This is only a preview of the May 2012 issue of Silicon Chip. You can view 27 of the 112 pages in the full issue, including the advertisments. For full access, purchase the issue for $10.00 or subscribe for access to the latest issues. Items relevant to "PIC/AVR Programming Adaptor Board; Pt.1":
Items relevant to "High-Temperature Thermometer/Thermostat":
Items relevant to "1.5kW Induction Motor Speed Controller, Pt.2":
Items relevant to "SemTest Discrete Semiconductor Test Set; Pt.3":
Items relevant to "Ultra-LD Mk.3 135W/Channel Stereo Amplifier, Pt.3":
Purchase a printed copy of this issue for $10.00. |
siliconchip.com.au
May 2012 1
ED MA
IT Y
IO
N
Pr
ice
WANT A FREE COPY OF OUR
va
lid
2012 CATALOGUE?
un
til
23
Place an order of $30 or more via our Techstore
/0
5/
website and type "FREE CATALOGUE" in the
20
12
comment box as you check-out.
MEGA
Offer valid until 31/5/2012.
80mm Silent Hydrodynamic
Bearing Case Fan
MAY
Designed to provide additional airflow without
increasing noise levels. Features balanced blades,
long-life hydro-dynamic bearing and ultra-low noise
making these fans perfect for a home theatre, PC, or
ventilating the amplifier
wedged into your
entertainment unit.
• 12VDC
80mm YX-2570 $19.95
90mm YX-2572 $24.95
120mm YX-2574 $28.95
IP67 True RMS Autoranging
Cat IV DMM with Wireless USB
SMD Electrolytic Capacitors
• Ideal for prototyping
• Mixed - pack of 50
• 20 values
RE-5959
NEW
1295
$
SMD Tantalum Capacitors
• Ideal for prototyping
• Mixed - pack of 30
• 20 values
RZ-6618
A quality true RMS multimeter with a wireless USB
computer interface and includes logging
software which allows computer based live
data whilst keeping your computer
completely isolated and protected. Double
moulded housing and IP67 rated.
• Non-contact voltage indicator, data hold
• Backlit, auto off
• Diode test and audible continuity
• Cat IV, 600V, 4000 count
• 10A current range
• Size: 170(L) x
NEW
79(W) x 50(H)mm
$
00
QM-1571
Goot Soldering Station
NEW
1295
$
Retractable 3m Alligator
Test Lead Set
Set of 3 heavy duty red,
black, and green leads
on a retractable 3m
reel, terminated with
insulated alligator clips.
109
1995
$
SAVE $5
• Reel size:
152(Dia.) x 20(W)mm
WT-5334 Was $24.95
This soldering station is a precision, Japanese
manufactured instrument with excellent
temperature stability and anti-static characteristics.
It has a digital temperature adjustment from 200 to
480°C at 65W and a
lightweight
soldering pencil.
See website or
catalogue
for full
specifications.
• Power: 70W
• Operating voltage: 24VAC
• Control size: 146(L) x
115(W) x 98(H)mm
TS-1440 Was $299.00
KIT BACK CATALOGUE
If you can’t find the kit you are
looking for, try the Jaycar Kit Back
Catalogue. Our central warehouse
keeps a quantity of older and
slow-moving kits that can no longer
be held in stores. A list of kits can
be found on our website. Just
Attention:
search for “kit back catalogue’.
Kit Builders
To order call 1800 022 888
24900
$
SAVE $50
25MHz Dual Trace Digital
Storage Oscilloscope
An ideal DSO for the advanced hobby user or
technician and is particularly suited to audio work.
Full data storage capabilities and USB interface so
you can store
traces on a flash
drive. Includes 2
x 10:1 probes,
EasyScope
software, and
USB cable.
• Channels: 2
• Input impedance: 1Mohm
• Size: 310(W) x
150(H) x 130(D)mm
QC-1932
59900
$
Stereo Compressor Kit
Refer: SC Magazine Jan 2012
Compressors are useful in eliminating the extreme
sound levels during TV ads, "pops" from
microphones when people speak or
bump / drop them, levelling signals
when singers or guitarist vary
their level, etc. Kit includes
PCB, processed case and
electronic components for
12VDC operation. 12VDC plug
pack required - use MP-3147.
• PCB: 118 x 102mm
KC-5507
5995
$
Semiconductor
Component Analyser
Intelligent semiconductor analyser that offers simple
identification and testing of a variety of 2 or 3-pin
devices. Type and lead identification as well as forward
voltage, test current and other
parameters for transistors.
• Automatic pinout
identification
• Gain and leakage current
measurement for BJTs
• Silicon and germanium
detection for BJTs
• Forward voltage and test current
• Size: 100(W) x 71(H) x 27(D)mm
QT-2216 Was $99.00
4900
$
SAVE $50
AVAILABLE LATE
MAY!
Jaycar CATALOGUE
on a Flash Drive
www.jaycar.com.au
Contents
SILICON
CHIP
www.siliconchip.com.au
Vol.25, No.5; May 2012
Features
14 The Australian Synchrotron
Located close to Monash University, this impressive circular, megavoltage
machine enables investigators to determine the structure and composition
of various materials with extremely high detail – by Dr David Maddison
22 Getting The Most From ADSL
PIC/AVR Programming Adaptor
Board, Pt.1 – Page 30.
30.
Most of us owe our high-speed Internet access to the freakish technology
of ADSL. But what exactly is it, how does it work and how do you solve slow
speed problems? – by Alan Ford
Pro jects To Build
30 PIC/AVR Programming Adaptor Board; Pt.1
Do you often need to program microcontrollers? This board, combined with an
In-Circuit Serial Programmer (ICSP), allows you to quickly program most 8-bit
& 16-bit PIC microcontrollers as well as 8-bit Atmel AVRs – by Nicholas Vinen
40 High-Temperature Thermometer/Thermostat
Need to measure or control temperature over a very wide range? This compact
unit hooks up to a K-type thermocouple, can measure temperatures from -50°C
to +1200°C and drives a relay for thermostatic control – by John Clarke
High-Temperature Thermometer/
Thermostat – Page 40.
68 1.5kW Induction Motor Speed Controller, Pt.2
Second article details the construction and testing and gives guidelines on its
use – by Andrew Levido
84 SemTest Discrete Semiconductor Test Set; Pt.3
Follow this article to build and test this useful unit. We also describe how to
fit a crowbar circuit to quickly discharge the HT after making high-voltage
measurements – by Jim Rowe
96 Ultra-LD Mk.3 135W/Channel Stereo Amplifier, Pt.3
Third article gives the specifications & performance – by Nicholas Vinen
Special Columns
61 Serviceman’s Log
The dodgy, dangerous, home-made stereo amplifier – by the Serviceman
Building the 1.5kW Induction Motor
Speed Controller, Pt.2 – Page 68.
78 Circuit Notebook
(1) PICAXE 433MHz Data Transmitter & Receiver; (2) Electronic Ballast
For Fluorescent Light Fittings; (3) Air-Compressor Controller For A SandBlaster; (4) Motor Protector Uses Missing Pulse Detector; (5) Maximite-Based
Ultrasonic Rangefinder
98 Vintage Radio
Breville 730 dual-wave 5-valve receiver – by Rodney Champness
Departments
2 Publisher’s Letter
4 Mailbag
66 Product Showcase
siliconchip.com.au
105 Order Form
106 Ask Silicon Chip
111 Market Centre
Building The SemTest Discrete
Semiconductor Test Set – Page 84.
May 2012 1
SILICON
SILIC
CHIP
www.siliconchip.com.au
Publisher & Editor-in-Chief
Leo Simpson, B.Bus., FAICD
Production Manager
Greg Swain, B.Sc. (Hons.)
Technical Editor
John Clarke, B.E.(Elec.)
Technical Staff
Ross Tester
Jim Rowe, B.A., B.Sc
Nicholas Vinen
Photography
Ross Tester
Reader Services
Ann Morris
Advertising Enquiries
Glyn Smith
Phone (02) 9939 3295
Mobile 0431 792 293
glyn<at>siliconchip.com.au
Regular Contributors
Brendan Akhurst
Rodney Champness, VK3UG
Kevin Poulter
Stan Swan
Dave Thompson
SILICON CHIP is published 12 times
a year by Silicon Chip Publications
Pty Ltd. ACN 003 205 490. ABN 49
003 205 490. All material is copyright ©. No part of this publication
may be reproduced without the written consent of the publisher.
Printing: Hannanprint, Noble Park,
Victoria.
Distribution: Network Distribution
Company.
Subscription rates: $97.50 per year
in Australia. For overseas rates, see
the order form in this issue.
Editorial office:
Unit 1, 234 Harbord Rd,
Brookvale, NSW 2100.
Postal address: PO Box 139,
Collaroy Beach, NSW 2097.
Phone (02) 9939 3295.
Fax (02) 9939 2648.
E-mail: silicon<at>siliconchip.com.au
Publisher’s Letter
Sound levels a sore point
on TV and elsewhere
It is fair to say that my recent Publisher’s Letters on
the topics of loud TV commercials and excessive sound
levels in cinemas and theatres have triggered a lively
response. We have more letters on this topic in the
Mailbag pages this month and since they are still coming
in, there will undoubtedly be a few more next month.
I must admit to be being baffled by some of the responses to the editorial about sound levels in theatres,
coming from the people who actually do the work. One response was along the
lines that before anyone should think of complaining to theatre management,
they should take a series of measurements around the theatre or whatever
venue is involved.
How unrealistic! One does not go to a theatre, on the one hand expecting to
enjoy the entertainment, and on the other expecting to be blasted and therefore
also carrying a sound level meter so one can dart around the theatre making
measurements. As my daughters would say, “Get real!” or worse, “Get a life!”.
If one was so concerned about being blasted, that would be substantial disincentive to buying a ticket in the first place.
It seems to me that whoever is responsible for the sound levels at cinemas
and elsewhere, whether it is the producer, management, disc jockey or whoever, simply does not realise that if people have to shout to communicate to
the person next to them, then the sound is just too loud; no test equipment
is required. That rule of thumb has been quoted by hearing experts over the
decades. That it seems to be largely ignored by people who should know better is a paradox.
Mind you it also seems to me that many people are simply inured to excessive sound levels and are too timid to even think about complaining. And
there is another group who are obviously well on the way to going deaf and
probably need the wick wound up a bit. But still on the same theme, if people
are moderately to severely deaf then it is also true that they are less able to
cope with excessive sound levels; in effect, they can’t hear the soft bits and
can’t stand the really loud bits. And inevitably there are some people at public
performances who are too drunk or stupid to care.
We already know that a substantial proportion of the population is deaf and
a lot of that deafness is due to being exposed to excessive sound levels. It is
because so many people are deaf that most public venues also provide hearing
loops so that people with hearing aids can listen to the performance. That is
an even bigger paradox, isn’t it? Public venues provide for deaf people and
then act as though the rest of the population should also be rendered deaf!
So if we already know that a significant portion of the population is already
deaf and even more people are likely to be deaf in the future, doesn’t that tell us
something? If the relevant authorities are ineffective at protecting the public’s
hearing, then individuals must act on their own behalf.
For my part, in the future I will always take earplugs with me whenever
I go to a venue where sound levels are likely to be high. I do the same thing
when I use noisy power tools, just as I wear eye protection. I suggest that you
do the same.
Leo Simpson
ISSN 1030-2662
Recommended and maximum price only.
2 Silicon Chip
siliconchip.com.au
SIOMAR
Battery Engineering
IRON PHOSPHATE
Lithium Iron Phosphate (LFP) is a special kind of rechargeable lithium battery that addresses the 4 major issues with current lithium
technologies: Safety, Life, Power, and Environmental Friendliness. The chemistry has similar charge requirements to lead acid
batteries and can therefore be more easily integrated into many lead acid applications than other chemistries.
SAFE
Lithium Iron Phosphate technology is an inherently safe battery chemistry "the safest lithium battery on the market" as shown
repeatedly by independent data generated by the Department of Energy, UL, UN, Sandia National Labs and other agencies. The cells
pass UN/DOT (UN38.3 transport safety testing) requirements with NO circuit board protection. This is not possible with traditional
lithium chemistries.
LONG LASTING
Traditional Li-ion chemistries already typically last 5 years or 500 charge / discharge cycles before the AH capacity tapers off to 80%
of what it was when the battery was new. This is a league above lead acid batteries which age dramatically with increasing
temperature and it is even better than NiCd and NiMH. Lithium iron phosphate is better still, achieving up to 2000 charge discharge
cycles.
POWERFUL
Cells can be optimized for long life, low or high temperature operation or high current without dramatically affecting service life. The
LFP300HPS (90AH) cell can deliver over 4,200 Amps. Lithium phosphate cells can now meet even the most demanding application
requirements — from starting a locomotive to powering an F1 racer, or cold starting a tractor at -20F.
CLEAN
Lithium Iron Phosphate can replace outdated battery chemistries like Lead-acid, NiCd, NiMH. They are designed to offer extremely
long cycle life, high energy capacity, outstanding power performance, and quick charge times, while simultaneously being
environmentally friendly.
These batteries use no harmful heavy metals and can be recycled. Use of phosphates in the system architecture reduces
environmental concerns in all stages of the battery’s life cycle. And given their long lasting nature, wearing one of these batteries out
takes a lot longer than most other batteries in the first place.
3.2V
50AH
3.2V
90AH
12V
10AH
24V
19AH
For more information, contact
SIOMAR BATTERY ENGINEERING
Phone (08) 9302 5444 or email mark<at>siomar.com
www.batterybook.com
Siomar Batteries design and custom make portable Power Solutions
siliconchip.com.au
May 2012 3
MAILBAG
Letters and emails should contain complete name, address and daytime phone number. Letters to
the Editor are submitted on the condition that Silicon Chip Publications Pty Ltd may edit and has the
right to reproduce in electronic form and communicate these letters. This also applies to submissions
to “Ask SILICON CHIP” and “Circuit Notebook”.
Enthused over the
Atwater Kent story
This is just a short note to congratulate you on the amazing article in the
March 2012 issue of SILICON CHIP on
Atwater Kent. The words create excellent stories of facts and history of the
man and his radios, and the restored
monochrome photos are truly magnificent, with their high contrast and
level of detail. You do manage to keep
raising the bar!
By the way, are the radios shown on
pages 98/99 actually here in Australia?
The photos are beautifully done. Geez,
I’d like a few of those radios if I could
afford them!
I’ve never before seen such a beautiful display of A-K radios in print. Great
work on the part of Kevin Poulter.
Graeme Dennes,
Bunyip, Vic.
Comment: Kevin took the colour photos in question. All the radios are from
a private collection in Melbourne.
Doubts raised about the Induction
Motor Speed Controller
With respect to the comments on
the Induction Motor Speed Controller
in the Publisher’s Letter, (April 2012)
and the admission that the 2kW Sinewave Inverter project published in the
October 1992 to February 1993 issues
was more complex than the VSD, I am
Creating a scene
in response to bedlam
At last! Someone with a voice has
spoken out on my pet hate; excessive
noise in theatres and live shows. For
years my wife and I have complained
bitterly to club managers about noisy
shows in their auditoriums. We go to
clubs to enjoy good company, chat
with friends, enjoy good food and
if possible enjoy the entertainment.
The incredible din created by
most entertainers make all of the
foregoing impossible! The most
puzzling aspect of this whole affair is
that presumably sound engineers are
4 Silicon Chip
still unsure why the delay.
Also, the third harmonic injection
is not new, so while Mr Levido has
done a great job, it is not his idea!
But my main point is, with no dv/dt
output filter, the earth currents will
not be sinusoidal, even if the phaseto-phase current is. Therefore, what
are you going to do about the motor
bearing failures that will result? And
what about the resulting possible Earth
Leakage Breaker tripping problem?
Note that even an output sinewave
filter will not eliminate nasty earth
currents (as I have discovered via an
internet Inverter/VSD forum).
Ray Hamono,
Bayswater Nth, Vic.
Comment: in response to your first
question, the delay in producing the
design was mainly due to our conservative approach.
We directed your other questions to
the designer, Andrew Levido, for his
comments, as follows:
The reader is correct in stating that
the injection of third harmonic is not
new. In fact the author first saw this
technique used 25 years ago and it
was not so new even then! The article made no claims that this was an
original idea.
The high level dv/dt present in the
output of induction motor speed controllers can cause a small currents to
trained with formal qualifications
and yet they do not understand the
basic operation of the human ear.
That is, at some point just before
pain threshold, sound becomes so
distorted that it is unintelligible.
Add to this the fact that this
level of noise would invoke legislative punishment in any factory in
Australia. As well as complaining
loudly, my wife and I also remonstrate visually by twisting the ends
of a full size serviettes and inserting
the twisted ends into our ears!
This has a twofold effect in that
it protects our ears as well as sends
flow to earth through the capacitance
between the stator windings and the
earthed frame of the motor. These
currents will be spiky, corresponding
with the switching edges of the PWM,
and not sinusoidal, but they will be
symmetrical. For small motors such
as those used with this project, the
magnitude of these currents will be
quite small.
There should not be any problems
with domestic earth leakage circuit
breakers which typically trip when
there is an imbalance in current between the Active and Neutral lines.
Despite these small current spikes, Active and Neutral currents will be balanced, so should not cause problems
with earth leakage circuit breakers.
The reader also correctly suggests
that AC drives have been implicated in
early bearing failure in motors under
some circumstances. Any potential on
the rotor with respect to earth will force
a current to flow via the path of least
resistance. In some cases, the potential
is high enough to break down the thin
insulating coating of oil on the bearings causing pitting or frosting, leading
to their premature failure.
This is more likely to occur in very
large motors where the capacitance
between stator and rotor is high but
has reportedly been seen in motors as
a very clear signal to all around us
that this is not tolerable. Hopefully
it also sends a message to those on
stage as well as the sound engineers
and club managers.
Sadly, we have had no noticeable
effect on any of these people that we
are aware of. However, entertainment in clubs is dying and perhaps
that is in protest against a very unpleasant environment.
Keep up the good work Leo and
try to get the motion picture industry
to fall into line if possible.
Bob Young,
Riverwood, NSW.
siliconchip.com.au
Background music
is often too loud
small as 10kW. I are not aware of it being a problem with motors in the 2kW
range where the capacitance between
stator and rotor is quite small.
The problem can be eliminated by
either insulating the steel bearings,
using ceramic bearings or by shorting
the rotor to earth via some lowerresistance path. Sometimes this occurs
naturally through the load, such as in
water pumps. In other cases, the shaft
can be grounded via conductive plastic
brushes. In other cases, the bearings
are packed with conductive grease.
Adding a dv/dt filter between the AC
drive and the motor can also help to
some extent.
Given the size and usage of the motors that will be used with this project,
it is unlikely that this problem will be
an issue in the domestic setting.
Background to the
18-bit DAC circuit
This is a response to the letter entitled “18-bit DAC Circuit Is Not Valid”
by Phil Denniss, on pages 10 & 12 of
the March 2012 issue. An explanation
siliconchip.com.au
I read Leo Simpson’s comments
on excessive sound levels in theatres
and cinemas with great interest.
But why stop there? How many of
us have we watched a TV program
or DVD and found the audio poorly
balanced with background music or
sound effects so loud that speech is
partially or completely obscured?
Poor quality audio has been a bugbear of mine for years and while my
experience is more with the home
viewing environment, it amounts to
the same thing. Enjoyment of program material is being significantly
impaired because of inept or just
of the genesis of the circuit will hopefully assist understanding it.
I started by looking at the Lab-Standard 16-Bit Digital Potentiometer from
the July 2010 issue of SILICON CHIP. It
seemed to me that the specified 0.1%
tolerance (1 in 1000) resistors are not in
fact sufficiently accurate to guarantee
the specified accuracy of the output
voltage (1mV in 10V or 1 in 10,000),
although statistically most units built
plain lazy audio mixing methods.
Some programs seem to take this
to extremes, with a flea farting two
streets away being more audible than
the dialogue between actors right
in front of the camera. Even interviews in news broadcasts can suffer
when background noise or chatter
overwhelms what the reporter and
interviewee are saying.
Poor microphone selection or
placement and sound technicians
obsessed with giving background
ambience and incidental music
precedence over speech seems to be
the norm now.
Chris Loader,
Brisbane, Qld.
would be sufficiently accurate.
This set me wondering if the digital
potentiometer could be redesigned to
use lower-tolerance resistors by using
some kind of on-board calibration.
This lead to a current-mode DAC
design that used non-binary weights.
The resistor values were assigned so
that there was a guarantee of overlap
in output voltages at the major transitions (eg, input going from 00001111
May 2012 5
DYNE
INDUSTRIES PTY LTD
Now manufacturing the
original ILP Unirange
Toroidal Transformer
- In stock from 15VA to 1000VA
- Virtually anything made to order!
- Transformers and Chokes
with Ferrite, Powdered Iron
GOSS and Metglas cores
- Current & Potential Transformers
DYNE Industries Pty Ltd
Ph: (03) 9720 7233 Fax: (03) 9720 7551
email: sales<at>dyne.com.au
web: www.dyne.com.au
Mailbag: continued
to 00010000) regardless of variation
within the resistors’ tolerance. A
precise and linear ADC, probably a
delta-sigma ADC referenced to an accurate voltage source, would be used
to calibrate the DAC.
Although I realised that this idea
would have limited application, the
concept of guaranteeing full coverage of the output voltage range using
imprecise resistors with non-binary
weights and the use of an ADC with
feedback to create a DAC, was novel,
so I submitted it to Circuit Notebook.
I limited it to 11-bit operation because
the current in the highest order resistor would have become excessive for
more bits.
Shortly after I submitted this idea
I realised that I could split the DAC
to reduce the resistor currents and
therefore allow operation with more
than 11 bits. I also submitted this idea
to Circuit Notebook but I omitted most
of the details that were in the previous
6 Silicon Chip
Praise for
Publisher’s Letters
I would like to congratulate Leo
Simpson for the continued great
layout, depth and range of articles
in the magazine plus his wisdom
as displayed in Publisher’s Letter.
I have to say I can’t remember ever
disagreeing with his sentiments and
expert commentary.
I particularly appreciate the fact
that he goes outside electronics from
time to time in order to bring a touch
of sanity into populist politicallydriven issues like carbon trading.
His comments on incandescent
lamps and solar panels are memorable too. I was listening to an “expert”
on Radio National awhile ago who
stated that it takes 57 years to pay
off a solar installation if the carbon
emissions of manufacturing, installation and transport etc are taken
into account. By then the expensive
batteries may have been replaced
three or four times and the panels
will be completely outdated and
most likely replaced years earlier
than their potential lifespan.
For all of this, there would have
been less of that nasty carbon
dioxide in the atmosphere if the
solar installations had never been
installed at all!
Having said that, the notion of
“free power” from the sun is very
alluring but the numbers just don’t
add up without massive crosssubsidisation, which means people
paying those subsidies have to use
more energy and emit more CO2 to
submission, assuming that the first
item would be published. However,
the editor decided to publish the second submission, not the first.
The January 2012 Circuit Notebook
circuit is designed for 1% tolerance
resistors (unfortunately the tolerance was not marked on the circuit
diagram).
As it stands, the circuit has two
flaws that affect its accuracy. First,
noise in the higher-value resistors
could potentially lead to output voltage fluctuations that exceed the lowest
bit voltage. However, the mean value
of this noise is zero so if the DAC is
to be used simply as a variable preci-
earn the money to subsidise the solar
industry.
When more efficient panels are
developed, perhaps then renewable
energy will be practical.
I worked in the coal industry for
20 years and now see a massive
increase in the industry with total
government support. From that, one
can only conclude that the people
we elect every few years, who approved these projects, are complete
hypocrites when it comes to issues
like carbon dioxide emissions and
carbon trading.
Geoff Mattick,
Gulgong, NSW.
Comment: the cost of solar panels
has now come down to the point
where we believe investment by
businesses and domestic consumers
can now be viable, even without the
benefit of green subsidies or beneficial grid-feed tariffs.
In fact, the introduction of smart
meters makes such an investment
more worthwhile, as the owner of
such an installation can then avoid
peak tariffs and also ensure power
reliability during blackouts. This
would assume that the installation
was battery backed up to ensure
power was available during blackouts at any time.
We do not accept the long payoff
times quoted for solar panel manufacture. In practice, the dollar cost of
the solar panel more or less reflects
the actual energy input to manufacture it. If not, the manufacturers
would go broke.
sion voltage reference the noise could
simply be filtered out.
Second, variations in the values of
the resistors due to thermal effects
could exceed the lowest bit voltage.
Thermal variations due to resistor current can be eliminated by modifying
the circuit so that instead of the left
ends of the weighting resistors being
switched between 0V and VDD, the
left ends are permanently connected
to VDD and the right ends are switched
between 0V and the inverting input of
op amp IC1. The current, and therefore
self-heating, in each resistor is then
constant regardless of whether it is in
use or not. In addition, to ameliorate
siliconchip.com.au
siliconchip.com.au
May 2012 7
NEW 30AMP 12V/ 24 DC PWM with
SOFT START FEATURE
The MXA087 PWM MODULE
is a new generation
microprocessor - based
DC Motor Speed
Controller for loads
up to 30Amps at
Fully
either 12V
Assembled
or 24VDC
and Tested
Value!!!$50.22 inc. GST Plus $7.50 Pack & Post
NEW CLOCK /TIMER KIT with
20 ON/OFF EVENT CAPABILITY
The FK949 12V Clock Timer kit displays current time
in 24hour mode in hours and minutes
or minutes and seconds. Up to
20 ON/OFF events may be
preset to drive a 500watt load
via its on-board relay
Build in 30
minutes!
Only $33.26 inc. GST
add $7.50 P & P
For more details or to buy
on-line see us at:
www.kitstop.com.au
P.O. Box 5422 Clayton Vic.3168
Tel:0432 502 755
Mailbag: continued
VU meters tell the story
In your March Issue, Neil Davis and Brian Wallace
both make interesting comments about the loudness of
commercials versus program material. I would make the
broad statement that almost all audio program material
broadcast these days has undergone a degree of compression. Only a few fine music stations claim not to.
Compression is carried out either in the initial recording and editing process, the transmission path or both.
I still have (after many years) one of David Tillbrook’s
the effect of variations in ambient
temperature, the entire circuit could be
enclosed in a temperature-controlled
oven.
Having made the second submission, I discovered that a team at Linear
headed by the late, great Jim Williams,
had been thinking along similar lines.
In Linear Application Note AN86, A
Standards Lab Grade 20-Bit DAC with
0.1ppm/°C Drift (http://cds.linear.com/
docs/Application Note/an86f.pdf) they
8 Silicon Chip
control units with LED VU metering. In his design,
David built simultaneous peak and average readings.
Putting this across a broadcast or telecast audio signal
shows that most material has little difference between
peak and average readings. This is evidence of serious
compression.
Then in order to have the commercial’s audio compete in the attention-grabbing stakes, often the VU LEDs
hardly move at all, except during pauses!
The following story demonstrates how extreme signal
processing can get. When I worked for a National Broadcaster in Canberra (again, years ago, before satellites), we
often recorded and broadcast a national program called
“Notes On The News”. As it left our studios, it was put
through a CBS “Audimax” compandor. The signal also
went through a limiter at the transmitter.
One day, I had a visit from a technician who lived
in Perth. He complained that when programs like ours
reached Perth, the VU meters peaked on 0dbm and stayed
there! He told me that after a bit of research on his travels
over to us in the east, he had found that the signal leaving
us was companded again in Sydney prior to passing on
to Melbourne where it copped another dose.
In Adelaide, there were, he told me, two passes; one
through Master Control as the signal went to air but
because the signal to Perth had to be delayed before
crossing the Nullabor, that signal was companded again
as it left the Adelaide Record Suite. No wonder the VU
meters did not move!
In our studio operations, we were required to record or
broadcast music at a level somewhat below that reached
by the voice of the announcer/presenter attached to that
program. This was so both would have a similar apparent loudness and therefore have transitions sounding
pleasant to the ear of the listener. I still use that principle
in my current life in Community Radio and have been
commended for it.
It seems to me that if a program and its commercials
have audio that disrupts the audience’s attention, then
the presentation folk in that TV or radio station have
annoyed and therefore lost that listener’s attention.
Surely, advertisers must be aware of this. If audience
attention is diverted by audio levels, then he/she cannot
be focusing on the commercial message.
Bruce Bowman.
Ainslie. ACT.
employ two 16-bit DACs with 8-bits
of overlap to effectively form a 24-bit
DAC. This DAC is used in a feedback
loop with a 24-bit delta-sigma ADC
to form a precise, stable DAC whose
output is referenced to a single fixed
precision voltage source.
I agree that modern components are
inexpensive and precise enough that
all this effort to design and build a
discrete circuit is unnecessary but the
circuit as published does have a cou-
ple of interesting ideas that someone
might some day find useful.
Andrew Partridge,
Toowoomba East, Qld.
Loud TV
commercials
I refer to the Publisher’s Letter in the
February 2012 edition, entitled “Loud
Television Commercials Will Continue
To Be Annoying”.
I enjoyed the whole article and
siliconchip.com.au
Comments on
Maximite triggering
found it to be an accurate reflection of
what is going on out there, with TV and
other volumes all over the place and
the ongoing onslaught of media advertising. Adverts continue to mostly be
annoying when you just want to get
back to the main program you have
been watching; adverts waste so much
time. But everyone is used to those
annoying commercials so they have
devised their different approaches to
circumvent them, as Leo pointed out
in his Publisher’s Letter.
I just basically ignore advertisements on TV and will go do something
else while around seven or so adverts
screen in each commercial break.
When it comes to the volume problem
I like everyone else am constantly
changing the volume setting on the TV,
depending on whether I am watching
TV (volume varies between stations)
or watching a DVD or an old VHS tape.
The volume varies again for a DVD and
a VHS tape and when I play an audio
CD on the DVD player.
You presented a new Stereo Compressor project in the January 2012
edition. I can see how the audio
outputs of a DVD player, a VHS tape
player and a cassette deck could be fed
to the TV via the stereo compressor.
But I can’t see how you can use the
stereo compressor with a TV itself;
the sound feeds directly to the speakers inside the TV. Perhaps the sound
can be fed out to external speakers
via the stereo compressor? If so then
how do we switch off the internal TV
speakers? Can we use the mute func-
Thank you for publishing my
contribution in the Circuit Notebook
pages of the March 2012 issue. However, I have some concerns regarding
the Editor’s Note at the conclusion
of the item.
If, as suggested, pin 1 of the Maximite is connected to pin 2 (Q-bar) of
IC1a, it is certainly possible to trigger
the interrupt this way but the trigger
event should occur on Q-bar going
low to high which marks the end of
the gating period. So the interrupt
should be a low-to-high transition
and not a high-to-low as suggested
in the note.
In addition, pin 1 of the Maximite
performs two different functions.
The first is in the initialisation
tion or does this cut off the TV’s audio
completely or does it only mute the
internal speakers? The TV audio may
always be available at the various RCA
outputs regardless of whether we use
the mute function or not?
I agree with Leo that people are
watching less TV these days as they
want some quality with their TV viewing experience. The ads and volume
problems do not add anything positive to the TV viewing experience so
people will inevitably vote with their
feet and their remote controls (or the
off buttons on the TV or those at the
wall socket).
Mark Eastaugh,
Armadale, WA.
Comment: the stereo compressor project is an effective solution only if the
PICAXE
routine where it is defined as an
output (line 10) and used to ensure
that IC1a is in the correct reset state
initially. After this initialisation, pin
1 is then redefined to be the interrupt
pin (line 30).
Clearly, if pin 1 of the Maximite
is disconnected from pin 4 of IC1a
and reconnected to pin 2 instead,
it cannot perform this initialisation
function and some other circuitry
must now be added to pin 4 to ensure
the reset state is attained at power
on. Any additional circuitry added
to pin 4 runs the risk of it interfering
with the primary task of the circuit
and that is to measure the charging
time of a 4.7nF capacitor through a
100kΩ resistor.
Jack Holliday,
Nathan, Qld.
TV is being used in a home theatre
system, ie, with external amplifier and
speakers. Some up-market digital TV
sets do have in-built “volume levellers” which go some way to coping
with differing audio levels from different TV stations. However, the stereo
compressor has little real effect on TV
commercials which have high levels of
audio processing. In those cases, the
only cure is to use the mute button on
the remote control.
Sound levels in cinemas
are excessive
The sound level in cinemas and
theatres is now grossly excessive.
Consider the sound system at the
Astor in Windsor, Melbourne, an
elegant Art Deco cinema built in the
Large Range of PICAXE Chips
Starter packs, kits and accessories
●PICAXE 08M, 14M, 18M, 18M2, 18X, 20M, 20X2, 28X1, 28X2 & 40X1 Chips
●Books
●Kits
●Starter Packs
●Software
●Project Boards
●Accessories
●Experimenter Kits
For the full PICAXE range, pricing and to buy now online, visit
www.wiltronics.com.au
Ph: (03) 53342513
Email: sales<at>wiltronics.com.au
38 Years Quality Service
IN STOCK NOW!
PICAXE is a registered Trademark of Microchip Technology Inc.
siliconchip.com.au
May 2012 9
Mailbag: continued
Banning incandescent lamps
was a silly idea
Because I was looking for a highpower motor controller, I burrowed
my way back to the April 2007 edition of SILICON CHIP and found the
article I wanted. I also happened to
re-read your Publisher’s Letter in the
same issue, entitled “Banning incandescent lamps will have a negligible
effect on greenhouse gasses”.
After five years of this nonsense
now, I wonder if the Federal Government could tell us exactly how
much “greenhouse gas” has been
saved on the planet as a whole
1930s. It now boasts a total of 11 kilowatts of sound, including 3.5 kilowatts
fed into eight 18-inch subwoofers.
On display in the foyer is the original Western Electric sound system.
State of the art for 1929, it comprises
a six-foot high 19-inch rack, with an
impressive output of 15 watts!
The original theatre seated 1700
people and I’m sure they had no difficulty hearing the sound.
James Goding,
Carlton North, Vic.
Good customer relations
from Telstra
SILICON CHIP printed a letter of mine
in the March 2012 issue which mentioned my frustration in dealing with
Telstra over our poor mobile phone
coverage.
I have a lifelong experience in radio
and microwave communications and
but more importantly, tell us how
much mercury has been bulldozed
into landfill sites when the “dead”
fluorescent lamps have been thrown
out and not recycled.
I guess it goes to prove that the Liberals (this was Malcolm Turnbull’s
bright idea) are just as idiotic as
Labor in the global warming – sorry,
“climate change” – scaremongering.
The only up-side to the Carbon Tax
that is going to save the world is that
it should ensure the total demise of
Labor at the next election.
John Brown,
Bibra Lake, WA.
could not understand why a mobile
phone tower only 2.2km away did
not service our suburb. After two
years, I and the TIO gave up. Telstra
provided no logical reason as to why
we had such poor phone coverage.
The team leaders and case managers
had no technical knowledge of mobile
phone systems, even though they were
handling my complaint.
Two years after this I decided to
write directly to the CEO of Telstra,
David Thodey, in the hope that he
may just read my letter. Well he did,
or at least one of his direct office staff
did and three days after sending the
letter I received a phone call from a
Telstra engineer who talked my language; radio.
In two sentences I now knew why
a 30-metre high phone tower 2.2km
away, with large gain antennas pointing at our house and with minimal
Some key features
Mixed Signal Oscilloscope
+ Signal Generator
terrain problems did not get to our
house. It was designed not to!
The antennas on the tower pointing
at us are phased so as to produce 9° of
downwards tilt. This limited the signal
from the phone tower to less than 1km.
It sure works, as you have to drive
to under 1km before suddenly your
phone works with full signal strength.
Beyond this distance, if you removed
a few trees, you could see the tower
but still get little or no signal.
The reason for the downwards tilt is
that we live in the hills east of Perth’s
metropolitan area and as the phone
tower is up high it interferes with
other cells on the coastal plane, so the
signal is prevented from spilling over
onto the wider area by electronically
down tilting the antennas. We are in
that direction.
At long last my decades of working in the radio propagation domain,
along with having an amateur license,
were vindicated. There was an understandable reason as to why we had no
mobile phone coverage, along with
all the other neighbours in the area.
There really was a reality about radio
waves after all.
Telstra fixed the problem for us in
two weeks! They came out to our house
and installed a fairly new repeating
system, that now gives us full signal
strength on our mobile phones, in and
around our house.
I’m not 100% sure how it works
but it appears to be an on-frequency
repeater. A 15-element Yagi antenna
on the roof points at the best phone
tower signal (not the one 2.2km away)
and this full-scale signal is connected
to a box that is WiFi linked to another
box further away in the house and this
Mixed signal with protocol decoding
+ Windows and Linux
+ 10ns resolution on 40 msec display
+ Disk save at 1.5 MSPS, to 500 GSamples.
+ Spectrum analysis – see 50Hz in 50 MHz.
+ Driver software for Visual C#, C++, Basic,
Labview, Delphi and C Builder with
examples.
+ Protocol decoding – SPI, I2C and UART
+ Isolated Ethernet option
www.cleverscope.com
10 Silicon Chip
siliconchip.com.au
siliconchip.com.au
May 2012 11
Mailbag: continued
Helping to put you in Control
Control Equipment
Serial Server
The SE5001 is a gateway for Ethernet (TCP/
IP) and RS232/RS485/
RS422 serial communications. It allows almost
any serial device to be connected to a
new or existing Ethernet network .
ATO-101 $129+GST
Solid State Relay
A 2amp 240VAC solid
state relay with a 6-24VDC
input. Features zero crossing and screw terminals.
KTD-273 $19.00+GST
RHT Controller.
Easy to configure it
features 3 relays for
temperature, humidity
and alarm. Comes with
a 3m long probe and RS485 comms
CET-112 $209.00+GST
Anemometer Datalogger We have upgraded
our anemometer monitoring and alarm card to
now include a datalogger
to record wind speed and direction
KTA-250-AL $349+GST
VLN3000 Breakout Board
Features an ambient light
sensor and an IR proximity
sensor with a 20cm range.
A I2C interface allows easy
interfacing with a microcontroller
SFS-300 $9.95+GST
Arduino Mega
2560R3 This Arduino is
fitted with an Atmega
2560 controller and
includes 256K Flash,
54 Digital I/O, 16 Analog inputs USB interface and heaps of
other features
SFA-104 $59.00+GST
Industrial grade Unmanaged 8 port Ethernet Switch. A 10/100MB
switch with a 5 year warranty this switch can
operate in temperatures
up to 70degC
ATO-005 $159.00+GST
Contact Ocean Controls
Ph: 03 9782 5882
www.oceancontrols.com.au
12 Silicon Chip
Solar power absurdities
“I want to receive the generous
feed-in tariff just when my solar generation is at a maximum, so please
wind down the grid voltage so that I
can do so.” Why should I get a better
price for my power than the power
station operators? There must be a
more equitable way of encouraging
local (and usually more expensive)
power generation, if house holders
must go to this trouble.
In current practice, you can’t just
switch off large or even multiple
power generators to suit the load,
let alone consumers who become
generators. Even the planned coal/
gas replacement for Victoria’s big
Hazelwood generators may not react
quickly enough when the clouds
part over the suburbs of Melbourne.
Electric cars with big enough
batteries have been contemplated
as generators and storage devices
connected to smart grids but why
do it? Why waste all those expensive
and short lived batteries! Yet it might
be feasible to encourage each household to be energy self-sufficient
given big enough batteries or other
power storage systems, and removing the need for air-conditioners by
local storage and usage of heat.
It seems that the combined salt/
solar 100MW base-load system at
Mildura might get further funding
second box repeats the signal from
the Yagi.
The two boxes cannot be placed
close together. If you do so, a red light
comes on telling you to move the boxes
further apart. What an amazing piece
of technology! Any number of people
can use their mobile phones in the
house, just like you are living next
door to a phone tower.
And the best news of all is Telstra
did this at no cost to us. Finally I can
give Telstra, and in particular their
CEO, a “well done”, for letting me
talk to someone who had real technical knowledge and as a result we now
have mobile phone coverage. Well
done Telstra.
William McGhie, VK6UU,
Gooseberry Hill, WA.
and yet it is not very sensible to consider significant up-scaling requiring “large lakes of fire” as has been
mooted in Europe. Reversible hydro
has been mentioned lately as having
storage recovery efficiencies of the
order of 70%. I suppose salt water
could even be used as the working
fluid, to avoid running out of potable
water. But why do it?
Maybe in 30 years time, it will be
“proven” that the oceans and water
vapour (latent heat of vaporisation in
the atmosphere), not CO2, are what
really stabilises the climate – that
leaves control of pollution as a very
reasonable thing to do. Yet there is
probably no hurry (I can already hear
the howls of protest!)
Meanwhile, the number of sunspots with their intense solar flares
are starting to once again have a
measurable effect on our climate,
as also might the seeding of clouds
by cosmic rays when the sunspots
were recently reduced in number.
Why do we try to make things so
complicated? Why are we so frightened? Modern technology offers
many new ways of doing things and
to do so efficiently but we do need
to keep things in perspective, hopefully simple and easy to understand
and control.
Brian Tideman,
Mulgrave, Vic.
Autonomous long-distance
trucks might be workable
Talk on the internet to any of the advocates of an ultra-fast fibre-to-everyhome National Broadband Network
and it is incomprehensible to them
that anyone could not want it. Anyone who doesn’t must, they think,
be misinformed or stupid. And you
could see exactly that attitude when
the engineer in charge of Google’s
autonomous vehicle project was interviewed. Wouldn’t everyone just
want to be able to get in their car in
the morning, press a button and have
it take them to work while they sit in
the back and read the paper? Yes, of
course. That was his attitude.
That is undoubtedly true for some
people. But for a lot of people, driving
siliconchip.com.au
Wasteful wind power
keeps them physically and mentally
occupied for the period that it takes to
get them to where they have to be and
they enjoy the challenge of it.
Autonomous vehicles are the next
big thing in cars. The last big thing was
electric cars. Every politician talked
about how car manufacturers should
be building them. But you can count
annual sales of electric cars on your
fingers.
General Motors’ Volt plant in the US
has laid off its workers for a couple of
months because few people are buying them. Another company got huge
media attention by talking about how
it would be building battery swap
stations so electric cars could go cityto-city despite their short range and
long recharge time. There hasn’t been
a single vehicle sold in this country
that could use them.
Advancing technology makes a lot of
things possible. But they have to make
sense to buyers, to provide them with a
benefit that’s worth the price. Holden’s
version of the Volt will be based on
the Cruze. It will be the same size car
with similar equipment and carrying
capacity. But it will cost three times
the price. The autonomous version
of cars will similarly carry a hugely
higher price tag. I have no doubt that
the car my children will be driving
will, as well as being electric, will also
be autonomous. But it won’t be either
of those because most people want it.
Let’s try to understand wind
power. Every wind farm needs
backup generators to supply power
when the wind fails. If there is no
wind, zero electricity is produced
by the turbines and all power comes
from base-load power stations or the
backup gas-fired generators.
If wind speed exceeds the design
capacity, the turbines are shut down
to prevent damage and all power
comes from the base-load or backup
generators. In freezing still air, the
wind turbines take electricity from
the backup generators to prevent
It will be electric because oil will keep
getting more expensive.
It seems to me that the developers
of autonomous vehicles, like Google,
are targeting totally the wrong market.
The news is full of stories about longdistance truck drivers who are speeding to make the destination before their
allowed hours behind the wheel run
out, or fatigued and taking drugs to
prevent that fatigue because they have
to work long hours.
That’s where the developers of
autonomous systems could get their
product onto the roads today. Big
trucks are expensive. A whole lot
of bulky and expensive hardware
wouldn’t be noticed and long-distance
haulage companies could give us what
we want. For our eBay purchases to
arrive faster and for our supermarkets
to be able to offer year-round fresh
damage from cold. And they draw
power to get reconnected.
When the wind blows strongly
all over the wind farm, the grid may
not be able to cope with the surge in
supply so the operator may be paid
to close down some turbines.
Now we find that wind power
probably increases the production of
carbon dioxide; not that this matters.
So why not scrap the wind turbines
and produce a steady supply of
low cost power from the base-load
generators?
Viv Forbes,
Rosewood, Qld.
food (no matter where in the country it has to be grown), they need to
be able to keep their vehicles going
for as long as takes to get anywhere
without any occupational health or
legal compliance issues. Making them
autonomous, even if that still required
having a human supervisor on board
to take over in case of the unforeseen
and unprogrammed, would make good
commercial sense.
It is not by accident but because it
made sense that the first autonomous
vehicles in this country were longrange aerial drones used for maritime
surveillance and the next will be longdistance ore trains.
Gordon Drennan,
Burton, SA.
Comment: 3915 Chevy Volt electric
cars were sold in the USA for the first
SC
quarter of 2012.
Australia’s Lowest Priced DSOs
Shop
On-Line
at
emona.com.au
Now you’ve got no excuse ... update your old analogue scopes!
Whether you’re a hobbyist, TAFE/University, workshop or service
technician, the Rigol DS-1000E guarantee Australia’s best price.
RIGOL DS-1052E 50MHz
RIGOL DS-1102E 100MHz
50MHz Bandwidth, 2 Ch
1GS/s Real Time Sampling
512k Memory Per Channel
USB Device & Host Support
100MHz Bandwidth, 2 Ch
1GS/s Real Time Sampling
512k Memory Per Channel
USB Device & Host Support
ONLY $
Sydney
Melbourne
Tel 02 9519 3933
Tel 03 9889 0427
Fax 02 9550 1378
Fax 03 9889 0715
email testinst<at>emona.com.au
siliconchip.com.au
Brisbane
Tel 07 3275 2183
Fax 07 3275 2196
362
Adelaide
Tel 08 8363 5733
Fax 08 8363 5799
inc GST
Perth
ONLY $
Tel 08 9361 4200
Fax 08 9361 4300
web www.emona.com.au
439
inc GST
EMONA
May 2012 13
The Australian
The Australian Synchrotron is one of the nation’s largest and most
significant scientific facilities. It is a powerful machine of great utility
that enables investigators to determine the structure and composition
of all materials, including living specimens, with extremely high detail.
By Dr David Maddison
T
he Australian Synchrotron, located adjacent to the
Monash University campus in Clayton, Victoria, was
completed in 2007 at a cost of $221 million.
Funding came from the Victorian Government with a
contribution of $157 million, with additional funding of
$50 million from other state government, university and
research organisations and a $14 million dollar contribu-
tion from the Commonwealth government. About 65%
of the initial funding was spent with local suppliers and
contractors. As well, substantial design input was made
by Australian scientists and engineers. The facility has an
annual operating budget of $25 to $30 million.
When not undergoing scheduled maintenance, the synchrotron runs 24 hours per day, year round, producing a
Bird’s-eye view of the Australian Synchrotron (bottom right). Some idea of the size of this facility can be gleaned by
comparing it with the oval in the grounds of the Monash University at left!
14 Silicon Chip
siliconchip.com.au
Synchrotron
What is a synchrotro
n?
wealth of scientific results and important industrial research.
It is one of about 50 similar devices around the world,
although not all are as new or as advanced. Typically 3,500
scientists visit the facility each year and work on more than
600 experiments.
In order to probe a material’s structure the Synchrotron
produces what is essentially very high quality light, tunable
over a wide variety of wavelengths from the microwave part
of the spectrum through to “hard” X-rays (see diagram).
Note that non-visible electromagnetic radiation such as
X-rays is also considered a form of light.
The beam is very intense with a brightness of around one
million times greater than that of the Sun and the X-rays
produced can be millions of times more intense than those
produced by conventional X-ray tubes.
The Synchrotron is a state-of-the-art, third generation
device. It was conceived at the outset to produce bright
The range of wavelengths produced that the Australian
Synchrotron. Image: Australian Synchrotron.
siliconchip.com.au
As described on the Au
stralian Synchrotron
website, in simple terms, a synchro
tron is a very large, cir
cular, megavoltage
machine about the siz
e of a cricket ground.
From outside, the
Australian Synchrotro
n, for example, looks
very much like a
roofed football stadiu
m. But on the inside
, it’s very different.
Instead of grass and
seating, there is a vast,
circular network of
interconnecting tunne
ls and high tech appa
ratus.
Synchrotrons are a typ
e of particle accelerato
r and when used
to accelerate electron
s, can produce inten
se beams of light, a
million times brighter
than the sun. The light
is produced when
high-energy electrons
are forced to travel in a
circular orbit inside
the synchrotron tunne
ls by ‘synchronised’ ap
plication of strong
magnetic fields with ve
ry powerful electrom
agnets.
The electron beams tra
vel at just under the
speed of light –
about 299,792 kilom
etres per second. Th
e intense light they
produce is filtered and
adjusted to travel into
experimental workstations, where the lig
ht reveals the innermos
t, sub-microscopic
structure of materials
under investigation, fro
m human tissue to
plants to metals and
more.
With this new knowled
ge that synchrotron sc
about the molecular str
ience provides
ucture of materials, res
earchers can invent
ways to tackle disease
s, make plants more
productive and metals more resilient – am
ong many other bene
ficial applications
of synchrotron science
.
More technical inform
ation about how the Au
stralian Synchrotron and other similar
facilities work is availab
le from the ‘ABOUT
US/Our facilities’ secti
on of the www.synchro
tron.org.au website.
X-rays and other wavelengths of light compared with the
first generation of such devices in which synchrotron radiation was utilised essentially as a by-product of particle
accelerators..
Other characteristics of the generated light are that it is
highly collimated meaning that the light rays in the beam
all travel parallel to each other as in a laser beam.
The light beam is also polarised and different polarisation
modes can be produced as required for different experiments. In addition, the light is also pulsed.
Information about the structure and composition of matter is revealed by the way the light beam interacts with the
object under investigation.
The beam may be absorbed, transmitted, refracted or diffracted by the object and by carefully measuring the beam
May 2012 15
,
Australian Synchrotron control room. Image: the author.
properties after it has interacted with the test specimen,
it is possible to determine its structure and composition.
The Synchrotron is used by Australian and New Zealand
scientists and industrial researchers and by many other
scientists from around the world. These scientists and
associated staff are extremely dedicated and enthusiastic
about their work in this facility. To accommodate the many
visiting scientists there is an accommodation block currently under construction.
Schematic view of the Australian Synchrotron. Image:
Australian Synchrotron.
16 Silicon Chip
Applications
Most experiments fall within three main categories.
These are (a) X-ray diffraction and scattering to determine
the crystal structure and other structural properties of
samples; (b) spectroscopic analysis down to nanometre
resolution (one millionth of a millimetre) to determine the
chemical composition of samples and (c) high resolution
imaging at any wavelength of light that can be produced at
the Synchrotron of biological and non-biological materials,
animals and, in the future, humans.
Some research highlights from the Synchrotron are as
follows:
• Determination of whether a proposed coating on electrode wires used on the Monash bionic eye would damage the wires. The rapidity with which the results were
obtained saved a large amount of development time
and money.
• Improvement of processes to extract pharmaceutical
substances from poppies by determining their chemical
structure from minute samples.
• Development of microbeam therapy to treat cancer.
• Imaging of lung function in newborn animals to better
understand breathing processes in premature babies.
• Discovery of new information about an immune system
protein leading to a better understanding of treating
diseases.
• Research on the life-cycle of the malaria parasite in blood
cells which will lead to the development of better drugs
to control the disease.
• Research on immune system T-cells to develop better
drugs to boost immune system function.
• Use of imaging techniques to develop new procedures
to accurately place cochlear implants and improve their
function.
• Development of techniques to accurately identify healthy
human egg cells for use in IVF procedures.
siliconchip.com.au
“MASSIVE” Computing Facilities
Part of MASSIVE1.
Image: the author.
Each experiment at the Synchrotron produces large
quantities of data which need to be stored and processed.
The Synchrotron facility has a supercomputer cluster to
perform this task.
MASSIVE (Multi-modal Australian ScienceS Imaging
and Visualisation Environment) provides the hardware,
software and personnel needed to service this task,
among others.
This facility, which actually consists of two machines
connected by a high-bandwidth link, is also accessible by
scientists working in areas outside the Synchrotron such
as in neuroimaging, geosciences and microscopy or any
other area that requires advanced image processing and
visualisation resources.
The great computer power allows three-dimensional
images to be generated and manipulated in real time,
enabling researchers to adjust their experiment and/or
the beam parameters without having to wait for postprocessing of image data.
MASSIVE1 is located at the Synchrotron facility and
MASSIVE2 is located next door at Monash University,
Clayton campus. The computer utilises both CPUs (Central Processing Units) and GPUs (Graphics Processing
Units) for its computing tasks. The CPUs are used for
regular computing while the GPUs are used for graphic
processing and can also be used for matrix and vector
operations for non-graphic tasks.
siliconchip.com.au
MASSIVE1 has a capacity five teraflops for traditional
CPU computing and 50 teraflops when using its GPU coprocessors and MASSIVE2 has a capacity of 10 and 100
teraflops respectively. (One teraflop is 1012 floating point
operations per second.) The specifications are as follows:
MASSIVE1 at the Australian Synchrotron
• 42 nodes with 12 cores per node running at 2.66GHz (504
CPU-cores total)
• 48GB RAM per node (2,016GB RAM total)
• 2 nVidia M2070 GPUs with 6GB GDDR5 per node (84 GPUs total)
• 58TB of fast access parallel file system (IBM GPFS)
• 4x QDR Infiniband Interconnect
MASSIVE2 at Monash University
• 42 nodes with 12 cores per node running at 2.66GHz (504
CPU-cores total) in two configurations, 32 nodes identical
configuration to MASSIVE1
• 48GB RAM per node (1,536GB RAM total)
• 2 x nVidia M2070 GPUs with 6GB GDDR5 per node (64GPUs
total)
• 10 nodes (visualisation/high memory configuration)
• 192GB RAM per node (1,920GB RAM total)
• 2 x nVidia M2070Q GPUs with 6GB GDDR5 per node (20 GPUs
total)
• 250TB of fast access parallel file system
• 4x QDR Infiniband Interconnect
May 2012 17
Distribution map of titanium (blue), niobium (green) and
thorium (red) in ilmenite, an iron titanate mineral and an
important source of titanium dioxide for pigment. It was
produced using the innovative Maia detector. The field of
view is 10 x 6 mm. Image: La Trobe University, CSIRO,
Australian Synchrotron.
• Development of techniques to track stem cells as they
repair the body. This information can be used to develop
methods of stem cell therapy.
• Search for gold in ore samples in which the gold cannot
be detected by normal techniques. This may lead to the
discovery of new gold deposits.
• Analysis of the structure of sheep leather, which has led
to methods to strengthen it so it can be used for shoes,
something which is not otherwise possible.
• Understanding how the runoff from acidic soils affects
Australia’s east coast fisheries and the development of
methods to control soil acidity.
• Understanding the reason for the buildup of scale in
pipes used in the bauxite industry and the development
of methods to alter processing conditions in order to
minimise scale formation.
• Exploration of materials for use in the electronics industry such as synthetic diamond films.
• Studying the distribution of nutrients in foods after
processing in order to assist in the development of plant
varieties which better retain their nutrients.
• Studying old paintings to look for underlying images,
determine paint composition or to establish authenticity.
Analysing the composition of glazes on ancient Egyptian
artifacts.
• Examining the internal structure of ancient fossils which
are too fragile to completely remove from their rocky
encasement and also imaging soft tissue impressions
therein.
• Analysing the structure of “green” cement and enabling
VicRoads to update their standards to allow for its use.
• Studying molecular structures which are suitable for
hydrogen storage for its use as an alternative fuel.
• Researching the interaction of carbon dioxide with
various materials that may be used for sequestration
of the gas.
• Development of a forensic method to identify soil from
crime scenes using extremely small samples.
• Studies of the chemistry of fingerprints to enable improved detection.
• Discovering why Phar Lap died by looking for toxins in
hair follicles from his preserved hide. This indicated
ingestion of arsenic in the last 30 hours of life.
• Studying the distribution of elements in mineral samples
(see picture above).
18 Silicon Chip
Image of animal lungs clearly showing detailed structure.
Such detail cannot be achieved with conventional imaging
techniques. Image: Australian Synchrotron
Apart from other areas of world-leading expertise indicated above, scientists at the Synchrotron are leaders in
determining the structure of proteins, an essential component of all life forms.
The structural determination of many proteins is extremely difficult or impossible by conventional techniques
but is assisted at the Synchrotron using the technique of
small angle X-ray scattering. Normally, high quality crystals
are required for this work but unfortunately, some proteins
do not crystallise well.
In these cases, the Synchrotron can be used to determine
the shape of the protein’s outer “envelope”. With this partial
information it is possible to infer the rest of the structure
with the aid of advanced computing methods.
A particularly difficult medical imaging problem is to
visualise lung tissue and the motion of the lungs during
breathing. Due to the high resolution of the beam and
the tunability of the X-rays, successful imaging has been
achieved by Australian research groups and the findings
have already found application such as in studies of cystic
fibrosis and asthma.
“Tricks” of light are used to image the soft tissue and air
spaces of the lungs whereby X-rays are refracted differently
from the tissue and the air. Tuned with the right parameters
a “phase contrast” image, which can be viewed in real time
if desired, can be produced to show the working lungs.
How the light beam is produced
In essence the function of a synchrotron is to generate a
beam of charged particles travelling close to the speed of
light. These can then subject them to an acceleration which
causes them to emit light radiation. This beam of particles
is maintained in a storage ring.
Electrons are typically used as the charged particles for
light generation and different magnet configurations are
siliconchip.com.au
Medical applications
A recently built facility at the
Synchrotron site is the flagship
Imaging and Medical Beamline.
This was built with a grant of $13.2
million from the National Health
and Medical Research Council
and grant of $1.5 million from the
Victorian Government.
It will be used for medical (and
other) imaging research as well as
treatment research, for example
on high precision irradiation of
tumours.
An interesting area of research
is to irradiate tumours in a “checkerboard” pattern which is possible
due to the fine control possible
with the X-ray beam. This has
been shown to destroy tumours
just as effectively as normal radiation treatment but with much less
damage to healthy tissue.
Other clinical research will
include observing how tumours
respond to treatment and the The new Imaging and Medical satellite building. The synchrotron beam is conveyed
possibility of watching specially to this building via a 150m long tunnel. Image: the author.
marked individual cells migrate
through the body in real time.
to the new building. The long tunnel is needed to allow the
For patient comfort, the facility will provide patients with a X-ray beam from the Synchrotron to expand in size from
clinic-like rather than a “laboratory” experience. Note that the the original dimensions of 1mm wide by 50 microns high to
present intention is for selected patients to visit for clinical produce the largest X-ray beam of any synchrotron in the
research and trials only – this will not be a general facility world, having a cross section of 50cm by 4cm.
for patient treatment.
This beam will enable images to be produced with a resoluThe facility also contains sections to house and conduct tion of one micron over large areas of a human or animal body.
research on animals. Of interest is a miniature combined Typical human and animal cells are 10-100 microns in size
CT and PET scanner for small animals such as mice (see so images of individual cells should theoretically be possible.
picture).
Images that are about one hundred times more detailed
The central feature of this facility is the beamline that ar- than a hospital CT scanner will be able to be produced and
rives via a 150m long tunnel leading from the Synchrotron monitored in real time.
Imaging and Medical beam tunnel, 150m long (under
construction). Note the black support structures which
will hold stainless steel tubing under vacuum that will
contain the X-ray beam. One small section of tube is
installed in this picture. Imaging will occur in a room at
the end of the tunnel. Image: the author.
siliconchip.com.au
Miniature CT/PET scanner for small laboratory animals
such as mice. For scale, compare with the size of the
small computer screen on the left. Image: the author.
May 2012 19
Right: prototype
sextupole magnet.
Image: the author.
Prototype magnet assembly on display in the Synchrotron
building showing the bending or dipole magnet (yellow)
which causes the generation of the synchrotron radiation
when electrons pass through its centre (in the direction
from one side of the picture to the other) at close to the
speed of light. The red and green magnets at each end are
quadrupole and sextupole magnets respectively and these
are used to focus and steer the beam. Image: the author.
used to make the electron beam either bend, “undulate”
or “wiggle”, causing the electrons to accelerate and emit
light. Note that in physics terminology “acceleration” can
mean a change in either speed or direction. In this case it
is the change in direction as the electron travels through
the bending magnet that constitutes acceleration.
The Synchrotron consists of the following main components: electron gun, linear accelerator, booster ring, storage
ring and beam-lines where the radiation is emitted into the
experimental “end stations” as shown in the diagram of
the Australian Synchrotron.
Generating the electrons and then boosting their speed
is a multi-stage process.
Electrons are first generated with an electron gun similar
to one in a cathode ray tube, only larger. The electron gun
produces electrons with an energy of 90keV.
After leaving the electron gun, electrons are injected into
the linear accelerator (LINAC) where the 90keV beam is
boosted to an energy of 100MeV. Electrons are energised
using a series of radio frequency (RF) resonant cavities
which operate on a similar principle to the magnetron in
microwave ovens.
When a radio wave of the appropriate frequency is
generated and enters a resonant cavity, a standing wave is
created, the intensity of which increases as more RF energy
is injected. Electrons in the beam absorb that energy and
their speed is increased. The electrons are travelling at
99.9985% of the speed of light as they leave the LINAC.
After leaving the LINAC, the electron beam enters the
booster ring where the beam is further energised from
100MeV to 3GeV with the use of a 5-cell RF resonant cavity. The booster ring also contains 60 combined focusing
and steering magnets.
The electrons are resident in the booster ring for half
a second during which time they complete one million
circuits of the 130m-long ring. A new cycle for the next
batch of electrons can be initiated every second.
In the final stage, electrons from the booster ring enter the
storage ring. This has a circumference of 216m and actually consists of 14 main sections each with a 4.4m straight
20 Silicon Chip
Below: end-view
of prototype
quadrupole
magnet. Image:
the author.
section and an 11m arc-shaped section.
Each arc section contains two bending magnets (also
known as dipole magnets) as well as six quadrupole (four
pole) and seven sextupole (six pole) electromagnets.
Each bending magnet generates synchrotron radiation
as the electrons pass through it at close to light speed. As
shown in the following diagram, the radiation (green) is
Radiation pattern
(green) as electron
traverses the bending
magnet (path shown in
red). Image: Australian
Synchrotron.
emitted at a tangent to the direction of the electron path
through the magnet. It is this radiation that is used in
experiments.
At each experimental station at the active beam-lines
there are beam-line optics that contain filters, monochromators, mirrors, attenuators and other optical devices that
help condition the beam to the required characteristics for
each experiment. Following these optics is the rest of the
experimental equipment such as a spectrometer or X-ray
diffraction apparatus.
All of the “end station” equipment sits in a radiationshielded “hutch” to protect staff from X-ray radiation.
siliconchip.com.au
Part of the storage ring of the Australian Synchrotron. Image: Australian Synchrotron.
The quadrupole and sextupole magnets are used to keep
the electron beam focused and to correct for any aberrations
in the beam. In all, there are 84 quadrupole magnets and
98 sextupole magnets in the storage ring. The sextupole
magnets also have extra windings to provide vertical or
horizontal corrections to the beam path.
Typically the electron beam is 50 microns wide with a
deviation from the desired path of no more than 5 microns
(one micron is one thousandth of a millimetre).
The magnets are water cooled and the temperature in the
main building and the beam tunnel is highly controlled to
minimise errors due to thermal effects in equipment and
the structure.
Two of the straight ring sections contain a total of four
RF cavity resonators in order to replace beam energy that
is lost due to synchrotron radiation.
The remaining twelve straight sections are able to accommodate “insertion devices”. These devices are used
to further increase the intensity of the light and impart it
with certain characteristics. There are two types of insertion devices. One is the “multipole wiggler” and the other
is the “undulator”.
In the wiggler, light cones are emitted at each bend in the
electron trajectory and these cones reinforce each other to
Multipole wiggler:
the green shading
represents the emitted
radiation and the red
line represents the electron
path. Image: Australian Synchrotron.
create an extremely bright, broad spectrum beam.
In the undulator, weaker magnets are used, resulting
in a more gentle bending of the electron’s path. In this
configuration some cones of light interfere with each
other cancelling out their energy, while others reinforce
each other. By adjusting the spacing between the magnet
poles it is possible to enhance some frequencies of light
to thousands of times the intensity of other frequencies,
allowing for an extremely intense beam at one particular
wavelength of choice.
Undulator: the
radiation pattern is
shown in green and
the electron path in red.
Image: Australian Synchrotron.
siliconchip.com.au
The electron beam needs to be maintained in an enclosure that is kept under an extremely high vacuum, in this
case 10-13 bar (10nPa) where 1 bar is equivalent to about
one atmosphere of pressure. The reason for this ultra high
vacuum is so that the electrons will not lose energy or be
scattered by residual gas particles.
As the electrons in the beam are travelling at very close
to light speed Einstein’s Theory of Relatively applies. Due
to relativistic effects, including time and length contraction,
from the electrons’ point of view, the time and distance
through which they travel appears much shorter than a
stationary observer would experience.
This means that the frequency of light emitted as the
electrons are accelerated through the bending, wiggler or
undulating magnets is many orders of magnitude greater
than would otherwise be the case if Relativity did not apply.
Beam-lines and future development
Currently there are nine beam-lines in use. These are used
for powder diffraction, X-ray absorption spectroscopy, small
and wide angle X-ray scattering, soft X-ray spectroscopy,
infrared spectroscopy, macromolecular spectroscopy and
micro crystallography, X-ray fluorescence microscopy and
medical imaging.
All these beam-lines are in constant heavy use and even
so, there is not enough beam-line time available to service
the demand for them. Fortunately, the Synchrotron was
constructed with future expansion in mind and a total of
29 additional beam-line positions are available.
The Synchrotron is subject to continual improvement
and there is a dedicated accelerator physics group who
are constantly working to better the device by improving
control systems, beam parameters and researching theoretical aspects of synchrotron devices.
Conclusion
The Synchrotron provides Australian researchers with
a powerful, world-leading set of tools for analysing and
imaging living or non-living matter in ways that are unSC
achievable by conventional techniques.
OPEN DAYS
The Australian Synchrotron has periodic Open Days. The last
one, in November 2011, attracted over 3,000 people. The next
Open Day is expected to be later this year. Keep an eye on the
Synchrotron website (www.synchrotron.org.au) for details.
May 2012 21
Getting the most from
ADSL
It’s a fair bet that most readers of SILICON CHIP enjoy their daily
fix of internet access by courtesy of ADSL. While some readers are
luxuriating with optical fibre – and an unhappy minority are still
using dial-up – most of us owe our ongoing communication to the
distinctly freakish technology of ADSL. But what exactly is ADSL
and what came before it? How does it work and why is it often called
a “freak” technology? Are your internet speeds painfully slow? Can
anything be done to speed them up? Do you curse your ISP? Read on!
By ALAN FORD
M
any of us remember the early days of text-only
bulletin boards (which could be regarded as the
fore-runners to today’s internet), to which we
connected via an acoustic modem.
Bulletin boards were set up in the early 1980s by special
interest groups, some businesses and even altruistic individuals. Most specialised in a particular subject or brand
and we connected to them by dialling a number specific to
that bulletin board. We then carefully inserted the telephone
handset into a contraption of cups and flexible joints.
We laid it on its side to prevent
the carbon granules in the
microphone
from
A
Radio Shack
(Tandy in
Australia) acoustic
modem from the 1980s.
These did not work well
with the carbon microphone used
in Australian telephones at the time.
22 Silicon Chip
coalescing, moved the cat out of the room to prevent the
heavy tread of its paws from interrupting data flow and
settled down to enjoy the lightning fast data transfer speed
of approximately 300 bits per second (bps).
Without delving into the esoteric realms of parity bits,
overhead or consideration of baud versus bits (don’t ask!),
you can take that as about 35 characters/bytes per second
(8 bits = 1 byte). In practice, various factors contributed to
delays (as they do today) and we would usually see text
characters emerging on our computer screens one by one
or in groups of a few at a time.
How did that old acoustic modem work? Computer
binary data streams (well, trickles) would be converted
in the modem (or “modulated”) and transmitted over the
telephone line as frequency shifted audio tones. At the
ISP’s end another modem would convert the sounds to data
(or “demodulated”), or vice versa. In fact, that’s where the
word “modem” comes from: it’s a MODulator/DEModulator.
That was fine for plain text but then along came graphics,
with Microsoft Windows a pioneer (but certainly not the
only one), as well as the World Wide Web. Now we needed
to access the Web with its rich images and sounds as well.
The direct modem
Enter the direct-wired modem, connecting the computer
electrically to the PSTN (Public Switched Telephone Netsiliconchip.com.au
Aaahhh – the way we were! This photo, taken in 1981, shows a youthful Dick Smith talking bits and bytes with an equally
youthful and then-hirsute Leo Simpson. But the main point about this picture is not so much the all-new System 80
computer and its external floppy disk drive, it’s that whizz-bang acoustic modem in which the telephone handset resides.
The problem with this (which obviously Dick and Leo didn’t understand) was that the modem needed to be turned on its
side to work properly, otherwise the carbon granules in the microphone would tend to coalesce – and cause data loss.
work), with much anguish on the part of the telcos (well,
Telecom Australia!).
They (Telecom) even took to placing adverts in the media
warning of the dangers of using unapproved (ie, not supplied by them!), mainly imported modems and the heavy
fines for doing so. It wasn’t too long before they realised the
horse had well and truly bolted so instead started issuing
approvals for imported equipment.
Wired modems used more complicated methods of coding and offered much faster communication than the simple
two-tone system of the acoustic device. Speeds increased
as modulation methods became more clever, until in theory
56kb/s could be reached.
In case you haven’t done the sum, that’s about 187 times
faster than the acoustic modem! Of course, we still were
using the PSTN speech path and bandwidth, so for a time
it was thought that 56kb/s was the limit. But we could now
get our images — even some (jerky) moving ones!
But as well as new Web applications needing
even more speed, there was another big disadvantage to the technology — the engaged line
syndrome. While we were using the ‘net’, the Mother-inlaw received the engaged signal and could not impart any
telephonic wisdom to the family. Neither could the kids
call their friends.
Today both would use email or even the dreaded text
messaging but we are getting ahead of ourselves. Or if
by chance we had the call-waiting facility activated, the
internet experience would be rudely but soundlessly interrupted; neither the Mother-in-law nor our Net aspirations would be fulfilled. A fortunate few might have had a
separate line installed for Net use but it was not a general
rule for households.
POTS and carriers
Then along came ADSL, really a “freak” technology and
After acoustic modems came direct-connect
modems such as this D-Link DFM-526E 56K. At the
time, everyone thought they were unbelievably fast
compared to acoustic models. But ADSL has consigned
them to the rubbish-heap of history!
siliconchip.com.au
May
ay 2012 23
in some ways it’s surprising it works at all. But it does work
if all or most of a large number of aspects are at or near
optimum, as I shall explain.
Hopefully my words will reduce the total of frustrated
users and prevent many of those newsgroup or forum posts
that sometimes use violent language to blame the ISP for
shortcomings that are entirely outside its control!
To appreciate how ADSL works it is helpful to go right
back to the basic telephone network — the Plain Old
Telephone System (POTS) in the mid part of last century.
At first, most lines were strung overhead, including long
distance lines, before the much later advent of coaxial and
tower-to-tower microwave links. The stringing of a dozen
or so wires between say Sydney and Melbourne was expensive and there were obvious limitations of space as the
bare wires could not be allowed to touch and short in any
foreseeable winds.
To have just a dozen connections between Sydney and
Melbourne seems ludicrous now and in fact it was ludicrous
then. So it was necessary to somehow concentrate several
speech channels down one pair of lines in order that the
best use be made of that expensive (and expensively erected)
copper — even galvanised iron in some places!
The solution was carrier telephony. A number of telephone channels were modulated onto several different
radio frequency (RF) carriers, sent down the overhead
wires and separated and demodulated at the other end.
Normal speech occupies a relatively narrow bandwidth;
typically in those days the speech path was designed for
a bandwidth of 200Hz-3kHz.
But the lines were capable of carrying frequencies of
several hundred kilohertz — radio frequencies but still
carried by line.
A typical carrier system in use in the 1950s was capable
of concentrating 17 RF channels down one pair of wires,
spaced by 4kHz, with the highest being 68kHz. Later systems used even higher frequencies.
ADSL and more carriers
Years later, internet engineers reckoned (correctly) that
they should be able to do the same sort of thing with internet signals. The ADSL method consists of modulating
a large group of separate RF channels, often called bins or
buckets, and sharing the data to be transmitted digitally
amongst them.
That’s the Digital Subscriber Link (DSL) but what about
the A for Asymmetric?
Think about how we typically use the internet. We type
a few characters of a website address and in return we get
pages of visual information and plenty of text too. So most
of the data traffic is downloaded and therefore most of the
bins are allocated to it.
An ADSL2+ capable line carries the normal speech and
telephony (POTS) signals in the first 4kHz of bandwidth,
followed by a guard (unused) band from 4 to 25kHz, and
then a large block of separate frequencies spaced 4.3125kHz
apart (up to 512 of them) above that for the internet data.
These separate channels are the bins and about 5% of them
are used for upload with the rest for download.
The number of bins and the allocation between upstream
and downstream varies according to which version (‘Annex’) of the standard is in use. (And before you pundits
TYPICAL ADSL2+ FREQUENCY ALLOCATIONS (not to scale)
ADSL BINS AT 4.3125kHz SPACING
POTS
0-4kHz
GUARD
UPSTREAM
25kHz
DOWNSTREAM
138kHz 142kHz
26 BINS (25 available)
2.2MHz
479 BINS (446 available)
NOTE: some bins are used for pilots or other special purposes
Here’s how ADSL is arranged on a standard PSTN telephone line. The bottom 4kHz is reserved for your phone calls,
followed by a number of channels (‘bins’) for uploaded data and a much larger number for downloaded data.
24 Silicon Chip
siliconchip.com.au
reach for your keyboards, I am simplifying the position for
the benefit of newbies).
Separating the information
How is all the information kept separate? First of all let’s
deal with telephony, because this touches on a great advantage of ADSL — the end of the ‘engaged line’ syndrome!
When ADSL is in use, each telephone should be provided
with a low pass filter that allows the DC signalling (such
as on-hook condition), AC ring current and audio frequencies (such as speech and DTMF dialling) to pass normally.
The filter passes the lowest part of the total passband
(up to 4kHz), to the telephone and keeps it separate from
the RF of the internet connection.
The internet is always connected but the telephone, duly
filtered, operates normally. Whether or not we are using
the net, the phone will still ring if someone is calling and
neither telephone party will hear the internet signals.
Now to the Net connection. All the bins, whether allocated to upload or download, are kept separate by the
special ADSL modem, a complex piece of technology now
relatively cheap. At the telephone exchange end the allocation of the bins is controlled by the Digital Subscriber
Line Access Multiplexer (DSLAM). The Modem-DSLAM
combination does more than keep all those bins separate. It
is also a smart self-training combination, passing information on a per bin basis according to how free of interference
each bin is.
Later we will see why this is one possible reason for a
slow internet connection. Meanwhile you can see that the
provision of a large number of separate but simultaneous
bins (= channels) offers a vastly improved speed.
Are your expectations too high?
It is not good for the blood pressure to pursue the unattainable. Because it involves transmitting RF down a copper pair, with corresponding attenuation and other effects,
your speeds will depend on your cable distance from the
nearest exchange and of course, the cable distance will be
more than the ‘crow flying’ distance.
Because ADSL gets progressively slower as the cable distance rises, it becomes marginal at 4km and will probably
not work at all at 5km, although there can be exceptions.
I am lucky enough to be 167 metres cable length from
my exchange and on ADSL2+ I enjoy at least 10Mbit/s
and sometimes nearly 16Mbit/s download speeds, at the
same time as 0.8Mbit/s upload speed, although the copper
cabling here is not very good.
To put that in perspective, the download speed is up
to over 283 times faster than a dial-up modem and 53,000
times faster than the old acoustic modems!
Because speed varies so much with cable distance and
quality of connection, it is not possible to lay down hard
and fast rules but there are many ISP and other sources
on the net where you can compare your speeds to others
in your area.
There’s a possible fly in the ADSL ointment. You may
be on a telephone line concentrator system, such as a RIM
(Remote Integrated Multiplexer), where many lines are
multiplexed and share a fibre (or even coax) link to the
exchange.
Since ADSL is itself multiplexed there can be clashes
and speed penalties. Your telco will tell you if you are on
a RIM or similar concentrator and whether you can expect
a good ADSL experience.
In fact, when you enquire about ADSL, one of the first
things that happens is that you are asked for your phone
number to check whether you are on a multiplexed system.
Unfortunately, that’s all that is checked – the line is not
physically checked to see if ADSL is possible until you
actually apply for the service.
Low speed, dropouts & throttling
Basically there are two types of trouble you can face
as an ADSL subscriber: low speed and drop-outs. What
about throttling? No, not the person at the ISP help desk,
the speed.
Because this check is so simple it ought to be the first
that you do. Many ISP plans have a data limit, after which
data speeds are deliberately restricted or “throttled”. This
is irritating but you will probably agree that the alternative
of receiving an unexpected bill for excess data usage would
be a tad more annoying! Check via your ISP’s website to
see if you are being throttled.
Telephone line problems
Bearing in mind that ADSL is really a freak technology
where we stuff RF signals down a copper pair that it was
not designed for, for optimum speed and especially for freedom from drop-outs the line needs to be in good electrical
condition. There are many joints in a phone line between
you and the exchange as separate ‘pairs’ of wire are connected at street cabinets or pits. If any of those joints are
faulty you will have problems.
Your Reliable Partner in the Electronics Lab
ab
LPKF ProtoMat E33 – small, accurate, affordable
Hardly larger than a DIN A3 sheet: The budget choice for milling, drilling and
depaneling of PCBs or engraving of front panels – in LPKF quality.
www.lpkf.com/prototyping
Embedded Logic Solutions Pty. Ltd. Ph. +61 (2) 9687 1880
siliconchip.com.au
Email. sales<at>emlogic.com.au
May 2012 25
DSL Glossary
ADSL
Asymmetric Digital
Subscriber Line.
ATM
Asynchronous Transfer Mode.
Authentication
Auto-negotiation
Bandwidth
Cross-talk
A digital subscriber line (DSL)
technology in which the
transmission of data from server
to client is much faster than the
transmission from the client to
the server.
A cell-based data transfer
technique in which channel
demand determines packet
allocation. ATM offers fast
packet technology, real
time, demand led switching
for efficient use of network
resources.
A security feature that allows
access to information to be
granted on an individual basis.
Procedure for adjusting
line speeds and other
communication parameters
automatically between two
computers during data transfer.
The range of frequencies a
transmission line or channel can
carry: the greater the bandwidth,
the greater the informationcarrying capacity of a channel.
Signal currents being induced
into neighbouring wires and
causing errors.
bit
(“BINary digiT”)
A single unit of data, where there
are only two possible states. The
smallest amount which can be
carried/transmitted.
bps
bits per second
A standard measurement of
digital transmission speeds.
Bridge
A device that connects two or
more physical networks and
forwards packets between them.
Broadband
Characteristic of any network
that multiplexes independent
network carriers onto a single
cable. This is usually done using
frequency division multiplexing
(FDM).
byte
DMT
Discrete Multitone
26 Silicon Chip
(usually!) 8 bits = 1 byte; origin
is the number of bits needed to
define one text character.
The leading method of signal
modulation for DSL service.
The usable frequency range is
separated into 512 frequency
bands (or channels) spaced
4.3125kHz apart.
DMT uses the FFT (fast Fourier
transform) algorithm as its
modulator and demodulator.
Downstream rate
The line rate for return messages or
data transfers from the network to
the customer.
DSL
Digital Subscriber Line
A technology for bringing highbandwidth information to homes
and small businesses over ordinary
copper telephone lines.
DSLAM
Digital Subscriber Line
Access Multiplexer
A device at the telephone exchange
which enables connection to
multiple customers simultaneously.
Encapsulation
The technique used by layered
protocols in which a layer adds
header information to the protocol
data unit (PDU) from the layer
above.
FTP
File Transfer Protocol
The Internet protocol (and
program) used to transfer files
between hosts.
HTML
Hypertext Markup Language
The most common page-coding
language for the World Wide Web.
HTML browser
(or web browser)
A browser used to traverse the
world wide web.
http
Hypertext Transfer Protocol. The
protocol used to carry world-wide
web (www) traffic between a www
browser computer and the www
server being accessed.
Internet address
An IP address assigned in blocks
of numbers to user organizations
accessing the Internet.
Internet
A collection of networks
interconnected by a set of routers
which allow them to function as a
single, large virtual network.
IP
Internet Protocol
The network layer protocol for the
Internet protocol suite.
IP address
The 32-bit address assigned to
hosts that want to participate in a
TCP/IP Internet. Written as four
numbers separated by dots.
ISP
Internet Service Provider
A company that allows home and
corporate users to connect to the
Internet.
LAN
Local Area Network
A data communications network
restricted to a small area (often
within one building or office)
siliconchip.com.au
Last mile
The final connection between
the nearest exchange and the
subscriber (for most people at the
moment a copper pair).
Line rate
The speed by which data can be
transferred over a particular line
type, express in bits per second
(bps).
Loopback
A diagnostic test that returns the
transmitted signal back to the
sending device after it has passed
through a network or across a
particular link. The returned signal
can then be compared to the
transmitted one.
MAC
Media Access Control Layer. A
computer’s interface to a physical
network.
Multiplexer
A device that can send several
signals over a single line. They are
then separated by a similar device
at the other end of the link.
Router
A system responsible for making
decisions about which of several
paths network (or Internet) traffic
will follow.
SNMP
Simple Network
Management Protocol
The network management
protocol of choice for
TCP/IP-based internets.
Split pair
Where the earth leg of one
twisted pair is inadvertently
swapped with the earth leg of
another during jointing, leading
to the noise cancelling effect of
the twist being defeated.
Spoofing
A method of fooling network end
stations into believing that keepalive signals have come from
and return to the host. Polls are
received and returned locally at
either end of the network and are
transmitted only over the open
network if there is a condition
change.
Synchronous connection
During synchronous
communications, data is not sent
in individual bytes, but as frames
of large data blocks.
TCP
Transmission Control Protocol
The major transport protocol in
the Internet suite of protocols
providing reliable, connectionoriented full-duplex streams.
NAT
Network Address Translation.
Allows multiple computers to share
one IP address.
Packet
The unit of data sent across a
packet switching network.
PAP
Password Authentication Protocol.
Pair
The pair of copper wires making up
an individual telephone circuit.
UTP
Unshielded Twisted pair
Port
The abstraction used by Internet
transport protocols to distinguish
among multiple simultaneous
connections to a single destination
host.
Two insulated copper wires
twisted together to reduce
potential signal interference
between the pairs.
Upstream rate
The line rate for message or data
transfer from the source machine
to a destination machine on the
network. Also see downstream
rate.
VC
Virtual Connection
A link that seems and behaves
like a dedicated point-to-point line
or a system that delivers packets
in sequence, as happens on an
actual point to point network.
In reality, the data is delivered
across a network via the most
appropriate route.
WAN
Wide Area Network
A data communications network
that spans any distance and is
usually provided by a public
carrier (such as a telephone
company or service provider).
POTS
Plain Old Telephone Service
Also known as PSTN – the public
switched telephone network.
PPP
Point-To-Point-Protocol.
Provides router-to-router and hostto-network connections over both
synchronous and asynchronous
circuits.
Protocol
A formal description of messages
to be exchanged and rules to be
followed for two or more systems
to exchange information.
RIM
Remote Integrated Multiplexer
Where many lines are locally
multiplexed and share a link to the
exchange. Often precludes ADSL.
Route
The path that network traffic takes
from its source to its destination.
siliconchip.com.au
May 2012 27
vals, it turned out to be a paging alarm system on another
subscriber’s line, with, you guessed it, a split pair.
Local RF Interference
Flat patch lead at top and standard UTP (unshielded
twisted pair) below. The twisted pair lead is preferable for
minimising noise and interference.
Pick up your phone and dial 1. In the few seconds before
the “number unobtainable” tone kicks in you should have
absolute or very near silence (possibly a very faint steady
hiss with some types of phone). There should definitely
be no intermittent crackles (caused by bad joints) and no
cross talk from other telephone users. Cross talk or hum
may indicate a split pair somewhere on the route between
you and the exchange. Is a split pair the same as split end?
A split pair is where the technician has in error used
someone else’s telephone earth line instead of yours. The
phone still works because all the earth wires are connected
together at the exchange but it is noisy, and ADSL does not
like noise. The pair of telephone wires that make up your
normal connection are twisted together in the multi-way
cable that usually runs underground in ducts. The twists in
the pair go a long way towards minimising noise because
the noise currents in the adjacent wires balance out. But
if by error your earth wire is really someone else’s, it is no
longer twisted with your other wire. The pairs are split
and noise will result.
Returning to crackly joints, these are often far worse in
extreme weather, by which I mean excessive heat or cold,
or heavy rain. If you hear line noise, even when all other
equipment is removed from your phone connection, there
is a problem either inside or outside your premises. If
inside you will need the services of a competent licenced
cabler. If outside, your approach needs to be to Telstra or
any other line retailer involved. Be careful here. You may
be slugged with a charge if no faults are found, so do the
other checks first. Also, you may be unlucky with your
technician. Many a case has been marked as ‘no fault found’
when there was a glaring one. I have no magic solution to
this human problem.
Patch leads & RJ plugs
How long is the patch lead joining your modem/router to
the telephone point in your house? Such leads are usually
not twisted, probably because it is cheaper to make flat ones.
Although it’s a comparatively short run, the interferencecancelling effect of a flat pair is much less than a twisted
pair. I would recommend that patch leads be no longer
than two metres. While we are considering the patch lead,
internet problems may also be due to a loose RJ plug on
the patch lead, where it has not been fully pressed home
in the modem or telephone point.
Problems at given times
Internet users often experience drop-outs at specific intervals (every four hours is common) and these have turned
out to be Securitel or similar auto-paging alarm units on the
same line. For optimum internet speed and freedom from
drop-outs, it is best to have only telephones (and possibly
faxes), all with filters, on a line that is to carry ADSL traffic.
In one recent case of drop-outs exactly at 4-hourly inter28 Silicon Chip
In one case I assisted with, the user experienced a dramatic reduction of internet speed at 6PM every evening.
He was convinced that his ISP was deliberately throttling
his speed at peak times, that there were insufficient ISP
servers etc.
Fortunately, he was not one to rush into un-researched
blame and in due course he found by experiment that the
problem was his new plasma TV which was switched on
at 6PM for the news each evening. The switch-mode power
supply was poorly screened and filtered (as is often the
case) and it radiated pulses throughout the house and into
his untwisted patch lead.
RF interference can even be caused by a faulty lamp,
including an incandescent one that is ‘singing’ just before
failing. To test for this, appliances should be completely
disconnected one by one (not just switched to standby).
(But do not emulate a certain friend who noted that if he
switched off at his main fuse panel, the interference certainly stopped — but along with everything else…).
The sheer field strength of a nearby radio or TV broadcaster will result in some induced signals in nearby telephone lines. Your telephones are fitted with capacitors to
reduce or eliminate rectified audio currents from interfering
with normal telephony but the RF itself will often interfere
with ADSL.
Why does RF interference matter?
We’ve looked at the modem/DSLAM combination and
how it dynamically negotiates the best use of the big block
of bins available to it. If there is interference on one or
more of these bins, caused by harmonics or fundamental
frequencies, the modem/DSLAM will agree together not to
use it/them and so the total number in use reduces, leading
to reduced overall speed.
Power interruptions
Some time ago, I was puzzled about drop-outs that I was
suffering after upgrading from ADSL1 to ADSL2+ (a more
‘fussy’ technology, because the frequencies are higher).
Previously, if power visibly failed for a few seconds (as it
often did in thunderstorms), the modem would reset itself
and consequently there was delay. This I understood but
since the upgrade to ADSL2+, the drop-outs also happened
at times when there was no flicker of the lights.
A small Uninterruptible Power Supply (UPS) for the
modem and router solved this problem. Indeed, almost
immediately after fitting the UPS it went into alarm mode
several times, although the lights did not blink.
Here was the solution: power occasionally dipped for a
few milliseconds, not enough for an incandescent lamp to
flicker (because of thermal inertia) but certainly enough to
cause the modem and router to reset.
Modems and routers
Although I have been lucky myself, there are many reported cases of these units becoming flaky. There is a tendency
to leave them on continuously and thus heat fatigue may
occur, particularly when there are underrated electrolytic
capacitors fitted, which is the case with much overseas
siliconchip.com.au
equipment or when the mains voltage is abnormally high.
Modem and router firmware may need upgrading as
manufacturers discover shortcomings in current versions.
Sometimes settings may be lost for one reason or another
and the modem or router may have to be reset to the factory configuration.
Although it is not always practical, the best way of eliminating the modem or router from your list of suspects is to
borrow another; as long as you are quite sure you have set
it up properly. Sometimes when the Net connection fails
it is necessary only to perform a simple power reset of the
modem/router, where you disconnect the low voltage power
for 20 seconds or so (to let the electrolytics discharge) and
then power it up again.
After a few minutes to let the ADSL connection become
established, you may find that you have normal service.
Computer
My experience was that by far the biggest contributor
to low internet speeds was the state of my own computer.
Running XP it had become somewhat cluttered, with
much software installed or de-installed over the years.
Consequently the machine often ‘froze’. By pressing CTRLALT-DEL on an XP machine you can see its operating state,
via Windows Task Manager. The CPU usage is the critical
figure. If this sits at 100% for more than a few seconds then
the machine is in a virtual ‘locked up’ state and all your
computing will experience considerable delays.
In my case, I had sufficient backups and original software
CDs/DVDs to be able to clean off my hard disk and do a
complete reload of everything. And I had another machine
to work on while this was in progress.
I admit that it took many hours and I know it’s not a luxury
open to all but in my case when it was finished my internet
speeds were revolutionised! When working at the computer
I now make a point of calling up Windows Task Manager
and then minimising it. On the task bar I can then see the
loading of the CPU as a partially green square all the time.
Possibly, you are not able to perform (or risk) a complete
reload from the start but at least, after taking good backups
and verifying them, check your machine for viruses or other
malware. Also, avoid having other applications running
when on the Net and defrag your disk from time to time.
General infrastructure failure
The telephone network that we rely on so heavily can
experience local or even national outages for a number
of reasons. You can be sure that every effort is made to
minimise the effects of this, especially if over a wide area,
because the telco concerned is losing valuable revenue.
A reputable ISP will have a page on its website that will
list outages and general locations (although admittedly this
is of no value when you can’t access it!). Phone help lines
may also be a source of information. (I hear you groan —
but remember that the help-desk jockey is a human being
reading from a script).
We now come to less likely reasons for poor speeds or for
drop-outs. Exchange DSLAMs are very reliable, since they
feed many subscribers, perhaps up to 1,000 or more. If one
is faulty it will give rise to many simultaneous complaints
and is likely to be swapped out quickly.
Exchange congestion is a fairly rare occurrence, though
some remote exchanges are notorious. Again, this is a leaksiliconchip.com.au
Portable Performance
for Challenging
Environments
Now you have the best of both worlds with the new THS3000 Series
of handheld oscilloscopes from Tektronix. Whether you need to measure
low voltage control signals or high voltage differential signals, the
THS3000 is a handheld oscilloscope that thrives in both environments.
Up to 200 MHz bandwidth and 5 GS/s sample rate
4 isolated channels and up to 7 hours of battery life
21 automated measurement & waveform math Validate your device operation and identify issues quickly
Data logging capability Find intermittent faults
• TrendPlot™ function
• capturing and replaying up to 100 display screens
or set-up pass/fail waveform limits
The THS3000 series of handheld oscilloscopes is a complete and
comprehensive solution providing you with the ability to make
safe, accurate measurements in both bench and field
environments
Visit TekMark Australia, Booth 5306 in NMW 2012 for
demonstration. Alternatively, contact us on 1300 811 355 or email
enquiries<at>tekmarkgroup.com
age of potential revenue for the telco and the likelihood is
that it will be fixed fairly quickly. Unfortunately, in these
cases ‘fairly quickly’ for a telco may mean a number of
months.
Incompetent or poorly resourced ISPs
I have placed this last because although it is possible I
have never experienced it. I can only think that in this case
subscribers would leave in droves and the ISP concerned
would fold.
How can you tell if your ISP is the guilty party? I would
like to say be guided by forum and newsgroup postings.
Unfortunately this is not completely reliable. In the first
place, among the millions of users there will be many
who are experiencing your problem, even in your own
immediate area, and are convinced they know the causes
(invariably they cite the ISP).
Some posters pop up under cover of a different user
name to their normal one, make a disparaging post and then
disappear, possibly returning under another user name and
agreeing with their own post.
Are they paid stooges from another ISP? Or are they
genuinely frustrated and distressed? There is no way of
checking. You could go by praise but to be fair, praise posts
could also be made by stooges!
I can only be certain about my own experience. To paraphrase a certain person recently in the news, I am a happy
little Vegemite now that I’ve cleaned up my computer and
fixed my power outage problem. And all without any helpdesk jockeys being harmed…
SC
May 2012 29
By NICHOLAS VINEN
PIC/AVR Programming
Adaptor Board
Do you frequently program microcontrollers with a serial
programmer? Want to streamline the process so you can quickly
do virtually any micro? Well now you can! Our new Programming
Adaptor Board, in combination with an In-Circuit Serial
Programmer (ICSP), allows you to program most 8-bit & 16-bit PIC
microcontrollers as well as 8-bit Atmel AVRs. It has a 40-pin ZIF
socket and is configured with just a few DIP switches.
M
OST EMBEDDED developers
program their microcontrollers
using an In-Circuit Serial Programmer
such as the Microchip PICkit3 or the
Atmel AVRISP MkII. These plug into
the USB port on your PC and a header
on the development board. The PC
software (eg, Microchip MPLAB or
Atmel AVR Studio) is then used to
30 Silicon Chip
program or re-program the microcontroller.
This is handy while developing the
project but you won’t always have a
complete circuit with a programming
header when you need to program a
micro.
It may be that the circuit operates at
230VAC mains potential and so you
can’t safely plug a programmer in.
Or perhaps the circuit connects the
micro’s programming pins to other
components which interfere with onboard programming. Maybe there just
isn’t room for the programming header
on the board because it wouldn’t fit
or there is one but you can’t get to it
once the board is mounted in its case.
siliconchip.com.au
So often, it’s just more convenient
to pop the micro out and take it to a
computer for programming.
In short, there are lots of reasons
why you might want to program a
micro but an in-circuit programmer
alone won’t do the job. That’s where
this board comes in. It forms a circuit
for the microcontroller to operate in
and provides the programming header
connection and power supply. Once
it’s set up and the micro is locked into
the ZIF socket, you fire up the serial
programmer and program the chip as
per usual.
At SILICON CHIP we used to wire up
a socket on Veroboard every time we
wanted to program a new chip but this
is a pain. There are so many different
pin configurations and power supply
requirements that you end up with
dozens of the things floating around.
You also have to bend the IC pins to
get it into a standard socket and then
it can be difficult to pull out without
mangling them.
Some programming adaptor boards
available on the internet use multiple
ZIF sockets to suit different micros.
Unfortunately, good ZIF sockets are
quite expensive so these boards usually use cheap ones which don’t last
very long. And you’d need an awful
lot of them to support a large portion
of the PIC range.
Features
This programming board supports
over 400 different 8-bit and 16-bit
PICs – around 90% of the currently
available range. It also supports about
45 different Atmel microcontrollers,
covering most of the popular ATtiny
and ATmega micros. It is capable of
programming the vast majority of
microcontrollers used in SILICON CHIP
projects in the last 10 years or so.
The programming adaptor board
has a power supply since not all ICSP
units can supply power to the micro.
It also has soft-power control with
over-current/short-circuit protection,
to prevent damage to the micro in case
something goes wrong.
The on-board power supply can
provide 3.3V or 5V, depending on what
the micro to be programmed needs. In
addition, the micro is always inserted
into the ZIF socket with pin 1 at upper
left, making it easy to use.
Design
Before drawing up the circuit, we
siliconchip.com.au
Main Features & Supported Microcontrollers
Features
•
•
•
•
•
•
•
Runs off a 9-12V DC plugpack or USB 5V power
Programs most Microchip PICs and Atmel AVR microcontrollers in DIP
Selectable 3.3V or 5V micro power supply
Easy configuration – chip type selected with 8-way DIP switch
Electronic fuse protects micro
Uses high-quality, reliable 40-pin universal ZIF socket
Compatible with PICkit3 and AVRISP MkII
Supported Microcontrollers
•
•
•
•
Virtually all PIC12s, PIC16s and PIC18s
Most PIC24s and dsPIC33s
Most ATtinys and ATmegas
Over 450 different microcontrollers supported – see panel on page 34
surveyed the entire range of 8-bit and
16-bit PICs and AVRs to figure out
what proportion of the range we could
support. There are nearly 500 different PICs in DIP packages with about
30 different pin configurations. The
AVR range is smaller, with less than
100 parts and eight different pin-outs.
Supporting them all is a huge ask
but we figured that with 17 different
pin configurations (13 for PICs and
four for AVRs) we could cover about
90%, including all the most popular
micros ranging from 8-pin up to 40pin DIP parts.
We have to connect different pins to
VCC and ground, depending on which
micro is inserted. We also need to route
the programming signals and voltages
to the appropriate pins. For AVRs,
it’s also useful to be able to drive the
clock pins with a square wave during
programming as they don’t automatically switch to the internal oscillator
in programming mode (unlike PICs).
Having sorted out what was required, the next question was how to
achieve it. Essentially, what we need
is a type of sparse crossbar or matrix
switch – think of a telephone exchange. We have a 40-pin socket, two
power supplies rails (0V and 3.3V/5V),
three or four programming signals/
voltages and a couple of clock signals
(we’ll explain that later). We need to
connect some combination of these for
a given micro and ideally this should
not involve a lot of effort for the user.
There are three obvious ways to
do it: using jumper shunts, relays or
electronic switches. We ruled the relay
option out almost immediately; we
would need at least 50 relays and it
would have been a huge PCB.
Jumper shunts would be a cheap
and cheerful solution but then you, the
user, would have to spend time reconfiguring the board one pin at a time,
based on a whole series of diagrams.
That would be a recipe for a disaster
and besides, technology is supposed
to make your life easier!
So we decided on electronic switching using Mosfets. They are quite small
and cheap and can easily be controlled
by digital logic, making configuration
a snap.
Circuit description
The resulting circuit is quite complicated, due to the large number of
different configurations and how many
pins need to be connected for each. So
we have broken it up into sections,
with Fig.1 showing the power supply
switching and Fig.2 showing the control logic and serial data multiplexing.
First, let’s examine IC1-IC3 in Fig.2.
We are switching the serial programming signals using CMOS 1-to-8
analog multiplexer ICs (4051B). There
are two such signals for PICs (PGD
and PGC) and three for Atmel AVRs
(MOSI, MISO and SCK). To simplify
the circuit, we join PGD with MOSI
and PGC with SCK; only one set is
used at a time. These three programming lines are connected to the “Z”
terminals on the ICs.
The active-low enable pins of these
three ICs (EN-bar) are joined together.
When they are pulled low, the “Z” terMay 2012 31
minals are connected to one of the “Y”
terminals. Which one depends on the
state of input pins A0-A2. If A0-A2 are
all low for one IC, its “Z” is connected
to “Y0”. If S0 is high and the rest are
low, giving a binary input of 1, “Z” is
connected to “Y1” and so on.
We have specified HEF
4 051Bs,
which are pin-compatible with the
original 4051Bs but have half the
on-resistance between connected
terminals (40Ω).This is important for
reasons that will be explained later.
The first three DIP switches in S1,
labelled DIP0-DIP2, drive the A0-A2
inputs of these three ICs. The Y0-Y7
terminals of each are connected so that
for each combination of DIP0-DIP2,
one of the programming headers is
connected to the appropriate pins for
one type of micro. EN-bar is driven
low simultaneously with the micro
power supply being switched on, so
that when the micro has no power, the
programmer is disconnected.
IC1-IC3 run from 16V, slightly
higher than the normally specified 15V
but below the 18V maximum. They
can therefore not only pass the 3.3V or
5V digital signals but also withstand
the 13.5V which can be applied to the
MCLR/VPP pin when programming a
PIC. In some cases, pin 1 is connected
to VPP and this pin is also connected
to IC1, so it must be able to withstand
this voltage.
Each programming pin connected to
IC1-IC3 is also wired to a dual Schottky
diode which is connected between the
supply rails (D6-D8). These prevent
the programmer (connected via CON1
or CON2) from driving the terminals
of IC1-IC3 beyond their supply rails
when the adaptor board power supply
is switched off.
Programming voltage
The 13.5V mentioned earlier comes
from the VPP pin of CON1, the ICSP
header for PICs. This is used to power
the micro’s internal flash programming
circuit.
Because the PIC draws some current
from this rail during programming,
we can’t use another HEF4051BT to
route it, since the 40Ω series resistance
would be an issue. Instead, we use
discrete analog switches comprising
surface-mount dual Mosfets Q18-Q21
– see upper left of Fig.1.
Each pair is connected drain-todrain. One of the source terminals is
connected to VPP on the programming
32 Silicon Chip
socket while the other is connected
to one of pins 1, 4, 9 or 10 on the ZIF
socket. The gates are tied together.
When the gates are held at 0V,
both Mosfets are off since the source
voltages are never below ground
(0V). The body diodes are connected
cathode-to-cathode so at least one is
reverse biased and no current can flow
through them either.
When the gates are pulled up
together to +16V, the gate-source
voltages will be in the range of 2.516V, depending on the source voltages.
These are in the range of 0-13.5V.
Even with just 2.5V between gate and
source, the FDS6912A Mosfets switch
on, applying VPP to the connected
pin. If the programmer pulls VPP low,
the micro pin will also go low as the
analog switch allows current to flow
in either direction.
When programming Atmel AVR
microcontrollers, the reset pin is also
used but the programmer only needs
to pull it low, to enable the micro’s
programming mode. We have provided
a reset pushbutton (S4 in Fig.2) which
also pulls this line to ground. Sometimes, when a micro is already running
code, you need to do this before you
initiate programming.
There is a further difference with Atmel chips. If they have been configured
to run from a crystal, ceramic resonator or external clock, they expect this
to be present during programming as
well as normal operation. This is in
contrast to PICs which automatically
switch to their internal oscillator when
programming mode is enabled.
So that you can still program chips
set up in this way (and many will be),
the adaptor board can supply a clock
signal to the micro. This works even
if it is expecting an external crystal; as
long as it gets a square wave on both
clock pins it will operate.
This facility is provided by IC4 and
IC5 which are also 4051Bs. When they
are enabled, they apply the 1MHz
square waves CLK and CLK-bar to the
XTAL1 (clock input) and XTAL2 (clock
output) pins of the micro. They are
automatically disabled while programming PICs; CLKENA-bar and hence
their EN-bar inputs are kept high.
Switching power
We also need to supply power to
the micro. Some micros have a single
pair of power supply pins (VCC and
GND) while others can have multiples
Fig.1: the ZIF socket and power supply
switching (multiplexing) section of the
programmer. The micro is placed in
the 40-pin ZIF socket and Mosfets Q1a/
b-Q25a/b connect the various pins to
VCC, GND and MCLR/VPP as required.
Some Mosfets also connect capacitors
between pairs of pins as necessaey. In
addition, Mosfets Q26-Q29 connect
serial programming lines PGC and
PGD to pins 39 & 40 respectively, for
high-speed programming of certain
PICs.
of each. These pins must have a low
source impedance at high frequencies
(1MHz+) or else the micro will not
operate correctly.
Normally, this is achieved by connecting power supply bypass capacitors between each pair of supply pins.
But we can’t put capacitors directly
between ZIF socket pins because while
they may be used to supply power
for one type of micro, another may
use the same pins for serial programming. Large value capacitors would
just shunt the programming signals
to ground.
Instead, we have connected multiple low-ESR bypass capacitors between the VCCS (micro power supply)
and GND rails around the ZIF socket.
We then switch those rails directly
to the appropriate socket pins using
low on-resistance Mosfets. The static
on-resistance for the FDS6912As is
about 0.02Ω and this is effectively in
series with the bypass capacitor ESRs,
for both VCC and GND pins. The total
supply impedance is therefore quite
low (<0.1Ω).
In most cases, a single Mosfet
switches power to one of the micro
pins. For example, Q10a (right side)
connects pin 36 to ground for MODE
2 while Q3b connects the same pin to
VCCS in MODE 3. In both cases, the gate
is pulled to +16V to turn the Mosfet
on and to 0V to turn it off.
However, for pins which share VPP
(~13.5V) and VCCS (3.3V/5V), two
Mosfets are connected drain-to-drain
for VCCS, just as they are to supply VPP.
For example, Q22a & Q22b (upper left)
connect pin 1 to VCCS. This is necessary to prevent the higher VPP voltage
from feeding back into VCCS when it
is turned on.
In total, there are 13 Mosfets connecting various pins to VCCS and 12 for
GND. Then there are an additional six
siliconchip.com.au
G
S
+2.5V
Q16b
D
MODE 9C
G
S
S
40
39
38
37
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
Q3b
Q24a
Q10b
D
Q4b
Q15b
D
Q25a
S
220nF
G
D
G
S
Q23b
Q5b
G
FROM IC3 PIN 1
S
G
S
S
MODE 1,6,7A,7C,9
G
MODE 6
D
FROM IC2 PIN 1
Q14a
Q23a
G
FROM IC1 PIN 1
FROM IC5 PIN 12
S
FROM IC4 PIN 12
S
VPP
PIC/AVR PROGRAMMER
siliconchip.com.au
G
S
S
Q11a
MODE 2
D D
G
D
MODE 8
S
MODE 1
S
D
G
Q26-29: 2N7002P
G
S
G
MODE 6
S
D
ZIF SOCKET & MULTIPLEXING
Vccs
D
VPP
G
Q8b
D D
Q2a
S
S
Q11b
MODE 6
S
Q2b
G
Q21b
Vccs
Q21a
Q20a
G
Q20b
S
G
MODE 7A
D
D
MODE 7
MODE 1,4,9
S
D
Q14b
Q7b
G
Vccs
FROM IC4 PIN15
SC
S
MODE 4
G
S
MODE 7B,7C
S
Q8a
G
D
10F
D
D D
D D
G
D
G
S
D
MODE 1
S
G
G
D
Q25b
Vccs
D
G
MODE 2
S
Q15a
Q12b
Q12a
G
S
G
D
S
D
S
G
MODE 3
S
D
G
Q1a
G
Vccs
D
S
MODE 1A
D
G
2012
FROM IC2 PIN 13, IC3 PIN 5
FROM IC1 PIN 13
10F
S
FROM IC3 PIN2,
IC4 PIN14
FROM IC5 PIN14
FROM IC1 PIN 14
S
S
MODE 4
MODE 6
G
FROM IC1 PIN 12
FROM IC3 PIN 12
D D
MODE 9A
MODE 5A
Q5a
S
G
MODE 4,7,9
G
FROM IC1 PIN 5, IC3 PINS 13,14
10F
220nF
FROM IC2 PIN2,
IC5 PIN15
Q9a
S
Q7a
G
MODE 9B,9C
MODE 5
S
D
MODE 2
G
Q3a
Q27
Q29
MODE 1,9
Q24b
FROM IC1 PIN 2
G
S
Q10a
Q6a
D
G
MODE 0,7,8
Q4a
S
MODE 3
G
470nF
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
D
G
S
S
Q1b
FROM IC3 PIN4,
IC5 PIN1
S
D
40-PIN ZIF
SOCKET
FROM IC3 PIN 15
MODE 2
10F
D D
S
D
MODE 0
FROM IC2 PIN 15,
IC4 PIN13
G
Q6b
FROM IC1 PIN 15
MODE 7
G
Vccs
D
Q13b
G
G
S
D
D D
PGD
S
Q13a
FROM IC1 PIN 4,
IC4 PIN1
VPP
S
G
MODE 3,5
Vccs
G
D D
Q28
G
Q9b
S
Q19b
S
G
D D
D
Q19a
D
PGC
S
4x
100nF
G
MODE 5
Q18a
Q22a
D D
G
Vccs
S
Q18b
Q22b
S
Q26
VPP
MODE 0,1,4,7,8,9 G
G
Q1-25: FDS6912A
Db
Db
Da
Da
Gb
Sb
SaGa
May 2012 33
Supported Microcontrollers
[x = any digit, (A) = with or without A suffix]
Microchip
PIC12F, PIC12HV: All [25]
PIC16F, PIC16LF, PIC16HV: All but PIC16F57 and PIC16F59 [149]
PIC18L, PIC18LF: All* [132]
PIC24E: All [8]
PIC24F: All** but PIC24F04KA20x, PIC24F04KLx00 and PIC24FJ16MC101 [29]
PIC24H: All but PIC24HJ12GP20x [14]
dsPIC33E: All [12]
dsPIC33F: All but dsPIC33FJxxGSxxx and those ending with -101 [26]
* PIC18F2331, PIC18F44J10 and PIC18F45J10 require an extra component in the ZIF
socket.
** PIC24FxxKA30x (12 types) require an extra component in the ZIF socket.
Atmel
ATtiny13(A)(V), ATtiny15L, Attiny25/45/85(V) [10]
ATtiny261/461/861(A)(V) , Attiny26(L) [11]
ATtiny2313(A)(V), Attiny4313 [4]
ATtiny48/88, Atmega48/88/168/328(P)(A)(V), Atmega8(A)(L) [21]
ATmega16/32(A)(L), Atmega164/324/644/1284(P)(A)(V) [20]
ATmega8535(L) [2]
Total: 463 fully supported, 15 programmable with additional components Note: some
parts no longer in production have not been checked but are likely to work.
dual Mosfets (ie, a total of 12) which
connect various capacitors between
pins as required for some micros.
These latter FETs are all configured
as analog switches, in series with the
capacitors. For example, Q12 (left side)
connects a 10µF capacitor between
pins 6 and 8, to filter the core supply
voltage of certain micros (dsPIC33s
and PIC24s).
Q16b supplies 2.5V to pin 6 if required. Q14a is unused and has its
gate and source tied to ground. Three
additional Mosfets are used in the
power supply, to be described later.
In total, there are 25 FDS6912A dual
Mosfets surrounding the ZIF socket.
This may seem like a lot but they are
relatively cheap.
Parasitic capacitance
The problem with all these Mosfets
is that even when switched off, they
are effectively still present in the
circuit. While the drain-source leakage current is very low and can be
ignored, the output capacitance is an
issue. This refers to the capacitance
seen at the drain pin which is the sum
of the drain-source and drain-gate
capacitances.
This capacitance is highest (about
34 Silicon Chip
1nF) when the Mosfet’s drain-source
voltage is zero. As the drain-source
voltage increases, it drops to about
150pF. With multiple Mosfets on a
single pin, this can add up and in
combination with the 40Ω resistance
of the analog multiplexer ICs, IC1-IC5,
it forms low-pass filters for the serial
programming and clock signals. This
limits the signals passed to a maximum
of about 1.5MHz.
In most cases, this is not a problem.
We tested the programming adaptor
board with a variety of PIC and AVR
chips (about 20 different types), using the PICkit3 and AVRISP MkII
serial programmers. We found the
programming speed was typically
around 0.5MHz and it worked reliably
in each mode.
There is one situation where the
parasitic capacitance is an issue and
this is when programming PIC18FJ devices. The PICkit3 uses a higher clock
frequency for these, of about 2MHz.
It is therefore necessary to have four
small additional Mosfets, Q26-Q29.
They form analog switches in parallel with IC1 and IC3, for routing the
PGD and PGC signals to pins 40 and
39 respectively.
For PIC18FJs then, the series re-
sistance to the programming signals
drops to a couple of ohms, allowing
the higher frequency signals to pass
through.
These four additional Mosfets are
2N7002Ps. The 2N7002 is a surfacemount version of the 2N7000. The P
suffix is important as it indicates a
lower on-resistance (1Ω compared to
2.5Ω) which is required for reliable
programming of PIC18FJs.
Control logic
All this switching is controlled
by the circuitry shown in Fig.2. To
program a particular micro, eight DIP
switches (S1) are set to the appropriate positions. Each DIP switch is connected to a pull-down resistor, so if the
DIP switch is off, the corresponding
line labelled DIP0-DIP7 is 0V and if
the switch is on, the line is at 16V.
DIP0-DIP3 configure the analog multiplexers IC1-IC5, described earlier,
They are also connected to the four
inputs (A0-A3) of CMOS 4028B BCDto-decimal decoder IC6. Depending
on the positions of DIP0-DIP3, one of
its 10 outputs (O0-O9) is high and the
rest are low. These outputs then drive
the gates of some of the Mosfets shown
in Fig.1, turning the appropriate ones
on for that mode.
For example, in MODE 5, output O5
of IC6 goes high and turns on Mosfet
Q9a, which connects pin 40 of the ZIF
socket to ground. Some Mosfets must
be turned on in more than one mode
and so the 10 mode lines are also fed
into nine OR gates: IC7a-IC7c, IC8aIC8c and IC9a-IC9c. In some cases,
these are cascaded. So when MODE 1,
4 or 9 is selected, the output of IC9b
(MODE 1,4,9) is high and this turns
on Mosfet Q4b, supplying 3.3V or 5V
to pin 32 of the micro.
But the scheme described above
only gives us 10 possible pin configurations and as we said earlier,
we need 17. The additional seven
configurations use the same power
and programming pins as the other 10
but involve different combinations of
capacitors connected between other
pins and, in one case, an additional
2.5V supply.
The extra control signals are derived
from the 10 mode signals and the
positions of two more DIP switches,
DIP4 and DIP5. This is achieved using
eight 2-input AND gates, IC10a-IC10d
and IC11a-IC11d, in combination with
inverter gates IC12a-IC12c and 2-input
siliconchip.com.au
Parts List
1 PCB, code 24105121, 116 x
127mm (available from SILICON
CHIP)
1 40-pin Universal ZIF socket,
0.6-inch wide pin spacing
(Element14 1169111)
1 40-pin production DIL socket,
0.6-inch wide pin spacing
1 220µH bobbin inductor (Jaycar
LF1104, Altronics L 6225)
1 1MHz or 1.008MHz crystal or
2-pin ceramic resonator (Rockby
10234, 13390 or 10233)
1 100kΩ 9-pin 8-resistor network
(Element14 9356827)
1 6-way pin header strip (CON1)
2 10-way shrouded vertical IDC
socket (CON2)
1 PCB-mount USB type B socket
(CON3)
1 PCB-mount DC socket (CON4)
1 2-way polarised header,
2.54mm pitch (CON5)
1 3-way pin header strip and
shorting block (LK2)
1 8-way DIP switch (S1) (Element14 1123941; Jaycar SM1024; Altronics S3060)
3 PCB-mount tactile pushbuttons
(S2-S4)
1 miniature PCB-mount SPDT
slide switch (S5) (Element14
1123875)
5 M3 x 6mm machine screws
1 M3 shakeproof washer
1 M3 nut
4 M3 x 12mm tapped Nylon
spacers
Semiconductors
5 HEF4051BT 8-way analog
multiplexers [SOIC-16] (IC1IC5) (Element14 1201291)
OR gate IC7d. The additional modes
are labelled A, B and C and are selected
by switching DIP4 on (mode A), DIP
5 on (mode B) or both on (mode C).
If MODE 9 is selected and both DIP4
and DIP5 are on (high), the output of
AND gate IC10a (MODE 9B,9C) is high,
as is the output of AND gate IC10c
(MODE 9A,9C). As a result, the output of IC10b (MODE 9C) is also high.
With MODE 9B,9C and MODE 9C both
high, Mosfets Q12a and Q12b connect
a 10µF capacitor between pins 8 and
6 while at the same time, Q16b turns
siliconchip.com.au
2 CD4028BM BCD-to-decimal
decoders [SOIC-16] (IC6,
IC17) (Element14 1753401)
2 HEF4071BT quad 2-input OR
gates [SOIC-14] (IC7, IC8)
(Element14 1085289)
1 CD4075BM triple 3-input OR
gate [SOIC-14] (IC9) (Element14 1739910)
2 HEF4081BT quad 2-input AND
gates [SOIC-14] (IC10, IC11)
(Element14 1085290)
1 HEF4069UBT hex inverter
[SOIC-14] (IC12) (Element14
1201295)
1 SN74HC04D hex inverter
[SOIC-14] (IC13) (Element14
1311424)
1 HCF4013BM1 dual D-type
flipflop [SOIC-14] (IC14) (Element14 1094187)
1 OP07CD precision op amp
[SOIC-8] (IC15) (Element14
1575526)
1 LM293D dual low-power
comparator [SOIC-8] (IC16)
(Element14 2292944)
1 7805T 5V 1A linear regulator
(REG1)
1 AP1117E33 3.3V low-dropout
linear regulator [SOT-223]
(REG2) (Element14 1825291)
1 SPX1117M3-L-2-5 2.5V
low-dropout linear regulator
[SOT-223] (REG3) (Element14
1831943)
1 MC34063ADG switchmode
controller [SOIC-8] (REG4)
(Element14 1211119)
25 FDS6912A dual independent
N-channel Mosfets [SOIC8] (Q1-Q25) (Element14
1095019)
4 2N7002P N-channel Mosfets
[SOT-23] (Q26-29) (Element14
1859848)
on, supplying 2.5V to pin 6. This suits
PIC18LF2xJ5x microcontrollers.
Inverter stages IC12e and IC12f are
unused, so their inputs are tied to +16V
to prevent oscillation. Two-input OR
gate IC8d is also unused and connected
similarly.
a 1MΩ biasing resistor and a 4.7kΩ
current-limiting resistor. The 1MHz
clock signal is then buffered by IC13b
and IC13f which are paralleled for
increased drive strength. This signal
then passes to IC5, to be connected to
the micro’s XTAL1 (clock input) pin,
when enabled.
This clock signal is inverted again,
by IC13c-IC13e (also paralleled) and
this signal passes to IC4, which routes
it to the micro’s XTAL2 (clock output)
pin, if enabled.
Both inverters must charge and
Clock generator
Fig.2 also shows the crystal oscillator circuit which is based on hex
inverter IC13, a 74HC04D. IC13a forms
the oscillator in combination with
crystal X1, two 33pF load capacitors,
Diodes & LEDs
1 1N5819 1A Schottky diode
(D1)
3 1N4148 small diodes (D2-D4)
1 1N4004 1A diode (D5)
3 BAT54S dual series Schottky
diodes [SOT-23] (D6-D8; Element14 1467519)
1 Green 3mm LED (LED1)
1 Yellow 3mm LED (LED2)
1 Red 3mm LED (LED3)
Capacitors
4 100µF 16V electrolytic
2 47µF 25V electrolytic
1 10µF 16V electrolytic
4 10µF 6.3V SMD X5R ceramic [3216/1206] (Element14
1833825)
1 470nF MKT
2 220nF 50V SMD X7R ceramic [3216/1206] (Element14
1362557)
11 100nF 50V SMD X7R ceramic [3216/1206] (Element14
1301906)
14 100nF MKT
1 470pF disc ceramic
2 33pF disc ceramic
Resistors (1%, 0.25W)
1 1MΩ
1 4.7kΩ
1 100kΩ
3 2.2kΩ
1 68kΩ
1 1.1kΩ
5 47kΩ
2 1kΩ
1 13kΩ
1 220Ω
1 1Ω (1% or 5%)
1 0.1Ω SMD [3216/1206] (Element14 1865244)
May 2012 35
2
6
4
3
1
8
5
10 12 14 16
7
100nF
S1
DIP
SWITCH
16
Vdd
DIP7
O7
DIP6
O6
11
DIP3
12
DIP2
13
DIP1
10
DIP0
8
5 MODE 9
A3
O3
A2
O2
A1
O1
A0
O0
Vss
8
RN1
8x100k
10
9 MODE 8
4 MODE 7
2
7 MODE 6
IC12c
IC10a
5
15 MODE 3
2 MODE 2
12
14 MODE 1
1
3 MODE 0
IC11d
4.7k
X1 1.0MHz
33pF
14
11
IC12a
IC11a
3
1
2
IC13b
5
9
11
IC13c
IC13d
IC13e
14
13
14
IC7c
2
1
IC9b
11
MODE 9
MODE 7
13
12
IC9c
MODE 8
6
1
7 IC8d
IC8a
6
3
SC
10
MODE 4,7,9
IC8b
4
MODE 0,1,4,7,8,9
MODE 5
9
MODE 3
8
IC8c
MODE 3,5
10
+16V
CLK
10
100nF
8
9
DIP4
MODE 7
10
MODE 0
IC12f
MODE 8 5
6
1
RESET
S4
3
D6
2
PIC/AVR PROGRAMMER
IC7b
4
1
14
IC7a
5
3
4
5
MODE 1A
MODE 0,7,8
MODE 0,8
7
+16V
Vdd
1
6
4
2
VPP
PGD
PGC
MODE 5A
IC11b
CON1
1
2
3
10
IC11c
6
MODE 1
12
PIC ICSP
2012
MODE 1,4,9
5
8
+16V
IC10, IC11: 4081B
IC7, IC8: 4071B
IC9: 4075B
IC12: 4069B
IC13: 74HC04D
6
7
2
IC12e
11
9 MODE 1,6,7A,7C,9
IC9a
MODE 9
MODE 4
CLK
4
14
4
5
3
100nF
11
MODE 7A
MODE 1,9
8
MODE 6
12
100nF
12
7
10
MODE 5
14
4
7
100nF
+16V
13
12
IC12b
11 3
+16V
MODE 1 8
7
33pF
IC7d
7
MODE 9 9
IC13f
3
7
6
13
2
100nF
13
MODE 9A
IC10d
MODE 7A,7C
1M
2
IC10b
MODE 7B,7C
13
Vcc
IC13a
MODE 9B,9C
MODE 9C
11
13
100nF
1
4
6
12
3
6 MODE 5
14
5
IC10c
1
IC6 O5
4028B O4 1 MODE 4
DIP5
DIP4
9
DIP4
DIP5
O9
O8
9 11 13 15
100nF
3
1
D7
2
3
D8
2
2
4
6
8
10
CON2
1 MOSI
3
5 RESET
7 SCK
9 MISO
AVR ICSP
CONTROL LOGIC
Fig.2: the control logic for the adaptor board is shown at left, while IC1-IC5 (HEF4051Bs) connect the serial programming
and clock lines to various pins on the ZIF socket (see Fig.1). 8-way DIP switch S1 selects the micro to be programmed and
the switch states are decoded using the various logic ICs, to drive the appropriate Mosfets and analog switches.
36 Silicon Chip
siliconchip.com.au
+16V
100nF
16
Vdd
CLKENA
6
DIP3
9
Y7
Y6
EN
Y5
A2
Y4
A0
Y2
4
2
5
ZIF SOCKET PIN 5
1
IC5
10
12
A1 4051B Y3
DIP2
11
CLK
3
Y1
Z
Vss
8
Vee
7
Y0
ZIF SOCKET PIN 13
15
ZIF SOCKET PIN 7
14
ZIF SOCKET PIN 9
13
+16V
16
Vdd
6
9
100nF
Y7
Y6
EN
Y5
A2
Y4
A0
Y2
Z
Y0
4
2
5
ZIF SOCKET PIN 4
1
IC4
10
12
A1 4051B Y3
DIP1
11
CLK
3
Y1
Vss
8
Vee
7
ZIF SOCKET PIN 12
15
ZIF SOCKET PIN 10
14
ZIF SOCKET PIN 8
13
ZIF SOCKET PIN 2
Power supply
+16V
ENABLE
DIP0
16
Vdd
DIP1
DIP2
6
9
100nF
Y7
Y6
EN
Y5
A2
Y4
4
2
ZIF SOCKET PIN 6
5
ZIF SOCKET PIN 39
1
ZIF SOCKET PIN 29
IC1
10
12
A1 4051B Y3
11
3
Y2
A0
Y1
Z
Vss
8
Vee
7
Y0
ZIF SOCKET PIN 35
15
ZIF SOCKET PIN 1
14
ZIF SOCKET PIN 40
13
ZIF SOCKET PIN 37
+16V
16
Vdd
6
9
100nF
Y7
Y6
EN
Y5
A2
Y4
A0
Y2
Z
Y0
4
2
5
ZIF SOCKET PIN 30
1
IC2
10
12
A1 4051B Y3
11
3
Y1
Vss
8
Vee
7
15
14
ZIF SOCKET PIN 38
13
+16V
16
Vdd
6
9
10
11
3
100nF
Y7
Y6
EN
Y5
A2
Y4
A0
Y2
4
2
5
1
IC3
12
A1 4051B Y3
Y1
Z
siliconchip.com.au
Vss
8
Vee
7
Y0
15
ZIF SOCKET PIN 31
ZIF SOCKET PIN 34
D6-7-8: BAT54S
3
1
2
Refer now to Fig.3 which shows the
power supply. The unit can run from
either a 9-12V DC plugpack or a USB
port. The plugpack is connected to
CON4 and this disconnects the USB
ground pin so that power can’t flow
back into the USB port. D5 provides
reverse polarity protection and REG1
then drops the supply voltage to the
required 5V. For USB, 5V is drawn
straight from the socket.
Either way, slide switch S5 acts as
the power switch and when on, green
LED1 lights up. The 5V rail is reduced
to 3.3V by REG2, a low-dropout (LDO)
linear regulator. These 5V and 3.3V
rails provide the two power options
for the micro.
The 5V rail also powers REG4, an
MC34063 switchmode regulator. This
switches current through inductor L1
(a 220µH choke) and in combination
with Schottky diode D1, generates the
+16V logic supply. This only needs to
deliver a few milliamps since the logic
is all static. The ratio of the 13kΩ and
1.1kΩ resistors sets the output voltage
to 1.25 x (13kΩ ÷ 1.1kΩ + 1) = 16.02V.
LK1 allows the power supply to be
tested before voltage is applied to the
rest of the circuitry. This is shorted for
normal operation.
Voltage selection
ZIF SOCKET PIN 3
14
13
discharge the parasitic capacitance
at the target pin at 1MHz. This could
be a couple of nanofarads. Their load
impedance can be up to 40Ω + (1 ÷ (2π
x 1MHz x 2nF)) = 120Ω, hence the use
of multiple inverters in parallel.
Unfortunately, 1MHz crystals are
not as common as 2MHz crystals.
The circuit will work with a 2MHz
crystal but the dissipation in IC4 and
IC5 increases due to the increased
current required to drive the load
capacitance at the higher frequency.
We did not experience any failures in
our prototypes but cannot vouch for
the long-term reliability of the circuit
if using such a crystal.
If you do use a 2MHz crystal, avoid
leaving the clock and micro power
enabled for long periods, when programming at 5V. This is not an issue
when programming PICs.
Mosfets Q17a and Q17b switch the
3.3V and 5V rails to the micro respectively; only one can be on at a time.
Q17a turns on when DIP6 is high but
Q17b is only indirectly controlled by
May 2012 37
38 Silicon Chip
siliconchip.com.au
Ct
GND
4
Cin5
REG4
MC34063
SwE
A
+2.2V
Vcc
100nF
K
A
1k
(Rshunt)
0.1
1.1k
13k
K
POWER SUPPLY
47k
D2
1N4148
CON5
1
2
IC15: OP07CD
2
1
K
A
D1
1N5819
LED1
+2.2V
1k
S5 POWER
100F
16V
L1
220H
SwC
8
DrC
7
Ips
6
220
GND
OUT
Vcc
1
IN
REG1 7805
PIC/AVR PROGRAMMER
470pF
3
100nF
47F
25V
K
6
5
3
2
2
3
7
3
1
6
4
IC16b
7
+5V
47k
IC16: LM293D
IC16a
8
1
8
TPG
1
+3.3V
A
DIP6
MODE7
S
D
K
47k
2.2k
G
A
K
11
10
13
12
A
O6
O7
O8
O9
Vss
5
6
7
4
9
S
D
R
Q
Q
2
1
100nF
Q
LED3
13
IC14: 4013B
IC14b
CLK
IC14a
14
Vdd
3
14
2
15
K
11
CLK
12
Q
10
R
Vss
7
9
8
47k
4
3
6
S
5
D
8
O1
O0
A1
A0
O2
O3
A2
A3
Vdd
16
IC17 O5
4028B O4 1
1N5819
D4
1N4148
100nF
100nF
DIP7
MODE9BC
1N4148
47k
S3 OFF
S2 ON
100nF
S
D
100F
16V
+16V
G
4
2
+16V
OUT
OUT
GND
TP1
IN
REG2
AP1117E33
100nF
68k
4
IC15
47F
25V
LK1
100F
16V
+5V
A
K
1N4004
K
A
2.2k
K
LED2
A
2.2k
K
4
2
Db
Db
Da
Da
1
GND
OUT
OUT
10F
LK2
A
K
A
IN
GND
8
LEDS
IC12d
OUT
7805
+2.5V
ENABLE
CLKENA
DIP4
Vcc
Vdd
Vccs
+16V
GND
TAB (OUT)
AP1117E33,
SPX1117M3-L-2.5
9
100k
S
D
GND
OUT IN
1
2
3
G
Q16a
Gb
Sb
SaGa
Q16,Q17: FDS6912A
IN
D3 1N4148
100nF
3
REG3
SPX1117M3-L-2.5
Fig.3: the adaptor power supply. Power comes from a 9-12V DC plugpack or a USB cable. From these, 16V, 5V, 3.3V and 2.5V rails are generated. 16V powers the
logic while the rest can supply the micro. IC14 controls power to the micro with IC15 and IC16 monitoring the current flow. If the current limit is exceeded, IC14
turns the power off and turns on red LED3.
2012
SC
4
1
100F
16V
CON3
USB POWER
CON4
A
D5 1N4004
Q17a
DC POWER
Q17b
DIP7. IC17, another 4028B BCD-todecimal decoder, is between the two.
We don’t want Q17b to turn on if
Q17a is on as this would short the 3.3V
and 5V supplies together. Q17b is also
disabled if the programmer has been
set up for a micro which will be damaged by 5V. So for Q17b to come on,
DIP6 must be off, DIP7 on and neither
Mode 7 (for dsPIC33s) nor Modes 9B
or 9C (for PIC24s) should be enabled.
Since Q17b’s gate is connected to
output O1 (pin 14) of IC17, it will only
turn on if input A0 is high and inputs
A1-A3 are low, giving a binary input
value of 1. This can only occur under
the conditions specified above.
Electronic fuse
Whichever supply voltage is selected, current then flows from Q17a
or Q17b through Rshunt (0.1Ω) and
then through Q16a, to the micro’s
VCCS (switched VCC) supply. Q16a is
the soft-power switch and this allows
power to the micro to be cut quickly
in an over-current condition.
This condition is detected by the
voltage across Rshunt rising to a
certain level. The voltage across it is
amplified by precision op amp IC15
and monitored by comparator IC16a.
With 100mA through Rshunt, there is
just 10mV across it. If IC16a monitored
this directly, its maximum offset voltage of ±9mV would mean an error of
up to ±90mA. That’s clearly too much,
given that we want a nominal current
limit of around 100mA.
By comparison, IC15 has a very
low maximum input offset voltage
(0.15mV). It is configured for a gain
of 69, ie, (68kΩ + 1kΩ) ÷ 1kΩ. This
reduces the error due to IC16b’s offset
voltage to around ±1.5mA. A 100nF
feedback capacitor provides a time
delay (of about 1ms) so that very brief
current transients do not trip the current limit.
This is necessary since when power
is first applied, the charging of the supply bypass capacitors causes a brief
current spike which could otherwise
cause a nuisance trip.
IC15’s output is relative to VCC and
is negative, ie, the more current that
flows through Rshunt, the lower IC15’s
output voltage is. The reference voltage it is compared against must also
be relative to VCC and this is generated with small signal diode D2 and
a 47kΩ load resistor. The drop across
this 1N4148 diode is quite predicable
at around 0.6V.
In combination with IC15’s gain, this
sets the current limit to about 90mA
(0.6V ÷ 69 ÷ 0.1Ω). If the micro draws
any more than this during programming, IC16a’s output goes high and
the supply switches off. This was
sufficient for programming all micros
that we tested.
There is an additional consideration; when the micro supply is off, input pins 2 and 3 of op amp IC15 are
outside its normal operating range (115V). Its output is therefore undefined
and it could switch power off before
VCC rises to a normal level.
Comparator IC16b prevents this. It
compares VCC against the 2.5V rail and
so its output remains low until VCC
rises above the 2.2V reference derived
from LED1. Since the outputs of IC16a
and IC16b are connected together, this
prevents the over-current signal from
being asserted until the supply voltage
is high enough for IC15 to monitor the
current through Rshunt.
Power control
IC14a is a flipflop which drives
the gate of Q16a and hence controls
power to the micro. Its pin 4 reset
input is driven by comparator IC16a,
mentioned earlier. If excessive current
flow is detected and IC16a’s output
goes high, the 47kΩ resistor pulls pin
4 of IC14a high and this resets IC14a,
cutting power to the micro.
IC14a’s “set” input (pin 6) is tied
to ground and its data input (pin 5) is
pulled high. It is therefore switched
on by a positive transition on clock
input pin 3. The clock pin is driven
by pushbutton S2 with an associated
47kΩ pull-down resistor, hence pressing S2 turns the micro power on.
Similarly, pushing S3 turns the
power off since this pulls the reset
input (pin 4) high via a 2.2kΩ resistor and diode D4. IC14a is also reset
initially by the 100nF capacitor from
D4’s anode to +16V, so micro power is
off when the unit is first switched on.
IC14b, the other half of the dual
flipflop IC, is used to indicate if an
over-current trip occurs. When the
output of comparator IC16a goes high,
it not only resets IC14a but also sets
IC14b via pin 8. This turns on red
LED3 to indicate a fault. This LED can
then be turned off using pushbutton
S3 (power off) since this pulls its pin
10 reset input high.
When the output of IC14a is high
and Mosfet Q16a is on, supplying
power to the micro, yellow LED2 is
also lit. IC14a also drives the input of
inverter stage IC12d, which enables
clock signal multiplexers IC4 and IC5.
DIP4 must also be on for the clock enable to be asserted as otherwise, pin 6
of IC12d remains low.
Pin header CON5 can be used to
monitor VCC externally and if necessary, provide an off-board micro power
supply. Three-pin header LK2 selects
whether the ICSP receives power at
the same time as the micro or when
the programming adaptor board is
switched on. It is usually left in the
position shown, with pins 2 and 3
shorted, selecting the former condition.
LDO regulator REG3 derives 2.5V
from VCCS (3.3V or 5V) when Q16a is
on. This is required when programming PIC18LF2xJ5x micros.
More to come
Next month, we will provide the
PCB overlay diagrams and the construction details. We’ll also detail the
set up and describe how to use the
SC
Programming Adaptor Board.
Issues Getting Dog-Eared?
Keep your copies of SILICON CHIP safe, secure and
always available with these handy binders
REAL
VALUE
AT
$14.95
PLUS P
&
P
Available Aust, only. Price: $A14.95 plus $10 p&p per order (includes GST). Just fill
in and mail the handy order form in this issue; or fax (02) 9939 2648; or call (02)
9939 3295 and quote your credit card number.
siliconchip.com.au
May 2012 39
Measure and control temperature
over a wide range with this . . .
By JOHN CLARKE
High-Temperature
Thermometer/Thermostat
Need to measure or control temperature over a very wide range?
Now you can do it with this compact unit which hooks up to
a K-type thermocouple. It drives a relay which can be used to
precisely control the temperature in ovens, kilns, autoclaves, solder
baths or at the cold end of the spectrum, fridges and freezers. It is
based on an Analog Devices AD8495 precision instrumentation
amplifier with thermocouple cold junction compensation.
N
OW WE KNOW that some digital
multimeters can measure temperatures with a K-type thermocouple
but that’s all they can do; they cannot
control the temperature in an oven etc.
In other words, they do not provide
an adjustable thermostat function.
In all the above examples, our new
High-Temperature Thermometer/
Thermostat can be used to measure
40 Silicon Chip
and control the temperature at the
same time. That’s because it has a relay
output that opens or closes at a preset
temperature.
The switched output can be used directly or in conjunction with a higherrated relay to control power to the element of a heater or the compressor of
a refrigerator. For heating, the power
can be switched on when the tempera-
ture is below the preset temperature
and switched off when it is above.
Alternatively, for cooling, power can
be switched on when the temperature
goes above the preset and off when it
goes below. The preset temperature for
this thermostat action can be adjusted
between -50°C and 1200°C.
It is important that the thermostat
function does not cause rapid on and
siliconchip.com.au
Features & Specifications
Main Features
•
•
•
•
•
•
•
•
•
K-Type thermocouple probe
Ground referenced or insulated
probe can be used
Measures -50°C to 1200°C (depending on probe)
Pre-calibrated temperature measurement
Optional calibration of span and offset adjustment
Thermostat switching at a preset temperature with adjustable hysteresis
High to low or low to high thermostat threshold
Relay output for thermostatic control
Relay contacts rated at 10A (30V AC/DC maximum recommended
switching voltage)
Specifications
Power supply: 12V <at> 100mA
Measurement range: -50°C to 1200°C (probe dependent)
Initial accuracy: ±4°C for -25°C to 400°C measurements (ambient between
0°C and 50°C)
Optional calibration adjustment for span: -4%, +5.27%
Optional calibration adjustment for offset: ±6.2mV equivalent to >±1°C
off switching of the heater, compressor
or whatever is being temperature-controlled. Hence the design incorporates
adjustable hysteresis. This allows a
preset temperature difference to apply
between switching power on and off.
The hysteresis is adjustable from less
than 1°C to more than 9°C.
The temperature is displayed on a
3½ digit LCD and while the unit can
display a temperature from -50°C to
1200°C the actual measurement range
will depend on the particular probe.
Some K-type probes will operate from
-50°C to 250°C, while others will operate from -40°C to 1200°C.
The High-Temperature Thermometer/Thermostat is housed in a small
instrument case and controls on the
front include a power switch and a
switch to select between measured
temperature and the preset thermostat
temperature. A LED indicator is for
power indication and a second LED
shows when the thermostat relay has
switched on.
At the rear of the case is the power
input socket for a 12V DC supply and
a socket for the K-type thermocouple.
Additionally, there is a terminal connector inside the case for connection
to the thermostat relay contacts. The
common (C), normally open (NO) and
normally closed (NC) contacts are
available for connection.
Inside the case there are jumper
siliconchip.com.au
Thermostat set point range: adjustable from -50°C to 1200°C
Thermostat hysteresis: adjustable from <1°C to >9°C
Cold junction compensation: optimised for 0-50°C ambient temperatures
links to select whether the thermostat
relay switches on above or below the
preset temperature for the thermostat.
There are also jumper selections to
select whether the Thermometer/
Thermostat is built pre-calibrated or
where the temperature calibration can
be accurately adjusted.
K-type thermocouple
As mentioned above, this design
uses a K-type thermocouple which
comprises a junction of two dissimilar
wires; in this case it uses an alloy of
chrome and nickel (called Chromel)
for one wire and an alloy of aluminium, manganese, silicon and nickel
(called Alumel) for the second.
These two wires are insulated and
make contact at the temperature probe
end only. The other end of the wires
are usually connected to a 2-pin plug.
Basically, a thermocouple’s operation relies on the principle that the
junction of two dissimilar metals
produces a voltage that is dependent
on temperature. A K-type thermocouple produces a voltage output that
typically changes by 40.44µV/°C.
This change in output per is called
the Seebeck coefficient and it refers
to the output change that occurs due
to the temperature difference between
the probe end and the plug end of the
thermocouple.
In practice, the Seebeck coefficient
for the K-type thermocouple varies
with temperature and is not precisely
40.44µV but this is a good average
value over the temperature range from
0°C to 1200°C.
If we know the temperature at the
plug end of the thermocouple, we can
calculate the temperature at the probe
since we also know the Seebeck coefficient. For example, if the plug end is
held at 0°C, the output will increase
by 40.44µV for every 1°C increase.
Similarly, the output will decrease
by 40.44µV for every 1°C drop in
temperature.
In practice, we do not keep the plug
end of the thermocouple at 0°C; it’s not
practical. Instead, we compensate the
thermocouple output by measuring the
temperature at the plug end and then
adding 40.44µV for every 1°C that the
thermocouple plug end is above 0°C
or subtracting 40.44µV for every 1°C
that the plug end is below 0°C.
May 2012 41
THERMOSTAT
PRESET
REF1: 2.5V
REFERENCE
VR1
IC2d
COMPARATOR
(IC2a)
A=3
+2.5V
–
IC1
AD8495
OUT
5mV/°C
2
S2
REF
+2.5V
1.25V
1
NO
COM
NC
K-TYPE
THERMOCOUPLE
+
RELAY
RELAY
DRIVER
(Q1, Q2)
1 = THERMOMETER
2 = THERMOSTAT
1/50
DIVIDER
(VR4, LK3-4)
A=1
~ 1.25V
100 V/°C
INHI
INLO
3.5-DIGIT LCD
PANEL METER
(200mV FULL SCALE)
BUFFER
(IC2b, IC2c,VR3
Fig.1: block diagram of the High-Temperature Thermometer/Thermostat. IC1
processes and amplifies the thermocouple’s output and drives the LCD panel
meter and comparator IC2a. Trimpot VR1 sets the thermostat temperature.
For example, if the thermocouple
plug is at 25°C, its output will be
1.011mV (ie, 25 x 40.44µV) lower than
it would be if it were at 0°C. By adding
an extra 1.011mV to the reading, we
obtain the correct result without having to keep the plug end at 0°C.
Note that there are several dissimilar
metal junctions within the connections between the thermocouple plug
and amplifier. These include the
Chromel to copper junction and the
Alumel to copper junction on the PCB
itself. These do not contribute to the
overall voltage reading provided they
are all kept at the same temperature.
As a result, the PCB has been
designed to help maintain similar
temperatures at these junctions by
making the copper connections all the
same size. Once the PCB is installed
inside its case, the inside temperature
should remain fairly constant for all
these junctions.
Note that if the thermocouple lead
needs to be extended, it’s necessary
to use the same K-type thermocouple
wire for the whole length between the
probe and plug.
Signal processing
Refer now to Fig.1 which shows the
block diagram of the High-Temperature
Thermometer/Thermostat. As shown,
the thermocouple signal is processed
using the Analog Devices AD8495 IC.
This is a precision instrumentation
amplifier with K-type thermocouple
42 Silicon Chip
cold junction compensation. Its output
is 5mV/°C.
The amplifier within the AD8495
is laser trimmed for a gain of 122.4.
This gain effectively converts the
40.44µV/°C output of the thermocouple to 4.95mV/°C. The output is
optimised for a 25°C measurement
where a gain of 122.4 gives a result of
123.75mV.
Within the AD8495, a 1.25mV offset is added to the amplified value,
giving a 125mV output at 25°C. For
temperatures other than 25°C, the
combination of the variation in the
Seebeck coefficient over temperature,
the 122.4 gain and the 1.25mV offset
provides an accurate 5mV/°C output
over the range of -25°C to 400°C. For
this range, the output is within 2°C.
Note that the specification panel
shows that the accuracy is ±4°C for
ambient between 0°C and 50°C and
-25°C to 400°C measurements. This
is different to the 2°C error for the
AD8495 because the display is showing a reading via a voltage divider that
is prone to extra tolerance errors.
It’s possible to calibrate the measurement to a finer accuracy if this is
required. Table 1 shows the expected
output from the AD8495 over a wide
range of temperatures and compares
this with the ideal 5mV/°C output.
How it works
Returning now to the block diagram
of Fig.1, the K-type thermocouple con-
nects directly to the AD8495 (IC1) at
the IN+ and IN- terminals. The resulting 5mV/°C output signal from IC1 is
then fed to the non-inverting input of
comparator IC2a and also to position
1 (Temperature) on switch S2. S2
selects between the Temperature and
Thermostat modes of operation.
In order to allow for negative temperature measurements, the output
from the AD8495 is offset by approximately 1.25V. This offset is derived
by a voltage divider connected across
a 2.5V reference (REF1) and buffered
using op amps IC2b and IC2c. The
buffered 1.25V signal is then applied
to the AD8495’s REF (reference) input.
This effectively “jacks up” the
AD8495’s output by 1.25V. As a result, a -50°C measurement now gives
an output that’s theoretically 250mV
below (-5mV x 50) the 1.25V reference
offset (ie, 1V). Without this offset, the
AD8495 would not be able to handle
negative temperature measurements
since its output cannot go below 0V.
Although the offset only needs to
be 250mV to allow for a -50°C measurement, a value of 1.25V is used
because of the LCD panel meter that’s
used to measure the voltage. This
meter requires an input that’s at least
1V above the 0V supply for correct
operation. According to Table 1, the
actual output from IC1 at -50°C is
228mV below the offset voltage. So
using an offset of 1.25V leaves us with
a comfortable 22mV margin above the
critical 1V level.
The 3.5-digit LCD panel meter
used to display the temperature has
a 200mV full scale reading (actually
199.9mV) for a reading of 1999. It’s
basically connected to measure the
voltage between IC1’s output (via a
divider) and the offset voltage. This
effectively removes the offset voltage
from the reading.
To prevent the meter from overranging and to get a reading in °C,
we need to divide IC1’s output by
50. For example, if the temperature
is 1200°C, the voltage between IC1’s
output and the 1.25V offset will be
6V (ie, 1200 x 5mV). Dividing this
by 50 gives 120.0mV and the panel
meter is configured to show 1200 (ie,
no decimal point).
Note that, in the full circuit, either
a fixed divide-by-50 attenuator or an
adjustable divide-by-50 attenuator
can be used. The desired attenuator
is selected using jumper links and
siliconchip.com.au
the adjustable one allows for accurate
calibration.
The display can either show the
measured temperature when switch
S2 is in position 1 or the preset temperature (for the thermostat operation)
when S2 is in position 2. VR1 sets the
thermostat temperature. As shown, it’s
connected to a 2.5V reference (REF1)
and the voltage at its wiper drives op
amp IC2d.
As a result, IC2d’s output can
range up to 7.5V, slightly more than
the 7.25V at IC1’s output when the
measured temperature is at the 1200°C
maximum (ie, 1200 x 5mV plus the
1.25V offset). This allows VR1 to set
the thermostat temperature anywhere
from -50°C to 1200°C.
IC2d’s output is fed to the inverting
input of comparator IC2a where it is
compared with IC1’s output. IC2a’s
output thus switches low when the
temperature is below the preset and
high when the temperature is above
the preset. This output then drives
a relay via transistors Q1 and/or Q2.
Links LK5 and LK6 can be selected
so that the relay either switches on
when IC2b’s output goes high or on
when it goes low.
Circuit details
Refer now to Fig.2 for the full circuit diagram of the High-Temperature
Thermometer/Thermostat. As well as
the AD8495 (IC1) and the LCD panel
meter, it includes an OP747 precision
quad op amp (IC2), a 7805 3-terminal
regulator, an LM285-2.5 precision voltage reference, transistors Q1 & Q2 and
various minor components.
IC1 is powered from a 12V DC plugpack supply via switch S1, diode D1
(for reverse polarity protection) and a
10Ω resistor. A 22V zener diode (ZD1)
clamps any over-voltage transients
while 100µF and 100nF capacitors are
used to bypass the supply.
In operation, IC1 draws just 180µA
to minimise internal heating (note:
internal heating would affect the
measurement of the ambient temperature used for the thermocouple
ice-point temperature compensation).
The K-type thermocouple connects
to its IN+ and IN- terminals (pins 8
& 1) via series 47kΩ resistors. These
resistors and their associated 100nF
ceramic capacitors prevent RF (radio
frequency) signals from being detected
by IC1’s sensitive input stages. The
resistors acts as RF stoppers, while the
siliconchip.com.au
Table 1: AD8495 Output vs. Temperature
Thermocouple
Temperature (°C)
Ideal Output <at> 5mV/°C
(mV)
AD8495 Output
(mV)
Display Reading (°C)
±1 Digit
-50
-40
-20
0
20
25
40
60
80
100
120
140
160
180
200
220
240
260
280
300
320
340
360
380
400
420
440
460
480
500
520
540
560
580
600
620
640
660
680
700
720
740
760
780
800
820
840
860
880
900
920
940
960
980
1000
1020
1040
1060
1080
1100
1120
1140
1160
1180
1200
-0.25
-0.2
-0.1
0.0
0.1
0.125
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
4.0
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
5.0
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9
6.0
-0.228
-0.184
-0.093
0.003
0.100
0.125
0.200
0.301
0.402
0.504
0.605
0.705
0.803
0.901
0.999
1.097
1.196
1.295
1.396
1.497
1.599
1.701
1.803
1.906
2.010
2.113
2.217
2.321
2.425
2.529
2.634
2.738
2.843
2.947
3.051
3.155
3.259
3.362
3.465
3.568
3.670
3.772
3.874
3.975
4.076
4.176
4.275
4.374
4.473
4.571
4.669
4.766
4.863
4.959
5.055
5.150
5.245
5.339
5.432
5.525
5.617
5.709
5.800
5.891
5.980
-46
-37
-19
0
20
25
40
60
80
101
121
141
161
180
199
219
239
259
279
299
320
340
361
381
402
423
443
464
485
506
527
548
569
589
610
631
651
672
693
713
734
754
774
795
815
835
855
875
895
914
934
953
973
992
1011
1030
1049
1068
1086
1105
1123
1141
1160
1178
1196
May 2012 43
44 Silicon Chip
siliconchip.com.au
+
–
CON2
10k
10k
10k
1
8
K
10
9
6
5
SENSE
5
100nF*
100 F
IC2c
IC2b
–Vs
3
2
8
7
* CERAMIC
REF
6
IC1
OUT
AD8495
100nF
–IN
+IN
7
+Vs
100 F
10
UNCAL
10
LK1 LK2
CAL
OUT
GND
V+
IN
REG1 7805
+2.5V
K
A
15k
VR1
100k
100 F
+5V
10k
VR4
100
LK3
CAL
UNCAL
LK4
100nF
THERMOSTAT
PRESET
LED1
POWER
470
HIGH TEMPERATURE THERMOMETER/THERMOSTAT
A
+2.5V
A
K
D1
100nF*
VR3
100
REF
OFFSET
10k
100
47k
A
100nF*
ZD1
22V
100nF*
47k
S1
POWER
100nF
20k
IC2d
4
1.8
39
1
THERMOMETER
13
12
100nF
B
10k
A
K
D1, D2: 1N4004
1k
51k
2
THERMOSTAT
470
IC2: OP747
S2
14
10 F
NP
9
7
5
COM
6
INLO
8
RFL
1
K
1
+5V
B
A
IN
E
CON3
A
ZD1
K
2
NO
NC
COM
OUT
K
A
NC
C
BC337
7805
B
A K
LEDS
GND
LM285-2.5Z/LP
Q2
BC337
GND
E
C
RELAY1
.
.8:8.8
10k
2.2k
D2
K
3.5-DIGIT LCD PANEL METER
VR2 1M
+
ROH
RFH
INHI
100k
LK6
L/H
11
IC2a
10k
10
3
2
E
Q1
BC337
C
H/L
LK5
LED2
A
Fig.2: the complete circuit diagram for the High-Temperature Thermometer/Thermostat. An accurate 1.25V reference is derived from REF1 via IC2b or IC2c and
this is applied to the REF input of IC1 to enable measurements down to -50°C. IC1’s output drives the LCD panel meter via a 50:1 divider and also drives the noninverting input of comparator IC2a. IC2a compares IC1’s output with the thermostat preset temperature, as set by VR1 & IC2d, and drives relay 1 via transistors
Q1 and/or Q2 when the preset limit is reached. Links LK5 & LK6 allow the relay to be driven on either a rising or falling temperature.
SC
2012
+5V
K
4.7k
REF1
LM285
-2.5Z/LP
–
+
CON1
K-TYPE
THERMOCOUPLE
12V
DC
INPUT
+12V
+11.4V
100nF capacitors effectively shunt any
remaining RF signal to ground.
In addition, the negative terminal
of the K-type thermocouple is tied to
ground via a 100Ω resistor. This prevents the probe from picking up noise
and mains hum, which would cause
erratic operation.
Note that the 100Ω resistor is included so that the circuit can be used
with both earthed and insulatedsheath thermocouples. Basically, the
thermocouple probe wires are housed
in a cylindrical metal sheath or rod.
Some units connect the negative thermocouple wire directly to this metal
sheath (an earthed probe), while others
fully insulate the metal sheath from
the thermocouple wires (an insulated
probe).
For an insulated probe, it doesn’t
matter whether the negative terminal
is connected directly to ground or
connected to ground via a 100Ω resistor. That’s because an insulated probe
can connect to a point that’s not at 0V
without affecting the operation of the
probe.
By contrast, an earthed probe does
require the 100Ω resistor. That’s
because the probe could make an
external connection to the 0V supply
rail and this might not be at exactly
the same voltage as the 0V rail inside
the unit.
This type of situation could easily
arise, for example, when measuring
engine heat or brake disc heat in a car
and the unit is being powered by the
vehicle’s battery. In this situation, the
probe point and the internal 0V rail
will be at slightly different voltages
due to current flowing in the vehicle’s
chassis. The difference in voltage may
only be small but the thermocouple’s
output only varies by about 40µV/°C,
so only small variations can mean a
huge error in temperature readings.
The 100Ω resistor eliminates this
problem by preventing significant
current flow between the thermocouple’s negative terminal and the 0V rail
within the thermometer.
Deriving the offset
The 1.25V offset for IC1 is derived
from REF1, a precision 2.5V voltage
reference, via a resistive divider. This
divider comprises four 10kΩ resistors
and a 100Ω trimpot (VR3).
As shown, the 1.25V midpoint of
the 10kΩ fixed resistive divider is fed
to pin 5 of IC2b, while the voltage on
siliconchip.com.au
VR3’s wiper is fed to IC2c. VR3 allows
the offset voltage to be varied over a
small range either side of 1.25V.
IC2b and IC2c are both connected as
unity gain buffer stages. When LK1 is
installed, IC2b provides a fixed 1.25V
offset for IC1 at its REF (pin 2) input.
At the same time, IC2c provides the
variable offset output to the panel
meter at its IN LO input.
Alternatively, if LK2 is installed,
IC2c drives both the reference input
of IC1 and the INLO input of the LCD
panel meter. In this case, the voltage
applied to both IC1’s REF input and the
panel meter’s INLO input are exactly
the same and this is the linking option
to use if you do not want to accurately
adjust the temperature calibration.
LCD panel meter
As stated previously, the LCD panel
meter measures the difference between its INHI (pin 7) and INLO (pin
6) inputs. In this circuit, IC1 drives
the INHI input via one of two 50:1
voltage dividers (one fixed, the other
variable) when S2 is in position 1.
IC1 is capable of delivering in excess
of ±5mA to a load but the fixed 50:1
divider draws just 115µA maximum
when IC1’s output is producing 7.25V
for a 1200°C measurement. This low
current minimises any internal heating of the IC.
The fixed divider is selected using
link LK4. It’s made up using a 51kΩ
resistor in the top section and 39Ω,
1.8Ω and 1kΩ resistors at the bottom.
Assuming the values are exact, the
division ratio is very close to 50:1.
However, resistor tolerances can shift
this to within a range of around 50.05:1
to 49.95:1.
The variable divider shares the 51kΩ
and 1kΩ resistors but uses a 100Ω trimpot in place of the 39Ω and 1.8Ω resistors in the fixed divider. This allows
the divider to be adjusted. It’s selected
by installing link LK3 instead of LK4.
The LCD panel meter itself is based
on an Intersil ICL7106 3.5-digit LCD
analog-to-digital converter (ADC).
Its INLO, COM (common) and RFL
(reference low) pins are all connected
together, ie, they are all fed with the
reference offset voltage at IC2c’s output. In addition, the ROH output is
connected to the RFH (reference high)
input and this sets the panel meter to
200mV full scale.
A 5V supply rail for the LCD is
derived from regulator REG1 (7805).
The OP747ARZ Quad
Precision Op Amp
The OP747ARZ quad precision op
amp specified here has features that are
not found in general-purpose op amps.
First, it features a low offset voltage
of 100µV maximum and the input bias
and offset currents are in the very low
nA range. Second, it can handle input
voltages ranging from the ground supply rail up to within 1V of the positive
supply. And third, the output can reach
close to each supply rail.
Taken together, these characteristics
make the op amps ideal for this circuit.
REG1’s input and output rails are
both filtered using 100µF electrolytic
capacitors, while LED1 in series with
a 470Ω current-limiting resistor provides power indication.
This regulated 5V supply also drives
the 2.5V reference (REF1), this time via
a 4.7kΩ resistor. As well as providing a
source for the offset voltage, the resulting 2.5V rail is also fed to the top of
VR1 which sets the thermostat preset.
VR1 is connected in series with a
15kΩ resistor across this supply and
its wiper provides an output which
ranges from 326mV up to 2.5V. IC2d
amplifies this by three, as set by the
20kΩ and 10kΩ resistors in the feedback path. The resulting voltage at the
output of IC2d can range anywhere
from 978mV up to 7.5V and that more
than covers the possible voltage range
from IC1, for temperatures ranging
from -50°C to 1200°C.
As described previously, op amp
IC2a is wired as a comparator. It monitors IC2d’s output and compares this
with IC1’s output. IC2a thus switches
its output high when the measured
temperature is above the preset temperature or low when the measured
temperature goes below the preset
(ignoring hysteresis).
Trimpot VR2 (1MΩ) and the 100kΩ
and 470Ω resistors provide hysteresis.
With VR2 set at 1MΩ, the hysteresis is
at its minimum and there is less than
1°C hysteresis. At the other extreme,
with VR2 set for 0Ω, the hysteresis is
more than 9°C.
Relay driver circuit
IC2a drives transistor Q1, which
in turn drives Q2, when link LK5 is
inserted. Alternatively, if LK6 is selected, Q1 is bypassed and IC2a drives
Q2 direct.
May 2012 45
1M
100
47k
100
100nF
100nF
22V
ZD1
100
10
LK1 LK2
LK3 LK4
1.8
51k
39
1k
15k
0V
+5V
100nF
A
S2
LED2
LED1
VR4
10k
REF1
2.2k
A
VR3
TEMPERATURE
THERMOMETER
/THERMOSTAT
100nF
LM285
-2.5Z/LP
470
HI
47k
BC337
10k
100k
10k
12150112
EPYT K
RHIGH
ETE M O MRE HT
100nF
100nF
20k
10k
IC2 (UNDER)
VR2
100nF
10
470
VR1
4.7k
100 F
4004
100 F
S1
100 F
100k
100nF
IC1 (UNDER)
ROH
RFH
RFL
InHi
InLo
COM
COIL
10k
REG1
7805
D1
Q1
RELAY1
4004
D2
LK5 LK6
10k
Q2
10k
10k
CON3
–
Thermocouple
K type
+ CON1 –
LOW
© 2012
BC337
10 F
NP
+
TO THERMOCOUPLE
SOCKET
NO
COM
NC
CON2
12VDC IN
RELAY
CONTACTS
13 12 11 10 9 8 7 6 5 4
2 1
3.5-DIGIT LCD PANEL METER (REAR)
Fig.3: follow this diagram to build
the unit but note that the first job is
to install surface-mount devices IC1 & IC2 on the underside of the PCB (see
below). You can omit the relay, CON3, S2 and transistors Q1 & Q2 if you intend
using the unit as a thermometer only and don’t need the thermostat function.
(UNDER SIDE OF PCB)
1
IC1
04105121
K TYPE
THERMOMETER
1
IC2
46 Silicon Chip
Fig.4 (left): this diagram and
the above photo show how
surface-mount devices IC1
& IC2 are mounted on the
underside of the PCB. Make
sure that both devices are
correctly orientated (pin 1 is
identified by a small dot on
the device body) and follow
the step-by-step procedure
described in the text to solder
them into position.
These two links select whether the
relay turns on for a low-to-high temperature transition (LK6 in place) or a
high-to-low transition (LK5 in place).
When LK6 is in circuit, Q2 turns on
when IC2a’s output goes high (ie, when
the temperature rises above the preset)
and this turns on relay 1. The relay
subsequently turns off again when
IC2a’s output switches low (ie, when
the temperature falls below the preset).
Conversely, when LK5 is in circuit,
Q1 inverts the logic. In this case, Q2
and the relay are normally on since
Q2’s base is pulled high. However,
when IC2a’s output switches high (as
the temperature rises above the preset),
Q1 turns on and pulls Q2’s base to
ground. As a result, Q2 and the relay
turn off and remain off until the temperature falls below the preset again.
LED2 lights whenever the relay
switches on to indicate that the set
temperature threshold has been reached. The associated 2.2kΩ resistor limits the current through LED2, while
diode D2 protects Q2 from damage
by quenching the back-EMF voltage spikes generated when the relay
turns off.
The relay provides both the usual
common (COM), normally open (NO)
and normally closed (NC) contacts, so
it can also drive a load on or off depending on the selection of the NO or
NC contacts. So it may seem that links
LK5 and LK6 are not really necessary
to reverse the switching sense.
However, there are reasons why you
may wish to select whether the relay
is normally powered or not, especially
when the relay contacts are required to
switch a heating or cooling operation.
One reason is that less current is
drawn by the circuit when the relay
is off and you might want to choose
the link and contact configuration that
draws the least power.
Another reason is that you might
want to ensure fail-safe operation if
power is cut to the circuit. By using
the COM & NO contacts to do the
switching, you can ensure that power
is not provided for heating or cooling
if the power to the Thermometer/
Thermostat fails.
Construction
The assembly is straightforward
with all parts except the probe socket
and the LCD panel meter mounted on
a PCB coded 21105121 (117 x 102mm).
This is housed in a plastic instrument
siliconchip.com.au
The thermocouple socket is connected to an
adjacent screw terminal block via two short leads.
Alternatively, the screw terminal block could be omitted
and a couple of flying leads soldered direct to the PCB.
case measuring 140 x 110 x 35mm.
Begin by carefully checking the PCB
for any defects. Check also that the
hole sizes are correct for each component to fit neatly. The corner mounting
holes and the regulator mounting hole
should all be 3mm in diameter.
Our prototype used a double-sided
PCB and Fig.3 shows the parts layout.
The first step is to install IC1 and IC2.
These are both surface-mount devices
(SMDs) and mount on the underside
of the PCB – see Fig.4.
To install these, you will need a
fine-tipped soldering iron, some fine
solder and some quality solder wick. A
magnifying lamp or at least a magnifying lens will also be handy.
It’s best to install IC2 first. This is
the 14-pin device with the wider pin
spacings. First, place the PCB copper
side up and apply a small amount of
solder to the top-right pad, then pick
the IC up with tweezers and position
it near the pads. Check that it is orientated correctly (ie, with its pin 1 dot
positioned as shown on Fig.4), then
heat the tinned pad, slide the IC into
place and remove the heat.
Now check the IC’s alignment
carefully using a magnifying glass. It
siliconchip.com.au
should be straight, with all the pins
centred on their respective pads and a
equal amount of exposed pad on either
side. If not, reheat the soldered pin and
nudge the chip in the right direction.
Once its position is correct, solder
the diagonally opposite pin, then
recheck its position before soldering
the remaining pins. Don’t worry too
much about solder bridges between
pins at this stage; they are virtually
inevitable and can easily be fixed.
The most important job right now is
to ensure that solder flows onto all the
pins and pads.
Once you’ve finished, apply a thin
smear of no-clean flux paste along
all the solder joints and remove the
excess solder using solder wick. You
should then make a final inspection
to ensure that there are no remaining
solder bridges and that the solder has
not “balled out” onto a pin without
flowing onto the pad.
If there are still bridges, clean them
up with more flux and solder wick.
Once IC2 is in place, you can install
IC1 in exactly the same manner.
Through-hole parts
The larger through-hole parts can
now be installed on the top of the
PCB. Start with the resistors and diodes, then install zener diode ZD1,
the MKT and ceramic capacitors and
the electrolytics. It’s a good idea to
check the value of each resistor using
a multimeter before installing it.
Take care with the polarity of the
electrolytics, the diodes and the zener
diode. They must be orientated as
shown on Fig.3.
Transistors Q1 & Q2 and the LM3852.5 precision voltage reference (REF1)
can go in next. REG1 can then be installed. This mounts horizontally with
its tab against the PCB, so you will
have to bend its leads down at right
angles to match its mounting holes.
Secure its tab to the PCB using an M3
x 6mm screw and nut before soldering
its leads. Don’t solder the leads before
securing the tab; you could crack the
copper tracks at the mounting screw
is tightened if you do.
Trimpots VR1-VR4 are next on the
list. These must all be mounted with
the adjustment screw to the right. Follow with the three 3-way pin headers
for links LK1-LK6, then install the
6-way and 2-way polarised headers
for the LCD panel meter connections.
May 2012 47
The cable gland on the rear panel allows an external lead to be fed into the case and connected to the relay contacts at
CON3. The LCD is secured to the front panel by running a couple of beads of silicone adhesive or hot-melt glue down
the vertical inside edges.
Be sure to orientate these headers as
shown, ie, with their vertical tabs
towards the panel meter.
Once they’re in, you can install the
two LEDs but first you have to bend
their leads down through 90° some
9mm from their bodies.
The best way to do this is to first cut
a cardboard spacer 9mm wide. This
is then be used as a template when
bending the LED leads. Make sure
that each LED is correctly orientated
before bending its leads – the (longer)
anode lead must be on the right when
looking at the lens.
Having bent their leads through 90°,
the two LEDs must be installed with
their leads 5mm above the PCB. This
is best done by pushing them down
onto a 5mm spacer, then soldering the
leads to the PCB pads.
Switches S1 & S2 are right-angle
types and are mounted directly on the
PCB. Push them down onto the board
as far as they will go before soldering
their leads. The PCB assembly can then
be completed by installing the relay,
48 Silicon Chip
the DC socket (CON2) and the 2-way
and 3-way screw terminals.
Connecting the panel meter
The panel meter is wired to the
6-way header plug and to the 2-way
header plug using short lengths of ribbon cable. These leads can be obtained
by separating an 8-way ribbon into
6-way and 2-way strips.
Cut these strips to 50mm in length,
then strip about 2mm of insulation
from the individual wires at one end
and crimp them to the header pins.
The pins can are then inserted into
the headers.
The other ends of these leads can
then be stripped and soldered to the
LCD panel meter pins. Check carefully
to ensure that each wire goes to the
correct pin on the panel meter and
that there are no shorts between them.
In fact, it’s a good idea to slip a short
length of heatshrink over each wire
before soldering it and then pushing
over the soldered joint to insulate it
from its neighbours.
Jumper links LK2 & LK4 should now
be installed and either LK5 or LK6.
Install LK5 if you want the relay to
switch on when the temperature drops
below the preset. Alternatively, install
LK6 if the relay is to switch on when
the temperature rises above the preset.
Final assembly
Fig.5 shows the front and rear panel
artworks. You can purchase finished
panels from SILICON CHIP or you can
download the artworks in PDF format
from our website.
Mounting the panel meter
The LCD panel meter is mounted by
sliding into its front-panel slot (which
is open at the top). Check that the top
of the meter sits flush with the top of
the panel. If it protrudes slightly, it will
be necessary to make the slot slightly
deeper until it does sit flush.
The meter is secured in place by
running a bead of silicone sealant or
hot-melt glue along the two vertical inside edges, adjacent to the front panel.
siliconchip.com.au
+
its mounting slot from the rear (terminal screws facing up) and fitted with
the supplied clip to hold it in place.
Once that’s done, the rear panel can
be slipped into the case and two short
wires run between the thermocouple
socket and the screw terminal block
on the PCB.
The lid can now be test fitted to
make sure everything is correct. Note
that it will be necessary to file the
RELAY CONTACTS
10A MAX & 30V MAX
K-Type Thermocouple
Relay
Temperature
Power
SILICON CHIP
THERMOMETER/
THERMOSTAT
Thermostat
RELAY
OUTPUT
12V DC
.
Once the meter is in place, the front
panel and the PCB assembly can be slid
into the case. The PCB is then secured
to the base using four self-tapping
screws that go into integral mounting
bushes. That done, the leads from the
panel meter can be plugged into the
headers on the PCB.
The rear panel carries a cable gland
(for the relay outputs) and the thermocouple socket. The latter is fed through
Fig.5: these front and rear panel
artworks can be copied and used as
drilling templates. Finished panels
are also available from SILICON CHIP.
two ridges at the front of the lid down
where they meet the panel meter.
Testing
To test the unit, first apply power
Table 1: Resistor Colour Codes
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
siliconchip.com.au
No.
1
1
2
1
1
8
1
1
1
2
1
1
2
1
Value
100kΩ
51kΩ
47kΩ
20kΩ
15kΩ
10kΩ
4.7kΩ
2.2kΩ
1kΩ
470Ω
100Ω
39Ω
10Ω
1.8Ω (5%)
4-Band Code (1%)
brown black yellow brown
green brown orange brown
yellow violet orange brown
red black orange brown
brown green orange brown
brown black orange brown
yellow violet red brown
red red red brown
brown black red brown
yellow violet brown brown
brown black brown brown
orange white black brown
brown black black brown
brown grey gold gold
5-Band Code (1%)
brown black black orange brown
green brown black red brown
yellow violet black red brown
red black black red brown
brown green black red brown
brown black black red brown
yellow violet black brown brown
red red black brown brown
brown black black brown brown
yellow violet black black brown
brown black black black brown
orange white black gold brown
brown black black gold brown
not applicable
May 2012 49
Parts List
1 PCB, code 21105121, 117 x
102mm
1 plastic instrument case, 140 x
110 x 35mm
1 12V DC 500mA plugpack
1 3.5-digit LCD panel meter (Jaycar
QP-5570 or similar)
1 front-panel label or 1 front-panel
PCB, code 21105122
1 rear-panel label or 1 rear-panel
PCB, code 21105123
1 K-type thermocouple probe (Jaycar QM-1292 -50°C to 250°C,
QM-1283 -40°C to 1200°C)
1 K-type thermocouple probe socket (Element14 Cat. 708-6386)
1 SPDT 10A 12V relay, Jaycar SY4050 or equivalent (RELAY1)
2 SPDT PCB-mount toggle switches (S1,S2) (Altronics S1421 or
equivalent)
1 PCB-mount 2.5mm DC socket
(CON1)
1 2-way PCB-mount screw terminal
block, 5.08mm spacing (CON2)
1 3-way PCB-mount screw terminal
block, 5.08mm spacing (CON3)
1 cable gland for 3-6.5mm diameter
cable
1 2-way polarised pin header,
2.54mm spacing
1 6-way polarised pin header,
2.54mm spacing
1 2-way header sockets to match
above header
1 6-way header sockets to match
above header
2 3mm LED bezels (optional)
3 3-way pin headers, 2.54mm spacing (LK1-LK6)
3 jumper shunts
4 No.4 x 6mm self-tapping screws
1 M3 x 6mm pan-head machine
screw
1 M3 nut
1 100mm length of 0.8mm tinned
copper wire
1 50mm length of 8-way ribbon
cable
Semiconductors
1 AD8495ARMZ precision thermocouple amplifier with
cold junction compensation (IC1)
(Element14 Cat. 186-4707)
1 OP747ARZ quad precision single
supply op amp (IC2) (Element14
Cat. 960-4405) (IC2)
1 LM285Z/LP-2.5 micropower
voltage reference diode (REF1)
(Element14 Cat. 966-5447; Jaycar ZV1626)
1 7805 5V 3-terminal regulator
(REG1)
2 BC337 NPN transistors (Q1,Q2)
1 22V 1W zener diode (ZD1)
2 1N4004 1A diodes (D1,D2)
1 green 3mm LED (LED1)
1 red 3mm LED (LED2)
Capacitors
3 100µF 16V PC electrolytic
1 10µF 50V non-polarised electrolytic
4 100nF ceramic
4 100nF MKT polyester
Trimpots
1 1MΩ top-adjust multi-turn trimpot
(code 105) (VR2)
1 100kΩ top-adjust multi-turn trimpot (code 104) (VR1)
2 100Ω top-adjust multi-turn trimpots (code 100) (VR3,VR4)
Resistors (0.25W, 1%)
1 100kΩ
1 2.2kΩ
1 51kΩ
1 1kΩ
2 47kΩ
2 470Ω
1 20kΩ
1 100Ω
1 15kΩ
1 39Ω
8 10kΩ
2 10Ω
1 4.7kΩ
1 1.8Ω 5%
Note: PCBs for this project are
available from SILICON CHIP.
The unit can be used with any K-type
thermocouple, eg, the Jaycar QM1292
or QM1283.
and off when the preset goes just over
or under the measured temperature.
VR2 can now be adjusted to give
the required amount of hysteresis
(clockwise for more hysteresis and
anticlockwise for less).
Calibration
If you wish, the unit can be left uncalibrated in which case its accuracy
will be as shown in the specifications
panel.
Alternatively, if you wish to calibrate the unit for improved accuracy,
the procedure is as follows:
(1) Remove jumper links LK2 & LK4
and install links LK1 & LK3 instead.
(2) Place the thermocouple probe in a
cup of distilled water brimming with
ice (note: the ice also needs to be made
from distilled water to ensure accuracy
and the ice-water mixture has to be
constantly stirred to maintain a 0°C
temperature).
(3) Adjust VR3 so that the thermometer reads 0°C.
(4) Place the thermocouple probe in
boiling distilled water and adjust VR4
for a reading of 100°C at sea level or
deduct 1°C for every 300m above sea
level.
That completes the calibration. The
lid can now be attached to the case and
the unit is ready for use.
Ambient temperature display
and check that the power LED lights.
The display should also show a temperature reading with S2 (Thermostat/
Thermometer) in position 1 (Temperature).
If it does, check the power supply
voltages on the board. REG1’s output
should be close to +5V, while pin 7 of
IC1 should be about 11.4V as should
50 Silicon Chip
pin 4 of IC2. REF1 should have close
to 2.5V across terminals 1 and 2.
Now check that the display shows a
temperature that’s close to the ambient
when the connected probe is exposed
to room air. Assuming it does, switch
S2 to position 2 (Thermostat) and
check that you can adjust the preset
using VR1. The relay should click on
There are a couple options available
if you just want the unit to measure
the ambient temperature.
First, you can use the thermocouple
as the sensor and simply sit it in free
air. Alternatively, you can disconnect
the thermocouple and short its inputs
on the PCB using a short length of wire.
The unit will then display the ambient
siliconchip.com.au
Controlling Mains Voltages
temperature (in °C) as measured by the
AD8495 itself.
Note that this will really be the temperature inside the case rather than the
room temperature. However, this will
be close to room temperature, since
there is little warming inside the case.
If you intend using this project simply as an ambient temperature thermometer or to measure temperatures
up to 199°C only, then the divider
resistors can be changed so that they
divide by five instead of 50. That way,
siliconchip.com.au
As presented in the diagram and photos, the Digital Thermometer/Thermostat is
capable of controlling external loads running at 30V DC and up to 10A. However, it
can control 230VAC loads, provided the relay and the wiring itself is rated for 250VAC
mains operation. This will mean that a larger case must be used to accommodate
the extra wiring and mains input and output sockets (note: the plastic case used here
is not suitable; it’s too small and the back is too flimsy to safely anchor mains cables).
The mains input wiring will need to include a mains fuse and we suggest an IEC
chassis-mount male socket that includes a switch and fuse (eg, Jaycar PP-4003). For
the output mains wiring, use a chassis or panel-mount female IEC socket (eg, Jaycar
PS-4176) or 3-pin mains panel-mount socket (eg, Jaycar PS-4094).
All mains wiring should be run in 250VAC 10A-rated cabling. Cable tie and clamp the
internal mains wires so they cannot possibly come adrift and contact any low-voltage
section of the circuit. It’s a good idea to secure the terminal block wires to the PCB; eg, by
using silicone sealant or a cable tie that loops through a couple of holes drilled through
the PCB adjacent to the terminal block.
A metal enclosure will need to be securely earthed. For a plastic case, any exposed metal
screws used to secure the IEC connector or other parts near to the mains wiring will also
need to be earthed. Nylon screws can be used as an alternative to earthing the screws.
The relay should be an Altronics S-4197 or exact equivalent, with contacts rated for
250VAC operation. Finally, for the 3-way terminal block, CON3, we recommend using a
Weidmuller type (Jaycar HM-3132) so that it has sufficient voltage rating.
the display can show the temperature
with a 0.1°C resolution.
To do this, change the 1kΩ resistor
to 12kΩ, the 39Ω resistor to 750Ω and
the 1.8Ω resistor to a 0Ω resistor (or
wire link). The 100Ω trimpot (VR4)
on the adjustable side of the divider
should be changed to 1kΩ.
Finally, the decimal point in front
of the righthand digit can be displayed
by connecting the LCD panel meter’s
DP3 pin to the +5V supply. The details
are shown on the instruction sheet
SC
supplied with the meter.
May 2012 51
Mssed an issue?
SILICON CHIP has available all back issues going back to 2003 and many issues before then.
(And if we can't supply a back issue, we can always supply a reprint of any particular article.
Project reprints also include relevant notes and errata). And it's not just for SILICON CHIP –
we can also supply reprints of articles from Electronics Australia/RTV&H and ETI!
The price for either a back issue or a project reprint is the same: $12.00 including P&P within
Australia; $15.00 inc P&P overseas.
Keep your SILICON CHIP collection intact – order your back issues today before they run out!
But there's an even better
way of ensuring you don't
miss an issue... subscribe!
A SILICON CHIP subscription positively guarantees that you will not only receive EVERY issue,
but there are several other advantages – for example, you get 12 issues for less than the price
of 11. Count the advantages of a SILICON CHIP subscription:
u
v
w
x
y
z
{
It's cheaper – you $ave money!
PRICE OF 12 ISSUES
OVER-THE-COUNTER
It's delivered right to your mail box!!
IN AUSTRALIA:
You can always be sure you'll receive it!!!
We pick up all the postage and handling charges!!!!
You will never miss an issue because it's sold out (or you forgot)!!!!!
You choose the length of subscription required: 6, 12 or 24 months.
11160
$
You can even choose to auto-renew your subscription at the end of the period!
Here's the deal:
SILICON CHIP : 52 in Australia; 55 in NZ*; 80 o'seas*
12 Months SILICON CHIP : 97 in Australia; 99 in NZ*; 140 o'seas*
24 Months SILICON CHIP : 188 in Australia; 196 in NZ*; 265 o'seas*
6 months
$
00
$
$
$AU
50
00
00
$AU
$AU
$AU
00
00
$AU
00
$AU
00
00
* VIA AIR MAIL
There's a handy order form on P105
52 Silicon Chip
siliconchip.com.au
ED MA
IT Y
IO
N
MEGA
MAY
Pr
ice
EtherMega, Mega Sized Arduino
va
lid
Compatible Board with Ethernet
un
til
The ultimate network-connected Arduino-compatible board: combining an ATmega2560
23
MCU, onboard Ethernet, a USB-serial converter, a microSD card slot for storing gigabytes
/0
5/
of web server content or data, Power-over-Ethernet support, and even an onboard
20
switchmode voltage regulator so it can run on up to 28VDC without overheating.
12
• ATmega2560MCU running at 16MHz
• 10/100base-T Ethernet built-in
• 54 digital I/O lines
• 16 analog inputs
• Prototyping area
• Size: 105(W) x 54(H) x 19(D)mm
XC-4256
More Arduino
Boards, Shields
& Modules on
page 8
11900
$
WANT A FREE COPY OF
OUR 2012 CATALOGUE?
High Power Wireless Outdoor
Router/Range Extender
802.11n
Place an order of $30 or more
via our Techstore website and
type "FREE CATALOGUE" in
the comment box as you
check-out.
Offer valid until 31/5/2012.
10-Way Headphone Listening
Centre with Microphone
Distributes audio signal across up to 10 headphones
and has a built-in amplifier which prevents loss of
sound quality. Each channel has its own volume
control. Supplied with a mains power adaptor, 1 x
dynamic microphone and 1 x 2 metre 3.5mm plug to
6.5mm plug stereo lead.
8” Colour LCD Doorphone
System with 4Ch Recording
• Output power:
220mW (32ohm)
• Headphone
Impedance:
16-64ohm
• Size: 191(W) x
95(H) x 45(D)mm
AA-0403
The next generation in video doorphone entry
systems. The 8-inch high resolution screen
connects up to two intercom/cameras and you
can optionally connect up to 4 additional security
cameras (QC-3639 available separately) to the
screen, giving you four-channel security. The
monitor can then display a multi-way split screen
view, or auto-switch between channels.
• Quad split screen or full screen display
• Total of 6 video inputs
• Hard drive capacity: 2.5"/ 500GB Hard drive
• Remote access over Ethernet or Internet
• Mains power supply included
Due Early May
• Monitor size: 210(H) x
250(W) x 35(D)mm
QC-3628
Additional Intercom Camera
to suit QC-3629 $79.00
NEW
• Weight: 200g
• Size: 114(H) x 74(W) x 29(D)mm
QT-2304
NEW
149
$
00
MEGA
SALE!
Install the included software, plug in the
encoder and you're ready to convert music from your
cassettes to digital MP3 or WAV format! Doubles as a
handy dictaphone and tape player with built-in speaker.
• Windows compatible
• Power via 3VDC adaptor or 2 x
AA batteries (not included)
• Size: 90(L) x 116(W) x 36(D)mm
GE-4053
Was $89.95 $
95
NEW
199
$
00
To order call 1800 022 888
5W VHF Marine
Radio Transceiver
Powerful 5W hand-held transceiver gives you
coverage of all International VHF marine
channels. The antenna is detachable so units
can be connected to a larger antenna mounted
on a boat. Includes Li-ion rechargeable battery
pack, AC adaptor, charging cradle and belt clip.
• 1W/5W switchable output power
• LCD backlit display
• Dual and triple watch function
• Channel scan function
• Auto power saver mode
DC-1096 Was
$199.00
11900
$
SAVE $40
Limited Stock.
Not available online.
A signal generator with the features of a bench top generator and a
portable size! This pocket signal generator will produce sine,
square, and triangle waveform signals yet it is only a little
bigger than a Smartphone. Output frequency adjustment is
from 1Hz to 1MHz with maximum amplitude of 8Vpp. It also
has a function to shift between two frequencies over an
adjustable period. With a backlit LCD, inbuilt rechargeable
battery, and durable rubber surround it is an ideal
instrument for testing on the go or in your workshop.
See website for specifications.
NEW
16900
Cassette Tape to MP3
Encoder with USB
1MHz Handheld Function Generator
• LAN port 10/100 (featuring PoE)
• PoE power injection adapter included
• 12VDC 1A Power supply included
• High data rate - up to 150Mbps
• Size: 225(H) x 77(W) x 59(D)mm
YN-8330
$
49
54900
$
Has a wireless power of 600mW and excellent point
to point range with the installed internal directional
antenna. If omni directional transmission is required
then you can connect an external antenna (sold
separately) via the SMA connector at the base.
SAVE $80
Digital DC Power Meters
Displays both continuous and peak voltage, current, and power. Cumulative amp
hours and watt hours consumed are also stored allowing you to monitor the system
over time. Suitable for DC operation from 5 to 60V.
An ideal addition to low voltage DC circuits on
boats, caravans, or solar systems.
• Size: 41(L) x 45(W) x 23(D)mm
DC Power Meter with Internal Shunt
MS-6170
$69.95
DC Power Meter to suit 50mV External Shunt
MS-6172
$74.95
Suitable DC shunts sold separately
Due Early May
From
6995
$
www.jaycar.com.au
MEGA MOTHER'S DAY - 13th May
1kg Digital Bench Scale
Waterproof Bathroom Clocks
Sports Stopwatch
• Requires 1 x AA battery
• Approx. 130(Dia.)mm
• Split time, alarm and
calendar function
• Includes 600mm lanyard
• Size: 55(W) x 65(H) x 15(D)mm
XC-0270
Truly convenient and waterproof analogue clocks.
The suction cup allows you to mount it on to any
smooth surface like a tiled wall or
mirror. Available in blue and pink.
Precision 1kg electronic
scale with resolution of
0.01g for when a high
degree of accuracy is
required. Weighs in grams,
ounces, pounds, grains,
carats and troy ounces.
Blue AR-1757 $9.95
Pink AR-1758 $9.95
• Auto power-off after 60 seconds
• Automatic calibration
• Backlit LCD
$
• Tare and counting function
• Mains powered
• Size: 175(W) x 75(H) x 260(D)mm
QM-7264
995
• Countdown range 99 hours 99
minutes 99 seconds
• Batteries included
• Size: 88(W) x 130(H) x 22(D)mm
XC-0271
Pack one of these on your next trip
and avoid nasty surprises at the
check-in counter.
2495
$
Non-Contact Digital Thermometer
Featuring an easy to read LCD. Handy for use in the kitchen
or the food service industry to ensure
proper cooking, grilling and
storage temperatures.
• Pocket sized with LCD
• Fast response time
• Temp range: -35 - 230°C / -31
446˚F
• Battery included
• Size: 74(L) x 40(W) x 20(D)mm
QM-7225
3995
Ultrasonic Cleaner
MEGA
SALE!
Massive 100W transducer produces
millions of microscopic bubbles that
penetrate and clean the most microscopic of
crevices, cleaning them thoroughly. Used for
automotive injectors,
jewellery, glasses, circuit
boards and more!
2
Small in size but won't cover up your pictures,
notes or shopping lists. These nifty fridge
magnets are strong enough to hold up to 10
sheets of paper.
• Pack of 5
• Size: 20(H) x
11(Dia.)mm
LM-1629
29900
Robust construction to take the inevitable bump
and occasional crash. 20 minute charge directly
from the remote unit gives about 10 minutes
flying time.
• Remote requires 6 x AA batteries
• Recommended for ages 8+
• Size: 160mm
long approx.
GT-3306
Was $29.95
SAVE $100
Add a bit of bling to mum’s desk with
these glittering pink and white
rhinestone-finished accessories.
Don't forget
Mother's Day
13th May!
Rhinestone Calculator
It has an 8 digit LCD and all the features
of a regular office calculator. Can be
battery operated or solar powered.
• Size: 160(L) x
39(W)mm
GH-1894
Was $19.95
7
$ 95
SAVE $12
$
995
SAVE $10
Rhinestone Hub
This USB 4 port hub is sure to add some style and
class to mums boring old desk. Simply
plugs into the computer's USB ports.
• 90mm lead
• USB 2.0 compliant
GH-1898
Was $29.95
1495
$
SAVE $15
Note: Products above are limited in stock and not available
online. Please ring your local store to check stock.
2495
$
Limited in stock. Not available online.
Water Misting Fan
• Requires 2 x
AA batteries
GH-1071
2995
$
This stapler takes size 56 staples and
will finish off any desk with style.
Powerful enough to provide a
significant breeze but safe
enough for the kids to use
with soft blades.
$
• Range 7-10 metres
• Transmitter size: 60(L) x
32(W) x 7(D)mm
XC-0354
Rhinestone Stapler
995
SAVE $5
Locate misplaced objects such as
keys, TV remote, glasses, wallet etc.
Simply tag your keys and press the
master transmitter (colour coded) and
the lost item beeps back. Supplied
with 1 x keyfob transmitter and 3
x separate receivers.
• Battery included
• Size: 145(L) x 100(W)mm
GH-1892 Was $19.95
Mini 3 Channel RC Helicopter
$
Wireless "Object Locator"
MEGA Desktop Bling For Mum
Rare Earth Fridge Magnets
3495
• Hot 1300°C adjustable flame
• Size: 150(H), base 69 x 69mm
TS-1660
• Large LED
display
• Mains powered
• Tank
capacity: 3L
• Size: 265(L) x
160(W) x 245(H)mm
YH-5410 Was $399.00
1995
$
Very versatile and ideal for brazing, silver soldering, jewellery
work, plumbing or general hobby use.
Butane gas refill:
NA-1020 $5.95
• Requires 2 x AAA batteries
• Capacity: 40kg
• Backlit LCD
$
• Overload and low
battery indication
• Size: 122(L) x
85(H) x 25(W)mm
QM-7232
$
Upright Gas Torch
95
Each
Digital Luggage Scale
Use it for cooking, parking, exercising, studying
or even timing the kids on the computer. It's
water resistant, has a memory setting for
frequently used values and the buzzer alerts
you to when your time is up.
For the
sporty
mum!
12
$
$
14900
Mum’s Kitchen Helpers
Countdown Timer
This water resistant handy sports
timer will be a useful addition
to mums sports bag.
895
$
To order call 1800 022 888
Deluxe Automatic
Soap Dispenser
Automatically dispenses a
measured amount of liquid
soap when you put your
hands under it. No more
touching soap bottles which
reduces the risk of transferring germs.
• LCD with auto cleaning mode
• Requires 3 x AA batteries
• Size: 195(H) x 85(W) x 160(D)mm
GH-1188
2995
$
All savings based on Original RRP. Limited stock on sale items.
Prices valid until 23/05/2012.
MEGA AUDIO & VIDEO
Earphones Adaptor with Mic
Turntable with Built-in
Pre-amp & PC Encoding
Converts old vinyl to modern digital
WAV, MP3 or OGG format using a
PC. It has a high quality magnetic
cartridge which ensures clear and
distortion free conversion. With a
built in pre-amp, it has a line level
RCA output for component or
amplifier integration.
9900
$
Simply connect this adaptor to both your iPhone®
and headphones and listen away. With its built-in
mic, you can answer a call without even having to
take your headphones off! Simply click to answer
or to play/pause the music.
• Music playback
• 3.5mm headphone jack
AA-2098
• 8-ohm impedance
• 25mm dome tweeter
• White in colour
NEW
• Mains powered
• Size: 372(W) x 352(D)
x 104(H)mm
GE-4134
6.5" Polycone Woofer
1295
$
2.4GHz Omni Antenna 12dB
USB Dynamic Microphone
Plug and play with any recording software or the
recording interface on your PC for superior sound
quality on your next podcast or home recording
masterwork.
• Easy connectivity
• High sound quality
• Type: USB unidirectional dynamic
• Termination: 3m cable,
USB 2.0 plug
AM-4103 Was $39.95
2995
$
SAVE $10
Suitable for boosting signal strength and range in
WLAN applications. The fibreglass construction
provides excellent weather proofing properties
making it ideal for outside
NEW
mounting.
$
00
• Power handling: 20W
• Size: 960(H)mm x 20(Dia.)mm
AR-3284
Kitchen Voice Recorder
Designed for box office style communication through
protective glass such as ticket booths, bank counters,
reception desks, or nightclub entry points. It features
an adjustable gooseneck and volume adjustable
rotary button at the base.
9900
$
• 3 folders up to 50 messages
• Digital clock
• Requires 2 x AA batteries
• Size: 130(L) x 66(H) x 17(W)mm
XC-0249 Was $69.95
Active VGA + Audio to HDMI
Converter
This converter box takes the
VGA output + stereo audio
signal from your PC, and
converts them to HDMI
format whilst maintaining
full high-definition resolution.
Mains adaptor included.
• Size: 90(L) x 68(W)
x 25(H)mm
AC-1609 Was $99.00
Use it for shopping
lists, as a reminder
for the kids or
even to record
the ingredients in
those quick
cooking commercials.
Strong magnetised backing.
NEW
• Requires 2 x
AA batteries
• Uni-directional electret
condenser
• LED Indicator (power, on-off,
low battery)
• Size: 98(H) x 66(W) x 25(D)mm
AA-4089
99
Suitable 5m lead (N plug to SMA plug) available
separately NEW WC-7822 $34.95
Microphone Intercom Speaker
Don’t forget
mothers
day!
3495
$
SAVE $35
Socket Wallplate - TV & FM
Standard power point size. Has two PAL sockets,
one for TV and one for FM. Simply run one cable
in from antenna.
• F59 rear connection
• Size: 71(W) x 115(H)mm
LT-3066
Also available:
89
$
00
SAVE $10
Flushmount 75 Ohm TV Wall
Socket with F59 Connection
LT-3065 $6.50
Powered CAT5 VGA Baluns with Audio
2-Way Ceiling Mount Indoor Speakers
Combining a coaxial woofer and dome tweeter these 2 way
ceiling speakers give excellent audio quality. Both models
feature custom designed crossovers and high performance
tweeters matched to the
woofers, providing much
better sound reproduction
over the full music range.
NEW
995
$
• Resonant Frequency: 70Hz
• Rated Power: 30WRMS
CS-2446 Was $99.00
8" Polycone Woofer
Sold as
a pair
7900
$
SAVE $20
• Resonant Frequency: 50Hz
• Rated Power: 50WRMS
CS-2448 Was $139.00
11900
$
SAVE $20
Kingray Masthead Amplifier
Accepts a single "mixed" antenna input, and
provides four amplified outputs for you to run to
each wall point in your home. All connections are
F-type for best signal quality. Housed in a metal
case to shield against noise. Includes power
supply.
• Wide input
range to suit
all analogue
and digital TV
signals
• Suitable for
combined VHF/UHF antenna input
• Replacement power
supply (LT-3256 $23.95)
• Size: 105(W) x 90(H) x 35(D)mm
LT-3253
NEW
7995
$
4-Input HDMI Switcher with
Digital Audio Splitter
Get high quality
picture and sound
while keeping your
older surround sound
system. This switcher
will select between 4 HDMI
inputs and also separates the
digital audio signal from for
connection to your older
amplifier system.
7900
$
SAVE $20
• 1 x HDMI output with coax and SPDIF audio output
• HDTV 1080p resolution
• Size: 155(W) x 70(D) x 23(H)mm
AC-1625 Was $99.00
Stereo 2.4GHz Digital Wireless Amplifier System
Transmits VGA and audio signals across a standard CAT5 cable for distances up
to 300 metres. Suitable for home or commercial applications where a standard
VGA cable can't reach or to run VGA signals through
existing wiring in a wall or ceiling. Supplied as a sender
$
00
and receiver pair with plugpacks included.
Send crystal clear audio from your Hi-Fi or portable music device to speakers up to
20m away without messy wiring. Connect your speakers to the spring terminals
and power using the included power supplies or by batteries.
Supplied with 2 x 150mm 3.5mm curly
cables to connect your audio source.
$
00
• Supported resolutions from 640x480 to 1920x1200
• Sender Size: 100(L) x 65(W) x 26(H)mm
• Receiver Size: 81(L) x 43(W) x 23(H)mm
AC-1671
• Class T amplifier design
• Power output: 15WRMS x 2 (into 4 ohms)
• Transmitter and receiver requires 8 x AA batteries each
• Size (transmitter and receiver): 156(L) x 45(H) x 95(W)mm
AR-1895
119
Better, More Technical
129
www.jaycar.com.au
3
MEGA AUTO ACCESSORIES
12V HD Digital Set Top Box
Portable Car Safe
Keep your mobile phones,
cameras, wallets, keys or GPS's
protected with this portable car
safe. Simply attached the included
1m alloy cable strap to your seat railing
and neatly tuck it under your seat
or place in the boot of your car.
Great for use on the road, this high definition set top
box will pickup all the digital channels on offer in
your locale. You can also plug in a USB drive and
record TV in MPEG2 format to watch at a later date.
NEW
24
$
95
• Size: 210(L) x
150(W) x 68(H)mm
HB-5455
Note: iPhone® and notes not included
Response Precision Car Amplifiers
• Output: HDMI, Composite, RF
• USB port for recording and playback
• Cigarette lighter cable included
• Size: 154(W) x 117(D) x 40(H)mm
XC-4921
NEW
6995
$
With improved heat sinks and
upgraded low-profile
chassis design,
each model
delivers
outstanding
performance package
that fits neatly under
your car seat.
Includes gold plated
power and speaker terminals
with variable low pass filters. In
addition, our class AB amps come with variable high
pass filters and pass through RCAs; while our class D
subwoofer amps feature variable
From
subsonic filter, phase shift and
$
00
master/slave operation.
5-in-1 Jump Starter
2 x 80WRMS Class AB Amplifier
4 x 50WRMS Class AB Amplifier
2 x 150WRMS Class AB Amplifier
4 x 100WRMS Class AB Amplifier
500WRMS Linkable Class D
Subwoofer Amplifier
1000WRMS Linkable Class D
Subwoofer Amplifier
4 x 100WRMS Full Range
Digital Amplifer
When the vehicle's battery falls below
11.2V, the Battery Protector will
automatically suspend power to any
connected 12V accessory, saving
enough battery power to start the car.
119
AA-0450
AA-0451
AA-0452
AA-0453
$119.00
$149.00
$169.00
$229.00
AA-0454 $179.00
AA-0455 $299.00
AA-0457 $259.00
Excellent for automotive or camping adventures.
Includes heavy duty insulated jumper
leads, a 400W inverter, LED
worki light, 12V power outlets,
status gauges, and even a
260 PSI air compressor!
Powered from the built-in
18Ah SLA battery and
comes with mains and
12V charging cables.
• Size: 295(H) x
270(L) x 215(W)mm
MB-3594
DC-1005
Was
$54.95
38 Channel UHF
CB Radio Mini
DC-1008
Was
$19.95
Advanced 2 Watt 38
Channel UHF Transceiver
DC-1047 Was $99.95
5995
$
SAVE 25
$
1 Farad Capacitor
High farad capacitors act as
surge current reservoirs for your
amplifiers and other electrical
equipment. Integrate these
capacitors into your audio system
to avoid voltage drops from high
transient current peaks.
14900
• Gold plated terminals
• Digital voltage display
• Size: 260(H) x 75(Dia)mm
RU-6754 Was $99.00
$
3 Watt 38 Channel
UHF CB Radio with
Scrambler & CTCSS
DC-1060
Was $169.00
• Includes glue for
installation
99
00
149
10" 200WRMS CS-2351
Was $249.00 Now $149.00 Save $100.00
12" 250WRMS CS-2353
Was $299.00 Now $179.00 Save $120.00
4
95
19Each
$
SAVE $20
7" Pad (pair) AX-3665
Was $39.95 Now $19.95 Save $20.00
12" Pad (single) AX-3666
Was $39.95 Now $19.95 Save $20.00
Note: Products above are limited in stock and may not available at all stores.
Please ring your local store to check stock availability. Not available online.
These subwoofers produce high sound quality and
outstanding performance. With dual voice coils, high
power handling and die-cast aluminium chassis,
they don't just deliver brilliant low register bass
clarity but also thump
From
tremendous SPLs like
$
00
only VIFA speakers can.
$100
SAVE
SAVE $20
Mainly designed for car audio,
but could be used in any
speaker application. These
pads are installed inside the
door skins opposite the
back of the speaker
drivers. They absorb
standing waves and
resonances so you get
maximum performance.
SAVE $4
SAVE $5
VIFA Subwoofers
7900
$
Sound Dampening Pads
1995
$
$
From
2495
$
SAVE $70
1495
2995
4" 15WRMS Speakers
CS-2310 $24.95
5" 17WRMS Speakers
CS-2312 $29.95
6" 22WRMS Speakers
CS-2314 $34.95
6 x 9" 27WRMS Speakers
CS-2316 $44.95
$
• Plugs directly into the cars
cigarette lighter socket
• LED indicator
• 10 amp fuse protection
• 1metre supplied lead
MS-6120 Was $23.95
SAVE $40
$
Ideal replacement for the
standard equipment
stereo speakers. All are
equipped with titanium
coated fibre woofers
and silk dome tweeters
for smooth high
frequency response.
Battery Protector
CRAZY CB RADIO CLEARANCE
38 Channel UHF CB
Dual Pack with Charge
Coax 2-Way Car Speakers
Twin Port Subwoofer Enclosures
MEGA
SALE!
Dual ported subwoofer enclosures with black carpet
covering. Designed for optimal performance with the
VIFA 10" and 12" subwoofers. All you need to do is to
add the driver of your choice.
10" CS-2526 Was $39.95
Now $19.95 Save $20.00
12" CS-2527 Was $49.95
Now $24.95 Save $25.00
Note: VIFA driver not included
To order call 1800 022 888
From
1995
$
SAVE $20
MEGA
SALE!
All savings based on Original RRP. Limited stock on sale items.
Prices valid until 23/05/2012.
MEGA SECURITY
3MP Mini HD Video Camera
Monitor areas wirelessly
with these PIR motion
detectors. The main
control panel is alerted
when movement is
detected and triggers
the announcer. Ideal for
retail stores, offices or work
shops. Two models available.
Detect covert cameras and
listening devices with this handy
little unit. It uses 6 pulsing LEDs
to reveal the location of a
camera by illuminating its lens
when you look through the lens
viewer from up to 10m away.
Earphones supplied.
• 350mAh rechargeable
battery included
• Supports up to
$
00
32GB MicroSD card
• Size: 23(H) x
SAVE $44
78(H) x 14(D)mm
QC-8005 Was $119.00
75
• Built-in wireless RF detector
• Requires 2 x AAA batteries
• Size: 85(H) x 56(W) x 18(D)mm
QC-3506 Was $99.00
CRAZY CAMERA CLEARANCE
Product
QC-3467
QC-3494
QC-3496
QC-3498
QC-3310
QC-3309
QC-3298
QC-3299
QC-3307
QC-3301
QC-3297
Wireless PIR Announcers
Camera Detector
Ultra portable, compact HD video camera and
recorder with 2GB of internal memory that will hold
up to 50 minutes of video (20 minutes in high
definition) or over 3000 photos. Recharges via
USB and will gives about 4 hours of use.
Pocket clip and desk stand included.
Description
Camera CCD Bullet B&W IP57
Camera CCD Colour in Metal Case Mini
Camera CCD Colour Pinhole in Metal Case
Camera CCD Dome Style Colour
Camera CCD Pro Style B&W
Camera CCD Pro Style Colour
Camera CCD Pro Style Colour ExView HAD
Camera CCD Pro Style Colour ExView Hi-Res
Camera CCD Pro Style Hi-Res Colour
Camera CCD Pro Style Hi-Res Day/Night
Camera Dome Vari-Focal 480TVL w/ Bracket
ORRP
$99.00
$99.00
$99.00
$279.00
$109.00
$179.00
$249.00
$349.00
$299.00
$299.00
$299.00
00
SAVE $20
An economical digital video recorder which
incorporates a 500GB hard drive, 4 channel
multiplexer, Ethernet functions, H.264 video
compression, and even delivers D1 (704 x 576)
resolution video (playback, live view, & recording) at
100 frames per second. Use it to record up to 4
cameras simultaneously with playback available
locally, via a network connection, or using an iPhone®
or Smartphone app*.
MEGA
SALE!
12V CCTV Power Supplies
QC-3310
QC-3297
2-Station Wired Intercom
• Operates on 9V battery
or 240V plugpack
• Supplied with 20m of connecting cable and staples
AM-4310
Note: Domestic use only. Warranty does
not apply if product is used industrially.
Note: *Free application available to
view live footage. Application based
searching and backup requires
advanced version at an additional cost.
29900
$
SAVE $100
2495
$
A truly versatile monitor with low power
consumption, wide viewing angle and NTSC and
PAL compatibility. Use it
to watch DVDs, PS2®,
XBOX®, etc. Unit comes
with an adjustable swivel
bracket with double
sided tape. Infrared
remote control included.
• Power input: 12VDC
• Resolution: 1140(H) x 234(V)
QM-3752
2.4GHz Rear View Mirror Reversing Camera
Incorporates a reversing camera that transmits video signals via the 2.4GHz
band to the monitor which can be mounted internally
or externally. The monitor fits securely
over your existing rear view
mirror and can be quickly
removed when needed.
• 3.5" LCD colour screen
• Range: up to 80m
• Size: 280(L) x 95(H) x
26(D)mm
QM-3795 Was $199.00
Better, More Technical
14900
$
SAVE $50
Mains power supplies suitable for CCTV
installations, with multi-channel outputs
for each individual camera. Housed in
a rugged lockable
steel enclosure
designed for
permanent
professional
installations. Must
be installed by a
licensed Electrician.
MEGA
SALE!
From
4 Channels MP-3850
$3995
Was $69.95 Now $39.95 Save $30.00
SAVE $30
8 Channels MP-3852
Was $129.00 Now $75.00 Save $54.00
A simple low cost solution for
communicating. Either station can
signal the other even when the
system is off. It can even
be used as a room/baby
monitoring system.
7" TFT LCD Colour Monitor
• Power supply: 19VDC 2.1A (included)
• Size: 343(W) x 240(D) x 68(H)mm
QV-8107 Was $399.00
LA-5172
shown
Due Early May
Note: Products above are limited in stock and may not available at all stores.
Please ring your local store to check stock availability. Not available online.
4Ch H.264 Network DVR with
D1 Resolution
5995
Battery Powered
LA-5172 $59.95
Battery and Solar Powered LA-5174 $79.95
QC-3467
SPECIAL SAVE $$$
$49.00 $50.00
$49.00 $50.00
$49.00 $50.00
$99.00 $180.00
$69.00 $40.00
$59.00 $120.00
$109.00 $140.00
$149.00 $200.00
$129.00 $170.00
$129.00 $170.00
$139.00 $160.00
From
• Detection range: up to 5m
• Unit size: 122(W) x 143(H) x 52(D)mm
PIR size: 75(W) x 120H) x 60(D)mm
79
$
NEW
$
GPS/GSM Tracking Device
The solution to locate and track the whereabouts of your
vehicle in real time via the Internet on a computer or
Smartphone. The device is easy to install with only 2 - 4
wires to connect. Insert a GSM Sim card (not included) and
hide the device away on a metal surface or carpet using the
included Velcro adhesive. It works by sending the vehicle's
GPS coordinates via the GSM network to
the free online tracking service, which
shows the location on Google Maps. It
can also SMS longitude and latitude
coordinates to up to 3 mobile phones.
See web for full features and specs.
• Size: 68(L) x 48(W) x 20(D)mm
LA-9011
NEW
11900
$
14900
$
8Ch CCTV Power Supply with Battery Backup
To round out any professional CCTV installation, some kind of power backup
protection needs to be provided in case any would be thieves decide to cut the power.
This handy CCTV power supply solves both the problem of supplying power to a
multiple camera installation and providing that
power backup.
• Requires backup battery: 12V 7Ah size
SLA (use SB-2486 $29.95)
• Size: 263(L) x
$
00
195(W) x 64(D)mm
MP-3855
SAVE $60
Was $149.00
89
www.jaycar.com.au
5
MEGA POWER
A high powered switchmode power supply with variable
voltage output from 1 to 16VDC and variable voltage from 0
to 40A. Features dual action (coarse/fine) microprocessor
controlled rotary encoder tuning for
smooth, precise and
fast settings, 3 user
defined voltage and
current presets, and
intelligent fan cooling
control. See website for
features and specifications.
• High RFI immunity and excellent EMI
• Overload, short circuit, over
temperature and tracking over
voltage protected
• Size: 200(W) x 90(H) x 215(L)mm
MP-3094
Rotating Surge
Protector with 2 x
USB Outputs
Universal Battery Charger
Laboratory Power Supply
A great universal battery charger for Li-ion battery
packs, AA, AAA and 9V Ni-MH and
Ni-Cd rechargeable batteries.
A USB port on the side
accommodates the
charging needs for
your iPhone®,
Smartphone or any
USB power device.
Features a 90° rotating design for
easy GPO switch access and 2
USB charging ports with 2.1A
(total) for fast charging your iPad®,
iPhone®, iPod®. This compact unit
also has attractive illuminated
indicators showing that power and
surge protection are operating.
NEW
• USB output: 5VDC, 500mA
• Size: 120(L) x 62(W) x 35(H)mm
MB-3639
39
$
95
Note: Batteries not included
NEW
39900
$
Dimmable Constant
Current LED Driver
A compact mains powered
unit, capable of driving 14 high power LEDs at a
constant current of 700mA
(10W max), whilst also being
dimmable, it is also an excellent driver for domestic
LED lighting projects or as a replacement for a
failed LED driver. See website for specs.
• Overload and short circuit protection
• Dimmable with leading edge or
NEW
trailing edge triac dimmer
• Size: 112(L) x 39(W) x 25(H)mm
$
95
MP-3365
29
Mains Power Meter with CO2
Measurement
This meter tells you the cost of electricity
consumption of an appliance plugged into it and
the amount of power used in kilowatt hours, as
well as how many cumulative kg of CO2 the
appliance is putting
into the atmosphere.
• Extra large LCD
for easy reading
• Size: 120(L) x
58(W) x
BUY 2 for
40(H)mm
$40.00
MS-6118
SAVE $19.90
NEW
2995
$
• Size: 112(H) x 57(W) x 42(D)mm
MS-4027
80W Standard Recreational
Solar Package
Clean renewable energy wherever you go. Solar-convert your
4WD or caravan to generate sufficient power to operate
several appliances - including your laptop, portable lighting,
CB radio and 12-24V camping electricals. Just add a battery
for your own self-sustained solar powered setup.
2995
$
Pure Sine Wave
Inverter/Chargers
MEGA
SALE!
Wind Turbine Generators
Always at the forefront of
alternative energy technology,
we’re pleased to offer a great
range of wind turbine generators.
All models feature external charge
controllers with three phase AC output.
Combining the functions of a pure sine wave
inverter, battery charger and automatic transfer
switch in one unit. When connected to the
mains, the connected batteries are charged.
If the mains is interrupted or exceeds the
allowable limits, power is drawn from the
batteries and mains power is provided by the
inverter. No more manual
From
switching from mains to
$
00
battery power!
• No of Blades: 3
• Includes: generator, blades, tail, hub,
nose cone, charge controller
200W 12VDC MG-4520
Was $499.00 Now $449.00 Save $50.00
300W 12VDC MG-4580
Was $699.00 Now $649.00 Save $50.00
300W 24VDC MG-4582
Was $699.00 Now $649.00 Save $50.00
NOTE: These wind generators are designed
for permanent terrestrial installations only.
Mounting tower and hardware not included.
Not suitable for marine use.
More Recreational Solar Packages
available that suits your needs. See
our friendly staff for details.
Spare Parts
Available
Separately
From
44900
$
• 12VDC, 20A
MEGA
SALE!
30000
$
SAVE $65
Non-Insulated Spade
Connectors NEW
1099
SAVE $400
• 6.3mm
• Pack of 10
1500W Inverter/Charger MI-5260
Was $1499.00 Now $1099.00 Save $400.00
2000W Inverter/Charger MI-5262
Was $1799.00 Now $1399.00 Save $400.00
275
$
/Pack
Socket PT-4630 $2.75
Plug PT-4631 $2.75
SAVE $50
IP67 Waterproof LED Flexible Strip Light
LED Lighting Strips
Flexible Adhesive LED Strip Lights
Made using the highest brightness 5060-SMD type LEDs, and feature 60 of these
LEDs per metre of stripping. Each strip comes as a 5m length, which can be broken
down into individual 5cm sections with 3x LEDs that can be
NEW
individually soldered to apply power. Sold by the section and
cut to your desired length. See website for full specifications.
$
95
Two colours available:
/metre
2
Cool White ZD-0570 $2.95
Warm White ZD-0572 $2.95
6
• Package includes: monocrystalline
solar panel, charge controller, 2 x
male and female PV connectors
ZM-9300 Was $365.00
To order call 1800 022 888
A 1m long fully waterproof, flexible LED strip light that is perfect for any outdoor
application needing reliable lighting. Uses 60 of the highest brightness 5060SMD type LEDs that are fully sheathed in a protective plastic casing to protect
from water, dust and damage. See website
for full specifications.
• Powered by 12VDC
• Size: 1000(L) x
10(W)mm
ZD-0579
NEW
4995
$
All savings based on Original RRP. Limited stock on sale items.
Prices valid until 23/05/2012.
MEGA HARDCORE
Electronic Flow
Rate Meter with LCD
Solder Paste SMD Syringe
Supplied with a reed
switch and piezo alarm, it
operates from 2 x AAA
batteries and a
battery holder is
included. When
Note: Batteries not included
used in conjunction
with the Digital
Flowmeter (ZD-1202
available separately), it will count
down in litres from a predetermined volume and
measure flow rate, and can be used in large-scale
applications such as irrigation (up to 99,500 litres).
Full data sheet and instructions available on
website.
$
95
• PCB/LCD size: 60(L) x 40(W)mm
ZD-1204 Was $69.95
$
1295
$
Features a large, easily read display and IP67
rating, making it waterproof.
SAVE 20
• True RMS
• CatIV, 600V, 4000 count
• Data hold & relative function
• Auto off & backlit display
• Diode test & audible
continuity
• Autoranging
• 10A current range
QM-1549
Digital Flowmeter ZD-1202
Was $49.95 Now $29.95 Save $20.00
Super Heavy Duty Heatsink
• Size: 125(W) x 75(L)mm
HH-8590 Was $19.95
NEW
IP67 True RMS Autoranging
CatIV DMM
49
This black anodised heatsink is
36mm high. Unique
recessed channels are
incorporated into the flat
side to accommodate screws,
panels, heatsink brackets etc.
Composed of a polyurethane base
designed to electrically insulate
and protect against dust and
moisture.
Ideal for surface mount work and rework.
Easy application, simply apply it to the
soldering pads, put your components in
place and heat it with your soldering iron.
• 15g
• Size: 120 (L) x 15(Dia.)mm
NS-3046
USB Datalogger
995
Resistance Wheel
Great for experiments or
selecting the best resistance
for a circuit. Select from 36
values ranging from 5 ohms
to 1M ohms.
Jacob's Ladder High Voltage
Display Kit MK2
Refer: Silicon Chip Magazine April 2007
With this kit and the purchase of a
12V ignition coil (available from
auto stores and parts recyclers),
create an awesome rising ladder
of noisy sparks that emits the
distinct smell of ozone. This improved circuit is suited
to modern high power ignition coils and will deliver a
spectacular visual display. Kit includes PCB, pre-cut
wire/ladder and all electronic components.
7995
14
SAVE $5
NEW
$
19
$
95
• Allow 15 minutes for
setting time
• Cures in around 4-5 hours
• 70ml
NM-2016
• Comes complete with leads
and insulated alligator clips
• Uses 0.25W resistor with
$
95
5% tolerance
RR-0700 Was $24.95
SAVE $5
NEW
$
Polyurethane
Potting Compound
Open Wall Mount
Rack Enclosures
These USB dataloggers log temperature and humidity
readings and store them in internal memory for later
download to a PC. The
measurement interval
is adjustable - simply
set up the recording
parameters then download
the data when you need it.
Ideal for mounting in
other enclosures, such
as road cases, but
can also be mounted
stand-alone. One side
is hinged so that patch panels
can be easily accessed at the
rear for reconfiguring patch sets.
• Windows compatible
• Sizes: 100(L) x 22(W) x 20(H)mm
QP-6013 Was $119.00
2U HB-5190 Was $49.95 Now $39.95 Save $10.00
4U HB-5192 Was $59.95 Now $49.95 Save $10.00
Simply insert your wires and squeeze it closed
to connect and crimp.
Quick, easy,
and secure.
Ferrule Crimp Terminals
• Pack of 6
9900
$
SAVE $20
100 Piece Driver Bit Set
The ultimate driver bit set. It has a
magnetic holder, adaptors, Phillips
bits, slotted bits, torx,
tamperproof, pin drive, and even
a wing nut driver - Fantastic.
See web site for full listing.
Commonly referred as a
bootlace crimp, these lugs
are designed to neatly
terminate cables before
inserting them in a spring
loaded or screw-down
terminal.
TD-2038 Was $19.95
• Pack of 20
White PT-4433 $2.95
Red
PT-4533 $2.95
Blue PT-4633 $2.95
995
$
SAVE $10
A range of non-insulated eye terminals for a variety of electronic
or automotive applications.
Pk 8
Pk 8
Pk 8
Pk 8
PT-4930 $2.75
PT-4932 $2.75
PT-4934 $2.75
PT-4935 $2.75
8mm 10mm2
8mm 25mm2
8mm 35mm2
8mm 50mm2
39
$
95
SAVE $10
Pk 8
Pk 4
Pk 4
Pk 2
NEW
295
$
/pack
Better, More Technical
From
PT-4936 $3.45
PT-4937 $3.95
PT-4938 $5.95
PT-4939 $4.95
4295
$
Quick Splice Connectors
NEW
275
$
/pack
149 Piece Pink Tool Set
275
$
• PCB: 170 x 76mmm
KC-5445
Red
PT-4537 $2.75
Blue PT-4637 $2.75
Yellow PT-4737 $2.75
NEW
Non-Insulated Eye Terminals
4mm 2.5mm2
6mm 4mm2
6mm 6mm2
8mm 6mm2
From
Contains a hammer, long nose pliers, multigrips,
tape measure, screwdrivers, shifting spanner,
shears, driver with 20 bits, 8-piece Allen key set, 6
jewellers screwdrivers plus an assortment of nails,
screws and other fasteners. An easy-to-follow Howto booklet is included
on each tool and
common
household tasks.
3495
$
SAVE $15
PT-4934
PT-4939
• Case size: 250(W) x
322(H) x 65(D)mm
TD-2075 Was $49.95
www.jaycar.com.au
Don’t forget
Mother’s Day
13th May!
7
ARDUINO KITS
LeoStick Arduino Compatible
A tiny Arduino-compatible board that's so small you can
plug it straight into your USB port without requiring a
cable! Features a full range of analogue and digital
I/O just like its larger cousins, and also has a
user-controllable RGB LED on the board
and an on-board Piezo/sound generator
so you can make your board light up
and play sounds without any extra
hardware at all!
• ATmega32u4 MCU with 2.5K RAM
and 32K Flash
• 6 analogue inputs (10-bit ADC) with
digital I/O, 14 extra digital I/O pins
XC-4266
NEW
29
$
95
LeoStick Prototyping Shield
Add your own custom parts to the LeoStick to build
projects or add more I/O connectors. Fits on the top of the
LeoStick and provides you a
free matrix of plated-through
holes for your own use.
• 64 general-purpose plated
holes for your parts
• All Arduino I/O headers
brought up for your use
NEW
• Includes male header pins
$ 95
• Gold-plated surface
XC-4268
7
Light Sensor Module for Arduino
This silicon light sensor outputs a voltage
proportional to incoming light. Perfect
for measuring light levels both
indoors and out, security sensing
and human feedback like waving
a hand over the sensor.
• +/-60° field of view
• Supply voltage: 3.0 to 5.5VDC
XC-4228
995
$
OLED Display Module for Arduino
High resolution, full colour OLED
display module! Perfect for
graphics, gauges, graphs, even
make your own video game
or interactive display.
• 16,384 full colour RGB pixels
in a 128 x 128 format
• Active display area 28.8 x
26.8 mm, (1.5 inch diagonal)
XC-4270
NEW
Mega Prototyping
Shield for Arduino
EtherTen, ArduinoCompatible with Ethernet
Fits the EtherMega (XC-4256) and Arduino
compatible "Mega" size boards so you can fit your
own parts for projects. Includes header pin sets.
This Arduino-compatible development board
includes onboard Ethernet, a USBserial converter, a microSD card
slot for storing gigabytes of
web server content or data,
and even Power-overEthernet support.
• ATmega328P
MCU running at 16MHz
• 10/100base-T Ethernet built in
$
95
• Used as a web server, remote
monitoring and control, home automation projects
• 14 digital I/O lines (6 with PWM support)
• 8 analog inputs
See website for full range of
XC-4216
69
Arduino compatible products.
IR Temperature Sensor
Module for Arduino
Connect this to your board and point it at a surface
or heat source to remotely measure its
temperature. This is our special version of the
industrial infrared remote thermometer units
with an onboard power supply,
communication support and a
software library and
examples supplied.
• 3.3 to 5V operation
• -33 to +220°C
measurement range,
1 second response time
XC-4260
1795
Real-Time Clock
Module for Arduino
Perfect for clock projects, dataloggers or anything
that needs to know the date and time. Keeps
accurate time for years using a
tiny coin-cell, and is very simple
to connect to your Arduino
project. A driver library allows
your program to easily set or
read the time and date.
NEW
• Battery included
XC-4272
2995
$
Arduino Compatible Relay Drivers
Drive up to 4 relays using logic-level outputs from
an Arduino or other microcontroller. Isolates your
microcontroller from the relay coils using FETs.
3495
$
• Size: 36(W) x 23(H) x
12(D)mm
XC-4278
NEW
1395
$
8 Channel Shield
Directly drive DC motors using your Arduino
compatible board and this shield, which provides
PWM (Pulse-Width Modulation) motor output on 2 Hbridge channels to let your board control the speed,
direction and power of two motors independently.
Perfect for robotics and motor control projects.
• Drives up to 2A per
motor channel
• All outputs are diode and
back-EMF protected
XC-4264
Drive up to 8 relays from an Arduino using just 2 I/O pins. It
communicates with your board using I2C, so you can even
stack several shields together to
drive 16, 24, or more outputs!
• Size: 52(W) x 66(H) x 12(D)mm
XC-4276
Both feature:
• Plugs straight into your Arduino-compatible board
• Individual LED status display on every
NEW
output channel
• LED status displays for external power
$
95
(and host power for XC-4276)
• Drive relay coils of 5VDC to 24VDC
(with external power supply)
• Works with a wide range of relays like SY-4052
34
NEW
2995
$
4995
NEW
$
4 Channel Module
NEW
H-Bridge Motor Driver
Shield for Arduino
$
• Over 300 general-purpose plated
holes for your parts
• Handy 5V and GND rails
• All Arduino I/O header pins
branched out for your use
• Gold-plated surface
• Reset button
XC-4257
YOUR LOCAL JAYCAR STORE - Free Call Orders: 1800 022 888
• AUSTRALIAN CAPITAL TERRITORY
Belconnen
Fyshwick
Ph (02) 6253 5700
Ph (02) 6239 1801
• NEW SOUTH WALES
Albury
Alexandria
Bankstown
Blacktown
Bondi Junction
Brookvale
Campbelltown
Castle Hill
Coffs Harbour
Croydon
Erina
Gore Hill
Hornsby
Liverpool
Maitland
Ph (02) 6021 6788
Ph (02) 9699 4699
Ph (02) 9709 2822
Ph (02) 9678 9669
Ph (02) 9369 3899
Ph (02) 9905 4130
Ph (02) 4620 7155
Ph (02) 9634 4470
Ph (02) 6651 5238
Ph (02) 9799 0402
Ph (02) 4365 3433
Ph (02) 9439 4799
Ph (02) 9476 6221
Ph (02) 9821 3100
Ph (02) 4934 4911
Newcastle
Penrith
Port Macquarie
Rydalmere
Sydney City
Taren Point
Tweed Heads
Wagga Wagga
Wollongong
Ph (02) 4965 3799
Ph (02) 4721 8337
Ph (02) 6581 4476
Ph (02) 8832 3120
Ph (02) 9267 1614
Ph (02) 9531 7033
Ph (07) 5524 6566
Ph (02) 6931 9333
Ph (02) 4226 7089
• NORTHERN TERRITORY
Darwin
Ph (08) 8948 4043
• QUEENSLAND
Aspley
Caboolture
Cairns
Capalaba
Ipswich
Labrador
Ph (07) 3863 0099
Ph (07) 5432 3152
Ph (07) 4041 6747
Ph (07) 3245 2014
Ph (07) 3282 5800
Ph (07) 5537 4295
Mackay
Maroochydore
Mermaid Beach
Nth Rockhampton
Townsville
Underwood
Woolloongabba
Ph (07) 4953 0611
Ph (07) 5479 3511
Ph (07) 5526 6722
Ph (07) 4926 4155
Ph (07) 4772 5022
Ph (07) 3841 4888
Ph (07) 3393 0777
• SOUTH AUSTRALIA
Adelaide
Clovelly Park
Gepps Cross
Reynella
Ph (08) 8231 7355
Ph (08) 8276 6901
Ph (08) 8262 3200
Ph (08) 8387 3847
• TASMANIA
Hobart
Launceston
Ph (03) 6272 9955
Ph (03) 6334 2777
• VICTORIA
Cheltenham
Coburg
Arrival
datesofofnew
new
products
in flyer
this were
flyer confirmed
were confirmed at the
Arrival dates
products
in this
HEAD OFFICE
at the of
time
of print.
Occasionally
thesedates
dateschange
change unexpectedly.
time
print.
Occasionally
these
320 Victoria Road, Rydalmere NSW 2116
unexpectedly.
Pleaselocal
ringstore
your local
store to
check
stock Prices valid from
Please
ring your
to check
stock
details.
Ph: (02) 8832 3100 Fax: (02) 8832 3169
Prices valid
from 24th
April
to 23rd
2012.are based on original RRP
details.
24th February
to 23rd
March
2012.
All May
savings
Ph (03) 9585 5011
Ph (03) 9384 1811
ONLINE ORDERS
Website: www.jaycar.com.au
Email: techstore<at>jaycar.com.au
Frankston
Geelong
Hallam
Kew East
Melbourne
Ringwood
Shepparton
Springvale
Sunshine
Thomastown
Werribee
Ph (03) 9781 4100
Ph (03) 5221 5800
Ph (03) 9796 4577
Ph (03) 9859 6188
Ph (03) 9663 2030
Ph (03) 9870 9053
Ph (03) 5822 4037
Ph (03) 9547 1022
Ph (03) 9310 8066
Ph (03) 9465 3333
Ph (03) 9741 8951
• WESTERN AUSTRALIA
Joondalup
Maddington
Mandurah
Midland
Northbridge
Rockingham
Ph (08) 9301 0916
Ph (08) 9493 4300
Ph (08) 9586 3827
Ph (08) 9250 8200
Ph (08) 9328 8252
Ph (08) 9592 8000
SERVICEMAN'S LOG
The dodgy home-made stereo amplifier
The old adage that a little knowledge is a
dangerous thing certainly applies when it
comes to mains-powered equipment. This
home-made amplifier was so dangerous and
the standard of assembly so dodgy that I
really shouldn’t have taken it on.
E
VERY NOW AND THEN, something lands on the workbench
that causes us to shake our heads in
wonder. This has happened to me
quite a few times over the years and it
never ceases to amaze me what some
people do. Sometimes they are simply
rank amateurs blundering about with
something they don’t understand and
shouldn’t touch. At other times, it’s
so-called professionals who just do
shoddy repair work.
Good workmanship is especially
critical with mains-powered equipment. Aside from any performance or
usability issues arising from shoddy
work, there is the very real danger of
killing someone, so it pays to know
what you are doing when dealing with
such equipment.
One memorable example was a
home-made stereo amplifier that an
acquaintance asked me to have a look
at. He’d been given the system as part
payment for some deal or other and
he’d used it for a few months until one
day the left channel stopped working.
I asked him to bring the whole kit
and caboodle in because I wanted
to test the speakers and the leads as
well. After all, many a serviceman has
jumped into an amplifier repair only
to find that the cause was literally
outside the box. In my case, I learned
a long time ago to start with the easy
stuff and work my way up from there.
The speakers were surprisingly
well-made and it looked like they
had been made from a professional kit
(either that or the woodworking skills
of the builder far outshone his or her
electronics abilities). I tested them for
continuity before removing the backs
to check that the crossovers and driver
units all looked and measured as they
should.
There were no fuses inside the
speakers, so that immediately ruled
out one possible source of trouble.
The crossovers were decent units
and appeared to be professionally assembled. However, I didn’t recognise
the speakers and there were no brand
names to give me a clue.
That said, they looked like a nice
piece of kit. The bass drivers were
beefy 150mm units and the tweeters
were expensive ribbon models which
looked very similar (but not the same
as) a couple I purchased from Jaycar a
Servicing Stories Wanted
Do you have any good servicing stories that you would like to share in The Serviceman column in SILICON CHIP? If so, why not send those stories in to us? In doesn’t
matter what the story is about as long as it’s in some way related to the electronics
or electrical industries, to computers or even to car electronics.
We pay for all contributions published but please note that your material must
be original. Send your contribution by email to: editor<at>siliconchip.com.au
Please be sure to include your full name and address details.
siliconchip.com.au
Dave Thompson*
Items Covered This Month
•
The dodgy, dangerous homemade stereo amplifier
• Intermittent MIDI keyboard
• Water pressure pump controller
• The freezer that really froze
*Dave Thompson, runs PC Anytime
in Christchurch, NZ.
while back. In short, it all looked good
so I buttoned everything back up and
connected the speakers to my workshop test amplifier using the supplied
cables and gave them good thrashing
. . . err, I mean workout.
They performed flawlessly and so,
having ruled out the speakers and
cables, I shifted my focus to the amplifier. I connected the speakers to it
and this indeed confirmed that there
was no sound from the left channel. In
fact, there wasn’t even a power “pop”
when the switch was turned on or off.
There was, however, a considerable
thump from the right channel when
power was applied. And just to prove
a point, I then swapped the speakers
over but the fault remained in the left
channel. There was also obviously no
switch-on muting or speaker protection, which was hardly surprising
given that this was an older home-built
amplifier and as I discovered, was
pretty much a bare-bones deal.
Everything was mounted inside a
standard rack-mount case (the type
with handles on each side at the front).
I removed the four case screws holding
the cover, lifted it clear and was immediately struck by how amateurish
(read “extremely dodgy”) the wiring
looked. For starters, the power cable
passed loosely through a badly-worn
rubber grommet and (unbelievably)
had a knot tied in it for strain relief!
In fact, I could see metal through the
grommet where the cable had rubbed
against it and the chassis over time. If
the amplifier hadn’t failed at this point
in time (due to some other fault), it
would almost certainly have made a
May 2012 61
Serr v ice
Se
ceman’s
man’s Log – continued
nice “bang” at some point in the future.
Unfortunately, this method of securing (if you could call it that) the mains
cable didn’t particularly surprise me as
I have seen similar installations in the
past. It may have been the way things
were done 60 or 70 years ago but it’s
no longer acceptable or legal. So, if
the amplifier was salvageable (and I
very much doubted it at that point),
this would have to be put right before
I returned it.
The next problem was the interconnecting wiring between the boards and
the various input and output sockets.
Basically, the inside of the case looked
like a bombed-out a pasta factory!
There was wiring everywhere, nothing
was tied back and the leads floated
everywhere, willy-nilly.
This is not only a problem in an
environment that can get hot during
use (eg, near the heatsinks) but can
also introduce all manner of electrical
noise and hum, something not typically welcome in a hifi set-up. So this
too would need tidying up if the job
were to go ahead.
However, the most obvious issue I
could see was the quality of the soldering; it looked like whoever built it
62 Silicon Chip
had used a plumber’s soldering iron,
bar solder and a couple of tins of flux
paste. It was an absolute mess and I
wondered how the amplifier had ever
worked at all.
And so I called my friend and told
him that I’d pretty much need to
rebuild the thing, aside from finding
and replacing any faulty components.
Of course, I fully expected him to
say chuck it in the bin and would
have been happy if had he done so.
However, much to my surprise, he
instead gave me a ceiling of “about a
hundred and fifty bucks” and said that
if I thought it was going to go over that
figure, he would reassess things.
I said that I’d see what I could do,
although I wasn’t at all enthusiastic
about taking the job on.
The power amplifier modules appeared to be based on the now-ancient
but very popular ETI-480 50W design.
In fact, I used a few of these in guitar
amplifiers many years ago and they
did a pretty good job. Having said
that, you wouldn’t bother building one
these days as the SILICON CHIP SC480
amplifier module gives much better
performance for about the same cost
and is much more suitable for hifi use.
The faulty left-channel amplifier in
the unit on my workbench must have
been working at some point so the first
thing I did was check the on-board
fuses. These checked out OK so I took
a closer look at those dodgy solder
joints. Most of them looked awfully
old and dry, a state of affairs likely
compounded by the use of inappropriate soldering tools when the modules
were originally assembled.
Many an intermittent problem has
been fixed by resoldering so I decided
to go over both boards. This involved
first removing as much of the old solder as I possibly could using a vacuum
desoldering tool. I then re-soldered
every connection on both boards.
There really aren’t that many solder
joints on these modules, so the whole
process only took about 30 minutes for
the two boards.
Once done, I tested the modules insitu and found they both now worked,
so my guess was right – a dodgy solder
joint was the culprit after all.
While I was at it, I also reworked
all the soldered joints for the wiring
connections between the boards and
the sockets. In the end, I replaced almost half this wiring in order to route
it around the edges of the boards and
used small cable ties to loom everything together. I then removed the
power cable and as I did so, I discovered that the Earth wire had simply
been stripped, twisted and clamped
under the head of one of the power
transformer mounting nuts!
Worse still, the earth wire’s insulation had been so roughly removed
that half the individual strands were
missing while others were scored
to the point of breaking away when
touched. Further inspection revealed
that the Active and Neutral wires were
similarly damaged so I cut everything
back and since the switch had spade
connectors, used a crimping tool to reterminate the Active and Neutral wires
with decent matching connectors.
Next, I replaced the worn-out grommet and drilled two new holes through
the bottom of the case – one to accept
the mounting bolt for an Adele-type
P-clip to clamp the cable and the other
for a dedicated earth point. After passing the mains cable through the new
grommet and the P-clip, I tightened
everything and connected all the terminals, tying the leads up with cable
ties where required.
The Earth wire was terminated in an
siliconchip.com.au
Faulty pump pressure controller
This next story comes from K. G. of
Yattalunga, SA and concerns repairs
to a pressure control unit in a water
pump. Here’s what happened . . .
Roger, a neighbour of mine, has
a pump to cycle rain water into his
house, as is quite common these
days. However, after a recent prolonged period of operation, the pump
wouldn’t restart and he quickly diagnosed a faulty run capacitor. The
fault was obvious, as the capacitor
had oozed some of its innards to
the outside and when I checked it
with a capacitance tester it showed
open circuit.
Roger fitted a new capacitor and
the pump was restored to normal operation but in reassembling the unit,
one of the threaded water connections was damaged. This was part of
the housing of the pump control unit
which is branded “Presscontrol”.
These controllers are widely used
for this type of service across several
brands of pump and it turns out that
my own pump, although a different
brand to my neighbour’s, uses the
same Presscontrol control unit. In
operation, the control unit senses
the pressure drop in the pipe when
a tap is turned on and switches on
the pump. Then, when the flow rate
drops below a threshold, the pump
turns off.
Because of the damage, Roger had
to buy a new complete unit as the
housing isn’t available as a separate
unit. A phone call to a local supplier
resulted in a trip to the establishment but contrary to the telephoned
information, they were in fact out of
stock. However the helpful sales assistant said they had a secondhand
unit which had a faulty electronics
board and as they had made the error
and could not supply a new unit from
eyelet lug and bolted to the chassis. I
used a couple of star washers, one on
either side of the eyelet, and tightened
it down with a “Nylock” style locking
nut to ensure a sound physical and
electrical connection. There is no way
this can now come loose, so a decent
earth connection is guaranteed.
Once all was tidy inside the case, the
siliconchip.com.au
stock, he could have the secondhand
one free of charge. However, it would
mean swapping his good board over
into the secondhand housing.
On the surface, this seemed to be
a simple task but like many such
tasks it turned out to be not quite
that simple.
The board is held in the housing on three plastic pillars which
protrude through the board and a
plastic cover is installed over the
board. There are spring clips which
go over the pillars and these press
down onto the cover and hold it and
the board in place against shoulders
on the pillars. These clips are round
and have “fingers” going in towards
the centre of the central hole. When
the clip is fitted, the fingers dig into
the plastic pillar, making it very difficult to remove.
Eventually, we managed to get the
clips off without damaging the pillars
so much that the board couldn’t be
held in place. Roger then completed
the re-assembly of the pump and
returned it successfully to service.
This left us with the broken housing from Roger’s original unit and
a faulty PCB from the secondhand
unit. Given the cost of a new spare
part, it was worth spending some
time to see if the board could be
repaired for use as a spare. And so I
took it home for further investigation.
When I got it onto the workbench
and had a good look, it turned out
to be not that complicated. A 4093
CMOS quad NAND Schmitt trigger
was the most complex part on the
board and I could recognise the power supply components. This power
supply runs directly from the mains
and uses a capacitor in series with
the Active and a rectifier diode, filter
capacitor and zener diode to produce
cover was replaced and the unit given
a good listening test. Its performance
was OK but its certainly not up to
today’s standards.
So why did I go to so much bother on
a repair that was simply uneconomic?
The answer is that I spent a lot longer
on this than I should have because
the client was someone I knew. That
a DC rail to run the CMOS device.
In addition, a relay with mains
rated contacts to switch power to
the motor was included and the
labelling on the relay indicated it
had a 24V coil.
I connected the PCB to the mains
using a special cord I keep for this
sort of job. This cord has a wellinsulated in-line fuse in case anything goes awry during testing and
of course, great care must be taken
to avoid contacting the high-voltage
Active conductors on the PCB.
On powering up the board, I
measured 8V DC on the 4093 which
I judged to be correct. I then checked
the voltage on the supply side of the
relay coil. This measured just 13V –
not enough to pull in the relay.
My next step was to check the ESR
of the main electrolytic filter capacitor (47µF 50V) and this turned out
to be well inside the recommended
limit. Similarly, the 100Ω resistor
in series with the mains Active
also tested OK. Then came the big
capacitor and I was immediately
suspicious. It was labelled 1µF 400V
DC but had none of the usual mainsrated labelling on it.
I tested its capacitance and it
measured just 235nF which is about
one quarter of its labelled value.
As a result, I replaced it with a 1µF
mains-rated unit which fortunately
fitted on the PCB, despite being quite
a bit larger than the original.
On powering up the board, I was
immediately greeted with the sound
of the relay operating and both the
green and orange LEDs lighting
whereas before, only the green one
lit. I then checked the voltage on the
supply pin of the relay coil and found
it to be exactly 24V.
So the exercise was well worthwhile. Roger and I now have a spare
controller board should either of our
controllers fail.
said, far better quality units are now
available for very reasonable money
and dinosaurs like this are really not
worth fixing.
It’s clear that whoever built this amplifier wasn’t very competent and nor
did they have the good sense to have
someone who was check their work.
I shudder to think how many other
May 2012 63
Serr v ice
Se
ceman’s
man’s Log – continued
home-made mains-powered devices
are out there in this kind of condition.
Midi keyboard
One of my “other” hobbies is music
and this is partly why I became interested in computers in the early 80s.
Australia led the way in the late 1970s
and early 1980s with music computers
like the much-lauded (and expensive)
Fairlight, used by such musical luminaries as Peter Gabriel, Duran Duran
and Jan Hammer (who produced the
original Miami Vice theme song on a
Fairlight CMI).
By today’s standards, the original
Fairlight range looks rather quaint but
we should not underestimate the impact that these early “music-specific”
computers had on the recording scene.
Music and electronics in general
have gone hand in hand since the
early Moog days and it’s no surprise
that the main remaining use of valve
technology is in audio (particularly in
guitar amplifiers).
Since I am also known as a musician, I sometimes get music-related
hardware in for repair. My last such job
involved a so-called MIDI keyboard.
MIDI stands for Musical Instrument
Digital Interface and is a standard
that’s used for controlling and connecting compatible electronic instruments
to each other and to computers (and
vice versa).
In MIDI devices, connections are
commonly made via old-style DIN-5
64 Silicon Chip
plugs, similar to those used by olderstyle AT computer keyboards.
This particular MIDI keyboard was
a reasonably common piece of musical hardware. And although it looks
like an electric piano, or one of those
portable keyboards with all the drum
sounds and rhythms programmed in,
it is not an instrument in itself. The
keyboard produces no sound (other
than a dull plastic “thunk” if you really
hammer the keys down) and is used
purely to control other MIDI-capable
devices.
Basically, it can be thought of as a
computer keyboard, except instead
of having letters and numbers, it uses
a piano-style set of black and white
keys. When this keyboard controls
a computerised piano, it becomes a
piano keyboard. However, it can also
be used to play computer-generated
guitars, drums, bass and any one of a
gazillion other sampled or waveform
generated/synthesised sounds.
In this case, the keyboard was
an Evolution MK-149 and it was in
for repair because its operation was
intermittent. It would work happily
for awhile and then suddenly stop
working, all at completely random
intervals, and this was driving the
owner mad.
Now I’m not a keyboard technician
but being a male electronics technician
gave me all the qualifications I needed
to at least pull it apart and have a look
around. Two dozen or so PK-style
screws held the top of the keyboard to
the bottom section and after removing
the screws, the case easily split apart.
The main keyboard section was basically a set of 48 plastic “keys” which
actuated PCB-mounted switches, with
everything mounted onto a single PCB
assembly. Since this all functioned
perfectly when the keyboard was
working, I figured that the problem
wasn’t likely to be there.
Fortunately, everything else, including the related circuitry, was
easy to see. I began by checking the
connections to and from each PCB
to the output sockets and all looked
solid. However, when checking the
MIDI In/Out DIN sockets, I noticed a
small spring which was completely
out of place. It was caught on one of
the soldered pins on the output socket
and to the naked eye this looked to be
bridging or shorting the adjacent pin.
When I looked at it under a magnifying glass, I could see a very slight
gap between the spring and the pin
and using my continuity tester, I confirmed that there was no connection.
However, with just the right amount
of vibration, the spring would flex
and short out the pins. This certainly
wasn’t right and bridging those pins
would be enough to cause the MIDI
side of things to stop working. I didn’t
know what each pin did data-wise but
it certainly wouldn’t do to have one
touching the other.
It was easy enough to remove the
spring but I then had a problem tracking down its original location. All
the keyboard springs were in place
and besides, they were much heavier
than this small spring. What’s more,
this particular keyboard model has no
springs for centring the modulation
and pitch wheels, so it couldn’t have
come from there.
After spending a good 10 minutes
going over everything, I still couldn’t
see anywhere the spring could have
come from. In the end, I concluded
that it was a “stray” and had probably
been there since the keyboard was assembled at the factory and was only
now causing problems.
The client reports the keyboard now
works as expected so problem solved!
The freezer that really froze
A. F. of Chinderah, NSW recently
did battle with a freezer that was
working overtime. Here’s his story . . .
I enjoy my repair work when my
siliconchip.com.au
customers become friends. And this was the case with
Enid who had an split-system air-conditioning unit that
had twice been repaired but had stopped working for
the third time. She had then decided to forgo having
it repaired a third time, as she just couldn’t afford the
repeated call-out costs.
When I heard about her plight, I agreed to look at the
unit without charge. I soon found that the circuit breaker
was open and I reset this without any further problems
showing up, even nine months later.
Because I hadn’t found a cause, I kept in regular touch
with Enid in case the unit went faulty again. In any case,
this is a good practice, as such calls sometimes result in
new customers due to word-of mouth recommendation.
One day, when I called, Enid invited me in for a cup
of tea, along with an offer for some fruit and ice-cream.
I was then asked if I would scoop out the ice-cream, as
Enid did not have the strength to do this.
When I tackled the job, I was not surprised that Enid
did not want the task – the ice-cream was as hard as a
block of solid lead. In fact, I only managed to scoop out
a small slither and it was obvious that her freezer was
running much too cold.
I listened to the sound of the deep freeze motor and
all the time that I was there, the motor didn’t stop. I
mentioned this to Enid and suggested that I install a
thermometer in the deep freeze to check its temperature.
The next time I was in the neighbourhood, I called in
and placed the remote sensor bulb of an indoor/outdoor
thermometer into the freezer compartment. It eventually
indicated -39°C so it was no wonder the ice-cream was
hard – freezer temperatures should be around -18°C.
While I was there, I also installed a mains power meter
but not once did the thermostat turn off the power over
a period of about one hour. The power consumption at
-39°C was about 60W and it was running continuously.
To speed up the repair, I left the food inside the freezer
and just tilted it until it was leaning against a solid bench
top. This allowed me to access the thermostat which was
mounted underneath, near the compressor. I then carefully
marked all the leads that were connected to the thermostat
and drew a diagram showing what went where.
When I subsequently removed the leads to the thermostat, I found that the contacts were still closed at the
-39°C temperature, which was obviously wrong. I then
had to destroy the thermostat to get to the contacts and
just as I expected, they were welded together.
I phoned our local appliance spare parts seller but
they wanted $85 for a new thermostat. A search on eBay
soon located a similar part for $33 including postage, so
I ordered one and it arrived four days later.
It now keeps the temperature at around -18°C. What’s
more, the ice-cream is now soft and easily scooped, so I
have made myself redundant at ice-cream serving.
While I had all my equipment connected to the freezer,
I watched the mains power meter for more information.
I did not tell Enid how much electricity her appliance
had been wasting and in any case, the cost is unknown
as I have no way of knowing how long the thermostat
had been faulty.
I wonder how many other freezers have welded thermostat contacts and motors that run 24 hours a day? Most
SC
people would not have the knowledge to check.
siliconchip.com.au
5
GOOD
REASONS
Switchmode
to use
– the repair specialists
to industry and defence
one
two
three
four
five
specialised
service
Benefit from our purpose-built facilities,
efficient and effective service. Since 1984
we have specialised solely in the repair of
all types of power supplies up to 50KVA.
turn-around
time
We provide three levels of service:
Standard (10 days) Standard Plus (4 days)
Emergency (24 hours)
access
access to
to techs
techs
and
and engineers
engineers
Talk directly to our highly skilled
Technicians and Engineers for immediate
technical and personal assistance.
quality
assurance
Accredited to ISO 9001 with SAI Global
and ISO 17025 with NATA. Documented,
externally audited management systems,
deliver a repeatable, reliable service.
convenience and
certainty
We provide fixed price quotes after assessment of goods and cost-effective maintenance,
tailored to meet individual customers needs.
TAKE ADVANTAGE OF OUR RESOURCES
SWITCHMODE POWER SUPPLIES Pty Ltd ABN 54 003 958 030
Unit 1/37 Leighton Place Hornsby NSW 2077
(PO Box 606 Hornsby NSW 1630)
Tel: 61 2 9476 0300 Fax: 61 2 9476 0479
Email: service<at>switchmode.com.au Website: www.switchmode.com.au
May 2012 65
PRODUCT SHOWCASE
PowerSTAR MPPT
Solar Regulator
from Roc-Solid
It’s one thing to put an array
of solar panels on your roof. It’s
another thing altogether to obtain
the maximum power from those
solar panels.
Far too many people wonder
why their “1kW” solar cell installation delivers perhaps only half
to two thirds of its rated output,
this with bright sunshine and
clean cells.
For many people, that’s because
their installer saved some money
and put in a cheap regulator. Not
only will they cheat you out of
power, in some cases they can allow the solar panel to overcharge
and therefore damage the very
expensive battery bank.
Roc-Solid have released an
Australian-designed MPPT solar regulator which not only
boasts that it will never allow your batteries to be cooked,
it also utilises Maximum Power Point Tracking which ensures that the maximum possible power is extracted from
your solar panels.
There are several regulators in the Roc-Solid range.
The PowerSTAR PS-2024-D MPPT Regulator is fully
user-configurable and has the ability to save configuration
settings on a PC. More than that, it is also a data logger,
storing performance data for 400 days. This data includes
amp-hours in, amp-hours out, maximum and minimum
battery voltages, time to float and days to next equalisation.
Output to PC is via a plug-in USB interface (as shown
at right of unit). Its flexible programming modes (either
user-defined or via the PowerSTAR settings manager) are
assisted by a large, easy-to-read on-board LCD screen
Overall the PS-2024-D is just 195 x 92 x 65m (including
mounting provision).
While the PowerSTAR PS-2024-D is intended for the
installation market, it is simple enough to allow the average user to retro-fit to an existing system.
A quick-start guide is included with the unit but a
comprehensive user manual (in PDF) can be downloaded
from the company website, as can the PowerSTAR Settings
Manager (Windows) software.
PowerSTAR regulators are sold through a network of
distributors throughout Australia. Your closest ROCSOLID distributor
Contact:
can be found on the
ROC-SOLID Technologies
‘Where to buy’ page
37 Belford Ave, Devon Park, SA 5008.
on the Roc-Solid
Website: www.roc-solidsolar.com.au
website.
New WES 2012 trade/wholesale catalog
Now in its 28th year, major trade/
wholesale electronics parts supplier
WES Components have released
their new 2012 Trade Catalog.
WES is constantly adding new
products to their range with the
result that the new catalog, a 1272page monster, has more than 150
new pages.
There are several notable differences between the new catalog and
previous versions, one being that
66 Silicon Chip
it is now in full colour and with an
improved page format.
However, the main difference is
that all prices exclude GST, as most
wholesalers quote GST-exempt prices.
It can be obtained either in hard
copy direct from Contact:
WES, or WES Components
o n - l i n e 138 Liverpool Rd, Ashfield NSW 2131
( P D F ) Tel: (02) 9797 9866 Fax: (02) 9716 6015
Website: www.wes.net.au
format.
siliconchip.com.au
Verbatim LEDs – 80% Energy Saving
Rohde & Schwarz and
HAMEG Instruments under one banner
Almost anywhere in the world, test and measurement
products from HAMEG Instruments are now directly available from Rohde & Schwarz. This is why these products
will receive a new logo that contains the names of both
companies – a move that is intended to further increase
the brand awareness and growth of HAMEG Instruments
in both the domestic and international marketplace.
HAMEG Instruments has been a Rohde & Schwarz subsidiary since 2005. After several years of separate branding,
the new dual logo containing both company names is an
important step toward integrating HAMEG products more
tightly into the Rohde & Schwarz portfolio.
Since the end of last year, HAMEG products have been
an integral part of
the portfolio of- Contact:
fered by the Rohde Rohde & Schwarz (Aust) Pty Ltd
& Schwarz sales Unit 2, 75 Epping Rd, North Ryde NSW 2113
Tel: (02) 8874 5111 Fax: (02) 8874 5199
organisation.
Website: www.rohde-schwarz.com.au
Verbatim are proving to be a major
player in LED lighting with an everincreasing LED product offering
for residential and commercial
applications.
The latest products released
include a 600lm AR111, G53 base
low voltage spot as a replacement
for 50W halogen spots, a 600lm
PAR30 high voltage flood light as a
replacement for 75Whalogen lamps
and a 800lm PAR38 high voltage
flood light suitable for 100W halogen
or incandescent PAR lamps.
Verbatim’s LED lamps have a lifetime
between 25,000 and 30,000 hrs and offer a significant energy savings of 80%
when compared to halogen lamps.
Contact:
Verbatim Australia
6 Weir St, Glen Iris, Vic 3146
Tel: (03) 9823 0999 Fax: (03) 9824 7011
Website: www.verbatimlighting.com.au
Roland DG launches online store
Roland DG Australia, a leading provider
of wide format print solutions, 3D milling and engraving technologies, today
announced the launch of the Roland DG
Australia Online Store.
Windspeed and direction monitoring
The KTA-250 is an
anemometer monitoring and alarming
card which allows
the wind speed and
direction to be measured with a Davis
Instruments DS7911
Anemometer without the need for the
entire weather station.
Monitoring of the
wind speed and direction can be done via the analog retransmission channels with 0-5V, 1-5V, 0-20mA or 4-20mA
outputs. As well as monitoring the speed and direction,
two alarm relays can be programmed to activate at a certain
speed or wind direction range, or combination of the two.
Wind speed and direction can also be monitored using the
Modbus protocol with either USB or 2-wire RS-485 connection. All settings are set using the Modbus protocol and
the provided software.
An optional LCD
screen and 64K Da- Contact:
talogging memory Ocean Controls
PO Box 2191, Seaford BC, Vic, 3198
are available.
Prices start from Tel: (03) 9782 5882 Fax: (03) 9782 5517
Website: www.oceancontrols.com.au
$179.00+GST.
siliconchip.com.au
Contact:
To support the recently launched iCreate
website for the Roland iModela, the Roland DG
Australia Store offers secure online purchasing
for the iModela, iModela accessories, and also
a range of Pantone and SAi software products.
Roland DG Australia
Allambie Grove Bsns Park, 14/25 Frenchs
Forest Rd, Frenchs Forest NSW 2086.
Tel: (02) 9975 0000 Fax: (02) 9975 0001
Website: www.rolanddg.com.au
Tektronix “Thunderbolt”
test solution
Tektronix, Inc. has a
comprehensive test solution for Thunderbolt,
a new, high-speed, multi-protocol I/O technology designed to provide
headroom for next generation display and I/O requirements.
The new solution includes a 20GHz DSA70000 Series
Oscilloscope, 12.5Gb/s BSA Series BERTScope, and a
DSA8300 Series Sampling Oscilloscope. More information
can be found at www.tek.com/webinar/10gpbs-thunderbolt-conformance-testing
The Tektronix solution serves the comprehensive needs
of Thunderbolt physical layer testing and spec conformance validation. Thunderbolt’s four channel 10.3Gbps I/O
architecture is the most significant advancement in PC I/O
design ever introduced into the consumer level electronics
industry and Tektronix is dedicated to delivery of a broad
portfolio of tools to facilitate the successful deployment of
this new technology working in concert with Intel.
The Tektronix
physical layer Contact:
electrical valida- TekMark Australia Pty Ltd
tion solution for Suite 302, 18-20 Orion Rd, Lane Cove 2066
Thunderbolt is Tel: 1300 811 355 Fax: (02) 9418 8485
now available. SC Website: www.tekmarkgroup.com
May 2012 67
Note: this updated article for the Induction Motor Speed Controller incorporates
all the changes to the original version (including the modified PCB), as described
in the December 2012 and August 2013 issues. The software is also revised.
1.5kW Induction Motor
Speed Controller, Pt.2
Pt.2: by ANDREW
ANDREW LEVIDO
Last month, we described the features of the 1.5kW Induction
Motor Speed Controller and explained in detail how it works.
This month we describe its construction and testing and give
some guidelines for use.
WARNING: DANGEROUS VOLTAGES
This circuit is directly connected to the 230VAC mains. As such, most of the parts and wiring operate at mains
potential and there are also sections running at 325-350V DC. Contact with any part of these non-isolated
circuit sections could prove FATAL. Note also that the circuit can remain potentially lethal even after the
230VAC mains supply has been disconnected!
To ensure safety, this circuit MUST NOT be operated unless it is fully enclosed in a plastic case. Do not connect
this device to the mains with the lid of the case removed. DO NOT TOUCH any part of the circuit unless the
power cord is unplugged from the mains socket, the on-board neon indicator has extinguished and at least
three minutes have elapsed since power was removed (and the voltage across the 470μ
470μF 400V capacitors has
been checked with a multimeter – see text in Pt.1).
This is not a project for the inexperienced. Do not attempt to build it unless you understand what you are doing
and are experienced working with high-voltage circuits.
68 Silicon Chip
siliconchip.com.au
NYLON
CABLE TIES
1 0 nF∗
S UP
SLE PRESSIO
EVE
N
CON5
RITE
CON4
NYLON
CABLE TIES
V
220nF X2
NYLON SCREW
WITH TWO
LOCK NUTS
NYLON
CABLE TIES
BOX FRONT
PANEL
(INSIDE VIEW)
150k
NE-2
NEON
WARNING!
Neutral Earth Active
4004
4004
4004
4004
EARTH
CON3
(COVERED)
150k
47nF X2
BR1 GBJ3508 (UNDER)
DANGEROUS VOLTAGES
W
CON2
47nF X2
FUSE1 10A
NYLON CABLE CLAMP
620k
16k
IC1 STGIPS20K60 (UNDER)
FLT1
YF10T6
CABLE GLAND
(REAR VIEW)
620k
TH1 SL32 10015
MINI
MUFFIN
FAN
8.2k
1
ORIENTATE FAN SO THAT IT
BLOW S AIR INTO THE C ASE
– SEE TEXT
8.2k
8.2k
D1 D2 D3 D4
∗ PART VALUES MARKED
IN RED ON PCB HAVE
BEEN CHANGED
FROM ORIGINAL
VALUES – SEE TEXT
8.2k
0.015Ω
2W
SECURE FAN USING 4 x 20MM NYLON M3 SCREWS, NUTS & SHAKEPROOF WASHERS
CON6
ZD2
RAMP
SPEED
FER
100nF
100nF
VR1
VR2
10k 100nF 10k
U
1.5k
100Ω∗
100nF
OPTO2
470 µF
(GPO MAINS OUTLET
MOUNTED ON
OUTER SURFACE)
ZD1
HCPL2531
OPTO3
10 µF
BC337
Q1
100nF
100nF
100nF 100nF
IC3
HCPL2531
10µF
100nF
10µF
dsPIC33FJ64MC802
10 µF
OPTO1
4N35
+
HEATSHRINK
SLEEVING
A
A
Rev
Run
Fault
A
PP
Ext
O/S
Flt
TO TH2
CON7
470 µF
LM317T
470 µF
100nF
100nF
REG1
D5 D6 D7 D8
ISOLATION BARRIER
T2 6V+6V 5VA
(UNDER)
+
4004
+
10k
1M ∗
470Ω∗
0.5W
4004
10µF
5.1V
IC2 LM319
10Ω
4004
100nF
100nF
100Ω
4004
470 µF 400V
(UNDER)
ISOLATION BARRIER
100Ω
470 µF 400V
(UNDER)
47k
470Ω
100Ω
470 µF 400V
(UNDER)
+3.3V
100Ω
100Ω
T1 6V+6V 5VA
(UNDER)
Vin
100Ω
4.7k 5W
4.7k
100Ω
4004
4.7k 5W
GND
1.5k
15V
180Ω
D9
110Ω
S1 – 4
4.7k 5W
RUN
680Ω
10105122
100Ω
4.7k
REV
100nF
100Ω
+
EST
5
+
GND
ICSP
1.5kW Induction Motor Speed Controller
100Ω
1
+
+
WARNING: ALL PARTS IN YELLOW AREA ON PCB OPERATE AT LETHAL VOLTAGE & LETHAL VOLTAGES
REMAIN FOR SOME TIME AFTER POWER IS REMOVED – SEE TEXT
Fig.8: follow this diagram to build the unit. Note that transformers T1 & T2, the three 470μF 400V electrolytic capacitors,
bridge rectifier BR1 and IC1 (the IGBT module) are mounted on the underside of the board.
B
EFORE GOING any further, we
must again remind readers that
this project is intended only for experienced constructors. Most of the circuit
operates at 230VAC mains potential
and it has portions operating at 325350V DC. Furthermore, the circuit can
siliconchip.com.au
remain potentially lethal even after
the 230VAC mains has been removed.
Construction begins with assembly
of the PCB. Be sure to use the revised
PCB which is coded 10105122. Note
that several component values were
changed after this board was de-
signed, so the screened overlay on
early versions of this revised board
may show the old values. The parts
layout of Fig.8 is correct.
Be sure also to use a PIC micro that’s
programmed with the latest veresion
of the software; ie, 1010512B.hex.
May 2012 69
200
25
5
60
85
105
45
170
ALL NINE HOLES ARE TAPPED M3
70 Silicon Chip
Fig.9: this full-size diagram shows the drilling details for the heatsink. It should be copied, attached to the heatsink with sticky tape and
used as a drilling template. Use a small pilot drill (eg, 1mm) to start the holes, then drill each one to a depth of about 8mm using a 2.5mm
drill. The holes are then tapped to 3mm. Use plenty of light machine oil to lubricate both the drills and the tap during this procedure and
withdraw these parts frequently from the hole being worked on to clear any metal swarf (if this is not done, the aluminium swarf will bind
to the tool and spoil both the tool and the job). Note: the drilling diagram is also available in PDF format from the SILICON CHIP website.
5
5
75
65
ALL DIMENSIONS
IN MILLIMETRES
TH2 MOUNTING
POSITION
5
Note that some components are
mounted on the underside of the
board and there are five surfacemount components to contend with.
These surface-mount components are
all passive (four 10μF capacitors in
2012/0805 packages and one 0.015Ω
2W resistor in a 6432/2512 package)
and are easy to install using a conven-
tional soldering iron with a small tip.
Start by loading these SMT components, then move on to the rest of the
components in reverse height order.
Don’t install any of the parts that mount
underneath the board at this stage.
Note that the 4N35 opto-coupler is
mounted the opposite way to the two
HCPL2531s. The 4.7kΩ 5W resistors
must be mounted 2-3mm proud of the
PCB to ensure free airflow on all sides.
The input surge-limiting NTC thermistor TH1 should be mounted such
that there is about 15mm of bare lead
exposed above the surface of the board.
This serves two purposes: first, it
prevents the solder joints from overheating, since this component runs
quite hot at full load. And second, it
allows the thermistor to be bent down
parallel with the PCB so that it will fit
inside the IP65 case and not foul the
lid. This can be seen in the photograph
on page 74.
However, don’t bend the thermistor
down at this stage because you need
access to the screw hole for the bridge
rectifier, BR1. The bridge rectifier
must be secured to the heatsink and
soldered to the PCB, before the thermistor is bent over.
Next you can begin mounting the
parts on the bottom of the board. Leave
the IGBT driver and bridge rectifier
off for now. The polarity of the large
electrolytic capacitors must be correct
– a mistake here would be disastrous
(not to mention messy and dangerous).
Heatsink assembly
Drill and tap nine M3 holes in the
machined surface of the heatsink as
shown in Fig.9. Make sure the holes
are carefully de-burred so that the
heatsink surface is completely smooth.
Next, use the PCB as a template to
bend the leads of the bridge rectifier
upwards so that the leads fit and the
mounting hole is directly under the
corresponding hole in the PCB.
The next step is to mount thermistor TH2 on the heatsink with its
leads twisted and poking upwards
so that they can be later soldered
directly to CON7’s pads. Before fitting the heatsink thermistor, smear
a small amount of heatsink compound on the mounting lug and
then attach it to the heatsink with an
M3 x 6mm screw and lockwasher.
Orientate the lug so that the thermistor
wires run to the right – see Fig.10.
Now apply a thin smear of heatsink
compound on the mounting surfaces
of the IGBT driver (IC1) and bridge
rectifier (BR1). Insert them in their
appropriate places in the circuit board
(from the bottom) but don’t solder
them yet. You can stop them from
falling out when you turn the board
upright by making a small bend in a
couple of the leads.
siliconchip.com.au
This view shows the underside of the PCB. Note the
aluminium brackets attached to either side of the heatsink.
Case and wiring
Since much of the printed circuit
board is at lethal potential, it is essential that the speed controller be
siliconchip.com.au
M3 x 10mm
SCREW WITH
FLAT WASHER
IGBT BRIDGE
MODULE
DIODE
BRIDGE
M3 x 6mm SCREW
WITH STAR
LOCKWASHER
2 x M3 x 10mm
SCREWS WITH
FLAT WASHERS
5 x M3 x 16mm SCREWS
WITH STAR LOCKWASHERS
& 9 mm METAL SPACERS
PCB
THERMISTOR TH2
TO
CON7
THERMAL GREASE
(HEATSINK)
THERMAL GREASE
NOTE: DIAGRAM NOT TO SCALE
Mount the PCB assembly on the
heatsink using M3 x 16mm screws,
star lockwashers and 9mm spacers,
as shown in Fig.10. Once the board
is firmly screwed into place you can
screw down the IGBT and diode bridges using M3 x 10mm screws. These
screws are inserted through the holes
in the PCB but the flat washers have to
be manipulated into place under the
board using tweezers.
Alternatively, you could glue them
in place on the devices with a drop of
superglue before assembly. Tighten
the screws carefully, making sure both
devices are flat against the heatsink.
Once everything is in place, solder
the pins from the top, clipping off any
excess very carefully. Finally, twist
and feed the heatsink thermistor (TH2)
wires up through the CON7 pads with
a pair of tweezers and solder them on
the top of the PCB. It doesn’t matter
which lead goes to which pad. Keep
these leads short, so that they cannot
possibly short against high-voltage
circuitry if they come adrift.
That completes the assembly of the
controller module.
Fig.10: diode bridge BR1, the IGBT module (IC1) and the 10kΩ thermistor TH2
are mounted on the heatsink as shown here. The PCB is attached to the heatsink
at five points on 9mm untapped spacers while the leads from the heatsink
thermistor are fed up through the PCB’s CON7 pads and soldered.
mounted in a fully enclosed case.
Whatever case you choose, you must
take care that the mains wiring is
fully compliant with the relevant
standards. If the case is metal, it must
be securely earthed.
Note that the Speed Controller dissipates around 28W at idle and over
50W at full power. So we recommend
that you either use a vented case or
drill a series of holes on one side and
fit a fan on the other side. We’ll show
how to do this with the specified case.
Obviously, with vents, the IP65 case
is not waterproof or dustproof but
the unit will run much cooler (and
therefore more reliably) with airflow.
Note also that if a plastic case is used,
May 2012 71
Parts List: Induction Motor Speed Controller
1 double-sided PCB, code
10105122, 200.5 x 125mm
1 front panel label (147 x 102mm)
1 diecast heatsink, 200 x 75 x
48mm (Jaycar HH8546, Altronics
H0536)
1 IP65 ABS case, 250 x 200 x
130mm (Altronics H0364A)
1 IP68 cable gland to suit 4-8mm
cable (Jaycar HP0724, Altronics
H4313)
1 surface-mounting single mains (3pin) socket
1 10A mains lead
1 ferrite suppression bead, 28mm
long, 15mm OD, 7mm ID (Jaycar
LF1260, Altronics L4802A)
1 60mm 12V DC fan (Jaycar
YX2505)
1 60mm fan grille (Jaycar YX2550)
2 6+6V 5VA PCB-mount trans
formers (Altronics M7052A)
2 10kΩ mini horizontal trimpots
(VR1, VR2)
2 PCB-mount 3AG fuse clips (F1)
1 10A 3AG fast-blow fuse (F1)
1 fuse cover for F1
1 SL32 10015 NTC thermistor (TH1)
(Element14 1653459)
1 10kΩ NTC thermistor with mounting lug (TH2) (Altronics R4112)
1 YF10T6 mains filter (FLT1) (Jaycar
MS4000)
1 NE-2 Neon lamp (Jaycar SL2690,
Altronics S4010)
2 3-way PCB-mount terminal
barriers, 8.25mm pitch (CON2,
CON3) (Altronics P2102)
3 3-way terminal blocks, 5/5.08mm
pitch (CON4-CON6)
1 4-way DIP switch (LK1-LK4)
1 5-way pin header, 2.54mm pitch
(ICSP)
1 2-way pin header, 2.54mm pitch
(CON7)
1 Nylon* P-clamp to suit 5mm cable
12 small cable ties
there must be no metal screws protruding through to the outside since
that would present a safety hazard.
We assembled our controller into
a plastic case measuring 200mm x
250mm x 95mm (Altronics H0363).
As shown in the photos, the PCB/
heatsink assembly is installed inside
the case using a pair of brackets cut
from aluminium angle. These brackets
are screwed to the heatsink using M3 x
10mm screws, nuts & shakeproof washers and secured to the short pillars in
72 Silicon Chip
1 Nylon* M4 x 15mm machine screw
(to secure P-clamp)
3 Nylon* M4 nuts
2 M4 x 20mm machine screws &
nuts
4 M4 shakeproof washers
4 M3 x 20mm machine screws
4 Nylon* M3 x 20mm screws (to
secure fan)
4 Nylon* M3 nuts
5 M3 x 16mm machine screws
6 M3 x 10mm machine screws
5 M3 x 9mm untapped metal
spacers
14 M3 star washers
3 M3 flat washers
8 M3 nuts
4 No.4 x 9mm self-tapping screws
1 250mm length mains-rated heavyduty green/yellow striped wire
1 200mm length mains-rated extraheavy-duty red wire
1 200mm length mains-rated extraheavy-duty dark-blue
wire
1 200mm length mains-rated extraheavy-duty white wire
1 300mm length 6-8mm diameter
heatshrink tubing
1 300mm length aluminium
L-shaped extrusion, 20 x 10mm
* Use genuine Nylon (polyamide)
parts rather than clear plastic
Semiconductors
1 STGIPS20K60 3-phase
IGBT bridge (IC1) (Mouser
511-STGIPS20K60, Digi-Key
497-10573-5-ND)
1 LM319 dual high-speed
comparator (IC2)
1 dsPIC33FJ64MC802 16-bit
microcontroller (Element14
1576842) programmed with
1010512B.HEX (IC3)
1 4N35 optocoupler (OPTO1)
(Altronics Z1647)
the base of the enclosure using No.4 x
9mm self-tapping screws.
Mounting the fan
Before installing the PCB, drill four
mounting holes in the front side panel
of the case for the fan and grille. The
fan goes right in the middle of the
panel and must be orientated so that
it blows air into the case. The airflow
direction is indicated with arrows
moulded into the plastic housing.
When drilling the holes, make sure
2 HCPL2531 high-speed dual
optocouplers (OPTO2, OPTO3)
(Element14 1021247)
1 LM317T adjustable linear regulator
(REG1)
1 3mm green LED (LED1)
1 3mm yellow LED (LED2)
1 3mm red LED (LED3)
1 BC337 NPN transistor (Q1)
1 5.1V 0.4W/1W zener diode (ZD1)
1 15V 1W zener diode (ZD2)
1 GBJ3508 35A SIL bridge rectifier
(BR1) (Mouser 833-GBJ3508-BP,
Digi-Key GBJ3508-BPMS-ND)
9 1N4004 1A diodes (D1-D9)
Capacitors
3 470µF 400V snap-in
electrolytic (Altronics R5448)
3 470µF 25V electrolytic
1 10µF 25V electrolytic
4 10µF 25V SMD ceramic
[2012/0805] (Element14 1867958)
1 220nF X2 250VAC (22.5mm pitch)
(Jaycar RG5238, Altronics R3127)
14 100nF monolithic ceramic
2 47nF X2 250VAC (15mm pitch)
(Jaycar RG5234, Altronics R3117)
1 10nF MKT or ceramic
Resistors (0.25W, 1%)
1 1MΩ
2 4.7kΩ
2 620kΩ
2 1.5kΩ
2 150kΩ
1 680Ω
1 47kΩ
2 470Ω 0.5W
1 16kΩ
1 180Ω
1 10kΩ
1 110Ω
4 8.2kΩ
11 100Ω
3 4.7kΩ 5W 5%
1 10Ω
1 0.015Ω 2W SMD resistor
[6432/2512] (Element14 1100059,
Digi-Key MCS3264R015FERCT-ND)
Note: additional components are
required for external motor run/speed/
direction control – see text and Fig.11.
that the fan (when mounted internally)
will sit all the way down against the
bottom of the case (so that the lid will
still fit). You can use the grille as a
template to locate the four 3mm holes,
one in each corner. You will also have
to make a 50mm-diameter cutout in
front of the blades, so that the fan can
draw air into the case.
While you’re making holes in the
box, drill a row of 6mm holes along
the bottom half of the case side panel
opposite the fan (see photo), to allow
siliconchip.com.au
Note: early prototype PCB shown.
This view shows how the PCB assembly is mounted on the heatsink. Be sure to mount the PCB in place
and tighten BR1 and the IGBT module (IC1) down on the heatsink before soldering their leads.
fresh air to be blown out of the box
when the fan is running. The more
holes you drill, the better the airflow
will be (to a point) but a row of 15
should be adequate.
If you are using a larger case than
that specified, you may want to consider using a 230VAC 120mm fan instead,
which will move substantially more
air and thus provide extra cooling.
Secure the fan and the matching
grille (with filter) in place using four
Nylon M3 x 20mm screws, nuts and
shakeproof washers.
Mains socket
If fitting a standard mains socket
for a single-phase motor, mark out
the three hole positions to the right
of the fan. You will need to rotate
it about 45°, ie, with screw holes at
upper-left and lower right. The screw
holes are 4mm while the central hole
needs to be large enough to comfortably fit four mains-rated wires through
(about 12mm diameter) and should be
smooth, ie, no jagged edges. Mount it
using M4 x 20mm machine screws
with shakeproof washers under each
head and nut.
The mains input cable enters via
a gland to the left of the fan and is
secured to the inside of the case with
a Nylon P-clamp. Use a Nylon screw
and nut to secure it (not metal) and fit
a second Nylon nut to lock the first
one into place, so that the P-clamp
assembly cannot possibly come loose.
Complete the mains wiring accord
ing to Fig.8, taking care that everything
is properly secured with cable ties.
Note that, for a plastic case, the Earth
lead from the mains cable goes direct
to the Earth terminal on the mains
socket (GPO). A separate earth lead
is then run from the GPO to the Earth
terminal on the PCB. Use green/yellow
mains-rated cable for this connection.
The ‘W’ and ‘U’ outputs from CON2
go to the Active and Neutral terminals
of the GPO socket. Use red and blue
mains-rated cable for these connections. Don’t forget the ferrite RF suppressor on these output leads. This
Table 1: Resistor Colour Codes
o
o
o
o
o
o
o
o
o
o
o
o
o
o
siliconchip.com.au
o
o
No.
1
2
2
1
1
1
4
2
2
1
2
1
1
11
1
Value
1MΩ
620kΩ
150kΩ
47kΩ
16kΩ
10kΩ
8.2kΩ
4.7kΩ
1.5kΩ
680Ω
470Ω
180Ω
110Ω
100Ω
10Ω
4-Band Code (1%)
brown black green brown
blue red yellow brown
brown green yellow brown
yellow violet orange brown
brown blue orange brown
brown black orange brown
grey red red brown
yellow violet red brown
brown green red brown
blue grey brown brown
yellow violet brown brown
brown grey brown brown
brown brown brown brown
brown black brown brown
brown black black brown
helps reduce the RFI radiated from the
motor cable.
With the mains wiring in place, you
can then wire up the fan. It runs off
the unregulated input to REG1 (about
6-7V) and so will run quite slowly (and
hence, quietly). DO NOT wire it across
the 15V HOT rail as the insulation of
the fan may not be adequate.
Because they run adjacent to highvoltage circuity, sleeve the fan leads
with a continuous length of 5mm
diameter heatshrink tubing. Route the
fan power cable around the right-hand
side of the board and solder the leads
to the cathode of D6 (red) and anode
of D7 (blue or black) – see Fig.8.
Use the hole immediately to the
right of CON7 and the lower-right
corner mounting post as cable tie
Table 2: Capacitor Codes
Value
220nF
100nF
47nF
10nF
µF Value IEC Code EIA Code
0.22µF
220n
224
0.1µF
100n
104
.047µF 47n
473
.01µF
10n
103
5-Band Code (1%)
brown black black yellow brown
blue red black orange brown
brown green black orange brown
yellow violet black red brown
brown blue black red brown
brown black black red brown
grey red black brown brown
yellow violet black brown brown
brown green black brown brown
blue grey black black brown
yellow violet black black brown
brown grey black black brown
brown brown black black brown
ay 2012 73
brown black black black M
brown
brown black black gold brown
This is the view inside the prototype. If you are going to use external controls, then
these should be mounted on the righthand side of the case well away from the mains
outlet socket the high-voltage circuitry on the PCB – see panel overleaf. Note the row
of ventilation holes towards the bottom of the rear panel. Use cable ties to secure the
high-voltage leads, the fan wiring and the ferrite cylinder as shown.
points to clamp the fan cable (enlarge
the hole next to CON7 if necessary).
This is most important as otherwise,
the solder joints could break and the
wire could easily float around inside
the case and cause havoc.
That done, attach additional cable
ties to ensure that all the wiring is
properly tied down so that even if
one of the wires breaks or becomes
disconnected from the PCB, it can’t
make contact with something that it
shouldn’t – see Fig.8 and the photos.
In particular, note how the sleeved fan
leads and the mains Earth wire to the
GPO are tied to the mounting holes at
the top rear of the fan.
Finally, double check your work,
especially the mains wiring.
Testing
To test the control electronics, take
74 Silicon Chip
a short piece of hook-up wire and
connect it between the RUN terminal
and one of the GND terminals. Ensure
that all the DIP switches are off (sliders to the left), and set both trimpots
to about 50%. Do not connect a load
at this stage.
With the unit on the bench, apply
power and observe the neon and LEDs
(it’s a good idea to wear goggles in
case there are any nasty surprises
when power is first applied). The neon
should come on almost immediately
and the green LED should begin flashing, as the microcontroller ramps up
the output frequency. After about 15
seconds, the flashing should stop and
the green LED should remain lit.
If this is the case, the micro is working fine. If there is a problem, switch
off, unplug the unit from the mains
socket and wait until the neon has
fully extinguished. You should then
wait a further three minutes and
check the voltage across the 470μF
400V electrolytics to make sure the
circuit is safe. You can then carefully
inspect your work for errors.
Avoid making any measurements
or troubleshooting this circuit while it
is live. Only the portion of the circuit
in the bottom right hand corner of the
board inside the marked isolation barrier is isolated. The rest is at 230VAC
mains potential and is lethal.
If you want to check the control
circuitry more thoroughly, first check
that the unit is disconnected from the
mains and that the 400μF 400V electrolytics have discharged, then feed
3.3V from an external regulated power
supply into terminals 1 and 3 of the
control terminal block (ie, at CON4).
You could also simultaneously feed
siliconchip.com.au
15V from a second supply into the
+15VHOT line (cathodes of D2 & D3)
to check the control circuitry on the
high-voltage side (the negative side of
this supply can be connected to the
anodes of D1 & D4).
In fact we debugged this circuit in
this manner, even adding a third supply at 60V DC feeding the DC bus and
some 10W load resistors. This way
you can check pretty much all of the
circuitry in a safe manner.
Using it
Once you’ve made some basic
checks, you are ready to put the controller to use. We will examine three
likely use scenarios: pool pump power
saving, driving a single-phase motor
with external controls and driving a
3-phase motor.
The first step is to ensure that
your motor is suitable for use with
a speed controller of this type – see
last month’s article for full details. In
summary, any induction motor with
a centrifugal switch is NOT suitable.
Check the name-plate to ensure the
motor is rated for 230V or 240V and
1.5kW (2HP) or less. 3-phase motors should be rated for 230/400V or
240/415V operation and 1.5kW or less.
Pool pump operation
In this mode, the controller operates
in stand-alone mode (ie, without exterFig.11 (right): this front
panel label should be
placed behind a Perspex
window which is then
affixed to the case lid
using silicone adhesive.
It can be downloaded
in PDF format from the
SILICON CHIP website.
Check List
Before switching on:
(1) Check that the electrolytic capacitors are all correctly
orientated.
(2) Check that the mains
wiring and the output wiring
from CON2 to the GPO are
correct and securely laced.
(3) Check that the heatsink
is correctly earthed (ie, use
a multimeter to check for
continuity between the heatsink surface and the Earth
pin of the mains plug). Make
sure that the Earth screw to
the left of CON3 is tight and
siliconchip.com.au
has a shakeproof washer
fitted under its head.
A row of ventilation holes must be drilled across the lower section of the
rear panel (22-23mm up from the bottom) to allow the air sucked in by the
fan to be blown out of the case. These holes should be about 6mm diameter.
nal controls) and is connected to the
output of the pool pump timer switch.
When the pump is switched on, it
ramps up to full speed, then runs the
pump at full speed for 30 seconds,
before ramping the pump down to a
lower speed for the rest of the filtration
period. When the timer switch disconnects the mains, the pump coasts to
a stop, ready for the next cycle. This
was explained in more detail in the
previous article.
To achieve this, the controller is
configured as shown in Fig.11(A). The
RUN terminal is hardwired to GND, so
that the motor will automatically start,
and the DIP switch for pool pump (PP)
mode is set to ON.
The speed pot should be set for
about 70% of full speed, which gives
a good compromise between efficient
filtration and power saving. You may
need to experiment with this setting.
The ramp speed is not critical –
about 25% of rotation seems to work
quite well.
Tool spin-up mode
This is a variation on pool pump
mode, where the motor spends less
SILICON
CHIP
1.5kW Induction
Motor Speed Controller
(1) Suitable for use with delta-connected 3-phase induction motors and
single-phase induction motors without a centrifugal switch
(2) Maximum Motor Rating: 1.5kW
(3) Maximum Mains Current: 8.7A RMS (230V)
(4) Prolonged low speed operation reduces fan cooling and may overheat
the motor
WARNING
DANGEROUS VOLTAGES INSIDE DURING OPERATION
& FOR SOME TIME AFTER POWER IS REMOVED
May 2012 75
PP
DIP
SWITCH
SETTINGS
A
W
V
(A) POOL PUMP 'STAND ALONE' MODE
A
W
V
FLT
RUN
R*
EARTH
O/S
SPEED RAMP
U
NEUTRAL
MOTOR
ACTIVE
EXT
GND
E
RUN
N
PP
DIP
SWITCH
SETTINGS
* SELECT VALUE OF RESISTOR (R)
IN SERIES WITH SPEED POT TO
SET THE MINIMUM SPEED
time at full power before dropping
to the set speed (half a second rather
than 30s). This feature can be useful
for lathes or other equipment which
start off-load and is activated with Pool
Pump enabled and a shorting block
across pins 3 & 4 of the ICSP header.
Single-phase motor
with external control
LINK
MOTOR
ACTIVE
FLT
SPEED RAMP
U
NEUTRAL
EARTH
O/S
GND
E
RUN
N
EXT
SPEED (10kΩ)
(B) SINGLE-PHASE EXTERNAL MODE
In this example, we want to run a
single-phase motor with external controls. Fig.11(B) shows how it’s wired.
The speed is controlled using an
external 10kΩ pot. The EXT DIP switch
must be set to ON, to tell the micro to
read the external pot instead of the onboard trimpot. In this case, we want to
be able to run the motor at higher than
rated speed, so the O/S (overspeed)
DIP switch is also set to ON. Resistor
R sets the minimum speed.
Now when the RUN switch is clos
ed, the motor will ramp up to the speed
setting of the external pot. When the
RUN switch is opened, the motor will
ramp down to zero.
The speed control pot and the RUN
switch must be mounted on the side of
the case near the isolated area.
3-phase motor operation
* SELECT VALUE OF RESISTOR (R)
IN SERIES WITH SPEED POT TO
SET THE MINIMUM SPEED
W
V
U
R*
EARTH
FLT
SPEED RAMP
NEUTRAL
MOTOR
ACTIVE
EXT
O/S
GND
A
REV
E
RUN
N
PP
DIP
SWITCH
SETTINGS
SPEED (10kΩ) RUN
REV
(C) 3-PHASE EXTERNAL MODE
Fig.11: these diagrams show how to use the controller in pool pump mode
(A), in single-phase mode with external controls (B) and in 3-phase mode
with external controls (C).
Safely Installing External Control Wiring
The wiring to any external front-panel controls (ie, speed pot & switches) must
be run using 230VAC-rated cable. This wiring must not be longer than necessary
to reach the controls and must be securely terminated at both ends and laced
together and to fixed tie points using cable ties. This will ensure that the leads
cannot possibly come adrift and contact the motor output terminals or any other
high-voltage circuitry outside the isolation barrier. Provided you do this, the external
controls are electrically isolated from the high-voltage components and are safe.
The controls themselves must be mounted on the righthand side of the case
near the isolated area, well away from any high-voltage components. The controls
should all be sleeved with heatshrink insulation and properly secured in place.
76 Silicon Chip
The final example (Fig.11(C)) is for
a 3-phase motor with external controls. This is similar to the previous
example. The motor must be wired for
230V operation in delta configuration.
Any 3-phase wiring should be run by
a licensed electrician.
One of the big advantages of 3-phase
motors is that they can be reversed
electrically. In this example, a reverse
switch is connected between the REV
terminal and ground. If the reverse
switch is opened or closed while the
motor is running, it will ramp down
to zero speed, pause for a short time
and then ramp back up in the opposite
direction.
Extended low-speed caution
Finally, we should warn against
running any induction motor, singlephase or 3-phase, at low speeds for
extended periods. Where fitted, the
internal fan will be ineffective at low
speed and so there is no cooling.
In fact, larger motors designed for
speed control often have separately
powered cooling fans for this reason.
However, these tend to be rated over
1.5kW and thus are not suitable for use
SC
with this speed controller.
siliconchip.com.au
STIC
FANTAIDEA
GIFT UDENTS
FOR SFT ALL
O S!
AGE
THEAMATEUR SCIENTIST
An incredible CD with over 1000 classic projects
from the pages of Scientific American,
covering every field of science...
THE LATEST
VERSION 4 –
WITH EVEN
MORE
FEATURES!
Arguably THE most IMPORTANT collection
of scientific projects ever put together!
This is version 4, Super Science Fair Edition
from the pages of Scientific American.
As well as specific project material, the CDs
contain hints and tips by experienced amateur
scientists, details on building
science apparatus, a large
database of chemicals and
so much more.
ONLY
62
$
00
PLUS $10 Pack and Post
within Australia
NZ P&P: $AU12.00,
Elsewhere: $AU18.00
“A must for every science student,
science teacher, science lab . . . or simply
for those with an enquiring mind . . .”
Just a tiny selection of the incredible range of projects:
! Build a seismograph to study earthquakes ! Make soap bubbles that last for
months ! Monitor the health of local streams ! Preserve biological specimens !
Build a carbon dioxide laser ! Grow bacteria cultures safely at home ! Build a
ripple tank to study wave phenomena ! Discover how plants grow in low gravity !
Do strange experiments with sound ! Use a hot wire to study the crystal structure
of steel ! Extract and purify DNA in your kitchen !Create a laser hologram ! Study
variable stars like a pro ! Investigate vortexes in water ! Cultivate slime moulds !
Study the flight efficiency of soaring birds ! How to make an Electret ! Construct
fluid lenses ! Raise butterflies as experimental animals ! Study the physics of
spinning tops ! Build an apparatus for studying chaotic systems ! Detect metals in
air, liquids, or solids ! Photograph an ant's brain and nervous system ! Use
magnets to make fluids into solids ! Measure the metabolism of an insect . . . !
and many, many more (a thousand more, in fact!)
See the V2 review in SILICON CHIP, October 2004. . . or read on line at siliconchip.com.au
This is the ALL-NEW Version 4 . . . it’s even BETTER!
HERE’S HOW TO ORDER YOUR COPY:
BY PHONE:*
(02) 9939 3295
9-5 Mon-Fri
BY FAX:#
<at>
(02) 9939 2648
24 Hours 7 Days
BY EMAIL:#
silicon<at>siliconchip.com.au
24 Hours 7 Days
BY MAIL:#
BY PAYPAL:#
PO Box 139,
Collaroy NSW 2097
silicon<at>siliconchip.com.au
24 Hours 7 Days
* Please have your credit card handy! # Don’t forget to include your name, address, phone no and credit card details.
BY INTERNET:^
siliconchip.com.au
24 Hours 7 Days
^ You will be prompted for required information
There’s also a handy order form inside this issue.
Exclusive in SILICON
Australia to: CHIP siliconchip.com.au
siliconchip.com.au
May 2012 77
CIRCUIT NOTEBOOK
Interesting circuit ideas which we have checked but not built and tested. Contributions will be
paid for at standard rates. All submissions should include full name, address & phone number.
PICAXE 433MHz data transmitter & receiver
along with a station identifier and
checksum, to a 433MHz ASK transmitter module (Jaycar ZW-3100).
The data is transmitted at 2400 baud
and the period between transmissions can be set in software in onesecond increments up to over 18
hours. The initial rate is 15 seconds.
During the break period, transmission is off to reduce power consumption. The transmitter and micro are
powered from via two alkaline cells
VR1
1
Vdd
5 P2
10k
7
1
ICSP
SKT
22k
2
3
P3
IC1
PICAXE
-08M2
P0
2 SER
IN
P1
P4
172mm
ANTENNA
4
6
Vcc
3
433MHz
TX
MODULE
DATA
Vss
10k
8
ANT
GND
0V
ANT
Vcc
DATA
GND
TRANSMITTER
172mm
ANTENNA
The 433MHz data transmitter
circuit is shown at left, while below
is the receiver circuit. Note that
the transmitter supply should be
regulated for accurate ADC readings.
433MHz Tx MODULE
* REGULATED 3V FOR ACCURATE ADC READINGS
Vcc
ANT
433MHz
RX
MODULE
GND
4
7
1
ICSP
SKT
22k
2
3
1
Vdd
P3
P0
2 SER
IN
P2
IC1
PICAXE
-08M2
P1
P4
Vss
8
10k
1
2
5
3
6
4
10k
3
5
6
470
8
4
18
SCL
Vdd
GP0
GP1
SDA
GP2
12
IC2 GP3 13
MCP23008 14
A2
A1
GP4
A0
GP5
GP6
RESET
INT
A
Vss
9
GP7
2
Vdd
RS
10
11
6
EN
16 x 2
LCD MODULE
CONTRAST
D7 D6 D5 D4 D3 D2 D1 D0 GND
1
14 13 12 11 10 9 8 7
VR1
10k
3
R/W
5
LCD
CONTRAST
15
16
17
433MHz Rx MODULE
LED1
LED
K
RECEIVER
78 Silicon Chip
5V
–
2x
4.7k
DATA
+
S1
470 F
16V
100nF
K
A
Vcc
DATA
DATA
GND
ANALOG
INPUT
VOLTAGE
+3V*
S1
100nF
ANT
GND
GND
Vcc
This design uses a pair of PICAXE08M2 microcontrollers and lowcost 433MHz ASK transmitter and
receiver modules to provide a wireless remote data monitor with LCD
readout.
The transmitter comprises a PIC
AXE08M2 that monitors DC voltages
via potentiometer VR1. The micro’s
internal ADC converts the DC voltage at pin 5 to a digital signal with
10-bit resolution and sends the data,
giving a nominal 3V. However, for
accurate analog readings the DC supply should be regulated as the digital
count will vary as the battery voltage
changes. Standing current is of the
order of 1mA and less than 10mA
during the short transmission bursts.
The receiver comprises a PICAXE08M2, a 433MHz ASK receiver
module (Jaycar ZW-3102), an MCP
23008 I2C port expander (available
from www.futurlec.com) and a
2-line LCD display.
The PICAXE monitors the data
from the 433MHz receiver module.
When a valid data packet is received
(station identified correctly and
checksum correct), the LED connected to port 4 (pin 3) is flashed
and the received data is processed
and sent via I2C protocol to the 8-bit
I2C port expander which interfaces
to a 2-line 16 character LCD display.
The display shows a text message,
the data and a packet counter to
confirm that all is working correctly.
siliconchip.com.au
The LCD module is operated in 4-bit
mode.
The 433MHz receiver has no muting facility. Hence, during periods of
no-signal, the AGC sets the gain to
maximum and there is a high level
of noise on the data line. To ensure
reliable reception, the transmitter
is switched on for a short duration
to allow the receiver AGC to operate and then briefly low again with
the receiver AGC still active, thus
eliminating output noise before the
data stream is sent.
The software can be easily modified for other purposes, for example to send text messages, water
level monitoring etc. A number
of transmitter and receiver pairs
could operate simultaneously using
different station identification and
slightly different transmit periods.
A PICAXE14M2 with software
changes could be used in lieu of the
PICAXE08M2 to avoid the need for
the port expander.
The range is at least 200 metres
outdoors and 25 metres indoors.
As the operating frequency is in the
Industrial, Scientific and Medical
band that does not require licencing, other devices such as door bell,
garage door and keyless car entry
transmitters could interfere with operation. However, these transmitters
are generally not continuous and a
data packet would only be lost if one
of these other units was transmitting
physically close to and concurrently
with the data transmission from this
circuit.
Data reception reliability is high
due to the use of station ID and
checksum. Data security is low as the
transmit packet can be monitored by
others. A simple encryption routine
has been implemented that would
puzzle anyone who had the time
and inclination to eavesdrop on the
data stream.
A 3-way pin header (ICSP SKT)
and two resistors provide a simple
PICAXE programming interface on
both the transmitter and receiver
circuits.
The software (433MHz Tx-Code.
bas and 433MHz Rx-Code.bas) can
be downloaded from the SILICON
CHIP website.
Phillip Webb,
Hope Valley, SA. ($60)
siliconchip.com.au
Cheap electronic ballast for
fluorescent light fittings
The easiest and cheapest method
to obtain an electronic ballast for a
fluorescent lamp fitting is to obtain
one from a compact fluorescent
lamp (CFL). Tests with a salvaged
electronic ballast from a defective
18W GEC CFL showed it to work
well for 600mm 18W and 1200mm
36W fluorescent tubes.
In fact, compact fluorescent lamps
are now so cheap that it is worth
buying a new one in order to get a
working electronic ballast. If you
don’t have any CFLs, purchase one
rated at 20W or 24W (eg, Philips
Tornado 24W or GE 20W).
To remove the electronic ballast
from the CFL, first desolder the
two terminal joints on the bayonet
lamp holder. Then use a hacksaw
to cut the base off the bayonet lamp
holder. Use side cutters to cut off the
plastic base to get access to the PCB
of the electronic ballast. It can then
be disconnected from the CFL tube.
You will need to mount the electronic ballast PCB securely inside
your lamp fitting and well insulated,
perhaps wrapped in duct tape, to
avoid shorts. Connect the filaments
of the fluorescent tube lamp to the
four terminals of the electronic
ballast.
No starter is required and switching on the fluorescent tube lamp will
be virtually instantaneous, with no
initial flickering. Overall power dissipation is reduced compared to a
conventional ballast and the bright-
Micha
is this m el Ong
of a $15 onth’s winner
0 gift vo
ucher fr
Hare &
Forbes om
An electronic ballast salvaged from
a compact fluorescent lamp.
ness from the fluorescent tube lamp
will be slightly higher.
If a Philips Tornado 24W CFL is to
be used, the 6.0µF 400V Aishi electrolytic capacitor on the PCB should
be replaced with a 6.8µF 400V or a
10µF 400V electrolytic capacitor as it
is prone to leakage. The 3.9µF 400V
electrolytic capacitor on the board of
the GE 20W CFL appears to be fine.
The wrecked CFL can disposed of
at the designated green bins available at shopping complexes.
Michael Ong,
Wembley, WA.
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$
$
contribution
$
$
$
$
$
$
$ As you can see, we pay $$$ for contributions to Circuit Notebook. But $
$
$
$ each month the best contribution (at the sole discretion of the editor) $
$
$
$ receives a $150 gift voucher from Hare&Forbes Machineryhouse.
$
$ That’s yours to spend at Hare&Forbes Machineryhouse as you see fit – $
$
$
$ buy some tools you’ve always wanted, or put it towards that big
$
$
$
$ purchase you’ve never been able to afford! Contribute NOW and WIN! $
$
email your contribution now to editor<at>siliconchip.com.au or post $
$
$
to PO Box 139, Collaroy NSW 2097 $
$
$
$
$
$
$
$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
May the best man win!
May 2012 79
Circuit Notebook – Continued
MOTOR
MAINS ACTIVE
MPX 5999D
PRESSURE
SENSOR
SOLENOID
+11.4V
+5V
10nF
A
1 F
LED1
3
8
2
3
470pF
IC1a
LM358
RELAY1
K
2.2k
1
Q2
BC327
+5V
10k
K
C
A
B
E
D4
Q1
BC337
K
A
2.2k
VR2 1M
COMPARATOR
+12V IN
LED2
K
C
10k
A
10k
B
A
10k
D3 1N4148
THRESHOLD 1k
E
RELAY2
4
VR1
K D2
1k
1
2
MAINS VOLTAGE
+11.4V
D1
POWER
A
K
RELAY1 & RELAY2: JAYCAR SY-4042
OR SIMILAR
+11.4V
S1
560k
100k
6
REG1 7805
IN
100 F
16V
100 F
16V
MPX 5999D
100nF
+5V
OUT
GND
10 F
10 F
1 2 3 4 5 6
A
D1,D2,D4: 1N4004
K
A
K
K
A
Air-compressor controller for a sand-blaster
This circuit was devised to control
a large volume air compressor which
is used for sand-blasting. This uses
a 4kW single-phase electric motor
and a solenoid valve that controls
the loading and unloading of the
compressor.
When starting, the compressor
motor needs to be powered for a few
seconds before the solenoid valve is
closed. This allows the motor time
to provide sufficient air pressure. In
operation, the compressor needs to
be controlled to provide between 95
psi (655kPa) and 100 psi (689kPa)
to produce a consistent sand blast.
To do this, a Freescale MPX5999D
differential pressure sensor is used
to monitor the compressor's pressure. This sensor is designed to
operate up to 1000kPa (150 psi) and
produces a DC output of about 3.3V
4
3
IC2
7555
5
2
LEDS
1N4148
8
7
at 655kPa and 3.4V at 689kPa.
A modified version of the Simple
Voltage Switch for Car Sensors (SILICON CHIP, December 2008) is used
to control the compressor motor.
Switching is via RELAY1 which has
contacts rated for 220VAC at 30A. A
second relay (RELAY2) is controlled
by a 7555 timer to provide the delay
before the solenoid valve kicks in.
The output from the pressure sensor is monitored by op amp IC1a,
connected as a comparator. The noninverting input (pin 3) is connected
to threshold trimpot VR1 via a 10kΩ
resistor. Hysteresis is included using
diode D3 and trimpot VR2. VR1 is
adjusted to give about 3.3V.
When IC1a’s output is high, VR2
and the 1MΩ resistor pull pin 3
slightly higher than 3.3V and VR2
should be adjusted to give 3.4V at
Issues Getting Dog-Eared?
1
100nF
7805
BC327, BC337
B
E
TIMER
GND
IN
C
GND
OUT
pin 3. When IC1a’s output is low,
pin 3 will then be at 3.3V.
When the pressure is low, pin 1 of
IC1a will be high. This drives transistor Q1 and RELAY1 to power the
compressor motor. When the pressure goes above 689kPa, the comparator goes low, the relay switches
off and the compressor motor stops.
IC2 is a CMOS 555 timer that provides the start-up delay. At power
up, when S1 is closed, the 555 timer
is triggered because the trigger (pin
2) is momentarily pulled low by a
100nF capacitor. This sets its pin
3 output high, so transistor Q2 and
RELAY2 are off. Thus no power is
provided for the solenoid.
The 10μF capacitor at pins 6 &
7 now begins to charge and after
about 8s, pin 3 goes low and Q2
and RELAY2 turn on and turn on
the solenoid valve.
SILICON CHIP.
Keep your copies safe
with our handy binders
Available Aust, only. Price: $A14.95 plus $10.00 p&p per order (includes GST). Just fill in and mail the handy
order form in this issue or ring (02) 9939 3295 and quote your credit card number.
80 Silicon Chip
siliconchip.com.au
siliconchip.com.au
GND
3.3k
10nF
5
1
2
6
IC1
555
100 F
25V
VR3
100k
E
X
B
K
A
100k
E
1.5k
B
C
Q1
BC548
1N4004
1.5k
1 F MKT
10k
100 F
16V
LED
560k
4
3
1
2
BRANDED
SIDE
UGN3503U
+5V
MOTOR
SHAFT
2
VR1
10k
3
7
2
3
S
S
UGN3503U
1 HALL SENSOR
N
N
+5V
MAGNETS
IC3
741
K
A
6
100nF
LED1
K
A
+5V
GND
E
E
Q3
BC548
1.5k
K
3
A
D2
1N4004
4
10k
7
8
1
2.2 F
25V
2
C
Q2
BC548
X
100nF
GND
IN
OUT
REG2 7805
A
D3
1N4004
5
IC2
555
6
8
7
VR2
20k
S1
MOTOR
ON
+12V
10nF
K
A
4
100 F
25V
3
GND
D1
1N4004
IN
OUT
REG1 7812
2200 F
35V
K
+
B
~
C
–
RLY1
12V
AC
B
C
BC548
CONTACTOR
230V
IN
OUT
7812, 7805
A
N
MOTOR
E
A
N
T1
~
This unit protects induction motors from
burnout in the event that they stall or are running too slowly, as may happen if the mains
voltage is low.
Two magnets are glued on opposite sides of
the motor shaft and these are monitored by an
adjacent Hall Effect sensor and fed to a missing pulse detector. If this detects no pulses or
pulses that are too slow, it removes power from
the motor.
The circuit is powered from a 12V transformer which feeds a bridge rectifier and a
2200µF capacitor. The unregulated DC is then
fed to 12V regulator REG1. This powers two 555
timers (IC1 & IC2) and a 12V relay with 250VAC
contacts. 5V regulator REG2 powers the Hall
Effect sensor and 741 op amp IC3.
The motor is turned on by switch S1 which
powers up both 555 timers. IC1 is connected as
a one-shot monostable with a 5-second pulse
output from its pin 3. This turns on transistor
Q3 and the relay, allowing the motor to come
up to speed. It also holds pin 4 (Reset) of IC2
high for this period, then swings low.
The pulses from the Hall Effect sensor are fed
to op amp IC1 which is configured as a comparator. Its output drives transistor Q1 via LED1
which flashes when pulses are present. Q1
drives Q2 via a 1µF capacitor. Q2 is connected
to pin 6 of IC2 which is connected as a missing
pulse detector. With Q2 out of the circuit, IC2
would run as a conventional monostable, with
a high output from its pin 3 until its pins 2 &
6 rise to about 9V whereupon pin 3 goes low.
With Q2 fed with pulses from the motor, it
continually discharges the 2.2µF capacitor at
pin 2 of IC2, preventing its pin 3 from going low.
If the motor pulses are too slow or non-existent,
IC2 times out, its pin 3 goes low, Q3 turns off
and the relay switches off, disconnecting power
from the motor. This also pulls pin 4 of IC2 low
so the circuit is then disabled. The motor can
only be re-started by turning off the power and
then turning it on again via switch S1.
While the relay should have 250VAC contacts, it is only used to switch a 250VAC contactor with ratings to suit the motor.
An induction motor with a rated speed of
1440 RPM will generate 48Hz with two magnets installed on the shaft. VR2 will need to be
adjusted so that IC2 is only triggered for low
motor speeds. VR1 is adjusted so that pin 2 of
IC3 is slightly above pin 3 when no magnet is
near the Hall Effect sensor.
The magnets should be installed so that their
south poles face the sensor.
Geoff Coppa,
Toormina, NSW. ($70)
BR1 W04
Motor protector uses
missing pulse detector
May 2012 81
Circuit Notebook – Continued
Maximite-based
ultrasonic rangefinder
This circuit for an ultrasonic rangefinder is based on the Maximite Microcomputer (SILICON CHIP, March-May
2011). Many of the additional parts
were salvaged from an old ultrasonic
burglar alarm kit but suitable equivalents should be available. With the parts
shown, the minimum sensing range
is around 30mm and the maximum
around 4-5m.
In brief, an ultrasonic transmitter
produces a burst of high-frequency
sound waves (40kHz) which then
bounce off the nearest hard object in
their path and return to be picked up
by the receiver unit. The delay between
transmission and reception indicates
the distance that the sound waves
travelled and thus the distance to the
reflecting object.
IC1d & IC1e form an oscillator with
40kHz crystal X1. The resulting square
wave is buffered by IC1b & IC1c and
then fed to IC1f via a 10kΩ resistor.
The output of IC1f is inverted again by
IC1a and these two out-of-phase signals
drive ultrasonic transmitter Tx to produce the sound waves. Because the two
ends of the transmitter are driven with
out-of-phase square waves, it is driven
at 18V peak-to-peak.
The reflected sound waves are pick
ed up by receiver Rx. Its output is
low-pass filtered by a 1kΩ resistor and
680pF capacitor and then attenuated
by VR2, which sets the sensitivity. The
signal is then AC-coupled via a 680pF
capacitor into IC2a, part of a CA3401
quad current feedback (Norton) amplifier.
These are like op amps but amplify
current rather than voltage. External
resistors convert the signal voltage into
a current and then the output current
back into a voltage. Here, they are used
as inverting amplifiers with a gain of
about 43.7 (1MΩ ÷ (27kΩ || 150kΩ)).
With three stages in series, the overall
gain is nearly 100dB.
Each stage works as follows. Each
inputs is the base of an NPN transistor
with its emitter connected to ground.
So current only flows into an input
when its voltage is at least one diode
drop above ground (ie, about 0.6V). The
current from the output is proportional
82 Silicon Chip
to the difference between the currents
flowing into the non-inverting and
inverting inputs.
So with the non-inverting inputs
tied to ground, current flows from the
amplifier output until the voltage at the
inverting input rises above 0.6V. We
can then calculate the quiescent output
voltage by considering the 1MΩ/150kΩ
divider between the output and inverting input, ie, 0.6V x (1MΩ + 150kΩ) ÷
150kΩ = 4.6V. This is close to half the
nominal 9V supply.
When a signal is picked up by the
receiver, the voltage is converted to a
current by the 27kΩ resistor connecting
it to the inverting input of IC2a and this
current is summed with the feedback
current and thus appears amplified at
the pin 5 output. A 2.2pF capacitor
across the 1MΩ feedback resistor reduces noise in the output by decreasing
the feedback network impedance at
high frequencies.
After passing through the three
gain stages, the amplified signal is
AC-coupled via a 10nF capacitor to
a window comparator based on IC3,
an LM393 dual comparator. The common output (at pins 1 & 7) goes low
whenever the signal swing exceeds
4.1V peak-to-peak. The window thresholds are set at 0.45V and 4.55V by the
2.2kΩ/20kΩ/2.2kΩ voltage divider between +5V and ground. The incoming
signal is biased to 2.5V by a 10kΩ resistor from the centre of the same divider.
A 4.7nF capacitor at the comparator
output ensures the output pulse has a
minimum duration. When IC3’s opencollector output(s) switch off, it charges
via a 10kΩ resistor from the 3.3V rail.
The comparator output signals the
Maximite when a reflected signal is
detected and also resets the RS flipflop
formed by IC4a and IC4d, two CMOS
NAND gates. The flipflop output controls IC4c, another NAND gate, which
gates the signal from the audio output of
the Maximite. This is set up to provide
a 200kHz square wave.
This allows the Maximite to accurately measure the time between ultrasonic transmission and reception. The
Maximite pin 10 goes low to send the
signal burst, turning off NPN transistor
Q1 which, when on, suppresses the
40kHz signal to the transmitter.
When pin 10 goes low it also sets the
aforementioned RS flipflop, bringing
pin 11 of IC4d high and thus pin 9 of
IC4c also. The audio pulses at pin 8 of
IC4c are then fed into the Maximite via
input pin 11 which is set up to count
the pulses. When the reflection is detected and the RS flipflop is reset, pin 9
of IC4c goes low and so the pulses are
no longer received. The counter stops
incrementing.
The Maximite senses when the flipflop is reset using its input pin 9 and
it can then read the counter value to
determine the delay (by multiplying the
result by 5μs, ie, 1 ÷ 200kHz). This can
then be used to calculate the distance
the sound waves travelled.
To do this final calculation it is necessary to measure ambient temperature,
as the speed of sound varies with it.
This is done by 100kΩ thermistor
TH1 (B=4500). TH1 forms a voltage
divider with a 100kΩ resistor and this
voltage is applied to analog input pin
8 of the Maximite. The Maximite uses
its internal analog-to-digital converter
(ADC) to measure this voltage and thus
the temperature.
The computed range is displayed on
a 16x2 alphanumeric LCD. VR1 allows
contrast adjustment while the backlight
is powered from 5V via a 36Ω currentlimiting resistor. The 5V supply is
provided by an LM7805 regulator (see
"Measuring Short Intervals With the
Maximite", Circuit Notebook, March
2012).
The unit is calibrated using S1 and
S2. Place the rangefinder 0.5m away
from and perpendicular to a wall. Set
S1 appropriately and press S2. Then
move the unit to be 3.5m from the
wall, change S1 and press S2 again.
The results from these measurements
are stored in a file and used when calculating future range readings. S1 can
then be set to the “RUN” position for
normal operation.
Note that when building the unit,
the leads connecting the ultrasonic
transmitters and receivers should be
run using shielded cable and kept as
short as possible. Don’t run them too
close to each other and be careful with
the layout and grounding of the amplification section because of its high gain.
Jack Holliday,
Nathan, Qld. ($70)
Editor’s note: the CA3401 is now obsolete but “new old” stock is still available
from Futurlec (www.futurlec.com.au).
siliconchip.com.au
40kHz
Rx
680pF
1k
+9V
10nF
100nF +5V
2.2k
150k
27k
10k
10k
10k
2.2k
680pF
VR2
5k
START
S2
10k
6
5
2
3
10k
1
6
4
1M
IC2a
14
7
IC3:
LM393
1
S1
100k
TH1
100k
2.2pF
IC3b
IC3a
8
RUN
CAL 3.5m
10k
CAL
0.5m
+3.3V
siliconchip.com.au
May 2012 83
5
680pF
100nF
4.7nF
9
14
7
8
150k
27k
20
18
19
17
15
16
11
10
10k
2
3
2.2pF
1M
IC2b
IC2: CA3401
AUDIO
OUT
3.3V
OUT
MAXIMITE
MODULE
9V
IN
+9V
4
6x
10k
10k
8
9
11
3
150k
27k
IC4c
IC4d
IC4a
680pF
10
13
12
2
1
+3.3V
6
5
7
4
IC1b
+5V
B
E
12
2.2pF
1M
IC2d
10
Vdd
1
27k
100k
11
3
13
8
7
IC2c
1(2* )
GND
10
IC1a
2
270k
8
1
GND
IN
9
IN
GND
5
R/W
IC1d
X1 40kHz
15
IC1e
IC1:
4049B
7 12
10k
14
IC1f
470 F
OUT
REG1 LM7805
16x2 LCD MODULE
IC1c
Q1
BC548
5 6
C
+5V
D7 D6 D5 D4 D3 D2 D1 D0
14 13 12 11 10 9 8 7
EN
RS
11
6
4
27k
4
IC4b
14
IC4: 4011B
100nF
OUT
E
B
C
BC548
3
7805
CONTRAST
9
680pF
680pF
100nF
100 F
GND
VR1
10k
40kHz
Tx
+ 9V
DC
– IN
Building our new
SemTest
Pt.3: By JIM ROWE
Now that we have looked at the full circuit of our new Discrete
Semiconductor Test Set, it’s time to describe its construction and
the setting-up procedure. We also describe how to fit a crowbar circuit
to quickly discharge the HT after making high-voltage measurements.
A
S SHOWN IN the photos, the
SemTest is built in an ABS
enclosure measuring 222 x 146 x
55mm. Apart from VR10 (the Mosfet
VGS pot) and the five pushbutton
switches (which mount directly on
the front panel), all the components
are mounted on one of two PCBs.
Both boards are double-sided, so
there is no need to fit any wire links.
Incidentally, our prototype photos
in the February & March 2012 issues
showed numerous link positions on
both PCBs. These have now been
84 Silicon Chip
incorporated into the copper patterns
on the top layers of both boards so
that is one less tedious task needing
to be done.
The main board (coded 04103121)
mounts in the bottom of the enclosure, while the display board (coded
04103122) sits behind the front panel
and is spaced 18mm from it. The two
boards are linked via three flat ribbon
cables fitted with IDC connectors.
Rotary switch S2 is mounted on
the lower PCB. Its control shaft is
42mm long, so that when the case is
assembled, it passes through clearance
holes in both the display PCB and the
front panel.
Power switch S1 and 12V input connector CON1 are both located on the
righthand end of the main board, towards the rear, and pass through holes
in the righthand end of the enclosure.
A small hole nearer the front of the
enclosure provides access to trimpot
VR2, which is used to set the micro’s
2.490V reference voltage.
Six similar holes along the top edge
of the righthand end of the enclosure
siliconchip.com.au
COMMON DISCRETE SEMICONDUCTOR DEVICE CONNECTIONS
DIODES
B
A
A
E
A
K
A
K
A
A
A
C
C
K
A
C103B, BT149
G
D
S
G
A
2N7000, VK10KN
K
BT169D, 2N5060
C106D
(TO-225)
(DO-247)
S
(TO-225)
E
D
S
(TO-264)
B
C
A
C
E
B
LEDS
C
(TO-218)
E
A
B
CATHODE
BAND
A
E
K
A
(TO-220)
G
C106D1, C122E
(TO-263A D-PAK)
PUTS
2N6027
G
S
K
(SOT-93/
TO-264)
G
S
A
TRIACS
D
IGBTS
A1
A2
C
FGA25N120ANTD
G
BT137F, SC141D, SC151D, TAG225
A2
C
E
B
(TO-5)
B
K
A
K
(TO-262)
D
(TOP-3)
K
A
D
BD135-6-7-8, BD139-140, BD681-2,
BF469-470, MJE340-350
G
D
G
D
(TO-220)
C
G
BS170, BS250
G
C
D
S
(TO-92/72)
E
PN100, PN200, C8050 ETC
A
K
S(1)
C
K
B
A
K
S
G
A
K
A
G
D(2)
B
(TO-220)
K
G
D
G1(4)
E
B
BC639, BC640, 2SC3242
K
(TO-220)
K
G2(3)
(TO-92/14)
K
SCRS
2N7002,
DMP2215L
BF998
(TO-92/17)
BC327-8, BC337-8, BC546-7-8-9,
BC550, BC556-7-8-9, BC560,
2N2222A, 2N3638
MBR735
K
MOSFETS
BJTS
K
K
C
E
G
C
E
(TO-3PN)
A1
C
(TO-3)
A2
(TOP-3)
G
ABOVE: this handy table shows the pin connections for many discrete semiconductor devices. The ZIF socket on the front
of the SemTest makes it easy to connect devices for testing.
are used to access various trimpots
mounted on the righthand end of the
display board.
Main PCB assembly
Use the layout diagram of Fig.10 as
a guide to assembling this board.
Begin by fitting all the smaller resistors, which should be of 1% tolerance.
Note that one of these resistors (which
mounts about 20mm above and to
the right of IC3) is marked “0Ω/68Ω”,
because its value depends on the
type of relay you use for Relay 1. If
you use a relay with a 12V coil, this
resistor can be replaced with a wire
link (or zero ohm resistor). With a 6V
relay (eg Jaycar SY-4058), the resistor
should be 68Ω.
The 1W and 5W resistors are next.
Mount the 5W resistors about 1.5mm
above the surface of the board, to allow some ventilation if they become
hot in operation.
siliconchip.com.au
Follow with trimpots VR1 and VR2.
VR1 (50kΩ) mounts near IC1 while
VR2 is a horizontal multi-turn 10kΩ
unit which mounts at lower right.
Once these are in, fit the capacitors.
The two 47µF 450V electrolytics need
to be laid on their sides and secured
with small cable ties.
Now fit the DC input connector
CON1, followed by power switch S2,
DIL pin headers CON2, CON3 and
CON4, the 40-pin DIL socket for IC4
and the 8-pin DIL sockets for IC1 and
IC3. Relay drivers IC5 and IC6 do not
need sockets and are soldered direct
to the PCB later during the assembly.
The six 1mm PCB terminal pins,
used for the various test points can
now go in. The four relays can then be
installed. Note that RLY7 and RLY8 are
mini-DIL reed relays, which should be
mounted with the orientation shown
in Fig.10.
The next step is to wind T1, the
step-up transformer for the SemTest’s
DC-DC converter. The winding and
assembly details are shown in Fig.11;
follow this exactly (or else!). Wind
each layer as closely and evenly as
possible; wind them all in the same
direction and cover each layer with a
layer of insulating tape (to both hold
that layer in place and provide insulation between it and the layer above it).
Before T1 is assembled don’t forget
the “gap” washer, cut from a small
piece of 0.06mm thick plastic sheet.
T1 can now be mounted on the
main PCB. It is held in place (as well
as being held together) by an M3 x
25mm long Nylon screw and nut.
Note that the primary start (S), tap
(T) and secondary finish (F) wires all
connect to the PCB, just to the right of
the transformer itself.
Semiconductors
Now for the semiconductors, startMay 2012 85
HT crowbar – a safety refinement
+HV
1
2
3
A
10nF
K
A
Vin
4
5
10k
D1 1N4004
V+
LK1
AG
100 F
16V
PUT1
2N6027
A
SCR1
TYN816
KG
K
K
100k
A
A
AG
K
HV DC CROWBAR
TYN816
2N6027
1N4004
SC
100
1W
GND
CON1
2012
330
1W
K
A
K
A
KG
CENTRE
LEAD CUT
SHORT IN
THIS PROJECT
Fig.7: the circuit monitors the converter’s power supply rail in the SemTest
& when Vin drops below 6V, PUT1 & SCR1 turn on to discharge the 47μF
capacitors across the high-voltage output.
S
INCE PRODUCING our proto
type SemTest presented in the
February & March issues, we have
developed a further refinement – an
add-on crowbar module which in-
Parts List
1 PCB, code 04105121, 56 x
40.5mm (available from SILICON
CHIP)
1 M3 x 6mm machine screw & nut
2 M3 shakeproof washers
1 100mm length red heavy duty
mains-rated hook-up wire
1 200mm length black heavy duty
mains-rated hook-up wire
1 200mm length yellow hook-up
wire
1 70mm length 30mm diameter heatshrink tubing (Jaycar
WH5658, Altronics W0919A)
Semiconductors
1 TYN816 SCR (SCR1; Altronics
Z1778)
1 2N6027 PUT (PUT1; Jaycar
ZT2397, Altronics Z1410)
1 1N4004 1A diode (D1)
Capacitors
1 100µF 16V electrolytic
1 10nF monolithic multi-layer
ceramic
Resistors (0.25W, 1%)
1 100kΩ
1 330Ω 1W (5%)
1 10kΩ
1 100Ω 1W (5%)
86 Silicon Chip
stantly kills the high voltage applied
to the ZIF socket at the conclusion
of any breakdown voltage test. As a
further safety measure, it also kills
the high voltage in the event that the
SemTest is inadvertently turned off
before a test has properly concluded.
This minimises the chance of the
user getting a shock from the test
terminals when removing the DUT
or a possible breakdown of the DUT
itself when the power is inadvertently removed.
The crowbar module is wired to
three points on the main (lower)
SemTest PCB. On our prototype,
these wires have been soldered to
specific component leads but the
final SemTest PCB has pads for
these wires.
The crowbar board senses the
11.4V supply rail to the MC34063
DC/DC converter IC1. This drops
very quickly to around 6V when a
test finishes or more slowly if the
unit is switched off during a test.
Either way, this is the trigger for the
crowbar to discharge the capacitor
bank from 600V to a few volts in
around 20ms.
Circuit description
Fig.7 shows the full crowbar circuit. It could potentially be used in
other devices but for use with the
SemTest, link LK1 is installed, to
short Vin (the sense input) and V+
(its power supply) together.
The +HV and GND terminals
at CON1 are connected across the
SemTest’s high voltage capacitor
bank. Fig.8 shows a fragment of the
SemTest circuit and demonstrates
how the crowbar module is connected. The V+ terminal goes to
pin 6 of IC1, which is at around
+11.4V when the DC/DC converter
is running and drops to 0V when it
is switched off.
While the DC/DC converter is
running, current flows from this
rail, through diode D1, charging the
100µF capacitor. As this capacitor
charges, the gate (AG) of programmable unijunction transistor PUT1
is pulled up too, via the 10kΩ and
100kΩ resistors. At the same time,
the anode (A) is pulled up via a 330Ω
resistor. The 10nF capacitor between
PUT1’s anode and gate is initially
discharged and this helps to keep the
gate at anode potential, preventing
false triggering if there are any initial
glitches in IC1’s power supply (eg,
due to relay contact bounce).
A PUT is essentially a small anodegate SCR. While a conventional SCR
is turned on when its gate is pulled
above its cathode, a PUT turns on
when its gate is pulled below its
anode, sinking current from the gate.
Both SCRs and PUTs remain on once
triggered until their anode-cathode
current flow drops below the “holding” current, in this case much less
than a milliamp.
As long as V+/Vin are held at
around 11.4V, the crowbar circuit
remains deactivated. But once Vin
drops precipitously, the 10nF capacitor begins to charge while the
100µF capacitor retains its charge,
by virtue of diode D1.
Once Vin drops below the ~6V
threshold, sufficient current flows
from PUT1’s gate to trigger it on. It
then dumps the charge in the 100µF
capacitor into SCR1’s gate (KG), via
the 330Ω current-limiting resistor.
This happens in less than 100μs if
Vin drops fast, as when a test ends
normally.
The 330Ω resistor limits the current into SCR1’s gate to around 25siliconchip.com.au
+11.4V
RELAY1
CROWBAR
3
+HV
V+
68 IC5 PIN18
1
GND
5
D2 UF4007
A
5W
80T
6
7
8
Vcc
Ips
DrC
10T
SwC
Ct
IC1
MC34063
GND
4
1nF
33k
1W
1.0k
1W
33k
1W
+OPV/+BV
T1
0.27
3
1.5k 5W
K
33k
1W
SwE
Cin5
TP4
1
B
C
E
2
E
2.2k
B
C
Q1
BC337
470nF
630V
390k
75k
1%
100k
390k
75k
1%
100k
Q3
IRF540N
G
Q2
BC327
390k
470nF
630V
S
+Vdevice
75k
1%
D
100
1.0k
47 F 1W
450V
75k
1%
390k
SET TEST
VOLTS
VR1
50k (25T)
+1.25V
100k
100k
47 F
450V
RELAY
2b
TO S2a
Fig.8: this diagram shows how the HV Crowbar module is connected to the SemTest circuit. Only three connections are
required, as indicated by the lines highlighted in red.
© 2012
CON1
+HV
04105121
100 1W
330 1W
+V
Vin
GND
D1
100F
4004
LK1
100k
10k
SCR1
TYN816
10nF
12150140
30mA, enough to trigger it reliably.
SCR1 then rapidly discharges the
high voltage capacitor bank through
the 100Ω resistor. The peak discharge current is 600V ÷ 100Ω = 6A.
PUT1 switches off as soon as it
has finished dumping the charge
of the 100µF cap into SCR1’s gate.
But SCR1 stays on until the current
through it drops below 40mA (its
holding current) so the capacitor
bank discharges to around 4V.
The specified TYN816 SCR is
rated for 800V & 16A. Do not use an
SCR with lower ratings.
PUT1
2N6027
HIGH VOLTAGE
RABW
CROWBAR
ORC EGATLOV HGIH
Fig.9: follow this layout diagram and photo to build the HV Crowbar. For the
SemTest, leave out the screw terminal block and install a wire link for LK1.
Refer to the overlay diagram,
Fig.9. Fit the two small resistors
first, followed by diode D1, with
its cathode stripe towards the right
side of the board. Use a lead off-cut
for LK1 and solder it in place. Then
install the two 1W resistors.
Wiggle the middle lead of SCR1
back and forth until it snaps off.
If there is any lead remaining,
remove it with side-cutters. Bend
the remaining two leads down and
insert them through the holes on the
PCB, then use the machine screw
to attach the tab with a shakeproof
washer both under the screw head
and under the nut. Do it up tightly
since the screw conducts the current
when the crowbar activates. Then
solder the two pins.
Fit the 10nF capacitor and then
PUT1, bending its leads out with
pliers to suit the pad spacing. Push
it down as far as it will go before soldering and trimming the leads. Then
mount the 100µF capacitor, with its
longer (positive) lead towards the
left side of the PCB. Bend its leads
so that it lays down flat on the board
before soldering them.
Don’t fit a terminal block for
CON1 since we have limited clearance to fit the unit into the SemTest.
Instead, solder a red wire to HV, a
yellow wire to V+ and a black wire
to 0V. Make sure there are no stray
copper strands.
Wire the unit up to the SemTest as
shown in the main overlay diagram
(Fig.10). Trim each lead so that you
don’t have a lot of extra length. The
photos show the best place to fit it.
Once it’s wired up, slip the crowbar module into the heatshrink tubing and apply gentle heat. Make sure
there is no exposed metal when you
are finished. Some silicone sealant
can then be used to hold the unit
in place, so it doesn’t rattle around
inside the case.
Once the SemTest unit itself is
complete, the HV crowbar must now
be tested for correct operation, as
described in the main article.
ing with the diodes and zener diodes.
Make sure that these are all installed
the correct way around. The same goes
for transistors Q1 & Q2. Make sure Q1
is a BC337 and Q2 is a BC327. Note
that IC7, the metering voltage reference
IC, is in the same TO-92 package as Q1
and Q2 – be careful not to install it in
the wrong position.
Two devices come in TO-220 packages – REG1, the 7805 5V regulator and
Q3, the IRF540N switching Mosfet.
Construction and testing
siliconchip.com.au
May 2012 87
300k
V4.11+
240k
160k
4D
15
16
+Vdev
C ON3
8.0MHz
X1
ULN2803A
27pF
27pF
7D
1
2
IC 6
C
ON2
V5+
S
T
F
75k
+
–
2
1
IC 4
Q3
IRF540N
47 F
Q1
BC 337
Q2
BC 327
390k
sgV
10k
vedI
+
100nF
V5+
1nF
1k
39
470k
VR1
50k
15
16
+
–
1DEL
D9
4148
2
1
IC 5
V01,R
680
12k
5.1k
30
RLY1
10k
100k
PUT1
2N6027
10nF
100 F
220 F
REG1
7805
1000 F
1000 F
FER+
ZD2
V52
6V2
RLY7
10k
2.7M
2.4k
S2
V001
SET TEST VOLTS
V05
DNG
2102 ©
ULN2803A
100 F
0.27 5W
C ON4
7S
3.0k
10k
IC 3
LM358
3.9k
LK1
4004
D1
330 1W
SC R1
TYN816
© 2012
4004
D1
S1
C ON1
NOTE:
W IRE LINK
FITTED
FOR LK1
C ROW BAR
MODULE (IN
HEATSHRINK
SLEEVING)
D3
TP1
4148
VR2
TPG
D4
10k
SET 2.49V REF
4148
IC 7
LM336Z-2.5
100nF
LOW ER BOARD
R AB W
OR C EBAR
GATL OV H GI H
HIGH VOLTAGE
C ROW
DRA OB R E W OL
GND
GND
Vin
V+
100 1W
04105121
22
+HV
C ON1
ET ER C SI D
R OT CU D N O CI M E S
T E S T S ET
47 F 450V
TP4
V+
C ROW BAR
20k
TPG
470nF 630V
390k
GND
C ROW BAR
IC 1
34063
390k
PIC 16F877A
12130140
100nF
470nF 630V
390k
47 F 450V
Vgs
Idev
POW ER
Fig.10: follow this parts layout diagram to build the main (lower) PCB assembly. Use a socket for IC4 and take care to ensure that all semiconductors
and electrolytic capacitors are correctly orientated. Take care also when installing the three IDC headers – they must go in with their key-way slots
positioned as shown. The two switches are mounted directly on the PCB but be sure to use the specified switch for rotary switch S2 to ensure that its
control shaft is long enough (see text and panel).
300k
RLY8
TPVdev
9
10
33k 1W
33k 1W
TPG
RLY2
33k 1W
T1
75k
D2
UF4007
75k
!VH+
75k
1.5k 5W
1.6k
HV+
100
10k
100k
C ROW BAR
680
ZD1
4V7
100k
2.2k
10k
12V IN
WARNING! HIGH VOLTAGES (UP TO 600V DC) CAN BE PRESENT WHEN THE
CIRCUIT IS OPERATING. CHECK TO ENSURE THAT THE 47F 450V
CAPACITORS HAVE FULLY DISCHARGED BEFORE WORKING ON THE CIRCUIT.
2.2k
100k
100k
56
1.0k 1W
300k
0 /68
1.0k 1W
C OM
10k
C OIL
10k
10k
NO
10k
NC
10nF
3.0k
100k
560
NC
560
56
C OIL
10nF
NO
2.4k
C OMMON
+
88 Silicon Chip
+
100nF
04105121
siliconchip.com.au
The view shows the completed main board assembly before the HV
crowbar module is added. It carries the PIC microcontroller (IC4), the
power supply components and the test voltage selector switch (S2).
Both are mounted with their leads bent down by 90° at a distance of 6mm from their bodies, so they pass down through the
corresponding holes in the board to be soldered. Both devices
are mounted on standard 19mm-square U-shaped heatsinks
and secured using M3 x 10mm machine screws and nuts.
Having installed the semiconductors, install crystal X1.
It’s mounted just to the left of IC4’s socket. That done, install
the 3-pole 4-position rotary switch. This switch must have a
42mm long shaft and the one to use is a metric switch made
by Lorlin (CK1051). We sourced ours from Element14 (Cat.
112-3697).
IC5 & IC6 can then be soldered in place and IC1, IC3 & IC4
plugged into their respective sockets. The main PCB assembly
can then be completed by wiring the HV crowbar PCB to it,
as shown in Fig.10
Display PCB assembly
Fig.12 is the component overlay for this PCB. Begin by fitting the resistors. As before, two of these are shown with a
value of 0Ω/68Ω, to suit 6V or 12V mini SPDT relays: with a
6V relay, use a 68Ω resistor; for a 12V relay, use a wire link.
The seven trimpots can now go in. VR11 is a 10kΩ mini
horizontal type near relay RLY3. The remaining six multi-turn
trimpots have values of 5kΩ and 10kΩ; don’t mix them up.
VR10, the 10kΩ dual-gang pot, is wired with short flying
leads and will be bolted to the front panel later. Note that
it should have its shaft cut to 15mm long, to suit the knob.
Follow with the two capacitors and the relays. Make sure
the two mini-DIL reed relays are correctly orientated, as you
would for DIL ICs.
Now fit the semiconductors. There are four TO-92 devices:
transistors Q4 & Q5 and voltage references IC8 & IC9; don’t
siliconchip.com.au
UPPER SECTION
OF FERRITE
POT CORE
BOBBIN WITH WINDING
(10T OF 0.8mm DIAMETER
ENAMELLED COPPER WIRE
WITH END BROUGHT OUT.
THEN START OF 0.25mm DIA
ECW TWISTED TO IT, BEFORE
WINDING 4 x 20T LAYERS
OF SECONDARY. NOTE
THAT ALL FIVE LAYERS
SHOULD BE COVERED
WITH INSULATING TAPE)
FINISH (OF SECONDARY)
TAP (END OF PRIMARY,
START OF SECONDARY)
START (OF PRIMARY)
'GAP' WASHER OF 0.06mm
PLASTIC FILM
LOWER SECTION
OF FERRITE
POT CORE
(ASSEMBLY HELD TOGETHER & SECURED TO
PCB USING 25mm x M3 NYLON SCREW & NUT)
Fig.11: here are the winding details for the step-up
transformer (T1) on the main PCB. Note the “gap”
washer which is cut from 0.06mm plastic sheet.
May 2012 89
This view shows the assembled display PCB with the ZIF socket and potentiometer VR10 removed for clarity. Note
that this is a prototype board and there are some differences between this and the final version depicted in Fig.12.
mix them up. Don’t fit LED1 at this
stage; do it just before the display PCB
is attached to the front panel.
The three DIL pin headers CON5,
CON6 and CON7 are next, followed
by the 8-pin DIL socket for IC2. Then
fit the four PCB terminal pins near IC2.
Next comes the ZIF socket. It’s not
mounted directly on the board but
needs to be “jacked up” so that it will
protrude through the matching hole
in the front panel. The ZIF socket
also needs to clear the front panel by
almost 8mm, to allow its actuator lever to swing down into the horizontal
position.
Fig.13 shows how two 18-pin DIL
sockets, piggy-backed together, are
used to mount the ZIF socket. Most of
the “jacking up” is done by an 18-pin
DIL IC socket with long wire-wrap
tails. However, because the machined
clips of this type of socket are not able
to accept the rectangular pins of the
ZIF socket, we have to use a “production” type 18-pin DIL socket (having
bent sheet metal clips) between the
two, as an adaptor.
The ZIF socket is plugged into this
90 Silicon Chip
intermediate socket first and the two
are then plugged into the machinedclip socket. After this the 3-socket
assembly is held together using fillets
of epoxy adhesive – see Fig.13.
When the epoxy cement has cured
you can fit the whole ZIF socket assembly to the display PCB. Note that
the assembly should be installed with
the actuator lever towards the LCD
module position on the PCB.
Make sure also that the bottom of the
ZIF socket itself is exactly 18mm (or
19.5mm if you are using a PCB front
panel) above the top surface of the PCB
before you solder the 18 wire-wrap
pins of the bottom socket to the pads
on the PCB.
You can ensure this by using an
18mm-wide strip of stout cardboard
underneath the assembly as a temporary spacer. It’s best to initially tacksolder one pin at either end, then do a
final check of the spacing and vertical
positioning. This will allow you to
make any last-minute adjustments that
may be necessary before soldering the
remaining 16 pins.
The next step is to mount the LCD
module – see Fig.14. The connections
between this module and the PCB are
made via a 16-way section of SIL pin
header strip, which should be fitted
to the PCB (long pin sides uppermost)
before the module is attached. Don’t
solder its pins at this stage, though.
The module itself is mounted on the
PCB on two M3 x 6mm tapped Nylon
spacers. These are secured using M3 x
15mm machine screws which pass up
from under the board, with a flat Nylon
flat washer under each screw head.
The LCD module is then carefully
slipped down over the screws, with
the SIL strip pins passing up through
the matching holes at bottom left.
M3 nuts are then fitted to the top
ends of the screws to fasten the module
in position, after which the bottom
ends of the SIL strip pins are soldered
to the display PCB pads underneath.
Finally, their top ends are soldered to
the pads on the top of the LCD module.
Use a fine-tipped iron for this job
and solder as quickly as possible to
prevent heat damage.
Once the LCD module is in position, fit LED1 to the display board. It’s
siliconchip.com.au
IC9
S7
TEST
ON/OFF
TPG
LM336Z-2.5
10k
VR4
TP3
e4Q
RLY6
1k
100
Q5
BC549
BC559
+
RLY6
RLY5
Q4
4148
D8
IC8
4148
IC2
LM358 D7
D6
TP2
1k
e5Q
TP+
RLY4
RLY3
SET
2.49V
LM336Z-2.5
10k
4148
6.8k
VR3
SET
8.75V
(TP+ –2.49V)
D5
4148
VR5
4.7k
620
G
LCD CONT
5k
SET
+500 A
5k
VR8
4.7k
620
SET
-500 A
5k
VR7
SET
-100 A
68k
5k
VR6
68k
SET
+100 A
56k
VR11 10k
UPPER BOARD
220 F
COIL
COIL
4YLR
RLY5
15
16
S6
ENTER
RLY9
RLY15
S5
S4
UP
E
RLY11
S3
MENU
E
B
C
G
S
D
G
S
sgV
DOWN
G
K
A
G
K
K
A
K
A
SOCKET)
2 2 1 3 0 1 4 0(DUT 2 1 0 2 ©
R OT CUD N O CI MES ETER CSID
DRA O
B RE P P U T E S T S E T
ZIF1
vedV+
NO
NC
2
16
CON6
0 /68
1
sgV
11YLR
RLY10
+11.4V
RLY12
(VR10a
CONNECTIONS)
COIL
RLY14
RLY12
COIL
NC
COM
NO
RLY14
COIL
COM
15
COM
NC
vedI
NO
vedV+
RLY10
2
COIL
NC
vedI
NO
COM
1
RLY13
COMMON
14 13 12 11 10 9 8 7 6 5 4 3 2 1 16 15
COMMON
1M
22
16X2 LCD MODULE
NC
NO
NC
ALTRONICS
CON7
RLY9
COM
NO
COIL
14 13
2
1
+11.4V
COIL
saibI-/+
0 /68
Z-7013 (B/L)
RLY15
NO
NC
RLY3
RLY16
+Vdev
NC
COIL
(JAYCAR QP-5515 LCD MODULE)
RLY16
COM
NO
WARNING! HIGH VOLTAGES (UP TO 600V DC) CAN BE PRESENT ON THIS PCB WHEN THE CIRCUIT IS OPERATING.
+11.4V
COIL
22
LED1
10k 1W
12V
ZD3
10k 1W
(VR10b
CONNECTIONS)
VR10a/b
(2x10k)
12V
ZD4
CON5
10k 1W
9
10
10k 1W
1
2
RLY13
1k
siliconchip.com.au
NC
COM
4.7k
If you are working from a kit, the lid
NC
COM
120k
Front panel
NO
100
If you are building the SemTest
from a kit, the case will probably be
already laser-cut and screen printed.
If you are working from scratch, you
will need to download the drilling/
cutting diagrams from the April 2012
downloads section of the SILICON CHIP
website and print these out to use as
drilling templates.
Take care when you are cutting the
rectangular holes in the lid of the case
for the ZIF socket and the LCD window
because any curved or out-of-square
edges will be painfully obvious when
your SemTest is finished. The best approach is to first drill a series of 2.5mm
holes around the inside perimeter of
each rectangle and then use small
jeweller’s files to complete the job.
The easiest way to prepare the six
notch holes along the upper edge of
the righthand side of the case (and the
matching edge of that end of the lid)
is to first temporarily fit the lid to the
case. You can then drill the holes in
both at the same time, using a 2mm
drill to first make pilot holes and then
enlarging these holes with a 4mm drill.
4.7k
Preparing the case
NO
120k
The details of these are shown in
Fig.15. The two 16-way cables are cut
from 120mm lengths of ribbon, with
15mm at each end to loop through the
top of the IDC connector, leaving approximately 90mm of ribbon between
the connectors.
The 10-way cable is made from a
190mm length of ribbon, with 15mm
again used at each end for the connector loops. This leaves approximately
160mm of cable between the connectors. When you’re fitting the IDC
connectors to each end of the cables,
make sure you fit them with the orientation as shown in the circled details
in Fig.15.
56k
Making the ribbon cables
100nF
mounted at lower left, with its cathode
flat side to the left. At this stage just
tack-solder its leads temporarily to the
board pads, with the lower surface of
the LED body about 16mm above the
board. This will enable you to adjust
its final height above the board after
it’s attached to the front panel.
Now plug IC2 into its socket at lower
right. That completes the assembly of
this board.
Fig.12: the display (top) PCB assembly. This PCB carries the ZIF socket, the LCD
and most of the relays and is connected to the main board via IDC cables.
is likely to be already screen-printed
with the label. If not, you can purchase
a PCB dress panel from SILICON CHIP.
It is secured to the front panel with
the same screws which mount the
display PCB.
Cut a 70 x 25mm rectangle of clear
plastic sheet and fasten this to the lid,
behind the 51 x 16mm rectangular
cut-out for the LCD viewing window.
This will protect the LCD from dust
and moisture. The plastic sheet can
be fastened to the underside of the lid
using cellulose tape around its edges.
Now mount pushbutton switches
S3-S7 on the front panel. That done, fit
the four M3 x 25mm machine screws
which ultimately attach the front PCB
to the rear of the front panel.
As shown in Fig.14, each screw
May 2012 91
18-pin ZIF SOCKET
S4
CON4 on the main board. The unit is
now ready for testing.
BOX LID (FRONT PANEL)
Setting up
18-pin CLIP-TYPE
DIL IC SOCKET
18-pin MACHINED-CLIP
DIL IC SOCKET WITH
WIRE-WRAP TAILS
EPOXY
CEMENT
UPPER (DISPLAY) PCB
Fig.13: the ZIF socket is mounted via two 18-pin IC sockets, with the parts
piggy-backed together and secured using epoxy cement before the assembly
is installed (see text). Note that the bottom of the ZIF socket should be 18mm
above the display PCB (or 19.5mm if you are using a PCB front panel).
is fitted with an M3 x 15mm tapped
spacer. The screws and spacers should
be tightened as securely as you can,
without causing the screw head to
distort the dress front panel. An M3
nut is then added to each screw at the
end of each spacer to bring the effective
spacer length close to 18mm.
Next, solder “extension wires” to
the connection lugs on pushbutton
switches S3-S7. The extension wires
for these switches should all be made
from 0.5mm diameter tinned copper
wire, with their lengths staggered between about 40mm and 60mm as this
will make it easier to later pass them
through their matching holes in the
upper PCB.
You now need to solder some short
flying leads (about 50mm long) to the
terminals of dual-gang potentiometer
VR10. The other ends of these leads
can then be soldered to the PCB as
shown in Fig.12.
That done, temporarily stick the
back of the pot to the display board
with its shaft sticking up, ready to
pass through the front panel. You can
use some BluTac or double-sided tape
for this job.
The three IDC cables should now be
plugged into CON5, CON6 & CON7.
The next operation is a bit tricky,
because you have to dress each of the
extension wires from switches S3-S7
so they all go through their respective
holes in the display PCB as it is moved
up towards the rear of the front panel.
You also have to pass the body of the
ZIF socket (with its actuator lever
vertical) up through its cut-out in the
panel, and make sure that LED1 and
VR10 are lined up to pass through their
clearance holes in the front panel.
When you have managed to mate the
two together, with the PCB fitted onto
the ends of the four mounting screws,
you can add a further nut to each screw
to hold it all together. Tighten each nut
to complete the job.
Once it’s in position, solder all of
extension wires from switches S3-S7
to their pads on the underside of the
PCB. Be sure to trim the excess leads
after the wires are soldered.
The main board can now be mounted in the case but don’t fit the lid/
upper board assembly to the case just
yet. It can be stood up near-vertically
just in front of the case, with the front
panel buttons and LCD display quite
accessible.
Now plug the free ends of the three
ribbon cables into CON2, CON3 and
M3 x 25mm MACHINE SCREWS
M3 x 15mm
TAPPED SPACERS
M3 NUTS
M3 NUTS
Be careful when testing this device,
as high voltages (up to 600V DC) can
be present on both PCBs (see panel).
Start by setting the voltage selector
switch S2 to its 50V position, then
connect the SemTest to a 12V DC
plugpack rated at 900mA or more
and turn on power switch S1. There
should be no test devices plugged into
the ZIF socket as yet. You should see
this initial greeting message in the
LCD window:
SC Discrete Semi
conductor Tester
which should be replaced after a
couple of seconds with this message:
Press Menu Select
button to begin:
If you only see a clear window or
two lines of 16 black rectangles, it
probably means that the contrast trimpot VR11 needs adjustment. Adjust
VR11 in one direction or the other
until you see the messages displayed
clearly and with good contrast.
Once this has been done, you can
use your DMM to check the voltages
at the input and output pins of REG1
(at upper right on the main board, just
to the left of CON1). With the DMM’s
negative lead connected to the TPG
pin just below D4 on the same PCB,
you should get a reading of about
11.4V on REG1’s upper input pin and
a reading very close to 5.00V at its
lower output pin.
Finishing the set-up
Your SemTest is now ready for the
final setting-up adjustments. Do the
adjustments in this order:
• Adjust trimpot VR2, at lower right
on the main board, to set the PIC miBOX LID (FRONT PANEL)
16x2 LCD MODULE
M3 x 6mm TAPPED
NYLON SPACER
UPPER (DISPLAY) PCB
NYLON FLAT WASHERS
M3 x 15mm MACHINE SCREWS
16-WAY SECTION OF SIL PIN HEADER STRIP
USED TO MAKE INTER-BOARD CONNECTIONS
Fig.14: this diagram shows the mounting arrangement for the LCD module. It’s mounted on four M3 x 6mm tapped
Nylon spacers, with the holes along one edge mating with the pins of a 16-pin SIL header strip that’s soldered to the
display PCB. Secure the LCD module in place before soldering it to the header pins along the top. The PCB itself is
mounted on the box lid using M3 x 15mm spacers, with M3 nuts used to provide additional spacing.
92 Silicon Chip
siliconchip.com.au
90mm
TWO CABLES REQUIRED
cro’s ADC reference voltage to 2.490V.
It’s adjusted while monitoring the reference voltage with your DMM, across
terminal pins TP1 and TPG, just below
D4. This calibrates the SemTest ADC
module’s voltage and current measurement ranges.
• Adjust trimpots VR3 and VR4, at
lower right on the display board. VR3
sets the voltage drop across IC8 to
2.490V, while VR4 is used to set the
drop across IC9 to the same figure. IC8
is the voltage reference for the +IBIAS
current source, while IC9 does the
same job for the -IBIAS current source.
To do this, connect the DMM leads
between TP+ (+) and TP2 (-) and adjust
VR3 to get a reading of 2.490V. VR4 is
adjusted while monitoring the voltage
between test point pins TP3 (+) and
TPG (-) with your DMM, again to get
a reading of 2.490V.
These adjustments effectively set
the lowest current level (20µA) for
+IBIAS and -IBIAS.
The next four set-up adjustments set
the higher current settings for +IBIAS
and -IBIAS, using VR5, VR6, VR7 &
VR8. To do these adjustments, you
need to fit two short lengths of hookup
wire into two of the device lead positions on the ZIF socket, and then set up
the SemTest for four different device
tests. Here’s the procedure:
• Take two short lengths of insulated
wire with about 15mm of insulation at
each end stripped off. Then with the
ZIF socket’s actuator lever upright,
introduce one end of each wire into
the socket’s “B” and “E” lead holes for
a BJT. (It doesn’t matter which of the
two “E” holes you use).
• Push the socket’s actuator lever
down into the horizontal position, to
lock these temporary base and emitter
leads in place.
• Switch your DMM to read low DC
current levels (say 200µA to begin) and
connect its test leads to the two wire
leads: the “+” lead to the base wire and
120mm LENGTH OF 16-WAY IDC RIBBON CABLE
(15mm LOOP IN CONNECTOR AT EACH END)
190mm LENGTH OF 10-WAY IDC RIBBON CABLE
(15mm LOOP IN CONNECTOR AT EACH END)
ONE CABLE REQUIRED
160mm
Fig.15: here’s how to make up the IDC cables. Be sure to orientate the
headers with the locating spigots facing exactly as shown – they face
outwards on the 90mm cables and inwards on the 190mm cables.
the “-” lead to the emitter wire.
Now we need to negotiate SemTest’s
menu system to reach a device test setup which will allow us to measure the
various IBIAS levels using the DMM.
Apply power and press the MENU
SELECT button for half a second or
so. You should then see the opening
device selection display:
Device to Test: ˄
1:Diode/Zener ˅
In case you’re wondering, those “^”
and “v” symbols at the right-hand ends
of the lines are meant to remind you
that you can scroll up or down through
a sequence of menu choices, using the
UP or DOWN buttons.
For the first of these IBIAS adjustments, we actually want to select some
BJT (NPN) tests, so press either of these
buttons briefly a number of times, until
you see this display:
Device to Test: ˄
3:NPN bipolar ˅
Since that’s the type of device we
want to set up for (even though there
is no actual device plugged into the
ZIF socket), confirm this by pressing
the ENTER button. This will cause the
display to change into:
Test parameter:˄
BVcbo (e o/c) ˅
As before, note the symbols at far
right on the display, indicating as
before that other tests can be selected
using the UP and DOWN buttons. So
press either of these buttons briefly a
few times until you see this display:
Test parameter:˄
hFE (Ib=20μA) ˅
This is the first test we want to set
up for in order to make these set-up
adjustments, so press the ENTER
button to confirm it. The display will
then become:
NPN bipolar:
hFE(Ib20μA)=0000
Now, after checking that you have
set voltage selector switch S2 to its
50V position, press the TEST ON/OFF
button to turn on the DC-DC converter
and take a measurement. LED1 should
be on, to indicate that the DC-DC converter is operating and providing test
WARNING: SHOCK HAZARD!
The DC-DC step-up converter used in this project can generate high voltages (up to 600V DC) and can also supply significant
current. As a result, it’s capable of delivering a nasty electric shock and there are some situations where such a shock could be
potentially lethal.
For this reason, DO NOT touch any part of the circuit while it is operating, particularly around transformer T1, diode D1 and
the two 47μF 450V electrolytic capacitors on the main (lower) PCB. Note, however, that high voltages can also be applied to the
display board (via CON6) during operation, so it’s not safe to touch certain parts on this board either. Exercise caution if testing
the unit with the lid opened and always allow time for the 47μF capacitors to discharge before working on the circuit.
Note also that high voltages (up to 600V DC) can be present on the component leads when testing for high-voltage breakdown.
DO NOT touch any leads while testing is in progress.
siliconchip.com.au
May 2012 93
Sourcing The Rotary Switch
As mentioned in the article, the 3-pole
4-position rotary switch (S2) must have
a 42mm-long control shaft, so that when
the case is assembled, it passes through
the clearance holes in the front panel
with enough length left over to attach the
control knob.
A Lorlin CK1051 switch is suitable
and this can be sourced from Element14
(Cat. 112-3697). Note that the shafts on
the switches usually available from the kit
suppliers will be too short for this project.
voltage. The LCD display will also
change, but don’t take much notice
of the hFE reading because there is no
transistor connected at present (it will
probably show an hFE reading of either
“00” or “01”).
Your DMM should now show a figure very close to 20.0µA (the default/
lowest IBIAS level).
Now press the TEST ON/OFF button
again, and hold it down for a second or
so until LED1 goes out, indicating that
the DC-DC converter has been turned
off. The LCD display will also return to
its “Press MenuSelect” message, ready
for another test. And when you press
the MENU SELECT button, you’ll find
that the SemTest has “remembered”
that you were testing an NPN bipolar
device and will offer the same device
test again:
Device to Test:˄
3:NPN bipolar ˅
Confirm this by pressing the ENTER button. Then use either the UP
or DOWN buttons until you get this
display:
Test parameter:˄
hFE (Ib=100μA) ˅
Press the ENTER button to confirm
and finally press the TEST ON/OFF
button again to turn on the DC-DC
converter and take a measurement. As
before though, don’t worry about the
hFE measurement on the LCD display
– pay attention to what the DMM is
showing, because this will be reading
the actual bias current. This should
be close to 100.0µA. Now adjust VR6
with a small screwdriver until it reads
100.0µA.
Once that’s done, press and hold
down the TEST ON/OFF button until
LED1 goes off. Then press the MENU
SELECT and ENTER buttons and then
UP or DOWN to get:
Test parameter:˄
hFE (Ib=500μA) ˅
94 Silicon Chip
Press ENTER to confirm, set your
DMM is set to read to over 500µA,
then press the TEST ON/OFF button.
Your DMM should now read close to
500µA. Adjust VR5 to get that exact
figure. Press the TEST ON/OFF button
once again until LED1 goes off.
That completes the two adjustments
for the +IBIAS current levels. Those
for the -IBIAS levels are next on the
list. This time we use the tests for an
PNP bipolar instead of an NPN and we
need to reverse the connections to the
DMM test leads.
Press MENU SELECT again and then
press the UP button once, to get:
Device to Test:˄
4:PNP bipolar ˅
Press ENTER to confirm and press
either UP or DOWN to select the
“hFE (IB=20µA)” test. Press ENTER
to confirm and then press TEST ON/
OFF. Your DMM should show close to
20.0µA, confirming the default/lowest
-IBIAS level. Now press and hold down
TEST ON/OFF to stop this test.
Now press MENU SELECT again and
you’ll find that the PNP bipolar tests
are still being offered. Press ENTER
to confirm and then the UP or DOWN
buttons until you get:
Test parameter:˄
hFE (Ib=100μA) ˅
Confirm this by pressing ENTER and
follow by pressing TEST ON/OFF to
start the test. Your DMM should now
be reading close to 100.0µA. Adjust
trimpot VR7 to bring the reading as
close as possible to that figure, then
press TEST ON/OFF to stop the test.
Set the DMM to read more than
500µA and then press MENU SELECT,
ENTER and the UP or DOWN buttons
until you have selected:
Test parameter:˄
hFE (Ib=500μA) ˅
Press ENTER and TEST ON/OFF
again and confirm that the DMM reads
close to 500µA. Adjust VR8 to obtain
that exact figure, then press TEST ON/
OFF again and you have completed
all the setting-up adjustments for the
SemTest’s IBIAS current levels.
One more adjustment remains:
using trimpot VR1 to set the DC-DC
converter output voltage levels. To do
this, check that S2 is set to 50V. Then
press MENU SELECT and UP or DOWN
until you get:
Device to Test:˄
7:SCR ˅
Press ENTER to confirm and either
UP or DOWN until you get:
Test parameter:˄
Vak on (OPV) ˅
Now press ENTER and TEST ON/
OFF. The second line of the LCD
should now read something like this:
Vak(OPV) = 49.6V
Adjust VR1 (just above the centre of
the main board) until the LCD reading
changes to:
Vak(OPV) = 50.0V
Finally, press the TEST ON/OFF
button once. This completes all the
set-up adjustments.
Final assembly
The front panel assembly can now
be lowered down onto the case. Make
sure that the three ribbon cables are
folded neatly into the space above the
lower PCB and not caught between
the edges of the case or lid. Fasten the
case together with four M4 screws into
the corner holes, then fit the knobs to
the rotary switch and the pot and the
assembly is complete.
Testing the HV crowbar
It’s now necessary to check that the
HV crowbar circuit is working correctly. To do this, power up the unit,
wait a few seconds and then press
the Menu Select button. You will get
a display like this:
Device to Test: ˄
1:Diode/Zener ˅
Press Enter and then the Up button.
The display will then show:
Test parameter:^
Irev(OPV) ˅
Press Enter again. Set the Device
Operating Voltage to 25V, using the
right-hand knob. Then press the Test
On/Off button to start the test.
Now carefully measure the voltage
across the top and bottom A & K terminals in the Diodes & LEDs section of
the test socket. You should get a reading close to 25V. If it’s much lower (say,
12V) then either the crowbar circuit
has triggered prematurely or there is
a fault in the DC/DC converter circuit.
You will need to switch off, open up
the unit and check the crowbar and
converter circuits for faults such as
incorrectly orientated components.
If you get a much higher reading
than 25V, there is a problem with the
DC/DC converter section. Switch off
and measure the voltage across the A
& K terminals until it drops to a safe
level. Then open the unit up and look
for the source of the problem.
Assuming all is well, press the Test
siliconchip.com.au
On/Off button to terminate the test.
You can now do a high-voltage test.
The procedure is similar to before except you want to do an IREV(BV) test.
So when you get to this stage:
Device to Test: ˄
1:Diode/Zener ˅
press enter twice and start the test.
Carefully measure the voltage across
the A & K terminals again. It should
be several hundred volts and it will
rise to close to 600V after a number
of seconds. Now press the Test On/Off
button again to terminate the test while
monitoring the voltage between the A
& K terminals. It should immediately
fall to just a few volts when the test is
terminated.
If it remains high and only decreases
slowly, the crowbar has failed to operate and you will need to wait for the
capacitors to discharge before opening
the unit up and checking for faults.
If the crowbar is not working (eg, if
it fails), a warning will be displayed on
the LCD immediately after performing
a high-voltage test. This indicates that
there is still a high voltage present at
the test socket. If you get this warning
then you should open the unit up and
repair the crowbar circuit.
Using the SemTest
The SemTest is used as follows:
STEP 1: place DUT in ZIF socket and
switch on.
STEP 2: Press Menu Select.
STEP 3: Use Up/Down buttons to select
device type and press Enter.
STEP 4: Use Up/Down buttons to select
test and press Enter.
STEP 5: For OPV tests, use righthand
knob to select test voltage.
STEP 6: Press Test On/Off to start test
(red LED on) and read off result.
STEP 7: Press Test On/Off again to finish test (red LED out).
STEP 8: check that the red LED is out
and that there is no high voltage warning on the LCD before removing DUT.
Exercise caution when testing components for high-voltage breakdown.
Up to 600V DC is present on the device
leads during such tests, so be careful
not to touch them!
The biggest problem in using the
SemTest is knowing the various lead
configurations of the devices it can
test. To that end, we have prepared a
connections chart showing commonly
used diodes, LEDs, BJTs, Mosfets,
SCRs and PUTs. It can be stuck on a
wall or to the underside of the SemTest
siliconchip.com.au
This view inside the completed prototype shows how it all goes together.
The two PCB assemblies are mounted in their respective case halves on
spacers and joined together via the three IDC header cables.
case for easy reference.
For less common devices, you’ll
need to look up the connections in a
data book or by downloading a data
sheet from the manufacturer’s website.
Finally, here are a few tips to guide
you when you’re doing some of the
more specific tests:
• When reading the forward voltage drop VF of a diode or LED or the
voltage drop VAK of an SCR when it’s
conducting, be aware that the accuracy
of this measurement is not very high
due to measuring circuit limitations.
So if you need to make really accurate
measurements of VF or VAK, you’ll
need to use an external DMM with its
leads connected across the device’s
“A” and “K” leads.
Remember that during the same
tests, it’s OK to increase the device
operating voltage to a higher setting in
order to see the voltage drop at higher
current levels.
• When you want to measure the hFE
of a BJT, start on the setting with the
lowest IBIAS level (ie, 20µA), because
this is the setting with the highest hFE
range. Only swing down to one of the
higher IBIAS settings if the hFE reading
you get is very low (ie, below 300).
This should only be necessary with
medium-to-higher power devices,
which often have their “peak” hFE at
higher currents.
• When you want to measure the IDS
vs VGS characteristic of a Mosfet to
get an idea of its transconductance or
“gm”, start by selecting the highest
device operating voltage which will
not exceed the device’s VDS ratings.
That’s because the VGS bias voltage
(adjusted via VR10) is derived from
the actual device operating voltage,
which inevitably tends to drop once
the device begins to draw drain-source
current (due to voltage drop in the
current limiting resistors).
If you don’t set the switch for a
reasonably high voltage to start with,
you’ll find that it won’t be possible
to provide much VGS once the device
starts to conduct.
Actually, although you need to set
the operating voltage within the device
ratings when you start this test, it’s OK
to increase the setting to 100V during
the test itself, if you need to do so in
order to achieve a higher VGS.
This won’t cause any problems if
you only increase the voltage setting
SC
once the device is conducting.
May 2012 95
Ultra-LD Mk.3 Amplifier
Tweaks & Performance
By NICHOLAS VINEN
Finally, we present the specifications for the new Ultra-LD Mk.3
Amplifier along with a couple of minor tweaks to the design to
maximise its performance.
T
HESE FIGURES and graphs show
the performance of the complete
Ultra-LD Mk.3 amplifier. The test
signal source was set to 2V RMS and
connected via the RCA inputs of channel 1 on the rear panel of the unit.
The performance was measured at the
speaker terminals on the rear of the
unit, with a resistive load connected
via 1m of twin lead. The volume control was set to deliver 100W into 8Ω
and 200W into 4Ω with the 2V RMS
input signal.
The overall performance of this
amplifier is much better than the vast
majority of commercial amplifiers,
even expensive models sold as “high
fidelity”. Distortion figures for commercial units are often quite vague;
those that do provide graphs typically
show quite a dramatic rise in distortion
above 1kHz. As you can see from the
graphs published here, our Ultra-LD
Mk.3 retains the low high-frequency
distortion characteristics of the individual modules featured in the July
2011 issue.. The signal-to-noise ratio
is also very good.
The left/right channel performance
differs, despite the fact that the amplifier modules are identical. This is
because one module is closer to the
power transformer than the other;
we purposefully arranged it this way
because otherwise, the transformer
would be close to the sensitive input
circuitry of the right-hand module and
that would be worse.
Specifications
Continuous power, both channels driven (THD+N < 0.1%): 100W into 8Ω; 135W
into 4Ω
Music power, both channels driven: ~150W into 8Ω; ~200W into 4Ω
Total harmonic distortion plus noise: <0.0025%, 1kHz, 20Hz-22kHz bandwidth,
90W (both channels driven)
Signal-to-noise ratio: -109dB (left channel), -115dB (right channel) with respect to
90W into 8Ω
Frequency response: +0,-0.3dB (8Ω), +0,-1.3dB (4Ω) 20Hz-20kHz
Channel separation: approximately 50dB, 4Ω and 8Ω, both channels
96 Silicon Chip
The performance of the right
channel is almost as good as that of
the module by itself, with very low
distortion up to 20kHz – see the red
and mauve traces in Fig.6. This graph
was produced using a wide analyser
bandwidth of 20Hz-80kHz, so that it
includes the first and second harmonics of the higher frequency test signals.
Despite this, distortion is down around
0.001% at 1kHz and below 0.004%
at 20kHz.
The left channel results are slightly
worse for the reasons explained earlier,
but still very good. The higher distortion for the left channel with both
channels driven is due to the increased
magnetic field around the transformer
as it delivers nearly twice the current.
Channel separation is virtually flat
with frequency and insensitive to load
impedance at -50dB.
Further refinements
We made a couple of additional
refinements to the amplifier design
in order to achieve this level of performance, not described in the previous articles. Both changes reduce the
amount of ripple from the power supply that couples into the signal earth.
First, we changed the 10Ω 0.25W
siliconchip.com.au
0.1
THD+N vs Frequency, 90W, 20Hz-80kHz Bandwidth, 8
04/10/12 14:32:38
+1
04/10/12 14:59:31
0
0.05
Left channel, both driven
Left channel, one driven
Right channel, both driven
Right channel, one driven
0.02
-1
8
4
-2
0.01
Relative Amplitude (dBr)
Total Harmonic Distortion + Noise (%)
Frequency Response, 10W, 4 & 8both channels identical)
0.005
0.002
0.001
-3
-4
-5
-6
-7
0.0005
-8
0.0002
-9
0.0001
20
50
100
200
500
1k
2k
5k
10k
-10
10
20k
20
50
100
200
500
Frequency (Hz)
Fig.6: distortion versus frequency into an 8Ω load at
90W per channel. The right channel has lower distortion
than the left channel due to its proximity to the mains
transformer and the hum/buzz coupling that results.
Measurements with a 400Hz high-pass filter show the
performance of the two channels is virtually identical if
hum is ignored.
0.1
THD+N vs Power, 1kHz, 20Hz-22kHz Bandwidth, 8
04/10/12 14:43:57
0.1
10k
THD+N vs Power, 1kHz, 20Hz-22kHz Bandwidth, 4
0.02
Total Harmonic Distortion + Noise (%)
Total Harmonic Distortion + Noise (%)
5k
20k
50k 100k
04/10/12 14:49:53
0.05
Left channel, both driven
Left channel, one driven
Right channel, both driven
Right channel, one driven
0.01
0.005
0.002
0.001
0.0005
Left channel, both driven
Left channel, one driven
Right channel, both driven
Right channel, one driven
0.01
0.005
0.002
0.001
0.0005
0.0002
0.0002
0.0001
0.5
2k
Fig.7: frequency response of the complete amplifier
which is virtually flat from 20Hz to 20kHz. Very little
bass roll-off is evident. The high-frequency roll-off is due
to the output RLC filter, which is necessary to isolate
the amplifier from the speakers and cabling, ensuring
stability. As a result, the 4Ω high-frequency roll-off is
significantly higher than for 8Ω.
0.05
0.02
1k
Frequency (Hz)
1
2
5
10
20
50
100
200 300
0.0001
0.5
1
Fig.8: distortion versus power for 8Ω loads. Again, the
right channel is noticeably lower in distortion than the left
channel. Note that the power supply limits the available
continuous power when driving both channels to around
100W while around 135W can be delivered if a single
channel is driven. Music power is about 150W even if both
channels are driven.
resistor on each amplifier module to
47Ω. This resistor is located to the right
of the RCA input socket and connects
the signal ground to the power supply
ground. If you have already built the
modules, it’s simply a matter of clipping off these resistors, removing the
lead stubs, clearing the holes with a
solder sucker and soldering the new
resistors in place.
The second change is in the amplifier power supply wiring. While it’s
siliconchip.com.au
2
5
10
20
50
100
200 300
Power (W)
Power (W)
Fig.9: distortion versus power for 4Ω loads. As is typical
for power amplifiers, the distortion is somewhat higher
when driving 4Ω loads than 8Ω loads, partly due to
the increased noise that results from the lower load
impedance. The power delivered is higher than for 8Ω,
with around 135W per channel available when both are
driven and about 200W with a single channel driven.
convenient to wire up each amplifier
module to its own terminal on the
power supply board, this results in a
relatively high ground resistance between the two modules. Performance
is improved if both are wired to the
same supply terminal, with a longer
cable running from the left-hand
module to the supply terminal on the
right side.
This requires one 3-wire supply lead
to be longer than previously speci-
fied, around 150mm. This is then run
around the capacitors at the bottom of
the power supply module to reach the
power connector for the left channel
amplifier.
Keep it as short as possible and use
heavy-duty wire as lower resistance
means lower distortion. Twist the
leads together before plugging it into
the connector on the module.
With these changes, your amplifier
SC
will give the best performance.
May 2012 97
Vintage Radio
By Rodney Champness, VK3UG
Breville 730 Dual-Wave
5-Valve Receiver
Manufactured in 1948, the Breville 730 tabletop receiver was housed in an attractive
timber cabinet and covered both the broadcast
and shortwave bands. It featured a wide
audio response and this, coupled with a large
loudspeaker, gave very good performance.
T
HIS 1948 BREVILLE 730 receiver
was obtained by a friend of mine
(Marc) in almost original condition.
In fact, it’s quite rare to come across a
set as original as this one, as most sets
have had some routine servicing and
parts replacement during their life.
Hopefully, any work that has been
done on a set will have been carried
out by a competent serviceman. An
incompetent servicemen or hobbyist
can leave a set with more faults than
it started out with and can sometimes
98 Silicon Chip
even destroy hard-to-get parts.
In the case of Marc’s Breville 730,
the only evidence of any service work
was on the band-selection indicator. In
fact, the condition of this 65-year-old
set is so good that it has obviously
been used in a lounge room for most of
its life. It had eventually failed when
the ECH35 converter developed a
short circuit (as confirmed by a valve
tester), after which it had been carefully stored away.
As a result, virtually no damage
has occurred to either the cabinet or
chassis, other than the normal ravages
of time.
Circuit details
The Breville 730 has a conventional
superheterodyne circuit that’s similar
to many other receivers of the era.
However, it does have some features
which, although not unique, are not
seen in many other receivers.
Fig.1 shows the circuit details. The
signal from the antenna is fed to an
input tuned circuit and the position
of the band-change switch determines
whether shortwave (6-18MHz) or
broadcast band tuning is selected.
As shown, the primary of the
shortwave antenna tuned circuit (top)
is in series with the primary of the
broadcast-band antenna tuned circuit.
Capacitor C2, a 100pF capacitor across
the broadcast-band coil, performs two
tasks: (1) it acts as a low impedance
to earth for the bottom end of the
shortwave antenna primary and (2)
it tunes the primary winding of the
broadcast-band coil to below the lowest frequency on this band.
This technique enhances the performance at the low-frequency end of
the broadcast band.
Note that the primary of the shortwave antenna coil has little effect on
the performance of the broadcast-band
antenna tuned circuit and may even
boost its performance slightly. This
circuit works well and simplifies the
band switching.
Converter stage
The selected output from the antenna tuned circuit stage is applied
to the signal grid of an ECH33/35.
This functions as a converter or local
oscillator stage.
In operation, the local oscillator
tuned circuits are also switched to suit
the selected band (either broadcast or
shortwave). Note, however, that there
is an error in the circuit diagram regarding the oscillator switch position
siliconchip.com.au
Fig.1: the circuit is a fairly conventional 5-valve superhet design,
although the 3-coil first IF transformer is somewhat unusual. The
IF stage is tuned to 446kHz and has switch-selectable bandwidth.
– the antenna switch is shown in the
broadcast position, while the oscillator
switch (immediately following C5) is
shown in the shortwave position.
Capacitor C6 acts as both a padder
and a phase change network when the
broadcast band is selected, to provide
positive feedback for the oscillator. It
also provides an earth return for the
shortwave oscillator primary feedback
winding.
By contrast, the shortwave oscillator
tuned circuit does not have a padder
capacitor attached to its tuned winding. Because of the relatively small difference between the oscillator and the
signal frequencies on shortwave, some
manufacturers left this component
out. Note also that double-spotting
or image reception is quite common
on shortwave receivers having no RF
stage and a 455kHz IF (intermediate
frequency).
Once the ECH35 has converted the
tuned RF signal to a 446kHz IF (not
455kHz as we normally expect), it
is applied to the first IF transformer.
This transformer is different to most
as it has three windings. The primary
is tuned to 446kHz and so is the secondary when the tone control (immesiliconchip.com.au
diately below it on the circuit) is in
its centre (normal) and bass positions.
By contrast, when the tone control
is switched to the wide range position, the third coil is switched into
circuit, in series with the secondary
tuned circuit.
In practice, I suspect that the secondary of this IF transformer is detuned to give a broad response through
the IF strip. In addition, I suspect that
the third coil is coupled to the first
tuned circuit so that the combination
of the primary and secondary tuned
circuits also broadens the response
(with a dip in the centre), so that the
receiver has an audio bandwidth of
up to 10kHz. This, combined with the
set’s large loudspeaker, would result in
good quality audio although it should
be noted that AM broadcast stations
later restricted their audio bandwidth.
Following the first IF transformer,
the signal is fed to a 6U7G IF amplifier stage and the resulting signal then
applied to the second IF transformer.
The IF signal is then fed to the detector diode in a 6G8G detector, AGC and
audio amplifier valve.
From there, the detected audio
signal is fed via R7 and C13 to vol-
ume control R12 and then to the grid
of the 6G8G. This circuit technique
enabled Breville to overcome the oftexperienced problem of “scratchy”
volume controls, caused when DC
from the detector is applied directly
across them.
Note that most radios use a triode
as the first audio amplifier but this set
uses a 6G8G pentode for additional
audio gain. The output from this stage
appears at the anode and is applied
to the grid of a 6V6GT audio output
valve. This in turn drives a speaker
transformer and an 8-inch (200mm)
loudspeaker.
In addition, the audio on the plate
of the 6V6GT is sampled via an RC
network and fed to the 6G8G’s cathode
to provide tone control and negative
feedback.
Record player terminals
The receiver is equipped with terminals which allow a record player (PU)
to be connected. However, this really
doesn’t work that well because there’s
no way of turning off the RF section of
the set when records are being played.
A combination of the latest broadcast
episode of “Biggles” and a recording
May 2012 99
The chassis of the Breville 730 was in quite good order although some corrosion was evident, especially on
the power transformer cover and at the top of the tuning gang.
of Tommy Dorsey playing over the top
of each other would hardly have been
satisfactory!
A simple switch would have solved
this problem, with one pole used to
switch the HT (high-tension) rail to
the RF stages on or off and another
pole to switch the input to the audio
amplifier between radio and turntable.
AGC & power supply
The automatic gain control (AGC)
signal is obtained from the plate of the
6U7G and is applied to the AGC diode
in the 6G8G. This diode is normally
biased off, as its anode is 1.5V negative with respect to the cathode of this
valve. As a result, it will not generate
any AGC voltage until the incoming
signal exceeds 1.5V.
This delayed AGC signal is applied
to both the converter (ECH35) and IF
amplifier (6U7G) stages. Both these
valves share a common cathode resistor (R2) and 2.3V of bias is obtained
before the AGC voltage is applied.
The power supply is standard for the
era and uses a power transformer plus
a 5Y3GT rectifier. The transformer’s
primary is tapped at 220V, 240V and
260V, while the secondaries consist of
a 6.3V winding for the heaters and dial
lamps, a 5V winding for the 5Y3GT’s
filament and a centre-tapped 270V
per side winding for the high tension
(HT) supply.
100 Silicon Chip
The output of the rectifier is filtered
using an 8µF electrolytic capacitor
(C27), a 12H (Henry) choke and a following 16µF electrolytic (C26).
Cabinet restoration
As mentioned earlier, the cabinet
was in quite reasonable condition.
However, as antique dealers have often
pointed out, timber items stored in
very dry environments can develop
cracks and this cabinet was no exception.
These cracks and splits were carefully repaired using an epoxy adhesive
(Araldite). And because the timber
was so dry, Marc applied linseed oil
to the inside of the cabinet using a
paintbrush. The outside of the cabinet
also received attention, with linseed
oil applied sparingly using a cloth.
The revitalised cabinet now looks
quite good despite the minimal attention paid to it. Further restoration
was not considered desirable in the
interests of originality.
The original speaker cloth was in
poor condition and so this was replaced with some open-weave brown
cloth obtained at a haberdashery. It
looks authentic even though it isn’t
genuine speaker cloth. In addition,
new rubber buffers were fitted to the
bottom of the cabinet, replacing the
old ones which had either perished,
become hard or had gone missing.
Finally, the cabinet features a celluloid strip which is mounted behind
the various controls and which carries the control legends. Although
yellowed with time, it is still original
and quite legible. These strips usually
deteriorate and fall to pieces over time
but this one is good for many years yet.
The control markings on the strip
are (left to right): On-Off-Volume; Tone
– Bass, Normal, Wide Range; Station
Selection; and Wave Change SW/BC.
Circuit restoration
The chassis is quite easily removed
from the cabinet. This involves removing the four control knobs and the
dial-light assembly, followed by the
four screws underneath the cabinet
which secure the chassis in place. One
of the dial lamps had to be replaced,
after which the inside of this assembly
was repainted white to ensure good
reflectivity.
One problem with many sets is that
the dial-scale is left behind (ie, still
attached inside the cabinet) when the
chassis is removed. The Breville 730
is no different in this regard but the
redeeming feature of this set is that the
alignment frequencies are marked on
the edge of the dial scale, along with
the position of the dial pointer when
the tuning gang is closed.
That certainly makes it easier to
get the dial pointer lined up with the
siliconchip.com.au
station markings correctly when the
chassis is reinstalled.
Once the chassis had been removed,
Marc could immediately see that some
of the wiring was in need of replacement as the rubber insulation had
perished. This particularly applied to
the dial-lamp leads as the insulation
had actually fallen off and the wires
were shorting.
Closer inspection of the wiring revealed several other leads that were
shorting due to perished rubber insulation. These leads were all replaced,
after which the dial-light supply line
was isolated and the valves removed.
This was necessary to allow highvoltage tests on the power supply, to
confirm that it was in a safe condition.
First, the insulation resistance between the mains and chassis and other
windings of the transformer was tested
using a high-voltage insulation tester.
These were all found to be in good
order, with over 200MΩ of resistance
in each case.
That done, the old power cord was
replaced with a new 3-core cable. This
was securely anchored to the chassis
using a cord clamp.
Marc then tested and/or replaced
a number of parts that are known
to cause problems. In particular, all
the paper capacitors were replaced
with modern polyester types, while
the electrolytic capacitors were also
siliconchip.com.au
This under-chassis view shows the receiver after restoration. The original paper
and electrolytic capacitors were all replaced, along with some of the wiring.
replaced due to their age and the fact
that they were visibly leaking. Several
resistors were also found to be out of
tolerance and were replaced.
The loudspeaker was the next on
the list. It had developed a number
of cracks along the speaker cone edge
and these were repaired with Selleys
“Quik Grip”.
Testing the valves
Marc’s next step was to use his
valve tester to check all the valves in
the receiver. All tested OK except the
ECH35 converter, the tester indicating
a short circuit inside this valve. This
would have completely stopped the
receiver from working and is probably what caused the original owner
to retire the set.
Marc had a working ECH35 which
could replace the dead ECH35 but its
conductive red paint shield (which
Philips call “metallisation”) had been
damaged. As a result, he decided to
make an experimental shield to replace the damaged one.
A little investigation showed that
the wire contacting the red shield and
the earth pin in the valve plug was
intact and accessible. A thin strand
of wire was therefore soldered to this
earth wire (without cracking the valve
envelope) and then spiral-wound
around the valve envelope. Some
“Wire Glue” (available from Jaycar,
Cat. NM-2831) was then applied to the
envelope to secure it in place.
If access to the earth wire is not
practical, a thin wire can be soldered
to the earth pin of the valve and then
extended up and wound around the
May 2012 101
This view shows the components associated with the band-change switch. The
two coils associated with the input tuned circuits are clearly visible.
This slide assembly is controlled by
the band-switch and indicates which
band has been selected.
envelope. Another possible earthing
shield can be seen at www.andersonproducts.com. The sales information
says that it is “carbon blended in a
non-toxic binder”.
Repairing the band indicator
The band-change switch has a lever
off to one side of the control shaft and
this controls a spring-loaded slide assembly via a length of dial cord. This
slide assembly has two small labels
which are alternatively visible through
a clear window on the righthand side
of the dial scale and indicate the band
selection (ie, broadcast or shortwave).
AWA used a somewhat similar idea in
their 7-band receivers of the same era.
102 Silicon Chip
This slide assembly wasn’t working
in the old Breville 730 as the control
cord had broken. It had been replaced
during the life of the set with single
conductor tie wire instead of dial cord
but this wire had eventually fractured
at the eyelet. Re-stringing the assembly
with dial cord soon got it working
again.
The tuning gang was also a little
worse for wear so it was the next job
on the list. First, a small hand blower
was used to remove the dust that had
accumulated between the plates. This
revealed that some of the plates had
corrosion on them, so these were carefully cleaned by pushing some fine
emery paper between them.
The chassis itself was in quite good
order and was simply cleaned using
the blower and a small brush.
Testing
Having fixed all the obvious faults
in the set, Marc then decided to power
the set up to see if there were any other
faults in the circuit. As it turned out,
the set started up normally and stations could be clearly heard. A quick
check then revealed that all the voltages were normal and no components
showed any obvious signs of distress.
Even at this early stage, the set’s performance was quite good and tweaking both the antenna and oscillator
circuits made it even better. In fact,
its shortwave performance is better
than average for a set of this calibre.
However, there were a couple of other
issues to be dealt with. One dial globe
was dead and more importantly, it
was obvious that the volume control
pot was worn out and would have to
be replaced.
The IF alignment were then checked
using a signal generator and a frequency counter (to adjust the signal
generator exactly to frequency). Because Marc had no information on
adjusting the first IF transformer with
its three windings, he decided to
proceed with the tone control in the
“Normal” position. Before adjusting
anything though, each IF transformer
was marked so that he could easily
return it to its original position should
his alignment technique with the
uncommon 3-winding IF transformer
go awry.
As it turned out, the alignment went
smoothly and the first IF transformer
was easy to adjust in the standard selectivity position (ie, Normal).
New volume control
With the set now performing well,
Marc decided to replace the worn-out
on/off-volume control. Unfortunately,
he was unable to obtain a direct replacement with a long shaft, so he
was forced to use one with a splined
shaft and make up an extension shaft
on a lathe.
This proved to be a complete success and the new volume control
worked smoothly, without crackles.
The chassis was then reinstalled in
its cabinet and the restoration was
complete.
Summary
Breville produced many fine radios
and the model 730 was one such set.
It performs well and the broad response of the IF amplifier stages (when
switched to “Wide Range”) means that
the set was able to reproduce a wider
range of audio frequencies than most
other similar sets.
The set is also easy to service, with
all parts readily accessible. However,
the inability to isolate the RF section
when a turntable is connected to the
audio amplifier section is a rather
puzzling omission, especially since it
would have been so easy to do. All that
would have been required is an extra
position on the band switch, which
could then have been labelled “Short
Wave”, “Broadcast” and “Gram”.
In summary, the Breville 730 is an
excellent receiver with many interesting features and is a worthy addition
to Marc’s collection.
SC
siliconchip.com.au
WANT TO SAVE 10%? S
C (PRINT EDITION)
AUTOMATICALLY QUALIFY FOR
REFERENCE $ave SUBSCRIBERS*
CHIP BOOKSHOP 10% A 10% DISCOUNT ON ALL BOOK PURCHASES!
SILICON
ILICON
HIP
(*Does not apply to website orders)
SELF ON AUDIO
PROGRAMMING and CUSTOMIZING THE
PICAXE By David Lincoln (2nd Ed, 2011) $65.00
by Douglas Self 2nd Edition 2006 $69.00
See
A collection of 35 classic magazine articles offering a dependable methodology for designing audio power amplifiers to improve performance at every
point without significantly increasing cost. Includes compressors/limiters,
hybrid bipolar/FET amps, electronic switching and more. 474 pages in paperback.
Review
A great aid when wrestling with applications for the PICAXE
series of microcontrollers, at beginner, intermediate and advanced April
2011
levels. Every electronics class, school and library should have a copy,
along with anyone who works with PICAXEs. 300 pages in paperback
SMALL SIGNAL AUDIO DESIGN
PIC IN PRACTICE
By Douglas Self – First Edition 2010 $88.00
by D W Smith. 2nd Edition - published 2006 $60.00
The latest from the Guru of audio. Explains audio concepts in easy-to-understand language with plenty of examples and reasoning. Inspiration for audio
designers, superb background for audio enthusiasts and especially where it comes to
component peculiarities and limitations. Expensive? Yes. Value for money? YES! Highly
recommended. 558 pages in paperback.
Based on popular short courses on the PIC, for professionals, students and
teachers. Can be used at a variety of levels. An ideal introduction to the world
of microcontrollers. 255 pages in paperback.
PIC MICROCONTROLLER – your personal introduc-
AUDIO POWER AMPLIFIER DESIGN HANDBOOK
tory course By John Morton 3rd edition 2005. $60.00
by Douglas Self – 5th Edition 2009 $81.00
A unique and practical guide to getting up and running with the PIC. It assumes no knowledge of microcontrollers – ideal introduction for students,
teachers, technicians and electronics enthusiasts. Revised 3rd edition
focuses entirely on re-programmable flash PICs such as 16F54, 16F84 12F508 and
12F675. 226 pages in paperback.
"The Bible" on audio power amplifiers. Many revisions and
updates to the previous edition and now has an extra three
chapters covering Class XD, Power Amp Input Systems and
Input Processing and Auxiliarly Subsystems. Not cheap and not a book
for the beginner but if you want the best reference on Audio Power Amps,
you want this one! 463 pages in paperback.
OP AMPS FOR EVERYONE
PRACTICAL GUIDE TO SATELLITE TV
By Carter & Mancini – 3RD EDITION $100.00
Substantially updates coverage for low-speed and high-speed applications,
and provides step-by-step walk-throughs for design and selection of op
amps. Huge 648 pages!
By Garry Cratt – Latest (7th) Edition 2008 $49.00
Written in Australia, for Australian conditions by one of Australia's foremost
satellite TV experts. If there is anything you wanted to know about setting
up a satellite TV system, (including what you can't do!) it's sure to be covered in this
176-page paperback book.
PROGRAMMING 32-bit MICROCONTROLLERS
IN C By Luci di Jasio (2008) $79.00
NEWNES GUIDE TO TV & VIDEO TECHNOLOGY
Subtitled Exploring the PIC32, a Microchip insider tells all on this powerful
PIC! Focuses on examples and exercises that show how to solve common,
real-world design problems quickly. Includes handy checklists. FREE CD-ROM includes
source code in C, the Microchip C30 compiler, and MPLAB SIM. 400 pages paperback.
By KF Ibrahim 4th Edition (Published 2007) $49.00
It's back! Provides a full and comprehensive coverage of video and television technology including HDTV and DVD. Starts with fundamentals so is
ideal for students but covers in-depth technologies such as Blu-ray, DLP,
Digital TV, etc so is also perfect for engineers. 600+ pages in paperback.
USING UBUNTU LINUX
by J Rolfe & A Edney – published 2007 $27.00
RF CIRCUIT DESIGN
Ubuntu Linux is a free and easy-to-use operating system, a viable alternative to Windows and Mac OS. Introduces Ubuntu, tells how to set it up,
covers the various Open Office applications and gives troubleshooting
hints and tips. Highly recommended. 222 pages in paperback
DVD PLAYERS AND DRIVES
by K.F. Ibrahim. Published 2003. $71.00
A guide to DVD technology and applications, with particular focus on design issues and pitfalls, maintenance and repair. Ideal
for engineers, technicians, students of consumer electronics
and sales and installation staff. 319 pages in paperback.
by Chris Bowick, Second Edition, 2008. $63.00
The classic RF circuit design book. RF circuit design is now more important
that ever in the wireless world. In most of the wireless devices that we use
there is an RF component – this book tells how to design and integrate in a very practical fashion. 244 pages in paperback.
See
Review
Feb
2004
PRACTICAL RF HANDBOOK
by Ian Hickman. 4th edition 2006 $61.00
A guide to RF design for engineers, technicians, students and enthusiasts.
Covers key topics in RF: analog design principles, transmission lines,
couplers, transformers, amplifiers, oscillators, modulation, transmitters
and receivers, propagation and antennas. 279 pages in paperback.
ELECTRIC MOTORS AND DRIVES
By Austin Hughes - Third edition 2006 $51.00
PRACTICAL VARIABLE SPEED DRIVES &
POWER ELECTRONICS
Se
Intended for non-specialist users of electric motors and drives,
filling the gap between academic texts and general "handbooks".
Explores all of the widely-used modern types of motor and drive
including conventional & brushless DC, induction motors, steppers, servos, synchronous and reluctance. 384 pages, soft cover.
e
Review
Feb
An essential reference for engineers and anyone who wishes
2003
to design or use variable speed drives for induction motors.
by Malcolm Barnes. 1st Ed, Feb 2003. $73.00
286 pages in soft cover.
AC MACHINES
BUILD YOUR OWN ELECTRIC MOTORCYCLE
By Jim Lowe Published 2006 $66.00
Applicable to Australian trades-level courses including NE10 AC Machines,
NE12 Synchronous Machines and the AC part of NE30 Electric Motor
Control and Protection. Covering polyphase induction motors, singlephase motors, synchronous machines and polyphase motor starting. 160
pages in paperback.
by Carl Vogel. Published 2009. $40.00
Alternative fuel expert Carl Vogel gives you a hands-on guide with
the latest technical information and easy-to-follow instructions
for building a two-wheeled electric vehicle – from a streamlined
scooter to a full-sized motorcycle. 384 pages in soft cover.
NOTE: ALL PRICES ARE PLUS P&P – AUSTRALIA ONLY: $10.00 per order;
eMAIL (24/7)
To
silicon<at>siliconchip.com.au
Place
siliconchip.com.au
with order & credit card details
Your
Order:
1-13
See
Review
March
2010
OR
FAX (24/7)
Your order and card details to
(02) 9939 2648 with all details
OR
NZ – $12.00 PER BOOK;
PAYPAL (24/7)
Use your PayPal account
silicon<at>siliconchip.com.au
OR
REST OF WORLD $18.00 PER BOOK
PHONE – (9-5, Mon-Fri)
Call (02) 9939 3295 with
with order & credit card details
OR
MAIL
Your order to PO Box 139
MayCollaroy
2012 103
NSW 2097
Or use the handy order form on P85 of this issue
*ALL TITLES SUBJECT TO AVAILABILITY. PRICES VALID FOR MONTH OF MAGAZINE ISSUE ONLY. ALL PRICES INCLUDE GST
SILICON
CHIP
PARTSHOP
Looking for a specialised component to build that latest and greatest SILICON CHIP project? Maybe it’s the PCB you’re after.
Or a pre-programmed micro. Or some other hard-to-get “bit”. The chances are they are available direct from the SILICON CHIP PARTSHOP.
As a service to readers, SILICON CHIP has established the PARTSHOP. No, we’re not going into opposition with your normal suppliers
– this is a direct response to requests from readers who have found difficulty in obtaining specialised parts.
•
•
•
•
PCBs nominated are normally IN STOCK and ready for despatch (you don’t have to wait for them to be made!).
Even if stock runs out (eg, for high demand), in most cases there will be no longer than a two-week wait.
One low p&p charge: $10 per order, regardless of how many boards you order! (Australia only; overseas clients – email us for a postage quote).
New project boards will normally be available within days of the magazine on-sale date: no waiting!
• Our PCBs are beautifully made, very high quality fibreglass boards with pre-tinned tracks, silk screen overlays and where applicable, solder masks.
• Best of all, those boards with fancy cut-outs or edges are already cut out to the SILICON CHIP specifications – no messy blade work required!
CODE
Price*
PROJECT
PUBLISHED
CODE
Price*
JAN 1993
06112921
$25.00
12V 20-120W SOLAR PANEL SIMULATOR
FEB 1994
01102941
$5.00
MICROPHONE NECK LOOP COUPLER
MAR 2011
04103111
$25.00
MAR 2011
01209101
PRECHAMP: 2-TRANSISTOR PREAMPLIER
JUL 1994
01107941
$5.00
$25.00
PORTABLE STEREO HEADPHONE AMP
APRIL 2011 01104111
HEAT CONTROLLER
JULY 1998
10307981
$25.00
$25.00
CHEAP 100V SPEAKER/LINE CHECKER
APRIL 2011 04104111
MINIMITTER FM STEREO TRANSMITTER
APR 2001
$25.00
06104011
$25.00
PROJECTOR SPEED CONTROLLER
APRIL 2011 13104111
MICROMITTER FM STEREO TRANSMITTER
$10.00
DEC 2002
06112021
$10.00
SPORTSYNC AUDIO DELAY
MAY 2011
01105111
$30.00
SMART SLAVE FLASH TRIGGER
JUL 2003
13107031
$10.00
100W DC-DC CONVERTER
MAY 2011
11105111
$25.00
12AX7 VALVE AUDIO PREAMPLIFIER
NOV 2003
01111031
$25.00
PHONE LINE POLARITY CHECKER
MAY 2011
12105111
$10.00
POOR MAN’S METAL LOCATOR
MAY 2004
04105041
$10.00
20A 12/24V DC MOTOR SPEED CONTROLLER MK2
JUNE 2011
11106111
$25.00
BALANCED MICROPHONE PREAMP
AUG 2004
01108041
$25.00
USB STEREO RECORD/PLAYBACK
JUNE 2011
07106111
$25.00
LITTLE JIM AM TRANSMITTER
JAN 2006
06101062
$25.00
VERSATIMER/SWITCH
JUNE 2011
19106111
$25.00
POCKET TENS UNIT
JAN 2006
11101061
$25.00
USB BREAKOUT BOX
JUNE 2011
04106111
$10.00
STUDIO SERIES RC MODULE
APRIL 2006 01104061
$25.00
ULTRA-LD MK3 200W AMP MODULE
JULY 2011
01107111
$25.00
ULTRASONIC EAVESDROPPER
AUG 2006
01208061
$25.00
PORTABLE LIGHTNING DETECTOR
JULY 2011
04107111
$25.00
RIAA PREAMPLIFIER
AUG 2006
01108061
$25.00
RUDDER INDICATOR FOR POWER BOATS (4 PCBs)
JULY 2011
20107111-4 $80 per set
GPS FREQUENCY REFERENCE (A) (IMPROVED)
MAR 2007
04103073
$55.00
VOX
JULY 2011
01207111
$25.00
GPS FREQUENCY REFERENCE DISPLAY (B)
MAR 2007
04103072
$30.00
ELECTRONIC STETHOSCOPE
AUG 2011
01108111
$25.00
KNOCK DETECTOR
JUNE 2007
05106071
$25.00
DIGITAL SPIRIT LEVEL/INCLINOMETER
AUG 2011
04108111
$15.00
SPEAKER PROTECTION AND MUTING MODULE
JULY 2007
01207071
$25.00
ULTRASONIC WATER TANK METER
SEP 2011
04109111
$25.00
CDI MODULE SMALL PETROL MOTORS
MAY 2008
05105081
$15.00
ULTRA-LD MK2 AMPLIFIER UPGRADE
SEP 2011
01209111
$5.00
LED/LAMP FLASHER
SEP 2008
11009081
$10.00
ULTRA-LD MK3 AMPLIFIER POWER SUPPLY
SEP 2011
01109111
$25.00
12V SPEED CONTROLLER/DIMMER (Use Hot Wire Cutter PCB from Dec2010 18112101)
$25.00
HIFI STEREO HEADPHONE AMPLIFIER
SEP 2011
01309111
$45.00
CAR SCROLLING DISPLAY
DEC 2008
05101092
$25.00
GPS FREQUENCY REFERENCE (IMPROVED)
SEP 2011
04103073
$55.00
USB-SENSING MAINS POWER SWITCH
JAN 2009
10101091
$45.00
DIGITAL LIGHTING CONTROLLER LED SLAVE
OCT 2011
16110111
$30.00
DIGITAL AUDIO MILLIVOLTMETER
MAR 2009
04103091
$35.00
USB MIDIMATE
OCT 2011
23110111
$30.00
INTELLIGENT REMOTE-CONTROLLED DIMMER
APR 2009
10104091
$10.00
QUIZZICAL QUIZ GAME
OCT 2011
08110111
$30.00
INPUT ATTENUATOR FOR DIG. AUDIO M’VOLTMETER
MAY 2009
04205091
$10.00
ULTRA-LD MK3 PREAMP & REMOTE VOL CONTROL
NOV 2011
01111111
$35.00
6-DIGIT GPS CLOCK
MAY 2009
04105091
$35.00
ULTRA-LD MK3 INPUT SWITCHING MODUL
NOV 2011
01111112
$25.00
6-DIGIT GPS CLOCK DRIVER
JUNE 2009
07106091
$25.00
ULTRA-LD MK3 SWITCH MODULE
NOV 2011
01111113
$10.00
UHF ROLLING CODE TX
AUG 2009
15008091
$10.00
ZENER DIODE TESTER
NOV 2011
04111111
$20.00
UHF ROLLING CODE RECEIVER
AUG 2009
15008092
$45.00
MINIMAXIMITE
NOV 2011
07111111
$10.00
6-DIGIT GPS CLOCK AUTODIM ADD-ON
SEPT 2009
04208091
$10.00
ADJUSTABLE REGULATED POWER SUPPLY
DEC 2011
18112111
$5.00
STEREO DAC BALANCED OUTPUT BOARD
JAN 2010
01101101
$25.00
DIGITAL AUDIO DELAY
DEC 2011
01212111
$30.00
DIGITAL INSULATION METER
JUN 2010
04106101
$25.00
DIGITAL AUDIO DELAY FRONT & REAR PANELS
DEC 2011
0121211P2/3 $20 per set
ELECTROLYTIC CAPACITOR REFORMER
AUG 2010
04108101
$55.00
AM RADIO
JAN 2012
06101121
$10.00
ULTRASONIC ANTI-FOULING FOR BOATS
SEP 2010
04109101
$25.00
STEREO AUDIO COMPRESSOR
JAN 2012
01201121
$30.00
HEARING LOOP RECEIVER
SEP 2010
01209101
$25.00
STEREO AUDIO COMPRESSOR FRONT & REAR PANELS
JAN 2012
0120112P1/2 $20.00
S/PDIF/COAX TO TOSLINK CONVERTER
OCT 2010
01210101
$10.00
3-INPUT AUDIO SELECTOR (SET OF 2 BOARDS)
JAN 2012
01101121/2 $30 per set
TOSLINK TO S/PDIF/COAX CONVERTER
OCT 2010
01210102
$10.00
CRYSTAL DAC
FEB 2012
01102121
DIGITAL LIGHTING CONTROLLER SLAVE UNIT
OCT 2010
16110102
$45.00
SWITCHING REGULATOR
FEB 2012
18102121
$5.00
HEARING LOOP TESTER/LEVEL METER
NOV 2010
01111101
$25.00
SEMTEST LOWER BOARD
MAR 2012
04103121
$40.00
UNIVERSAL USB DATA LOGGER
DEC 2010
04112101
$25.00
SEMTEST UPPER BOARD
MAR 2012
04103122
$40.00
HOT WIRE CUTTER CONTROLLER
DEC 2010
18112101
$25.00
SEMTEST FRONT PANEL
MAR 2012
04103123
$75.00
433MHZ SNIFFER
JAN 2011
06101111
$10.00
INTERPLANETARY VOICE
MAR 2012
08102121
$10.00
CRANIAL ELECTRICAL STIMULATION
JAN 2011
99101111
$30.00
12/24V 3-STAGE MPPT SOLAR CHARGER REV.A
MAR 2012
14102112
$20.00
HEARING LOOP SIGNAL CONDITIONER
JAN 2011
01101111
$30.00
SOFT START SUPPRESSOR
APR 2012
10104121
$10.00
LED DAZZLER
FEB 2011
16102111
$25.00
RESISTANCE DECADE BOX
APR 2012
04105121
$20.00
12/24V 3-STAGE MPPT SOLAR CHARGER
FEB 2011
14102111
$15.00
RESISTANCE DECADE BOX PANEL/LID
APR 2012
04105122
$20.00
SIMPLE CHEAP 433MHZ LOCATOR
FEB 2011
06102111
$5.00
1.5kW INDUCTION MOTOR SPEED CONTROLLER
APR 2012
10105121
$35.00
THE MAXIMITE
MAR 2011
06103111
$25.00
HIGH TEMPERATURE THERMOMETER MAIN PCB
MAY 2012
21105121
$30.00
UNIVERSAL VOLTAGE REGULATOR
MAR 2011
18103111
$15.00
HIGH TEMPERATURE THERMOMETER F&R PANELS
MAY 2012
21105122/3 $20.00
PROJECT
PUBLISHED
AM RADIO TRANSMITTER
CHAMP: SINGLE CHIP AUDIO AMPLIFIER
OTHER ITEMS CURRENTLY IN THE SILICON CHIP PARTSHOP:
$20.00
* All prices P&P – $10 Per order within Australia. (Overseas customers please email us for a P&P quote)
G-FORCE METER/ACCELEROMETER SHORT FORM KIT
AUG 2011/NOV 2011
$44.50
(contains PCB (04108111), programmed PIC micro, MMA8451Q accelerometer chip and 4 MOSFETS)
TENDA USB/SD AUDIO PLAYBACK MODULE (TD898)
JAN 2012
$33.00
TENDA USB/SD AUDIO PLAYBACK MODULE (TD896)
JAN 2012
$33.00
2-WAY JST CONNECTOR LEAD
JAN 2012
$3.45
RADIO & HOBBIES ON DVD-ROM (Needs PC to play!)
n/a
$62.00
3-WAY JST CONNECTOR LEAD
JAN 2012
$4.50
AMATEUR SCIENTIST VOL4 ON CD
n/a
$62.00
AND NOW THE PRE-PROGRAMMED MICROS, TOO!
Micros from copyrighted and contributed
projects may not be available.
As a service to readers, SILICON CHIP is now stocking microcontrollers and microprocessors used in new projects (from 2012 on) and some
selected older projects – pre-programmed and ready to fly! Price for any of these micros is just $15.00 each + $10 p&p per order
PIC18F2550-I/SP
PIC18F4550-I/P
PIC16F877A-I/P
dsPIC33FJ128GP802-I/SP
Batt Capacity Meter (Jun09), Intelligent Fan Controller (Jul10)
GPS Car Computer (Jan10), GPS Boat Computer (Oct10)
6-Digit GPS Clock (May-Jun09), Lab Digital Pot (Jul10) Semtest (Feb-May12)
Digital Audio Signal Generator (Mar-May10), Digital Lighting Controller (Oct-
Dec10), SportSync (May11), Digital Audio Delay (Dec11)
PIC16F88-E/P
Projector Speed (Apr11), Vox (Jun11), Ultrasonic Water Tank
Level (Sep11), Quizzical (Oct11), Ultra-LD Preamp (Nov11)
PIC18F27J53-I/SP
USB Data Logger (Dec10-Feb11)
PIC32MX795F512H-80I/PT Maximite (Mar11), miniMaximite (Nov11)
PIC18LF14K22
PIC18F14K50
ATTiny861
PIC12F675
ATTiny2313
ATMega48
PIC18F1320-I/SO
dsPIC33FJ64MC802-E/SP
Digital Spirit Level (Aug11), G-Force Meter (Nov11)
USB MIDIMate (Oct11)
VVA Thermometer/Thermostat (Mar10), Rudder Position Indicator (Jul11)
UHF Remote Switch (Jan09), Ultrasonic Cleaner (Aug10),
Ultrasonic Anti-fouling (Sep10)
Remote-Controlled Timer (Aug10)
Stereo DAC (Sep-Nov09)
Intelligent Dimmer (Apr09)
Induction Motor Speed Controller (Apr-May12)
*Note: P&P is extra ($10 per order). Prices listed include GST and are valid only for month of publication of this list; thereafter are subject to change without notice. 05/12
When ordering, be sure to nominate BOTH the micro required and the project for which it must be programmed.
SILICON CHIP Order Form
Your Name:
Your Address:
State:
Postcode:
Country:
Telephone No:
Fax No:
Email Address:
Please supply:
Qty
Item
Price
Item Description
P&P
if extra
Total
Price
TOTAL $A
Thank you for your order.
Payment options:
EFT/Bank Deposit: Silicon Chip BSB 012-243 A/C 2636-80001 Please confirm transfer by email to silicon<at>siliconchip.com.au or fax 02 9939 2648
PayPal: From your PayPal account: “Send Money” to silicon<at>siliconchip.com.au
Cheque/Money Order/Bank Draft: payable to Silicon Chip (Australian dollars only) Mail to Silicon Chip PO Box 139 Collaroy NSW 2097 Australia
Credit Card (see below; Visa and Mastercard ONLY): Fax to 02 9939 2648, telephone 02 9939 3295 or mail or email to above address.
If paying by Visa or Mastercard please enter your details below (we DO NOT accept Amex, Diners or other credit cards)
Card
No:
Cardholder Name:
To
eMAIL (24/7)
Place
siliconchip.com.au
silicon<at>siliconchip.com.au
Your
with order & credit card details
Order:
-
OR
-
FAX (24/7)
This form (or a photocopy) to
(02) 9939 2648 with all details
-
/
Expiry
Date:
Signature:
OR
PAYPAL (24/7)
OR
Use PayPal to pay
silicon<at>siliconchip.com.au
PHONE – (9-5, Mon-Fri)
Call (02) 9939 3295 with
your credit card details
MAIL
ay form
2012 105
ORMThis
to PO Box 139,
*ALL ITEMS SUBJECT TO AVAILABILITY. PRICES VALID FOR MONTH OF MAGAZINE ISSUE ONLY. ALL PRICES IN AUSTRALIAN DOLLARS AND INCLUDE GST WHERE APPLICABLE.
Collaroy NSW 2097
05/12
ASK SILICON CHIP
Got a technical problem? Can’t understand a piece of jargon or some technical principle? Drop us a line
and we’ll answer your question. Write to: Ask Silicon Chip, PO Box 139, Collaroy Beach, NSW 2097 or
send an email to silicon<at>siliconchip.com.au
Courtesy light circuit
cycling on and off
I recently assembled an interior light
delay kit (SILICON CHIP, June 2004). I
wired it up with a test light, switch and
12V power supply on my bench before
I tried installing it into my vehicle.
The problem I am having is that
after dimming the light turns back
on, then times out, then dims again
etc. All the polarised components are
installed correctly and I have used the
proper capacitors for 12V operation.
Reversing wires to the switch makes
no difference. Any suggestions? (D. S.,
Calgary, Canada).
• For the circuit to drive the lamp
to cycle on and dim and then back
on again, capacitor C1 must be being
discharged each time. That normally
only occurs if the door switch closes
again or if 12V is removed.
Alternatively, there may be a problem when using a test light to check the
circuit operation. That’s because the
test lamp may not have a sufficiently
low resistance which might be because
it consists of a LED and series resistor
instead of a light bulb.
The circuit does rely on power being supplied via the cold filament. So
to test operation correctly, you may
need to connect a 100Ω 5W resistor
in parallel with your test light. This
resistor would also be required if the
courtesy lights in the vehicle are LEDs
instead of incandescent lamps.
Using the SportSync as
an audio delay line
I was wondering if the SportSync
Delay for Digital TV (SILICON CHIP,
May 2011) could be modified to work
as an audio delay line or a voice pitch
shifter?
I’m an analog man myself and do not
know much about digital circuitry. But
based on Figs.1 & 2 in the May 2011
article, I wonder if this could be done
in a future article of SILICON CHIP? (K.
S., Scranton PA, USA).
• Yes, the SportSync project could be
used as a generic audio delay line as
long as 40kHz, 12-bit mono quality is
good enough. The sample rate could
be increased but the hardware isn’t
present for a second channel and the
micro’s internal ADC/DACs limit the
voltage resolution. While the voltage
resolution is nominally 12 bits, when
you take noise into account the performance is more like 11 bits.
The software could be modified
to provide smaller delay adjustment
steps and better repeatability as well
as a shorter minimum delay setting.
For use as a voice pitch shifter, with
different software it could be configured to perform any DSP task as long
as 40MIPS is enough CPU power. The
same sound quality proviso as stated
above would apply.
What causes chip
failure on credit cards?
I received my first credit card with
an embedded chip around two years
ago. In that time the chip has failed
three times and I have been sent a
replacement card each tim.
I have used cards with the normal
magnetic strip for over 30 years and
have never had a failure; they are replaced every three years or so at the
expiry date. The card is always stored
in my leather wallet and the copper
colouring of the embedded chip does
show some signs of wear as it turns
to a silvery colour but looks in good
condition.
Considering the amount of information that can be stored digitally in
the chip it seems ludicrous that they
should fail so easily. Can you enlighten
me as to why I am experiencing these
problems as no amount of questions
Adapting The Radiator Coolant Alarm To Detect Water In Diesel Fuel
Over the years I have built several
of the coolant level kits (SILICON
CHIP, June 1994) for my cars and utes
which has saved me many dollars; a
great kit, an oldie but a goody.
I have now joined the grey nomads
with a 20-year old Coaster camper
van which is going to receive the
water level alarm. There have been
serious articles in 4WD Action concerning the effect of water in diesel
fuel and the damage that this can
cause.
I would like to adapt the kit to
run off 24V and to detect the presence of water in fuel, either diesel or
petrol, to give it a greater audience.
106 Silicon Chip
The alarm would be a back-up to
any existing unit in the vehicle. I
have not made or adapted any filter
system at this stage for the sensor but
have a few ideas in mind.
Any assistance would be appreciated. (P. G., Manly, Qld).
• The Coolant Level Alarm circuit
could be run from 24V if 16V zener
diode ZD1 is changed to 33V and the
1kΩ resistor feeding the 4.7V zener
diode (ZD2) is changed to 2.7kΩ.
The electrolytic capacitors should
be upgraded from 16V to 35V rating.
The circuit’s measurement of conductivity should be suitable for detecting the difference between water
and fuel. The comparator threshold
level can be changed to suit the
water/fuel conductivity voltage by
setting the voltage divider at pin 6
of comparator IC1a. The 6.8kΩ and
10kΩ voltage divider resistors across
4.7V zener diode ZD2 should be
removed. Instead, a trimpot across
ZD2, with the wiper connected to
pin 6, should allow suitable voltage
variation.
Note that we have not tested these
suggested modifications. Note also
that the sensor’s ability to withstand
petrol and diesel should be checked
to ensure that the plastic of the sensor is not dissolved or softened.
siliconchip.com.au
directed at the card providers gives
me any plausible explanations? (M.
T., Donvale, Vic).
• We really have no idea as to why
these credit cards fail but we would
be prepared to bet that it is due to a
failure of the encapsulation protecting
the chip.
Changing the antifouling cut-off voltage
I have built the Ultrasonic Antifouling kit (SILICON CHIP, September &
November 2010). Is it possible to alter
the turn-off voltage as 11.5V is way too
low? As indicated in the attached link
(http://marine-electronics.net/techarticle/battery_faq/b_faq.htm) you will
see a battery is 50% discharged at
12.24V. I want to maximise my battery
life by not deep-discharging it.
Can I alter to the kit to stop at this
voltage? (M. V., via email).
• As you state, the website quotes that
a battery is 50% discharged at 12.24V.
Note that this is the open-circuit voltage of the battery (one that has been
allowed to stand without power draw)
and at a specific gravity electrolyte
measurement of 1.265 when fully
charged for a wet lead-acid battery
and also at 26.7°C. They also state that
the open-circuit voltage will vary for
gel cell (SLA) and AGM batteries, so
you would need to check the manufacturer’s specifications.
Trying to gauge battery capacity
simply by measuring battery voltage
is open to a wide interpretation. Typically, a battery is deemed to be fully
discharged at 10.8V when discharged
at the 10-hour rate, according to Exide
battery data. The 11.5V we have chosen is a conservative threshold that
ensures that the battery is not fully
discharged. In fact, if you measured the
battery voltage after the anti-fouling
driver had switched off, you would
find that the voltage would bounce up
to pretty close to 12V.
On the other hand, if you did adjust
the unit for a higher cut-off voltage,
you would find that the anti-fouling
driver will be switched off most of
the time. That’s because a battery will
generally provide 12V or less under
load unless it is being charged with
sufficient current to raise this voltage.
Assume you set the cut-off voltage
for 11.8V. Once the battery reached this
voltage, the anti-fouling unit would
switch off and then not turn on again
siliconchip.com.au
Toroidal Inductor For The 100W 12V Converter
I have a question concerning the
12V 100W converter published in
May 2011. Unfortunately, I cannot
seem to find a source for the toroidal
inductor core outside of Australia.
I would very much appreciate any
specifications you can supply me
such as inductance, current rating,
etc so that I may be able to wind a
similar inductor using a different
core.
I have a core that I salvaged from
a large AC to DC inverter that is
slightly smaller (by less than 10%)
than the size that is stated in the
article. Do you think that it could
be used if a few extra turns of wire
were added?
Any input is greatly appreciated,
as I would like to build many of
these for friends and colleagues
that are constantly on the go with
until the battery voltage had risen
substantially above 12V because of
the hysteresis in the setting.
(Step 1). The switch-off voltage is
determined by the resistor between
pin 5 of IC2 and the positive supply
rail. Let’s call this R1. It is 20kΩ in
the original design, giving an 11.5V
threshold. The formula to calculate
the value for a different threshold is:
R1 = ((Switch off threshold voltage x
10kΩ) - 38.4kΩ) ÷ 3.83
(Step 2). To calculate the switch-on
voltage (12V in the original design):
Switch on threshold voltage = 4 x
(R1 + 10kΩ) ÷ 10kΩ
For example, to switch off at 12.24V
R1 should be 2.2kΩ. Switch on voltage
would then be 12.8V.
Bypass switch needed
for Stereo Compressor
For my use the Stereo Compressor
featured in the January 2012 issue appears to have a major flaw. It needs a
“bypass” facility for when it is not in
use. It appears to be easy to implement.
Two relays could be easily added and
these would be activated when the
unit is turned on.
While I could do this on the current PCB, in the current box it would
be a squeeze. Can your designers see
any problem with just using one relay
which simply connects the input to
their mobile PCs. (J. K., Baltimore,
Maryland, USA).
• The toroid is a powdered-iron
type with dimensions of 42 x 22 x
17mm and an Al value of 90. The
formula for the number of turns
based on the Al value is number
of turns = 1000 x the square root of
(the inductance in mH divided by
the Al value).
The maximum current for the
wound inductance is calculated from
the 16mJ value for the core, where
this = (inductance L) x (the maximum
instantaneous current squared).
The windings and core Al value
are not critical and a slightly smaller
core could be used if necessary, but
the core must be powdered-iron,
not ferrite. You can also obtain the
specified core from www.jaycar.com
(Cat. LO-1246).
the output when the unit is powered
off? Would having the un-powered
unit still connected to the audio line
be likely to have any serious effect?
(G. S., Mildura, Vic).
• The compressor’s input cannot be
directly connected to the output when
power is off since the input would then
be directly driving the output of IC3
via the 150Ω resistor, ie, the left signal
would be driving IC3b’s output and the
right channel input would be driving
IC3a’s output. This would severely
load the input signal.
However, a single relay that incorporates double-pole change-over
contacts could be used to provide for
a change-over of signal.
For each channel (left and right
channels), connect one set of relay
contacts with the normally open (NO)
contact to the compressor output and
the normally closed (NC) contact to
the input and use the common (C)
connection of the relay as the signal
output. This will then select between
the output of the compressor when
the relay is on and the input to the
compressor when the relay is off.
Note that the signal levels may
be different between when the compressor is on and when the signal is
diverted directly from the input. The
level difference is not due to compressor action but because of the level
and volume settings available for the
May 2012 107
Checking The Polarity Of PA Loudspeakers
As a school janitor/groundsman
I am expected to know and do
everything (actually even before it
happens!) The school’s 100V line
PA amplifier system is now in need
of “tweaking” after the installation
of a 250W Redback amplifier and
separate Redback mixer. It all works
reasonably well but the issues are
with the horn-type speakers, some
of which are probably 20 years old.
They want to play pacifying music
over the system and currently it is
not “pacifying” due to the poor audio quality. I foolishly mentioned the
availability of quality TOA music
horns as suitable replacements that
would help. However, I have had
various conflicting advice concerning the “polarity” or correct “phasing” when connecting them up. I
would have thought the marked
wire would be the “common”, ie,
negative, and the un-marked wire
would be the “positive”.
Some of these speakers are in
sight of each other and I don’t want
to have them competing with each
other. I have tried a multimeter at the
back of the amplifier but it does not
read on AC or DC (when at rest, no
music being played but amplifier on)
so this would prevent me checking
“polarity” at each speaker location.
How do I check the polarity please?
The wires are a total mishmash
draped through roof crawl spaces
etc and change type and colour
compressor which are not effective for
the direct signal bypass.
Confusion about
power factor
I’ve been reading your past articles
about power factor correction and
electricity saving equipment that has
recently been a hot topic in the media.
I just wanted to ask one question. Energex say that we are charged in kWh
which is amps x volts x power factor.
When testing these types of products
I presume that you have tested what
the energy meter is ACTUALLY measuring.
Can you tell me what you discovered when you tested the meter and
108 Silicon Chip
from start to finish at times. (R. M.,
Brisbane, Qld).
• There is no easy answer to your
question, especially as it is an old installation and you cannot be certain
that the speaker wires are connected
“stripe to stripe” where there are
joins. Nor can you be certain they’re
not corroded, have faulty insulation,
or are open or short circuit, etc.
Unfortunately, in most PA horns
you cannot easily get access to the
diaphragm (once you do you’ve
probably destroyed the driver!) so
you cannot check forward or backward movement as you can with a
typical speaker.
You are right that the usual practice is simply to connect the striped
or marked wire of the figure-8 cable
to the negative terminal so then all
speakers will be “in phase”. But if
the speakers are separated by a reasonable distance, say 20m+, correct
phasing isn’t as important as you
might imagine. Sure, it’s better to
have them all in phase but PA speakers, especially horns, don’t have a
good bass response anyway – maybe
this is the reason for your perceived
poor sound quality.
You won’t do any damage by having one or more speakers connected
out of phase. You are also right that
you should not be able to read any
voltage at the amplifier speaker terminals when not playing anything.
Not that it will mean anything but
whether it actually records power
factor?
I’ve been reading a few comments
in forums of people who say they have
tested it and found that they don’t
measure power factor at all. They say
they don’t even have the means to
measure it. If this is the case then these
power saving products would work to
reduce what we are being billed. My
partner is arranging for one of the sales
guys to visit our home so we can find
out more about these products and we
would value your opinion.
Many people have seen their bills
jump after having a smart meter installed. Have you tested these at all?
(M. E., Regents Park, NSW).
• Regardless of what Energex told
you should be able to read an AC
voltage when there is something
playing but definitely not DC – if
you can something is wrong with
the amplifier!
If you have access to the individual speaker lines, we’d be much more
inclined to check their impedance
with an AC impedance meter. From
this, if you know what the power tap
is on your speakers and how many
are connected in parallel (the correct way to wire 100V speakers) you
can tell if you have a problem. In a
typical school-type PA system, we’d
expect to see the speakers set to their
5W or 10W tap, so each should have
an impedance of about 2kΩ or 1kΩ.
If you cannot access individual
lines, check the overall impedance
(when not connected to the amplifier) and use Ohm’s Law to work out
if the number of speakers in parallel
equates to roughly the right impedance; eg, 10 speakers on 10W tap
(1kΩ each) should be somewhere
around 100Ω, plus a bit for the
speaker cables themselves. A wildly
different figure than this probably
means a faulty speaker or line.
Note that the resistance as measured with a multimeter will not
show the correct reading, as you
are measuring the DC resistance.
It’s OK for a really rough guide but
really only tells you if you have, for
example, an open-circuit or shorted
speaker line.
you, domestic electricity customers
are not charged for power factor. Reducing the power factor of loads in
your home will make no difference
to your bill. Actually, on the basis of
our measurements, it will lead a to a
very slight increase in your power bill.
We have several types of power
meter, some of which measure power
factor while others don’t. It doesn’t
matter whether they measure power
factor or not because the calculation
for power in watts only takes into
account the voltage and current components which are actually in phase.
So you should get the same reading,
regardless of whether the power meter
used has a readout of power factor
or not.
siliconchip.com.au
Smart meters are a different proposition entirely because they vary the
electricity tariff according to the time
of day. In fact, most people who have
smart meters installed will experience
a substantial increase in their power
bills unless they are able to operate
their major power-consuming appliances in the off-peak periods. Most
people cannot do that because, for
example, they need to operate their
air-conditioner or heaters at the time
of peak power tariff.
In fact, smart meters are only
“smart” for the electricity retailers.
They are very bad, not smart at all, for
consumers. However, even if you do
have a smart meter installed, you are
not charged for power factor.
Dual battery
switcher wanted
I was wondering if SILICON CHIP has
ever put together, or might consider
putting together, a “Dual Battery Isolator Relay Controller” for “small” automotive applications. I’m not referring
to the large high-current dual-battery
systems used for caravans or campers
with deep cycle batteries of 200Ah or
Reed Switch & Frequency-Activated Switch
With reference to your Frequency-Activated Car Switch project
(SILICON CHIP, June 2007), is it
possible to use a reed switch with
a couple of magnets instead of the
sensors that were suggested? (K. W.,
via email).
• A reed switch can be used but
a 10kΩ resistor should be used to
pull the voltage up or down when
more. I’m referring to just using a small
12V sealed lead-acid battery, say 5Ah,
as the power source sink for a car PC or
headrest DVD player. Call my desired
target solution a “12V In-Car Smart
UPS” if you will.
The 12V automotive power supplies
in car PCs have not, in my experience,
been 100% reliable at shutting down
the host operating system. I keep hearing of situations where a blocked shutdown, hibernation or even a delayed
sleep results in the car PC staying on
overnight until the main battery goes
dead.
Having an isolated dual-battery sys-
the reed switch contacts are open.
If the reed switch is wired to
GND and the frequency switch
input, connect the 10kΩ resistor
between the input and 7.4V supply.
Alternatively, if the reed switch is
wired between the 7.4V supply and
the frequency switch input, connect
the resistor between the input and
GND.
tem would prevent this from hurting
your attempt to commute the next day.
Also, on a very long drive when
the kids are watching DVDs on an
after-market car DVD player (headrest,
ceiling, standalone etc), as soon as the
car is turned off at a petrol station or
a toilet stop, their playback gets shut
down too. If you’re playing a normal
DVD then most headrests have a
resume-from-last-position feature. But
when playing AVI or DiVX files from
a CD/DVD or thumb drive, the poor
kids have to skip a gazillion times
to find their place again once we are
back under way. Having the headrests
Radio, Television & Hobbies: the COMPLETE archive on DVD
YES!
NA
R
MO E THA URY
N
E T
QUARTER C NICS
O
OF ELECTR !
HISTORY
This remarkable collection of PDFs covers every issue of R & H, as it was known from
the beginning (April 1939 – price sixpence!) right through to the final edition of R, TV & H
in March 1965, before it disappeared forever with the change of name to EA.
For the first time ever, complete and in one handy DVD, every article and every issue
is covered.
If you’re an old timer (or even young timer!) into vintage radio, it doesn’t get much more
vintage than this. If you’re a student of history, this archive gives an extraordinary insight
into the amazing breakthroughs made in radio and electronics technology following the
war years.
And speaking of the war years, R & H had some of the best propaganda imaginable!
Even if you’re just an electronics dabbler, there’s something here to interest you.
• Every issue individually archived, by month & year
• Complete with index for each year
• A must-have for everyone interested in electronics
Please note: this archive is in PDF format on DVD for PC.
Your computer will need a DVD-ROM or DVD-recorder (not a CD!)
and Acrobat Reader 6 or above (free download) to enable you to view this
archive. This DVD is NOT playable through a standard A/V-type DVD player.
Exclusive to
SILICON CHIP
ONLY
62
$
00
+$10.00 P&P
HERE’S HOW TO ORDER YOUR COPY:
BY PHONE:*
(02) 9939 3295
9-4 Mon-Fri
BY FAX:#
(02) 9939 2648
24 Hours 7 Days
<at>
BY EMAIL:#
silchip<at>siliconchip.com.au
24 Hours 7 Days
BY MAIL:#
PO Box 139,
Collaroy NSW 2097
* Please have your credit card handy! # Don’t forget to include your name, address, phone no and credit card details.
siliconchip.com.au
BY INTERNET:^
siliconchip.com.au
24 Hours 7 Days
^ You will be prompted for required information
May 2012 109
Hand Capacitance Affects Motor Speed Controller
Controller
I have a question regarding the
Full-Wave Motor Speed Controller
published in the May 2009 issue. It
operates basically as I would expect,
except for one minor anomaly.
If I set a relatively low speed with
my hand resting against the metal
case, then remove my hand from
the case, I see a noticeable speed increase. I’ve tried varying the setting
of the internal trimpot but it doesn’t
seem to affect this behaviour. It’s fine
if I adjust the speed and make sure I
don’t touch the case.
continue to play – or even pause – until
the car begins moving again would
solve that issue too.
Using just a 12V relay for dual-battery operation means that the batteries
are connected together as soon as the
ignition is on, possibly adding load
to the main battery while it is coldcranking if the second battery is fully
discharged. I don’t think that’s ideal.
I think automotive diode solutions are
another option but I’m not sure if the
diode voltage drop would be an issue.
The ideal solution would be to have
a little “smart” controller between the
vehicle ignition line and the trigger
input of a suitable small 12V relay.
This could delay activating the relay
for a set period of time after the vehicle main battery/alternator voltage
has reached and stabilised above a
“turn on” level and could shut down
when the battery/alternator voltage
drops below a “turn off” level. So the
batteries are connected only when the
car is running.
This “smart” controller might also
Do you have any suggestions as
to how to minimise this behaviour?
The mains wiring is well clear of
the pot, so I don’t think it’s related
to that. (P. P., Inverell, NSW).
• This is probably caused by the
case not being properly earthed.
Make sure that both the lid and the
case are earthed to the mains earth
via the earth pin on the IEC connector. Check also that the power
lead has an earth wire connecting
between the earth pins of the connectors at each end.
be used to trigger high-current relays
in a larger solution as well if needed,
but would be ideal for the smaller
solution proposed here. Incidentally,
it would also be an ideal ‘trigger’ for
turning a car PC on and off – the “On”
occurring after cold-cranking has
completed and the “Off” occurring at
vehicle shut down.
Being able to use a small lead-acid
battery lends itself to limited space
and costs constraints and satisfies the
requirements for the situations listed
previously. Why are existing solutions
not appropriate? The extended run
time of a large deep-cycle battery is not
required, therefore we can opt out of
both the large-space and high-current
challenges it poses. Off the shelf
dual-battery isolators are not ideal
in a small solution, because they are
size and cost-geared to large current
applications.
So a “Dual Battery Isolator Relay
Controller” on its own would be ideal.
Alternatively, a project that includes
a battery, power control and even
perhaps “Vehicle Power Socket and
5V USB” outputs ready to go – ie, a
complete “12V In Car UPS” – would be
just fantastic! (L. P., Rutherford, NSW).
• A dual-battery isolator was published in Circuit Notebook in March
2007. This used a 40A relay. Horn
relays would be suitable even though
they have a rating higher than you are
after but they are rugged and ideally
suited for automotive use.
Pushbutton engine start
circuit wanted
I have searched the net for a DIY
engine start/stop button but with no
success. All I can find is a pushbutton
instead of the key, so you still need
to insert the key into the ignition and
turn ACC on, so why bother? I was
hoping to do a pushbutton start as in
late BMWs (keyless), so no key will
be needed.
My car is not equipped with an immobiliser so the process will be easier.
What I’m looking for is a way to get into
the car using the alarm remote control,
depress the brake pedal then push the
button and the car will start. If you
don’t press the brake pedal it will only
turn ACC on. If the engine is already
running, then pressing the button will
stop the engine; that’s the idea.
• You can forget about that idea if
you have a reasonably modern car
with a key start. The problem is that
the ignition keys in modern cars have
an inbuilt RFID transponder. The key
must be in the lock before the car will
start.
There is no way you can overrule
this unless you can hack the software
in the car’s body computer. If your car
is older, as indicated by its lack of an
. . . continued on page 112
WARNING!
SILICON CHIP magazine regularly describes projects which employ a mains power supply or produce high voltage. All such
projects should be considered dangerous or even lethal if not used safely.
Readers are warned that high voltage wiring should be carried out according to the instructions in the articles. When
working on these projects use extreme care to ensure that you do not accidentally come into contact with mains AC voltages
or high voltage DC. If you are not confident about working with projects employing mains voltages or other high voltages,
you are advised not to attempt work on them. Silicon Chip Publications Pty Ltd disclaims any liability for damages should
anyone be killed or injured while working on a project or circuit described in any issue of SILICON CHIP magazine.
Devices or circuits described in SILICON CHIP may be covered by patents. SILICON CHIP disclaims any liability for the infringement of such patents by the manufacturing or selling of any such equipment. SILICON CHIP also disclaims any liability
for projects which are used in such a way as to infringe relevant government regulations and by-laws.
Advertisers are warned that they are responsible for the content of all advertisements and that they must conform to the
Competition & Consumer Act 2010 or as subsequently amended and to any governmental regulations which are applicable.
110 Silicon Chip
siliconchip.com.au
MARKET CENTRE
Cash in your surplus gear. Advertise it here in SILICON CHIP
ELNEC IC PROGRAMMERS
Battery Packs & Chargers
High quality
Realistic prices
Free software updates
Large range of adaptors
Windows 95/98/Me/NT/2k/XP
C O N T R O L S
Tough times
demand innovative solutions!
CLEVERSCOPE
USB OSCILLOSCOPES
2 x 100MSa/s 10bit inputs + trigger
100MHz bandwidth
8 x digital inputs
4M samples/input
Sig-gen + spectrum analyser
Windows 98/Me/NT/2k/XP
IMAGECRAFT C COMPILERS
ANSI C compilers, Windows IDE
AVR, TMS430, ARM7/ARM9
68HC08, 68HC11, 68HC12
GRANTRONICS PTY LTD
www.grantronics.com.au
FOR SALE
PCBs MADE, ONE OR MANY. Any
format, hobbyists welcome. Sesame
Electronics Phone (02) 8068 2713.
sesame<at>sesame.com.au
www.sesame.com.au
LEDs! Nichia, Cree and other brand
name LEDs at excellent prices. LED
drivers, including ultra-reliable linear
driver options. Many other interesting
and hard-to-find electronic items!
www.ledsales.com.au
questronix.com.au – audiovisual experts solve home, corporate security
and devotional installation & editing
woes. QuestAV CYP, Kramer TVone
(02) 4343 1970 or sales<at>questronix.
com.au
PCBs & Micros: Silicon Chip Pub
lications can supply PCBs and programmed micros for recent (and some
not so recent) projects described in the
magazine. See our advert in this issue
for further details. Phone ( 02) 9939 3295
or email silicon<at>siliconchip.com.au
siliconchip.com.au
Siomar Battery Engineering
www.batterybook.com
Phone (08) 9302 5444
Made in Australia, used by OEMs world-wide
splat-sc.com
SERVICE TECHNICIAN (CASUAL)
SYDNEY BRANCH (AUBURN)
We require the services of a bright and enthusiastic
person to help service our product returns. Duties will
include fault finding, repairs & RA processing. Computer
literacy & an extensive knowledge of electronics is a
must. Australian residents only.
To apply contact NSW Manager Peter Deeb
peter.deeb<at>altronics.com.au or post applications to
15 Short St, Auburn, NSW 2144.
KIT ASSEMBLY & REPAIR
KEITH RIPPON KIT ASSEMBLY &
REPAIR:
* Australia & New Zealand;
* Small production runs.
Phone Keith 0409 662 794.
keith.rippon<at>gmail.com
GEOFF COPPA KIT ASSEMBLY AND
TROUBLE SHOOTING SERVICE.
Phone Geoff on 0414226102.
coppamitchell2<at>bigpond.com
WANTED
CUSTOMERS WANTED: Truscotts
Electronic World – large range of semiconductors and passive components for
industry, hobbyist and amateur projects
including Drew Diamond. 27 The Mall,
South Croydon, Melbourne. Phone
(03) 9723 3860. www.electronicworld.
com.au
WANTED: EARLY HIFIs, AMPLIFIERS,
Speakers, Turntables, Valves, Books,
Quad, Leak, Pye, Lowther, Ortofon,
SME, Western Electric, Altec, Marantz,
McIntosh, Tannoy, Goodmans, Wharfedale, radio and wireless. Collector/
Hobbyist will pay cash. (07) 5471 1062.
johnmurt<at>highprofile.com.au
DESIGN SERVICES
Sethna Electronics is a small business
that specializes in the design and manufacture of electronic based products.
We do low volume manufacturing and
reverse engineering. Contact 042 5239
254 or email nbsethna<at>gmail.com to
discuss your requirements.
ADVERTISING IN MARKET CENTRE
Classified Ad Rates: $29.50 (incl. GST) for up to 20 words plus 85 cents for
each additional word. Display ads: $54.50 (incl. GST) per column centimetre
(max. 10cm). Closing date: 5 weeks prior to month of sale. To book, email the
text to silicon<at>siliconchip.com.au and include your name, address & credit
card details, or fax (02) 9939 2648, or phone (02) 9939 3295.
May 2012 111
Advertising Index
Altronics....................loose insert,111
Amateur Scientist CD..................... 77
Cleverscope................................... 10
Dyne Industries................................ 6
Embedded Logic Solutions............ 25
Emona Instruments........................ 13
Geoff Coppa................................. 111
Grantronics.................................. 111
Hare & Forbes............................ OBC
Circuit Ideas Wanted
High Profile Communications....... 111
We pay up to $100 for contributions
to Circuit Notebook, or you could win
a$150 gift voucher.
Silicon Chip Publications, PO Box 139,
Collaroy, NSW 2097.
Instant PCBs................................ 111
the motor noise can be reduced. You
set it to the frequency where the motor
makes the least noise associated with
the PWM switching.
The June 1997 controller would
require the oscillator components to
be altered to adjust the frequency. We
recommend using a 100nF capacitor
at pin 5 of IC1 and a resistance value
ranging from 10kΩ up to 100kΩ at pin
6; a 100kΩ trimpot could be used.
While noise will never be completely eliminated with such a pulse width
modulated controller, the noise can be
reduced by selecting the optimum frequency. The frequency can be adjusted
from 100Hz to 1kHz or thereabouts.
Note that the higher frequencies may
not provide a linear speed control but
SC
this depends on the motor.
LED Sales.................................... 111
DOWNLOAD OUR CATALOG at
www.iinet.net.au/~worcom
WORLDWIDE ELECTRONIC COMPONENTS
PO Box 631, Hillarys, WA 6923
Ph: (08) 9307 7305 Fax: (08) 9307 7309
Email: worcom<at>iinet.net.au
Ask SILICON CHIP
. . . continued from p110
immobiliser, your idea may well be
feasible but we have not produced any
suitable circuitry.
Motor is noisy with
PWM speed control
I built the 12V 10A speed control kit.
It works fine but there is a very loud
2kHz tone coming from the motor. It
is just a 12V fan but the tone is louder
than the motor! Can you give me some
advice please? (G. F., via email).
• We are not sure if you are referring
to the controller from June 1997 or
from June 2011. The June 2011 controller has frequency adjustment so that
Issues Getting Dog-Eared?
Keep your copies safe
with these handy binders
Available Aust. only. Price: $A14.95
plus $10 p&p per order (includes
GST). Just fill in and mail the handy
order form in this issue; or fax (02)
9939 2648; or call (02) 9939 3295 and
quote your credit card number.
Buy five and get them postage free!
112 Silicon Chip
Jaycar ................................ IFC,53-60
Keith Rippon................................. 111
Kitstop.............................................. 8
LHP.NET.AU................................. IBC
Microchip Technology....................... 7
Mikroelektronika............................. 11
Ocean Controls.............................. 12
Quest Electronics......................... 111
Radio & Hobbies DVD.................. 109
RF Modules.................................. 112
Roc-Solid......................................... 5
Sesame Electronics..................... 111
Sethna Electronics....................... 111
Silicon Chip Binders..................... 112
Silicon Chip Bookshop................. 103
Silicon Chip Order Form.............. 105
Silicon Chip Partshop................... 104
Silicon Chip Subscriptions............. 52
REAL
VALUE
AT
Siomar Battery Engineering...... 3,111
P&P
Switchmode Power Supplies.......... 65
$14.95
PLUS
Splat Controls.............................. 111
Tekmark......................................... 29
Truscotts Electronic World............ 111
Verbatim Lights.............................. 51
Wiltronics......................................... 9
Worldwide Elect. Components..... 112
siliconchip.com.au
May 2012 113
|