seen when
Fig.2.13. Breadboard assembly of Fig.2.11
the time constant is really slow.
modified to include five more LEDs.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1154
Maxfield & Montrose Interactive Inc
 7($&+,1 

F9>

4$

4" Q [

[ Q
F?<DC
FB3
B B B
B54 ! # %
F
93!
3?=

2<;
3! B"

 4! B$

[ Q 4#

Q [


21DD5BI
&F

2 F?ED! F?ED" F?ED#

C55 D5HD

Fig.2.14 and Photo 2.7, layout for the oscillator experiment, (Fig.2.15). Note the new link be-
tween IC1 pins 4 and 5.
See if you can establish junctions of the LEDs and their Vout3 and Vin. See also Photo
what the capacitor voltage is respective resistors. Also note 2.7.
when the LED on-offness fully the voltages at the outputs of Connect up the power.
changes from one state to the the two gates – do they actually What you should see now is
other. reach 0V and +6V? that all four LEDs appear to be
What effect do two LEDs glowing, but at a reduced bril-
MORE LEDS have on the output voltages of liance level. In fact, they are all
the gates? Compare with the rapidly switching on and off, but
Just for fun, connect up an- voltages produced without LEDs too fast to differentiate between
other inverter gate (IC1b) and connected. We shall discuss them. In the author’s test model,
five more LEDs (D2 to D6) plus this in another Tutorial. Also see the rate was in excess of one
the extra resistors (R3 to R7 –
:
if you can draw the circuit dia- million cycles per second
also of 470 ) as shown in gram for Fig.2.13. (1MHz)!
Fig.2.13.
The clever thing we/you
Now you will find that D1 FLASHY have done is to use IC1c to in-
and D2 alternate in their on-off vert the output of IC1b, and
states. This is due to D2 being We wonder if you realize then to use the output of IC1c
connected to the +6V power how easy it is now to put the ca- as the power supply for the
supply, whereas D1 is con- pacitor charging/discharging un- resistor-capacitor chain.
nected to the 0V line. der automatic control for perpet-
ual repetition of the cycles? One With the correct combina-
The action of D3 and D4 will tion of R1 and C1 values, this
way to do it, using an additional
be seen to be the opposite of has the effect of repeatedly
inverter gate, IC1c, is shown in
D1 and D2 (as will D5 and D6). switching the voltage feeding
the circuit diagram of Fig.2.15
Which brings us to an interest- into R1 between +ve and 0V.
(we’ll discuss the change of IC
ing point about inverter gates. Here’s why:
type number from 74HC04 to
When two are used in series, as
74HC14 in a moment). When power is first
done here, a double inversion
Using the values shown, switched on, the voltage at the
occurs and so the final output
reconstruct your breadboard as- input to IC1a and the output of
logic level is the same as seen
sembly as illustrated in Fig.2.14 IC1b will be low (double inver-
by the input to the first gate.
(deleting D5, D6, R6, R7), and sion), and the output of IC1c will
What we’d also like you to be high (another inversion).
still using the 74HC04 device.
do is to make a note of the volt- This output is now supplying
Note that a crocodile-clipped
age that actually occurs at the +ve to R1, and C1 starts to
link is made between point
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1155
Maxfield & Montrose Interactive Inc
 7($&+,1 

like the 74HC04, has six invert-

Q Q ers within it and its pins are ar-
4" 4$
ranged in the same order. Note
[ [
B!  the symbol within each of the
B# B% inverter outlines in Fig.2.15 that
93!Q
$' : 93!R
$' : 93!3
&F
indicate its Schmitt trigger sta-
'$83!$ '$83!$ '$83!$
21DD5BI tus.
!$
" F?ED! F?ED"
FB3 ! # $ % &
F?ED#
With power disconnected,
' find a 74HC14 device from your
B" B$
bag of components and substi-
 $' : $' : tute it into the 74HC04 position
  C55 D5HD
3!

Q Q
on your breadboard.
4! 4# When power is re-applied,
[ [
F
you will see a considerable dif-
Fig.2.15. Circuit diagram for the oscillator experiment. ference in the rate at which the
LEDs now flash. Indeed, you
charge up (as it did when you tremely erratic, although it may should be able to count the
connected it directly to the +ve indicate a voltage at around the flashes quite readily. This dra-
voltage line). 3V mark (half-way between the matic change in the flash rate is
6V battery supply and 0V). entirely due to the switchover
We said earlier that inverter
occurring only at the guaranteed
gates have a threshold voltage,
Schmitt trigger logic levels.
above which an input level of SCHMITT TRIGGER
Logic 1 is assumed. Eventually, Using your meter you can
as C1 continues to charge, the As the circuit stands, its fre- now track the voltage level at
voltage at the input of IC1a will quency of oscillation is some- the R1/C1 junction at which the
rise above the threshold, and what unpredictable. We said logic changes occur. Also meter
IC1a’s output will fall to Logic 0. earlier that the 74HC04 has a the output of IC1c (pin 6 –
As a result, the output at IC1b midway input voltage level Vout3). You will see that it is
will immediately go high, and range in which the inverter is repeatedly switching between
the output of IC1c go low. not too sure which logic level is Logic 1 and Logic 0.
being applied to it. It is at this
This action, in an instant, midway level that the circuit is
causes C1 to start discharging rapidly switching over from one TIME OUT
through R1. Eventually, there state to another. What we ide-
comes the point when the dis- Before Part 3, think up
ally need is for the circuit to some timing and capacitor
charging voltage falls to the switch over only at the input lev-
Logic 0 level as seen by the in- value situations and see if you
els which are guaranteed to be can solve them using the vari-
put to IC1a. It now once more Logic 1 and Logic 0.
switches its output back to Logic ous software options and a cal-
1, IC1b output switches back to To achieve this exactitude culator. Also see if you can get
Logic 0, and IC1c switches to with an ordinary inverter gate the oscillator to run so that its
Logic 1 again. such as the 74HC04 would re- output at IC1c changes at exact
quire the use of additional cir- intervals of your choosing, say
The cycle has now been cuitry. However, there is a simi- once per second or once per 10
completed, and starts all over lar inverter type that automati- seconds. Until next month, 'bye
again. Thus it continues, ad- cally responds only to those in- for now.
infinitum, until something stops put voltages which are at the
it, such as you disconnecting guaranteed logic levels, ignor-
the power! ing those input voltages which
What you have created with lie between the two thresholds.
this simple component arrange- Such an inverter is known as a
ment, is an oscillator. Schmitt trigger inverter.
For interest, try to take a One type of Schmitt trigger
voltage reading at IC1c pin 6. inverter is the 74HC14 which,
You will find that is probably ex-
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1156
Maxfield & Montrose Interactive Inc
 7+( -2< 2) (/
E\ 0,.( &(5$16.,
Everything you need to know about electroluminescent recently when EL cable came
cable, but were too afraid to ask. onto the scene.
Electroluminescent cable offers
instruments in a Lexus glowing hobbyists, car buffs, inventors,
Have you ever seen a car home designers, modelers, sign
drive by at night with a really cool with a beautiful blue-white color
at night. Electroluminescent makers, and anyone with a
neon glow coming from creative streak, the ability to
underneath? While on trips to Los lamps have been around for a
long time. They offer add neon-like light effects to
Angeles and San Francisco I whatever they want.
witnessed such road fantasies, manufacturers a very even light
and decided that this was source that draws very little
something I wanted to know more power, and they last a very long EL CABLE
about. time, making them ideal for
hand-held electronics. The cable, also called wire,
Determined sleuthing is about the same diameter as
revealed the secret to this “neon” But EL lamps are hard to 16 gauge speaker wire (2 2mm).
¬

that was safe enough to attach to work with. They require custom
It is made of a soft, flexible-but-
the undersides and grillwork of inverters and support circuitry
solid copper wire that is covered
cars – a cool new lighting and cannot be cut to size or
with a special phosphor. Around
material called easily applied to a wide range of
this phosphor core are two tiny
electroluminescent cable. I surfaces. And they come in only
strands of wire a little thicker
tracked down a source and two colors… icky green or
than a human hair. These
bought some (at a ridiculously bluish ice-white. While suitable
“transmitter wires” as they are
expensive price) and proceeded for large manufacturers, they
called, are woven around the
to experiment. I quickly found are very unfriendly to hobbyists
core phosphor wire for its entire
more uses than just tricking up and inventors. That changed
length. All this is encased in
my car. This stuff has endless
possibilities and applications.

“EL” FOR SHORT

This article is as an
introduction to electroluminescent
cable – “EL” for short. (In later
articles I’ll cover more advanced
topics – complete with
schematics – such as Sound-to-
EL and programmable
sequencers.) You’ve probably
already encountered primitive
examples of electroluminescent Fig.1. Coral colored EL cable was glued to the inside of these
technology like the lighted display
5-inch tall plastic numerals. The numerals were then mounted
in your cell phone. There’s a very
good chance the back-light for it on a piece of 1”x 8” painted board and that in turn was
is coming from an mounted to the house. The project took 6 hours
electroluminescent lamp. Or you from start to finish. The cost of materials for this
may have noticed the dash project was $48.00.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1157
Maxfield & Montrose Interactive Inc
 6SHFLDO 3URMHFW
weatherproof, and even
waterproof if all connection and
termination points are properly
sealed. EL is flexible and easy to
apply. And it’s fun. Best of all, it’s
unique… people stop what they
are doing and stare when they
see EL wire for the first time.
Especially when they see it on
something that their mind tells
them can’t have neon… like roller
blades.

SURPRISINGLY
BRIGHT
EL is surprisingly bright.
Though it can’t compete against
sunshine or a well lit room, it
does quite well in twilight or
darkness. It can be used both as
an outliner, or as a back-light. In
terms of hours, EL lasts up to
Fig.2. Two meters of Blue EL cable shown attached to a driver 50,000 hours if run from
that also holds two AA sized batteries. This driver makes it household AC, making it ideal for
very easy to create portable EL projects such as costumes. A home theater and household
friend is mounting EL to a pair of skis so his night-time skiing remodeling such as recessed sofit
is cooler than cool. lighting. Cranking up the Hz
increases the brightness – and
vinyl sheathing to which makes EL work with lower voltage
color is added. When the sources like standard 9-volt
wire has AC current batteries – however it does
applied to it, the core wire shorten its lifetime. At 2000Hz,
glows a bright blue-white. EL will last approximately 2,500
The light is then shifted by hours. It never actually burns out,
the sheathing into one of but does burn down, becoming
four colors: yellow, blue, less bright.
purple or orange. A fifth Like most things today, EL is
color, aqua, is simply the available on the web. It costs
natural blue-white color anywhere from $3 to $4 (US
shining through clear Dollars) per foot. You can buy kits
sheathing. that include 10 feet of EL, a
EL cable can be cut to driver (an inverter that transforms
length with wire cutters. It the input voltage, such as 9V DC,
can be spliced with non-EL to the required AC output
wire. It’s cool to the touch. voltage), vinyl glue (for applying
It can be powered by a to smooth surfaces) and
Fig.3. Showing the blue EL cable wide range of power mounting tape (for mounting the
while powered up. The AA batteries sources such as 1 5V, ¬ inverter). A good web site is
can power 7 feet for approximately 7 2V, 9V batteries, or 12-V
¬
www.softneon.com, which
40 hours. This driver also features a automotive power. Even includes a comprehensive FAQ
blink mode that flashes the EL on 120V AC household (Frequently Asked Questions)
current can be used (albeit area, as well as plenty of
and off once per second. examples, and web pages on EL
carefully). It is CE
approved in Europe. It is splicing and termination.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1158
Maxfield & Montrose Interactive Inc
 6SHFLDO 3URMHFW
SoftNeon also offers a complete
store making it simple to buy EL
online.

NEXT TIME
My next EL article will
describe how you can build a
simple sequencer for EL,
allowing you to power different
EL segments in sequence. This
is excellent for animating a
series of EL caricature-frames
such as a jumping frog. As you
may recall, during the heyday of
glass neon in the 40s and 50s,
sequenced neon was the rage.
Now you can do it easily and
Fig.4. EL is very flexible and can be mounted even in tight affordably yourself and amaze
places like the inside of this plastic number 5. A Dremel was your family and friends.
used to remove excess plastic from inside the numerals where
needed. It took about 10 minutes to mount the EL inside each
numeral using vinyl super-glue. When I did this again for a
friend’s house-warming present, I painted the inside of the nu-
merals with white gloss paint and let it dry overnight before
mounting the EL inside. This made the reflected glow almost
50% brighter than my first version.

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1159
Maxfield & Montrose Interactive Inc
 SURFING THE INTERNET

By Alan Winstanley

JET SCREAM dial the Internet instead (it still own contributions therein. The
does, by the way). reader wondered if my mood
Several examples have could be gauged from the way I
arisen recently which illustrate Remember readers, that if
you ever have hardware or soft- signed off my messages: per-
how useful the Internet can be, haps a curt “ARW” signature
as well as how it can also be the ware problems, there’s a
chance you’re not alone, and signified a certain amount of
bane of my life. Recently a grumpiness (never), whilst the
friend asked me if I could help often all you need to do is
search for the answer on the full moniker – usually bashed
with a problem with his HP out in some haste I must say –
LaserJet, an old but very sturdy Internet. The problem is usually
where to start. Have a look at meant that I was feeling a tad
LaserJet 3, which was display- more affable that day. Who,
ing the dreaded message on its the brand new Help web site op-
erated by CNET at me?
LCD Call Service Error 50.
www.help.com One thing that does admit-
I own an LJ 3P as well, but tedly test my patience at times
my own experiences of out-of- So to help my friend with
his LaserJet 3, I started by typ- is when I’m on the receiving end
warranty service had left a dis- of some intemperate emails
tinctly bad taste. My ScanJet 4C ing “laserjet” into Deja News
(www.deja.com) to see what from users of our web or FTP
scanner suddenly broke down at site. However, acknowledging
quite a critical time, so my first other folks have said in the
past. After reading through the principle that customers are
port of call was the HP web site. always right (even when they
It had a list of dealers who could many of the newsgroup mes-
sages archived there, several are completely wrong, and no-
(reportedly) repair scanners, one ever won an argument with
and the site also had a map fea- web sites caught my eye.
a customer anyway), your hum-
ture, which displayed a It was not that long before I ble scribe bites his tongue and
zoomable road map of the UK turned up Parts Now Inc. sallies forth with an ever-helpful
showing my nearest dealer. (www.partsnowinc.com) and reply.
I printed that off, threw ev- also All Laser Service in Califor-
Following the launch of
erything in the car and went off nia (www.all-laser.com) and –
Teach-In 2000, my E-mail has
in search of a scanner savior. bingo – there was a page de-
been alive with requests for
Two weeks later the HP dealer voted to Call Service Error 50!
help from readers who are new
charged the equivalent of nearly The problem could be a fuser,
to electronics, new to comput-
100 UK Pounds not to mend it, triac, thermistor, or halogen
ing, or new to the Internet (or
because after putting in about a lamp, they say. Using an ohm-
new to all three). Although I’m
day’s work it transpired that they meter we can now hopefully pin-
happy to oblige, surprisingly
can’t be fixed anyway. (“There point the fault and fix this one
there has been more than one
is no repair path,” the official ourselves, if we can get the
unfavorable comment about the
jargon explained.) spares.
FTP site, some users having
The replacement scanner, apparently been foiled by the
an HP 6250, has had more than HELP US TO HELP process of File Transfer Proto-
its fair share of installation diffi- col.
culties, and again the Internet
YOU
As one reader put it, “yours
proved vital in finding the I was once quite amused by is the one web site in the world
patches and fixes from the HP a quip made by a reader in rela- which has defeated me,” actu-
web site. Their slow but moder- tion to our Chat Zone service (at ally referring to the FTP file
ately useful scanner forum also the web site for the printed edi- server, which I must say is ex-
helped explain why, when I tried tion of EPE at tremely reliable and has tons of
to share the scanner on my net- www.epemag.wimborne.co.uk, bandwidth at its disposal. The
work, the host PC would try to Ed), or more specifically, my problem is that Internet users,

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1160
Maxfield & Montrose Interactive Inc
 1HW :RUN
especially newcomers, are pro- less, it should be accepted that ture backup. You could alterna-
gressively being spoiled by a browser upgrade will be re- tively have a look at EPE On-
world wide web sites, to the total quired sooner or later – maybe line’s web site
exclusion of the other ways of every 12 to 18 months or so. (www.epemag.com) and fetch
making information readily Presently the ideal answer files from our web server hosted
available over the Internet. is really to use proper FTP soft- in the USA.
So when Teach-In 2000 is ware, which will be second na- Occasionally, readers are
launched and I receive several ture to seasoned users. I regret so stuck that they ask that I
complaints about the hopeless- it when users take umbrage at send files to them by email in-
ness of our FTP site, the weary the suggestion that an upgrade stead, and I will of course try to
writer starts to feel rather exas- is required, or that we are trying oblige, although it is always
perated. I have described the to bar MSIE 4.0 users from the more encouraging to know that
processes of FTP several times FTP site. Try to upgrade from users have tried to help them-
in the past. The first problem is the obsolete browser if possible. selves to begin with.
that FTP is FTP, not the hyper- Internet Explorer 5 has the If you have any queries,
text transfer protocol associated honor of being the first Mi- comments or (whisper) com-
with web servers. crosoft browser I could actually plaints, please feel free to share
recommend. them with other readers in the
BROW-BEATEN EPE Chat Zone at
Web browsers have varying HELP-LINE www.epemag.wimborne.co.u
k or email them to me at
degrees of success or tolerance Here at EPE HQ we want all
[email protected].
when accessing FTP sites, and readers to enjoy such splendid
in my own experience, Microsoft series as Teach-In 2000, so we
Internet Explorer 5 is far more try to lend a hand where we
obliging with the process of can, often going well beyond the
anonymous FTP than version call of normal duty as many
4.0 ever was. Furthermore, ev- readers will confirm. When
ery instance of “extended re- things don’t seem to go right, it
sponse” server error messages is very easy to dash off an ur-
(generated for whatever reason) gent or intolerant email on the
arose, in my experience, due to spur of the moment, just be-
the use of MSIE 4.0, never any- cause a user has experienced
one’s favorite browser. As I some frustration or other.
state on the EPE web site, such It’s also very easy now to
error messages are browser is- send an impatient “chaser” the
sues, not related to our server. next day, which merely adds to
All such problems seem to our volume of work. It isn’t our
have gone after adopting MSIE fault if a user’s browser is flaky,
5.0, which deals with anony- or if a beginner is frustrated with
mous FTP in an orderly fashion. the complex techniques of oper-
If some readers are nervous ating a personal computer, or
about upgrading their Microsoft has never heard of FTP before
browser they have every reason now.
to be so. Sometimes it goes You don’t need to fetch a
smoothly, at other times a hefty browser upgrade from the
wheel might fall off in the pro- Internet either (another reader
cess, causing major headaches complaint), as browser up-
for the user who has usually grades are regularly included in
done nothing wrong at all. computer magazine cover-
For evidence of this, one mounted CD ROMs, and fur-
only has to read the Microsoft or thermore, the upgrade is there
W95/98 newsgroups. Neverthe- on an indestructible CD for fu-

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1161
Maxfield & Montrose Interactive Inc
 A ROUNDUP OF THE LATEST EVERYDAY NEWS
FROM THE WORLD OF ELECTRONICS

Pirate-Proof CDs
Long hoped for by the music industry, uncopyable CDs are now a reality.BARRY
FOX REPORTS.

New technology spells bad takes advantage of the fact that

news for people who use a PC the standard for music CDs, NOTABLE PARADOX
to copy music CDs or send known as the Red Book, was
set before the standard for CD- Paradoxically C-Dilla’s
them over the Internet. British breakthrough and Macrovision
company C-Dilla has found a ROMs, known as the Yellow
Book. takeover come just as the music
way to let a music CD play on a industry’s Secure Digital Music
home hi-fi, but not on a PC’s The ROM standard provides Initiative group has agreed with
CD-ROM drive. more powerful error correction the electronic companies to al-
Computer software compa- for data than is needed for mu- low owners of CDs to “rip”
nies, including Microsoft and sic. ROM drives are designed to copies into a PC
Lotus, already use C-Dilla’s handle either music or data (www.sdmi.org). There had
SafeDisc system to stop people discs. AudioLok adds false error
previously seemed no foolproof
copying ROM data discs. correction code to a music disc.
way to stop copying altogether.
An ordinary music CD player
SafeDisc puts the program simply ignores this extra code Says Paul Jessop, Director
material in an encrypted wrap- and plays the disc as normal. of Technology at the music in-
per, which can only be un- But a ROM drive reads the false dustry’s world trade body the
wrapped when a digital key code and rejects the disc as un- International Federation of the
code on the disc matches an playable. This stops the owner Phonographic Industry:
authorization code entered into sending the music over the In- “Although in general the
the PC. The key code is ternet or copying it onto a blank recording industry welcomes
pressed into the disc so that a disc. people listening to CDs on com-
ROM drive can read it but a
A prototype AudioLok disc puters, the ability to make discs
CD-recorder cannot copy it. So
duly played on a CD music that cannot be copied on com-
only the original disc will run the
program. player but refused to play or puters may be of considerable
copy on a PC. Peter Newman interest to some record compa-
says he is confident that he can nies.”
UNREPEATABLE also stop a consumer music
DREAMS CD-recorder making a copy, be- CHINESE AND
cause these devices are already
The record industry has designed not to copy CD-ROMs. CHIPS
been dreaming of just such an He expects AudioLok to be
anticopy system for 30 years, NEC Corporation and part-
ready for commercial launch in
since the Beatles claimed that ner Shanghai Hua Hong
a year.
their LP Sergeant Pepper could (Group) Co. Ltd., have officially
Macrovision of the US has opened their joint-venture semi-
be played but not copied. Like
now bought C-Dilla for around conductor plant, the largest in
the many systems that followed,
$18 million. Macrovision devel- China. Concentrating production
Pepper was as easily copied as
oped the systems which film on Dynamic Random Access
played.
and TV companies already use Memories (DRAM), production
Peter Newman, who to stop people copying videos. capacity is expected to expand
founded C-Dilla in 1991 and in- Now the company can offer to 20,000 wafers per month by
vented SafeDisc, has finally the same option to the music the end of year 2000. Currently
found the answer. AudioLok industry.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1162
Maxfield & Montrose Interactive Inc
 1(:6

supplying their home market,

the company proposes to even-
MAGENTA LOGS ON
tually manufacture for world
markets.
NEC pioneered the concept
of C&C, the integration of Com-
puters and Communications.
They employ in excess of
150,000 people around the
globe.
For more information visit
www.nec-global.com

MICROMOUSE
GRAND PRIX
Astonishing – July next year Specialist kit and component suppliers Magenta Electronics have
will see the action of an exciting introduced their own superb kit version of John Becker’s 8-Channel
new challenge for enterprising Analog Data Logger (EPE Online Aug/Sep ‘99).
11 to 18 year olds. The Micro- An enhancement welcomed by the author is the direct mounting of
mouse Grand Prix 2000 is being the liquid crystal display directly onto the main PCB using rigid stand-
organized by the UK’s key engi- off pillars. This makes the assembly considerably more robust than the
neering association Young Engi- original prototype. The inclusion of the switches on the PCB is also an
neers, in collaboration with the enhancement.
IEE (Institution of Electrical En- As is typical of Magenta’s kits, the quality of the components and
gineers), Europe’s largest pro- of the redesigned PCB is excellent.
fessional engineering society.
For more information contact Magenta Electronics Ltd., Dept EPE,
Teams of up to four are be- 135 Hunter Street, Burton-on-Trent, Staffs, DE14 2ST, UK.
ing invited to build and race
their own robot, i.e. a small ve- Tel: +44 (0) 1283-565435
hicle capable of finding its own Fax: +44 (0) 1283-546932
way round a course at high Email: [email protected]
speed. Entrants to this chal-
lenge will have specialized sup- Web: www.magenta2000.co.uk
port through the Young Engi-
neers website at both Fixers and those needing a
www.youngeng.org for chas- WECAN- Fixer. They say: “In our proper
sis design, electronics, steering work we build websites and
control, digital and pro-
grammable technology. There
FIXIT4U publish on CD-ROM, and offer
consultancy on the matching of
are several classes in which you Recently faxed through to Content to End User’’.
can compete. us is information about an inter-
For more information con-
esting website,
Preliminary race days take tact David Hall, Chameleon HH
www.wecanfixit4u.com. The
place around the UK in March Publishing Ltd., Dept EPE, The
site offers a free directory of fix-
2000, with the Grand Prix itself Quarry House, East End, Wit-
ers, repairers, restorers, and
in July. For more information ney, Oxon OX8 6QA, UK.
conservators in the UK. The
contact Fiona Hunt at Young Tel: +44 (0) 1993-880223
company invites readers to not
Engineers' Press Office on +44
only use the site to search out Fax: +44 (0) 1993-880236
(0) 171-823-3799 or Christina
the services they need, but also
Dagnall of IEE Media Relations Email:
to have their own skills listed as
on +44 (0) 171-344-5445. [email protected]
well. They aim to provide a
Please mention EPE Online
global directory searchable by Web: www.wecanfixit4u.com
when phoning.
item and hope to bring together
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1163
Maxfield & Montrose Interactive Inc
 1(:6

GET STUFFED AT
CYNTHIA’S
The worldly pleasures of this planet’s
First Robotic Bar and Restaurant are sampled by
John Becker.

Not that we’d ever suggest Accepting your order (they

you stop chatting up your fa- do obey at least one of Asi-
vorite local bar wench, but from mov’s Laws of Robotics, para-
behind her bar Cynthia’s really phrased as – Thou shalt not
PICS AND LCDS got what it takes to get you harm or through inactivity cause
New from Lascar Electron- drooling! Ah, Earthlings, we harm to occur to a human – and
ics is a 5-digit starburst liquid have a tale to tell of sensuous it would harm you to do without
crystal display module with an cosmic delights and entertain- your beverage, wouldn’t it?),
on-board Flash PICMicro. It has ment that you’ll enjoy when Cynthia and Rastus pivot round
been designed as the building Cynthia responds to every to the vast array of drinks on
block for thousands of display finger-tip’s request! She’s well optic behind them, and fill your
and instrumentation applica- programmed to serve you! glass to the correct measure.
tions. Using standard software “But who is Cynthia?”, we While you’re at the bar, it’s
routines that can be down- hear the cry from our valiant you who are likely to be chatted
loaded from Lascar’s website, readers, thrusting hot soldering up by Cynthia and Rastus. We’d
users can tailor the module’s irons hard into their holders. like to say that the tone of chat
function to suit their own needs. Gather round – Cynthia’s the respects all Laws of Polite Con-
Because of the Flash memory most amazing anthrobot you’re versation – but we can’t lie to
on board, users can re-program likely to meet this side of the you, can we? You just have to
the module as often as they galaxy and, together with cyber- accept that the occasional
like. Readers will know that EPE partner Rastus, is the star fea- “questionable phraseology”
Online has so far concentrated ture of a new theme bar and might occur! But it’s all in good
on PIC-controlling “intelligent” restaurant that’s just opened in fun and humor, and has nothing
alphanumeric LCDs. Here now London. that would not be heard in one
is an opportunity to make use of of the old “Carry On” films.
the more basic types of LCD in
conjunction with PIC devices. CYBER CHAT
CULINARY CALLING
Lascar’s module features Cynthia and Rastus are two
eight analog inputs and up to 22 2-meter high robots, each with After you’ve been cajoled
digital I/O ports, capable of driv- their own cavernous and glitter- by others of human persuasion
ing or interfacing with other sys- ing bar area from which they to vacate your place at the bar,
tems. serve the cocktails and other you have yet more delights to
drinks you’ve ordered through pursue – culinary ones. In other
For more information con- words, it’s along the glittering
tact Lascar Electronics, Dept their 75-option keypads. Rastus
is a bit of a DJ as well. As with corridor to Cynthia’s restaurant.
EPE, Module House,
Whiteparish, Salisbury, Wilts any Earthly (or even Earthy!) The centrally-illuminated
SP5 2SJ, UK. bar tender, these two cyberoids dining tables have call-buttons
respond to your orders with inset, offering choice of the ser-
Tel: +44 (0) 1794-884567 varying degrees of good or bad vice required: drinks, food, gen-
Fax: +44 (0) 1794-884616 grace (depending on their eral, and bill. Hi-tech is a key-
mood, and the state of their pro- word even in the way you are
Email: las-
gram cycle – which in turn re- attended at table. Attired in
[email protected]
flects the state of mind of their fetching Millennial black and
Web: original designers and program- silver togas, reassuringly human
www.lascarelectronics.com mers!). staff use handheld electronic

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1164
Maxfield & Montrose Interactive Inc
 1(:6

order pads. Your origins. There is the

order is keyed in profound feeling
and transmitted by that Cynthia is a
short range radio distant relative of
through to the “Marvin the Para-
well-equipped noid Android”, the
kitchens. All robot who, reluc-
“plastic” financial tantly, was involved
transactions are in “Hitch-Hiking
via a commercial Through The
networked EPOS Galaxy”.
system.
Described as EVOLUTIONARY
“Multi-national”, RELATIVITY
the selection,
As to whether
quality and com-
Cynthia is an an-
petitive pricing of Richard Becker (Cynthia’s Conceptual Parent) requests
cestor of Marvin, or
the food is compa- Roger Gay (Cynthia’s Behaviorist) to hand over the decla- his descendent, will
rable to that ration that Cynthia will always converse in a polite and so- probably never be
served in many cially acceptable manner. At the time that this reviewer de- ascertained, the
good restaurants
around the planet, parted, no such undertaking had been received (but Cyn- space-time chrono-
thia had sweetly growled “B****r-off Human” synclastic-para-
and there’s a spe-
fundi-bulum of the
cial menu for
and ascending orbit through the polyverse is far too
“mini-droids”!
transparent stems. Cynthia’s multitemporal to ever establish
and Rastus’ dominance at the who’s whose relative and in
ORBITAL SPACE-WAYS end of their respective supra- what order from the Event Hori-
Being in the comfort zone of spacial caverns enhances the zon, but there’s a family like-
a vast orbiting space station, feeling of outer-worldliness; and ness there somewhere
Arthur C. Clarke 2001 style: hints at their possible “genetic” (relatively speaking)!
that’s the futuristic atmosphere To drop back out of warp-
at Cynthia’s Cyberbar and time(!), “family” is involved in
Restaurant. You forget that it’s this Cyber-venture in another
all more down to Earth, set be- way, this commentator’s family,
low London Bridge, in the mass in the shape of his brother,
of broad tunnels and brick-built Richard Becker.
caverns that pervade that area Those of you who recall
of London. earlier days of electronic hobby-
The walls, floors and ceil- ing will probably remember that
ings are covered in a silvered Practical Electronics, in Novem-
metal skin and well interlaced ber 1981, published a robotic
with great expanses of mirrors. arm, Genesis, designed by
Any camera flash has a half-life Richard. This was very much a
of a thousand years (or so it “first” not only for Dick but also
seems). So does the image of for PE, which at that time was
the drinkers and diners – echo- edited by Mike Kenward (now
ing down to the ends of the uni- our Owner, MD, and Editor-in-
verse. Genesis in various forms as Chief).
Myriads of light emitting published in PE, Nov '81. Through his company Pow-
diodes enhance the entire Rumor has it that the white- ertran Cybernetics, Dick built up
lengths of the “populated” ar- coated android was a work- a worldwide market for his edu-
eas. Those in the bar stools ap- cational and light industrial
pear to be in constant twinkling
ing facsimile of today’s robotics products, which be-
editor-in-chief Mike Kenward.
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1165
Maxfield & Montrose Interactive Inc
 1(:6

came increasing more sophisti- processor used (well, a modern

cated at each generation. To cut 16-bit derivative of it anyway).
YOUNG AMATEUR
short a length of history, Dick “Why change a working sys- AWARDS
went on to found Cybernetic In- tem?”, says Dick, having long
ago optimized software- Sixteen-year-old Mark
struments Ltd, of which Cyn-
hardware interfaces for all his Haynes from Harlow, Essex was
thia’s Cyberbar is a division.
automation products. recently announced as the win-
ner of the Radio Communica-
CHARTING SPACE-TIME In fact there are 12 slave tions Agency’s Young Amateur
microprocessors, one for each of the Year Award 1999. Mark
Cynthia herself (though of the motors, all under control
we’re not really sure of her/his received first prize of 500 UK
of a master processor. A PC- Pounds, a certificate from
gender!) became a twinkle in compatible computer is in over-
Dick’s imaginative eye a good Stephen Byers, Trade and In-
all charge of the system, includ- dustry Secretary, and will be
ten years back. Sworn to se- ing the drinks ordering keypads
crecy, this author has seen invited to a conducted tour of
and Cynthia’s speech genera- the RA’s Monitoring Station in
great wall lengths in one of tion.
Dick’s large factory units in- Baldock, Herts.
creasingly covered by hundreds Cynthia’s inner organs are a Mark gained his Novice Li-
of mechanical drawings. Each sight to behold! Her body is cence at the age of 12 and be-
represented a part of Cynthia packed with thoroughly popu- came the youngest radio ama-
Mk1, manufactured and assem- lated printed circuit boards and teur of his home town. In July
bled when time permitted be- stacked in awe-inspiring regi- he organised and ran a special
tween other commitments. mentation. The scene behind event station commemorating
the drinks array leaves one al- the 175th anniversary of the
Genesis and many more of most dumb with admiration at
Dick’s earlier robotic arms RNLI.
how complex a system is re-
ranges were hydraulically con- quired to select and serve the If you would like to become
trolled (water for blood!). Cyn- correct drink on demand. Me- involved in Amateur Radio, con-
thia’s motion, though, is gener- chanical and electronic inter- tact The Radio Society of Great
ated by precision stepper mo- faces abound, thick neural-like Britain, Lambda House, Cran-
tors operated under tight closed- cable harnesses snaking their bourne Road, Potters Bar,
loop control. They are operated way amongst them. Herts, EN6 3JE, UK. Tel: +44
with varying degrees of resolu- (0) 1707-659015.
tion, from a basic 200 steps up
to around 12,800 when in micro- PLANETARY
step mode. The various limb CO-ORDINATES
motions are on a double axis,
Children are welcome
horizontal and vertical move-
(there’s even a dance floor, and
ment. The sophistication of the
soon there’ll be an amazingly
control software ensures that
fascinating technology-
movement is smooth, with dif-
orientated gift shop). Undoubt-
ferent rates of acceleration and
edly, all of you within sub-orbital
deceleration being applied de-
distance of Cynthia and Rastus
pending on the position of mo-
will by now be utterly consumed
tion. There was no need to give
with desire to drink with them,
Cynthia third-axis (rotational)
and to dine with their human
limb and wrist movement.
entourage.
Here’s how: the address is
INNER SPACE Cynthia’s Cyberbar, 4 Tooley
One might expect that the Street, London SE1 2SY, UK.
entire system would be gov- Tel: +44 (0) 171-403-6777
erned by the latest in microcon-
trollers. Not at all – that well- Fax: +44 (0) 171-378-1918.
proved and time-honored fa- Email: [email protected]
vorite, the 8051, is the micro- Web: www.cynbar.co.uk
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1166
Maxfield & Montrose Interactive Inc
 John Becker addresses some of the general points readers have raised. Have
you anything interesting to say? Email us at [email protected]!
Ed); we had little if any training inches off the bench then drop it
WIN A DIGITAL on these. We were told “Not to back on, apparently to see the
MULTIMETER bother checking these they don’t effect, if any, on the screen!
go faulty!” (I’ve heard that one Then each of the valves was
A 3 1/2 digit pocket-sized
before). “Anyhow, they will given a good tap with the thick
LCD multimeter, which measures
never replace the good old end of a screwdriver (is this
AC and DC voltage, DC current,
valve.” I did manage to finish where the term bottle bashers
and resistance. It can also test
the course and get a job came from?). This was to see if
diodes and bipolar transistors.
working for Granada Television, the valve was on the verge of
Every month we will give a Digital
not to be mistaken with death, or noisy! There were
Multimeter to the author of the
Granada Broadcasting station. I many more such nonsense acts
best Readout letter.
was literally working on a of violence...
conveyor belt. We had to pick We were on piecework as
* LETTER OF THE up the sets (Murphy model 5?) well. We did have a much
from one end and refurbish better set to work on, namely
MONTH * them. This meant changing the the KB, I don’t know the model
on/off switch (the one being number, but this was a hard-
TELE BYGONES replaced was operated by lifting wired set, i.e. no PCB. The KB
the lid), replacing the frame had a frame output transformer
Dear EPE, output valve and frame bias, at the top left-hand side looking
Whilst looking for a clean the rotating channel from behind. The quality
computer magazine, I came switch, replacing the two valves controller thought this was the
across EPE, glanced through, (I think these were EC80 and best place to tie his label, to tell
then decided to purchase it. It ECC81). the world that this set had
brought back a lot of memories. AC power was rectified with passed his inspection. All was
Many moons ago (mid 1960s), I a valve. This we replaced by a well until the team responsible
did a Government training half-wave rectifier, using two for re-casing got their hands on
course, to become a Television diodes, and then replaced the it and, yes you’ve guessed it,
and Radio Engineer, with the smoothing capacitors. If any pulled the tag off along with the
promise of a job at the end. other valves had to be replaced, transformer wires. Oh the good
About half way through, our we had to gently ease the CRT old days!
instructor had to go on a three- forward in order to do so. Part I left this type of electronics
day course. He returned, with of the HT was enclosed in an oil to work with hearing aids. This
the biggest TV receiver container, if the HT arced over was sub-sub-miniature work,
imaginable, the first color TV then you had to inject more oil! and very rewarding. A resistor
any of us had seen. When it Never try to align anything one was just bigger than a pinhead.
was switched on it worked for small turn – this could lead to Now I believe they use ICs.
just a short time, when all at replacing of IF can, or whatever
I have now ordered a two-
once, a bang, then smoke. The it was you were turning. This
year subscription to EPE, and
cause was something to do with happened due to the age of the
set. When you had done all the also twelve months back issues.
HT. I remember seeing what I shall also be sending for three
seemed to be three or four very work that needed to be done, it
was time to ask the quality CDs to help me. I have a lot of
large wirewound resistors burnt catching up to do!
out. These were the days when controllers to inspect the set.
the transistor was first brought The first thing they did was to Keith Barlow
about (actually invented in 1947. lift the set up three to four Bury, Lancs, UK

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1167
Maxfield & Montrose Interactive Inc
 5HDGRXW
Isn’t history fascinating! ally Big Thank You to Mr W.F. tempt is made to Install the pro-
Welcome to the modern world Ritchie, of Fraserburgh, Ab- gram with its zipped files being
of electronics, Keith. Good luck erdeenshire, who went to a lot in the wrong directory. When
with your “catching up”. It’s good of trouble to send us a great the program is Installed from our
to hear from you. deal of information on the sub- 3·5-inch disk as available from
ject, including tables of data. the Editorial Office, it seems un-
Sadly, it is too lengthy to repro- likely that this can happen.
TOOLKIT MK2 AND WIN duce here. However, readers who have
98 One of the many interesting downloaded the program from
Dear EPE, points made is that, whereas our web site may inadvertently
I would like to comment on the changeover to a Gregorian find that their unzipped files are
the parallel port interface soft- calendar (in honor of Pope Gre- in the wrong directory. The files
ware for the PIC Toolkit Mk2 gory XIII) began in 1582, Britain need to be installed in directory
(May-Jun '99). SETUP showed did not adopt it until 1752, which C:\ PIC, as expected by various
that the ACK bit was not work- caused considerable contro- file-accessing commands within
ing. However, inspection with versy because, in that year, 11 the suite of programs.
DEBUG and direct manipulation days had to vanish (“Give us There is normally a text file
of the port registers showed that back our 11 days!”). Greece on the FTP/Toolkit Mk2 path
the bit reflected the correct was even later – it was the last that explains this
value. modern nation to make the (pic_toolkit_instal.txt), al-
change, in 1923. though there was a brief period
Single stepping past the
INP instruction in TK- Mr Ritchie also says that “At during which the file “went miss-
SET06.BAS worked too, so I the age of 78 years I make do ing”. It should be there now –
suspected a timing issue. I with my Hewlett Packard 48GX follow its instructions (if it’s not,
duplicated the INP instruction to calculator, which . . . has a built- advise the Webmaster for that
add a small delay and it worked in clock and calendar covering a site).
OK. period from the start of the Gre- The ASMCNV directory
gorian calendar to 31st Decem-
This was using a 233MHz (folder) referred to must be cre-
ber 9999”. Astonishing!
Pentium running Win 98 ated as C:\ ASMCNV. (It should
(although the same problem not be created in the PIC direc-
occurred in DOS mode). DOS ERROR 76 tory, where it will not be found.)
Pete Shew Dear EPE, Incidentally, readers with
via the Net I have built PIC Toolkit Mk2 QBasic or QuickBASIC can find
(May-Jun '99) and the first part out what a particular DOS Error
of the Setup, where the volt- number means by entering the
Very useful comment Pete, ages and parallel port are program-writing area of these
thanks. All my software is tested checked is OK. However, when programs and typing in the com-
on four machines of varying I press enter to carry on with the mand, for example, ERROR 76.
ages and abilities, using Win 3.1 Setup, I get the message: Then run this one-line program,
and Win 95, at speeds up to “Setup program unforeseen MS upon which the program will halt
120MHz. I do not have a Win 98 DOS ERROR 76’’. Can you and display a text box contain-
machine and so any info on that please help me? ing the relevant error message.
system’s operation is useful in
this context. Anthony Marshall
via the Net
LOGGING EXCELS
LEAPING CALENDARS
Dear EPE,
(AGAIN) The ERROR 76 message
(“path not found”) is generated I’ve just read Part 2 of the
In several recent Readout 8-Channel Analog Data Logger
by MS DOS when it cannot find
columns, we have discussed (Aug-Sep '99) and have to say I
a particular file or directory
calendars in relation to the Mil- like it. I have always wanted to
named by the user. In Toolkit’s
lennium. We want to say a re- build one using a good old Z80
Setup this could occur if an at-
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1168
Maxfield & Montrose Interactive Inc
 5HDGRXW
CPU and a ZNxxx but even responsibilities, and the neces- plete evaluation of any mag-
when I had the chips I never got sity to earn a living, and a move netic material is a complex sub-
around to it! Your design is to Canada in 1954. ject, full of mind boggling
made so easy with the Only now, in retirement, maths! But it seems to me that
PIC16F877. have I the time to “convert” my- if we know its composition and
Being able to upload the self from the valve to solid designed frequency, this would
data to a PC is really essential, state. What a fascinating sub- give confidence to incorporate
and your comments on using ject it is! I find my main inter- the item in a project with rea-
Excel to view the data are very ests are receivers and test in- sonable chance of success.
good. struments and have recently I look forward to perhaps
I have also downloaded ver- built a number of them, includ- seeing an article on the above!
sion V2.3 of Toolkit Mk2 – I ing some from EPE. Now I have Please keep the practical pro-
have used the dissemble func- subscriptions to five electronics jects and informational articles
tion to recover a program from magazines from the USA, and coming.
a PIC16C84 that I had lost the two from England, and of the
seven your magazine stands B.J. Maloney
.ASM text for, it will make Alberton, Canada
rewriting a lot easier! head and shoulders above them
all. I can only say it is simply the
Mel Saunders BEST! We know that many readers
via the Net The series of articles on os- have responded favorably to
cillators by Raymond Haigh is Raymond’s oscillator discus-
excellent, the subject is dealt sions. Toroids-wise, though, we
It’s good to know that a de-
with in depth and gives down-to- suspect that an article relating
sign that took me so much time
earth schematics. The articles to them would be too esoteric to
to research and implement is
give one confidence to go appeal to most readers. How-
providing you and many other
ahead and build each type of ever, perhaps readers might
readers with a useful tool. There
oscillator, and indeed I have al- care to tell us we’re wrong!
were four “learning curves” in-
ready built some of them, and I
volved – getting to know the intend to construct many more. We appreciate your praising
PIC16F877, the serial memo- words. With our international
ries, serial communication be- I also have another request, readership continuing to grow,
tween PIC and PC, plus Excel and perhaps some of your in a large part due to our EPE
(to which I had previously only columnists or readers, better Online editions on the Internet,
had brief exposure when we versed in electronics than I, it’s good to learn what readers
produced the CD-ROM for PIC- could provide a method and a worldwide think of us.
tutor). circuit, for the testing and evalu-
ation of toroids. This compo-
nent, as is well known, exists on TRANSISTOR PROBLEM
A WAVE FOR OSCILLA- the surplus market in a vast ar- Dear EPE,
TORS ray of sizes and material com-
position, and they can be I’m having some trouble
Dear EPE, bought very cheaply. But not getting my PIC Toolkit Mk2
This is the first letter I have knowing its composition, i.e. (May-Jun '99) to work and hope
ever written to an electronics iron dust or ferrite, and what fre- you can be of some help.
magazine. To establish, very quency it was designed for, one I am not getting the correct
briefly, my background, I was would hesitate to incorporate it voltage measurements and I
brought up with the vacuum in any project. believe it has to do with two of
tube starting in the early thirties. Perhaps the circuit would the components: the power sup-
Although I was keen to build re- take the form of a BH curve ply and TR1.
ceivers, money was scarce, and tracer, where, instead of a 50Hz First, when the parallel port
I could only read about their or 60Hz input, a standard signal bit DA4 is high, the voltage on
construction. Then came WWII generator could activate the cir- PIC pin 4 (MCLR) should be
and I tried to get into Signals, cuit, and the output displayed 12V, but I am only getting 8 2V.
¬
but it was not to be. I served in on the scope.
the infantry. Then came family Now the 78LC05 power con-
I realize that a full and com- verter I used is only rated to
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1169
Maxfield & Montrose Interactive Inc
 5HDGRXW
supply 100mA and the LED I with those shown in Fig.4 of the interested in PICs at the pro-
chose draws 20mA. Could there published article (you may need gramming as well as the appli-
not be enough current going to to “twist” the device if they are in cations levels.
the MAX665 Flash memory pro- a different order). A couple of years (or so)
grammer for it to supply 12V? One alternative to the back, I ran an experiment with
Should I replace both the power
BC549 (a device which is part of one of my published PIC pro-
converter to get more juice and
my regular design stock) is the jects. I discussed at length one
the LED to consume less?
BC109. Another is the 2N3704, aspect of the program that con-
Second, when the parallel but note that this device has a trolled the project, in order to
port bits DA3 and DA4 are high, pinout of ECB whereas the see what reader response
the voltage on MCLR should be BC549 and BC109 have CBE. would be to that extended dis-
0V, but I still get 8 6V. When
¬
In reality, practically any general cussion.
DA3 is high alone, I get 4 5V on
¬ purpose npn transistor should The result was astonishing
MCLR, so something is working. work if correctly orientated, it’s – many readers expressed their
I used a different npn transistor only being asked to switch a gratitude for the discussion. So
for the reset instead of the very small current on and off. much so, that I felt justified in
BC549 specified because I Having sent the above info suggesting to Mike the PIC Tu-
could not find that listed in any direct to Fred, he subsequently torial series, which we subse-
of my catalogs here in the USA. emailed back: “Thanks a ton, it quently ran from March to May
Instead I used a generic npn works fine!!” '98. It was one of the most suc-
with a maximum collector cur- cessful series that EPE has
rent of 600mA.
ever run. Demand for back is-
I have just started working PICKING UP ON ED
sues (photocopies only now) of
with PIC micros and I really en- Dear EPE, the series still continues. The
joy your magazine because of In your Editorial of Nov '99, success of the series also
your concentration on them. I you say “It seems that some of prompted us to further develop
purchased your PIC Toolkit PCB our readers are definitely not the concept and produce the
and put together this program- interested in PICs’’. It’s not the PICtutor CD-ROM and associ-
mer to see what I could learn PICs they’re not interested in it’s ated hardware (see the “Books
and save a few bucks. I appreci- the endless discussions on code and CDs” section of the EPE
ate your work. Thanks very that put them off!. I feel the Online website, Ed.).
much. same. But as to actually indulging
Fred Ramsing Why not steer clear of code in “endless discussions of
University of Nevada and talk in terms of Basic pro- code”, we don’t do so as a regu-
Reno, USA gramming with which a vast lar part of PIC projects. By-and-
number of your readers must be large, the only discussion of
familiar? I notice that some code is when it is pertinent to
The LED is unlikely to be companies offer PIC Basic explaining how a particular de-
the cause of the problem since compilers. Why bother with the sign should be operated. Other-
its current is limited by resistor grief of learning code? Please wise, extended discussions
R13 and does not depend on enlighten me. about code from a program-
the LED’s actual rating (which mer’s point of view have been
states the maximum current at Murray Cameron confined to such educational
which the device can be safely via the Net features as the PIC Tutorial and
operated, not the current at PIC16F877 Mini Tutorial (and
which it always works). Readout! ). Even my 8-Channel
It seems probable that the Analog Data Logger of Aug/Sep
transistor is to blame, perhaps Ah, Murray, you’ve misunder- '99 (which for the first time intro-
because its pins are not orien- stood Editor Mike’s statement! duced the PIC16F877 as part of
tated correctly. Check the data By “some’’ is meant that “a few’’ a project) did not significantly
sheet (or supplier) for the pinout – a minority in fact – of our discuss code, that being left to
of the device and ensure that readers are not interested. The the Mini Tutorial.
the pin designations correspond vast majority most definitely are Regarding PIC Basic com-

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1170
Maxfield & Montrose Interactive Inc
 5HDGRXW
pilers, I am sure that for short There is no problem with pin
lengths of code writing they are 10 on IC6 being connected to Good to hear from you
probably an excellent asset for ground. This is required to make again George. Yes, I’ve thought
some readers who do not have IC6 continuously transmit a sig- about Sundials for Dull Days
the inclination to delve into writ- nal. On IC5 the same pin should and think that it is feasible, al-
ing PIC code directly. For my- be left floating as shown in the though precautions would need
self, though, the type of designs diagrams, since the pin is carry- to be taken to ensure that only
I create are not suited to compi- ing information which is not re- the sun’s light (obscured or
lation from one language to an- quired in this design. In other clear) would be responded to
other in this context. There is words the diagrams in the arti- and not other, brighter, sources.
usually a large overhead of ex- cle are correct. Should I get the craving to do
tra code that is generated when Sundial Mk2, I might try this ap-
such transformations take Tom Web and Max Horsey
via the Net proach, and attempt the use of
place, accompanied by a rela- just three sensors and a bit (?)
tive reduction in processing of triangulation through the PIC
speed. OVERCAST SUNDIALS software – probably more of a
For my purposes, I need the Dear EPE, problem than I appreciate at the
compactness and optimum moment . . . Still, where’s the
John Becker’s Musical Sun- fun without the challenge?! Ra-
speed of sub-routine processing
that writing in “machine-code” dial (R.A. Evans, Readout Nov dio Bygones will, of course, con-
'99) could well have its uses. tinue to be published in its own
can achieve. Writing in PIC is as
Sun time differs from clock time
second-nature to me as writing right.
by up to a quarter of an hour. It
in any of the other several
may be of interest to know the
computer-type languages that I
difference. Human beings can’t
know and use. There are a lot of
read a sundial when the sky is
readers who are similarly adept
overcast (hence the well known
and who delight in PICs in par- sundial motto “I count only the
ticular. Projects based on PICs sunny hours’’), but electronics
have turned out to be amongst might. Even when the sun is
the most successful projects we hidden, more light must on av-
have published in our 28 (nearly erage arrive from its direction
29) years of existence! than from other parts of the sky.
An integrating light detector
CHILD GUARD QUERY might show where it is.
Dear EPE, An electronic sundial could
be remote-indicating, allowing
In Child Guard (Sep '99), lazy people like me to monitor it
IC5 and IC6 both have their ad- without going out. If it measures
dress pins connected the same light intensity it might warn you
way. However, pin 10 on IC6 is to use sun lotion. Naturally, any
connected to ground, whilst on such device should be solar
IC5 it is left floating. Is this a powered! Perhaps readers
mistake, or is the diagram cor- could suggest an appropriate
rect? high tech motto.
Martin Male P.S. Interested to see that
via the Net Radio Bygones now emanates
from Wimborne Publishing. If
you go on absorbing other
mags, your abbreviated title will
We referred Martin’s question to be as long as the original name!
the author, Tom Web, who
replied: George Short
Brighton, East Sussex, UK

Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1171
Maxfield & Montrose Interactive Inc
 with DAVID BARRINGTON Pounds (overseas readers add
1 UK Pound for postage). For
those who wish to program their
Some Component Suppliers for EPE Online Con- own PICs, the software can be
structional Articles downloaded Free from the EPE
Online Library at
Antex www.magenta2000.co.uk www.epemag.com
Web: www.antex.co.uk Maplin (UK)
Web:
CPC Preston (UK) www.greenweld.co.uk Magnetic Field Detector
Tel: +44 (0) 1772-654455 – Starter Project
Maplin (UK)
EPE Online Store and Library Web: www.maplin.co.uk Just a couple of pointers
Web: www.epemag.com regarding purchasing of
Magenta Electronics (UK) components for the Magnetic
Electromail (UK) Tel: +44 (0) 1283-565435
Tel: +44 (0) 1536-204555 Field Detector, this month's
Web: starter project. The first
ESR (UK) www.magenta2000.co.uk concerns the 100mA “center
Tel: +44 (0) 191-2514363 Microchip zero” meter … some readers
Fax: +44 (0) 191-2522296 Web: www.microchip.com may have difficulty in locating
Email: [email protected] one. The meter used in the
Web: www.esr.co.uk Rapid Electronics (UK)
Tel: +44 (0) 1206-751166 prototype came from Maplin
Farnell (UK) (code RW98G).
Tel: +44 (0) 113-263-6311 RF Solutions (UK)
If you have trouble tracking
Web: www.farnell.com Tel: +44 (0) 1273-488880
down the UGN3503U Hall effect
Web: www.rfsolution.co.uk
Gothic Crellon (UK) sensor, the above company list
Tel: +44 (0) 1743-788878 RS (Radio Spares) (UK) one as order code GX09K. They
Web: www.rswww.com also supplied the OP77G
Greenweld (UK)
Speak & Co. Ltd. precision opamp, code UL05F.
Fax: +44 (0) 1992-613020
Tel: +44 (0) 1873-811281 The alternative TL071CP low-
Email: [email protected] noise opamp should be stocked
Tel: +44 (0) 1703-236363 by most component suppliers.
Fax: +44 (0) 1703-236307
Email: [email protected] Microchip
Web: Ginormous Stopwatch –
Web: www.microchip.com
www.greenweld.co.uk Giant Display
RF Solutions (UK)
Magenta Electronics (UK) Tel: +44 (0) 1273-488880 This month we complete the
Tel: +44 (0) 1283-565435 Web: www.rfsolution.co.uk Stopwatch project with the
Web: construction of a Giant Digital
Display module. Most of the
component supply “bugs” were
PIC Micro-Probe cut away the copper tracks as ironed out last month.
necessary. The rest of the
The component listing for The high voltage 4N25
components should be readily
the PIC Micro-Probe calls for a optocoupler, code AY44, and
available.
piece of “IC holder” type the ULN2003 Darlington array,
stripboard, with a central The PIC used in this project code AD93B, are listed by
channel, devoid of copper, should be the 10MHz version. Maplin. The BD681 Darlington
running across the copper For those who want a “plug-in transistor may be hard to find,
tracks. This will cost you around and go” preprogrammed but the suggested alternative
5 UK Pounds, but for just under PIC16F84, one is available from TIP141 and TIP142 should be
2 UK Pounds you can use a Magenta Electronics for the readily available. Note the
piece of standard stripboard and inclusive price of 5.90 UK differing pinouts for the TIP
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1172
Maxfield & Montrose Interactive Inc
 6KRS 7DON
devices (Fig.2 last month). range, order code V33M. Multimeter and components, Kit
Pre-programmed PICs are Note that to make up the 20 879.
available from the author for the megohm resistor (R10) you will FML Electronics (Tel +44
sum of 10 UK Pounds each (for need two 10 megohm types. (0) 1677-425840) – Basic
either the Display module or the Once again, the “series” pads component sets.
Stopwatch) or 50 UK Pounds for have been included on the N. R. Bardwell (Tel +44 (0)
six in any combination, with free PCB. 114 255-2886) – Digital
postage to anywhere in the The last mentioned Multimeter special offer.
world. Payments should be company also supplied the
made out to Mr. N. miniature light-dependent
Stojadinovic. His email address resistor (LDR), code AZ83E,
PLEASE TAKE NOTE:
is: [email protected] and the high power warning Demister One-Shot, Nov
or write to: Mr. N. Stojadinovic, buzzer, code FK84F. Although '99
PO Box 320, Woden ACT, most of components suppliers
2606, Australia. In Fig.4, on the PCB
should be able to offer component layout diagram, the
A programmed PIC16C54 is something similar. You could, of “body” outlines of capacitors C1
also available from Magenta course, use the good old and C2 should be transposed --
Electronics for the inclusive standard ORP12 LDR if you see the photograph on page
price of 5.90 UK Pounds wish. 1033. The electrolytic, shown as
(overseas readers add 1 UK Even though the a circle, should connect to the
Pound for postage). For those semiconductors are specific IC1 pin 8 copper track (+) and
who wish to program their own versions, they should be in the common GND track (–). The
PICs, the software can be plentiful supply. The PCB is actual annotations are correct.
downloaded Free from the EPE available from the EPE Online
Online Library at Store (code 7000249) at
www.epemag.com www.epemag.com
The two printed circuit
boards are available from the Teach-In 2000
EPE Online Store (code
7000247 (Digit) and 7000248 If you have only just picked
(Port Conv.)) at up on our new Teach-In 2000
www.epemag.com series with this issue, and being
a newcomer to electronics, you
may feel a bit apprehensive
Loft Guard about ordering the various parts
Most of the components for the demonstration
called-up for the Loft Guard “exercises”. Fear not, some of
project should be readily component suppliers have put
available from your usual together component and
supplier. The only problems that hardware packs especially for
are likely to crop up may be the new series. A few more will
finding the high value resistors. be added as the series
progresses, but we do not
The single 100 megohm
expect that to be until at least
resistor (R7) was only found
part seven.
listed under the “cermet film”
range stocked by Electromail or To date, participating
RS (quote code 158-222). As advertisers are as follows and
the article points out, you could readers are, advised to contact
use three 33 megohm resistors them for more details:
(in series); the PCB is also ESR Electronic
designed to accept these. This Components – Hardware/Tools
resistor (33M) came from the and Components Pack.
Maplin “high voltage” metal film
Magenta Electronics –
Copyright © 1999 Wimborne Publishing Ltd and EPE Online, December 1999 - www.epemag.com - 1173
Maxfield & Montrose Interactive Inc