This is only a preview of the August 2025 issue of Silicon Chip. You can view 46 of the 104 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. Articles in this series:
Articles in this series:
Items relevant to "Modules: Thin-Film Pressure Sensor":
Articles in this series:
Items relevant to "USB-C Power Monitor, Part 1":
Items relevant to "RP2350B Development Board":
Items relevant to "Mic the Mouse":
Items relevant to "Ducted Heat Transfer Controller Part 1":
Purchase a printed copy of this issue for $13.00. |
AUGUST 2025
ISSN 1030-2662
08
9 771030 266001
The VERY BEST DIY Projects!
Mthiec
Mouse
$1300* NZ $1390
INC GST
INC GST
Ducted Heat
Transfer
Controller
RP2350B
Pre-Assembled
Development
Board
USB-C
USBPower Monitor
Measure current, voltage, power, energy & time on all modern USB-C devices
NEXT-LEVEL PRINTING
HAS LANDED AT JAYCAR
Jaycar now stocks an extensive range of ELEGOO®
FDM & Resin 3D printers and materials, from entry-level
favourites to powerhouse machines trusted by enthusiasts.
• 7 NEW Filament Printers
• 6 NEW Resin Printers
• Wide range of Filament and Resin
EXPLORE THE
ELEGOO® RANGE (AU)
Explore our great range of ELEGOO® 3D Printing gear, in stock on our website,
or at over 140 stores or 130 resellers across Australia and New Zealand.
jaycar.com.au 1800 022 888 | jaycar.co.nz 0800 452 922
EXPLORE THE
ELEGOO® RANGE (NZ)
Contents
Vol.38, No.08
August 2025
11 SpaceX, Part 2
There’s more to cover on SpaceX’s enormous Starship rocket, and after
that, we’ll look at their launch and recovery facilities, some notable SpaceX
missions to date and then consider their future.
By Dr David Maddison, VK3DSM
Space technology
23 Amplifier Cooling, Part 1
Designing the chassis for an amplifier (especially a large one) is crucial to
achieving the best possible cooling performance and therefore a long life.
We also look at upgrading the cooling in existing amplifiers.
By Julian Edgar
Electronic system design
34 A Thin-Film Pressure Sensor
s
p
a
c
e
X
Part 2:
Page 11
Thin-film pressure sensors are an affordable way to measure force and
weight, albeit without high accuracy.
By Tim Blythman
Low-cost electronic modules
52 Rigol DHO924S Oscilloscope
Rigol’s new DHO900 series has a plethora of modern features in a compact
package. We describe what these new features are, and what it’s like to use.
Review by Tim Blythman
Test equipment
67 The Boeing 737 MAX story
We explain how the failure of a single electronic part led to two fatal air
crashes.
By Brandon Speedie
Aerospace technology
38 USB-C Power Monitor, Part 1
Our Power Monitor measures current, voltage, power, energy and time on all
modern USB-C devices, and even legacy USB devices with an adaptor. It’s a
self-contained unit, and is powered by a 400mAh rechargeable battery.
By Tim Blythman
Test & measurement project
RIGOL DHO924S
Oscilloscope
2
Editorial Viewpoint
4
Mailbag
73
Subscriptions
82
Circuit Notebook
84
Serviceman’s Log
91
Vintage Radio
98
Online Shop
100
Ask Silicon Chip
103
Market Centre
104
Advertising Index
104
Notes & Errata
46 RP2350B Development Board
Like a Raspberry Pi Pico 2 on steroids, this breadboard-friendly module has
47 available I/O pins, including eight that can measure analog voltages. It’s
also optimised for overclocking!
By Geoff Graham & Peter Mather
Single-board computer project
60 Mic the Mouse
Mic the Mouse doesn’t eat much, just the occasional 3V lithium cell, and it
makes a good project for playing pranks on family and friends.
By John Clarke
Toy project
74 Ducted Heat Transfer Controller
Improve the energy efficiency of your home by transferring warm or cool air
between rooms automatically using this smart controller.
Part 1 by Julian Edgar & John Clarke
Home automation project
Page 52
1. High speed transmission opto-couplers
2. A wireless battery charger
3. 3-way latch using SCRs
Silvertone Model 18 AM/FM radio by
Associate Professor Graham Parslow
SILICON
SILIC
CHIP
www.siliconchip.com.au
Publisher/Editor
Nicholas Vinen
Technical Editor
John Clarke – B.E.(Elec.)
Technical Staff
Bao Smith – B.Sc.
Tim Blythman – B.E., B.Sc.
Advertising Enquiries
(02) 9939 3295
adverts<at>siliconchip.com.au
Regular Contributors
Allan Linton-Smith
Dave Thompson
David Maddison – B.App.Sc. (Hons 1),
PhD, Grad.Dip.Entr.Innov.
Geoff Graham
Associate Professor Graham Parslow
Dr Hugo Holden – B.H.B, MB.ChB.,
FRANZCO
Ian Batty – M.Ed.
Phil Prosser – B.Sc., B.E.(Elec.)
Cartoonist
Louis Decrevel
loueee.com
Founding Editor (retired)
Leo Simpson – B.Bus., FAICD
Silicon Chip is published 12 times
a year by Silicon Chip Publications
Pty Ltd. ACN 626 922 870. ABN 20
880 526 923. All material is copyright ©. No part of this publication
may be reproduced without the written
consent of the publisher.
Subscription rates (Australia only)
6 issues (6 months): $70
12 issues (1 year):
$130
24 issues (2 years):
$245
Online subscription (Worldwide)
6 issues (6 months): $52.50
12 issues (1 year):
$100
24 issues (2 years):
$190
For overseas rates, see our website or
email silicon<at>siliconchip.com.au
* recommended & maximum price only
Postal address: PO Box 194,
Matraville, NSW 2036.
Phone: (02) 9939 3295.
ISSN: 1030-2662
Editorial Viewpoint
Supplier price increases
I had hoped we could avoid raising the magazine’s
cover price or subscription rates this year. Unfortunately, several factors beyond our control mean we
must do so to continue producing a magazine with
the same level of quality and content.
The new prices (listed below) will take effect from
the next issue, or for subscribers, on renewals from the
1st of September. If you’d like to lock in the current,
lower rates, you can extend your subscription now and save a few dollars.
While I dislike price increases as much as anyone else, if you compare
the price-to-content ratio of Silicon Chip to other magazines, I think it’s still
a good deal.
Over the past couple of years, we’ve worked hard to reduce our business
operating costs to avoid price rises. However, those savings have been more
than wiped out by increased supplier costs. That left us with a decision:
reduce the amount of content in each issue, or raise prices to maintain it. I
believe most readers would prefer the latter, so that’s the path we’ve chosen.
The costs of both printing and postage have increased substantially this
year. I negotiated a favourable printing deal at the end of last year that I
hoped would last for at least 12 months. However, I have recently been told
it is no longer available. As a result, our printing costs increased by 27%
virtually overnight.
It looks like there may soon be less competition in the magazine printing
industry in Australia. Of course, less competition and less choice means
higher prices.
Australia Post is also increasing prices again. In January 2020, you could
mail a regular letter for $1.00. By the time you’re reading this, they will be
charging $1.70. You don’t need to be a mathematical wizard to calculate
that’s a 70% increase in around five years – way above the rate of inflation.
As a comparison, the change in the Silicon Chip cover price since 2013
(from $9.95 to $14.00) fairly well matches the rate of inflation over those years.
Also, companies are reducing the amount they are spending on sponsorships and advertising. One advertiser pulled out after being hit by the ridiculous US tariff situation. Others are simply tightening their marketing budgets
for various reasons. When advertising revenue drops, the only practical way
to make up the shortfall is by charging more for the magazine.
I’m hopeful that our recent cost-cutting measures, such as relocating from
the large Brookvale office to more affordable premises, will help limit further price increases in future. But, in the end, we’re still at the mercy of our
suppliers. I want to keep this magazine going for many years to come, and
that means it has to remain financially viable – without reducing quality.
So, from the September 2025 issue onwards, the cover price of the print
edition will be AU$14.00 or NZ$14.90; the online issues will not increase
(see prices in the masthead). The subscription prices will change as shown
in the table below.
by Nicholas Vinen
Subscription Prices, effective 01/09/2025
New Prices Print (AU)
Printing and Distribution:
24-26 Lilian Fowler Pl, Marrickville 2204
2
Silicon Chip
Combined (AU) Print (NZ)
Combined (NZ) RoW
6-month $72.50
$82.50
$85
$95
$105
12-month $135
$155
$160
$180
$200
24-month $255
$280
$300
$335
$390
Australia's electronics magazine
siliconchip.com.au
MAILBAG
your feedback
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 has the right to edit, reproduce in electronic form, and communicate these letters. This also applies to submissions to “Ask Silicon Chip”, “Circuit Notebook” and “Serviceman’s Log”.
Low tempco voltage references with zener diodes
I was interested to read in the latest Silicon Chip magazine about stable voltage references (Precision Electronics
pt8, June 2025; siliconchip.au/Article/18312). There is a
very interesting aspect that most don’t know about.
That is because it was deployed in alternator voltage
regulators, as a temperature-stable comparator, in the days
before op amps were in widespread use (from the mid1960s to the 1970s).
The designers had figured out the negative tempco (temperature coefficient) of a transistor base-emitter (B-E) junction was around -2mV/°C. They also knew that zener diodes
that broke down below around 5V tended to have a negative tempco too. But those with a breakdown of around
7-10V or more had a positive tempco.
Often in alternator regulators, the ‘comparator’ that monitored the alternator’s output operated at around 14.2V.
They monitored the voltage with a resistor divider and a
series zener diode, plus a transistor’s base-emitter junction.
When the voltage rose above 14.2V, this switched on the
input transistor, switching off other transistors to cut off
the drive to the rotor’s field coil.
When the alternator output voltage fell below 14.2V, the
rotor’s field was switched on again. It was a switch-mode
system, with a switching frequency, due to the electromagnetic delays in the alternator itself, typically being a
few hundred hertz.
In any case, the ‘trick’ was to make it temperature stable,
so a zener diode was chosen that had a positive tempco
that exactly cancelled the negative tempco of the input
transistor’s B-E junction.
Many manufacturers just had the zener diode and the
transistor on the same PCB in close proximity. However,
Lucas electrical industries in the UK did better. They wised
Un-potted alternator
regulator PCBs
discovered inside
the defunct Lucas
Electrical plant
in the UK after it
closed down. The
zener+transistor
four-legged device
acted as a perfectly
temperaturecompensated
comparator.
4
Silicon Chip
up and realised that the perfect zener diode and transistor
combo should be in the same package for thermal reasons,
so they created the ‘four legged device’ (photo below).
Most technicians have never heard of the four-legged
device because all of Lucas’ alternator voltage regulators
were potted in a difficult-to-remove black resin. As far as
I know, the Lucas semiconductor plant was the only one
to make it. I have removed one and subjected it to tests.
Say you have a transistor with a grounded (common)
emitter and a collector load resistor, and provide just
enough base current to bias it into Class-A, so the collector voltage is half the supply voltage. If you then heat the
transistor, the collector voltage crashes to near-zero in
short order because of the negative tempco of the transistor’s B-E junction.
When I did this test with Lucas’ four-legged device, and
put it in an oven at 130°C, the collector voltage did not
change at all. They had perfectly balanced it, and I was
beyond impressed, especially because of its very simple
nature.
When the Lucas factory in the UK closed down, an archivist managed to get in there and copy some documents and
photos, including two of their alternator regulator PCBs
that were not potted in resin yet. That’s how we managed
to get the photo.
Dr Hugo Holden, Buddina, Qld.
The cost of grid-scale battery storage
Matthew Prentis (Mailbag, May 2025) is making an error
in evaluating the cost of grid-scale battery storage. The spot
price and the asset LCOE, while both usually measured in
$/MWh, are different metrics. Batteries make much of their
money from relative pricing, ie, the arbitrage between high
and low spot prices.
The LCOE of lithium-ion storage is currently between
$100 and $200 per MWh in Australia.
If there is sufficient interest, I would consider covering energy sector subsidies in a future article, though this
would cover all fuel types. Renewable subsidies are only
a relatively small fraction.
Brandon Speedie, Alexandria, NSW.
An alternative to the Solar Diverter project
Regarding the Hot Water Solar Diverter project in the
June & July 2025 issues (siliconchip.au/Series/440), I had
an electrician install a timer within the meter box that
switches on the water heater for a set time from midday.
In my case, 90+% of the time, the panels are producing
maximum solar feed-in during this period. Also, midday
is an off-peak time to buy power in Tasmania.
Australia's electronics magazine
siliconchip.com.au
A 5kW system more than covers the 3.6kW hot water
heater’s power needs. As you are probably aware, these
timers are the same size as the breakers, so it’s a very quick,
reliable and cheap option. We are a two-person household,
so we only need ours on for approximately two hours per
day. I also installed a mains LED across the element, making it easy to check how long it’s actually on.
Geoff Young, New Town, Tas.
The error is only small, but noticeable, and would not
be detectable with a 2kW load on. I purposefully switched
it off during the day because I noticed a peak charge on
a circuit where I expected none. In my case, the error is
about 0.6kWh in a 30-day period (equivalent to a load of
approximately 10W). I wonder if some of your readers may
have a similar problem.
Wolf-Dieter Kuenne, Bayswater, Vic.
Solar inverter registering a load when there is none
Synchronous vs asynchronous motors
I have discovered that there is a small error in my
Landis+Gyr E360 grid-tied solar inverter when the controlled load circuit is on and I have solar energy available
for export.
I discovered this when the load was disconnected by
turning the circuit breaker off during the day while exporting solar energy. I found that while exporting over 1.5kW,
I got a recorded energy draw on my meter for the circuit
that had no load connected.
It occurs because the meter has 2 current transformers:
one that measures the controlled load (now disconnected)
and one measuring the export power or normal day circuit.
What I believe is happening is that the two CTs are in close
proximity, giving mutual induction into the controlled load
circuit CT, and thus a positive reading for that circuit. That
circuit does not register power export, only import.
It requires both the controlled load switch to be on and
measurements enabled for readings to be given. The controlled loads are now being switched on when solar energy
is available to use solar energy for what used to be powered using off-peak loads at night.
I would like to comment on a Mailbag letter by Robert
Budniak published in the March 2025 issue on running
induction motors at reduced voltages. In his summation,
he ignored that the original Mailbag letter from Ian Thompson referred to squirrel cage AC induction motors, which
are asynchronous and not synchronous types.
One must take into account the rotor slip in an AC induction motor to accurately calculate the torque. There are also
power factor differences. The short video at https://youtu.
be/tl1N6_flY5k does an excellent job of graphically illustrating the differences.
Andre Rousseau, Auckland South, New Zealand.
Running Linux within Windows
I have a machine set up with both Windows 10 and
Ubuntu Linux. It is a 4th Generation Intel Core i7, so not
a new system, but it has reasonable memory capacity and
an SSD for storage.
A Microsoft feature called “Windows Subsystem for
Linux”, or WSL2 for short, is used to run Ubuntu Linux.
After activating WSL2 and installing Ubuntu from the
ourPCB
LOCAL SERVICE <at> OVERSEAS PRICES
AUSTRALIA
PCB Manufacturing
Full Turnkey Assembly
Wiring Harnesses
Solder Paste Stencils
small or large volume orders
premium-grade wiring
low cost PCB assembly
laser-cut and electropolished
Instant Online Buying of Prototype PCBs
www.ourpcb.com.au
6
Silicon Chip
Australia's electronics magazine
0417 264 974
siliconchip.com.au
Tilting head to 900
Digital Readout
Speed Display
2 Speed Gearbox
Digital Depth Display
YOUR TURNING SOLUTION
3MT Spindle
Work Light
Dovetail Vertical Z-Axis
Geared Headstock
D1-4 Camlock
Spindle
3MT Tailstock
3MT Tailstock
Induction Hardened
TU-3008G-20M
OPTI-TURN LATHE & MILL
DRILL COMBINATION
Sawrf Drip Tray
PACKAGE DEAL
K149
5,049 inc GST
$
SAVE $275 OFF RRP
TU-3008G - Opti-Turn Bench Lathe
PACKAGE DEAL
This new TU-3008G model geared head lathe of very modern
design has many convenient facilities ideal for the enthusiastic
model engineer for small component manufacturer. A very rigid
180mm cast iron bed width, induction hardened and ground
slideways gives a high level of precision turning. The lathe capacity
is 700mm between the centres and 300mm swing over the bed
and with a generous 38mm spindle bore allowing larger stock to be
placed through the headstock. The workhorse that drives this lathe
has an upgraded powerful 2hp (1.5kW) motor providing smooth
running.
Other very useful features included are cross and longitudinal power
feeds that's not normally found of small conventional lathes, quick lock
tailstock and both metric and imperial thread cutting, ranging from 8-56
tpi or 0.2-3.5mm pitch. Also included are a 4-way indexing tool post with
16mm tool shank capacity, fine adjustment on the tailstock for taper
turning and the rear splash guard. A 160mm 3-jaw chuck with reverse
jaws is also supplied.
ORDER CODE: L691
TU-3008G - Opti-Turn Bench Lathe
PACKAGE
CONTENTS
L691 + M647
ORDER CODE: M647
The BF-20AV mill head attachment has more throat & height capacity
than the the BF-16AV and allows for larger work to be machined, it
also is specifically designed to suit the TU-2506V & TU-3008G Optimum
bench lathes; converting them into a universal combination lathe/
mill when mounted together. In addition this compact setup has the
advantage of that it requires less valuable floor space then the two
separate conventional machines.
View and purchase these items online: www.machineryhouse.com.au/SIC2507
SYDNEY
BRISBANE
MELBOURNE
(03) 9212 4422
(08) 9373 9999
PERTH
ADELAIDE
1/2 Windsor Rd,
Northmead
625 Boundary Rd,
Coopers Plains
4 Abbotts Rd,
Dandenong
11 Valentine St,
11/20 Cheltenham Pde
Woodville
(02) 9890 9111
(07) 3715 2200
Kewdale
Specifications & prices are subject to change without notification.
(08) 9373 9969
06_SC_280723
For full specifications visit www.machineryhouse.com.au/k149
Microsoft store, you end up with an icon on the desktop
called “Ubuntu 20.4”. Double-clicking this icon brings up
an Ubuntu window in four seconds.
From then on, you can swap between Windows 10 programs, and the Ubuntu window instantly. The main reason I wanted to install Ubuntu was so I could run a small
bash script that compresses Raspberry Pi SD Card images.
I use several Raspberry Pis for different activities, like
gathering solar generation and uploading it to www.
pvoutput.org, monitoring water levels in tanks, switching
a bore pump on/off as required etc.
The bash script “pishrink.sh”, which only runs on Linux,
will compress the image from a RPi 16GB or 32GB SD card
down to about 4GB in size. This 4GB image file can then
be stored on a hard disk.
Every so often, a Raspberry Pi stops with a corrupted
SD card. The easy fix is to use the official RPi imager software to copy the compressed backup image from the hard
disk onto a new SD card, plug the SD card in to the RPi,
and all is good.
While WSL2 is good, it has limitations. A recent enhancement is the addition of the “--mount” option, which allows
a hard disk that has been formatted to ext4 (one of the main
Linux file system types), connected to a SATA port on your
computer, to be accessible from Windows.
Unfortunately, Microsoft has made the “--mount” option
only available for Windows 11, not Windows 10 (except
for the people using the Windows 10 Insider Build Versions). The end result is that you are forced to upgrade to
Windows 11 to use this option.
As a test, I installed a new SSD on this PC and loaded
Windows 11 onto it. I was able to connect and access an
ext4-formatted hard disk, then extract the files I wanted.
Since then, I have gone back to Windows 10, as all my other
PCs use Windows 10, and I have limited need to access an
ext4 hard disk any more.
Sid Lonsdale, Cairns, Qld.
A trick for finding solder bridges
I am currently assembling the RGB LED ‘Analog’ Clock
from May 2025 (siliconchip.au/Article/18126), and soldering the 60 RGB LEDs is indeed a fun job.
I used an old trick to check for almost-invisible solder
bridges between the slim pads that I thought might be a
help to other readers. A sharp blade slid between the pads
will easily let you feel if there is any bridge present. If there
is, it can be removed with solder wick.
David Coggins, Beachmere, Qld.
Building a replica Philbrick K2-W valve op amp
I noticed some errors in the circuit diagram of the
Philbrick K2-W valve op amp in the article on the History of Op Amps (August 2021 issue; siliconchip.au/
Article/14987). The 500pF capacitor and parallel resistor are actually returned to ground (0V), not the -300V
rail as shown.
There’s a second 7.5pF capacitor missing, which connects between the top of the 2.2MW resistor (the control
grid of the third triode) and the top of the 120kW resistor
(the cathode of the fourth triode).
I was already working on this circuit, so your article
arrived at a fortuitous time. I’ve made a copy of the device
with modern closer-tolerance and more stable resistors
(the original were 10% carbon types). I started with Philbrick’s original 1953 circuit and made some minor resistor changes.
It worked (first time!) surprisingly well, configured as
a -2× inverting amplifier with Philbrick’s suggested offset
null “Bias” control. The output is close to the expected
values, within the limits of component tolerances and
measurement error.
Next, I was able to obtain a couple of original devices.
One example worked with limited accuracy, the other was
way off. I cut open the latter device; it turned out to be
built in the point-to-point style, typical of the era. Do you
know of anyone else who’s replicated this historic device,
or is mine unique?
Godfrey Manning G4GLM, Edgware, UK.
Comment: the 500pF capacitor and 9.1kΩ resistor going
to -300V instead of 0V was our error in redrawing the circuit, but the original circuit we have doesn’t show the
7.5pF capacitor.
We wonder if it was a change made in later production
versions, or perhaps on request by the customer for extra
stability, depending on the device’s configuration in use.
It may have been required for stability with lower gain, but
would have limited the bandwidth in higher-gain applications.
The left photo shows the replica K2-W valve op amp, and the adjacent photo is an internal shot of an original Philbrick
K2-W after it was removed from its housing.
8
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
Modern op amps are sometimes available in a unity-
gain-stable version and higher-bandwidth version, with
the only difference being whether a dominant pole capacitor like this is present. For example, the OPA134 (8MHz,
unity gain stable) is very similar to the OPA604 (20MHz,
minimum gain of five times).
Good experiences with JLCPCB assembly
I’ve just read the editorial in the June 2025 issue. I know
of the problem with the bad CH334F chips, being an active
member (and administrator) on The Back Shed forums.
I understand where you are coming from, and I totally
agree – it could have been a disaster from your board-
ordering point of view.
Having said that, I think that you were simply unlucky
with the timing. JLCPCB have been stellar in all my SMD
assembly orders with them (hundreds now), and of all the
boards I have ever ordered, there was never an SMD assembly problem. The CH334F problem does appear to be the
exception, rather than the rule, and it was unfortunate that
you happened to be caught up in it.
Don’t let that put you off ordering assembled PCBs, as
99% of the time, it goes well. The other way to ensure you
are not bitten by any potential problems such as this is to
order smaller batches, instead of putting all your eggs in
one basket, so to speak.
That also helps to spread out the budget for the purchase
of assembled boards, if you get them in several smaller
batches, which is what I have been doing for years now.
Love the magazine, keep up the good work.
Graeme Rixon, Rictech Ltd, New Zealand.
We’re still lagging behind in HD TV broadcasting
Monochrome TV broadcasting commenced in Australia
in 1956, with colour added in 1975. The conversion to digital broadcasts in 2001 required the purchase of a new TV
or a set-top box. Most broadcasters had to buy new transmitters, along with studio equipment.
New Zealand started converting from analog PAL TV to
DVB-T with MPEG-4 compression in 2011 and finished
in 2013. In Australia, Freeview 2022 made HEVC (H.265)
decompression compulsory, but not DVB-T2. Satellite coverage for blackspots still uses the older DVB-S standard,
which only supports standard definition (SD) video.
With MPEG-4, HD video takes the same bandwidth as SD
video encoded using the older MPEG-2 format. HE-AAC
audio provides a similar reduction in data usage compared
to the older MP2 scheme.
DVB-T can transmit around 23Mbit/s, whereas DVBT2 supports up to 34.5Mbit/s in the same 7MHz bandwidth. Unfortunately for viewers, older TVs and boxes
are not compatible with DVB-T2, so a new TV or set-top
box is required.
For broadcasters, around 2778 transmitters would need
their modulators replaced; fortunately, the expensive highpower sections are unaffected.
The AS 4933:2015 standard already requires support for
MPEG-4 video and HE-AAC v2 audio. Yet the industry has
dragged its feet for nearly a decade, with the first complete
MPEG-4 conversion only occurring in Tasmania in October 2023. WIN TV and Seven Regional have now upgraded
most areas except remote Australia, Sydney, Melbourne,
and Brisbane.
siliconchip.com.au
Australia's electronics magazine
August 2025 9
SBS now uses MPEG-4 nationally, except for Channel 3, which remains SD. The ABC has been broadcasting MPEG-4 HD on Channel 20 for some time, and will
convert ABC News and ABC Entertains to MPEG-4 HD
in June 2025.
Once Seven completes the MPEG-4 conversion in the
remaining cities, there will be little justification for simulcasting the same primary programs in blurry SD MPEG-2 on
the ABC, SBS, and Nine Network. Network Ten is already
broadcasting in HD on Channel 1.
A range of updated standards now governs digital TV
in Australia:
• AS 4933:2024 applies to TVs and set-top boxes sold in
Australia. These must support MPEG-4 video and HE-AAC
audio.
• AS 4599:2025 defines transmission characteristics for
both DVB-T and DVB-T2.
• AS 5362:2024 covers DVB-T2 receivers, which enable
Ultra HD (UHD) or more HD channels over the same 7MHz
bandwidth.
• Other standards relate to outdoor antennas (AS 1417)
and coaxial/MATV systems in buildings (AS 1367).
These new standards will replace the 2015 versions,
which become obsolete in November 2026. For most
viewers, this means their current equipment is likely
compatible. However, some older TVs or boxes may need
replacement.
Viewer Access Satellite Television (VAST) provides
free-to-air coverage in remote areas and black spots
via Optus satellites. These use DVB-S2 and HEVC, and
receivers must meet government-specified standards. The
system is funded by the Commonwealth Government.
The latest version of the ETSI EN 302 307-2 (2024-08)
standard extends DVB-S2 to support UHD broadcasts via
satellite. This is a global standard used widely across
Europe and elsewhere.
While we are still upgrading to standards from around
2015, much of the world has moved on. DVB-T2 is already
used in 84 countries, covering 3.6 billion people.
Modern TVs often already support DVB-T2 and HEVC.
You can now buy DVB-T2/HEVC-capable set-top boxes for
around $45, and new VAST receivers with DVB-S2 and
HEVC support are also available.
If we upgraded 16 sites of five transmitters each per year,
we could complete a national switchover in 10 years. Each
broadcaster would also need to update its program feed with
HEVC video and E-AC-3 (Enhanced AC-3) audio encoders.
After the switchover, support for DVB-T could be phased
out. HEVC is backward-compatible with MPEG-4, so this
transition could be managed smoothly.
Large-screen UHD (4K) TVs have been available for several years, and retailers often use high-resolution promotional videos to sell them. It’s not always clear which TVs
can receive DVB-T2 broadcasts, even though most can
stream UHD video-on-demand thanks to built-in HEVC
decoders.
Australia has now been using DVB-T for over 20 years,
making an upgrade to DVB-T2 increasingly urgent. Instead
of being technology leaders, broadcasters have fallen
behind. It’s now up to electronics manufacturers to drive
progress, and the broadcasters must catch up.
SC
Alan Hughes, Hamersley, WA.
icomretail.com.au
10
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
Last month, we introduced
SpaceX’s Falcon 9, Falcon
Heavy, Super Heavy and
Starship launch vehicles and
described their engines and
capabilities. This second and
final instalment will cover
their launch sites, some of
the more notable missions
and what they are planning
for the future.
Part two
by Dr David Maddison
VK3DSM
Starship’s
seventh
test
flight
Image source: SpaceX / <at>
Space_Time3 via X (Twitter).
siliconchip.com.au
Australia's electronics magazine
August 2025 11
Fig.30: a rendering of what SpaceX’s
HLS might look like on the Moon.
Fig.31: a rendering of the lunar
Starship version with landing legs.
Fig.32: a concept from 2019 for a
Starship CLPS vehicle.
hen we left off in the previous issue, we had just
described how Starship is
launched atop the massive Super
Heavy launch vehicle, powered by 33
Raptor engines.
While Starship is still in the testing
phase, it is intended to be able to deliver
cargo and crew to the Moon and ultimately, Mars. It may even be refuelled
in orbit, allowing a much heavier cargo
to be sent to distant planets.
After we look at some of these aspects
of Starship, we’ll go through some of the
more notable SpaceX missions to date,
then look at two of their larger competitors and what they have done lately.
Like last month, uncredited images are
from SpaceX or public domain sources.
The main variants of Starship envisioned are the Human Landing System (HLS), for landing on the Moon
(Figs.30 & 31), the propellant tanker
(see Fig.35), the propellant depot
(Fig.36) and a cargo version.
The version of Starship intended
for Mars settlement will have heat
shielding and flaps for guidance – see
Figs.33 & 34.
The CLPS Lander
will refuel. It turns out that one reason SpaceX chose methalox as a fuel
is that it can be manufactured on Mars.
Methane fuel and oxygen for oxidiser can be produced on Mars from
CO2 in the atmosphere and hydrogen from water, which is now known
to exist on Mars beneath the surface
and elsewhere. The reaction used to
make methane is the Sabatier reaction,
CO2 + 4H2 → CH4 + 2H2O. The fuel
could be manufactured using electricity from solar energy or nuclear
reactors.
Hydrogen can also be extracted from
water by electrolysis, which provides
a supply of oxygen at the same time.
Alternatively, the hydrogen could possibly be transported from Earth in a
tanker spacecraft.
Starship fuel depots could also be
sent from Earth and placed in Martian orbit to later fully refuel Starship
for a return trip. The fuel sent would
be methane and oxygen. Or hydrogen
could be transported for manufacturing methane on the Martian surface.
The Perseverance rover, which
landed on Mars in 2020, successfully
W
The SpaceX Commercial Lunar
Payload Services (CLPS) lander is a
part of a contract to NASA to provide
lander services to deliver payloads to
the Moon as a precursor to landing
astronauts on the Moon. Payloads have
already been delivered to the Moon
using Falcon 9 rockets. SpaceX has
also proposed a Starship variant for
these missions (see Fig.32).
How will Starship get to
Mars, land and leave?
The most likely way Starship will
go to Mars is as follows. Starship will
be launched into Earth orbit and then
be refuelled from a Starship tanker or
fuel depot. Then, an energy-efficient
path known as a Hohmann transfer
orbit will be used to take Starship to
Mars in 7–10 months.
Starship will enter the Martian
atmosphere using aerodynamic drag
to slow down, then flip to a vertical
position for a propulsive landing using
its Raptor engines.
Once landed on Mars, there is a lot
of speculation about how Starship
Fig.33: an artist‘s impression of Starships at a Martian
settlement.
12
Silicon Chip
Fig.34: another artist’s concept of a Mars settlement.
Source: www.spacex.com/updates/
Australia's electronics magazine
siliconchip.com.au
Fig.35: a proposed method of inorbit refuelling of Starship.
Fig.36: refuelling in orbit from
another stripped-down Starship.
Fig.37: the glass-coated silica-fibre
tiles that protect Starship’s exterior.
performed the Mars Oxygen In-Situ
Resource Utilization Experiment
(MOXIE) to produce oxygen from the
Martian atmosphere, although not
methane. Carbon monoxide (CO) is a
byproduct of the reaction used in that
experiment; it can be reacted with
water or hydrogen to produce methane.
Starship more tolerant of a failure of
the heat shield than the Shuttle was.
The heat shield on the Dragon capsules is phenolic-impregnated carbon
ablator (PICA-X). The material ablates
or burns away, carrying excess heat
with it.
SpaceX also coats most vehicles with a heat-resistant, protective white paint for thermal control,
thought to be a formulation known
as AZ-93 (www.aztechnology.com/
product/1/az-93).
complexity of landing legs on the
booster. However, landing legs will be
used for landing Starship on the Moon
and Mars, at least until a Mechazilla
is built in those places.
Starbase has two Orbital Launch
Mounts; Starships intended for re-
entry to Earth will not need landing
legs.
Thermal protection systems
For re-entry, Starship uses several
types of thermal protection:
1. Its silica-fibre-based hexagonal
tiles can withstand a temperature
of 1400°C; they are similar to what
the Space Shuttle used and have
a similar consistency to Styrofoam. They are coated with a special heat-resistant black glass layer
(see Fig.37). There are 18,000 tiles,
which is 6,000 fewer than the Space
Shuttle used.
2. There is a secondary ablative layer
under the primary tile heat shield
for extra protection.
3. The Starship skin is made of stainless steel, which is far more resistant to heat than the aluminium of
the Space Shuttle, and will make
Launch pad & recovery
Due to the enormous power of the
Starship engines, a lot of damage was
done to the launch pad and surrounding area in early tests, requiring modification of the launch support structure.
Fig.38 shows the water deluge system
(flame deflector) used to absorb some
of the energy of the rocket exhaust.
For recovery, the Super Heavy
booster is caught in the “chopsticks” of
the Orbital Launch Mount or “Mechazilla” launch tower, in a remarkable
feat of guidance and control. This is
done to avoid the extra weight and
Fig.38: a full pressure test of Starship’s launchpad flame
deflector on the 29th of July 2023.
siliconchip.com.au
Spaceports
Starbase in Boca Chica, Texas
(Fig.39) is the main site for launching
the Starship rockets, including those
that will be launched to the Moon
and Mars. It is also the headquarters
of SpaceX, and a production and test
site for Starship.
Apart from Starbase, the other
launch sites used by SpaceX are:
Kennedy Space Center (Launch Complex 39A or LC-39A, leased from
NASA) in Florida – previously used
for the Apollo and Space Shuttle programs. It is now used by SpaceX, mostly
for Falcon Heavy launches, including cargo and crewed missions with
Dragon, and more complex missions.
Fig.39: part of Starbase, showing a Starship on display.
Source: SpaceX.
Australia's electronics magazine
August 2025 13
Fig.40: a Falcon Heavy being prepared at Vandenberg
Space Force Base. Photo by Jack Beyer via X.com.
Cape Canaveral Space Force Station
in Florida has multiple launch pads,
including Cape Canaveral Space
Launch Complex 40 (SLC-40), which
has been leased and upgraded by
SpaceX since 2007 for launching
Falcon 9 rockets.
It has made at least 230 launches. It
launched its first crewed mission in
September 2024. It also has landing
pads for Falcon 9 and Falcon Heavy
reusable boosters: Landing Zones 1
and 2 (LZ-1 and LZ-2).
Vandenberg Space Force Base (Space
Launch Complex 4 or SLC-4E) is in
California, and is used to launch satellites into polar orbits of the Earth and
Sun-synchronous orbits using Falcon
9 and Falcon Heavy (see Fig.40).
Fig.41: a Falcon 9 lands on the 52 × 91m platform of a
drone ship off the coast of the Bahamas.
Drone ships
The drone ships used for Falcon 9
and Super Heavy booster recoveries are
ocean-going barges, correctly known
as autonomous spaceport drone ships
(ASDSs) – see Fig.41. They have been
made autonomous for the recovery of
Falcon 9 boosters. The landing platform
is about 52 × 91m, while the Falcon 9
v1.1 landing leg span is 18m.
They are towed into position with a
tug, then kept in place by autonomous
station-keeping. After a landing, crews
board the ASDS and secure the rocket.
One of the ASDSs uses a robot called
the “octagrabber” to secure it.
Why not use parachutes?
Port Canaveral in Florida is used as a
base for the drone ships that operate
in support of booster recoveries in the
Atlantic Ocean from launches at Kennedy Space Center and Cape Canaveral
Space Force Station.
The Port of Long Beach is a base for
the drone ship doing recoveries in the
Pacific Ocean from Vandenberg Space
Force Base.
The Space Shuttle used parachute
recovery for its main boosters, so why
does SpaceX use propulsive recovery,
which is much harder to perfect?
The difference is that the Shuttle
jettisoned its boosters at a relatively
low altitude and speed, whereas the
SpaceX boosters are not jettisoned
until near orbital velocity. The speed
and energy involved preclude a parachute recovery.
The second stage of Falcon is not
reused, as it’s too complicated to
Fig.42: deploying Starlink satellites.
Source: NASAspaceflight.com
Fig.43: the Sora-Q mini-rover from
Hakuto-R. Photo by テレストレラッソ.
14
Silicon Chip
Australia's electronics magazine
recover. That’s a reasonable compromise because the second stage is a relatively simple and inexpensive structure. The trunk of the Dragon capsule
is not recovered either.
Unlike the Space Shuttle, which
was more what you might call ‘refurbishable’ than ‘reusable’ (it cost about
as much to refurbish between flights
as building a new one), the SpaceX
boosters are economically reusable.
From a cost point of view, the Shuttle was a disaster, but the genuine
reusability of the SpaceX boosters
helps to significantly reduce the cost
of launches.
Very little needs to be done to a
landed booster for its reuse. It’s pretty
much just checked over and refuelled,
then it is ready to go!
Starlink’s role in SpaceX’s
operations
According to the video at https://
youtu.be/lgt4zSD9UUc, SpaceX plans
to use the Starlink satellite network to
maintain communications with Crew
Dragon capsules during the re-entry
phase when the plasma layer surrounding the vehicle normally causes
a communications blackout.
Fig.44: the Intuitive Machines-1
Odysseus lander.
siliconchip.com.au
Fig.45: the Intuitive Machine-2 Athena lander carries the
Micro Nova Hopper. Source: Intuitive Machines.
There is no other published information that we could find about the
extent to which SpaceX platforms use
or do not use Starlink.
Notable SpaceX missions
Hakuto-R Mission 1
On the 11th of December 2022, a Falcon 9 was launched to deliver the Japanese Hakuto-R Moon lander (Fig.43),
but unfortunately, an error in the lander’s radar altimeter caused it to keep
hovering at an altitude of 5km until it
ran out of fuel and crashed.
Hakuto-R Mission 2 (Resilience)
Hakuto-R Mission 2 was launched
on the 15th of January 2025 to deliver
a payload to the Moon, including a
lunar micro rover developed by ispace
as a technology demonstrator for reliable transportation and data services
on the Moon. This mission shared the
same Falcon 9 launch vehicle as Blue
Ghost Mission 1 (see below).
Intuitive Machines-1
On the 15th of February 2024, a
SpaceX Falcon 9 launched the first
commercial mission to successfully
soft-land on the Moon. It was also the
first American-made spacecraft to land
Fig.46: Polaris Dawn launched in the dark, carrying Jared
Isaacman, Scott Poteet, Sarah Gillis & Anna Menon.
on the Moon since the 1972 Apollo mission. The Odysseus lander (Fig.44) carried a variety of instruments. It landed
on its side, but the instruments functioned and it was judged a success.
Intuitive Machines-2
Also known as Polar Resources Ice
Mining Experiment-1 (PRIME-1), this
lander, called Athena (Fig.45), was
launched on the 27th of February 2025
using a Falcon 9 rocket and landed on
the Moon on the 6th of March.
It carried The Regolith and Ice Drill
for Exploring New Terrain (TRIDENT),
to drill for ice as a source of water for
future habitation. The MSolo mass
spectrometer was included to measure
the amount of ice in the drill samples,
as well as the Micro Nova Hopper.
Unfortunately, the mission failed as
the spacecraft landed on its side, like
the Odysseus mentioned above.
The Polaris program
Polaris (https://polarisprogram.
com/) is a private space flight program established by Jared Isaacman,
now nominated to be the next NASA
Administrator.
The program was established under
a contract with SpaceX. Isaacman’s
first flight as a private astronaut on
a Crew Dragon spacecraft was on the
16th of September 2021, to raise money
for St. Jude Children’s Research Hospital.
The first flight under the Polaris
program was on the 10th of September
2024, on Crew Dragon, taking the occupants to an apogee of 1400km, higher
than any human has been in orbital
flight since the flight of Gemini 11 in
1966 (with an apogee of 1368km) – see
Fig.46. Two other flights are planned
under the Polaris program.
Blue Ghost Mission 1
On the 2nd of March 2025, Firefly
Aerospace’s Blue Ghost Mission 1
lander landed on the Moon, having
been launched by a SpaceX Falcon 9
(see Fig.47). Among ten science investigations that spacecraft will perform
will be receiving GPS signals using
the Lunar GNSS Receiver Experiment
(LuGRE) to investigate extending the
navigational capability of GPS to the
Moon and beyond.
We wrote about using GPS beyond
Earth orbit, including near the Moon,
in our October 2020 issue (siliconchip.
au/Article/14597).
There are also the Next Generation
Visiting Starbase
What is Max Q?
As of the time of writing, you can
visit Starbase and the surrounding
areas. We suggest you look at the
following links if you want help planning a trip to go there:
During a rocket launch, including
those of SpaceX, one often hears the
expression that the vehicle is going
through Max Q (or “max q”).
This is the time of maximum
aerodynamic drag on the vehicle
and maximum stress, when something is most likely to go wrong.
The engines are frequently throttled
back during Max Q to minimise the
structural load on the vehicle.
• https://siliconchip.au/link/ac5m
• https://siliconchip.au/link/ac5n
• https://everydayastronaut.com/
how-to-visit-Starbase/
• https://siliconchip.au/link/ac5o
siliconchip.com.au
Fig.47: a rendering of the Blue Ghost
lander on the moon’s surface.
Australia's electronics magazine
August 2025 15
Fig.48: at 1.2m in diameter, the Dragon cupola is the largest Fig.49: four astronauts wearing Starman suits in the
Dragon capsule to protect against depressurisation.
window in space, made from layers of polycarbonate.
Retroreflectors (NGLR), targets for
Earth-based lasers to accurately measure Earth-Moon distances. The first
laser retroreflectors were placed on
the Moon by Apollo 11 astronauts in
1969, followed by Apollo 14 (1971)
and Apollo 15 (1971). They are still
in use today.
This mission shared the same Falcon 9 launch vehicle as Hakuto-R
Mission 2, launching on the 15th of
January 2025. This was the first commercial venture to fully successfully
land a spacecraft on the Moon.
International Space Station
rescue mission
Due to technical problems with the
Boeing Starliner that was docked with
the ISS, astronauts Butch Wilmore and
Suni Williams were unable to return
to Earth at their scheduled date of the
14th of June 2024 (their mission was
originally meant to last for eight days).
The problems with Starliner were
not solvable in any reasonable amount
of time, so SpaceX offered a rescue
mission but that offer was not accepted
by the previous US Administration.
However, the new US Government
accepted the offer, and they launched
a rescue mission on the 14th of March
2025, docking on the 16th.
The spacecraft was a Crew Dragon
launched by a Falcon 9. It delivered four new astronauts and finally
returned to Earth on the 18th of March
2025, carrying Wilmore, Williams
and two others. The full video of the
re-entry and splashdown is available at www.spacex.com/launches/
mission/?missionId=crew-9-return
For a shorter version of the
video, see https://x.com/SpaceX/
status/1902116771806732511 or
https://youtu.be/fd-bMz4fGN4
Fram2
Fram2 was a private mission paid
for by Maltese billionaire Chun Wang.
He and several of his guests, including
Australian Eric Philips, were launched
by a Falcon 9 on the 31st of March 2025
and they splashed down in the Pacific
Ocean on the 4th of April.
After stage separation, the booster
landed on the drone ship named “A
Shortfall of Gravitas” in the Atlantic
Ocean.
Their Dragon capsule was inserted
into a polar retrograde orbit, the first
time astronauts have ever been put
into polar orbit. The capsule communicated with Starlink via a laser beam,
Fig.50: a Crew Dragon with its Trunk attached prepares
to dock with the ISS. The white part is the IDA.
16
Silicon Chip
the same way Starlink satellites communicate with each other. A cupola
for viewing was placed beneath the
nose cone (Fig.48), in the area normally used for docking with the ISS
and exiting Dragon.
There was an amateur radio station
onboard transmitting SSTV (slow scan
TV on 437.550MHz) images as part of
a high school and university competition. Among a variety of 22 experiments, the crew took the first x-ray
of a human ever in space. The mission websites are https://f2.com/ and
https://fram2ham.com/ plus there is a
video at www.spacex.com/launches/
mission/?missionId=fram2
Dragon to the Moon
As an alternative to the hugely
expensive, delayed and problematic
Boeing Space Launch System (SLS)
and Lockheed Martin Orion spacecraft
for landing people on the moon, Dr
Robert Zubrin of the Mars Society and
Homer Hickam have suggested sending a modified Crew Dragon, incorporating features from Red Dragon,
to the Moon.
This mission would involve both the
Falcon 9 and Falcon Heavy, but there
would be no landing. That mission
Fig.51: the Dragon capsule as it was about to dock with the
ISS on the Crew-5 mission.
Australia's electronics magazine
siliconchip.com.au
would resemble Apollo 8 (1968), orbiting the moon but not landing.
An alternative mission that would
involve landing would be to launch
Crew Dragon into low Earth orbit,
with astronauts then transferring to
Starship HLS (which never lands on
Earth), fuelled in Earth orbit, to land on
the Moon. The return to Earth would
be a reverse of that.
Space suits
The Starman suit, also known as the
intravehiclar activity (IVA) suit, is custom made for the astronaut who will
wear it; the helmets are 3D-printed to
the required shape. This suit protects
against depressurisation only; it has
no radiation protection, so it cannot
be used outside the spacecraft. Astronauts regard these suits as very comfortable.
Astronauts can be seen wearing
these suits in Fig.49.
SpaceX also has a space suit for
extravehicular activities (EVA). This
suit is also suitable for use inside the
spacecraft, and among its many features is a heads up display within the
helmet to display parameters such as
pressure, temperature and humidity
etc – see Fig.52.
Docking adaptors
With increasing space activity, it
is important to have standard docking interfaces between spacecraft.
One standard is the International
Docking System Standard (IDSS).
The NASA Docking System (NDS) is
NASA’s implementation of this system; it is used on the ISS, the Boeing
Starliner, the Orion spacecraft and
Crew Dragon 2.
The ISS used to use the Russian-
developed docking standard of APAS95 (as did the Soyuz, former Space
Shuttle and former Mir space station),
but the International Docking Adapter
(IDA) was brought to the ISS by SpaceX
Dragon and used to convert those
adaptors to the NASA Docking System,
which complies with the International
Docking System Standard. An IDA is
shown in Fig.50.
Fig.51 depicts a Dragon capsule
as it is about to dock with the ISS.
Note the Draco thruster firing and the
open docking hatch (nose cone) of
the Dragon with the docking interface
inside. If you want to try your hand at
docking with the ISS with a simulator,
visit https://iss-sim.spacex.com/
Starlink
Starlink is a subsidiary of SpaceX,
with SpaceX launching thousands of
Starlink satellites to provide satellite-delivered internet services almost
worldwide (and now telephony). Starlink can provide download speeds
of up to 200Mbps, with uploads of
10–40Mbps and latencies of 25–80ms.
As of the 27th of February 2025,
there were 7052 working Starlink satellites in orbit at about 550km altitude
(see Fig.53). They can be seen at night
with the naked eye, making them a
concern to astronomers.
SpaceX has permission to launch a
total of 12,000 satellites (Fig.54), and is
seeking permission to increasing that
number to as high as 30,000.
On the 5th of December 2024, Elon
Fig.53: the Starlink constellation at the time of writing.
Source: https://satellitemap.space/?constellation=starlink
siliconchip.com.au
Fig.52: the SpaceX EVA suit can be
worn outside a spacecraft.
Musk wrote, “The first Starlink satellite direct to cell phone constellation is now complete. This will
enable unmodified cellphones to
have internet connectivity in remote
areas.” (https://x.com/elonmusk/
status/1864571206004838425).
For more about Starlink, see our article on it in the June 2023 issue (https://
siliconchip.au/Article/15815).
SpaceX’s software
SpaceX’s software (and hardware)
obviously must be reliable, especially those used for flight operations.
They use Linux-based systems for
flight computers; flight software and
other systems are written in C++. A
stripped-down version of Linux is
Fig.54: a depiction of Starship delivering the next
generation of Starlink satellites.
Australia's electronics magazine
August 2025 17
used; it is tailored to the demands of
spaceflight.
SpaceX maintains its own Linux
kernel with the PREEMPT_RT patch
installed to enable real-time processing for applications like engine control and navigation (standard Linux is
not real-time capable). They also use
custom drivers.
The flight software runs on triply
redundant dual-core x86 processors,
all performing calculations in parallel. If the result of one core disagrees
with the others, it is ignored. This provides fault-tolerance without having
to use expensive radiation-hardened
computers.
LabVIEW by National Instruments
is used for data logging and monitoring of various parameters. A variety
of different software is used for web
applications. For Enterprise Resource
Planning (ERP), they use a proprietary
system called WARPDRIVE for all sorts
of day-to-day management functions.
Siemens NX is used for computer-
aided design (CAD), engineering analysis and manufacturing processes. It
creates 3D models and can perform
simulations to predict performance,
including structural analysis and aerodynamics.
Teamcenter is used for managing
product data such as CAD files, documentation, CNC code etc. It maintains revisions and allows collaboration between different departments.
NX and Teamcenter operate together
and help reduce SpaceX’s costs and
improve reliability of products.
• Is developing the New Glenn heavylift orbital launch vehicle.
• Is involved in the Blue Moon
human-capable lunar lander for the
NASA Artemis program, which can
land people and 3600–6500kg of
cargo to the lunar surface (depending on version).
• Is working on the Blue Ring spacecraft for refuelling, transporting and
hosting satellites.
• Is working on the Orbital Reef low
Earth orbit space station to support ten
people; it is expected to be operational
in 2027. It will support both commercial space activities and tourism.
Like Blue Origin, Virgin Galactic
(founded by Richard Branson) also
offers space tourism services. It
is believed to charge US$450,000
(~$750,000) for a sub-orbital trip into
space, with around 700 people on the
waiting list. They have made seven
commercial passenger-carrying flights,
the last being on the 8th of June 2024. It
reached an altitude of 87.5km. They’re
working on a new space plane, the
Delta-class (Fig.56).
Other private space ventures
Videos to watch
While this article has been primarily about SpaceX, there is news on two
other private space ventures involving
crewed vehicles.
Blue Origin (www.blueorigin.com)
is owned by Jeff Bezos. It is providing commercial sub-orbital passenger
flights into space on the New Shepard
sub-orbital rocket system (Fig.55). Its
last flight at the time of writing was
on the 25th of February 2025, when
it took six paying passengers to an
altitude of around 100.5km. You can
watch a replay of the flight at https://
youtu.be/zXRzcSw_bdc
The cost per passenger is unknown.
So far, they have made ten passenger
flights. In addition to space tourism,
Blue Origin:
• Produces engines for other spacecraft.
• How SpaceX Reinvented The
Rocket Engine:
https://youtu.be/nP9OaYUjvdE
• The Real Reason SpaceX
Developed The Falcon 9:
https://youtu.be/LmK18kPfMjA
• How SpaceX Reinvented The
Rocket:
https://youtu.be/7vE95eBX6M0
• Why The Raptor Engine Is Ahead
of Its Time:
https://youtu.be/6cwue7jMkww
• What Really Happened to Starship:
https://youtu.be/tlAo_6CG9o8
• SpaceX Upgrades Everything
Inside Crew Dragon:
https://youtu.be/dThdld_f0Rk
• Does the SpaceX Crew Dragon
have a toilet:
https://youtu.be/GT5Sm6v4oqo
• Lunar Lander Missions on SpaceX
18
Silicon Chip
Fig.55:
Blue
Origin’s
New
Shepard
suborbital
rocket
system.
Source:
Blue
Origin
SpaceX’s future
SpaceX has dramatically decreased
the cost of delivering cargo to space,
and will likely continue to do so. Elon
Musk’s vision is to have a fleet of rockets with a turnaround time the same
as passenger aircraft. He also wants a
fleet of 1000 Starships continuously
running 100–150 tonnes of cargo and/
or passengers into Earth orbit, the
Moon or Mars.
Australia's electronics magazine
Fig.56: Virgin Galactic’s latest Deltaclass spaceplane.
Source: Virgin Galactic
Rocket: www.youtube.com/live/
XOLnPRCpdYU
• China Tested Mechazilla
Chopstick Clone:
https://youtu.be/ohREX1PDYY0
• This Is the End of Boeing:
https://youtu.be/7f56Qldi_Fo
SC
siliconchip.com.au
August
Tech
BUYS
SAVE $34
A0323 65W
15000mAH
125
$
altronics.com.au
SAVE $420!
649
$
ONLY WHILE STOCKS LAST!
SL4576W 100Ah
100Ah LiFePO4 Slimline Lithium Battery
Ultra slim 75mm profile with full current discharge capability and a 5 year warranty. Space
at a premium in your camper, caravan or 4WD? This compact battery is perfect for remote
power solutions without taking up precious cargo space for your gear. Pre-fitted with
Anderson input and output connections and battery capacity gauge. 600 x 275 x 75mm.
Watch TV on the
go with a 12V
telly.
SAVE $100
Great picture quality!
299
$
23.6” 1080p screen and
HD tuner. Designed for
12V vehicles, 4WDs,
caravans and campers.
Easily wall mountable
with VESA mounting
points. USB PVR
recording in-built .
Includes
travel
adapters
SAVE $20
79
$
Handy USB
Note Taker
A0319A 20W 10000mAH
Record CD quality audio
with excellent audio pick
up for taking recording
interviews & more.
16GB storage.
Recharge anywhere you go!
Introducing our new global travel power banks. Available in either
20W or 65W power delivery with in-built mag safe compatible wireless
charging they come with travel adaptors for travel all over the world. All
supplied in a handy zip up case.
Wi-Fi RGB Strip
Lighting Kit
This kit includes 5m of RGB
strip lighting, power supply
and controller unit allowing
you to create colourful lighting
effects around your home.
Controller features a music
sensor input allowing the
lighting to trigger to music
being played in the room.
Great for home entertaining.
Works with Alexa and Google
Assistant. 60 LEDs per metre.
S 8866
75
With beat
triggering!
79
$
MacBook Docking Hub
SAVE $24
$
D 2367
SAVE $20
X 0705A
Plugs into MacBook Pro/Air USB C connections and provides a full suite of peripheral
ports. 100W PD passthrough charging.
Mini BT Speaker
& Wireless
Charger
SAVE 27%
X 3227*
55
$
A MagSafe compatible
charging pad for
iPhone and Apple
Watch, plus a
D 2035*
Bluetooth
SAVE 14%
speaker
with LED
$
lighting.
30
SAVE $10
M 8868B
39.95
$
65W 3 Port Mains Charger
High efficiency compact GaN charger
with 65W PD and QC3.0.
Your electronics supplier since 1976.
Shop in-store at one of our 11 locations around Australia:
WA » PERTH » JOONDALUP » CANNINGTON » MIDLAND » MYAREE » BALCATTA
VIC » SPRINGVALE » AIRPORT WEST QLD » VIRGINIA NSW » AUBURN SA » PROSPECT
Sale ends August 31st 2025.
Build It Yourself Electronics Centre®
Hot AV Deals.
C 5162
SAVE $59
220
$
C 5161
SAVE $40
149
$
14” Go-Anywhere
Portable Digital TV
SAVE $80
209
Perfect for the car or
caravan! Powered off internal
rechargeable battery, your
vehicle battery or mains
plugpack. Also fitted with USB
connection for recording TV.
$
With internal battery
- use it anywhere!
S 8864A
TOP
VALUE!
Why pay
$300 or
more?
C 9044
HALF PRICE!
20
$
Boom Box & Wireless PA Systems
Need instant sound for your next big get together?
Pick up a new Bluetooth entertainer box - available in small or
large systems. They not only sound great, but offer a wireless
micro phone for PA use, plus TWS pairing to a second unit (of the
same type) for added volume. Offers up to 3-8 hours use from a
single charge (depending on volume).
Affordable
Bluetooth
Earbuds
Flexible Sports Headphones
Bluetooth 5.0 for great range and audio
quality. Great sound and amazing 16 hour
battery life . Comfortable & compact design
Got a serial
earbud loser
in the family?
Buy them these
affordable BT
5.0 earbuds with
great sound
SAVE 35%
- for less!
C 9032B
25
$
A 2655
SAVE $40
SAVE $30
99
229
$
$
A 2548
HOT
DEAL!
USB Gooseneck
Microphone
Great for gaming,
livestreaming or work.
Quality omnidirectional
mic insert. Mic gain and
mute control knob with
LED lighting.
HALF PRICE!
4 Ch. USB Mixer With EQ & FX
Want to get into recording podcasts, voice overs
or making your own audio samples? This mini
USB mixer connects directly to your PC or Mac
and is powered directly from USB.
14 Channel Mixing Desk With Bluetooth
A compact mixer with top mount inputs for easy connection.
Features combo XLR & 6.35mm jacks, DSP effects unit, USB MP3
recording and much more! Its a great small to medium venue mixer
at an affordable price!
Top
quality
audio!
D 0981
34.95
$
SAVE 25%
44
$
A 2809A
12V/240V HD Set Top Box
This mini digital TV receiver features HDMI
output for connection to any monitor. Runs off
a 12V power source making it perfect for use in
caravans etc. USB recording & playback. Includes
plugpack, car adaptor & IR remote.
Includes
10m cable &
hardware
L 2003
A 3217D
SAVE $17
SAVE $30
78
99
$
$
Caravan/Boat Television Antenna
Run 4K HDMI over Cat5/6 Cable
Thousands sold. Get crystal clear TV reception
wherever you travel! Omnidirectional 360° design
requires no adjustment .
Transmits 1080p up to 70m and 4K<at>30Hz up to 40m with
zero delay and uncompressed transmission. Supports audio
formats up to PCM 7.1, Dolby True HD and more.
SCAN TO FOLLOW US!
Stay up to date on latest releases,
exclusive specials and news on
our socials.
Like our service?
Review your store on Google.
Every review helps us serve you better.
Power &
charge.
AV Sav-
100Ah
SAVE $145
580
$
SL4576BT
120Ah
SAVE $300
699
$
X 0604C
Get powered up for
spring camping!
SL4578BT
M 8070B
SAVE $19
SAVE 17%
50
33
$
$
Bluetooth
Car Audio Adapter
Works as a hands free system for
calls plus plays audio from your
phone. USB C 20W PD & QC3.0
chargers in-built.
240V from your cup holder.
Simply plugs into a 12V cigarette lighter
socket & provides mains power!
150W rated (300W surge).
Power up your campsite
with premium lithium batteries.
Bluetooth wireless monitoring in-built.
LiFePO4 rechargeables offer longer service life & weigh HALF as much as SLA
batteries. LiFePO4 also provide more usable life per cycle, allowing for longer run
times. All have high discharge battery management system on board for safe and
reliable use. 5 year warranty.
M 8538
Now with
15W
charging!
SAVE $13
D 0521
46
SAVE $40
$
125
A 0292
$
SAVE $20
79
$
SAVE 17%
With
LED
torch
D 2322A*
29
$
110W Laptop Battery Bank
A 0291
Surprisingly compact battery bank with
USB power delivery over 100W for
large laptops! Features 2 PD Outputs
and 2 QC3.0 outputs for simulataneous
charging. Total capacity 30,000mAh.
SAVE 33%
Lithium/NiMH Chargers
Ideal for lithium cells from 10340 to
26650 in size and NiMH cells from
AAAA to C size. USB powered.
SAVE 40%
9
Build wireless
charging into your desk
Maintain & Protect Your
LiFePO4 Battery
Great for the workbench! An ultraslim desk mount 15W wireless fast
charger. Needs 60mmØ hole. Includes
power adaptor & USB cable.
Suitable for charging and maintaining
12.8V LiFePO4 batteries from mains
power at home. Suits permanent
connection. Includes ring terminal
and crocodile clips.
SAVE 27%
M 8625A
19.95
$ .95
46
$
SAVE $10
29
$
$
SAVE 24%
9
$
M 8632A
M 8628B
3xUSB Car Charger
30W USB PD Car Charger
Ultra small USB
charger!
Keep everything charged up
in the car with this handy 7.2A
triple USB charger.
Huge 30W PD output from a tiny
car charger! QC3.0 plus USB type
C power delivery.
Amazingly small with
2.4A USB output from
type A ports.
SAVE 40%
P 0695
USB Panel Mount Socket
With volt meter! Aluminium
body suits 29mm panel hole.
12-24V DC input.
P 0668A
USB Charger With Fascia
Great for 12V vehicles. 2.1A max
output. Suits 29mm panel hole.
NO STRESS 30 DAY RETURNS!
GOT A QUESTION?
Not satisfied or not suitable? No worries! Return it in
original condition within 30 days and get a refund.
Ask us! Email us any time at:
customerservice<at>altronics.com.au
Conditions apply - see website.
15
$
Useful Gadgets.
S 9843B
Battery Free
Door Bell
SAVE
$30
SAVE $70
129
$
Never change
batteries again!
Kinetic action with
wireless chime
unit. 58 ring tones.
150m range.
99
SAVE 24%
30
$
A 0329
$
M 8867
Also includes magnetic
balljoint bracket.
Cable Free Wi-Fi Security
This battery powered 1080p camera can be
installed indoors or out - no cables required!
Offers 4-6 months of motion detect recording. When it’s flat, just take it off the wall &
recharge via USB. Suits sheltered outdoor use.
140W USB Car Charger
Power up all your devices while on the camping or
just on the open road! Provides 140W USB C Power
Delivery (PD3.0/3.1). In addition it also features a
secondary 20W USB-C PD port for charging phones
etc. Fitted with 2m cable so you can run it from a battery into your tent.
SAVE 34%
SAVE 25%
2 for
$
Boost your TV
signal!
Handy
problem
solver!
Simple to install
in-line booster for
added gain to your
existing antenna.
USB powered.
SAVE 19%
16
$
L 2047
P 8164*
39
$
40
X 0226
SAVE 15%
2 for
$
33
SAVE $30
X 3229
Wardrobe Sensor
Light Kit
At less than $20 you can
afford to put these 4xAAA
battery powered sensor
lights in every cupboard!
1m length.
A light for camping,
fishing, anything!
Compact and USB rechargeable!
Standard torch PLUS 5W LED flood
light in one convenient folding design.
149
$
Dual GPO Charging
Station
M 8196
Great for the workbench!
Two top mount GPO’s for your
appliances, USB charger (20W
PD+2x USB A) and 10W QI
wireless charger.
Getaway Power Generator & Light
With powerful LED lantern, torch & emergency
dynamo! This compact portable charging solution
keeps your devices powered up. The internal 83Wh
battery (equivalent to 7Ah at 12V) is paired with
three USB outputs, including an 18W USB C PD port
and two 18W QC3.0 type A ports.
Charge any
device, anywhere.
X 0212A
X 2385
SAVE 20%
SAVE 30%
29
39
$
$
Aluminium USB Torch
Durable all metal 5 Watt USB
rechargeable torch. Can be
used as an emergency power
battery bank. 182mm long.
SAVE 14%
Wall Mountable
6W Solar Outdoor Light
Needs no wiring. Ground spike or wall
mount. Flood/spot modes. 6-7hr run
time. IP67 rated.
Power Saver
Mains Switch
17
$
A 0346
Cut standby power usage by
switching appliances off at the
wall. Extra outlets $16.50 (A0347)
This high capacity
12,000mAh
battery bank has
charging cables
in-built for USB
C, Micro USB and
Apple Lightning
equipped devices.
It even includes
the recharging
cable!
SAVE 27%
36
$
D 0510A
Sale Ends August 31st 2025
Shop in-store at one of our 11 locations around Australia:
WA » PERTH » JOONDALUP » CANNINGTON » MIDLAND » MYAREE » BALCATTA
VIC » SPRINGVALE » AIRPORT WEST QLD » VIRGINIA NSW » AUBURN SA » PROSPECT
Or find a local reseller at: altronics.com.au/storelocations/dealers/
Shop online 24/7 <at> altronics.com.au
© Altronics 2025. E&OE. Prices stated herein are only valid until date shown or until stocks run out. Prices include GST and exclude freight and insurance. See latest catalogue for freight rates.
B 0008
Keeping a high-power amplifier cool is vital to its
longevity. Designing the chassis properly is
important for achieving the best possible
cooling performance. It’s even possible
to improve the cooling of existing
amplifiers if necessary. This
photo shows the Silicon
Chip 500W Power
Amplifier from AprilJune 2022.
Part 1 by Julian Edgar
Cooling
Audio Amplifiers
L
ow-power amplifiers are easy to
cool; a reasonably modest heatsink is sufficient for cooling to occur
through natural convection in the air.
That’s satisfactory in many domestic situations. But if it’s a powerful
amplifier that you push really hard, or
it’s mounted in a hot location, things
aren’t so easy!
I recently ran into major problems
with amplifier cooling. First, the two
amplifiers were working at higher
power levels than I’d ever previously
used them. Second, rather than being
located inside a cool house, they were
stacked in a much hotter roof space.
The outcome was fried amplifiers...
So it’s important to design an amplifier for proper cooling – and if it’s
already built, you might need to make
some adjustments to fix a less-thanideal design. This series will cover
both aspects.
Requirements
Amplifiers generate heat in three key
siliconchip.com.au
areas. The most important heat generators are the output devices, whether
they are transistors or ICs. Perhaps 3/4 of
the heat generated by a typical amplifier is created by these components.
However, significant heat is also generated by the power supply, mainly
in the bridge rectifier, the transformer
and assorted other devices like voltage
regulators (if present).
Cooling an amplifier falls into two
categories: specific cooling, typically by thermally connecting certain
high-temperature components to a
large heatsink, and general cooling,
typically by allowing ventilation or
forced air through the enclosure.
Where possible, these two requirements should be kept separate. For
example, if the main heatsink is buried deep within the case (which is
not at all uncommon), the heatsink
will warm nearby components. Conversely, if the heatsink is mounted
on the outside of the case, this heat
can move straight to the wider
Australia's electronics magazine
environment, so it won’t impact interior case temperatures so much.
Another option is to mount the output devices on a tunnel heatsink with
a fan sucking air in through a vent on
one side of the case and blowing the
warm air out a vent on the other side.
Unless that warm air is being sucked
back in somewhere else, it will have
minimal effect on other components
in the amplifier.
Heatsinks
Heatsinks work in two quite different ways. As it names suggests, a
heatsink absorbs heat. As it does, its
temperature rises. Say we are using
a huge 1kg block of aluminium as a
heatsink. The specific heat value of
aluminium is 0.9J/°C/g, so to raise
the temperature of our block of aluminium by 1°C requires 900J (0.9J ×
1°C × 1000g).
That’s equivalent to 900W of power
for one second, 450W for two seconds
or 225W for four seconds. So after 60
August 2025 23
seconds at 225W, the heatsink temperature will have risen by 15°C. If the
ambient temperature is 25°C, our 1kg
heatsink will already be at 40°C after
just a minute!
If we ran our very powerful amplifier (that we are assuming dissipates
225W) in only 10-minute bursts, we’d
be fine. But running it for an hour,
the transistors will get hot enough to
burn out. So our heatsink will be quite
inadequate.
You can see that the name ‘heatsink’ is a bit of a misnomer; what
we call heatsinks primarily work as
heat exchangers. Heat exchangers are
devices that transfer heat, often to the
air (or sometimes to water, or even oil).
Heat exchangers
While we have referred to amplifiers throughout this article, any piece of
equipment that needs to dissipate a lot of heat will benefit from these techniques. This includes inverters, speed controllers and electronic loads.
Heat exchangers shed their heat in
three different ways. The first is conduction. If you run an amplifier at full
power, switch it off, then pick it up
and moved it, you might find that the
shelf it was sitting on is warm.
The amplifier has heated the shelf
primarily through conduction –
although that’s more likely if the prototype amplifier is yet to gain feet, and
there was a big contact area between
the amplifier and the shelf.
Conduction is important to amplifier cooling in two ways. First, the heat
source (output transistor, output IC,
bridge rectifier etc) needs to conduct
heat to the heat exchanger (heatsink).
You could have the best heatsink in
the world, but if the device can’t transfer heat into it fast enough, the device
could still fail.
The heat transfer depends on
numerous factors such as the device’s
packaging, which will act as an insulator to some extent, but must be present to transfer the heat onto a large, flat
surface suitable for clamping to the
heatsink. It also depends on how flat
the surfaces are and how firmly they
are pressed together.
Because perfectly flat surfaces are
unlikely, thermal paste is usually
applied between them, to help fill in
the gaps. But it isn’t a perfect heat conductor either.
Thermal paste should not be used to
bridge large gaps – the mounting surfaces of both the electronic device and
heatsink need to be as flat as possible.
Ensure the compound is still runny;
if it has started going hard, discard it.
Second, in many amplifier designs,
the case itself can act as a heat
Australia's electronics magazine
siliconchip.com.au
A car amplifier I built with the cover removed (shown at the bottom). The
smallest possible enclosure dimensions were required, preventing the use
of conventional finned heatsinks. The front, rear and bottom aluminium
panels of the case all act as the heatsink. They are bolted together with
generous flanges coated with heatsink grease. The car amplifier fan is
controlled by an off-the-shelf module (lower right) that can be easily set to
different temperatures using DIP switches.
The bottom sheet
metal panel of this
car sound amplifier
was replaced with clear
acrylic. A fan has been
added under the sheet (a thin fan was required) and it draws air out of the
case. Air is admitted to the case through the chamfered holes shown inset,
positioned above added finned heatsinks.
Cooling other equipment
24
Silicon Chip
An amplifier I built that uses
thermostatically controlled fan
cooling. The temperature controller
and display are on the front panel.
A fan in the centre of the top panel
draws air out of the amplifier, aiding
natural convectional flow. There is
a similarly sized vent on the bottom
panel. At the rear of the amplifier,
the main heatsink is positioned
horizontally, with a fan blowing air
along the fins. The fans switch on at
40°C. Despite working hard during
some hot days, in 10 years, the
250W amplifier’s heatsink has never
exceeded 45°C.
exchanger. That’s especially so if the
enclosure is made from aluminium,
which is a decent conductor of heat
(good electrical conductors are also
usually good heat conductors). To do
this effectively, the various enclosure
panels need to be in intimate contact
so the heat is readily conducted to all
parts of the enclosure.
When a heat exchanger conducts
its heat to the adjacent air, it takes
very little time for that thin layer of
air against the heat exchanger fins
to warm up. Once the temperature
difference between the air and the
heat exchanger drops to nothing, the
heat transfer stops. The trick is to
move that air away, replacing it with
cooler air.
This can occur due to natural convection; the warmed air is less dense
and so it rises, being replaced with
cooler air that is drawn in from below.
Convectional flow is largely vertical,
so for a heatsink to work effectively
by convection, it requires vertical fins
along which the air can slide, and no
obstructions above or below those fins.
The amount of heat that will be
exchanged with the air in a given
period is heavily dependent on the
exposed surface area of the heat
exchanger – more is better. Increased
surface area is provided by using fins
and having a textured (rough) surface
to each fin.
Fins in most large amplifier heat
exchangers are often relatively thick
and few. Having numerous very thin
An amplifier during construction.
The two finned heatsinks have been
mounted face-to-face to form a tunnel.
One fan is used at each end of the
tunnel – one blowing & one sucking.
An efficient fan-forced heatsink
design – note the fins on the fins,
giving a massive surface area. This
main heatsink is external to the case,
preventing heat being shed from
this heatsink and warming internal
components.
siliconchip.com.au
Australia's electronics magazine
August 2025 25
This 250W amplifier was originally
cooled just by convection. However,
this proved insufficient, so two fans
were added (see below). Note how the
fan shrouds (upturned baking dishes)
cover the top of the heatsink fins,
drawing air past them. With a setup
like this, nothing can be placed on top
of the amplifier!
Note the heavy gauge aluminium angle used to thermally
link the output devices to the exterior finned heatsinks,
and how the rear and bottom panels are aluminium and
are thermally connected to also act as heatsinks. Heavy
aluminium angle is also used to cool the two bridge
rectifiers.
fins is more effective, but thinner
fins are more easily damaged. A good
example of this is an air conditioner,
which will usually have lots of very
thin fins, but if you bump it, they will
be squashed.
Convectional airflow can also be
used to cool the interior of the amplifier – the ‘general’ cooling we mentioned earlier. To achieve this, we
need to take a similar approach to heat
exchanger cooling – placing vents on
the top and bottom of the amplifier
enclosure and then ensuring there are
no restrictions to that gentle natural
air movement.
Vents in amplifier enclosure side
panels do very little unless there is
forced airflow (ie, fans).
One major downside to vertical
convectional flow is that it is easily
impeded by stacking equipment on
top of each other, using mounting feet
that are too short, and decorations (like
flower pots) that may be placed on the
top of exposed amplifiers to make them
look better. We’ve also seen cats lying
on top of amplifiers to keep warm – it
may be great for the cat, but not the
amplifier!
The final heat exchange mechanism
is radiation; however, this is the least
important. Black heatsinks will radiate heat more effectively than silver
or light-coloured heatsinks, but the
26
Silicon Chip
difference is relatively small unless the
heatsinks are getting very hot. Black
anodised heatsinks are around 6-8%
more effective than silver ones under
normal circumstances.
So it’s clear that while amplifier
heatsinks are heatsinks, more importantly, they are heat exchangers with
the air. Conduction and convection
are critically important in cooling
heat exchangers. Convectional flow
requires careful design and construction, especially in giving free vertical movement to cooling air. Heat
exchangers should have the maximum
possible exposed surface area.
Fans
As we suggested above, convectional flow can be thought of as being
quite fragile – easy to disrupt and
requiring specific heat exchanger fin
orientation. Rather than relying on
convection, we can use a fan or fans –
either to aid the natural convectional
flow, or to replace it. Let’s look first at
aiding convectional flow.
Say we have a commercial amplifier
that is running very warm. Its heatsink
is located in the middle of the enclosure, with its fins orientated vertically.
There are grilles in the top and bottom
enclosure covers, and convectional
flow is supposed to provide the cooling. To increase this convectional flow,
Australia's electronics magazine
we can add a fan to either the top or
bottom of the case.
If it’s on the top, it should draw air
out of the enclosure and blow it up. If
it’s on the bottom, it should draw cool
air from below the amp and blow it
into the enclosure. Either way, because
it is aiding natural convectional flow,
the result will be much more effective
than, say, attaching a fan to the side of
the heatsink itself.
In some cases, the new top or bottom
fan can be fitted within the enclosure –
even a quite small fan will, in my experience, massively improve flow over
purely convectional air movement. If
the amplifier is too tight inside to do
this, and the amplifier is not normally
able to be seen, cutting a hole in its lid
and adding an external fan sitting on
top will work well.
Rather than aiding convectional
flow, you can instead decide to organise the heat exchanger purely to
suit the fan. For example, the heat
exchanger fins can be horizontal. The
key criterion is that the air movement
provided by the fan must pass along as
much of the exposed area of the heatsink fins as possible.
For example, two long finned heatsinks can be mounted facing one
another, forming a heatsink tunnel. A
fan at one end blows into the tunnel,
while one at the other end extracts
siliconchip.com.au
heat from the tunnel (one fan may be
enough to do both jobs). The electronic
devices bolt to the outside of the heatsinks. For its size, this approach is one
of the most efficient ways of cooling
an amplifier.
This is the approach used in our
Variable Speed Drive Mk2 (November & December 2024; siliconchip.au/
Series/430) and it proved very effective.
Fans should always move air along
heat exchanger fins – we want air to
slide along the fins, pick up heat, then
depart. We don’t want air to just be
turbulently whizzing around!
It’s also important to consider what
happens to the warm air after it has
picked up the heat from the fins. We
don’t want it to end up pushed against
a solid panel where it will splash back
and heat up other components. We also
don’t want it to circulate around back
to the input side of the fan, or the air
will just end up getting hotter and hotter. Ideally, it should go straight out of
the case once it’s warm.
Conventional PC-type axial fans are
the most common and cheapest fans
available, and they are also easily salvaged at no cost from many discarded
consumer items. There’s also the significant advantage (for use in amplifiers) that many silent or almost-silent
types are available that still move a
reasonable amount of air.
However, squirrel cage (cross-flow)
fans can move a huge amount of
air, can be very quiet (or very loud,
depending on their design) and their
long, thin shape lends itself to low-
profile amplifiers. In the past, this type
of fan has been quite expensive, but
they’re now cheaply available from
Chinese suppliers, including low-
voltage designs.
However, if you decide to use one
of these fans, be prepared do so some
sheet metal work – they typically don’t
just bolt into place, but instead need
some baffles made.
The flow of air through an enclosure
needs to adequately cool the various
hot components. This will not occur
if the airflow can take a ‘short-cut’
route, for example, passing straight
from an inlet grill to the adjacent outlet fan. However, it can be difficult to
picture where the airflow will go just
by looking at the amplifier. Two airflow visualisation techniques can be
used, though.
The first is to stick short (eg, 10mm)
siliconchip.com.au
An older hifi amplifier heatsink,
pictured with normal and thermal
cameras. The thin fins give an
excellent surface area, while the thick
metal base conducts heat along the
heatsink from the widely separated
output devices. The temperature is
only about 10°C over ambient, even
after testing at high loads with the top
cover in place.
Sometimes individual components
can run very hot. Typically, they
have been fitted with small heatsinks,
but they seem quite ineffective. This
component is running at nearly 49°C
with a 20°C ambient temperature.
Australia's electronics magazine
August 2025 27
Sizing inlets and outlets
Any fan that draws air out of an amplifier must have an equivalent inlet vent
area. For example, if a 90mm diameter fan is fitted (a cross-sectional area of
about 6000mm2), the inlet vent area must also be about 6000mm2. This inlet
can comprise a single 90mm diameter opening, or multiple openings that add
up to the same cross-sectional area.
However, note that as the diameter of the inlets decreases, their restriction
to airflow increases – so if the inlet area comprises mesh with small openings, the total of the openings will need to be greater.
There is no immediate disadvantage in oversizing the ventilation inlet area,
although having too many vents may make it difficult to control the airflow
patterns.
If the inlet vent area is too small compared to the outlet fan area, the result
will be a reduction in air pressure inside the case. This can make the fan(s)
less effective, increase noise and dust collection and sometimes result in
uneven cooling. In general, it’s preferable to have neutral or a slightly positive
pressure inside the case.
pieces of cotton thread inside the
amplifier and then temporarily replace
the lid with a sheet of clear glass or
plastic (don’t leave the lid off – the
airflow direction will be quite different with the lid removed). With the
fan switched on, the direction that
the cotton pieces point will show the
directions of airflow.
Ensure that the power supply capacitors have fully discharged before
opening the amplifier. The same
applies after you have finished your
flow testing and need to remove the
threads.
The other approach, which works
very well, is to again temporarily
replace the top cover with a clear sheet,
but this time use a source of smoke,
like an incense stick, to make the airflow visible. Light the incense stick,
allow it to flame for a few moments,
then blow it out. A thin stream of
smoke will be released from the end
of the stick.
Allow the smoke to be drawn in by
the fan and watch where the airflow
goes by looking at the smoke pattern. If
the amplifier has multiple inlet openings, place the incense stick in front
of each in turn. It’s almost certain that
the internal airflow will show unexpected patterns. We will use this technique next month when modifying an
amplifier’s cooling.
If the cooling airflow is bypassing
key components, the easiest solution
is to place one of more baffles or guides
to redirect the airflow. Cardboard
can be temporarily used during flow
testing. Then, when effective baffle
designs have been developed, it can
be replaced with aluminium sheet
or, if there is insufficient clearance to
live areas, with Presspahn, acrylic or
a similar insulating material.
Fan control
Because of the noise, people
often object to the use of fans in hifi
amplifiers. After all, who wants a quiet
passage ruined by the whirr of a fan?
Two approaches can be used to overcome this objection.
The first is to use a thermal switch to
switch on the fan only when the heat
exchanger temperature is too high. A
normally open mechanical temperature switch, closing at say 40°C, is the
simplest way of achieving this. However, such switches are not as widely
available as they once were, and so it
may be easier to use an electronic temperature switch.
These prebuilt boards are available
with relays, remote sensors and adjustable temperature setpoints. They are
very cheap, and some have panel temperature displays – which can be reassuring to watch when your fan-cooled
amplifier is belting out the tunes!
One disadvantage of this approach
is that, unless your fan(s) are totally
silent at full speed, you may notice
them switching on and off. Also, given
that the ambient temperature may
vary, and amplifiers dissipate power
even when idle, it’s almost certain
that the fans will be on (and running
at full pelt) some of the time when the
amplifier is in use.
Another approach, which works
very effectively, is to have the fan(s)
operate at a slow speed whenever the
amplifier is switched on. Experiment
with suitable series resistor values
until you find one that slows the fan
to the point of inaudibility, but still
allows the fan to flow a reasonable
amount of air.
You can then use the temperature
switch to short out the resistor, changing the fan to full speed when an elevated temperature occurs. Because the
heat exchanger is always fan cooled,
When selecting amplifier and power supply modules, look carefully at the heatsinking. This bridge
rectifier heatsink has vertical fins (good), but the bottom of the heatsink is completely
blocked to convectional airflow (bad).
While designed to be mounted
horizontally, mounting this
amplifier module with the heatsinks
fins vertical and the board
slightly raised to give bottom
clearance will dramatically
improve cooling. The two bridge
rectifiers on the right need to be
raised on extension wires to give
clearance for fitting heatsinks, with
their fins aligned with those on the main
heatsink.
28
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
albeit at a low speed, it will take a lot
longer for the heat exchanger to reach
the ‘fan full speed’ temperature.
Note that some fans use bearings
that require a certain minimum speed
before the bearing operates properly.
If the bearing squeaks or makes any
other noise at low speed, increase the
minimum fan speed.
Another great option is to use our
Fan Controller & Loudspeaker Protector (February 2022; siliconchip.
au/Article/15195), which controls the
speed of up to three PWM-
capable
fans. You can set it so that the fans are
off at low temperatures, switch on at
low speed as the temperature rises,
then increase in speed until the temperature stabilises.
This gives you the best of all
worlds: complete silence (passive
cooling) when possible, effectively
silent fan-forced cooling under most
conditions, and highly effective cooling when the ambient temperature is
high and/or the amplifier is producing a lot of heat.
While it’s a little on the expensive
side, Jaycar’s YX2584 is a good example of a fan that runs basically silently
at full speed. It’s a 120mm, 12V DC
type with maglev bearings (that run
virtually forever; the rated life is
100,000 hours) and it flows 1795L/min
with a noise level of 25dBA. Even in a
quiet environment, you’d be unlikely
to notice that noise.
You could also consider a fan from a
manufacturer like Noctua or BeQuiet!,
both known for fans with a good balance between airflow and noise.
That’s all we have space for this
month. Next month, we’ll show you
how to test an amplifier at high loads
SC
and improve its fan cooling.
Measuring heatsink temperature under full load using an infrared thermometer.
A good amplifier cooling system should keep the heatsink temperature less than
25°C above ambient – in this warm room, this reading is just on that limit.
Both Thermalright and Noctua make excellent fans. Although Noctua’s are very
reliable, they are much more expensive compared to other manufacturers.
Squirrel cage fans, sometimes call cross-flow fans, work well for amplifier cooling,
especially where the enclosure is not very tall. Air can be drawn-in through
one or more vents, then discharged through a rear slot against which the
fan is positioned. These fans can be quiet and flow a lot of air. Both
mains-powered and low-voltage DC designs are available.
Modules like the one shown to the right can
easily have the overly small heatsink
unbolted and a very much larger
heatsink substituted. I use
four of these modules in
an amplifier with a
fan-cooled heatsink
about ten times as big
as the one provided!
siliconchip.com.au
Australia's electronics magazine
August 2025 29
EXPLORE THE
ELEGOO® RANGE (AU)
TL4976
TL4984
TL4830
EXPLORE THE
ELEGOO® RANGE (NZ)
TL4842
NEW DIMENSIONS. NEW POSSIBILITIES.
ELEGOO® 3D Printers and Materials - Now In Orbit at Jaycar
USE CASE
FILAMENT PRINTERS
USE CASE
RESIN PRINTERS
EVERYDAY PRINTS
Neptune 4
Neptune 4 Pro
Neptune 4 Plus
TL4970
TL4972
TL4974
$399
$549
$649
DESKTOP
RESIN PRECISION
Mars 4
Mars 5
Mars 5 Ultra
TL4824
TL4830
TL4832
$349
$399
$599
LARGE FORMAT
BUILDS
Neptune 4 Max
OrangeStorm Giga
TL4976
TL4982
$799
$4499
PROFESSIONAL
RESIN OUTPUT
Saturn 4 12K
Saturn 4 Ultra 12K
Saturn 4 Ultra 16K
TL4840
TL4842
TL4844
$729
$899
$1049
PREMIUM
HIGH-SPEED PRINTS
Centauri
Centauri Carbon
TL4984
TL4980
$549
$699
SEE IN-STORE OR ONLINE FOR OUR EXTENSIVE RANGE OF
FILAMENT AND RESIN AT GREAT VALUE
Explore our great range of ELEGOO® 3D Printing gear, in stock on our website,
or at over 140 stores or 130 resellers across Australia and New Zealand.
jaycar.com.au 1800 022 888 | jaycar.co.nz 0800 452 922
Prices shown in $AUD, and correct at time of publication but are subject to change. Jaycar reserves the right to change prices if and when required.
THINK GIGANTIC.
PRINT GIGANTIC.
Meet the OrangeStorm Giga by ELEGOO®. 800×800×1000mm
of industrial-grade print capacity, now at Jaycar.
ONLINE
ONLY
PORTABLE
7" HD TOUCHSCREEN
4499
$
Can be detached
from the body for
handheld operation
TL4982
POWERFUL
DOUBLE-SIDED
COOLING FANS
EXPANDABLE
TO 4 NOZZLES *
Can comprehensively
enhance the model
printing results
Print four identical
models at once, each
in different colours.
*
Nozzle kit sold separately
CATERPILLAR
CABLE TRACKS
Prevents the cables
from being worn or
broken and affecting
your prints
POWERFUL
EXTRUSION SYSTEM
High-quality nozzle
capable of reaching
300°C. Dual-gear
direct extruder.
GIGANTIC IN SIZE UP TO
800 X 800 X 1000MM
Print oversized models
without splitting.
Fast and efficient
up to 300mm/s
AUTO LEVELLING
One-click automatic
levelling and
non-contact high
precision sensor
•
EFFICIENT HEAT BED
Four independent
PEI magnetic
high-temp platforms.
Up to 100°C
SEE IN-STORE OR ONLINE FOR OUR
EXTENSIVE RANGE OF FILAMENT AND
ACCESSORIES AT GREAT VALUE
LEARN MORE: (AU)
jaycar.com.au/p/TL4982
LEARN MORE: (NZ)
jaycar.co.nz/p/TL4982
Explore our great range of 3D Printing gear, in stock on our website,
or at over 140 stores or 130 resellers across Australia and New Zealand.
jaycar.com.au 1800 022 888 | jaycar.co.nz 0800 452 922
Prices shown in $AUD, and correct at time of publication but are subject to change. Jaycar reserves the right to change prices if and when required.
FROM FIRST LAYER TO
FINISHED MASTERPIECE.
Explore the full range of ELEGOO®
Filament Printers now at Jaycar.
Neptune 4
TL4970
EXPLORE THE
ELEGOO ® RANGE (AU)
Neptune 4 Pro
TL4972
Neptune 4 Plus
TL4974
Neptune 4 Max
TL4976
Centauri
TL4984
Centauri Carbon
TL4980
EXPLORE THE
ELEGOO ® RANGE (NZ)
OrangeStorm
TL4982
Printer Type
Build Volume (mm)
Nozzle Size & Type
Printing
Nozzle
Speed (max.) Temp (max.)
Bed Temp
(max.)
Connectivity
Cat Number
Price
Neptune 4
Cartesian Gantry Frame
225x225x265
0.4mm Standard
500mm/s
300°C
110°C
USB, LAN, Wi-Fi^
TL4970
$399
Neptune 4 Pro
Cartesian Gantry Frame
225x225x265
0.4mm Standard
500mm/s
300°C
110°C
USB, LAN, Wi-Fi^
TL4972
$549
Neptune 4 Plus
Cartesian Gantry Frame
320x320x385
0.4mm Standard
500mm/s
300°C
100°C
USB, LAN, Wi-Fi^
TL4974
$649
Neptune 4 Max
Cartesian Gantry Frame
420x420x480
0.4mm Standard
500mm/s
300°C
85°C
USB, LAN, Wi-Fi^
TL4976
$799
OrangeStorm
CoreXY Open Cube Frame
800x800x1000
0.4mm# Standard
300mm/s
300°C
100°C
USB, LAN, Wi-Fi
TL4982
$4,499
Centauri
CoreXY Open Cube Frame
256x256x256
0.4mm Hardened Steel
500mm/s
320°C
100°C
USB, LAN, Wi-Fi
TL4984
$549
Centauri Carbon
CoreXY Enclosed
256x256x256
0.4mm Hardened Steel
500mm/s
320°C
110°C
USB, LAN, Wi-Fi
TL4980
$699
^
12 FILAMENT OPTIONS:
• RAPID PLA+
• PLA
• PLA MATTE
• GALAXY PLA
• SILK PLA
• PLA+
Wi-Fi dongle sold separately. Also supports other nozzle sizes. See individual specs.
• RAPID PETG
• PETG PRO
• TPU
#
• ASA
• CARBON FIBRE
SHOP ONLINE OR IN-STORE FOR OVER 115 ELEGOO® FILAMENT ROLLS
ACROSS 12 MATERIAL TYPES AND A WIDE RANGE OF COLOURS.
Explore our great range of 3D Printing gear, in stock on our website,
or at over 140 stores or 130 resellers across Australia and New Zealand.
jaycar.com.au 1800 022 888 | jaycar.co.nz 0800 452 922
Prices shown in $AUD, and correct at time of publication but are subject to change. Jaycar reserves the right to change prices if and when required.
SHARP,
SMOOTH,
SPECTACULAR.
Smooth layers. Sharp details. Spectacular results.
Get the ELEGOO® Resin Range at Jaycar today.
Mars 4 9k
TL4824
Mars 5
TL4830
Mars 5 Ultra
TL4832
Saturn 4
TL4840
Saturn 4 Ultra 12K
TL4842
Saturn 4 Ultra 16K
TL4844
Build Volume (mm)
Resolution
Precision
Layer Thickness
Print Speed
Special Features
Cat Number
Price
Mars 4
153.36 x 77.76 x 175mm
9K
18µm
0.01–0.2mm
30–70mm/h
• Laser-Carved Build Plate
TL4824
$349
Mars 5
143.43 x 89.6 x 150mm
4K
35μm
0.01–0.2mm
Max 70mm/h
• One-click levelling
TL4830
$399
Mars 5 Ultra
153.36 x 77.76 x 165mm
9K
18μm
0.01–0.2mm
Max 150mm/h
• One-click levelling
• AI camera
TL4832
$599
Saturn 4
218.88 x 122.88 x 220mm
12K
19×24μm
0.01–0.2mm
Max 70mm/h
• One-click levelling
TL4840
$729
Saturn 4 Ultra 12K
218.88 x 122.88 x 220mm
12K
19×24μm
0.01–0.2mm
Max 150mm/h
• One-click levelling
• AI camera
TL4842
$899
Saturn 4 Ultra 16K
211.68 x 118.37 x 220mm
16K
19×24μm
0.01–0.2mm
Max 150mm/h
• Auto-levelling
• AI camera
TL4844
$1,049
GREAT RANGE OF ELEGOO®
WASH & CURE MACHINES
AT GREAT VALUE
EXTENSIVE RANGE OF
ELEGOO® RESIN AVAILABLE
AT GREAT VALUE
2 MODELS AVAILABLE
13 TYPES AVAILABLE
Explore our great range of 3D Printing gear, in stock on our website,
or at over 140 stores or 130 resellers across Australia and New Zealand.
jaycar.com.au 1800 022 888 | jaycar.co.nz 0800 452 922
Prices shown in $AUD, and correct at time of publication but are subject to change.
Jaycar reserves the right to change prices if and when required.
EXPLORE THE
ELEGOO® RANGE (AU)
EXPLORE THE
ELEGOO® RANGE (NZ)
Using Electronic Modules with Tim Blythman
Thin-Film
Pressure Sensor
Being able to sense force and pressure is handy as it allows properties
like weight to be measured. While industrial-grade pressure sensors
are available at higher prices, thin-film pressure sensors use a simpler
technology and are much cheaper.
F
orce and pressure sensors are
used in industrial applications.
In addition to directly measuring pressure (such as in a gas reaction vessel),
they can measure liquid volumes and
weight.
Pressure can be related to liquid volume since the height of a liquid column and its density dictate the amount
of pressure it exerts. If you can apply
the pressure over a known area, the
applied force can also be known and
thus the weight-derived force due to
gravity can be determined.
Just about any product you can buy
by weight or volume has been precisely
measured out using a sensor such as a
strain gauge. These are among the more
common types used for this purpose
since they have the necessary accuracy.
Of course, accuracy comes at a cost,
and many projects don’t need the kind
of accuracy these devices provide.
That said, strain gauge sensors and
their interface electronics are readily
available to the hobbyist if that level
of accuracy is needed.
Thin-film sensors
So-called thin-film pressure sensors are also known as force-sensitive
resistors; simply put, they are devices
that change their resistance when force
is applied to them. This makes them
quite easy to use since a simple resistive voltage divider is sufficient to get
a reading using an ADC (analog-to-
digital converter).
The force-sensitive resistor consists
of a polymer containing conductive
particles. The polymer is applied as a
thin film (hence the name) to an array
of conductive electrodes. As pressure
is applied, the conductive particles
touch the electrodes and each other,
reducing the resistance. Fig.1 shows
the construction of a typical device.
We’ve seen similar sensors created
by sandwiching a layer of conductive
foam (such as used for packaging DIP
ICs) between two blank PCBs or similar conductive plates. As the foam is
compressed, its resistance decreases.
Thin-film pressure sensors have
hysteresis and thus poor accuracy;
error figures of around 10% or higher
are typical. Not only does the reading
vary quite a bit, but it will also depend
on the sensor’s recent history.
So they are not suitable for precise
measurements. However, they are
often used as touch sensors since touch
sensing does not require a high degree
of accuracy. As long as the touch force
can be coupled to them, they can work
behind a protective surface in harsh
conditions.
Some force-sensitive resistors are
constructed as long, thin devices with
three terminals, like a potentiometer.
A touch moving along the length of
the resistor is analogous to moving
the pot’s wiper, so the touch position
can be estimated.
Sensor modules
Fig.1: pressure applied to the sensor brings together conducting particles
within the substrate and closes the gap between the active area and
substrate, reducing the sensor’s resistance. The black region of the sensor
is a high-resistance polymer that’s embedded with carbon. The silver areas
are conductive electrodes that expand the sensor’s active area. The sensor
electrodes are connected to terminals that are soldered to a module PCB
featuring a resistor, mounting holes and a 3-way pin header.
It is possible to purchase bare
force-sensitive resistors, but they are
also available attached to a module
with a pin header, making them easy
to interface to a microcontroller board
such as an Arduino main board.
We tried the Duinotech XC3738
Arduino Compatible Thin-Film Pressure Sensor from Jaycar Electronics.
It consists of a sensor attached to a
module PCB. The PCB has a three-way
header and a single 510kW resistor,
marked as R1. There is an unpopulated
Australia's electronics magazine
siliconchip.com.au
34
Silicon Chip
Fig.2: The circuit on the Duinotech Thin-Film
Pressure Sensor module is a simple voltage
divider. As the sensor is in the upper half,
the output voltage increases as pressure is
applied. There is an empty footprint for a
capacitor, which we recommend fitting.
Fig.3: the module provides
an analog voltage related
to its supply voltage, so
its connections are simple
enough. The V (or +) pin
should be fed from a voltage
that matches the ADC
reference used to measure
the voltage from the S
pin. Our sample code uses
analog input pin A0.
Screen 1: Test Sketch
930.00
933.00
930.00
929.00
798.00
901.00
907.00
917.00
920.00
916.00
923.00
918.00
924.00
926.00
925.00
927.00
925.00
928.00
926.00
51000.00
49196.14
51000.00
51603.88
143796.99
69056.60
65226.02
58953.11
57097.83
59574.24
55254.60
58333.33
54642.86
53423.33
54032.43
52815.53
54032.43
52209.05
53423.33
The output from the test sketch shows
the raw 10-bit ADC reading and a
calculated sensor resistance based on
the module’s nominal 510kW resistor
value. Even with a steady weight,
there is some drift.
footprint for a capacitor on the module; this is marked C1. Fig.2 shows its
simple circuit.
A 5V or 3.3V supply is applied
between the V and G (alternatively
labelled + and −) pins. Since the sensor’s resistance decreases as pressure
is applied, the voltage at the S pin will
increase with more pressure.
Circuit and software
Fig.3 shows the simple circuit we
used to test the module with an Uno
R4. Since the Uno R4 has socket headers and the module has plug headers,
we made the connections using plugsocket jumper wires. We expect that
almost any Arduino board with an analog input can be substituted.
The “XC3738_test.ino” sketch uses
the ADC to read the voltage at its A0
pin and displays the raw 10-bit ADC
reading (from 0 to 1023) and the calculated force-sensitive resistor resistance
(siliconchip.com.au/Shop/6/502). This
was a simple way to get a feel for
how the module responds to being
squashed and squeezed.
When no pressure was applied, we
got a reading of 25, indicating a sensor
resistance of around 20MW. We could
get a reading over 1000 with firm pressure between our fingertips, indicating
a resistance near 10kW.
siliconchip.com.au
As you can appreciate from Fig.1,
the sensor is quite thin, and it’s not
immediately clear how it could be
used to weigh an object or vessel. We
measured the sensor tip with callipers
to be around 0.3mm thick.
The Jaycar website offers a basic
data sheet, and we found some more
detailed data sheets for similar devices
from Interlink Electronics (www.
interlinkelectronics.com). That firm
appears to be one of the pioneers of
this technology. The sensor on the
XC3738 looks quite like Interlink’s
FSR 400 sensor.
We also found an Integration Guide
on the SparkFun Electronics website
with numerous tips for this type of sensor (siliconchip.au/link/abx5). This
guide doesn’t exactly correspond to
the Duinotech sensor, but we found
it very helpful.
They state that the sensors should
not be exposed to sharp surfaces. They
are not waterproof and have an air
vent that runs parallel to the external
leads, allowing their internal pressure
to equalise.
The guide seems to focus on measuring weights and notes that a pressure measurement would require the
vent to be in contact with air at atmospheric pressure. So we will concentrate on applications that measure
weight rather than pressure.
Testing
The guide notes that the sensors are
tested by applying force via a silicone
rubber ball.
We recommend adding small rubber feet (see Fig.1) to help spread the load
on the sensor and protect it from impacts. We
also added a 100nF capacitor to the
vacant C1 footprint on
the module.
August 2025 35
Rubber is recommended in designs
where some degree of movement is
expected. It also protects the sensor
from sharp edges and impacts while
spreading the force uniformly across
the active area.
With that in mind, we found some
self-adhesive rubber feet about 5mm
in diameter, similar in size to the sensor’s active area. We attached one to
each side of the sensor’s tip.
Screen 1 shows the output of the
XC3738_test sketch with a half-full
(half-empty?) glass resting on the modified sensor. The ADC reading is moving around a bit; the sensor measures
around 50kW.
We then rigged up a container to balance on the sensor to see if it could be
used to measure weight.
The blue trace in Fig.4 shows the
results of our first experiment. The
curve indicates quite a narrow working
range, with a notable offset from zero
grams before a meaningful reading is
registered. The values near the centre
of the graph tended to drift around a
bit, even with a steady weight, sometimes by up to 100 ADC steps.
To test the hysteresis, we noted the
values as we filled and then emptied
the container, but due to the large
amount of drift, we couldn’t draw any
firm conclusions about hysteresis.
Many microcontroller ADC peripherals recommend a source impedance
of no more than 10kW. The data sheet
for the RA4M1 microcontroller on the
Uno R4 suggests 6.7kW at most.
The divider on the Thin-Film Pressure Sensor module is typically dominated by the 510kW resistor, so it
would usually have a much higher
impedance than the recommended
value.
That could lead to ADC readings
being affected by noise and even the
ADC sampling process. The typical
solution is to fit a capacitor here to
provide a low-impedance voltage
source; we generally use a 100nF part
for this role.
Such a value results in a time constant of around 50ms, which we figure should not affect any weight-
measuring applications. It might be a
bit high if you are using the module as
a touch sensor to detect brief touches,
though.
So we fitted a 100nF M3216 (1206
imperial) SMD capacitor to the C1
footprint on the module, visible in our
photo. We then repeated the weight
experiment and recorded the red curve
in Fig.4.
We still noted quite a bit of drift
around the middle of the graph. Overall, the response is similar, although
the values span a wider range; the
capacitor clearly makes a positive difference. The useful working range in
either case is approximately 150-300g.
There is some response to changing
weights above this range, but it is not
as distinct. We wonder if replacing
the resistor with a lower value might
provide better resolution at higher
weights at the cost of losing resolution
at lower weights.
In use
The narrow working range sounds
quite limiting, but it could be expanded
with the appropriate arrangement of
levers and pivot points. With the sensors being relatively cheap, a second
Fig.4: the blue curve
shows the raw 10-bit
ADC readings from the
sensor with different
weights applied. The red
curve shows the effect of
fitting a 100nF capacitor
to the module on the
readings. As you can see,
the module has a useful
response between about
150g and 300g when fitted
with rubber feet.
36
Silicon Chip
Australia's electronics magazine
or third sensor could be added to
share the load and thus the measured
weight.
Many electronic scales use an array
of four strain gauges to ensure the
weights are measured consistently,
even if they are unevenly distributed.
The thin film pressure sensors do
not produce a change in reading near
zero, which is not ideal. Adding an
extra weight could help offset the
reading, allowing it to measure lower
weights.
That said, the accuracy is not great,
and we suspect that the sensors will
be more useful in indicating a full or
empty state (with perhaps a handful
of steps in between) than a precise
weight.
The integration guide noted earlier
also suggests that calibration is necessary if precision is needed. This section of the guide also states that temperature compensation may also be
included in the calibration, with an
expected resistance change of up to
10% with temperature.
The guide mentions that humid conditions (95% RH) can change the sensor’s resistance, so this should also be
considered if the sensor is used in a
moist or humid environment.
Conclusion
Thin-film pressure sensor modules
such as the Duinotech XC3738 are
handy for detecting changes in weight
or pressure, but they are not wellsuited to precision applications. They
are more realistically useful when you
want to detect the presence or absence
of weight.
We recommend adding a capacitor and rubber feet to the sensor to
help in weight-measuring applications. Without the rubber feet, we’re
not sure how it would be possible to
apply a meaningful force to the sensor. The capacitor helps ensure it has
the correct source impedance to suit
a typical ADC.
The module’s response is expected
to vary under different conditions
and between different units. Individual calibration is probably the best
way to counteract any of those sorts
of variations. So, these devices are
better suited to one-off projects than
production devices.
The XC3738 Arduino Compatible
Thin-Film Pressure Sensor is available
from Jaycar Electronics; see:
www.jaycar.com.au/p/XC3738 SC
siliconchip.com.au
FREE Download Now!
Mac, Windows and Linux
Edit and color correct using the same
software used by Hollywood, for free!
DaVinci Resolve is Hollywood’s most popular software! Now it’s
easy to create feature film quality videos by using professional
color correction, editing, audio and visual effects. Because DaVinci
Resolve is free, you’re not locked into a cloud license so you won’t
lose your work if you stop paying a monthly fee. There’s no monthly
fee, no embedded ads and no user tracking.
Creative Color Correction
Editing, Color, Audio and Effects!
Designed to Grow With You
DaVinci Resolve is the world’s only solution that combines editing, color
DaVinci Resolve is designed for collaboration so as you work on larger jobs
correction, visual effects, motion graphics and audio post production all in
you can add users and all work on the same projects, at the same time. You can
one software tool! You can work faster because you don’t have to learn multiple
also expand DaVinci Resolve by adding a range of color control panels that
apps or switch software for different tasks. For example, just click the color
let you create unique looks that are impossible with a mouse and keyboard.
page for color, or the edit page for editing! It’s so incredibly fast!
There’s also edit keyboards and Fairlight audio consoles for sound studios!
Professional Editing
DaVinci Resolve 19 ............................................................... Free
DaVinci Resolve Micro Color Panel .............Only $809
DaVinci Resolve is perfect for editing sales or training videos! The familiar track
layout makes it easy to learn, while being powerful enough for professional
DaVinci Resolve’s color page is Hollywood’s most advanced color corrector
and has been used on more feature films and television shows than any other
system! It has exciting new features to make it easier to get amazing results,
even while learning the more advanced color correction tools. There’s
PowerWindows™, qualifiers, tracking, advanced HDR grading tools and more!
editors. You also get a library full of hundreds of titles, transitions and effects
that you can add and animate! Plus, DaVinci Resolve is used on high end work,
Learn the basics for free then get more
creative control with our accessories!
so you are learning advanced skills used in TV and film.
Learn more at www.blackmagicdesign.com/au
Download free on the
DaVinci Resolve website
NO SUBSCRIPTIONS • NO ADS • NO USER TRACKING • NO AI TRAINING
USB--C
USB
Part 1 by Tim Blythman
Power Monitor
It is always handy to know what voltage and current a USB device is using. Now that USB-C
is prevalent, it’s time for a USB-C Power Monitor. It can be used for all modern USB-C
devices, as well as legacy USB devices, with the right adaptor.
T
here are many situations where it’s
helpful to see how much power
or current a USB device is drawing,
or what voltage is being applied to
it. This design makes that very easy,
and it works with virtually all modern
USB-C devices.
Say you have a fast charger and a
smartphone or tablet that’s capable of
fast charging. While the situation is
getting better, there are cases where
they are incompatible and will not
actually fast charge, or will, but only
at a modest rate. With this device, you
can see exactly how much power is
being transferred.
Another situation is if you are developing a USB device and you want to
see how much power it draws when
performing certain functions, or
whether it’s correctly signalling the
power supply to give it a certain voltage or amount of power.
Ideally, you want a compact device
that can just be plugged between two
devices, with a screen to show the voltage, current, power, energy and more.
That’s exactly what this device is.
It doesn’t even matter which way
around you connect it – it can monitor
current and thus power flow in either
direction! It doesn’t need an external
power supply, either. Its controls are
simple yet intuitive; it uses just three
pushbuttons and an OLED screen.
A bit of history
This project is, in a sense, an update
of the USB Power Monitor from the
December 2012 issue (siliconchip.au/
Article/460). It was a small PCB with
a USB-A plug at one end and a USB-A
socket at the other.
This allowed it to be fitted inline at
any place you might connect a USB-A
plug, including a computer or USB
power supply.
That USB Power Monitor displayed
information on an LCD screen, such
as the USB bus voltage and current
drawn at the socket, as well as calculating power. It took its power from the
USB supply upstream of the current
measuring shunt, and could measure
down to 1μA.
An update published in the Circuit
Notebook section of the October 2013
issue (siliconchip.au/Article/4999)
added the ability to measure energy
consumption by accumulating the
power usage over time. This update
did not require any changes to the
hardware; it was a simple firmware
upgrade.
While the earlier Monitor can still
be useful, its legacy Type-A connectors rule out the option of working
with newer USB 3.x features, such as
Power Delivery (PD) or SuperSpeed
USB data transfer.
USB-C is now very widespread. We
have recently switched to using USB-C
sockets on practically all projects that
need a USB connection. Last year,
USB-C was legislated in the European
Union as the standard charging port
for mobile phones and similar gadgets.
The demand for higher currents,
higher voltages and the consequent
higher power levels means that a
USB-C Power Monitor is necessary
for some scenarios. USB-PD (power
delivery) is currently rated to provide
up to 48V at 5A, well beyond what the
older USB Power Monitor can tolerate
or measure.
Of course, legacy USB 2.0 devices
with the older USB-A and USB-B connectors can work with the newer Monitor with simple adaptors. However,
there is a lot more to the USB-C Power
Monitor than just adding USB-C connectors to the older design!
We’ll describe some of the technical
features that apply to USB-C and how
they have affected our design. If you
Short-form Kit (SC7489, $60): this kit includes all the non-optional parts listed except the case, lithium-ion cell
and glue. It will also include the FFC (flat flexible cable PCB) for joining the two PCBs.
38
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
Features & Specifications
● Main screen reports current, voltage, power, energy (in J or Wh) & time
● Configuration channel (CC) status screen
● All 24 USB data lines pass through
● Self-contained with 400mAh rechargeable lithium battery
● Internal battery means no extra load on the USB circuit under test
● Compact case is only 80 × 40mm
● Automatic offset trimming
● Voltage measurements: up to 60V with 10mV resolution
● Current measurements: up to ±5A with 1mA resolution; 10μA resolution below
~25mA
● Power: up to 300W with 1mW resolution (limited by V and I)
● Energy: up to 999999J (1mJ resolution) or up to 999Wh (10μWh resolution)
● Battery consumption: <20mA, giving 20 hours of usage per charge
● Sleep mode: <10μA drawn from battery, less than typical self-discharge
want more background on USB technology, we have a panel listing several
Silicon Chip features and projects that
involve USB on a more technical level.
Early USB was simple
The original USB 1.0 specification
dates back to 1996, and uses much the
same hardware and electrical arrangement as the later USB 1.1 and USB
2.0 standards. Two conductors (Vbus
and GND) are provided for a nominal 5V power supply. Another two
conductors (D+ and D−) providing a
bi-
directional differential signalling
pair. A separate shield connection is
also present.
These connections are carried
through the original USB Power Monitor, with only the Vbus line interrupted
by a current measuring shunt. All these
earlier specifications have been carried
forward into the newer versions of the
USB specifications, although many
aspects are now considered ‘legacy’.
The design of the plugs and sockets
enforced upstream and downstream
ends. Assuming the standards have
been followed, this means that current
can only flow one way, and thus the
USB Power Monitor’s simple design
only needed to handle current flowing in one direction.
The supply voltage was fixed at a
nominal 5V, meaning that it could
directly power a 5V microcontroller
without even needing a regulator. In
fact, all the ICs on the original USB
Power Monitor run directly from the
USB 5V supply.
Given USB’s broad usage as a power
supply and for charging, it was also
useful when no USB communication
takes place. For example, you could
siliconchip.com.au
use it to check how much current a
mobile phone used while charging.
The USB-C standard
USB-C is a 2014 design that allows
all USB protocols, so far up to USB4,
to be transmitted. Importantly, USB-C
only specifies things like the connectors and cables; it permits many other
protocols aside from USB. Some other
communication technologies that can
be carried over USB-C include Thunderbolt, PCIe, HDMI and DisplayPort
(including digital video & audio).
The USB-C connector is symmetrical and can be inserted in two ways (no
more fumbling to figure out the right
way!). It can also be used at both ends
of a cable, so the cable itself does not
enforce an upstream or downstream
end; current and data can flow either
way in a USB-C cable. In fact, USB-C
allows the power source or sink role
to be separate from the USB host or
device function.
USB-PD (USB power delivery) provides a means to negotiate voltages
up to 48V and currents up to 5A. So
clearly, there are many factors that
need to be considered in designing a
USB-C capable power monitor.
Because of all these features, a
USB-C receptacle can have up to 24
conductors, plus a shield. Some of
these are symmetrically arranged
duplicates. For example, there are
four ground pins and four Vbus pins
to share the current load, as shown in
Fig.1. The plug is mirrored relative to
the socket.
In the receptacle, there are two each
of the D+ and D− legacy USB data pins.
There are four more differential pairs,
two in each direction. These are the
same pairs that were introduced with
USB 3.0, and the blue-coloured connectors that marked the new SuperSpeed transfer rate.
The four remaining conductors are
for signalling related to USB-C’s features. The two CC (configuration channel) lines are the channel over which
USB-PD communication takes place.
Our article “How USB Power Delivery works” (July 2021; siliconchip.au/
Article/14919) gets into the details of
the power delivery protocol.
There are now chips specifically
designed to handle USB-PD communications, and these are often found in
the so-called ‘trigger’ or ‘decoy’ modules that can request a specific voltage
using USB-PD’s digital signalling. The
photo overleaf shows an example of
such a module.
Early versions of USB-PD supported
up to six power profiles at various currents and voltages of 5V, 12V or 20V.
Subsequent revisions added support
for 9V, 15V, 28V, 36V and 48V. A more
recent protocol, known as PPS (Programmable Power Supply), allows
voltages to be requested up to 21V
in 20mV steps and current limits in
50mA steps.
Fig.1: USB-C’s design incorporates rotational symmetry, removing the need
to flip the connector like earlier USB plugs/sockets. This requires extra
conductors, with many being duplicated. The configuration channel lines
provide cable rotation detection for correct operation.
Australia's electronics magazine
August 2025 39
The main PCB of
the Power Monitor
connects to the
USB Breakout
board using a
flexible PCB
(FFC).
A typical USB-PD trigger
board has a USB-C socket, a USBPD interface IC and an output connector.
This example sets the requested PD voltage by solder
jumpers; some can be controlled digitally, with an I2C
serial interface or similar.
The simplest signalling is done via
pullups and pulldowns on the CC lines
with specific resistor values; this can
be used to offer or request the legacy
5V supply voltage. We have used this
arrangement in our projects using
USB-C sockets for 5V power.
Fig.2 shows the basic arrangement
using the resistor-based signalling,
which is quite simple and elegant.
Only one CC line is typically connected through in a cable; this allows
both devices to determine the orientation of the conductors if the cable has
been flipped.
The source will only apply Vbus if
it sees that a sink has connected with
the correct pulldown resistance. The
pulldown can be applied by a simple
resistor to ground, so the signalling
will work even on a device that has
no power to begin with.
Cleverly, Vbus switching can be
achieved by not much more than a
logic-
level P-channel Mosfet. The
Mosfet’s gate is pulled up until a sink
is connected, switching the Mosfet off.
The sink’s Rd resistor pulls the gate
low and turns the Mosfet on, allowing the source to provide current to
the sink.
The current that the source applies
to the CC line also encodes a Vbus current capacity, and the sink can detect
this by the voltage across the Rd resistor on the CC line. This allows sink
devices to limit their current draw to
what the source can provide.
The second CC conductor can
40
Silicon Chip
also supply power to the electronics
embedded in an ‘e-marked’ cable. In
this role, it is known as Vconn, indicated by the resistors labelled Ra in
Fig.2.
E-marking involves embedding a
chip inside the cable. One use of such
a chip is to indicate to the source that
the cable can handle 5A. Any cable
that is not identified as such is limited to 3A. The CC lines are also used
to communicate other USB-C features,
such as support for alternative modes.
Finally, there are the two SBU (sideband use) lines, which are used by an
audio adaptor accessory mode and the
USB4 protocol. It’s also possible to use
them for a specific custom purpose
through USB-PD negotiation.
In summary, USB-C offers a lot more
variety in voltages, currents and protocols than the legacy standards and connectors. So the USB-C Power Monitor
has more to do than just measure the
current and voltage.
We need to be able to measure voltages up to 48V and monitor currents
up to 5A in either direction. There are
signals, such as those on the CC lines,
that we can monitor for connectivity
and possibly take control of.
We considered including a USB-PD
chip to allow monitoring of the power
delivery communications, but ultimately decided against doing so. Most
of these chips are in leadless packages
(eg, QFN) that are difficult to solder.
We also thought it would be best
to avoid a situation where multiple
chips are trying to communicate over
the same configuration channel bus,
lest that lead to improper voltages
being requested, potentially causing
damage to equipment connected to
the Monitor.
USB-C specifications
It is an unfortunate fact that the
USB-C Power Monitor (or any design
like it) cannot meet the USB-C specifications. As you can see from the
photos, it has a USB-C socket at one
end and a USB-C plug at the other. As
such, it is effectively a USB-C extension cable.
Such cables are simply forbidden by the specification. Consider
what would happen if a 5A e-marked
cable were connected to a 3A extension cable. The signalling for the 5A
cable would be passed through the 3A
cable, and the source would behave as
though both cables are capable of 5A
when they might not be.
Extension cables also diminish the
signal integrity due to the extra junctions between the plugs and sockets
and the cable length. This means that
the higher speed communications are
more likely to be compromised.
Fig.2: the basic signalling used on the CC lines requires sources and sinks to
use specific resistor values (or current sources/sinks) to determine rotation
and the current required. Before a source can apply power on Vbus, it must
detect that a sink is correctly connected.
Australia's electronics magazine
siliconchip.com.au
Providing two USB-C sockets on the
Monitor (effectively turning it into a
cable joiner) would be possible, but
would result in much the same concerns.
The specification also allows cables
to omit some internal conductors,
since the cable orientation sensing can
deal with that scenario. Connecting
two such cables with a joiner could
result in some signals not being passed
through at all.
We performed several speed tests
with different devices through our
Power Monitor, comparing the performance between having the Monitor inline or not. We did not find a
device for which it made a difference.
It takes some powerful gear to run the
high-speed (many GHz) tests needed
to validate USB 3.2 communication,
and we do not have access to such
test equipment.
However, by making the USB-C
Power Monitor as short as practically
possible, we minimise the possibility
of introducing signal integrity problems. Still, we cannot guarantee they
can’t happen.
So, while the USB-C Power Monitor can’t comply with the USB-C
specifications, we have worked hard
and performed some thorough testing
to ensure that it works with as many
devices as possible.
Circuit details
The circuit shown in Fig.3 is split
over two PCBs joined by a 7-pin header
(CON4) at each end. The smaller PCB
(inside the dashed box) has a USB-C
plug (CON1) and a USB-C socket
(CON2). All the Vbus and GND connections are joined at each end to simplify routing; this is actually required
in the USB-C specification.
Apart from a few, all the other lines
are connected straight through, forming the forbidden USB-C extension
lead. The Vbus, CC1 and CC2 lines
each have a resistor between CON1
and CON2. The 15mW shunt resistor
in series with Vbus is for measuring
the current flow.
The resistors inline with CC1 and
CC2 allow us to determine the source
and sink nature of whatever is connected to CON1 and CON2. The extra
resistance is about the smallest we
could use to reliably detect that difference, while being low enough to
not affect the CC detection thresholds
according to the specifications.
siliconchip.com.au
Parts List – USB-C Power Monitor
1 double-sided green PCB coded 04102251, 78 × 11 × 0.8mm
1 double-sided black PCB coded 04102252, 80 × 40 × 0.8mm
1 flat flexible PCB (FFC) coded 04102253, 18 × 40mm
OR 5cm of 7-way ribbon cable
OR 30cm of light-duty flexible cable
1 80 × 40 × 20mm enclosure [Hammond 1551KBK, Altronics H9004]
1 small lithium-ion rechargeable pouch cell with protection circuitry [Altronics S4723]
1 Amphenol 12401981E412A straddle-mount 24-pin USB-C plug (CON1)
[DigiKey, Mouser]
1 Würth 632723100011 24-pin SMT+through-hole USB-C socket (CON2)
[DigiKey, Mouser]
1 5-pin header, 2.54mm pitch (CON3; optional, for ICSP)
1 USB-C SMD power-only socket (CON5) [GCT USB4135 or equivalent]
1 M2016/0806 size 4.7μH 1A inductor (L1) [Murata LQM2MPN4R7NG0L]
1 128×32 pixel 0.91-inch I2C OLED module (MOD1)
3 Adafruit 5410 reverse-mount SMD tactile switches (S1-S3) [DigiKey, Mouser]
1 tube of neutral-cure silicone sealant or similar flexible adhesive
1 piece of foam-cored double-sided tape (to secure BAT1)
Würth 632723300011 is an alternative but it might be harder to solder
🔸
🔸
Semiconductors
1 PIC16F18146-I/SO 8-bit micro programmed with 0410225A.HEX, SOIC-20 (IC1)
1 INA296A3 or INA282 current monitor, SOIC-8 (IC2)
1 AD8541A or NCS325 rail-to-rail CMOS op amp, SOT-23-5 or SC-70-5 (IC3)
1 MCP73831T-2ACI/OT Li-ion charge regulator, SOT23-5 (IC4)
1 MCP16252T-I/CH boost regulator, SOT-23-6 (REG1)
1 3mm bi-colour red/green LED (LED1)
1 BAT54C common-cathode schottky diode, SOT-23 (D1)
Capacitors (all SMD M2012/0805 size X7R ceramic)
5 10μF 16V
5 100nF 50V
Resistors (all SMD M2012/0805 size ⅛W, 1% unless noted)
2 1MW
1 22kW
2 220W
1 390kW
3 10kW
1 100W
1 150kW
6 5.1kW
1 10W
1 120kW
2 1kW
1 15mW M6331/2512 size 3W
Articles on USB technology
If you wish to delve into the technical details of USB, the following Silicon
Chip articles may be of interest:
● USB: Hassle-Free Connections To Your PC by Peter Smith (November
1999): siliconchip.au/Article/4436
● The History of USB by Jim Rowe (June 2021):
siliconchip.au/Article/14883
● How USB Power Delivery (USB-PD) works by Andrew Levido (July 2021):
siliconchip.au/Article/14919
● El Cheapo Modules: USB-PD chargers by Jim Rowe (July 2021):
siliconchip.au/Article/14920
● El Cheapo Modules: USB-PD Triggers by Jim Rowe (August 2021):
siliconchip.au/Article/14996
The following projects may also be of interest:
● USB Power Monitor by Nicholas Vinen (December 2012):
siliconchip.au/Article/460
● USB Cable Tester by Tim Blythman (November & December 2021):
siliconchip.com.au/Series/374
● USB-C Serial Adaptor by Tim Blythman (June 2024):
siliconchip.au/Article/16291
Australia's electronics magazine
August 2025 41
Typical currents in the CC lines are
in the hundreds of microamps (from
source current Ip to sink resistor Rd),
so the drop across the 220W resistors
is in the tens of millivolts. This is the
precision of the voltage thresholds in
the specifications.
When a CC line is allocated to a
Vconn role, it can supply 5V power
at up to 200mA to electronics in an
e-marked cable. Clearly, 220W is too
high to allow 200mA to be passed successfully, although our testing didn’t
find any cables that had problems
with this.
A typical e-marked cable will try
to source Vconn from both ends of the
cable using diodes or similar to prevent Vconn being carried through the
cable. So if you run into problems,
try reversing the Monitor; that should
allow Vconn to be sourced from the
other end.
CON4 breaks out the important signals back to the main PCB for monitoring. There is a ground connection, and
two wires for Vbus, CC1 and CC2, fed
from each side of their respective resistors. CON4’s connections are arranged
symmetrically so that the entire PCB
can be rotated 180°.
This means that if you prefer the
USB-C plug on the left and the USB-C
socket on the right, you can wire the
boards in this fashion during the construction phase.
Main circuit
A microcontroller is required to
make measurements and drive the
display to report them. We’re using
an 8-bit PIC16F18146 (IC1), since it
has some useful internal peripherals,
including a 4.096V reference, an 8-bit
buffered digital-to-analog converter
(DAC) and a 12-bit analog-to-digital
converter (ADC).
IC1 has the usual 100nF bypass
capacitor on its power and ground
pins (1 and 20 respectively). A 10kW
resistor pulls up the MCLR pin (pin
4) to allow normal operation, unless
a programmer is connected at ICSP
header CON3. Pins 1, 4, 18, 19 and 20
are taken to CON3 for programming
and debugging.
Pins 11, 12 and 13 connect to tactile switches S1, S2 and S3. They are
configured with internal pullups, and
the micro detects the pin level changing to low when the switch is pressed,
closing the circuit to ground. A pinchange interrupt allows the micro to be
42
Silicon Chip
woken up from deep sleep by pressing
any of the switches.
These 8-bit PICs have a good output
pin drive strength and so can directly
power other circuit elements, allowing them to be switched off when the
Monitor needs to be in low-power
sleep mode.
Pin 16 powers OLED module MOD1,
a 128×32 pixel monochrome display,
so it can show a few lines of text or
similar. IC1’s pins 14 and 15 provide
a bit-banged I2C serial interface to control the OLED. IC1’s pin 2 is used to
provide power to IC2 and IC3, as well
as to control the ENABLE pin on REG1.
With both pin 2 and pin 16 low
and the microcontroller in low-power
sleep, only IC1 and REG1 draw power
from BAT1. The nominal 5V rail falls
slightly below the cell voltage (because
of the diode and resistor). The greatest current draw is the 6μA flowing
through REG1’s feedback divider, with
IC1 and REG1 drawing less than 1μA
each, for a total under 8μA.
Analog circuitry
IC2 is a current shunt monitor that
amplifies the voltage across the 15mW
current measuring shunt connected
via CON4. It has a gain of 50, so its
output voltage is 0.75V for every amp
through the shunt.
IC2 is equipped with two reference
inputs (REF1 and REF2); the output
voltage is offset against the average of
the voltage at these two pins. The twopin reference feature makes it easy to
set up a mid-rail reference for bidirectional current sensing, although we
don’t use that here.
Instead, we feed both reference pins
with a voltage supplied from pin 17
of IC1. This is derived from an 8-bit
DAC connected to a 4.096V internal
reference. By setting the DAC, we can
change the IC2 reference to be anywhere between 0V and 4.096V.
This allows us to nearly double the
span available for readings. For currents in one direction, we set the reference to near (but not quite) 0V and
we have almost 4V of range, allowing
up to 5A to be measured. If the current reverses direction, the reference
is taken near 4V, allowing similar magnitudes to be measured in the opposite direction.
IC3 is a single op amp configured to
amplify the output from IC2 (relative
to the same reference) by a factor of
100. The 100nF capacitor across the
Australia's electronics magazine
Fig.3: the circuit is split into two
sections connected by CON4s; one
section has a USB-C plug (CON1)
and a USB receptacle (CON2), wired
straight through apart from some
resistors. The other section has
microcontroller IC1, which measure
voltages, currents and so forth and
displays them on the OLED module.
feedback resistor provides low-pass
filtering of the amplified signal. IC1’s
pin 9 is used for ADC readings of the
low range (amplified) voltages from
IC3, while pin 3 samples the higher
range voltages directly from IC2.
The ADC peripheral can perform
differential readings, so the reference
is simply used as the second channel
for these readings. This scheme gives
us more dynamic range to accurately
read currents in both directions over
the two ranges.
The 150kW/10kW divider connected
to one of the Vcc pins of CON4 is
used for measuring the Vbus voltage.
The 100nF capacitor on its lower leg
provides a lower source impedance
to charge up pin 10’s ADC sampling
capacitor. With a 4.096V reference,
the divider allows up to 65V to be
measured, comfortably above the 48V
siliconchip.com.au
currently allowed by the USB-C specification.
CC sensing
The remaining resistors are used to
drive and sample the two CC lines.
Normally, pins 5, 6, 7 and 8 are set as
analog inputs to monitor the state of
the CC lines. The series 5.1kW resistors do not noticeably affect the sensed
voltage when these pins are inputs.
The state of the pins can determine
what current the source can provide.
By monitoring for the slight voltage
difference across the 220W resistors,
it can also determine which end is
the source and which end is the sink.
The 5.1kW resistors also allow the
USB-C Power Monitor to behave as
a power sink by driving one or more
of these pins to a low logic level
(0V). This allows you to check the
siliconchip.com.au
capabilities of a source, even if a sink
is not connected.
It is only a very limited use of the
CC signalling. But USB-C now allows
scenarios which seem improbable; for
example, a mobile phone attempting to
charge a laptop computer. So we think
the ability to identify such situations
could be helpful.
Power supply
One interesting problem is that we
realised a USB-C cable does not always
provide power, even if it is connected
to a suitable source such as a computer
or power supply. Thus, we needed a
way to power the USB-C Power Monitor independently.
We chose to use a lithium-ion
rechargeable battery (BAT1). This has
the advantage that the power supply is
decoupled from the USB-C circuitry,
Australia's electronics magazine
and the USB-C Power Monitor does
not load the circuit under test, apart
from a voltage divider totalling 160kW,
which draws about 31μA at 5V.
We used similar circuitry to our
other recent projects that include a
Li-ion cell. The most recent was the
Compact OLED Clock and Timer (September 2024 issue; siliconchip.au/
Article/16570).
In this case, 5V power for charging
comes in via CON5, a power-only
USB-C socket with the 5.1kW resistors necessary to identify it as a power
sink. The MCP73831 charger, IC4, has
a 10μF smoothing capacitor on its
input pin and another on its output
to the battery.
The MCP73831 monitors the
charging current and voltage of BAT1
and provides a multi-stage charging
regime. It has a status output and the
August 2025 43
The Breakout PCB is used to provide the USB connections and will be described in more detail in next month’s issue.
charging current can be programmed.
All in a tiny 5-pin SOT-23 package!
The charge current is set to 45mA
by the 22kW resistor connected to IC4’s
pin 5 (PROG). The STAT pin drives
one side of bicolour LED1; it is low
during charging and is high when it
is complete. The arrangement of 1kW
resistors allows the LED to light up
red during charging and green when
charging is complete.
Dual diode D1 is a common-cathode
schottky type. Power from the battery
and CON5 are connected to its anodes,
so there is no load on the battery when
power is available at CON5.
The remainder of the circuitry
is powered from D1’s cathode, and
because there is no draw on the battery while charging, IC4 can charge it
fully. This arrangement also means
that the USB-C Power Monitor can be
powered directly from CON5 even if
a battery is not fitted.
The cathode of D1 is connected
to MCP16252 boost regulator REG1,
which also has 10μF bypass capacitors at its input (pin 6) and output (pin
5). This part has an ENABLE input at
its pin 3. When this pin is low, the
REG1’s input is connected directly to
its output, providing a low-quiescent-
current mode with no voltage boost.
When ENABLE is taken high, the
boost regulator operates. It has internal N-channel and P-channel Mosfets.
The SW pin (pin 1) is pulled low by
the N-channel Mosfet, drawing current
through inductor L1. When this Mosfet switches off, the P-channel Mosfet
switches on, dumping current from the
inductor into the output and boosting
the voltage.
Using a Mosfet as an active switch
is more efficient than using a diode,
since there is negligible voltage drop
across the switch when it is on. The
390kW/120kW divider at the output
provides feedback for the regulator
and sets the output to a nominal 5.2V.
After the 10μF capacitor on REG1’s
output is a 10W resistor and a further
10μF capacitor to provide a degree of
filtering for the rest of the circuit. At
the circuit’s nominal 20mA draw, the
10W resistor drops about 0.2V, leaving
close to 5V.
Software
As we noted, the PIC16F18146 has
a handy set of peripherals, so we’ll
describe those next, along with some
aspects of the software operation. The
capabilities of the ADC are critical to
this project.
The PIC provides an internal reference that can be set to a nominal
1.024V, 2.048V or 4.096V, and can
be used to set the ADC scale. The
4.096V reference is used for most
of the analog readings and its measured value (in millivolts) is stored
in non-volatile memory at the time
of manufacture.
Being able to use the 4.096V range
is the main reason for the boost converter (REG1). The 4.096V reference
will not be functional if the supply is
only 3.7V, as might be the case for a
Li-ion cell that is nearly flat without
this boosting.
The analog peripheral is actually
described as an ‘Analog-to-Digital
Converter with Computation’ (ADCC)
module. It can produce a 12-bit result
and can also accumulate multiple
results. The accumulator has 18 bits, so
the software configures it to accumulate 64 12-bit results, giving a notional
18-bit result.
Since the total ADCC error is around
two bits, the lower two bits are discarded, and the software works with
convenient 16-bit numbers. It takes
about two milliseconds to perform
the 64 samples, so the processor sets
it running and uses an interrupt routine to store the readings, then commence the next sample while it continues with other tasks.
A set of 13 samples is taken in
round-robin fashion; the two current
readings are taken in differential mode
using the pin 17 VREF voltage as the
second channel. The voltages across
the 220W resistors in the CC lines are
also measured in differential mode;
this allows us to note the sign and
determine the direction that the current is flowing in those lines.
All these differential readings are
also validated by taking corresponding single-ended (absolute) readings.
For example, this allows us to detect
when the low-range current reading
approaches the rails, an indication
that we should shift the reference (by
setting the DAC) to allow more headroom or use the high-current range
instead.
A second 8-bit DAC is internally
connected to IC1’s supply. The DAC
is set to code 32, or 1/8 of the supply.
The ADCC sample of this channel uses
the 1.024V reference, meaning that we
can accurately check that the supply
voltage is high enough for the 4.096V
reference to work and that the other
readings are correct.
A timer peripheral is also used to
provide an internal clock. This is set to
The “>” button cycles through various
screens, while the up/down buttons are
used to adjust values and modes.
44
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
Screen 1: the main screen reports the Vbus voltage and
current (including direction). Power and accumulated
energy are also calculated; a timer is available too.
Another screen can be used to monitor and test the
condition of the configuration channel lines.
count at an integer fraction of a second
and make the calculations easy (for an
8-bit processor) without using computationally complex floating-point values. The time each sample set takes to
complete is also recorded.
This is necessary because the most
accurate ADCC results can be achieved
when the ADCC is run from its own
internal oscillator and completely
decoupled from the micro’s other
internal clocks.
The appropriate scaling factors are
applied to provide the numerical values that are needed. Derived values
like power (current multiplied by voltage) and energy (power multiplied by
time) are also calculated. This part of
the software also sets flags indicating
which of the low or high current ranges
is valid and should be used.
These measurements continue to
run any time the micro is active and
not sleeping; the results are displayed
according to the screen that the user
has selected. The main screen, for
example, reports the same basic statistics as the older Monitor: voltage, current (now including direction), power
and total energy.
The energy totaliser is paired with
a timer; both can be started, paused
and reset together. The totaliser also
runs in the background if other screens
are selected, providing seamless operation.
Another screen allows monitoring
and control of the CC lines, while a
third can be used to put the Monitor
into low-power sleep mode or check
the battery voltage. You can also access
the configuration screens by a long
press of S3.
Most of the configuration is involved
with trimming the calibration factors
to suit the components used in a specific build, although there are also the
options to adjust the screen brightness
and choose between watt-hours (Wh)
or Joules (J) for the units displayed for
energy. We’ll look more at the user
interface later.
Hardware
The two PCBs fit into a compact 80
× 40 × 20mm enclosure; we used a
siliconchip.com.au
Hammond 1551KBK, which is available from Altronics (Cat H9004). One
PCB replaces the lid, so the final unit
is only 18mm high, but is slightly longer than 80mm because of the plug
protruding at one end.
Both PCBs require reasonably good
soldering skills to build. Fully featured USB-C plugs and sockets are
only available with very tight pin
spacings of around 0.5mm. When the
Breakout PCB is completed, it can be
tested in isolation by using it as an
extension for a USB-C cable. Thus, it
is possible to detect problems early
in the process.
The PCB hosting CON1 and CON2
must be 0.8mm thick because the plug
(CON1) is a so-called ‘straddle-mount’
that clips over the edge of the PCB. It
sounds fussy, but the mounting style
allows the part to be precisely placed
before soldering begins, and we actually found it easier to work with than
the socket (CON2).
The upper PCB is populated with
M2012 (imperial 0805) sized passives,
which are 2mm long. This is about
the minimum size we consider easy
to handle. There are a couple of SOIC
chips and a few SOT-23 variants with
three, five and six pins.
Since the top PCB is also the lid,
there are a couple of unusual constructions steps. If you have built
the likes of the Compact Clock and
Timer (or read its article), these will
be familiar.
The two PCBs must be joined with
wires. Our early prototypes used pluggable headers, but we found that these
had enough variation in resistance to
interfere with the performance of the
current-measuring shunt.
We designed a flat flexible cable
(FFC) that can be soldered to the PCBs
to give a simple and elegant connection. You can also use hookup wire
for this, or a section of 7-way ribbon
cable, if you find the FFC is too expensive or difficult to source (we’ll supply
the FFC with our kits – see the panel).
We had originally planned to use
a 10440-sized lithium-ion cell as the
battery. These are the same physical
size as a AAA cell, and there is just
enough room to fit a AAA cell holder
to allow simple and safe connection
of the cell.
However, we think that our final
design, using a small, rectangular
pouch cell like Altronics’ Cat S4723,
works better. The slim shape is a better fit for the available space, and the
nominal 400mAh capacity is higher
than we have seen in 10440 lithium
cells. This cell measures 38 × 25 ×
6mm and includes protection circuitry.
Many small pouch cells are available from various online sellers. So
you might find an alternative with
slightly different dimensions that
still fits.
Until next month
This is a very detailed project, and
we have run out of room to describe
it in this issue. The second part next
month will include the construction,
setup, calibration and use of the USB-C
SC
Power Monitor.
Silicon Chip kcaBBack Issues
$10.00 + post
$11.50 + post
$12.50 + post
$13.00 + post
January 1997 to October 2021
November 2021 to September 2023
October 2023 to September 2024
October 2024 onwards
All back issues after February 2015 are in stock, while most from January 1997 to
December 2014 are available. For a full list of all available issues, visit: siliconchip.com.
au/Shop/2
PDF versions are available for all issues at siliconchip.com.au/Shop/12
We also sell photocopies of individual articles for those who don’t have a computer
Australia's electronics magazine
August 2025 45
RP2350B
development board
Think of this as the Pico 2’s bigger sibling – more pins, more I/O, more speed,
more storage and more memory. It’s perfect for breadboarding, too. Also, like the
Picos, it can be purchased pre-assembled!
T
he Raspberry Pi Pico is well known
to our readers, and has been
designed into many of our projects
as a drop-in, self-contained compute
module. It has been a runaway success
for the Raspberry Pi Foundation, with
over 4 million of this handy module
sold since its launch in 2021.
In August 2024, they improved on
it with the Pico 2, based on the new
RP2350A processor, which has also
been very popular.
The RP2350A processor also has
a lesser-known sibling called the
RP2350B, which is often overlooked.
This has the same features as the A
variant, but comes in a larger 80-pin
QFN package with 48 I/O pins vs the
30 I/O pins of the RP2350A.
The Raspberry Pi foundation does
not offer a module based on this chip,
but recently they have made it available for individual sale – so now
we can design our own RP2350B
based module. Thus, we present the
RP2350B Development Board.
This is similar to the Raspberry Pi
Pico 2, but it uses the RP2350B, with
nearly all of its 48 I/O pins available
for experimenters – a vast improvement over the Pico 2. This board can
be used as a general-purpose compute module when you need a lot of
I/O pins, or as a development board
while designing a circuit around the
RP2350B chip.
Features
The RP2350B Development Board
is designed to suit solderless breadboards with two rows of 32 pins on a
2.54mm/0.1-inch pitch. This is similar
to the standard Raspberry Pi Pico, and
this layout can also be used as a plug-in
module in your own PCB designs.
In our module, 47 of the RP2350B’s
48 I/O pins are routed to the edge connectors. This is useful when you need
a lot of I/O; for example, when driving
a high-performance LCD panel with
a parallel interface, or constructing
your own multi-key keyboard. The
one I/O pin not available, GPIO00,
is used as the PSRAM IC chip select
(CS) signal.
On the edge pins of the module, we
have also included numerous grounds,
plus +3.3V and +5V outputs.
One extra benefit of the RP2350B is
that eight of the I/O pins are capable of
analog measurements (vs four on the
RP2350A), and these are also available
for use in your programs. The diagram
opposite lists the full capabilities of
each I/O pin.
The module is self-contained,
including a 3.3V regulator. The input
power supply is nominally 5V, but
∎ Processor: Raspberry Pi RP2350B
∎ Cores: two ARM Cortex-M33 and two Hazard3 RISC-V
∎ Clock Speed: default 150MHz; overclockable up to around 400MHz
∎ Flash memory: 16Mbytes
∎ RAM: 520kiB, expandable to over 8MiB
∎ I/O pins: 47 (eight with analog capability)
∎ I/O connectors: two rows of 32 pins, 2.54mm/0.1-inch pitch, 25.4mm/1-inch
separation
∎ Power supply: 5V nominal (4.5-12V) <at> 95mA (150MHz clock)
∎ Size: 82 × 28mm
Words by Geoff Graham | Design by Peter Mather
RP2350B Assembled Board (SC7514, $30): includes a fully-assembled PCB
with nearly everything from the parts list, except for the optional components
a range of 4.5-12V is acceptable. A
USB-C socket is provided for power
and loading the firmware. The firmware loading process works exactly the
same as with the Raspberry Pi Pico, ie,
you hold down the BOOT switch while
plugging the USB into your desktop or
laptop computer.
The flash memory used for storing
programs and data in this design has
a capacity of 16Mbytes (the Pico 2 has
4Mbytes). The PicoMite BASIC interpreter occupies just 2Mbytes, which
leaves plenty of flash free to create an
internal “disk drive” with a capacity
of about 14Mbytes.
Our design also supports an 8Mbyte
PSRAM chip. This sits on the same
quad SPI bus as the flash memory,
and can be used to add to the internal
RAM of the RP2350B.
Overclocking
The RP2350B Development Board
is designed for overclocking, which
means running the processor cores at
a higher clock speed than specified in
the data sheet. This enables the module to be used in high-performance
applications, such as generating DVI/
HDMI video.
The default speed for the RP2350B
is 150MHz, but some people have
claimed to have overclocked it to over
600MHz. A more reasonable goal is
the 372MHz needed to generate DVI/
HDMI video.
To support the faster speeds, our
design uses an adjustable linear regulator to generate the digital core supply voltage (DVDD). This powers the
chip’s core digital logic, and in our
design, can be accurately adjusted
from 1.1V to over 1.4V. Higher voltages allowing the CPU to run faster.
In the Pico 2, this voltage is provided by an on-chip switching regulator, which is not suited to high
siliconchip.com.au
PWM
PWM0B
SERIAL
I2C
SPI
COM1 RX I²C SDC
SPI
I2C
GPIO47
SPI2 RX
I²C2 SDL
Pin
Pin
5V
5V
3.3V
3.3V
GND
SERIAL
PWM
PWM11B
GPIO01
GPIO46
SPI2 CLK I²C2 SDA
PWM1A
I²C2 SDA SPI CLK
GPIO02
GPIO45
I²C SDL
COM1 RX
PWM10B
PWM1B
I²C2 SDC SPI TX
GPIO03
GPIO44
SPI2 RX
I²C SDA
COM1 TX
PWM10A
GND
GND
GPIO04
GPIO43
SPI2 TX
I²C2 SDL
GPIO05
GPIO42
SPI2 CLK I²C2 SDA
I²C SDL
COM2 RX
PWM8B
SPI2 RX
I²C SDA
COM2 TX
PWM8A
PWM2A
COM2 TX I²C SDA
PWM2B
COM2 RX I²C SDC
SPI RX
PWM3A
I²C2 SDA SPI CLK
GPIO06
GPIO41
PWM3B
I²C2 SDC SPI TX
GPIO07
GPIO40
GND
GND
GPIO08
GPIO39
SPI TX
I²C2 SDL
GPIO09
GPIO38
SPI CLK
I²C2 SDA
PWM4A
COM2 TX I²C SDA
PWM4B
COM2 RX I²C SCL
SPI2 RX
PWM5A
I²C2 SDA SPI2 CLK
GPIO10
GPIO37
PWM5B
I²C2 SCL
SPI2 TX
GPIO11
GPIO36
GND
GND
PWM6A
COM1 TX I²C SDA
SPI2 RX
GPIO12
PWM6B
COM1 RX I²C SCL
GPIO13
PWM10B
COM2 TX
PWM10A
GPIO35
SPI TX
I²C2 SDL
GPIO34
SPI CLK
I²C2 SDA
PWM7B
I²C2 SCL
GPIO15
GPIO32
GND
GND
GPIO16
GPIO17
COM1 RX I²C SCL
SPI RX
PWM9A
COM1 RX
PWM8B
SPI RX
I²C SDA
COM1 TX
PWM8A
GPIO31
SPI2 TX
I²C2 SDL
PWM7B
GPIO30
SPI2 CLK I²C2 SDA
PWM7A
I²C2 SDA SPI CLK
GPIO18
GPIO29
PWM1B
I²C2 SCL
GPIO19
GPIO28
GND
GND
GPIO20
SPI RX
PWM9B
I²C SDL
PWM1A
SPI TX
PWM11A
I²C SDA
GPIO33
COM1 TX I²C SDA
PWM11B
SPI RX
GPIO14
PWM0B
PWM9A
COM2 RX
I²C2 SDA SPI2 CLK
PWM0A
PWM9B
I²C SDL
PWM7A
SPI2 TX
PWM11A
I²C SDL
COM1 RX
PWM6B
SPI2 RX
I²C SDA
COM1 TX
PWM6A
GPIO27
SPI2 TX
I²C2 SDL
PWM2A
COM2 TX I²C SDA
PWM2B
COM2 RX I²C SCL
GPIO21
GPIO26
SPI2 CLK I²C2 SDA
PWM3A
I²C2 SDA
GPIO22
GPIO25
I²C SDL
COM2 RX
PWM4B
PWM3A
I²C2 SCL
GPIO23
GPIO24
I²C SDA
COM2 TX
PWM4A
3.3V
3.3V
SPI TX
clock speeds because of the electrical
noise it generates. Additionally, it is
difficult to implement, as it requires
a specialised inductor, which is hard
to find.
In our design, the DVDD voltage
is provided by a TPS7A7002DDAR
linear regulator (REG34) that is both
inexpensive and does not generate
any electrical noise. The onboard
trimming resistor (VR1) is used to
set the DVDD voltage. Note that it is
important that this is set before power
is applied to the board. If DVDD is
accidentally set too high, it can damage the RP2350B chip.
We have also used an integrated
crystal oscillator to generate the base
clock of 12MHz for the RP2350B. This
is different from Raspberry Pi Pico 2,
which uses a simple crystal for this
purpose. The integrated crystal oscillator provides a more stable clock with
much less jitter. Jitter can be a problem
siliconchip.com.au
when the base clock frequency is multiplied many times in the RP2350B to
give the core CPU clock.
Development environments
All the familiar development environments used with the Raspberry Pi
Pico 2 can be used with this board.
This includes:
∎ The official Raspberry Pi C SDK
for C/C++ development, which can
be used from the command line on a
desktop or laptop computer, or within
popular integrated development environments like Visual Studio Code (VS
Code), Eclipse and Clion.
∎ MicroPython, which is a full
implementation of the Python 3 programming language running directly
on the Development Board. This
includes an interactive prompt to
execute commands immediately via
a USB serial port.
∎ Our own PicoMite firmware,
Australia's electronics magazine
SPI2 RX
PWM5B
PWM5A
ANALOG PIN
which implements a feature-rich
BASIC interpreter (MMBasic) with
support for audio, LCD panels, SD
cards, game controllers, HDMI/VGA
video and PS2/USB keyboards. This
firmware includes its own full-screen
editor so programs can be developed,
tested and run on the development
board in a highly productive environment.
Circuit details
Fig.1 shows the full circuit for the
RP2350B Development Module. At the
centre is the RP2350B processor. All its
general-purpose I/O pins are routed to
the two connectors at the edges of the
The RP2350B has
the same features
as the RP2350A
in the Pico 2 but
has more pins,
including 48
GPIOs.
August 2025 47
Fig.1: the circuit diagram for the RP2350B Development Board/Module. USB-C socket CON2 is used both for supplying
power (5V DC) and communicating with the RP2350B. The USB 5V supply is regulated to 3.3V by REG21 to power
oscillator XO4, PSRAM IC33 and flash chip IC6. The RP2350B's nominally 1.1V core supply is generated by adjustable
regulator REG34, so it can be increased for overclocking.
48
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
PCB (CON35/36), except for GPIO00,
which is used as the chip select signal
for the PSRAM (IC33).
One other pin (GPIO25) is special as
it drives the onboard LED, LED2. If you
do not need this indicating function,
the I/O pin can be used as a general
purpose I/O with the extra load presented by the LED and its 1kW current-
limiting resistor.
The input power for the board is
a nominal 5V DC; the RP2350B and
the other chips on the board run from
3.3V, which is supplied by a simple
AMS1117-3.3 low-dropout linear regulator.
The official Raspberry Pi Pico modules have a much more complex design
for the power supply, using a switching regulator, but this can cause significant electrical noise that interferes
with analog measurements and sensitive circuits such as audio input/
output.
Most designs do not need the wide
voltage range of the switching regulator, so our design avoids the noise
problems and still provides a useful
input supply voltage range of 4.5-12V.
The RP2350B needs another voltage
supply called the digital core supply
(DVDD), which we mentioned earlier
as essential for overclocking. This
powers the chip’s core digital logic. In
our design, it is provided by REG34, an
adjustable linear regulator controlled
by trimpot VR1.
IC6 is a W25Q128JVSIQ 128Mbit
(16Mbyte) flash memory chip made
by Winbond Electronics. It uses a
quad SPI interface and is designed for
true XIP (execute in place) operation,
which allows the RP2350B to execute
its program directly from this chip.
The W25Q128JVSIQ can operate
with high clock speeds on the SPI
interface (133MHz); this is important
when overclocking the RP2350B. The
RP2350B also has a built-in SRAM
cache, which operates to mitigate the
effect of the relatively slow quad SPI
bus interface.
The BOOT pushbutton switch (S16)
pulls the chip select line (CS) low on
the flash memory chip, which essentially disables the flash memory. When
power is applied, the RP2350B will
interpret the disabled memory as a signal to enter its bootloader mode, which
is used to load a new firmware image.
IC33 is an optional external PSRAM
chip (APS6404L-3SQR-SN), which
sits on the same quad SPI bus as the
siliconchip.com.au
Parts List – RP2350B Development Board
1 double-sided PCB coded 07107251, 82 × 28mm
2 momentary SMD tactile pushbutton switches (S15, S16)
[XKB Connectivity TS-1187A-B-A-B]
1 USB-C USB 2.0 data + power socket (CON2) [Kinghelm KH-TYPE-C-16P]
1 50kW 3.8 × 3.6mm SMD trimpot (VR1) [Bourns TC33X-2-503E]
2 32-pin male headers, 2.54mm pitch (optional)
Semiconductors
1 128Mbit QSPI flash memory, SOIC-8 (IC6) [Winbond W25Q128JVSIQ]
1 Raspberry Pi RP2350B microcontroller, QFN-80 (IC28)
1 APS6404L-3SQR-SN 8MiB PSRAM, SOIC-8 (IC33; optional)
1 12MHz oscillator module, 3.2 × 2mm SMD-4 (XO4)
[TOGNJING XOS32012000LT00351005]
1 AMS1117-3.3 low-dropout 3.3V linear regulator, SOT-223-3 (REG21)
1 TPS7A7002DDAR adjustable low-dropout voltage regulator, SOIC-8
(REG34)
1 white SMD LED, M1608/0603 size [KT-0603W]
Capacitors (all SMD multilayer ceramic capacitors)
3 10μF 50V X5R, M3216/1206 package [Samsung CL31A106KBHNNNE]
18 100nF 16V X7R, M1206/0402 package [Samsung CL05B104KO5NNNC]
1 10μF 10V X5R, M1608/0603 package
[Samsung CL10A106KP8NNNC]
Resistors (all SMD M1206/0402 ±1% unless noted)
1 1MW
2 1kW
2 20kW
1 33W
5 10kW
2 33W (0603 size)
2 5.1kW
2 150W (0603 size)
The finished RP2350B
Development Board shown at
75% of actual size. The 32-pin
headers are not included with
the assembled board
flash memory chip. This has a capacity of 64Mbits (8Mbytes) and it can be
used to augment the internal RAM of
the RP2350B. How it is actually used
depends on the running program.
For example, PicoMite BASIC will
automatically add it to the general-
purpose RAM seen by the BASIC interpreter, allowing for very large arrays
to be defined.
Because the PSRAM must communicate over a serial interface, it is a lot
slower than the internal RAM of the
RP2350B. It also can limit the amount
of overclocking that the board is capable of; however, it should still reach
the speeds needed for generating DVI/
HDMI video.
The internal RAM is normally more
than enough for most applications, so
for this and performance reasons, the
PSRAM location is not populated in
our design. However, it can be easily
added to the BOM (Bill of Materials)
for automated assembly or, as it is in
an easy-to-solder 8-pin SOIC package,
you can add it yourself.
Australia's electronics magazine
This chip is available from Mouser
for around $3 in small quantities:
https://au.mouser.com/ProductDetail/
878-APS6404L-3SQR-SN
Building it
The RP2350B chip comes in an
80-pin QFP package, which is designed
for automated surface-mount soldering. It can be hand soldered, but this is
a challenge, even for someone skilled
in SMD soldering.
So, practically speaking, you have
two options for obtaining this module. You can purchase it fully assembled from the Silicon Chip Online
Shop, or you can use an SMD assembly service such as JLCPCB’s to build
the board.
We recommend JLCPCB (https://
jlcpcb.com) for the automated assembly, as they have proved reliable and
reasonably priced in the past. JLCPCB
also source the components at a good
price and they do everything, including making the board, making the solder stencil, applying the solder paste,
August 2025 49
Resistance (TP1-DVDD)
DVDD
6.0kΩ
1.10V
9.0kΩ
1.15V
12.0kΩ
1.20V
15.0kΩ
1.25V
18.0kΩ
1.30V
21.0kΩ
1.35V
24.0kΩ
1.40V
27.0kΩ
1.45V
30.0kΩ
1.50V
33.0kΩ
1.55V
36.0kΩ
1.60V
Fig.2 (above): the
overlay diagram
for the RP2350B
Development
Board. The
table at the top
of the page can
be used as a
reference for
overclocking
the RP2350B
IC.
placing the components and reflow
soldering them.
Their minimum quantity for assembly is two boards. However, if you do
want a number of boards, it is hard to
see why you would want to undertake
the hand assembly of this board when
the automated assembly option is relatively cheap.
To order boards from JLCPCB, you
will need to download three files from
siliconchip.au/Shop/10/2832 (they
all come in one download package).
The package includes a ZIP file with
the Gerber files that contain the PCB
design, a Bill of Materials spreadsheet
listing all the parts required, and a
CPL spreadsheet that contains the
placement information for the pickand-place robots.
Ordering assembled boards from
the JLCPCB website is reasonably simple. Go to https://jlcpcb.com and create an account with them. Then, drag
and drop the “RP2350B Development
Board Gerbers.zip” file onto the “Add
gerber file” box to the left of the Instant
Quote button on their front page.
The website will process the file and
display an image of the PCB. Click on
the switch marked “PCB Assembly”,
then click “NEXT” until you reach a
screen that prompts you to drag and
drop the “RP2350B Development
Board BOM.xlsx” and “RP2350B
Development Board CPL.xlsx” files.
After supplying those files, you
can continue and then accept all the
defaults. However, you may wish to
change the quantity of PCBs made
(minimum of 5), and the number that
you want assembled (minimum of 2).
Note that boards purchased from
the Silicon Chip shop or assembled
by JLCPCB will not include the two
32-pin headers required for use with
a solderless breadboard. You will have
to add these yourself (if required).
Adjusting the DVDD voltage
When you receive the boards, there
is one adjustment that you need to
make: using VR1 to set the digital core
supply voltage (DVDD).
The RP2350B Development Board has all its
components mounted on the top, with the pin
designations listed on the reverse. It is available
fully-assembled, apart from the two pin header
strips. The module is quite small at 82×28mm
(shown here enlarged for clarity). It is designed to
suit solderless breadboards, with two rows of 32
pins on a 2.54mm/0.1in pitch, or it can be used as
a plug-in module in your own PCB designs.
50
Australia's electronics magazine
Important: the potentiometer must
be adjusted before applying power
to the board. Leaving it in a random
position may damage the RP2350B
chip.
Set your multimeter to the resistance
mode and, with the board unpowered,
place the leads across the test points
marked DVDD and TP1. Adjust potentiometer VR1 to give a reading of 6kW.
This will set DVDD to 1.1V, which
is the standard voltage for a clock
speed of 150MHz (the default for the
RP2350B). This should also work for
clock speeds up to 250MHz.
If you wish to overclock the
RP2350B, you need to do two things:
increase DVDD and set the desired
clock speed in the program by setting
CPU registers. Typically, a DVDD of
1.3V will allow the RP2350B to run
up to 400MHz, with intermediate values suitable for clock speeds between
250MHz and 500MHz. The table at left
lists some resistance values and the
resultant DVDD voltages.
A maximum of 1.4V for DVDD
should be safe. However, if you wish,
you can try higher voltages with a risk
of damaging the RP2350B processor.
When the board is powered, this setting can be checked by measuring the
voltage between the DVDD test point
and any GND point.
Note that overclocking the RP2350B
Development Board is not guaranteed,
although all the samples we tested
have reached a speed close to 400MHz.
Also note that when a PSRAM chip is
fitted, the maximum overclock speed
will typically be slightly reduced.
Using the module
Power for the board is supplied via
the USB-C connector. This is the normal mode when you are developing
a program, as the board will be connected to a desktop or laptop computer that is used to load or edit the
program.
When the board is used as an embedded controller (ie, not connected to a
computer), power can be supplied via
a 5V pin on the edge connector. This
supply can be 4.5-12V. In this case, you
cannot use the USB connector at the
same time, as that would cause a conflict between the two power supplies
and possibly damage your computer.
To prevent this possibility, you can
use a switch or jumper to isolate your
power supply whenever the USB connector is used.
siliconchip.com.au
◀ The RP2350B Development Board is
ideal when you need many I/O pins.
One example is driving a high quality
LCD panel with a parallel interface;
this can require 15-23 I/Os, difficult
for a Pico 2 to accommodate but easy
with our module.
Some development environments,
such as MicroPython and the PicoMite
BASIC interpreter, use the USB connector with a terminal emulator on a
computer to edit and manage the programming environment.
Other, hosted development environments, such as the C/C++ compiler,
will build the program on the desktop
or laptop computer, which then needs
to be transferred to the module.
To load this firmware, you simply
hold down the BOOT button while
restarting the module (the RESET button is good for that) and then copy the
firmware to the pseudo USB drive that
is created on your computer by the
RP2350B chip.
How you use the RP2350B Development Board will depend on the
firmware that you have running on it.
The PicoMite BASIC interpreter can
be downloaded from siliconchip.au/
Shop/6/833
This is a complete OS with a Microsoft BASIC compatible interpreter and
extensive hardware support, including
HDMI/VGA video, PS/2 and USB keyboards, touch-sensitive LCD panels,
SD cards and much more.
For a full
description of
the PicoMite firmware, read the article
on the Pico
Mite 2 (February
2025; siliconchip.au/Article/17729)
or visit the author’s website at https://
SC
geoffg.net
Songbird
An easy-to-build project
that is perfect as a gift.
SC6633 ($30 plus postage): Songbird Kit
Choose from one of four colours for the PCB (purple, green, yellow or red). The kit includes nearly all
parts, plus the piezo buzzer, 3D-printed piezo mount and switched battery box (base/stand not
included). See the May 2023 issue for details: siliconchip.au/Article/15785
siliconchip.com.au
Australia's electronics magazine
August 2025 51
RIGOL DHO924S
Digital Oscilloscope
Review by Tim Blythman
It’s been a while since we have reviewed an oscilloscope, and Rigol’s new
DHO900 series has a slew of modern features with a compact footprint. This
review covers these new features and looks at what it’s like to use out-of-thebox. We’ll also have a brief look at their DM858 digital multimeter.
4 channels
125MHz or 250MHz bandwidth
Parallel, UA
- RT, I2C, SPI, LIN, CA
- N protocol analyser
Sampling rate of 1.25Gsps (shared between active channels)
Optional 25MHz, 10V peak-to-peak waveform generator
12 bits analog resolution
7in touchscreen LCD (1024 × 600 pixels)
Sine, square, ramp, noise, arbitrary waveforms (from CSV file)
Sample memory of 50 million points
Period, frequency, rise time, fall time, duty, count, delay, phase and more measurements
Maths functions, FFT, Bode plot (using the A
- FG as a source), pass/fail
52
USB device, HDMI, LA
flash drive) interfaces
- N, USB host (mouse,
Silicon Chip
Australia's electronics magazine
15V DC, 3A
- (65W USB-C PD) PSU
siliconchip.com.au
A
fter seeing this oscilloscope at Electronex and being impressed,
Emona loaned us a unit to
review. The first thing we noticed
about the DHO924S is its compact size.
For comparison, the Rigol MSO5354
that we reviewed in February 2019
(siliconchip.au/Article/11404) measures 37 × 20 × 13cm (9.62L), while
the DHO924S is only 26 × 16 × 8cm
(3.33L). The enclosure is also a striking black colour.
The DM858 digital multimeter
appears to use the same case as the
DHO924S, so it is a similar size, shape
and layout. We also borrowed one of
those; our review of it is in a panel
later in the article.
The DHO900 series
The DHO900 series comprises four
models: DHO914, DHO914S, DHO924
and DHO924S. S-suffixed parts
include an arbitrary function (waveform) generator output, while the
DHO914 variants offer 125MHz bandwidth against the DHO924’s 250MHz
bandwidth. The DHO924S has the
highest specifications in the series.
Rigol’s DHO naming refers to a
high-resolution digital oscilloscope;
it has a 12-bit ADC (analog-to-digital
converter), while many ‘scopes we
have seen only offer 8-bit resolution.
There are four analog inputs, so that’s
the ‘4’ in the naming scheme.
The DHO924S also has a 16-input
logic analyser, but an optional active
logic probe is required to use this feature, so we did not test it. We find
that four channels are ample for most
scenarios, and the analog channels
can also be used as inputs to digital
features, such as the serial protocol
decoder.
There is also a DHO800 series,
which offers 12-bit resolution in a mix
of two- and four-channel options but
lacking the logic analyser option. They
have a lower bandwidth and lack some
of the other features of the DHO900
‘scopes, but are similar in that they
have the same compact form factor.
The legs pivot out during use or
fold flat for storage. When the legs are
extended, the unit leans back about
20°. The legs and base of the unit have
chunky rubber feet and the ‘scope does
not feel like it will tip or slide away
while the controls are being operated.
There is an Earth socket for a banana
plug, and the brass insets are threaded
for M4 VESA mounts with a 100mm
spacing. With the DHO924S being so
thin, it becomes practical to mount
it to an adjustable monitor arm, so it
no longer takes up any desk space.
Apparently, users are also 3D-printing
a variety of adaptors to suit the VESA
mounts to adapt to their ‘scopes.
The four input BNC connectors are
along the bottom of the ‘scope next to
the ground and test signal clips. The
connector for the optional active logic
probe is under the display, along with
a USB-A socket and the power button.
Compared to the likes of the Rigol
MSO5354, there is only one pair of
vertical scale knobs, despite there
being four channels. The active channel is selected by the same numbered,
colour-coded pushbuttons that are
used to switch channels off and on.
We found this to be intuitive enough
and, as it offers more room for the
main display, it seems like a reasonable compromise.
The oscilloscope package includes
four switchable probes (more on them
later), a USB-A to USB-B cable, an
Earth lead terminated with banana
plugs and a USB-C PD (power delivery)
power supply, which indicates that it
can deliver 5V, 9V, 12V, 15V or 20V.
The oscilloscope uses 15V, according
to our USB-C Power Monitor.
The USB-C plug has a pin and latch
arrangement that secures it. That
means that the power supply cannot be
easily used to power other devices, but
does not stop another suitable power
supply from being used to power the
oscilloscope instead.
Being an isolated power supply
necessitates the inclusion of the Earth
lead, since USB-C does not provide a
means for Earthing. In other words, the
oscilloscope is floating unless the separate Earth lead is connected.
This is an interesting pitfall that
may not be immediately apparent to
those accustomed to ‘scopes that are
normally Earthed via their mains lead.
Also, since it isn’t directly mains-
powered, it is not possible to trigger
off the mains waveform without a separate connection.
After powering on the ‘scope, it took
almost a minute to start up and display the expected screen. Along the
way, the message “Android starting”
appeared, hinting at the underlying
software. The DHO924S we tested ran
Android 7.1.2, which dates from 2019.
The LCD has a touch panel, which
works as you might expect for an
Android device. You can tap on virtual
buttons, drag the cursors around and
even perform a pinch-to-zoom. There
is a “Touch Lock” button to disable
the touch panel.
Out of the box experience
The photos show the front and back
of the DHO924S. The back panel has
two BNC connectors; one of these is
for the arbitrary function generator, the
other an auxiliary output. By default,
the latter emits a pulse when a trigger
event occurs. There are also Ethernet,
USB-B, HDMI and USB-C connections.
siliconchip.com.au
There are several ways to connect to a PC or external monitor. The VESA
mounts allow the unit to be mounted on a monitor arm, freeing up bench space.
While this is a DHO800 series ‘scope, the back of the DHO900 series is identical
except for having a black case.
Australia's electronics magazine
August 2025 53
Four PVP2350 350MHz switchable passive probes are included, along with the
accessories shown here. The ‘scope also includes a power supply & USB cable.
The ‘R’ icon at the bottom-left corner of the display opens a menu that
includes features beyond what you
might expect from a traditional ‘scope,
such as settings, operating system utilities and the like. Screen 1 shows the
contents of this menu.
The front panel looks much like any
other ‘scope, with the familiar adjustment knobs for vertical and horizontal
scale, trigger and RUN/STOP controls,
along with other controls to operate the
custom features that are seen in modern oscilloscopes. The “Quick” button
can be programmed to perform one of
several different actions; by default, it
saves a screenshot.
The two ‘Flex Knobs’ near the top
do not have a fixed use; their function changes depending on the items
selected in various menus. The functions are marked on-screen by small
‘1’ and ‘2’ icons.
The Flex Knobs allow all manner of
values to be adjusted instead of being
manually entered into a keyboard on
the touch panel. This will be handy
when values just need to be tweaked
by a small amount. These knobs can
also adjust the cursors when they are
turned on.
Probes
Screen 1: despite the numerous features, it’s easy to find most menu options. A
good place to explore is the R menu at the lower left corner of the screen. This
also gives a good overview of the advanced features.
The DHO924S comes standard with
four Rigol PVP2350 350MHz passive
probes, which are switchable between
10:1 and 1:1 attenuation (as usual, the
full bandwidth is only available at
10:1). Each probe includes an assortment of colour-coding rings and a
ground spring, as well as the requisite
compensation adjustment tool.
The leads are 1.2m long, and their
slim cords are light enough to not take
up too much space. There is no probe
detection (for automatic probe attenuation setting), so these must all be
set manually.
Using it
Screen 2: once we had the ‘scope’s IP address from this screen, it was easy to
connect to the Web Control interface. The Utility menu also includes the setup
and self-calibration options.
Since we were keen to try out
the more modern features of the
DHO924S, we hooked up an Ethernet
cable. We found the IP address of the
‘scope from the Utility menu, as seen
in Screen 2. Typing that into a browser’s address bar gave us a Web Control
Page, allowing us to open a Web Control window that shows the oscilloscope’s screen.
Screen 2 was captured on our PC
using the Web Control window; it’s
identical to what appears on the
Australia's electronics magazine
siliconchip.com.au
54
Silicon Chip
◀ Screen 3: hooking the ‘scope’s
function generator back to one of its
inputs shows off its sensitivity. The
2mV peak-to-peak square wave is the
lowest amplitude that it can deliver.
Screen 4: tapping on each channel’s
vertical settings shows the full signal
path and its associated parameters.
You can still see the waveform under
the transparent window, allowing the
trace to be adjusted with ease.
‘scope’s screen and even gives access
to the controls that would otherwise
require the touch panel. Any device
with a browser and WiFi connection
should work. We had no trouble controlling and viewing the ‘scope on an
Android mobile phone’s browser.
This makes it much easier to explore
the features of the ‘scope, although we
think it’s a bit of an omission that there
aren’t controls for the various hardware buttons and knobs. Still, most
settings can be set via menu panels.
The HDMI interface just works. We
plugged in a HDMI cable, connected it
to a monitor and the ‘scope’s display
appeared on the screen without having to change any settings. The output
appears to be identical to that on the
LCD, scaled up to use the entire viewable area of the monitor.
We’ll delve further into the various
interfaces a bit later, including some
software that Rigol offers. The relevant downloads can be found at www.
rigolna.com/download/
There is a self-calibration mode that
is recommended to be performed after
the ‘scope has warmed up to operating temperature (after about 30 minutes). The self-calibration process took
about 24 minutes to run on the unit we
were testing.
The sampling rate is shared between
the four channels, since there is only
one analog-to-digital converter IC.
Using two channels halves the available sampling rate, while using three
or four channels will reduce it to a
quarter. So the maximum sampling
rate can only be achieved if only one
channel is in use. Similarly, the sample memory is also shared between the
channels in use.
Noise performance
Using the inbuilt arbitrary function
generator, we looped a 2mV peak-topeak 1kHz square wave signal back
into the ‘scope with no attenuation and
with the bandwidth limited to 20MHz.
The result is seen in Screen 3. Some
of this noise will be from the function
generator, but clearly, the DHO924S
has no trouble resolving signals below
1mV, which is pretty impressive.
Inputs
The front of the DHO924S is compact, thanks to the
channel vertical controls being shared. The channel to
adjust is selected by one of the numbered buttons. Above these are
the Flex Knobs, which adjust values depending on the current sub-menu.
siliconchip.com.au
Australia's electronics magazine
Opening the menu options for the
inputs reveals the signal flow diagram
seen in Screen 4. Some of the options
can be adjusted by tapping on the diagram itself. You can use the Flex Knobs,
as indicated by the yellow hexagons.
August 2025 55
Screen 5: measurements are added to
the Result panel at right. This panel
can scroll up and down, so more than
five results can be added. The options
relating to the horizontal (time) axis are
shown here.
Screen 6: this display has both reference
waveform (orange) and pass/fail mask
(blue) active. The auxiliary output at the
rear of the ‘scope can be set to produce
a pass/fail signal.
Screen 7: a Bode plot expands to take
up most of the available screen space.
Most of the windows are movable and
adjustable, so you can customise the
viewport to suit your preferences.
Screen 8: the ‘scope’s help system
includes a PDF copy of the manual that
can be viewed on the 7-inch display.
56
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
For numerical values, a pop-up keyboard can also be used to enter values.
The main parameters are also displayed on the channel widget near
the bottom of the screen, and you can
see the appropriate parameters for the
arbitrary function generator displayed
on the G channel too.
Maths and measurements
Handy features on most modern
‘scopes include various mathematical
functions and trace measurements; the
DHO924S is no exception. The measurement menu can be accessed from
a fixed button on the right-hand side
or via the ‘Measure’ button at the topright of the screen, on the touch panel
or even via the main ‘R’ menu button.
The area at top-right actually hides
several different buttons that can be
accessed by swiping left or right. Fortunately, most menu items can be
accessed in different ways, so you
can choose whatever option is most
intuitive.
There are 41 different measurements
available; Screen 5 shows some of
these active, as well as the horizontal
(time-based) measurements that are
available. The vertical options include
voltage-based (peak-to-peak, RMS etc)
measurements, while the remainder
are time and phase delay measurements between two input channels.
The Setting option allows thresholds to be set. These default to 10%,
50% and 90% levels of the waveform,
which worked quite well for us during
our tests. You can also view various
statistics, such as average, minimum
or maximum of the measured parameters, or view a histogram of the measurements as they are gathered.
Four ‘Math’ channels are available,
including operations such as summing or differencing two channels.
Single-channel operations such as logarithm, exponent, derivative, integral
and square root are available. FFT (fast
Fourier transform) or spectrum analysis is also possible.
Functions
The menu shown in Screen 1 gives
an idea of the DHO924S’s built-in functions. When measurement or maths
windows are open, they can be dragged
around and reorganised (something
we’re not used to on a ‘scope).
A Reference waveform can be captured (“Ref”) from an active channel to be visually compared with the
siliconchip.com.au
changing input. A more rigorous signal check can be performed using the
Pass/Fail function. This requires a
mask against which the active channel is compared; the results of the
Pass/Fail can optionally be fed to the
auxiliary BNC connector on the rear
of the scope.
The mask can be loaded from a file
or can be easily created by applying
margins (in time and voltage) against
a sampled input channel. A typical
example of such a mask is the so-called
eye diagram used to verify high-speed
signals, such as USB or HDMI. Screen
6 shows the Reference and Pass/Fail
functions.
The protocol decoding feature supports Parallel, UART, I2C, SPI, LIN
and CAN signals, although you would
probably need the optional active logic
probe to do much with a parallel bus.
There are many parameters available
to adjust, including polarity, parity
and bit order, although the defaults
look to be sensible for commonly used
configurations.
There is a button to swap SDA and
SCL in I2C mode, so it’s reasonable to
just hook the ‘scope up without worrying too much about which wire is
which. It can decode up to four separate buses at once, which is more than
sufficient for most applications.
The Bode plot function is only available on models with an arbitrary function generator, since the generator is
used to provide the input waveform.
The Bode plot window expands to
take up most of the display, as seen
in Screen 7.
The Auto function is the same
automatic configuration utility that
is found on other ‘scopes to quickly
set up the timebase, voltage scale and
trigger selection based on the active
signal. It can also be accessed from
the hardware button in the top right
corner of the ‘scope.
The Display settings can select trace
persistence and change other parameters as trace and grid intensity, as
well as window transparency. On the
bottom row are functions related to
the operating system functions of the
unit. The Help feature is actually an
on-screen viewable PDF version of the
manual, which can be seen in Screen 8.
Silicon Chip
PDFs on USB
¯ A treasure trove of
Silicon Chip magazines on a
32GB custom-made USB.
¯ Each USB is filled with
a set of issues as PDFs –
fully searchable and with
a separate index – you just
need a PDF viewer.
¯ Ordering the USB also
provides you with download
access for the relevant
PDFs, once your order has
been processed
¯ 10% off your order (not
including postage cost) if
you are currently subscribed
to the magazine.
¯ Receive an extra
discount If you already
own digital copies of the
magazine (in the block you
are ordering).
EACH BLOCK OF ISSUES COSTS $100
NOVEMBER 1987 – DECEMBER 1994
JANUARY 1995 – DECEMBER 1999
JANUARY 2000 – DECEMBER 2004
JANUARY 2005 – DECEMBER 2009
JANUARY 2010 – DECEMBER 2014
JANUARY 2015 – DECEMBER 2019
OUR NEWEST BLOCK COSTS $150
JANUARY 2020 – DECEMBER 2024
Other connections
OR PAY $650 FOR THEM ALL (+ POST)
The front USB-A socket can be used
for either a mouse or flash drive (a
USB hub allows both to be connected
WWW.SILICONCHIP.COM.
AU/SHOP/DIGITAL_PDFS
Australia's electronics magazine
August 2025 57
simultaneously). The mouse is used
as you might expect, to interact with
items on the screen. We think that support for a keyboard or numeric keypad
might be handy for parameter entry,
since this can sometimes be awkward
to do with an on-screen keyboard.
A USB flash drive can be used to
transfer screen captures or waveforms
for the arbitrary function generator. It
can also be used to upgrade the firmware. The ‘scope has internal storage
and can connect to an SMB (network)
file server via Ethernet.
We often use screen captures or
‘scope grabs for articles in the magazine, so we thought we would use a
USB flash drive to transfer the necessary files for this article. But it turns
out that the web control and network
interface mean that is unnecessary,
since we can download captures
directly to a PC.
The Web Control’s Print Screen
tab has buttons to take a screenshot
or record a video; the image or video
is displayed in the browser window
and can be simply downloaded onto
the PC from there. We noted in our
review of the MSO5354 that its web
control response was a bit slow; in
comparison, the DHO924S feels much
snappier.
The USB-B socket on the rear of
the ‘scope can be used to control the
DHO924S, but we found that the Ethernet connection was more useful,
since the PC does not need to be near
the ‘scope as is required for a USB
connection.
Other notes
There is a sleep mode which can
be used instead of a full shut-down.
This has the advantage that the ‘scope
only takes about 20s to be ready from
sleep, and also retains its last operating state. The main downside is that
the power supply must remain active
to retain this state.
Using our USB-C Power Monitor,
we recorded a peak of 2.6A at 15V,
consistent with the 3A maximum recommended in the user guide. During
sleep, we recorded a draw of around
230mA at 15V, while a full shutdown
reduces this to about 1mA.
The web control view would occasionally reset, changing the window size and stealing the focus from
another window if it did not already
have focus. If the network connection
is lost, a message is displayed in Chinese. The same message is shown if a
second browser window attempts to
connect to the web control.
Software
SCPI (Standard Commands for Programmable Instruments) is a standard
for the control of test equipment and
instruments. The DHO924S presents
an SCPI interface on Ethernet port
5555, and it can also be accessed via
USB.
In the April 2023 issue, we wrote
about the free TestController software (siliconchip.au/Article/15740),
which can interface with SCPI-capable
devices. Rigol provides a Programming
Guide which outlines the SCPI commands specific to the DHO900 series
if you wish to control the DHO924S
this way.
Rigol also provides software that
can connect to its ‘scopes and other
hardware. The Ultra Sigma software
can connect to the DHO924S via Ethernet or USB, and has an SCPI Panel
Control, which can send commands
and read data using the SCPI interface.
Screen 9 shows the main window for
Ultra Sigma with both the DHO924S
and DM858 connected.
Ultra Sigma appears to be only available for Windows, although the web
control should work on a browser
under most operating systems. We
have heard, but it has not been confirmed, that Rigol will release updated
PC software later this year.
Overall impressions
With the numerous menu options,
it was easy enough to find a specific
function and everything feels intuitive. The ‘scope feels like it has all the
features we might need and probably
a few we don’t realise we need yet.
All the controls of the DHO924S feel
quite snappy and responsive, whether
using the knobs and buttons, the touch
panel or the web control interface,
although the waveforms will freeze
while dragging.
We didn’t often touch the wrong
item on the touch panel, since most
objects are quite large, but it happened
at times, and felt slightly fiddly. Using
a mouse was much more precise, so
that is a good option. We mostly used
the web controls for much the same
reason.
The noise level, bandwidth and
number of channels makes this oscilloscope suitable for a wide range of
jobs, for which an expensive high-end
‘scope would have been required in
the not-too-distant past.
Conclusion
If we were looking for a ‘scope right
now, the DHO924S would definitely
be on the shortlist. The web control
and Ethernet interfaces make it very
easy to capture screen grabs and other
waveforms and analyses. It also makes
it easy to control the many functions
of the ‘scope.
The unit is compact and light. It’s
responsive and intuitive to use, and
most of the specifications easily exceed
the ‘scopes that we currently use.
Screen 9: the Ultra Sigma program can interface with many Rigol instruments.
Here, both the DHO924S ‘scope and DM858 benchtop multimeter are connected.
Ultra Sigma includes an SCPI control panel.
58
Silicon Chip
Australia's electronics magazine
The DHO924S is available from
Emona Instruments for
$1448 + GST: https://emona.
com.au/products/electronic-testSC
measure/dho-924s.html
siliconchip.com.au
DM858 Digital Multimeter Review
We haven’t had much need for a benchtop multimeter,
with the handheld variants being sufficient for most
purposes. A benchtop multimeter falls between a
handheld multimeter and an oscilloscope, including
a 5.5 digit display capability and features like those
you might normally see on an oscilloscope.
The DM858 comes with a power supply and a
pair of CAT II multimeter leads, as well as a pair of
alligator clip adaptors that screw onto the ends of
the probes. Like the DHO924S, the power supply is
a 65W USB-C PD (power delivery) PSU. The DM858
only requires 12V at up to 10W, with our USB-C Power
Monitor indicating a typical draw of just 7W.
The back panel connectors are almost the same
as the DHO900 series ‘scopes, with only the HDMI
Using the DM858 is easy for anyone who has used a multimeter. The extra
socket missing. It has a USB-B socket for connection
banana sockets allow four-wire (Kelvin) resistance measurements.
to a computer, and an RJ45 jack for Ethernet. One
BNC socket is for an external trigger input, while the other can output a pulse after each measurement. Since it is much the same case,
the same 100mm VESA mounts are present.
Apart from the HDMI output, most of the user interfaces are the same as the DHO924S; the Ethernet and USB connections on the back
of the unit can be used for remote access. Web Control for the DM858 works similarly, as does access to the SCPI interface over Ethernet.
The front USB-A socket supports a USB flash drive or a mouse, and it has five banana sockets on the front panel, along with the fuse
for high-current measurements.
Three of these sockets are used as you might expect, with one common input used in conjunction with another input
for high-range current measurements. The third input is used for all other measurements, such as voltage, resistance,
capacitance and so forth.
The other two connections can be used for Kelvin (four-wire) resistance measurements. The Kelvin technique is
often used to measure low resistance values, since it eliminates contact and lead resistance that might interfere with
the measured resistance.
In use
The DM858 takes about a minute to boot up and it then shows a DC voltage reading. The default, slow update rate is
easy to follow. There are two faster update rates available.
There are buttons for typical multimeter measurements, such as voltage, current (DC and AC), resistance, continuity,
diode, capacitance and frequency; standard operations are fairly obvious. Some features, including four-wire resistance
measurement and diode mode, are accessed by pressing the Shift key.
The overall interface is very similar in feel to the DHO924S, with a 7-inch touch panel offering menu items above and
below the main display. An ‘R’ menu in the bottom-left corner offers a range of functions that duplicate some of the
controls, besides allowing access to the system controls and settings.
Other features
Despite being labelled a multimeter, the DM858 can display a
slow-moving trace, but it’s more like a chart recorder than an
oscilloscope. The fastest update rate is around 10Hz. There are
simple ‘Math’ functions, such as applying an offset to a reading,
as well as statistical results such as minimum, maximum, average
and standard deviation. Voltages can be converted to dB values.
The DM858 can also interface to sensors such as thermocouples and thermistors, with several inbuilt probe types and
presets being provided. An adjustment for cold junction temperature can be added.
Other custom sensors can be monitored by supplying a list
of measured value (resistance, voltage, current etc) and display
value (such as temperature) pairs. This makes it possible to set
the meter up as a custom display to suit just about any type of sensor.
Summary
We found the DM858 easy to use. It offers numerous handy features above those of most multimeters. We would
make good use of the Web Control interface to allow remote viewing and operation. The DM858 digital multimeter
can be purchased from Emona Instruments at: https://emona.com.au/dm858/dm-858.html
The DM858 is small enough that it can be mounted using a VESA mount. It also has an Ethernet connection on
the rear, as shown in the photo above.
siliconchip.com.au
Australia's electronics magazine
August 2025 59
Stops
ises if ‘scar
ed’
e
ey
Fl
mo
c
i
Mthe e
s
u
Mo
es
at
ul
g
in
h
as
g no
Em
J
’s
oh n Cla rke
kin
ma
use
s
sound
t
o
en
H
e
r
d
ar
fi nd if hidd
Draw
s little p
ow
Mice make good pets but you still have to take care of them. This
little critter doesn’t eat much (just the occasional lithium cell) and won’t
make a mess or escape from its cage!
B
uilt on a mouse-shaped printed
circuit board (PCB) and using
relatively few parts, Mic the Mouse
is ideal for fun and can be used to
play pranks on family and friends.
Mic the Mouse only produces squeaking sounds when everything is quiet.
Make a noise, and he goes quiet. This
makes him difficult to locate. He will
start squeaking again, but only after a
period of silence.
Mic the Mouse (Mic for short) is
best described as mousy coloured and
mousy shaped. To further add to the
realism, there is provision for whiskers. He sits vertically, with a slight
lean backwards, and is supported at
the rear using a stand that attaches to
the back.
Elephants and mice have been
known in folklore and animations
to have a unique relationship (see
siliconchip.au/link/ac63), which is
why the rear stand is reminiscent of
an elephant’s rear end. Or perhaps it’s
just because that’s a suitable shape to
hold the mouse up. Nobody knows
what happened to the front half of the
elephant – except maybe Mic!
It is also well-known that mice have
a unique relationship with humans.
The poem by Robert Burns, entitled
“To a mouse”, begins: “The best laid
plans of mice and men, often go awry
and leave us nought but grief and pain
for promised joy”. This couldn’t be
more true with Mic the Mouse.
Hunt him down you try, but
Mic is elusive. He won’t reveal his
60
Silicon Chip
whereabouts if you make any noise.
But wait quietly and he will begin
his merry squeaks again. Hide Mic in
a cupboard, on a shelf, or simply in
plain sight, and have others become
horrifically aware that there is a mouse
in the house. But where? You may be
confronted by Mic’s eye flashing in a
terrifying manner.
Previous designs
Way back in August 1990, we published an electronic cricket called
Horace (siliconchip.au/Article/6925).
Horace was similar to Mic the Mouse,
except it chirped cricket sounds rather
than producing mouse sounds. It only
chirped when there was quiet, and was
quiet himself when there was ambient noise.
It utilised an electret microphone,
quad op amp IC and a piezo transducer. This was powered by a 9V battery and its current draw was 3mA.
With a 600mAh capacity, Horace could
run for about 200 hours or about 8
days continuously. The reason for the
“Horace” name has been lost in time.
We published various updated
cricket designs in December 1994, July
2011, June 2012 and October 2017, culminating in Silicon Chirp (April 2023).
The reason for Mic’s name is a bit
more straightforward. Firstly, it keeps
up the tradition of alliterative names
for animal characters like Dorothy the
Dinosaur, Peppa Pig and so on. The
other reason is that Mic uses a microphone to listen for sounds.
Australia's electronics magazine
To make Mic the Mouse a similar size to a real mouse, we need to
use a few tricks. A 3V lithium cell
is much smaller than a massive (by
comparison) 9V battery. However, a
3V cell does not have as much capacity (200mAh, 600mWh) compared to
a 9V battery (600mAh, 5.4Wh), so the
current consumption needs to be kept
as low as possible.
To get a reasonable cell life, we need
to reduce the current consumption
from the 3mA of Horace the cricket
down to at least 1mA to get the same
cell life. However, we can do a lot better than that; we reduced the current
draw to an overall average of 105μA.
That’s a reduction of around 28 times
compared to Horace!
This was achieved by using a lowpower microphone and a microcontroller to mini mice (sorry, minimise)
the current draw by only powering the
microphone when required. Also, we
only flash the LED eye momentarily to
mini mice the power used.
Like a real mouse, it sleeps to reduce
its current consumption to an absolute
minimum; it is woken up periodically
by a timer. It’s a bit like hibernation, a
trick Mic has stolen from bears.
The low current microphone we use
is a MEMS (micro-electromechanical
systems) type, as used in phones. It is
supplied in a tiny package that measures just 2.75 × 1.85 × 0.95mm and
requires reflow soldering as the contacts are underneath the package. That
makes it difficult to hand-solder.
siliconchip.com.au
Coin cell warning!
This project contains a small lithium ‘coin’ cell that represents a serious health
risk should the cell be swallowed by a child. Young children are most at risk.
Read the information sheet at www.schn.health.nsw.gov.au/factsheets on
the dangers of small cells.
Ensure that the cell is kept secured using the cell capture screw and nylon
spacer and that it is tightened fully to prevent undoing by hand. Keep this project away from small children. Also, keep unused cells in a secure place away
from children, such as in a locked medicine cupboard.
New cells should be kept within the original secure packaging that requires
scissors to open until required for use.
If you have any older button/coin-cell powered devices that provide easy
access to the cells, store them in a safe place when not in use. Alternatively,
devise a method to make the cell access more difficult, such as gluing the cell
compartment shut so that a child can’t open it.
Fortunately, the MEMS microphone
is available pre-soldered on an inexpensive module, which also includes
an amplifier and 3V regulator.
Circuit details
Mic’s complete circuit is shown in
Fig.1. It’s based around microcontroller IC1, a PIC16F15214-I/SN, powered
by a 3V lithium cell (BAT1). Power is
applied via a slider switch. Mic does
not draw much current, typically only
about 0.36μA while asleep. This rises
to around 1mA when monitoring for
ambient sounds and 1.6mA while
making noise.
Diode D1 is included as a safety
measure to prevent damage to IC1
should the cell be inserted incorrectly.
The cell holder doesn’t stop you from
inserting the cell with the incorrect orientation (it should be positive side up).
With the positive side down, the
cell will be shorted out by contact
with the sides and top spring contacts.
However, during insertion, there could
be a brief period when there is no
contact with the cell holder sides, so
the circuit could be supplied with a
reversed polarity voltage that could
damage IC1. In that case, D1 clamps
this voltage to a low level.
The cell will lose some of its capacity if left connected in reverse for more
than a few seconds, but that’s better
than damaging the chip.
IC1 is clocked by an internal 4MHz
oscillator. Its power supply pin is
bypassed with a 1μF capacitor. IC1’s
job is to supply power to and monitor
the MEMS microphone module (MIC1)
output, drive the piezo transducer to
make mouse sounds, flash the LED
used for Mic’s eye and also check if
S2 is pressed.
The MEMS microphone module
(MIC1) is powered via IC1’s RA4 digital output, which goes high (near to
the cell voltage) when required. When
powered, the A output on the module
(pin 3) provides an amplified signal
from the MEMS microphone.
The circuit of the microphone module that includes the MEMS microphone is shown within the dashed
box in Fig.1. Its onboard regulator,
U1, supplies 3V to the MEMS microphone itself (U3) and provides a bias
voltage to pin 1 of the op amp, U2. U1
is a low-dropout regulator, so with a
3V input, its output won’t be much
below that.
The output from the MEMS microphone (U3) is amplified by op amp U2,
which is configured with a gain of 50.
The non-inverting input is held at half
supply using the two 10kW divider
resistors across the 3V supply. So,
with a 3V supply to the MEMS module, the DC output from U2 is typically at 1.5V. A MEMS module output
signal when subject to noise is shown
in Scope 1.
For data on the MEMS microphone
and module, see siliconchip.au/link/
ac64 and siliconchip.au/link/ac65
The current consumption of the
MEMS module is typically 287μA at
3V. That’s the total of U1 (7μA typical, 15μA maximum), U2 (80μA typical, 185μA maximum) and U3 (50μA
typical, 150μA maximum). The voltage divider comprising the two 10kW
resistors in series across the 3V supply
also contributes 150μA. We measured
our module’s draw as 330μA.
The PIC16F15214 (IC1) monitors the
microphone signal using its AN5 analog input. We have AC-coupled MIC1’s
output to that pin using a 100nF capacitor and biased the voltage to 0V via
a 10kW resistor. This has the signal at
the AN5 input (pin 2) normally sitting
Fig.1: the components inside the dashed cyan box are on the MEMS microphone module. IC1 monitors its output and
determines when to flash the eye LED and create squeaking noises using the piezo sounder.
siliconchip.com.au
Australia's electronics magazine
August 2025 61
Scope 1: this oscilloscope trace shows the output from the
MEMS microphone module after AC coupling to the AN5
input (pin 2) of IC1. The signal level goes above 100mV
peak.
close to 0V and swinging above and
below that.
A diode clamp internal to pin 2 of
IC1 will limit negative excursion to
-0.3V, while the 1kW series resistor limits the current in the clamping diode.
We do this so that pin 2 sits near
0V with no signal (ie, silence). Also,
while the 10kW/10kW resistive divider
in the module theoretically causes the
signal to sit at exactly half the supply
voltage, the supply voltage can vary
because it’s coming from a microcontroller output which has a fairly significant output impedance of around
116W.
That means that supply to the
MEMS module can be anywhere
between about 33-58mV less than the
IC1 supply due to the voltage drop at
the RA4 output. The MEMS module
current draw also varies, so it is difficult to predict the MEMS module’s idle
output voltage with sufficient accuracy
to allow for threshold detection of any
small signal that is superimposed on
it. Re-biasing the signal to 0V solves
this difficulty.
Noise detection
To detect ambient noise, we convert
the voltage at the AN5 input into a 9-bit
digital value every 1ms (1000 times
per second). The digital value ranges
from 0 to 511. If this exceeds a specific threshold value, it is detected as
noise. This threshold can be adjusted
between 1 and 10 in ten steps, corresponding to an analog range of about
12-65mV (assuming a 3V supply).
62
Silicon Chip
Scope 2: the RA0 (yellow) and RA1 (cyan) output
waveforms when producing mouse sounds. The signal
bursts vary randomly in length, with variable periods of
silence in between. It isn’t obvious from this that the signal
frequencies and duty cycles also vary.
Lower threshold settings give Mic
a greater sensitivity to noise. More on
this later.
Mouse sounds
IC1’s RA0 and RA1 output pins
drive the piezo transducer to produce
the mouse noises. The piezo is driven
in bridge mode by the two outputs,
increasing the AC voltage across it to
produce a louder sound.
When the RA0 output is high, RA1
is low and vice versa. In one condition, there is +3V across the piezo
transducer, and in the other, -3V. This
results in a 6V peak-to-peak square
wave. A 100W resistor limits the peak
current into the transducer’s capacitive load as the outputs switch.
The mouse sounds comprise various frequency bursts with variable
length gaps in-between. The signal
frequency varies between bursts and
also within each burst. Scope 2 shows
three such bursts.
While not visible in Scope 2, there
is considerable detail within each
signal burst. At the beginning of each
burst, the duty cycle starts off quite
low. This means the piezo transducer
is driven only with brief pulses, resulting in a low volume. As the duty cycle
increases, the output from the piezo
transducer also increases. The duty
cycle is increased a little each cycle
until it reaches a 50%.
A similar change in the duty cycle
occurs at the end of each burst. The
duty cycle is reduced on each cycle
until it reaches zero, so that the volume
Australia's electronics magazine
falls back to zero. This gives the signal bursts soft starts and soft finishes,
preventing loud clicking sounds from
being produced by the piezo transducer at the beginning and end of
each burst.
We also use lower duty cycles to
reduce the volume level within each
burst instead of having a constant
level. A varying volume level sounds
more natural.
The greatest volume available from
the piezo transducer is when it is
driven at 50% duty, as shown in Scope
3. The push-pull drive from the RA0
and RA1 outputs is visible too. This
is necessary to provide a sufficient
sound level from a supply voltage of
just 3V or less.
The drive frequency also varies over
the burst period. If we were to just
have the same frequency throughout,
it would sound just like bursts of a
single tone, like Morse Code. By having a frequency mix, the bursts sound
considerably less electronic in nature.
Mic’s eye
LED1 is driven via the RA2 output
via a 1kW current-limiting resistor.
This LED is made to flicker when Mic
is producing sound. The LED is also
used to indicate the threshold level
used to detect ambient noise during
the setup process. It flashes between
one and ten times to indicate the chosen threshold value. The LED also
flashes briefly when Mic is powered
up to acknowledge this.
Pushbutton S2 is used to set this
siliconchip.com.au
Scope 3: an expanded view of the drive to the piezo
transducer, showing how the ~3V peak square wave signals
from RA0 & RA1 (yellow and cyan) combine to form a 6V
peak-to-peak square wave across the transducer (red trace).
The duty cycle here is near 50%.
threshold. IC1 detects that S2 is closed
by monitoring its pin 4 digital input
(RA3). When S2 is pressed (closed), the
voltage is close to 0V. When the switch
is open, an internal pullup current in
IC1 keeps the RA3 input high. The S2
switch closure is only checked during
power-up; if it’s low (closed) then, the
threshold setup process starts.
Power control
Much of the design work went into
minimising the current draw from the
small 3V cell. Shutting down the circuit is the major way to do this. When
IC1 is in sleep mode, its oscillator is off
and the power supplies to the MEMS
module and LED are also switched off.
A separate ‘watchdog’ timer starts
running in sleep mode, to wake IC1
periodically. This varies between 4,
8, 16, 32, 64, 128 and 256 seconds in
a randomised order.
To extend the sleep periods and save
more power, IC1 is sent back to sleep
immediately upon waking 30 times.
This provides an off-time between two
minutes (when there is a 4s watchdog
timeout) and about two hours for the
256s timeout. During this period, the
current consumption is very low; we
measured this at 0.36μA with a 3V
supply.
IC1 itself draws just 0.9μA in sleep
mode, including the watchdog timer
and oscillator current draws.
After these 30 sleep periods, IC1
powers up the MEMS microphone
module and checks for ambient
sound. During this period, its current
siliconchip.com.au
Scope 4: this is similar to what’s shown in Scope 3, except
the duty cycle is lower, at around 20%. This reduces the
output sound level.
consumption is about 1mA. This is
mainly due to the MEMS module consumption at about 330μA, and IC1
drawing around 660μA while running.
There needs to be a 2-16 minute
quiet period (again a randomised
value; it’s either 2m, 4m, 8m or 16m)
before it is deemed to be quiet enough
for the mouse to make noises. Should
noises be detected during the listening
period, IC1 will go back to sleep for
another randomly chosen sleep period.
If no sound was detected, Mic the
Mouse will begin to make mouse
sounds. During this time, his current
consumption is around 1.6mA. These
sounds run over a variable-length
period between 100ms and two minutes; a typical duration is around
30 seconds. If noise is detected in
between making the mouse noises,
Mic will go back
to sleep and
stop making
noises.
All the components are located on
for mounting the stand.
Australia's electronics magazine
There is a brief 5ms delay between
each mouse sound ceasing and the
beginning of monitoring ambient
noises at the AN5 input. This wait
is to prevent the MEMS microphone
from picking up sounds from the piezo
transducer.
Adding up the total current draw
taking into account the typical sleep,
checking for ambient sound and the
mouse sounds operation periods, we
estimate the overall current draw is
an average of 105μA.
It is checking for ambient noise
(drawing 1mA) around 9% of the
time, making mouse sounds (1.6mA)
around 1% of the time and in sleep
mode (0.36μA) 90% of the time.
Considering a typical 3V cell has a
capacity of 200mAh, we expect Mic
the Mouse to operate on the one cell
for around 1905 hours. That’s 79
the rear of the PCB along with the slots
August 2025 63
Mic the Mouse with his stand shown separately. Note the
use of a spacer to secure the coin cell.
days if Mic is left on continuously.
If power is switched off, the current
draw from the cell becomes close to
zero, with the only draw being cell
leakage and diode D1’s reverse leakage. These are very low and in the
nanoamps (nA) region.
If you handle the cell with your fingers across the insulating ring between
the positive and negative contact areas,
the leakage current can be higher due
to skin oils bridging the terminals.
Cleaning the cell with methylated
spirits or similar will prevent this extra
leakage from occurring.
1 double-sided plated-through white PCB coded 08105251, 96 × 53mm
1 double-sided-plated through white stand PCB coded 08105252, 48 ×
31mm
1 Fermion MEMS microphone module (MIC1) [Core Electronics SEN0487]
1 30 × 5.5mm passive piezo transducer (PB1) [HYR-3006/AT3040]
1 SIL SPDT mini vertical slider switch (S1) [SS12D00G3]
1 4-pin 6.2×6.5mm tactile switch (S2) [SKHMQME010 or similar]
1 CR2032 surface-mount folded phosphor bronze PCB mount cell holder
(BAT1) [BAT-HLD-001 or similar]
1 CR2032 3V lithium cell
1 3-pin header, 2.54mm pitch (usually supplied with the MEMS microphone)
1 260mm length of white 0.8mm diam. bamboo cord [Spotlight 80325284]
5 M3 × 10mm nylon or polycarbonate screws (cheese or countersunk head)
4 M3 nylon or polycarbonate hex nuts
2 M3 nylon, polycarbonate or metal hex nuts
1 M3 × 6.3mm tapped nylon standoff/spacer
Semiconductors
1 PIC16F15214-I/SN 8-bit micro programmed with 1810525A.HEX (IC1)
1 SMD 75V 500mA fast signal diode, such as 1N4148WS or LL4148 (D1)
1 3mm standard brightness red LED (LED1)
Capacitors (all SMD M2012/0805 or M3216/1206)
1 1μF 50V X7R
1 100nF 50V X7R
Resistors (all SMD M2012/0805 ⅛W or M3216/1206 ¼W)
1 10kW 1%
2 1kW 1%
1 100W 1%
component overlay diagram is shown
in Fig.2.
Check that the tabs on the stand fit
into the Mouse slots before assembly.
If it is difficult to fit the two together,
a small amount of filing may be necessary. The stand should be removed
while installing parts on the Mouse
PCB.
If you are going to use countersunk
screws, the front of the PCB will need
its holes countersunk so that the screw
heads fit neatly, almost flush with the
PCB face.
Begin by installing the microcontroller (IC1), which comes in an 8-pin
SOIC SMD package. You will need a
soldering iron with a fine tip, a magnifier and good lighting. First, place
the chip with its pin 1 locating dot to
the lower right and with the IC leads
aligned with the pads. Then solder a
corner lead and check that it is still
aligned correctly.
If it needs to be realigned, remelt
the soldered connection and move
the IC to align it again. When correct,
solder all the remaining pins. Any
solder that runs between the IC pins
can be removed with solder-wicking
braid (ideally with the aid of a little
flux paste).
Continue by installing the resistors.
These will have value codes printed on
them, with the last number indicating
how many zeroes follow. For the resistors used, the codes will be 101, 100R
or 1000 for 100W, 102 or 1001 for 1kW
and 103 or 1002 for 10kW.
Two resistors and one capacitor are
located beneath the MEMS microphone module, so these need to be
Complete Kit (SC7508, $37.50): includes everything except the CR2032 cell
siliconchip.com.au
Construction
The parts for Mic the Mouse fit on
a double-sided plated-through PCB
coded 08105251, measuring 96 ×
53mm, with a white solder mask and
black labelling.
The rear stand plugs into the component side of the Mouse PCB to support
it; it is also a PCB, coded 08105252,
that measures 48 × 31mm. The main
Parts List – Mic the Mouse
installed before the MEMS module is
in place. The 100nF and 1μF capacitors can be soldered in next; their orientations do not matter. These will
not be marked with values, but the
1μF capacitor is likely to be thicker
than the other.
Diode D1 can now be installed, taking care to orientate correctly. There is
sufficient tinned copper area to allow
MiniMELF/SOD-80 or SOD-323 package devices to be soldered in.
S1 is a through-hole slide switch
but you should insert its pins into
the allocated holes high off the PCB
so the leads don’t protrude through
to the other side of the PCB. You can
then solder the switch pins to the
top side of the pads, not the underside, keeping the visible side of Mic
unmarred by solder joints. The on-
position for this switch is marked on
the silkscreen.
Switch S2 is surface-mounting tactile pushbutton switch, so solder its
four corner pins to the pads.
The two mounting holes on the
MEMS module need to be drilled out to
3mm to allow the module to be raised
off the PCB using nuts as spacers, and
secured with M3 machine screws and
nuts. The MEMS microphone module is connected electrically using a
standard 3-pin 0.1-inch/2.54mm pitch
header.
Solder this header initially on the
component side of the mouse PCB,
with the lead ends flush with the non
component side, like with the slide
switch. After that, slide the black plastic spacer off the pins.
Before soldering the MEMS microphone module, attach it to the mouse
PCB using M3 nylon or polycarbonate
screws and nuts, with the screw heads
on the non-component side and one
nut securing the screw to the PCB on
the component side. The MEMS module is then placed on the screws and
two more nuts added to hold the module in place.
Do not use metal nuts as they could
cause short circuits. With the module
attached with the screws, you can then
solder the three pins to the pads on the
MEMS microphone module.
The cell holder is mounted with the
cell entry side towards the mouse's
ears. That allows the cell capture screw
to keep the cell in place, preventing
small children from removing it. This
complies with Australian Standard
AS/NZS ISO 8124.1:2002.
While Mic the Mouse is not really
a project for very young children, it
could be used in a household where
young children live or visit, who could
potentially swallow button cells if they
find one and manage to remove it.
For our project, the cell is held
within a compartment, with the exit
blocked by a 10mm M3 screw that is
inserted from the non-component side
of the PCB and secured on the cell
holder side with a 6.3mm-long nylon
tapped standoff. When tightened with
a screwdriver, the standoff cannot easily be removed by hand. An alternative
to the standoff is two M3 nuts, with
the top one used as a lock nut.
The cell holder is a half-shell type;
its metal contacts the positive side of
the cell. A tinned copper area on the
PCB completes the cell holder, providing the negative connection to the cell.
LED1 is a leaded component,
with its leads bent so that they are
U-shaped, returning past the LED
body. The LED’s lens is positioned
over the mouse’s eye hole; it does
not protrude through the hole fully.
Solder the leads from the component
side and make sure the (longer) anode
lead is soldered to the pad on the PCB
marked “A”.
The wires for the piezo can be soldered to the PCB (the positions are
marked ‘piezo’). The wires will need
to be cut shorter than supplied. The
wires will probably be red and black,
but it does not matter which colour
wire goes to which PCB pad. Typically, including in this case, the
transducer is not used as a polarised
component.
You will need to drill the mounting holes on the piezo unit out to
3mm to suit the M3 screws. The piezo
transducer is then secured with two
10mm-long M3 screws and two nylon,
polycarbonate or metal nuts.
Now insert the CR2032 cell into
its holder, secure it with the screw
and M3 tapped standoff and switch
on the power with switch S1. If all is
well, the eye LED will momentarily
flash to acknowledge power has been
connected. The eye also very briefly
flashes at the end of each sleep cycle.
Programming IC1
That test assumes IC1 has already
been programmed, which it will be if
you buy it from us, either by itself or
as part of a kit. If you intend to program the PIC yourself, the firmware
(1810525A.HEX) can be downloaded
from siliconchip.au/Shop/6/2698
If the chip has already been soldered to the board, but is unprogrammed, you will need to wire up
a programming adaptor to the PCB,
such as a PICkit. Since there is no
in-circuit serial programming (ICSP)
header, you will need to make the
Fig.2: there are about 14 different components mounted on the PCB; don’t miss the three that are under the MEMS
microphone module. The five pads numbered 1-5 in red are the points you can solder wires to for in-circuit programming
of IC1. They correspond to pins 1-5 of a PICkit programmer or similar.
siliconchip.com.au
Australia's electronics magazine
August 2025 65
five connections separately. They go
to pads marked 1-5 on the PCB and
in Fig.2; these correspond to the pins
on the PICkit programming header (1
= MCLR etc).
Sensitivity to sound
As mentioned previously, sensitivity to ambient sound can be adjusted
so that you can select the sound level
that Mic reacts to, over a range of 1-10.
Lower values provide higher sensitivity to sound, ie, Mic will detect lower
noise levels. Higher values mean less
sensitivity, so more noise is required
to silence Mic.
To adjust sensitivity, switch it on
using S1 while holding down S2. This
initiates the adjustment mode, where
Mic’s eye blinks out the sensitivity setting. There is one blink for each sensitivity level. You can test each sensitivity level after the flashes have finished;
you have up to 16 seconds to test each
level. The eye will flash in response to
your making noises.
If the eye continuously flashes due
to the detection of background noises,
the setting is too sensitive, and a higher
value should be selected.
To change to the next sensitivity
level, press S2 before the 16 seconds
are up. This will cause the eye to flash
out the next sensitivity level. You can
then test this sensitivity level for up
to 16 seconds. Once the sensitivity
value has reached 10, the next value
will be 1 again.
The selected sensitivity is stored in
flash memory, and will be remembered
if the power is switched off.
If you wait out the 16 seconds after
releasing S2, Mic will begin to make
squeaking sounds. This is a quick way
to have Mic make some sounds for
testing. While making these sounds,
Mic also checks whether there is ambient sound. If detected, any mouse
sounds will cease, causing him to go
to sleep.
On a normal power-up without S2
pressed, mouse sounds will begin after
about four minutes from switch-on.
This period will also depend on
whether there is ambient noise present
that would prevent Mic from sounding. Further mouse sounds could occur
up to two hours later.
Adding some whiskers
Versatile
The whiskers are made using white
0.8mm bamboo cord. The whiskers can
be up to about 30mm in length, so cut
each length of cord to about 65mm,
allowing two whiskers to be formed by
folding the length in half. Then insert
each end into two adjacent holes in the
whisker region, from the component
side of the mouse.
Coat the cords with a thin smear
of PVA glue so that they will become
stiff when dry. You will need to orientate the whiskers by having the mouse
body supported on a stand so that the
PCB sits horizontally, with the whiskers hanging downward until the glue
is dry.
Finally, the rear stand can be
attached at the component side with its
two protrusions placed into the slots
on mouse PCB. The piezo wire leads
will add some holding force to keep
the stand in place.
Modifications
If you want to reduce the volume of
the mouse squeaks, increase the resistance of the resistor in series with the
piezo transducer. Increasing it from
100W to 1kW will reduce the apparent volume by about 50%. Higher values will provide an even lower volSC
ume level.
Battery
Checker
This tool lets you check the condition of most
common batteries, such as Li-ion, LiPo, SLA, 9V batteries, AA, AAA,
C & D cells; the list goes on. It’s simple to use – just connect the battery to the
terminals and its details will be displayed on the OLED readout.
Versatile Battery Checker Complete Kit (SC7465, $65+post)
Includes all parts and the case required to build the Versatile Battery Checker, except the optional
programming header, batteries and glue
See the article in the May 2025 issue for more details: siliconchip.au/Article/18121
66
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
The Boeing 737 MAX & MCAS
A predictable disaster by Brandon Speedie
Image source: Aka the Beav, www.flickr.com/photos/87117889<at>N04/23514088802 (CC-BY-2.0)
Boeing’s MAX version of their venerable 737 aircraft has had its share of problems, from
two deadly crashes in 2018 and 2019 to the latest drama with the door plug falling off in
flight. This article explains how the failure of a single electronic part led to two fatal crashes.
A
merican aircraft manufacturer Boeing
launched the 737 MAX in 2017 to much
fanfare. It is the first narrow-body
aircraft to be made predominantly of
composite materials, which are lighter
than the magnesium and aluminium
alloys of its predecessor: the 737 NG
(Next Generation).
The MAX also features new high-
bypass turbofan engines from CFM
International, as well as reprofiled
winglets to reduce drag. All of these
improvements aim to increase fuel
efficiency, one of the main operating
expenses of a passenger flight.
Unfortunately, the high bypass turbofan engines are also bulky; bulk that
wasn’t compensated for with the rest
of the aircraft design. Under load, the
aircraft was inherently unstable and
ultimately unsafe.
Lion Air Flight 610
On 29th October 2018, a recently
built 737 MAX took off from Jakarta.
Shortly after taking to the air, its pilots
were bombarded with warnings on
the flight deck, and the aircraft began
siliconchip.com.au
to pitch downward into a dive. The
pilots wrestled with the controls to
try to maintain altitude, which would
briefly arrest their descent, only for
the nose of the aircraft to pitch down
again moments later.
This wrestle between the pilots and
the aircraft continued for 13 minutes
after take-off, before the plane crashed
into the water off the coast of Jakarta,
killing all 189 passengers and crew.
Upon recovering the ‘black box’
flight recorder, investigators found
an automated system was overriding
pilot input, despite the autopilot being
disengaged.
Ethiopian Airlines Flight 302
On the 10th of March 2019, another
recently-
b uilt 737 MAX departed
Addis Ababa in on route to Nairobi.
Similarly to the Lion Air crash, the
pilots were immediately bombarded
with warnings after takeoff. The nose
of the aircraft again pitched down,
despite the pilots straining to pull back
the control yoke. The aircraft crashed
into a field six minutes after take-off,
Australia's electronics magazine
killing all 157 passengers and crew.
The black box showed an almost
identical scenario to the Lion Air
flight: repeated nose up commands
from the pilots, which would then be
overruled by an automated system that
placed the aircraft into a dive.
The Airbus A320neo
To understand how this automated
system was permitted to overrule
human input, we need to look across
the Atlantic to Boeing’s main competitor, the European Union’s Airbus.
Eight years prior, Airbus announced
a new aircraft that would be in direct
competition with Boeing’s 737. This
new variant is called the neo (New
Engine Option) which was groundbreaking for its ability to accept two
different engines: the CFM LEAP 1-A
or the Pratt and Whitney GTF. Airlines
loved the choice, as it gave them the
flexibility to select the highest fuel
efficiency engine for a given configuration.
The A320neo sold faster than any
aircraft ever before. Boeing knew their
August 2025 67
Fig.1: typical drive (red)
and receive (cyan &
green) waveforms for a
resolver. Original Source:
AD2S1210 data sheet
Fig.2: the configuration of a resolver. A sinusoidal excitation
applied between R1 and R2 inductively couples a current
into the rotor. The resulting magnetic field induces voltages
in orthogonal receive coils S1-S3 & S2-S4, which will vary
in response to the rotor position. Original Source: Analog
Devices – siliconchip.au/link/abwg
existing 737 NG could only be fitted
with the older CFM56, which was
much hungrier on fuel. They didn’t
have a product that could compete.
Enter the 737 MAX
Boeing executives scrambled to
come up with a solution. Many engineers considered the 737 to be in need
of a replacement; its original design
was over 40 years old. There were
existing plans to replace the 737 with
a brand new plane.
However, Boeing couldn’t afford
the lengthy time to market for a new
design, so they instead decided to
update the 737 NG so that it could
accept a new high efficiency engine
from CFM International. The CFM
LEAP 1-B is a high-bypass turbofan,
meaning that most of the air that flows
through the engine bypasses the turbine and is ejected without being used
in combustion.
This configuration is highly efficient, but requires a significantly larger
diameter than the CFM56 that was
used by the 737 NG.
To fit the larger engine to the 737,
Boeing engineered a compromise.
Ideally, the engines should be mounted
centrally to give the most stable flight
characteristics. But even with longer
landing gear, the new LEAP was too
large to fit under the wings. Boeing
had to mount the new engine higher
and further forward than was optimal.
This caused the aircraft to tend to
nose-up under thrust, giving the 737
MAX significantly differing flight characteristics to its predecessor.
This was enough of a departure from
the previous design that 737 pilots
would need training in the new handling characteristics. Boeing knew
that airlines would prefer to avoid
additional flight training. Removing
pilots from the air to spend days in a
simulator is costly and disruptive. To
avoid this requirement, they instead
decide to write some software to compensate. Unfortunately, this code was
reliant on a single point of failure: the
AoA sensor.
Angle of attack (AoA) sensor
Protruding externally from the side
of the aircraft’s nose is a small fin (see
Photos 1 & 2). This winglet rotates
with the direction of airflow during
Photo 2: an
angle of attack
sensor on the 737
MAX (below the
antennas near
the nose). Like
most other jets,
the 737 MAX has
a second sensor
on the other side
of the nose for
redundancy.
Source: Business
Insider –
siliconchip.au/
link/abwh
Photo 1: An angle of attack sensor
showing a winglet that aligns with
the airflow direction. Source: https://
bluemarble.ch/wordpress/tag/aoavane/
68
Silicon Chip
flight, thereby giving an indication
of the relative angle of the wing with
respect to oncoming air. This is known
as the wing’s angle of attack (AoA), an
important indication for the pilot to
ensure they don’t exceed the aircraft’s
performance envelope.
When flying level at cruise altitude, the plane should have a shallow
angle of attack, meaning low lift and
low drag from the wings. At take-off,
the wings will have a higher angle of
attack as the aircraft pitches into a
climb, providing more lift but with
greater drag.
Should the pilot attempt to climb
too aggressively, the angle of attack
could exceed a critical threshold, at
which point the wing will begin to
experience flow separation. The resulting turbulence results in a sudden loss
of lift, a dangerous situation known
as a stall.
Given the angle of attack sensor is
located in a vulnerable position on
the side of the aircraft’s nose, it is
commonly damaged by bird strikes or
debris. It is also vulnerable to freezing
up in icy conditions (there is a heater
to prevent that but it can fail or be
Australia's electronics magazine
siliconchip.com.au
overwhelmed). Therefore, many passenger planes have an AoA sensor on
each side of the nose to provide redundancy in case of damage or a fault in
one of them.
The resolver
The AoA winglet is attached to an
angular position sensor known as a
resolver. This sensor is similar to a
rotary encoder, except it is analog,
in contrast to the digital quadrature
output of the encoder. Resolvers are
favoured in high-reliability applications due to their rugged build quality.
Its theory of operation compares to
an induction motor – see Fig.2. An
excitation signal is applied to the signal coil, typically in the order of 10kHz
and 10V. This excitation induces a current in the rotor, which in turn induces
a signal in the two receive coils. These
receive coils are perpendicular, so they
are 90° out of phase of each other, as
shown in Fig.1.
Given a sinusoidal excitation,
the received signals will be complimentary sine and cosine pairs. If the
rotor’s angular position changes, the
coupling between the excitation signal and the two received signals will
change, ie, their mutual inductance
varies. This property can be used to
sense the angular position of the rotor,
using the scheme shown in Fig.3.
Effectively, this is a phased-locked
loop (PLL) that includes the resolver
itself, facilitating an angular accuracy
better than 0.01°.
The Boeing 737 MAX that was involved in the Lion Air flight 610 crash. Source:
PK-REN – www.flickr.com/photos/pkaren/45953419622/ (CC-BY-SA-2.0)
An Airbus A320neo aircraft. Source: BriYYZ – www.flickr.com/photos/
bribri/28915135713/ (CC-BY-SA-2.0)
Circuit Analysis
Fig.4 shows an example resolver
sense circuit based on the Analog
Devices AD2S1210 “resolver to digital converter”.
An advantage of this circuit is it
combines both the excitation and
sensing circuitry into a single IC. This
allows the sensed signals to be used
as feedback to adjust the phase of the
excitation signal and therefore null out
any angular position errors.
The excitation signal is derived from
the nominal 8.192MHz crystal clock,
which is internally divided down to a
range between 2kHz and 20kHz, as set
by an internal configuration register.
The synthesised waveform is sent to
the digital-to-analog converter (DAC),
which drives complementary outputs
EXC and EXC at around 3.6V peak-topeak, giving a total voltage swing of
7.2V peak-to-peak.
Fig.3: a block diagram of a resolver sense circuit. A DAC synthesises a
sinusoidal waveform from the reference oscillator, which excites the drive
coil, ultimately inducing a flux in the rotor. A “type II tracking loop” is used
to cancel errors in the sensed angular position, which allows the AD2S1210
IC to achieve excellent accuracy. Original Source: Analog Devices –
siliconchip.au/link/abwg
siliconchip.com.au
Australia's electronics magazine
August 2025 69
Fig.4: a simplified
circuit diagram of
the AD2S1210-based
resolver sense
circuit. The reference
oscillator is derived
from the 8.192MHz
crystal. The EXC
outputs have weak drive strength and need to
be amplified by op amps and complementary emitterfollower transistors pairs to match the low input impedance of the resolver
sense circuit. Original Source: AD2S1210 data sheet
The output DAC has weak drive
strength (100μA), which is a poor
match for the low input impedance of
the resolver excitation coil, typically
around 100W. Two external pushpull current amplifiers are needed.
These amplify the complimentary
EXC outputs to drive most resolvers
with ease.
The EXC voltage is applied to the
inverting input of the op amp via a
10kW input resistor. The non-inverting input is supplied with +3.75V,
derived from a 22kW || 10kW voltage
divider tapping off the 12V rail. This
provides a DC offset to avoid the need
for a separate negative supply rail. The
output of the opamp drives a push-pull
output made up of complementary
BC846B and BC856B pairs.
Biasing for this pair is provided by
the 2.2kW and 3.3W resistors, in combination with diodes D1 and D2. The
voltage gain is set by the ratio of the
10kW input resistor and 15.4kW feedback resistor. A 120pF parallel capacitor provides some high-frequency filtering to improve stability.
Additional filtering is provided
by the supply bypassing capacitors,
parallel 4.7μF and 10nF types. The
5V supply and ground are separated
for the digital, analog and reference
supplies, further improving noise
immunity.
The two sense coils are connected
to the SIN, SINLO, COS, and COSLO
inputs on the AD2S1210 via input
protection circuitry. Series resistance
and zener diodes provide circuit protection, while the anti-aliasing capacitors low-pass filter the sensed voltage to make it suitable for driving the
downstream receive circuit.
Optional voltage dividers formed
using added resistors Ra and/or Rb
can be used to attenuate the voltage
if its amplitude is too great to suit the
differential ADC on the AD2S1210.
As the resolver output is analog, its
angular resolution is only limited
by the resolution of this ADC. In the
AD2S1210, up to 16 bits are provided,
which gives an impressive 0.005° resolution.
Once digitised, the sine and cosine
inputs are compared to the excitation
signal using a so-called Type II tracking loop.
This feedback loop constantly
adjusts the excitation phase to minimise the angular position error. The
calculated position is made available
Australia's electronics magazine
siliconchip.com.au
70
Silicon Chip
to the flight computer over a digital
interface, which can be a 4-wire serial
or 16-bit parallel interface.
For more on how this circuit works,
see siliconchip.au/link/abwf
On the 737 MAX, the flight computer erroneously received the wrong
angle of attack from the resolver, ultimately causing two plane crashes
(another was narrowly avoided by an
alert copilot).
Air Crash Investigation
In the aftermath of the Lion Air
crash, investigators discovered an
irregularity with the resolver attached
to the left side AoA sensor.
In the weeks prior, the angular position readings had shown intermittent
errors. Detailed analysis revealed a
crack in the resolver, which presented
as an open circuit when the aircraft
was out of service and the resolver
cooled below 60°C.
This wasn’t detected by maintenance staff while the aircraft was in
service due to the action of the AoA
heater, which caused the resolver to
expand and close the circuit, restoring normal operation.
The Ethiopian airlines investigation
revealed a similar problem with the
left side AoA sensor, likely caused by
a bird strike 44 seconds after lift-off.
Wind tunnel tests revealed an impact
with a bird weighing 226 grams at
170 knots was enough to snap off the
AoA winglet, and leave the resolver
misoriented.
In both crashes, bad readings from
the left side AoA resolver caused some
automated software to activate: the
Manoeuvring Characteristics Augmentation System.
The MCAS (Manoeuvring
Characteristics Augmentation System)
Modern passenger airliners are ‘flyby-wire’ systems, meaning that the
pilot’s control yoke is not directly
connected to the control surfaces on
the jet by wires or hydraulics like in
older aircraft. Pilot inputs (like pressure on or movement of a control stick
or yoke) are read by sensors and fed to
the flight computer.
Software ingests these readings,
along with other sensors on the aircraft
such as airspeed, air density, temperature, and so on. It then commands the
appropriate movements of the control
surfaces to affect the aircraft’s attitude,
matching the pilot’s commands.
siliconchip.com.au
Fig.5: a vertical airspeed comparison of Lion Air flight 610 and Ethiopian
Airlines Flight 302. You can see how the pilots were fighting with MCAS to try
to gain altitude. Original Source: https://w.wiki/AGgf
MCAS is an addition to the normal
flight software on the 737 MAX to compensate for the suboptimal positioning
of the engines.
As mentioned earlier, the compromises to the design forced by reusing
the existing airframe created a nose-up
tendency under thrust. This could
allow pilots to inadvertently approach
a stall condition. As that did not happen with previous 737 models, pilots
migrating to the MAX from an earlier
model would not be expecting it.
Boeing reasoned that they could
write software to compensate for the
resulting tendency to lift the nose
under thrust, by programming in
opposing control movements. That
would make the plane feel similar to
operate to its predecessor, avoiding
the need to retrain pilots.
Australia's electronics magazine
This software (MCAS) uses the
angle-of-attack sensor to determine if
the aircraft is pitching nose up. If the
plane is reaching the critical AoA, the
flight computer operates the motorised
‘speed trim’, actuating the rear aileron
to pitch the nose back down again.
The speed trim is an existing system
on the 737 that allows pilots to ‘trim’
the aircraft to a neutral attitude, by
providing an adjustable offset to the
rear aileron to compensate for uneven
weight distribution. This avoids the
need for them to constantly press on
the control stick to stop the aircraft
from pitching up or down.
Boeing deliberately decided not to
mention MCAS in their flight manuals. Pilots were not briefed or trained
in its operation, as they wanted to be
able to sell the aircraft to airlines as not
August 2025 71
needing any pilot retraining. In combination with two other fatal flaws, that
turned out to be a big mistake.
Grounding
Following the Ethiopian Airline
crash, many countries around the
world moved to ground the 737 MAX.
The USA eventually followed, taking
the unprecedented step of banning
all 737 MAXes from flying until Boeing could confirm their airworthiness
with the FAA.
The grounding lasted 20 months,
during which time Boeing was forced
to wind back the influence of the
MCAS software and train pilots on its
use. New simulator sessions were also
conducted to provide pilots familiarity
with the differing flight characteristics
of the plane. Boeing was ultimately
penalised US$20 billion in fines and
compensation, and lost an estimated
US$67 billion in cancelled orders.
Conclusion
It is now mandatory for the MCAS
system to use two AoA sensors. This
brings MCAS in line with other critical flight systems, which must not
have a single point of failure. We still
can’t quite figure out why MCAS only
Undelivered Boeing 737 MAX aircrafts at Boeing Field in Seattle. Source:
SounderBruce – https://w.wiki/AGhr (CC-BY-SA-4.0)
used the data from one sensor when
two were already fitted to the aircraft!
It seems like a baffling oversight.
Apparently, Boeing believed that
MCAS was not ‘safety critical’. Early
iterations of the MCAS system could
not move the aileron enough to cause
a loss of control, but that was changed
before the first aircraft were delivered,
without revisiting the decision not
to use the data from the second AoA
sensor.
If its existence had initially been
disclosed to the pilots, simply having an off switch for the MCAS system
while leaving the trim motors under
manual control might also have saved
SC
hundreds of lives.
PIC Programming Adaptor
Our kit includes everything required to build the Programming Adaptor,
including the Raspberry Pi Pico. The parts for the optional USB power
supply are not included.
Use the Adaptor with an in-circuit programmer such as the Microchip
PICkit or Snap to directly program DIP microcontrollers.
Supports most newer 8-bit PICs and most 16-bit & 32-bit PICs with
8-40 pins.
Tested PICs include: 16F15213/4, 16F15323, 16F18146, 16F18857,
16F18877, 16(L)F1455, 16F1459, 16F1709, dsPIC33FJ256GP802,
PIC24FJ256GA702, PIC32MX170F256B and PIC32MX270F256B
Learn how to build it from the article in the September 2023 issue of
Silicon Chip (siliconchip.au/Article/15943). And see our article in the
October 2023 issue about different TFQP adaptors that can be used with
the Programmer (siliconchip.au/Article/15977).
Complete kit available from $55 + postage
siliconchip.com.au/Shop/20/6774 – Catalog SC6774
72
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
Subscribe to
JULY 2025
ISSN 1030-2662
07
The VERY BEST DIY Projects
!
9 771030 266001
$13 00* NZ $13 90
INC GST
INC GST
The SmartProbe
The perfect device for troubles
hooting
circuits. Probe voltages and
test continuity
with one hand, then put it
in your pocket
Solar USB Charging
Charge your USB devices
without paying a
cent for electricity
Hot Water Solar
System Diverter
Get the most out of your solar
panels –
this issue has the assemb
ly and testing
instructions
...and much more in this
Australia’s top electronics magazine
issue!
SpaceX
THE IR LATEST DEVELO
Silicon Chip is one of the best DIY electronics magazines in the
world. Each month is filled with a variety of projects that you can
build yourself, along with features on a wide range of topics from
in-depth electronics articles to general tech overviews.
PMENTS, INCLUDING STA
RSHIP
Published in
Silicon Chip
If you have an active subscription you receive 10% OFF
orders from our Online Shop (siliconchip.com.au/Shop/)*
Rest of
World
New
Zealand
Australia
* does not include the cost of postage
SmartProbe; July 2025
Length
Print
Combined
Online
6 months
$70
$80
$52.50
1 year
$130
$150
$100
2 years
$245
$280
$190
6 months
$82.50
$92.50
1 year
$155
$175
2 years
$290
$325
6 months
$100
$110
1 year
$195
$215
All prices are
in Australian
dollars (AUD).
Combined
subscriptions
include both
the printed
magazine and
online access.
2 years
$380
$415
Prices are valid
for the month
of issue.
Hot Water System Solar
Diverter; June-July 2025
SSB
Shortwave Receiver;
June-July 2025
Renew or extend your subscription before
September 1st to get in before the prices go up (see page 2).
To start your subscription go to
siliconchip.com.au/Shop/Subscribe
Part 1 by
Julian Edgar & John Clarke
This smart controller can
improve the energy efficiency
of your home. It can transfer
warm or cool air between
rooms automatically when
needed.
Ducted Heat
Transfer Controller
T
his device controls a mains-
powered fan that is used to transfer heat between rooms via ducts.
The controller can be used manually,
automatically, or based on a timer.
The wall-mounted LED gives an indication of the temperature difference
between rooms.
» Powered by the 230V AC mains
» Operates during all seasons without changes
» Three different operating modes
» Adjustable temperature difference and hysteresis
» Optional adjustable timer
» Optional fire alarm feature
» Wall plate button with sound and LED indicators
» Sensor disconnection indication
» Temperature difference options: 1, 1.5, 2, 3, 4, 5, 6, 8, 10 or 11°C
» Hysteresis options: 0.5, 1, 1.5, 2, 3, 4, 5, 6, 8 or 10°C
» Timer options: 15m, 30m, 1h, 2h, 3h, 4h, 5h, 6h, 8h, 12h or
multiples thereof
» Modes: manual, timed, or automatic
» Fire alarm function: switches on RLY2 and rapidly pulses the piezo
buzzer and LED when the temperature rise of either sensor is >8°C
per minute or 70°C is exceeded (this does not replace a smoke alarm!)
» Maximum total fan current: 10A
reasons reason for this. The first is that
the heater has been shut down – the
damper closed to reduce the airflow.
The second reason for smoke emissions is burning green wood that has
high moisture levels.
With current heaters that must meet
emissions standards, a wood heater
burning dry wood at full power produces no visible emissions. But the
key point is ‘at full power’ – throttling
the heater output reduces its efficiency
and increases emissions.
That’s where a fan-forced heat transfer duct comes in. It is much better to
keep the wood heater burning furiously and transfer some of that heat
to other rooms in the house than it is
to shut the heater down. Since most
homes using wood heating have only
one heater, using a transfer duct also
works to warm more than just the
room where the heater is located (see
Fig.1).
The second reason for using a
ducted heat transfer system is in
houses that use passive solar heating.
In southern Australia, windows facing
north can be used to warm the house
in winter. The sun shines in through
these windows, heating the wall and
floor surfaces of the room, and subsequently the air within.
Because the sun is higher in the
sky in summer, projecting eaves can
shade these windows in summer, so
Australia's electronics magazine
siliconchip.com.au
Advantages
The most common reason for using
a ducted heat transfer system is when
the source of heat for the house is confined largely to one room. There are
two likely situations where that would
occur: a wood heating stove is located
in one room, or passive solar heating
occurs largely at one end of the house.
While in some jurisdictions, wood
heating is frowned upon (for example, the Australian Capital Territory is
phasing out wood heaters), wood heaters can be environmentally acceptable
and, in some areas, cost little to run.
Wood heaters are effectively carbon
neutral; the carbon dioxide absorbed
by the trees during their growth is
released when the wood is burnt.
Wood heaters have a bad reputation for emissions – we’ve all seen
wood heater flues emitting a stream of
smoke for many hours. There are two
Features & Specifications
74
Silicon Chip
WARNING: Mains Voltage
Air return paths are required
A heat transfer duct works by moving air – that is, pushing air from one room
to another. But unless the air has a return path, the duct will not be very effective. Without a return path, air pressure will rise in the destination room, slowing the transfer of air. It’s therefore best to leave some internal doors open so
that good circulation can be achieved.
the northern windows don’t heat the
house when you don’t want them to.
In the northern hemisphere, this is
reversed – you want southerly windows.
However, the number of rooms in a
house that can face north is quite limited, so this type of heating can usually work in only one or two rooms.
That’s especially the case if the house
was never designed with passive solar
heating in mind. In this case, a ducted
heat transfer system can be added to
move solar heat to other rooms.
The problem with
commercial options
Fan-forced heat transfer ducts are
commercially available for installation in new or existing builds (see the
photo overleaf). Typically, they comprise flexible ducting and one or two
mains-powered fans. Common duct
and fan diameters are 150mm, 200mm,
250mm and 300mm. The fan and duct
are usually mounted in the roof space
with the inlet and outlet grilles located
in the ceiling.
Generally, these require you to
switch them on manually when
desired. You can certainly do that, but
it’s a little trickier than it first appears.
One thing that makes it tricky is that
the temperature differences can
be very small. For example, in a house that
uses passive solar heating, the temperature difference from the ‘warm’
northern room to the southern ‘cool’
part of the house may initially be only
2°C. That difference may increase
quite slowly – over hours.
Without either walking back and
forth to feel the temperatures, or consulting room thermometers, the best
time to turn on the fan isn’t at all
obvious. That’s if you’re even home
at the time!
Luckily, this Transfer Controller can
do the work for you.
Also, you may want the fan to operate for some time after you go to bed –
you’re no longer in the heated lounge
room, and you want that residual heat
distributed through the house. Or you
want to be manually in charge of when
the fan operates, but with a monitoring
LED showing when the heated room
is warmer by, say, 3°C than the room
at the other end of the duct.
Our Controller can perform all these
functions.
In long ducts, more than one fan
may be needed. The controller can run
fans up to a total power consumption
of 2300W (10A at 230V). Since most
This Direct Heat Transfer Controller
operates directly from the 230V AC
mains supply; contact with any live
component is potentially lethal. Do
not build it unless you are experienced
working with mains voltages.
duct fans are quite low in power, it
can likely drive however many fans
you need. If running multiple fans in
the duct, ensure they both blow in the
same direction!
Operating modes
The main function of the Ducted
Heat Transfer Controller is to switch
on the fan in the duct – the output is
simply on or off.
However, when it activates that fan
depends on the mode. Each mode
is selected by switch BCD4 on the
printed circuit board (PCB) – as with
the other set-up features, it is expected
that this will be set and then not frequently changed.
In all modes, the user interface is
a neat wall-mounted, spring-return
rocker-type pushbutton with a white
monitoring LED visible around the
periphery of the button, and a beeper
mounted behind. The other two inputs
are temperature sensors – one in the
room at each end of the duct.
Mode 0 is manual mode. In this case,
the pushbutton is used to switch the
fan on and off.
Mode 1 provides manually triggered timed operation. Pressing the
pushbutton switches the fan on for
a specified period. Each quick press
of the button adds (for example)
one hour of operation, so
one press gives one
Fig.1: a Ducted Heat Transfer System takes the heat from one room and distributes it to one or more other rooms. A fan in
the duct is used to move the air, and our controller determines when the fan switches on. Source: Vent-Axia.
siliconchip.com.au
Australia's electronics magazine
August 2025 75
This 150mm Ducted Heat Transfer
System uses a single fan to distribute
the air to two other rooms. Note
that this duct is uninsulated –
not a good idea.
Source: JPM Brands
Switch
BCD1 (temp.
difference)
BCD2
(hysteresis)
BCD3 (timer
period)
BCD4 (mode)
0
1°C
0.5°C
15 minutes
Manual
1
1.5°C
1.0°C
30 minutes
Timed
2
2°C
1.5°C
1 hour
Automatic
3
3°C
2°C
2 hours
Automatic
4
4°C
3°C
3 hours
Automatic
5
5°C
4°C
4 hours
Automatic
6
6°C
5°C
5 hours
Automatic
7
8°C
6°C
6 hours
Automatic
8
10°C
8°C
8 hours
Automatic
hour, two presses gives two hours etc,
up to a maximum of five presses. A
BCD switch preset determines the base
period, from 15 minutes to 12 hours.
Mode 2 is fully automatic. In this
mode, the fan operates when the
temperature difference between the
two ends of the duct exceeds a preset
threshold.
In addition to mode selector switch
BCD4, the PCB has three more adjustments.
BCD1 is used to set the temperature
difference that needs to occur before
the fully automatic mode (Mode 2)
switches on the fan. This can be set to
1, 1.5, 2, 3, 4, 5, 6, 8, 10 or 11°C.
BCD2 is used to set the hysteresis.
This is the difference between the
switch-on and switch-off temperatures. This can be set to 0.5, 1, 1.5,
2, 3, 4, 5, 6, 8 or 10°C. It must be set
lower than the temperature difference.
Let’s imagine the temperature difference is set to 4°C and the hysteresis is set to 1°C. If the heated room
is at 20°C and the unheated room is
at 16°C (a difference of exactly 4°C),
the fan will switch on. It will stay on
until the difference in temperature
decreases to 3°C; eg, the unheated
room warms to 17°C.
In use, if the fan switches on and off
too frequently, increase the hysteresis
setting. On the other hand, if the temperature of the room at the other end
of the duct varies up and down too
much, decrease the hysteresis.
BCD3 sets the timed period that
occurs in Mode 1 with each button
press. In the example above, I suggested that each press gives a onehour extension of the on-time. However, each button press can actually
be set to be 15m, 30m, 1h, 2h, 3h, 4h,
5h, 6h, 8h or 12h.
Refer to Table 1 for all the BCD
switch settings.
9
11°C
10°C
12 hours
Automatic
Monitoring LED and beeper
Table 1 – BCD switch settings
While we have described the function of the controller as operating a fan-forced
duct that transfers warm air to a cooler room, the system can also transfer
cool air to a warmer room. In fact, no changes are needed to do this because
the system operates based on the temperature difference between the two
rooms, rather than how much cooler the room is at the far end of the duct.
For example, say you have the difference in room temperature set to 3°C
and the Mode set to 2 (Automatic). When the room at the end of the duct is
3°C warmer than the room at the beginning of the duct, the fan will switch on,
transferring cooler air to the hotter room. Of course, the source room needs
to be the same room in both winter and summer.
In addition to the pushbutton
switch, the wall-mounted indicator is
equipped with one LED and a beeper.
The beeper operates in the same way
in all modes: a single beep indicates
switch on (a short press of the button)
and a triple beep indicates switch off
(achieved by a longer press of the button). The triple beep comprises a single
beep followed by a quick double beep.
The LED can show different information in each mode.
In manual mode, if the fan is off, the
Australia's electronics magazine
siliconchip.com.au
What about transferring cool air?
76
Silicon Chip
LED is off, possibly flashing on briefly.
If the fan is on, the LED is on, possibly
flashing off briefly.
If it’s flashing briefly every two seconds when the fan is off, that indicates
the measured temperature difference
is greater than or equal to the set temperature difference, so you might want
to switch it on. Similarly, if it’s briefly
flickering off while the fan is on, that
means the temperature difference has
fallen below the set difference (including hysteresis), indicating you may
want to switch it off.
Manual Timed operation (BCD4
position 1) has LED behaviour that is
the same as the manual mode.
Automatic mode (BCD4 position 2)
has different LED behaviour. If the system has been disabled, the LED flashes.
If the fan is on, so is the LED. If the fan
is off, again, so is the LED. A summary
of these modes is shown in Table 2.
Other potential uses
This device can also control a powered ventilator or fan; for example,
one that ventilates a hot roof cavity in
summer. In this use, one temperature
sensor is placed in the roof cavity (or
other hot area needing ventilation) and
the other outside in an area protected
from the weather (eg, under the eaves).
In this application, the best settings
will probably be Mode 2 (automatic),
with the temperature difference set
higher than you would use for internal house use (eg, 10°C with 5°C of
hysteresis).
Another use is for solar air heaters.
While uncommon in Australia, these
have been widely used in solar homes
in the United States. In this approach,
air is heated by a flat plate collector –
a little like a traditional solar water
heater but with air rather than water
heated through contact with the plate.
When the air in the heater rises
sufficiently in temperature, a fan can
be used to move that heated air into
the house through conventional air-
conditioning ducts. In this application, one sensor would be placed so
that it is exposed to the air in the heater
(but shielded from direct sunlight),
while the other would be placed inside
the house. The temperature difference
would be set quite low (eg, 3-4°C, with
perhaps 2°C of hysteresis).
Parts List – Ducted Heat Transfer Controller
1 polycarbonate IP65 enclosure, 171 × 121 × 55mm [Altronics H0478, Jaycar HB6218]
1 double-sided, plated-through PCB coded 17101251, 151 × 112mm
1 lid panel label (84 × 65mm) and side panel label (64 × 10.5mm)
1 3VA 9+9V PCB-mounting mains transformer (T1) [Altronics M7018A]
1 FRA4 250V 30A AC SPST relay with 12V DC coil (RLY1) [Jaycar SY4040]
1 PCB-mounting 250V 10A AC SPDT relay with 12V DC coil (RLY2)
[Altronics S4160C, Jaycar SY4066]
4 PCB-mounting 10-position BCD switches (BCD1-BCD4) [Altronics S3001] OR
4 2×4-pin headers and 12 jumper shunts
1 2-way header, 2.54mm pitch (JP1)
1 jumper shunt (JP1)
2 15A 300V 2-way screw terminals, 8.25mm pitch (CON1, CON2) [Altronics P2101]
1 2-way screw terminal, 5/5.08mm pitch (CON3)
1 3-way screw terminal, 5/5.08mm pitch (CON4)
3 8P8C RJ45 PCB-mounting horizontal sockets (CON5-CON7) [Altronics P1448A]
1 IEC mains input socket with integral fuse [Altronics P8324, Jaycar PP4004]
1 mains lead with IEC plug
1 surface-mounting mains socket (GPO) [Altronics P8241, Jaycar PS4094]
1 20-pin DIL IC socket (optional, for IC1)
1 fast-blow 10A M205 fuse (F1)
Hardware
2 M4 × 10mm panhead machine screws with matching hex nuts
2 M3 × 15mm panhead nylon machine screws
5 M3 × 6mm panhead machine screws
3 M3 brass hex nuts
1 200mm cable tie and 8 100mm cable ties
1 3-6.5mm diameter wire entry cable gland
Wire & cable
1 200mm length of black 7.5A hookup wire
1 50mm length of light-duty red hookup wire and light-duty black hookup wire
assorted lengths of 10A mains-rated green/yellow striped wire (150mm length);
brown wire (200mm length); and blue wire (100mm length)
3 Cat 5, Cat 5E or Cat 6 patch leads, lengths to suit installation
assorted lengths of clear heatshrink tubing (70mm length, 5mm diameter;
30mm length, 4mm diameter; and 50mm length, 1mm diameter)
Semiconductors
1 PIC16F1459-I/P microcontroller programmed with 1710125A.HEX, DIP-20 (IC1)
1 7805 1A 5V linear regulator, TO-220 (REG1)
3 BC337 NPN transistors, TO-92 (Q1-Q3)
1 W02(M) or W04(M) 1.5A 200V/400V bridge rectifier (BR1)
16 1N4148 200mA 75V diodes (D1-D16)
3 1N4004 1A 400V diodes (D17-D19)
Capacitors (16V PC radial electrolytic, unless specified)
2 470μF
1 100μF
2 100nF 63V/100V MKT polyester
Resistors (all ¼W, 1%)
5 10kW
2 2.2kW
4 1kW
1 470W
Control panel parts (per panel)
1 double-sided, plated-through PCB coded 17101253, 51 × 67mm
1 Clipsal Iconic 3041G single Gang Switch Grid Plate ●
1 Clipsal Iconic 3041C-VW single Gang Switch Plate Cover (Skin Only) ●
1 Clipsal Iconic 40FR-VW Fan Dolly Rocker Vivid White ●
1 Clipsal Iconic 40MBPRL-VW 10A Momentary Bell Press Switch Mechanism with LED ●
1 panel label, 45 × 30.5mm
1 top-entry 8P8C vertical RJ45 socket (CON10) [Altronics P1468]
1 3-16V self-oscillating piezo buzzer [Altronics S6104]
1 2-way vertical polarised header, 2.54mm pitch, with matching plug and pins (CON11)
1 2-way terminal block, 5/5.08mm pitch (CON12)
1 8P8C double adaptor (only required if using two control panels) [Altronics P7052A]
● available from electrical wholesalers, including www.sparkydirect.com.au
The complete circuit for the Ducted
Heat Transfer Controller is shown in
Temperature sensor parts
2 60 × 60 × 20mm vented enclosures or similar [Jaycar HB6116]
2 double-sided, plated-through PCBs coded 17101252, 20 × 37.5mm
2 8P8C RJ45 PCB sockets (CON8, CON9) [Altronics P1448A]
2 DS18B20 temperature sensors (TS1, TS2) [Altronics Z7280 or Z6386]
siliconchip.com.au
Australia's electronics magazine
Circuit details
August 2025 77
Fig.2. Microcontroller IC1 monitors
the temperatures via sensors TS1 &
TS2, which connect to the main board
via 8-way Cat 5 (or similar) cables and
RJ45 plugs/sockets. In each case, pin
4 carries the digital signal, pin 8 the
5V supply for the sensor and pins 5 &
7 are grounds.
TS1 & TS2 are Maxim DS18B20
1-wire digital thermometers. Just one
data line (DQ) is required for serial
communications. A minimum of one
extra connection for the common
ground connection is also required.
Power for the sensor can be derived
from the data line, but we include a
Enabling the fire alarm feature
The Ducted Heat Transfer Controller can also be configured as a fire alarm.
Because the system has temperature sensors that would normally be placed
at divergent ends of the house, monitoring of these sensors provides a widespread back-up system to the legally required smoke detectors.
When this function is enabled by shorting the pins of JP1, each temperature sensor is monitored for both the temperature and the rate of temperature change. If the temperature exceeds 70°C and/or the rate of temperature
change exceeds 8°C per minute, the beeper and LED rapidly pulse. Relay RLY2
is also energised, which can power a low-voltage warning siren, switch on
low-voltage lights etc.
If the fire alarm goes off, a short press of the wall-mounted button will
switch off the buzzer, but the LED will continue to flash. A long press will
switch off the buzzer, LED and RLY2, and the system will be re-armed to monitor again for fire.
Note that this is a mains-powered system with no battery back-up. It should
always be used in conjunction with traditional battery-powered or battery-
backed smoke detectors.
We suggest that this function be activated in all installations since it’s unlikely
to ever be triggered unless there is a fire.
78
Silicon Chip
Australia's electronics magazine
direct 5V supply connection (Vdd/V+)
since we have enough wires and this
makes signalling easier.
Two-way communication between
the microcontroller and temperature
sensor is possible since the DQ pin
is an open drain with a pull-up resistance of 2.2kW. Open drain means that
the drain of a Mosfet connects to this
pin, so when the Mosfet is on, the pin
is pulled to 0V, while if it is off, it is
pulled up by the resistor.
A Mosfet at either end of the wire
can be used to pull it down to 0V, so
a signal can be sent by the device at
either end of the wire. The microcontroller uses its RC2 and RB4 I/O pins
to request temperature readings and
get them from the sensors.
The DS18B20 has a temperature
reading accuracy of ±0.5°C from -10°C
to +85°C. Temperature readings are
available in 0.125°C steps, but for this
project, we measure the temperature
in 0.5°C increments.
BCD switches
The four BCD switches that select
the various mode, temperature and
timer features have internal contacts
siliconchip.com.au
Fig.2: microcontroller IC1 reads the positions of BCD switches 1-4
(or the alternative jumper sets) to determine is jobs. It then reads
the temperatures from sensors TS1 & TS2 connected via Cat 5/5E/6
cables and determines when to energise relay RLY1 to connect mains
power to the fan(s).
that connect the “1”, “2”, “4” and “8”
terminals to ground in a combination
that totals to the switch setting. For
example, if the switch is set to the 9
position, the “1” and “8” terminals
will be connected to ground but the
other two won’t.
This allows IC1 to sense 16 possible
positions for each switch using four
wires (although these switches only
have 10 positions).
Rather than the common (C) terminal of each switch being connected to
ground, they are connected to a separate pin on microcontroller IC1. This
way, the micro can pull them high one
at a time, and use the same four lines
(RA1, RC5, RA0 & RC4) to read the
position of the selected switch.
Isolation diodes D1-D16 are required
because, while the other switches can
be set to have their common terminals
floating while one switch is sensed,
those switches could still end up
effectively shorting two or more of
the sense lines together, depending
on their positions.
We need the diodes to ensure the
switches don’t affect each other during
the sensing procedure.
siliconchip.com.au
During switch sensing, any open
BCD switch will be pulled low to 0V
via one of the 10kW pull-down resistors.
BCD switches can be expensive, so
we have provided an alternative system using a 2×4-pin header with up
to four jumpers placed on it to replace
each BCD switch.
Fig.3 shows how the jumper settings
equate to BCD settings. Since these
settings are rarely (if ever) changed,
there’s little disadvantage in using
jumpers on headers instead.
Control Panel
The wall-mounted control panel
for the Ducted Heat Transfer Controller comprises switch S1, LED1 and a
piezo buzzer.
This is all incorporated in a Clipsal
sprung-return rocker switch plate that
includes an indicating LED. The piezo
buzzer is an addition to the switch
Fig.3: this shows the simple
binary codes you need if
using jumpers instead of
the BCD switches. IC1 also
monitors switch S1 and the
four selection switches, BCD1
to BCD4. In response to these
settings and temperature
readings, the microcontroller
can sound the piezo buzzer,
light LED1 and switch on
RLY1 to drive the duct fan.
IC1 can also switch on RLY2
if the fire alarm feature is
selected with JP1 and is then
activated.
Australia's electronics magazine
August 2025 79
The Ducted Heat Transfer Controller is housed in a polycarbonate IP65
enclosure (upper right photo). An IEC mains cord supplies power and the duct’s
fan plugs into the power outlet on top. The temperature sensor and control
panel connections are made using RJ45 sockets and Cat 5/5E/6 cables.
The Controller is easy to build, with only through-hole components
used. Care must be taken with the mains voltage wiring, though.
The faceplate (upper left photo) incorporates a momentary rocker
switch, piezo buzzer and a white LED that lights the periphery of the
switch. The wall plate can be mounted vertically or horizontally – this
one is configured for vertical mounting. The ‘floppy ears’ can be easily
removed (they’re not needed for normal mounting).
The room temperature sensors are each located in small, ventilated
wall enclosures (photo shown at right).
80
Silicon Chip
Australia's electronics magazine
plate to complete the control panel.
The control panel connects to the main
board via another Cat 5/5E/6 cable and
RJ45 plugs/sockets.
LED1 is driven from the RB6 output of IC1 through a 470W resistor to
ground. The LED current is around
4.25mA, assuming a voltage drop of
3V across the white LED.
Switch S1 is connected between
GND and the RB5 input of IC1,
with this input pulled to 5V via
a 1kW resistor when the switch is
open. If the switch is closed, RB5
will be pulled to GND and IC1 can
detect that.
The piezo buzzer is powered from
12V using transistor Q3 to switch the
negative side to ground. When the
buzzer is required to sound, the RC7
output of IC1 is driven high to switch
on Q3 by delivering current to its base
through a 1kW resistor.
Relays RLY1 & RLY2 are switched
on via the RC3 and RC6 outputs of
IC1, respectively. Both use a 1kW
base resistor to drive a transistor
to power the relay coil. Transistor
Q1 is used for RLY1 and Q2 for
RLY2. Diode D17, across RLY1’s
coil, and D18, across RLY2’s
coil, quench the back-EMF voltage from the coil when these are
switched off.
RLY2 is uncommitted and is
intended to drive a low-voltage
siren for the optional fire alarm
function. RLY1 connects the incoming mains Active to the fan socket
when the fan should be powered.
The output socket’s Neutral and Earth
pins are permanently wired to the
input socket.
Power for the circuit is derived via
a mains transformer that produces
a 9V AC output. This is rectified by
bridge rectifier BR1 and filtered by
two 470μF capacitors, giving close
to 12V DC. This is used to power the
two relays and the piezo buzzer. REG1
is a 5V regulator that drops its 12V
input to 5V to supply IC1 and the
DS18B20 temperature
sensors.
Next month
That’s all we
have space for this
issue. Next issue,
we’ll cover building
the unit and setting
SC
it up.
siliconchip.com.au
Table 2 – smart remote push button/LED/buzzer
Mode
Push button/buzzer
Fan status
Faceplate LED
‘0’
Manual fan on/off
Short press
– beep
– on
Runs when
fan manually
switched on
Fan off
Temp difference < set point
LED off
Longer press
– double beep
– off
Fan off
Temp difference > set point
LED flashes momentarily on once every 2s
Fan on
Temp difference < set point
LED flashes momentarily off once every 2s
Fan on
Temp difference > set point
LED fully on
‘1’
Manual fan timed
operation
Quick press or presses = on for set Runs for
period of operation, e.g. when timer timer period
is set for 30m, 1 quick press runs
when set
fan for 30m, 5 quick presses sets
‘on’ period at 150m (2.5h)
Longer press
– double beep
– off
Fan off
Temp difference < set point
LED off
Fan off
Temp difference > set point
LED flashes momentarily on once every 2s
Fan on
Temp difference < set point
LED flashes momentarily off once every 2s
Fan on
Temp difference > set point
LED fully on
‘2’ or more
Automatic
Short press
– beep
– system on
Runs when
System disabled
temperature
LED flashing
difference
exceeds
Fan off
preset level
LED off
when system
activated
Fan on
LED on
Longer press
– double beep
– system switched off
Fire alarm
activated (JP1
shorted and fire
detected)
Buzzer sounds rapidly and LED
flashes rapidly at 5Hz
Fan off
N/A
Short press, buzzer sound is off,
LED flashes rapidly
Long Press, LED and buzzer off
and retests for fire
siliconchip.com.au
Australia's electronics magazine
August 2025 81
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.
High speed transmission using regular opto-couplers
This circuit shows how you can
achieve high speeds with low power
consumption using standard, inexpensive opto-couplers.
It can be difficult to get high speed
out of readily available and inexpensive opto-couplers. For example, the 4N35’s data sheet says you
can achieve rise and fall times below
10μs if you drive the opto-coupler in
a not-very-practical way.
The example given in the 4N35 data
sheet is the same as shown in circuit
(a) but using a 100W pull-up resistor
to a 10V supply, and choosing the LED
drive resistor such that the phototransistor’s output current is 2mA.
This means the output signal will
vary from 9.8V to 10.0V as the LED
switches on and off, so you’ll need
more circuitry to turn this into a useful signal. Also, 10V may be difficult
to achieve if you’re using a battery to
supply the output circuit, 2mA will
run your battery flat more quickly than
82
Silicon Chip
you’d like, and the LED resistor will
vary from one circuit to the next due
to opto-coupler variability.
Circuit (b) here overcomes these
problems. It works with any supply
voltage, and with as little as 10μA of
pull-up current. As shown, it will couple a 19,200bps RS-232 signal with a
pull-up current of 14 microamps.
You can see the results in the Scope
1, which shows a lowercase “u” transmitted via RS-232. From top to bottom,
the four traces shown are:
1. The output of circuit (a) using a
220kW pull-up resistor in place of the
1kW resistor shown
2. The output of circuit (b) using the
220kW pull-up resistor
3. The output of circuit (b) with the
optional high-current output stage
4. The RS-232 input signal
The main reason opto-couplers are
slow is because they have a phototransistor output and the phototransistor
switches into saturation. Like ordinary
Australia's electronics magazine
transistors, phototransistors are slow
to come out of saturation. In the top
oscilloscope trace, the phototransistor switches on fast enough, but off
very slowly.
One way to help the phototransistor switch off faster is by connecting
a resistor between the base and emitter. This improves the switch-off time
significantly, but the Miller Effect then
comes into play, slowing the rise time
and preventing the rise time from
being as fast as the fall time.
A better way to achieve high speed
is to short the phototransistor’s base
and emitter together, effectively turning the phototransistor into a photodiode. This gives much better speed,
and the photodiode behaves like a
microscopic solar cell. When the LED
is on, the photodiode generates about
450mV at about 10μA.
By stacking two opto-couplers
together, we get 900mV, enough to
drive a small-signal transistor. This
is the BC546 shown in circuit (b),
although any small-signal NPN transistor will probably work. The 1nF capacitor is known as a ‘speed-up capacitor’.
When connected this way, it helps the
BC546 to switch off quickly.
The pull-up resistor can supply
anywhere from 10μA to 1mA. The
optional extra stage allows loads up
to 100mA.
Note the LED connected in anti-
parallel across the opto-coupler’s
LED pins. This is added because the
opto-coupler’s LED will be damaged
if more than 5V is applied across the
LED in reverse. It limits the reverse
voltage to no more than 2V. An ordinary diode could be used.
Circuit (c) uses a similar principle,
but a micropower comparator turns
the 450mV output of the photodiode
directly into a logic-compatible output.
Note that you can’t use a dual
opto-coupler in place of the two
opto-couplers, as there is no base
connection available in dual opto-
couplers, preventing you from shorting the base and emitter of each
siliconchip.com.au
Wireless battery charger
Three-way latch
This circuit shows a simple way
to wirelessly charge a small battery.
It is built around NPN transistor Q1
(2N3866), diode D1 (1N4148), LED1
and a few other components.
Transistor Q1 forms a 100MHz RF
medium power oscillator. Inductors L1
and L2 are made with four close turns
of 0.9mm diameter enamelled copper
with a diameter of 6mm. Coil L2 is kept
near L1 (2mm apart). There is no electrical connection between L1 and L2, but
RF signals from L1 get induced to L2.
The voltage and current obtained
from L2 are sufficient to drive LED1
or charge a 1.2V NiCd rechargeable
cell after being rectified and filtered.
After assembling the circuit on a
PCB, enclose it in a suitable plastic
box. Place L1 and L2 next to each other
such that LED1 glows when switch S1
is in the TEST position. When S1 is
in the CHARGE position, the cell will
be charged.
The alternative section shown in the
cyan box can replace the right-hand
section to make it work even if the
polarity of one coil is reversed.
Raj. K. Gorkhali,
Hetauda, Nepal. ($60)
phototransistor together. There are
also some single opto-couplers that
don’t have a base connection either.
In case you’re wondering if you
could stack ten or more opto-couplers
together, and use the resulting voltage as a supply for an electrically isolated circuit, yes, you can. But as this
arrangement can only supply microamps, you’ll need good design skills
to make a circuit that can run from so
little power.
Also, special opto-coupler chips are
available that do this internally; for
example, the PVI5050 and the PVI1050
(among others). Such chips are commonly used in solid-state relays, with
the output driving a Mosfet.
Russell Gurrin,
Highgate Hill, Qld. ($120)
Scope 1: a lowercase letter “u” transmitted via RS-232 using the different
circuits shown above.
siliconchip.com.au
Australia's electronics magazine
In the configuration shown, it
drives three LEDs individually.
When one SCR is switched on via
the momentary switch through the
10kW resistor, the anode voltage
of that SCR drops suddenly. The
100nF capacitors send a negative
pulse to the anode of the adjoining
SCRs, causing them to switch off.
This circuit can be extended to
a higher number of SCRs, although
you may need to increase the value
of the capacitors as they are effectively in series (eg, for coupling a
pulse from SCR1 to SCR3 or vice
versa). The circuit is reset by disconnecting power.
The LEDs can be replaced by
opto-couplers to control external
equipment, eg, to drive relays to
make a ganged latching switch.
Joe Curulli,
Perth, WA. ($60)
August 2025 83
SERVICEMAN’S LOG
Mirror, mirror on the door
Dave Thompson
Our five-year-old car was coming to the end of its warranty. Typically,
everything goes wrong about a week after that, with the previously
perfect car suddenly becoming one of those jalopies from the old silent
movies, where everything literally falls off, leaving the hapless driver
sitting on the road with a steering wheel in their hands.
This meant we were faced with a relatively major firstworld decision. Having only purchased used cars all my
life, and having to deal with the headaches those cars inevitably brought, being able to own a brand-spanking-new
car was a real luxury.
Anyway, we ended up with a new car, which of course
means learning all the new tech onboard and what all
the buttons and switches do. Modern cars are apparently
more advanced than the 1970s moon lander, and that’s a
fact (which I read on the internet, so I am sure it is true).
I can almost believe it with the radar, the cameras and
all the electronic doodads and gizmos (you can tell I’ve
spent a lifetime in electronics). But it is a learning curve,
and the manuals that come with these cars are like those
olde world phone books we used to get. It took me half
an hour to find out how to change the clock for daylight
saving time!
Of course, once I discovered the appropriate section in
the book, it took mere seconds to figure out how to do what
I wanted. But my point is that modern cars are hugely complex machines. Saying that, they are apparently not clever enough to switch
clocks over automatically, like
computers and phones have
done for decades when daylight saving clicks in. Or at
least, our one isn’t.
Maybe it’s because they are
all built down to a price these
days. I also note that the car’s
clock, built into the instrument
display, is not tied to the multimedia/GPS display system,
so often there are two slightly
dissimilar times being shown.
That’s not a big deal, of
course, but it seems silly that an
integrated GPS/computer system and the car’s basic display functionality don’t talk
to each other. Or maybe I’m
just an older guy pushing the
wrong buttons...
I have to say this thing is a
nice place to be, and a real step
up from the original Mini I used to
84
Silicon Chip
drive, which almost broke my spine every time I went out
in it due to the still quake-damaged roads here. It had all
the onboard technology of a particularly low-spec wheelbarrow.
Then my 1997 MG-F was great, until it kept breaking
down and I was getting far too old to be seen in it, especially with the top down. I won’t even mention the inelegance of me getting in and out of such a low car in public!
A computer on wheels
So this car is a lot larger than our previous cars, and
although it is bristling with the sort of technology and cameras you would expect to see on the latest Apache helicopter gunship, I still try to rely mainly on my old-fashioned
(yet increasingly fading) eyesight to make sure nothing is
nearby, or that I am not running over or hitting anything
that could cause any legal problems.
This is the way I’ve always done it, but I can feel the
ever-increasing pull of using the onboard tech to compensate for my flagging senses. I’ve also noticed that many
younger drivers just rely on the cameras and radar rather
than actually looking out the
windows, which is very disconcerting.
Backing down a driveway
by a school and relying solely
on the camera display in the
centre console is frankly
frightening.
This was reinforced
recently by me driving for
an appointment in town with
a very challenging car park.
It isn’t that the spaces are
overly tight, like so many
are now (especially when
driving a bigger car than
a Mini!), but because the
building has all these huge
square concrete pillars
holding it up. It’s like it is
on stilts, and there is car
parking below.
I always avoid parking
under structures like this
after experiencing some
Australia's electronics magazine
siliconchip.com.au
pretty terrifying quakes here,
watching similar pillars flex
like they were made of rubber and start cracking while
we were trapped in stand-still
traffic trying to get out of a mall
carpark. Once you see a road
waving and breaking apart,
things are never the same.
So, I try to park outside
places like this, and fortunately, there were half-a-dozen
spaces on the outside edge of the
building. The only free space in the
line was beside a pillar, so that’s
the one I took. It was tight, but
using the cameras and my vision,
I got in OK, despite the proximity
warnings going off madly and filling
the car with beeping noises.
I went off happily to my appointment, knowing I had a close parking
spot, and all was well in the world. I
was out in good time and back to the
car. In the meantime, someone had replaced the car that
was originally to my right. I’m not suggesting they snuck
in and swapped it, but obviously the original parker had
left and someone else came in.
Boxed in
This ‘new’ car was a large Range Rover (probably owned
by some medical staff if the personalised plate was anything to go by), and they had parked it quite close to my
door. It was a struggle to squeeze my no-longer-so-nimble
frame into the car, but I did it without marking his car with
my door and vice versa.
I fired up the steed, and of course, the usual cacophony
of beeps and sirens went off. Radar to the front shows I
am nosed into a decorative hedge, and to the sides, there
were other cars. It is so distracting, especially as I’m the
sort who has to switch off the radio to see better! There
were also people driving behind me to watch out for,
so a lot was going on.
I was most concerned about this huge square pillar
beside me, on the left. I didn’t want to hit it, and
hitting the Range Rover was not an option I
wanted to explore. I very gingerly backed
out, trying to take in everything in around
me. Then it happened; I touched the
very outside of my wing mirror on that
$%#%<at>! post.
Ironically, there was a
rubber buffer on each corner, but that was just out
of my sight line. The mirror
has a transparent side-light
indicator plastic piece that
protrudes from it, which I
can’t see from the cabin. I
just touched it on the pillar, blowing the plastic parts
of the mirror off. As you could
imagine, the air was blue.
siliconchip.com.au
I stopped, as there were bits on
the ground (I saw them fly off) and
I didn’t want to run over anything. It
seemed that I had popped the coloured
plastic back off the mirror, and two clear
plastic lenses that were inside the assembly onto the car park floor. Great.
I looked at the bits in my hands
and decided that this was too hard
while blocking a car park accessway.
I did pop the body-coloured
plastic cover back on, and it simply clipped into place, but I could
see that one of the clips was missing and one of the clear lenses
had broken in two. Those extra
bits just went in the back seat.
Of course, I felt really good
about myself at this point. Nice
new car, broken bits. Excellent. So that was an interesting trip home, and of course, there’s
the inevitable explanation to the longhaired general about how and why something got broken.
That went as you would expect. But, I pulled my sleeves
up, as I am a serviceman, and something must be able to
be done about this!
As I mentioned, I’d already popped the coloured plastic
housing of the mirror back onto the body. What was missing from that were the clear plastic bits that made up the
rest of it. I retrieved them from the back seat to see what
damage I’d done. There are two clear pieces; one is thinner and was broken in half. The other was complete, but
the concrete pole damaged its end.
If I had a 3D printer, I might have been able to make
something that would work, but it was apparent that my
skills ran out the moment I broke it. On top of all that, the
paint on the edge of the mirror had been scraped, and the
white undercoat showed through, so at the very least that
would have to be sanded and touched up. All in all, this
was looking like a ‘too hard basket’ job for me.
It’s a shame because the rest of the mirror was
fine; the bit under the cover was undamaged, and
the guts of it were pristine.
There is so much tech in
these things. The glass itself
has LEDs built into it; cameras, radar sensors, and even
the positioning motors and
cabling are all packed into
the housing somehow.
My old Mini’s wing mirror required me to get out in
all weathers to adjust it, then
get back in the car, then get
out again to adjust it further.
Half the time I couldn’t see
anything in it anyway due
to fogging, rain and the
wrath of God.
Essentially, it was hopeless,
but these new ones on modern
Australia's electronics magazine
August 2025 85
Items Covered This Month
• Objects in mirror are closer than they appear
• Repairing the cable on a National fan
• Fixing a young laser printer
• The decorative fix
• A voltmeter that only looked perfect
Dave Thompson runs PC Anytime in Christchurch, NZ.
Website: www.pcanytime.co.nz
Email: dave<at>pcanytime.co.nz
Cartoonist – Louis Decrevel
Website: loueee.com
cars are something to behold. Except mine, of course, now
that it’s broken.
Right to replace
That’s the thing, though; it isn’t really broken. Sure, a
few cosmetic bits are broken, but it could be repaired and
repainted. Still, I expect any repairer is just going to take
it off and throw it in the skip. It has a few blemishes, but it
still works; the radar and cameras in it still operate properly, and I can still see stuff behind me.
The problem is that repair guys don’t seem interested in
replacing small parts (if they’re even available!) or repainting the housing. They just charge it up to insurance and
put on a new one. Everyone in the system is making a killing, except for the poor chump paying their insurance and
knocking wing mirrors on concrete poles.
A friend of mine has another new Japanese brand car
and had a similar incident in his driveway. He went
through the same process, and was told it would be at
least a month to replace the mirror and would cost the
$1500 excess as well.
That seemed ridiculous to him, and so he went and
took his car to a local body shop, or collision repair centre
(whatever they are called these days) and for 50 bucks and
a dozen beers, the guys there plastic welded it, resprayed
it and refitted it within a day. I couldn’t even tell it had
been repaired.
So where is all this juicy insurance money going? It seems
like a huge rort. But then again, I suppose it always has
been. The system is set up to rip money from somewhere.
Rants aside, I thought I might go down the same road and
went to a local place. I do like to support local businesses,
as many locals have supported me over the years. I pulled
up into their car park and asked them about the mirror.
The guy looked at it and the parts I presented, then
hummed and hawed and said it would be cheaper to
replace the whole thing. I asked for a breakdown of what
it would cost to touch up and cover and replace the plastic
lens parts, and he basically said those parts are not available. They’d have to organise a whole new mirror, but not
to worry; insurance would cover it.
This seems incredibly wasteful, just to throw this monthsold part into the bin (if that’s really what they do with it).
So it seems I will be going down the mainstream route and
playing into the system.
I don’t approve, but in this case there is nothing I can
do. I can’t get the parts, I can’t repair what I have due to
bits missing and even if I could do it, it would likely look
shoddy and vex me for as long as we owned the car.
86
Silicon Chip
A wasteful system
This reeks of this whole ‘right to repair’ debacle. In the
old days, they would simply repair something like this with
all their skills and return it to the car, all without creating
a mountain of waste.
If that ‘old’ mirror doesn’t end up under the bench for
spares, where does it go? In a skip and then a landfill? What
a waste of resources, with all that tech in it, all for the sake
of a few dollars’ worth of plastic parts.
I understand these companies don’t want to invest capital into parts only to have them sitting on a shelf somewhere gathering dust, and then when the model involved
isn’t about anymore, those components are wasted anyway,
selling very rarely. But I am sure there was a law here that
any cars sold in this country had to have a 20-year supply
of spare parts.
That could be just folklore, and any spare parts laws
might be legally bent to mean just tyres or complete wing
mirrors or even bulbs for the headlights. I recall going to
local parts places with dad and asking for a flange valve
regulator for a 1959 Standard 10, and the guy would just
go to the shelf and get one. Then say, oh, sorry, you want a
right-hand one, then go back and get it. I think those days
are gone. And that’s a shame.
This is exactly the same as computer manufacturers now.
The only real parts you can source now are from used models bought by companies who disassemble the machines
for parts and sell them on. Unfortunately, in my experience at least, what I get from them is often not what was
shown on the website.
Long story short, I had to take it in and got a loaner
car for a few days while they swapped out the mirror. Of
course, it looks exactly the same; we haven’t had it long
enough for any sun-faded colour mismatching. Still, the
whole experience, from the shame of doing it in the first
place and the having to resort to the repair system, gave
me some pause for thought.
National fan cable repair
My wife recently found an old National fan at the local
tip shop recently and asked if I wanted to get it. These
old fans are very reliable, and I thought it had a very good
chance of working, so we did. The only visible problem
was that someone had cut the plug off the end of the figure-
eight power cable.
Australia's electronics magazine
siliconchip.com.au
I knew I had a plug at home that was designed for this
type of cable, so we got the fan and headed home. We have
several old fans at home that are not in very good cosmetic condition, but they are reliable. Some must be over
50 years old now.
The fans available in department stores these days are
incredibly unreliable. They have a non-resetting thermal
fuse buried deep within the motor windings, so if the fan
motor gets too warm, the fuse blows and that’s the end of
it. There is no way to repair them, short of replacing the
motor, which is not available. They are cheap junk.
I couldn’t find the figure-eight plug I knew I had, and
decided not to fit a regular plug to the cable, as it would
not be secure due to the smaller size of the cable. Instead, I
would just replace the cable with one I pulled from another
defunct appliance.
I found a nice cable with a two-pin plug that matched
the fan well. It would look original and better than the old
figure-eight cable that the fan came with.
The fan’s bottom panel was held on with five screws, one
in each of the four rubber feet near the corners and one in
the middle of the base. With the screws removed, I set the
panel aside. The old cable was secured with a cable clamp
and wired into a terminal block.
After unscrewing the four screws, I had the old cable
free, and I prepared the new cable ready to install. A short
time later, I had the new cable installed, replaced the bottom panel, and the fan was ready to use. It worked straight
away, so the cut-off plug was the only thing wrong with
it. This fan only has two speeds, unlike most, which have
three speeds.
It’s amazing what shows up at the tip shop from time
to time. However, I have found that in more recent times,
that there seem to be fewer people throwing things out and
more people shopping at the tip shops. This means fewer
goodies are available for purchase, although it’s still possible to find a variety of useful items. It’s just a matter of
being there at the tight time to nab a bargain.
Bruce Pierson, Dundathu, Qld.
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? It 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
cars and similar.
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.
purchase. There was a CD-ROM used to install it, which
I had forgotten about, so I immediately inserted it to reinstall the drivers.
It also checked for updated firmware, but the updates
made no difference; the Brother printer continued refusing to print.
I checked the print queue after trying to print documents
and also several test prints, but there was nothing in the
print queue. So it would appear that the data flowing from
the computer was not being processed because of some
fault at the receiving end.
The disc supplied had an easy setup for wireless printing so I figured that if I could re-route the data via wireless,
I could bypass the input from the USB cable and establish that the printer was functioning and maybe able print
pages wirelessly.
After going through the time-consuming setup and having to restart everything, including the household server,
a page was finally printed successfully, but it was not the
end of the problem!
When I tested a photo or graph, I got nothing, or sometimes a fraction of a printed page and it took forever to
print just a letter.
Brother HL-L3230CDW laser printer repair
Normally, this is a very reliable printer. In fact, it is the
best colour laser printer I have owned. After only about
two years of use, I was shocked to get a weird message and
a failure to print. Every time I pressed “print”, the printer
lit up the green “data” LED, but after about 10 seconds, the
yellow “error” light began to flash. Then it just stopped and
showed “ready” on the LCD panel.
Checking all the connections and restarting both the
printer and computer changed nothing. The troubleshooting option on Windows 10 printers and scanners proved
worthless. I checked on the Brother website and downloaded the updates, but still no cigar!
So I brought in the previous colour laser printer, an HP
CP1025nw, which I had stored in the large Brother cardboard box to keep it in good condition to use as a backup.
It was retired in good condition and with plenty of ink,
but it was a bit slower than the Brother and sometimes
had problems grabbing paper when I was doing big runs
for my labels.
When I lifted the box down, apart from a huge friendly
Huntsman spider that jumped out, I noticed that the
Brother box had a picture of the items included with the
siliconchip.com.au
Australia's electronics magazine
August 2025 87
Silicon Chip kcaBBack Issues
$10.00 + post
$11.50 + post
$12.50 + post
$13.00 + post
January 1997 to October 2021
November 2021 to September 2023
October 2023 to September 2024
October 2024 onwards
All back issues after February 2015 are in stock, while most from January
1997 to December 2014 are available. For a full list of all available issues,
visit: siliconchip.com.au/Shop/2
PDF versions are available for all issues at siliconchip.com.au/Shop/12
We also sell photocopies of individual articles for those who don’t have a computer
As luck would have it, I accidentally pressed “print” via
the USB connection and bingo! It printed OK.
I was just about to pack up the backup printer back
into its snug box when I noticed that a document failed
to print and I got the same error message as I received on
day one. Why?
Checking the cable connections again, I noticed that
the particular cable I was using looked a bit old and had
square and faded plugs at both ends, so I thought it was a
good idea to replace it. Everything then worked perfectly!
Replacing the old cable, the fault returned. How could I
have not done this before?
I think it is a lesson that we think that USB cables are bulletproof, even though we jerk them in and out and stretch
them when we move computers and printers around. My
daughter has wrecked so many chargers by pulling USB
connectors out at funny angles, and so it seems my printer
cable that had been used for years on five or six different
printers ultimately suffered the same fate.
My initial analysis of the situation removed any thoughts
about a simple cable, because my previous experience was
that computers have computer problems and printers have
printer problems and the humble USB cable was too humble to worry about!
These days, when you buy a new printer, the USB cable
is rarely included. I think in future I will buy brand new
ones instead of cheap old ones from op shops!
Allan Linton-Smith, Turramurra, NSW.
Decorative village repair
A friend asked me if I could look at a family heirloom
that stopped working. It is a ‘village’ house with many
decorative lights. A light source shines through a colour
wheel driven by a 12V AC motor. The light is diffused onto
many fibre light pipes scattered throughout the display. It’s
simple and effective.
Someone in the past had ‘fixed’ it by shoehorning a 20W
halogen lamp into the house; the original lamp was an 8W
MR8 halogen lamp, with it and the motor powered by a
12V AC 1.25A plugpack.
The overloaded plug pack eventually failed. AC plugpacks are not very common, and halogen lamps were phased
out years ago! After many phone calls, then 45km of suburban traffic, I sourced both original rated items from two
widely spaced dealers who had old supplies. The lamp
was the MR11 size and would fit OK.
Later, I sourced an MR11 LED replacement lamp from
Bunnings and fitted a diode in series with it (hidden inside a
red sleeve) to convert the AC to DC. The diode also reduces
the power to the lamp for a longer life with little effect on
the light output for the display.
A good deed by the non-technical resulted in hours of
time and parts to repair!
Victor Duffey, Rosanna, Vic.
HP410C voltmeter repair
I am an avid collector of old HP and Tektronix test equipment, so when a friend offered me their HP410C in exchange
for some Marconi RF coil standards, I jumped at the deal.
It was in fantastic physical condition, but had some problems that deserved my ministrations...
Firstly, it was inaccurate; the meter reading was either
too low or too high depending on what range I set it to. I
tried a quick calibration, which highlighted the many problems that I was about to find.
One of the main PSU capacitors was bad. It was a rather
large 2400μF 20V part. I replaced it with a 4700μF Kemet
electrolytic capacitor, which mounted on one of my oval
capacitor adaptor PCBs for retro work, as the original ‘big
can’ type is not cheaply available anymore. This capacitor filters the 6V rail that supplies the heater to the probe
valve diode, among other things.
At this stage, I discovered the big T03 germanium series
pass transistor was worn out to the point that it measured
as two leaky diodes! Luckily, Rockby had 2N1544s on sale
a while ago, which turned out to be the exact equivalent of
the original Motorola 1850-0098 PNP germanium transistor.
Shown at left is the decorative village, and the original MR8 lamp is visible in the adjacent photo.
88
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
The last part I replaced in the PSU was a 38V zener diode
which decided to avalanche at 32V these days (I guess 50
years is a bit much to ask).
At this point, I thought I was really getting somewhere.
However, there were three other problems. The first was
simple: a 100W resistor at the AC probe telephone-style plug
that sits across the heater rail was burnt out and making
intermittent contact, causing the needle to behave erratically when measuring AC voltages. When this was replaced,
the 410C settled down nicely.
Next, FSD (full-scale deflection) on the lowest of the
three ranges was impossible to get right. I found a 6MW
resistor with a tolerance of 0.5% on the attenuator switch
that was reading 6.3MW. Luckily, I had a junk 410C with
one that was OK, as I don’t think a 6MW resistor was in
my parts box. Now I was getting really close, but the most
interesting repair was to come!
The final hurdle was the moving meter itself. As some
people know, HP eventually made their own meters, and
they were individually calibrated by having someone print
the label on each and every face when testing the meter
movement.
This incredible feat of engineering, combined with the
taut-band movement, made HP simply the best money
could buy; even AVO meters didn’t come close, as they
just selected from several differently scaled faces to ‘best
suit’ the chosen movement.
This meter had lost magnetism in the ‘permanent’ magnet. When I should have gotten an FSD reading at 1mA, I
only received about 85% of full scale.
Due to this, and possible slackening of the taut band,
the other readings were well off. So there was no way to
make it work by calibrating the 410C to this particular
meter movement.
I decided to have a go at re-magnetising the meter. I carefully disassembled the inner workings of the moving coil
meter and removed the magnetic core. I wound this with
several turns of thick multi-strand wire and then shorted
this across a large 12V AGM battery. However, that was
not enough to make an improvement.
I went for broke and got a second 12V AGM battery and
put them in series for 24V DC at considerable current; I estimate that the surges were over 100A! This had the desired
effect, and the permanent magnet was now strong enough
to make the needle read FSD <at> 1mA.
There was still the small detail of the meter now being
different enough to not line up with the previous graduation markings. I decided to calibrate this in a similar way
to the HP of old by connecting my current calibrator to the
meter and running it at 0.1mA, 0.2mA, 0.3mA all the way
to 1mA. At each point, I marked the meter face to show
where the graduations should be.
I handed this to a friend who knows far more about vector diagrams than myself; he created a new meter scale to
accurately reflect the current calibration of the meter. This
was printed on a quality vinyl sticker and placed over the
original face.
After that, the calibration went smoothly, with all attenuation scales are bang-on, and the unit cleaned up like
new. Yes, it was a lot of work, but very interesting at the
same time. It’s quite satisfying to have repaired a classic
and still very usable meter.
SC
Deon Vandenberg, Torquay, Qld.
siliconchip.com.au
Australia's electronics magazine
August 2025 89
SOnline
ilicon Chip
Shop
Kits, parts and much more
www.siliconchip.com.au/Shop/
Rotating Lights
April 2025
Dual Mini LED Dice
August 2024
USB Power Adaptors
May 2025
SMD LED Complete Kit
SC7462: $20
TH LED Complete Kit
SC7463: $20
SMD LED Complete Kit
SC6961: $17.50
TH LED Complete Kit
SC6849: $17.50
siliconchip.au/Article/16418
siliconchip.au/Article/18112
This kit includes everything needed to build
the Rotating Light for Models, except for a
power supply and wire.
Includes either 3mm through-hole or 1206sized SMD LEDs. Choice of either white or
black PCB. CR2032 coin cell not included.
You can choose from one of four USB sockets
(USB-C power only, USB-C power+data, mini-B
or micro-B). The kit includes all other parts.
siliconchip.au/Article/17930
Compact HiFi Headphone Amplifier
Complete Kit
SC6885: $70
Complete Kit
with choice of USB socket
SC7433: $10
Capacitor Discharger
December 2024
December 2024
& January 2025
siliconchip.au/Series/432
This kit includes everything required to build the Compact HiFi Headphone Amplifier. The case is
included, but you will need your own power supply.
Programmable Frequency Divider
Complete Kit
SC6959: $60
Feb25: siliconchip.au/Article/17733
Includes all onboard components, except for a power supply and the optional programming header.
Short-Form Kit
SC7404: $30
siliconchip.au/Article/17310
Includes the PCB, resistors, semis, mounting
hardware and banana sockets. Case, heatsink,
thermal switch and wiring are not supplied.
→ Subscribers receive a 10% discount on all purchases, except for subscriptions (postage is not discounted).
→ Prices listed do not include postage. Postage rates within Australia start at $12, rates are calculated at the checkout.
Vintage Radio
Silvertone Model 18 AM/FM mantel
radio from 1952
The Silvertone model 18
is an excellent example
of a 1952 US radio.
The plastic case shows
some heritage from the
Bakelite era. From left-toright, the knob controls
are for volume, AM-FM
selection and tuning.
By Associate Professor Graham Parslow
M
any Bakelite radios of similar
appearance were made in the
1940s. However, thermo-mouldable
plastics, which were new in the early
1950s, were significantly cheaper and
faster to produce.
The old celluloid dial covers
that degraded to become brittle and
opaque were replaced by clear polystyrene that is naturally transparent.
One downside to polystyrene is that
organic solvents, including acetone,
can degrade the surface, leaving permanent splotches.
The dial cover on this radio is
in excellent condition after seven
decades. Brass and gold-tone features
were common on US radios at this
time, as embodied by the dial cursor,
the stylistic “S” and the knobs.
US dials were calibrated by frequency, not call signs, due to the sheer
number of stations that would clutter
the dial.
Transmissions at the same frequency should not be offered as the
reason. There is no problem with
putting multiple stations at the same
frequency on the dial because they
would be at considerable geographical separation.
Silvertone is a house brand
This label attached to the interior of the Silvertone 18 case has company
information, the serial number and some basic servicing guidelines for the set.
siliconchip.com.au
Australia's electronics magazine
The label glued to the bottom identifies the radio as Silvertone catalog number 18, October 1952. The
list price was US$37.95. Sears, Roebuck and Co is an American chain of
department stores founded in 1892
by Richard Sears and Alvah Roebuck.
The company began as a mail-order
catalogue company, progressing to
retail locations from 1925, beginning
in Chicago.
The 110-storey Sears Tower in Chicago (now known as the Willis Tower)
was the tallest in the world in 1974.
Sears filed for Chapter 11 bankruptcy
in October 2018, but a restructure
allowed them to continue trading at a
reduced scale.
August 2025 91
Fig.1: the FM tuner is across the top of the
circuit diagram, with the AM section below
it and the power supply at lower middle. A
3PDT switch selects between FM and AM
modes; it also switches valve V4 so that
it operates as an FM IF amplifier in FM
mode and an AM IF amplifier in AM mode.
Capacitor C38 allows the mains cord to
operate as an FM antenna for strong stations.
Sears and Roebuck had used the Silvertone brand going back to the 1930s.
Howard W. Sams & Co were the radio
manufacturers in this case. Likewise,
in Australia, Myer stores contracted
manufacturers of convenience to produce the in-house Aristone branded
radios.
Commendably, this radio has the
circuit pasted onto the side of the
92
Silicon Chip
case. The service notes provided by
the manufacturer can be downloaded
from siliconchip.au/link/ac1x
From the ten pages of impressively
detailed service notes, one page is
shown here that itemises components
on the top of the chassis. The photo
of the top of the chassis also shown
overleaf shows the tuning capacitor
shield in place.
Australia's electronics magazine
Another page of the documentation
shows the itemisation of components
beneath the chassis. I have not seen
documentation from any Australian
radio manufacturer of the period as
comprehensive as this.
The history of FM radio in the
United States
Edwin Armstrong (born December
siliconchip.com.au
18, 1890) served with the US army
in France during WW1. During this
period, he developed the superheterodyne receiver system. The superheterodyne radio shifts the high-
frequency radio signal of interest to a
lower ‘intermediate’ frequency. The
original aim was to get the frequency
down to a range better suited to amplification by early triode valves.
siliconchip.com.au
It had the serendipitous effect of
achieving precise tuning at the broadcast frequency, that otherwise needed
a cascade of tuned circuits (called TRF
circuits, TRF standing for tuned radio
frequency).
In this radio’s circuit (Fig.1), that
mixing is performed by ‘converter’
valve V3 (a 6BE6 heptode). It has two
control grids, one connected to pin 1
Australia's electronics magazine
and one at pin 7. The incoming signal, tuned by variable capacitor A6
and the secondary of transformer L8,
is applied to pin 7.
At the same time, a transformer-
coupled oscillator circuit, developed
by Armstrong and named after him,
acts as the local oscillator, which
tracks at a higher frequency than the
tuned signal (a fixed interval above).
August 2025 93
The oscillator is formed by transformer L9 and tuning capacitor A5,
and its output is applied to the control grid at pin 1. Feedback to maintain oscillation comes from the valve’s
cathode, at pin 2.
The alternative Hartley oscillator
is cheaper, using a coil with only one
tapped winding, but the Armstrong
oscillator has proved highly reliable
and is more commonly chosen.
The output at the anode (pin 5) contains the amplified RF signal, oscillator
signal, plus their sum and difference
products. It is the difference product,
at the intermediate frequency, that
passes through the following tuned
stages to ultimately be demodulated
to produce audio.
Introducing FM radio
Armstrong began developing FM
(frequency modulation) based radio
in 1928, and argued strenuously for
its adoption to replace AM. The time
was not ripe, and early attempts to
commercialise FM in 1941 in the USA
faltered, in part due to allocating only
40 channels spanning 42–50MHz.
In 1945, the FM band was reassigned to 88–106MHz. By 1952, limited FM stations were transmitting,
so most buyers would pick a cheaper
AM-only radio. Stereo FM broadcasting in the US took off in the 1960s,
and by the 1970s became the dominant music source, relegating many
AM stations to talk-only programs.
FM has the virtue of rejecting the
majority of the electromagnetic interference (EMI) that plays havoc with
AM transmissions. As a result, the
signal-to-noise ratios are far superior
to AM. Another virtue is high fidelity,
transmitting the full range of human
hearing, due in part to greater frequency separation between stations.
Unfortunately, the 5-inch (127mm)
speaker in this radio does little justice
to the potential for high fidelity.
Power supply
The top view of the Silvertone 18 chassis and a matching diagram
which has every component (on this side) labelled. Note the large tuning
capacitor shield, which was removed in the diagram.
94
Silicon Chip
Australia's electronics magazine
US radios of the period were commonly transformerless, using serieswired valve heaters with the high-
tension rectified directly from the
mains. Mains-direct radios are hazardous to work with, particular if an
auto-transformer (variac) is used. Happily, this radio uses a transformer to
interface with the US mains that is
nominally 117V.
This radio was restored using
a step-down transformer. Fortunately, the transformer on this
radio was substantial enough to
not heat up excessively with a
50Hz source rather than 60Hz.
The mains supply in the
USA is nominally 110120V at 60Hz. 220-240V
is also available for
larger appliances, due
to there being two out-ofphase 110-120V conductors
in the grid.
Because of the lower frequency of our mains, even if
the voltage is adapted using a
step-down transformer, some US
equipment will not be happy running at 50Hz.
There is typically less iron in their
transformer cores, as less is required
given the higher operating frequency.
siliconchip.com.au
They can therefore saturate at lower
than expected currents and overheat
when running from 50Hz mains.
Thankfully, the transformer in this
set seems to have a generous core and
that did not appear to be a problem
during my testing, with the transformer remaining cool.
The rear of the radio shows the
AM loop antenna with a threescrew connection strip below. The
connections are for an external
AM antenna, external FM antenna
and Earth. In my location, the loop
antenna was all that was needed for
AM.
The AM receiver
The tuning capacitor, designated
A6, tunes from 540kHz to 1600kHz,
while tuning capacitor A5 tunes the
oscillator from 995kHz to 2055kHz
to generate a 455KHz intermediate
frequency (IF). As mentioned earlier,
the IF signal emanates from the anode
of the 6BE6 mixer valve (V3).
The first IF transformer is A3/A4
(L12), delivering the IF signal to V4, a
6BA6 IF amplifier. The second IF transformer, A1/A2 (L14), passes its output
to detector valve V6, a 6T8. When the
switch is set to AM, detected audio
passes via L14 to the volume control
potentiometer, R1.
The dial for this radio is printed onto a metal
sheet, and is in excellent condition given its age.
The FM receiver
The FM section can work on high
signal stations without an external
aerial due to coupling the RF input to
the mains lead via 100pF capacitor C38
(ie, the mains lead acts as an antenna
if the station is strong enough).
The untuned RF signal is amplified
by V1, a 6BA6. The desired signal in
the 88-108MHz band is tuned by variable capacitor A13 and heterodyned
with the output from the oscillator,
tuned by A12 and L6. The intermediate frequency is 10.7MHz, so the
oscillator tracks 10.7MHz above the
tuned RF signal.
The first half of the 12AT7 (V2) is
the converter, with the RF signal fed to
the grid (pin 1). The same grid receives
local oscillator input generated by the
A12 cluster on the circuit diagram.
The second half of the 12AT7 is an IF
amplifier.
V4 is an additional IF amplifier
that does double service as an AM IF
amplifier, depending on the band that
is selected. The 6BA6 designated V5 is
another IF amplifier. V6, the 6T8 ratio
siliconchip.com.au
The back view of the chassis shows the loop antenna and the external
antenna terminal (the yellow wire). The second terminal is for an optional
FM antenna and the third is Earth.
detector, generates detected audio at
R21 that passes via the FM selector
switch to the volume control pot, R1.
A ratio detector has two diodes conducting in opposite directions connected to a centre-tapped transformer
secondary. In this case, both diodes are
within a single envelope (V6). Ratio
detectors have the significant advantage for FM demodulation that they
do not respond to AM signals, making
them more resistant to interference.
Australia's electronics magazine
If this was a full-wave rectifier, the
output would be one polarity, but a
ratio detector passes positive signals
at one diode and negative at the other.
The output is the sum of the diode voltages and the centre tap voltage. The
output signal from across the diodes
is filtered by a high-value capacitor, C3
(4µF) in this radio. It is bled to Earth
by R20 (1.5kW).
The combination of the capacitor’s opposition to voltage changes
August 2025 95
The other two 6BA6s also measured
as open circuits on their heaters. Previously swapping 6BA6s within the
radio had done nothing; it now became
clear why. It was shades of the movie
True Lies – they were all bad!
Inserting two replacement 6BA6
valves into the FM section instantly
produced FM reception. Why would
three identical valves all fail? My best
guess is that some transient surge blew
the most vulnerable heater filaments
in the 6.3V AC line. Perhaps someone
connected the radio to 230-240V.
There was a crack in the case that was repaired using two-part epoxy car filler.
This photo was taken before the epoxy was refined using an angle grinder.
and the resistive loading produces a
nearly constant amplitude for the output. This action gives FM its superior
immunity to electrical interference.
The set applies AGC for AM operation as usual, but it also applies AGC
in FM operation. FM usually relies
on the last IF stage being driven into
overdrive and acting as a limiter to
deliver a constant-amplitude signal
to the ratio detector.
The use of AGC implies that the last
stage does not provide limiting for all
FM signal levels, so it needs the AGC
to provide the same volume for all
stations. For FM, the audio output is
converted to a DC level by 220kW resistor R23 and 5mF capacitor C17 and is
used to bias the input signal and the
signal applied to the first IF amplifier,
V4. That path is disabled when V4 is
used as an AM IF amp.
(the final working power was 49W). So
something was not drawing (enough)
power.
Valve V4 (6BA6) is common to both
the AM and FM functions. With AM
selected, injecting a 400Hz-modulated
signal at the IF frequency of 455kHz to
the grid produced nothing. However,
a signal to the anode passed through
as 400Hz audio. So it was a matter of
working backwards to find the fault.
Fortunately, it soon became evident
what the problem was.
V4’s 10kW screen resistor had
0V across it, as did its 68W cathode
resistor. Everything indicated a non-
conductive valve V4. This is classic
for a non-functional cathode heater.
Sure enough, its heater pins 3 and 4
were open-circuit. Replacing V4 from
my stock got the AM function working, but not FM.
Case restoration
Examination of the crack at the bottom of the right-hand side revealed
that a substantial chunk was missing.
I fixed this gap using two-part epoxy
car filler.
To achieve this, I covered a section
of aluminium sheet with a 90° flange
in masking tape to make removal of the
former easy after the epoxy had set. I
laid the radio on its side and filled the
gap coarsely with epoxy. The external
surface of the radio had the epoxy set
flat by conforming to the former.
Once it had hardened, I used an
angle grinder to profile the inside
to size. I then painted over the pink
epoxy with satin black paint, and it
effectively disappeared.
Conclusion
The Silvertone 18 is a high-quality
set with good performance. Its style is
of its time, but offering FM was defiSC
nitely ahead of its time.
All three
6BA6s in the
set had opencircuit heaters.
Replacing them
was all I had to
do to get the set
working.
The audio section
After the volume control, a 6T8 triode acts as a preamplifier for the 6V6
beam tetrode, V7. This 6T8 triode is
actually part of V6; the 6T8 encapsulates one triode and three diodes.
The 5-inch (127mm) speaker has an
impedance of 3.5W. There is no negative feedback from the speaker transformer secondary, so the distortion
due to the transformer is not reduced.
Electrical restoration
When I got it, the radio was dead. I
ruled out the usual causes of complete
failure in my preliminary assessment.
The AM/FM switch checked out OK
and the audio circuitry worked from
the slider on the volume pot. The
valves were all well-seated and HT was
good at 201V/182V across the π filter.
The initial power draw was 39W
96
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
R&S®ZNB3000 Vector Network Analyzer
FAST FORWARD TO RESULTS
The R&S®ZNB3000 is the instrument you need for RF component production. This latest addition
to the Rohde & Schwarz network analyzer portfolio offers best-in-class RF performance, combining
high measurement accuracy with exceptional speed. With its high throughput rate, it is especially
suitable for high-volume production and short ramp-up time environments.
For more information visit: www.rohde-schwarz.com/solution/ZNB3000
SILICON
CHIP
.com.au/shop
ONLINESHOP
HOW TO ORDER
INTERNET (24/7)
PAYPAL (24/7)
eMAIL (24/7)
MAIL (24/7)
PHONE – (9-5:00 AET, Mon-Fri)
siliconchip.com.au/Shop
silicon<at>siliconchip.com.au
silicon<at>siliconchip.com.au
PO Box 194, MATRAVILLE, NSW 2036
(02) 9939 3295, +612 for international
You can also pay by cheque/money order (Orders by mail only) or bank transfer. Make cheques payable to Silicon Chip.
08/25
YES! You can also order or renew your Silicon Chip subscription via any of these methods as well!
The best benefit, apart from the magazine? Subscribers get a 10% discount on all orders for parts.
PRE-PROGRAMMED MICROS
For a complete list, go to siliconchip.com.au/Shop/9
$10 MICROS
$15 MICROS
ATmega328P
ATtiny45-20PU
PIC10LF322-I/OT
PIC12F617-I/P
110dB RF Attenuator (Jul22), Basic RF Signal Generator (Jun23)
2m VHF CW/FM Test Generator (Oct23)
Range Extender IR-to-UHF (Jan22)
Active Mains Soft Starter (Feb23), Model Railway Uncoupler (Jul23)
Battery-Powered Model Railway Transmitter (Jan25)
PIC12F675-I/P
Train Chuff Sound Generator (Oct22)
PIC12F675-I/SN
Tiny LED Xmas Tree (Nov19)
PIC16F1455-I/P
Railway Points Controller Transmitter / Receiver (2 versions; Feb24)
Battery-Powered Model Railway TH Receiver (Jan25)
PIC16F1455-I/SL Battery Multi Logger (Feb21), USB-C Serial Adaptor (Jun24)
Battery-Powered Model Railway SMD Receiver (Jan25)
USB Programmable Frequency Divider (Feb25)
PIC16LF1455-I/P New GPS-Synchronised Analog Clock (Sep22)
PIC16F1459-I/P
K-Type Thermostat (Nov23), Secure Remote Switch (RX, Dec23)
Mains Power-Up Sequencer (Feb24 | repurposed firmware Jul24)
8CH Learning IR Remote (Oct24), Heat Transfer Controller (Aug25)
PIC16F1459-I/SO Multimeter Calibrator (Jul22), Buck/Boost Charger Adaptor (Oct22)
PIC16F15214-I/SN Silicon Chirp Cricket (Apr23), Mic The Mouse (Aug25)
PIC16F15214-I/P Filament Dryer (Oct24), Tool Safety Timer (May25)
PIC16F15224-I/SL Multi-Channel Volume Control (OLED Module; Dec23)
NFC IR Keyfob Transmitter (Feb25), Rotating Light (Apr25)
PIC16F18146-I/SO Compact OLED Clock & Timer (Sep24), Flexidice (Nov24)
Versatile Battery Checker (May25), RGB LED ‘Analog’ Clock (May25)
USB-C Power Monitor (Aug25)
PIC16LF15323-I/SL Remote Mains Switch (TX, Jul22), Secure Remote Switch (TX, Dec23)
STM32G030K6T6 Variable Speed Drive Mk2 (Nov24)
W27C020
Noughts & Crosses Computer (Jan23)
ATSAML10E16A-AUT
PIC16F1847-I/P
PIC16F18877-I/PT
High-Current Battery Balancer (Mar21)
Digital Capacitance Meter (Jan25)
Dual-Channel Breadboard PSU Display Adaptor (Dec22)
Wideband Fuel Mixture Display (WFMD; Apr23)
PIC16F88-I/P
Battery Charge Controller (Jun22), Railway Semaphore (Apr22)
PIC24FJ256GA702-I/SS
Ohmmeter (Aug22), Advanced SMD Test Tweezers (Feb23)
ESR Test Tweezers (Jun24)
PIC32MX170F256D-501P/T 44-pin Micromite Mk2 (Aug14), 4DoF Simulation Seat (Sep19)
PIC32MX170F256B-50I/SP Micromite LCD BackPack V1-V3 (Feb16 / May17 / Aug19)
Advanced GPS Computer (Jun21), Touchscreen Digital Preamp (Sep21)
PIC32MX170F256B-I/SO
Battery Multi Logger (Feb21), Battery Manager BackPack (Aug21)
PIC32MX270F256B-50I/SP ASCII Video Terminal (Jul14), USB M&K Adaptor (Feb19)
STM32L031F6P6
SmartProbe (Jul25)
$20 MICROS
ATmega32U4
ATmega644PA-AU
PIC32MK0128MCA048
Wii Nunchuk RGB Light Driver (Mar24)
AM-FM DDS Signal Generator (May22)
Power LCR Meter (Mar25)
$25 MICROS
PIC32MX170F256B-50I/SO + PIC16F1455-I/SL
Micromite Explore-40 (SC5157, Oct24)
PIC32MX470F512H-120/PT Micromite Explore 64 (Aug 16), Micromite Plus (Nov16)
PIC32MX470F512L-120/PT Micromite Explore 100 (Sep16)
$30 MICROS
PIC32MX695F512H-80I/PT Touchscreen Audio Recorder (Jun14)
PIC32MZ2048EFH064-I/PT DSP Crossover/Equaliser (May19), Low-Distortion DDS (Feb20)
DIY Reflow Oven Controller (Apr20), Dual Hybrid Supply (Feb22)
KITS, SPECIALISED COMPONENTS ETC
MIC THE MOUSE KIT (SC7508)
(AUG 25)
RP2350B DEVELOPMENT BOARD
(AUG 25)
USB-C POWER MONITOR KIT (SC7489)
(AUG 25)
433MHz RECEIVER KIT (SC7447)
(JUN 25)
VERSATILE BATTERY CHECKER KIT (SC7465)
(MAY 25)
RGB LED ‘ANALOG’ CLOCK KIT (SC7416)
(MAY 25)
USB POWER ADAPTOR COMPLETE KIT (SC7433)
(MAY 25)
Includes all parts except a CR2032 cell (see p64, Aug25)
Assembled Board: a pre-assembled PCB with all mandatory parts fitted,
optional components are sold separately below (SC7514; see p49, Aug25)
- 40-pin header (two are required, SC3189)
- 8MiB APS6404L-3SQR-SN PSRAM SOIC-8 IC (SC7530)
Includes all non-optional parts except the case, cell & glue (see p39, Aug25)
Includes the PCB and all onboard parts (see p66, Jun25)
Includes everything in the parts list (including the case), except the optional
components, batteries and glue (see p30, May25)
$37.50
$60.00
$20.00
$65.00
Includes everything in the parts list and a choice of one USB socket: USB-C power only;
USB-C power+data; Type-B mini; or Type-B micro (see p80, May25)
$10.00
(APR 25)
Includes an assembled PCB, separate Raspberry Pi Pico 2 and front/rear panels $120.00
ROTATING LIGHT FOR MODELS KIT
(APR 25)
Complete kit which includes the PCB and all onboard components (see p60, Apr25):
- SMD LEDs (SC7462)
$20.00
- Through-hole LEDs (SC7463)
$20.00
433MHz TRANSMITTER KIT (SC7430)
Includes the PCB and all onboard parts (see p75, Apr25)
PICO 2 AUDIO ANALYSER SHORT-FORM KIT (SC6772)
(APR 25)
(MAR 25)
The Pico Audio Analyser kit from Nov23, but with an unprogrammed Pico 2
USB PROGRAMMABLE FREQUENCY DIVIDER (SC6959)
(FEB 25)
NFC PROGRAMMABLE IR KEYFOB (SC7421)
(FEB 25)
CAPACITOR DISCHARGER KIT (SC7404)
(DEC 24)
PICO COMPUTER
(DEC 24)
FLEXIDICE COMPLETE KIT (SC7361)
(NOV 24)
MICROMITE EXPLORE-40 KIT (SC6991)
(OCT 24)
DUAL-RAIL LOAD PROTECTOR (SC7366)
(OCT 24)
PicoMSA PARTS (SC7323)
(SEP 24)
COMPACT OLED CLOCK & TIMER KIT (SC6979)
(SEP 24)
Complete kit: includes all components (see p85, Feb25)
$60.00
Complete kit: includes all required items, except the cell (see p67, Feb25)
$25.00
$30.00
(DEC 24)
$1.00ea COMPACT HIFI HEADPHONE AMP (SC6885)
Complete kit: includes everything except the power supply (see p47, Dec24)
$70.00
$5.00
Includes all the parts except the power supply. When buying the kit select either a BZ-121
GPS module or Pico W (unprogrammed) for the time source (see p66, May25)
$65.00
PICO/2/COMPUTER (SC7468)
siliconchip.com.au/Shop/
$20.00
$50.00
Includes the PCB and all components that mount on it, the mounting hardware
(without heatsink) and banana sockets (see p36, Dec24)
$30.00
For full functionality both the Pico Computer Board and Digital Video Terminal kits are
required. Items shown unbolded are optional (see p71, Dec24)
- Pico Computer Board kit (SC7374)
$40.00
- Pico Digital Video Terminal kit (SC6917)
$65.00
- PWM Audio Module kit (SC7376)
$10.00
- ESP-PSRAM64H 64Mb SPI PSRAM chip (SC7377)
$5.00
- DS3231 real-time clock SOIC-16 IC (SC5103)
$7.50
- DS3231MZ real-time clock SOIC-8 IC (SC5779)
$10.00
Includes all required parts except the coin cell (see p71, Nov24)
Includes all required parts (see p83, Oct24)
Hard-to-get parts: includes the PCB and all semiconductors except the
optional/variable diodes (see p73, Oct24)
Hard-to-get parts: includes the PCB, Raspberry Pi Pico (unprogrammed),
plus all semiconductors, capacitors and resistors (see p63, Sep24)
Includes everything except the case & Li-ion cell (see p34, Sep24)
*Prices valid for month of magazine issue only. All prices in Australian dollars and include GST where applicable. # Overseas? Place an order on our website for a quote.
$30.00
$35.00
$35.00
$50.00
$45.00
PRINTED CIRCUIT BOARDS & CASE PIECES
PRINTED CIRCUIT BOARD TO SUIT PROJECT
Q METER MAIN PCB
↳ FRONT PANEL (BLACK)
NOUGHTS & CROSSES COMPUTER GAME BOARD
↳ COMPUTE BOARD
ADVANCED SMD TEST TWEEZERS SET
DIGITAL VOLUME CONTROL POT (SMD VERSION)
↳ THROUGH-HOLE VERSION
MODEL RAILWAY TURNTABLE CONTROL PCB
↳ CONTACT PCB (GOLD-PLATED)
WIDEBAND FUEL MIXTURE DISPLAY (BLUE)
TEST BENCH SWISS ARMY KNIFE (BLUE)
SILICON CHIRP CRICKET
GPS DISCIPLINED OSCILLATOR
SONGBIRD (RED, GREEN, PURPLE or YELLOW)
DUAL RF AMPLIFIER (GREEN or BLUE)
LOUDSPEAKER TESTING JIG
BASIC RF SIGNAL GENERATOR (AD9834)
↳ FRONT PANEL
V6295 VIBRATOR REPLACEMENT PCB SET
DYNAMIC RFID / NFC TAG (SMALL, PURPLE)
↳ NFC TAG (LARGE, BLACK)
RECIPROCAL FREQUENCY COUNTER MAIN PCB
↳ FRONT PANEL (BLACK)
PI PICO-BASED THERMAL CAMERA
MODEL RAILWAY UNCOUPLER
MOSFET VIBRATOR REPLACEMENT
ARDUINO ESR METER (STANDALONE VERSION)
↳ COMBINED VERSION WITH LC METER
WATERING SYSTEM CONTROLLER
CALIBRATED MEASUREMENT MICROPHONE (SMD)
↳ THROUGH-HOLE VERSION
SALAD BOWL SPEAKER CROSSOVER
PIC PROGRAMMING ADAPTOR
REVISED 30V 2A BENCH SUPPLY MAIN PCB
↳ FRONT PANEL CONTROL PCB
↳ VOLTAGE INVERTER / DOUBLER
2M VHF CW/FM TEST GENERATOR
TQFP-32 PROGRAMMING ADAPTOR
↳ TQFP-44
↳ TQFP-48
↳ TQFP-64
K-TYPE THERMOMETER / THERMOSTAT (SET; RED)
MODEM / ROUTER WATCHDOG (BLUE)
DISCRETE MICROAMP LED FLASHER
MAGNETIC LEVITATION DEMONSTRATION
MULTI-CHANNEL VOLUME CONTROL: VOLUME PCB
↳ CONTROL PCB
↳ OLED PCB
SECURE REMOTE SWITCH RECEIVER
↳ TRANSMITTER (MODULE VERSION)
↳ TRANSMITTER (DISCRETE VERSION
COIN CELL EMULATOR (BLACK)
IDEAL BRIDGE RECTIFIER, 28mm SQUARE SPADE
↳ 21mm SQUARE PIN
↳ 5mm PITCH SIL
↳ MINI SOT-23
↳ STANDALONE D2PAK SMD
↳ STANDALONE TO-220 (70μm COPPER)
RASPBERRY PI CLOCK RADIO MAIN PCB
↳ DISPLAY PCB
KEYBOARD ADAPTOR (VGA PICOMITE)
↳ PS2X2PICO VERSION
MICROPHONE PREAMPLIFIER
↳ EMBEDDED VERSION
RAILWAY POINTS CONTROLLER TRANSMITTER
↳ RECEIVER
LASER COMMUNICATOR TRANSMITTER
↳ RECEIVER
PICO DIGITAL VIDEO TERMINAL
↳ FRONT PANEL FOR ALTRONICS H0190 (BLACK)
↳ FRONT PANEL FOR ALTRONICS H0191 (BLACK)
ARDUINO FOR ARDUINIANS (PACK OF SIX PCBS)
↳ PROJECT 27 PCB
DATE
JAN23
JAN23
JAN23
JAN23
FEB23
MAR23
MAR23
MAR23
MAR23
APR23
APR23
APR23
MAY23
MAY23
MAY23
JUN23
JUN23
JUN23
JUN23
JUL23
JUL23
JUL23
JUL23
JUL23
JUL23
JUL23
AUG23
AUG23
AUG23
AUG23
AUG23
SEP23
SEP23
SEP23
OCT22
SEP23
OCT23
OCT23
OCT23
OCT23
OCT23
NOV23
NOV23
NOV23
NOV23
DEC23
DEC23
DEC23
DEC23
DEC23
DEC23
DEC23
DEC23
DEC23
DEC23
DEC23
DEC23
DEC23
JAN24
JAN24
JAN24
JAN24
FEB24
FEB24
FEB24
FEB24
MAR24
MAR24
MAR24
MAR24
MAR24
MAR24
MAR24
PCB CODE
CSE220701
CSE220704
08111221
08111222
SC6658
01101231
01101232
09103231
09103232
05104231
04110221
08101231
04103231
08103231
CSE220602A
04106231
CSE221001
CSE220902B
18105231/2
06101231
06101232
CSE230101C
CSE230102
04105231
09105231
18106231
04106181
04106182
15110231
01108231
01108232
01109231
24105231
04105223
04105222
04107222
06107231
24108231
24108232
24108233
24108234
04108231/2
10111231
SC6868
SC6866
01111221
01111222
01111223
10109231
10109232
10109233
18101231
18101241
18101242
18101243
18101244
18101245
18101246
19101241
19101242
07111231
07111232
01110231
01110232
09101241
09101242
16102241
16102242
07112231
07112232
07112233
SC6903
SC6904
Price
$5.00
$5.00
$12.50
$12.50
$10.00
$2.50
$5.00
$5.00
$10.00
$10.00
$10.00
$5.00
$5.00
$4.00
$2.50
$12.50
$5.00
$5.00
$5.00
$1.50
$4.00
$5.00
$5.00
$5.00
$2.50
$2.50
$5.00
$7.50
$12.50
$2.50
$2.50
$10.00
$5.00
$10.00
$2.50
$2.50
$5.00
$5.00
$5.00
$5.00
$5.00
$10.00
$2.50
$2.50
$5.00
$5.00
$5.00
$3.00
$5.00
$2.50
$2.50
$5.00
$2.00
$2.00
$2.00
$1.00
$3.00
$5.00
$12.50
$7.50
$2.50
$2.50
$7.50
$7.50
$5.00
$2.50
$5.00
$2.50
$5.00
$2.50
$2.50
$20.00
$7.50
For a complete list, go to siliconchip.com.au/Shop/8
PRINTED CIRCUIT BOARD TO SUIT PROJECT
WII NUNCHUK RGB LIGHT DRIVER (BLACK)
SKILL TESTER 9000
PICO GAMER
ESP32-CAM BACKPACK
WIFI DDS FUNCTION GENERATOR
10MHz to 1MHz / 1Hz FREQUENCY DIVIDER (BLUE)
FAN SPEED CONTROLLER MK2
ESR TEST TWEEZERS (SET OF FOUR, WHITE)
DC SUPPLY PROTECTOR (ADJUSTABLE SMD)
↳ ADJUSTABLE THROUGH-HOLE
↳ FIXED THROUGH-HOLE
USB-C SERIAL ADAPTOR (BLACK)
AUTOMATIC LQ METER MAIN
AUTOMATIC LQ METER FRONT PANEL (BLACK)
180-230V DC MOTOR SPEED CONTROLLER
STYLOCLONE (CASE VERSION)
↳ STANDALONE VERSION
DUAL MINI LED DICE (THROUGH-HOLE LEDs)
↳ SMD LEDs
GUITAR PICKGUARD (FENDER JAZZ BASS)
↳ J&D T-STYLE BASS
↳ MUSIC MAN STINGRAY BASS
↳ FENDER TELECASTER
COMPACT OLED CLOCK & TIMER
USB MIXED-SIGNAL LOGIC ANALYSER (PicoMSA)
DISCRETE IDEAL BRIDGE RECTIFIER (TH)
↳ SMD VERSION
MICROMITE EXPLORE-40 (BLUE)
PICO BACKPACK AUDIO BREAKOUT (with conns.)
8-CHANNEL LEARNING IR REMOTE (BLUE)
3D PRINTER FILAMENT DRYER
DUAL-RAIL LOAD PROTECTOR
VARIABLE SPEED DRIVE Mk2 (BLACK)
FLEXIDICE (RED, PAIR OF PCBs)
SURF SOUND SIMULATOR (BLUE)
COMPACT HIFI HEADPHONE AMP (BLUE)
CAPACITOR DISCHARGER
PICO COMPUTER
↳ FRONT PANEL (BLACK)
↳ PWM AUDIO MODULE
DIGITAL CAPACITANCE METER
BATTERY MODEL RAILWAY TRANSMITTER
↳ THROUGH-HOLE (TH) RECEIVER
↳ SMD RECEIVER
↳ CHARGER
5MHZ 40A CURRENT PROBE (BLACK)
USB PROGRAMMABLE FREQUENCY DIVIDER
HIGH-BANDWIDTH DIFFERENTIAL PROBE
NFC IR KEYFOB TRANSMITTER
POWER LCR METER
WAVEFORM GENERATOR
PICO 2 AUDIO ANALYSER (BLACK)
PICO/2/COMPUTER
↳ FRONT & REAR PANELS (BLACK)
ROTATING LIGHT (BLACK)
433MHZ TRANSMITTER
VERSATILE BATTERY CHECKER
↳ FRONT PANEL (BLACK, 0.8mm)
TOOL SAFETY TIMER
RGB LED ANALOG CLOCK (BLACK)
USB POWER ADAPTOR (BLACK, 1mm)
HWS SOLAR DIVERTER PCB & INSULATING PANELS
SSB SHORTWAVE RECEIVER PCB SET
↳ FRONT PANEL (BLACK)
433MHz RECEIVER
SMARTPROBE
↳ SWD PROGRAMMING ADAPTOR
DATE
MAR24
APR24
APR24
APR24
MAY24
MAY24
MAY24
JUN24
JUN24
JUN24
JUN24
JUN24
JUL24
JUL24
JUL24
AUG24
AUG24
AUG24
AUG24
SEP24
SEP24
SEP24
SEP24
SEP24
SEP24
SEP24
SEP24
OCT24
OCT24
OCT24
OCT24
OCT24
NOV24
NOV24
NOV24
DEC24
DEC24
DEC24
DEC24
DEC24
JAN25
JAN25
JAN25
JAN25
JAN25
JAN25
FEB25
FEB25
FEB25
MAR25
MAR25
MAR25
APR25
APR25
APR25
APR25
MAY25
MAY25
MAY25
MAY25
MAY25
JUN25
JUN25
JUN25
JUN25
JUL25
JUL25
PCB CODE
Price
16103241
$20.00
08101241
$15.00
08104241
$10.00
07102241
$5.00
04104241
$10.00
04112231
$2.50
10104241
$5.00
SC6963
$10.00
08106241
$2.50
08106242
$2.50
08106243
$2.50
24106241
$2.50
CSE240203A $5.00
CSE240204A $5.00
11104241
$15.00
23106241
$10.00
23106242
$12.50
08103241
$2.50
08103242
$2.50
23109241
$10.00
23109242
$10.00
23109243
$10.00
23109244
$5.00
19101231
$5.00
04109241
$7.50
18108241
$5.00
18108242
$2.50
07106241
$2.50
07101222
$2.50
15108241
$7.50
28110241
$7.50
18109241
$5.00
11111241
$15.00
08107241/2 $5.00
01111241
$10.00
01103241
$7.50
9047-01
$5.00
07112234
$5.00
07112235
$2.50
07112238
$2.50
04111241
$5.00
09110241
$2.50
09110242
$2.50
09110243
$2.50
09110244
$2.50
9049-01
$5.00
04108241
$5.00
9015-D
$5.00
15109231
$2.50
04103251
$10.00
04104251
$5.00
04107231
$5.00
07104251
$5.00
07104252/3 $10.00
09101251
$2.50
15103251
$2.50
11104251
$5.00
11104252
$7.50
10104251
$5.00
19101251
$15.00
18101251
$2.50
18110241
$20.00
CSE250202-3 $15.00
CSE250204 $7.50
15103252
$2.50
P9054-04
$5.00
P9045-A
$2.50
DUCTED HEAT TRANSFER CONTROLLER
↳ TEMPERATURE SENSOR ADAPTOR
↳ CONTROL PANEL
MIC THE MOUSE (PCB SET, WHITE)
USB-C POWER MONITOR (PCB SET, INCLUDES FFC)
AUG25
AUG25
AUG25
AUG25
AUG25
17101251
17101252
17101253
SC7528
SC7527
NEW PCBs
$10.00
$2.50
$2.50
$7.50
$7.50
We also sell the Silicon Chip PDFs on USB, RTV&H USB, Vintage Radio USB and more at siliconchip.com.au/Shop/3
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. Send your email to silicon<at>siliconchip.com.au
Hot Water System Solar
Diverter questions
I enjoy all your articles and projects,
with their explanations. I was very
interested in the Hot Water System
Solar Diverter project in the June 2025
edition (siliconchip.au/Series/440).
However, this project has left me with
questions that other readers may also
be wondering about.
1. The diverter is supposed to obtain
power export data from the inverter.
How does an inverter ‘know’ how
much of its power output is being
self-consumed and how much is
being exported? I understand that the
inverter measures the energy it produces. However, is it able to differentiate between its energy that is self-
consumed and that which is exported
to the grid?
2. The graph in Fig.1 on page 37
shows consumption/self-consumption
rising and falling sharply during daytime. Consumption loads during the
day are typically lower than in the
evenings. Unless the household is
frequently switching big loads on and
off, consumption would not swing
as the graph suggests. The consumption seems to follow the pattern of the
clouds that affect solar production. Is
there more explanation for this graph?
3. The ‘diverter’ is just a switch that
turns the hot water heater on and off,
even with the smarts in the software. I
had imagined the diverter would have
a changeover capability, so the heater
can be fed from the mains or solar (or
neither in the case of batteries).
4. If the diverter simply supplies
power to the hot water heater only
when there is excess solar output,
Why is the RGB LED ‘Analog’ Clock PCB round?
I built the Mesmeriser LED clock from one of your kits in 2005/2006. I had much
joy building the clock (my background is in electronics), and it ran faithfully on
my office wall for nearly twenty years. It failed about a year ago. I don’t remember
exactly what happened to it, but I decided it was ‘time’ to retire it.
So, I am excited to see Nicholas Vinen’s ‘timely’ new version in Silicon Chip’s
May 2025 issue, and I’ll have to buy that kit and start all over.
But it occurred to me that as well as a typical round clock face, why not a square
clock face, or tilt the square 45° for a diamond shape? Why stop there? Pick almost
any shape you can think of: oval or rectangle (horizontal or vertical), triangle etc.
A square or vertical rectangular shape could be built into a proportionately sized
grandfather clock case to sit on a hall table.
I have no idea of the cost to make a different shaped PCB, that may make the
options prohibitive. But if it can be done cheaply, LEDs chasing around a non-round
shape could appeal to some, myself for one. Hell, I might have to buy a different
shape for every room in the house!
Alternatively, I could make my own odd-shaped clock face and extend the
relevant LED connections from the supplied round PCB. Some of the more urgent
jobs around the house might just have to wait! Keep up the great work. (G. M.,
Pukekohe, New Zealand)
● The cost is generally based on the area of the rectangle that the PCB fits inside,
so a square clock PCB that’s 200 × 200mm would cost the same as the 200mm
diameter circular one we decided on.
Essentially any PCB shape is possible, but the design would have to be redone
in the new shape, with the LEDs painstakingly arranged and wired up. Still, a skilled
PCB designer could probably redo it in a few hours.
Ordering large PCBs is expensive, so we have to pick the shape that most people
would want for a clock. We figured, given the option, most people would choose
the circle, so that’s what we stuck with. Ordering two or three different batches
of boards in different shapes would have increased the amount of work and cost
substantially.
100
Silicon Chip
Australia's electronics magazine
wouldn’t there be cold water on
low-solar days? Am I missing something here? (N. K., Kedron, Qld)
● Our replies below:
#1. A typical grid-feed inverter has
a current transformer in its grid interface. It measures the energy flowing
into or out of the grid from the premises the same way an electricity meter
would. Inverters that lack such a transformer will calculate local consumption as production minus export.
#2. The consumption follows the
production in this case because that is
the purpose of the HWS Solar Diverter.
It determines how much excess solar
generation is available and adjusts the
average HWS element power to use as
much of it as possible without drawing from the grid. So the plot shows
it doing its job, despite the constantly
varying generation levels during that
day.
#3. The heater can be fed by mains
or solar, since the two are merged at
the grid-tied inverter. If the heater is
switched on when there is excess solar
production, it’s powered by solar. If it’s
switched on when there’s little to no
solar production, it’s powered by the
grid. That’s the same as any appliance
in a home with a grid-tied inverter.
#4. The unit has a HWS temperature sensor and can command the unit
to draw power from the grid if necessary. This is the feature described on
the first page of the article as “Automatic override if the HWS temperature
is still cold by the end of the solar day”.
RGB LED Clock time
zone is set manually
I have almost completed the RGB
LED ‘Analog’ Clock project (May 2025;
siliconchip.au/Article/18126); I just
need to get the Raspberry Pi Pico W
time source working properly. I previously set up one of these for my Compact OLED Clock and Timer project
without any difficulty.
This time, the unit sets the time
at GMT, 10 hours behind the correct
time. I set the latitude/longitude to my
siliconchip.com.au
location, and although the info from
the NTP insists that I am at a location 100km away, it is still within the
correct time zone. I also set the IPAPI
parameters. I can’t understand this.
Have I set something wrong? Can you
please assist? (D. C., Beachmere, Qld)
● The RGB LED Analog Clock
doesn’t use the latitude/longitude data
to set the time zone, so the location
data should not matter.
In fact, the same applies to the
Compact OLED Clock; it defaults to
GMT+10 (since that applies to the
majority of our Australian readers).
Having the correct time zone in this
case is little more than the coincidence of the defaults matching your
time zone.
With that said, the RGB LED Analog Clock should default to the same
GMT+10 time zone, and we are unsure
why that is not the case.
The time zone can be manually set
and the full details are on page 75 of
the project article. Briefly, a long press
on the A button enters the time zone
setting mode. Short presses on A or
B will adjust the time zone earlier or
later in 15-minute increments. A long
press on B will toggle daylight savings (assuming it is a one-hour offset),
while a second long press on A will
exit this mode.
How to retain a car CD
player’s memory
Discovering that our household had
no working CD players, I decided to
use an old car audio head unit that I
had taken out of one of our many vehicles. I sourced a suitable 12V power
supply, and have made a timber enclosure for it.
Like all car audio units, its settings
are retained by the continual presence
of 12V DC. I’d prefer not to have my
unit on all the time, so I decided to
include some sort of battery backup
to feed the ‘settings supply’.
I thought at first that supercapacitors
or the like might be a suitable store.
The spec sheet doesn’t mention it, but
testing shows the current draw when
‘off’ to be as much as 20mA. Given
that, I assume that supercapacitors are
out, as I was hoping for at least a few
days of off-time without the settings
getting lost.
I looked around online, but did not
find any products or circuit designs
that I consider being definitively
siliconchip.com.au
Differential Probe capacitor confusion
I’ve run into a problem with the PCB for this project (February 2025 issue; siliconchip.
au/Article/17721). The pads for C16 are a short circuit. Could you please check
your stock of boards for a short circuit between the pads of C16? It could be that
I bridged these pads with solder. However, unlike C15’s pads which I’ve been able
to clean up, I cannot remove the short between the pads of C16.
I check all SMD components as I go for continuity and shorts; that’s how I found
this problem. I’ve removed C16 and cleaned the pads, but still had a short. I then
removed C15 as a sanity check, and it’s fine. (B. P., Jeir, NSW)
● Given this board’s relatively small clearances between tracks, or tracks and
ground pours (6 thou/0.15mm),
a fault isn’t completely out of the
question, although it would be
very unusual. This is a standard
clearance required by many finepitch SMD ICs. Modern PCB
manufacturing is pretty reliable, and
most boards are electrically tested
by the manufacturer.
We wonder if you may have
accidentally soldered the two
capacitors horizontally rather than
vertically, as shown by the red
outlines in the accompanying
diagram. That would be easy to do
as the nearest components are also
horizontal and there are no outlines
marked on the PCB (since there isn’t
much space).
If you did that, the upper capacitor
would be between two ground pads
and thus would appear shorted.
If that’s the case, it should be
possible to carefully desolder
the components, rotate them,
and resolder them to the board
correctly.
suitable. This is probably indicative
of the fact that I’m misguided in my
approach! What do you recommend I
use to feed the 20mA supply? (A. J.,
Mindarie, WA)
● A 12V sealed lead acid (SLA) battery or a compatible LiFePO4 12V battery that’s charged using a 12V battery
charger would do the job. Something
like a 4.5Ah rating would provide a
few days of ‘settings’ storage, although
a 1Ah battery should be suitable if
only a couple of days without power
is normal. The batteries and battery
chargers are available from Jaycar and
Altronics.
Head units will happily run from up
to 14.4V (as they would see when the
car’s engine is running), so you could
use a single power supply to run the
player and charge the battery, although
this does rely on enough active usage
over time to keep the battery charged.
You could use a 15V DC regulated
Australia's electronics magazine
supply with a single series diode to
obtain ~14.3V to run the player, then
another series diode to drop it to
~13.6V to float charge the battery. A
low-value resistor in series with the
battery can limit the initial charging
current to avoid overloading the supply (eg, a 2.2W 5W resistor should
keep the maximum charging current
under 1A).
Will the VGA PicoMite
work with a Pico 2?
I recently read about the VGA
PicoMite and saw that you are selling
a kit for it. I am wondering if the kit is
compatible with Raspberry Pi Pico 2. If
I use a Pico 2 instead of the included
Pico, and install the latest PicoMite V6
firmware, will everything will work
as expected?
Also, I want to upgrade the default
Pico with the latest version of the
August 2025 101
PicoMite firmware. Will the kit still
work? (J. C., via email)
● Geoff Graham responds: I have
just updated my web page to clarify
this. Yes, the VGA PicoMite hardware
will work with either a Pico or Pico 2,
as long as the appropriate firmware is
installed (see the PicoMite 2 article
from February 2025 at siliconchip.au/
Article/17729). There’s nothing stopping you from upgrading the firmware
in the supplied Pico.
How were EPROMs
programmed in 1997?
Dr Hugo Holden’s article in the January 2025 issue about retrieving data
from old microcontrollers piqued my
interest. I built the colour TV pattern
generator from your June and July 1997
issues. It is still working well, but it
would be a shame if the EPROM failed.
I am curious to know what programming setup was used at the time.
I have looked on eBay etc and
noticed that there are some EPROM
programmers available, but they don’t
appear to support the device used in
the colour pattern generator. I realise
that the technology is dated now, but it
would be interesting to build a project
that runs on modern PC software that
can talk to these old chips.
As mentioned in Hugo’s article, it
can save some specialised gear from
the scrap heap. (G. C., Toormina, NSW)
● We used a basic EPROM programmer driven by a computer programmed
in BASIC. Unfortunately, the details of
that setup are lost in the mists of time.
The Windows-based EPROM programmer by Jim Rowe that was published in late 2002/early 2003 (see
siliconchip.au/Series/110) would be
able to program these devices. However, the software would need to run
within a DOSBox emulator on a modern Windows computer. You would
also need a USB to Centronics interface converter.
An easier solution would be to purchase the XGECU T48 Universal Programmer that we reviewed in April
2023 (siliconchip.au/Article/15735).
Its software runs natively on Windows 10/11.
GPS Time Source not
getting valid data
I’m having problems with the Clayton’s GPS Time Source project (April
102
Silicon Chip
2018; siliconchip.au/Article/11039).
I’m using the ESP8266 D1 Mini module, as you used in the article on page
58 of that issue. I compiled the code
using Arduino IDE V2.3.5, set for an
ESP8266 “LOLIN(WEMOS)D1 R1”.
The code is “NTP_client_for_
ESP8266_GPS_V13skt.zip”, downloaded yesterday from the Silicon
Chip site. It compiles OK and programs the ESP8266, although there are
many warnings in the compile window. After programming, I managed
to set it up for my home router SSID
and password OK.
However, the time it sends to the IDE
serial port seems to be incorrect. The
time was approximately 04:01 UTC
according to my PC, but I got the following serial data:
$GPRMC,001632.009,V,3746.000,S,14
453.000,E,0.00,000.00,010118,,,*22
$GPGGA, 001632.009,3746.000,S,1445
3.000,E,0,04,1.0,0.0,M,0.0,M,,*7B
$GPGSA,A,1,,,,,,,,,,,,,,1.0,1.0,1.0,*2D
$ESP82,connected,SSID Telstra******
chan 6,10.0,0.52,0,0,0*03
Do you have an idea what’s causing
this? (G. P., Narre Warren South, Vic)
● The output that you’ve included
looks normal, but suggests that the
Time Source has not been able to
acquire the time successfully through
NTP. The V in the $GPRMC sentence
means that the data is ‘void’ and is
not yet valid.
The time it is reporting is 00:16:32
on 1/1/2018, which is 16 minutes
after the default time programmed
into the sketch when it starts. The
010118 in the output (near the end of
the $GPRMC line) is the date field. So
the time is wrong because it is using
a default.
We suggest you start by rebooting
the D1 Mini module to force it to retry.
If you have another WiFi network, that
might be worth trying, too.
We’ve heard reports of ESP8266
modules not working in cases where
there are 2.4GHz and 5GHz networks
with the same name. What appears to
happen is that the router kicks the modules off the 2.4GHz network to see if it
will join the 5GHz network instead. Of
course, the ESP8266 only has a 2.4GHz
radio, so this does not work.
Some readers have successfully
renamed their 5GHz networks as a
work-around. For example, I’ve added
a ‘_5G’ suffix to the 5GHz SSID of my
home network.
Australia's electronics magazine
We don’t think that the warnings
are a concern since the project compiles successfully. We are sure that
these warnings are due to changes to
the board profile since the last update
from a few years ago. We suggest using
these versions of the board profiles:
V11 should be used with ESP8266
Boards Manager Profile version 2.7.4
and earlier.
V12 should be used with ESP8266
Boards Manager Profile version 3.0.0
and later (tested with V3.0.2).
The newest ESP8266 board profile
is version 3.1.2, which we haven’t
tested, so it would be worth trying
with a 3.0.x version. You can select a
specific version and downgrade to it
in the Boards Manager.
Troubleshooting
Turntable Driver
I have just built the Precision Turntable Driver (May 2016; siliconchip.
au/Article/9930). I tested it today and
got the following very strange results.
Initially, it produced 230V AC after
adjusting trimpot VR1. Pins 5 and 14 of
IC1 read 4.95V DC. I was able to power
my turntable (with its AC synchronous
motor) for a few minutes.
Since my turntable runs about 5%
fast (about 35.4 RPM, for some reason),
I tried repeated presses of the ‘slower’
button. It did not reduce the speed
back towards 33.3 RPM. However, the
‘faster’ button did increase the speed
incrementally, up to nearly 40 RPM.
Shortly after this, I noticed the
power LED was flashing slowly, in a
sort of slow pulsing fashion with the
LED never completely going dark. This
was accompanied by a faint tapping
sound that was synchronised with the
LED flashing. I plugged the turntable
back in, but it appeared to get no power
and didn’t spin, unlike the initial trial
described above.
I examined the PCB very carefully
under bright light to make sure I had
no solder bridges or short circuits, but
of course I don’t know if all the semiconductors, capacitors etc are OK. (P.
L., Kaleen, ACT)
● Based on the photo supplied,
there are some long component pigtails extending from the PCB. You
should check that they don’t short
to the enclosure (or, even better, trim
them). Otherwise, the construction
looks good.
continued on page 104
siliconchip.com.au
MARKET CENTRE
Advertise your product or services here in Silicon Chip
KIT ASSEMBLY & REPAIR
FOR SALE
DAVE THOMPSON
(the Serviceman from
Silicon Chip) is available to help you with
kit assembly, project
troubleshooting, general electronics and
custom design work.
No job too small. Based in Christchurch,
New Zealand, but service available Australia/NZ wide.
Email dave<at>davethompson.co.nz
LEDsales
KEITH RIPPON KIT ASSEMBLY &
REPAIR:
* Australia & New Zealand;
* Small production runs.
Phone Keith: 0409 662 794
keith.rippon<at>gmail.com
LEDS, BRAND NAME AND GENERIC
LEDs, filament LEDs, LED drivers,
heatsinks, power supplies, kits and
modules, components, breadboards,
hardware, magnets. Please visit www.
ledsales.com.au
PMD WAY offers (almost) everything
for the electronics enthusiast – with
full warranty, technical support and free
delivery worldwide. Visit pmdway.com
to get started.
Lazer.Security
PCB PRODUCTION
PCB MANUFACTURE: single to multilayer. Bare board tested. One-offs to
any quantity. 48 hour service. Artwork
design. Excellent prices. Check out our
specials: www.ldelectronics.com.au
WE HAVE QUALITY LED’S on sale,
Driver sub-assemblies, new kits and
all sorts of electronic components,
both through hole and SMD at very
competitive prices. check out the latest
deals at www.lazer.com.au
Micromite Explore-40
October 2024
Complete Kit
SC6991: $35
siliconchip.au/Article/16677
Includes the PCB and all onboard parts. Audio
Breakout board and Pico BackPack are sold
separately.
FOR FREE:
To give away, one decommissioned and
partly disassembled YUKI KP-480 pick
and place machine.
Contact: Graham – 0458 071 074
Location: Hobart, Tasmania
ADVERTISING IN MARKET CENTRE
Classified Ad Rates: $32.00 for up to 20 words (punctuation not charged) plus $1.20 for each additional word. Display ads in
Market Centre (minimum 2cm deep, maximum 10cm deep): $82.50 per column centimetre per insertion. All prices include GST.
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 phone (02) 9939 3295.
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.
siliconchip.com.au
Australia's electronics magazine
August 2025 103
Advertising Index
Altronics.................................19-22
Blackmagic Design..................... 37
Dave Thompson........................ 103
Emona Instruments.................. IBC
Hare & Forbes............................... 7
Icom Australia............................. 10
Jaycar............................. IFC, 30-33
Keith Rippon Kit Assembly....... 103
Lazer Security........................... 103
LD Electronics........................... 103
LEDsales................................... 103
Microchip Technology.............OBC
Mouser Electronics....................... 3
OurPCB Australia.......................... 6
PCBWay......................................... 5
PMD Way................................... 103
Rohde & Schwarz........................ 97
SC Micromite Explore-40......... 103
SC PIC Programming Adaptor... 72
Silicon Chip Back Issues..... 45, 88
Silicon Chip Battery Checker.... 66
Silicon Chip PDFs on USB......... 57
Silicon Chip Shop...........90, 98-99
Silicon Chip Songbird................ 51
Silicon Chip Subscriptions........ 73
The Loudspeaker Kit.com.......... 89
Wagner Electronics....................... 9
Errata and on-sale date
Vintage Radio – Emerson 888,
May 2025: Ian Batty quoted his
favourite “weak signal” station
as 3WV at Warrnambool. Fellow
HRSA member Bob Forbes
informed him that the station at
Warrnambool is actually 3WL
(1602kHz), ABC South Western
Victoria. His reference is actually
3WV (594kHz), ABC Radio
Wimmera at Horsham, around
300km from Rosebud.
Next Issue: the September 2025
issue is due on sale in newsagents
by Thursday, August 28th. Expect
postal delivery of subscription
copies in Australia between August
27th and September 12th.
104
Silicon Chip
It seems like the plugpack power
supply is not delivering power or there
is a short circuit or something drawing a lot of current in the circuit. Start
by checking the plugpack output. Is it
capable of delivering 2A? If it seems
OK, perhaps there is a component
drawing extra current. It would get
hotter than expected.
The fact that the power LED flashes
suggests that the plugpack is switching
on and off, and is possibly overloaded.
We recommend using an Altronics
M8945A (15V DC, 2.4A) or M8945B
(15V DC, 3.5A) plugpack to power the
Precision Turntable Driver. Jaycar’s
MP3492 could also be used, although
its 2A rating is only just high enough.
As for the inability to reduce the
speed, we think it is likely a fault with
the button or its wiring.
Editor’s note: the reader later replied
that the plugpack was at fault.
RIAA Preamplifier
wanted
I have been asked to restore a 1950s
chest-type valve radiogram that has
great sentimental significance to its
owner. Unfortunately, the existing Collaro record changer looks to be beyond
help, with perished rubber parts. The
only solution is to retrofit a newer turntable/record changer.
I envisage using something like a
later model Garrard fitted with a magnetic cartridge. The radiogram won’t
have sufficient audio gain to be driven
directly from a magnetic cartridge, but
I can easily add a solid state preamplifier, hidden inside the cabinet of the
radiogram.
Have you published a design for
a mono (or stereo) RIAA magnetic
phono preamp that can provide up to
300-400mV output? If so, do you have
PCBs available for it?
The equalisation curves are different between LPs and 78s. I don’t know
what the record playing expectations
of the radiogram’s owner are as yet.
I am assuming that 78 magnetic cartridges are still available, should the
owner plan on playing 78 records. I
am aware that the output level of the
preamp depends on the output level
of the cartridge.
Assuming there is a suitable design,
can I tweak the preamp’s output level
by changing the amount of feedback on
the preamp’s IC, or is the feedback loop
entirely dedicated to the equalisation
Australia's electronics magazine
components? Also, what are its power
supply requirements? (P. W., Auckland, New Zealand)
● We published a Magnetic Cartridge Preamplifier in August 2006
(“Build A Magnetic Cartridge Preamplifier”; siliconchip.au/Article/2740).
This is a stereo preamplifier and
its gain should be suitable for your
required output level. This level
depends upon the cartridge signal output with record groove modulation.
The PCB is available from our
Online Shop (siliconchip.com.au/
Shop/8/860). You can adjust the gain
by changing the components. We
also have a SPICE simulation file that
can be used to check the response
if changes are made. Changes to the
gain shouldn’t be necessary normally
unless a low output cartridge is used,
such as a moving coil type.
It can run from a 12V AC 250mA
plugpack. The circuitry has onboard
12V regulators to provide the required
±12V rails.
How to make a whistle
filter
I am trying to build one of your old
projects, but it uses parts that are no
longer available. Can you tell me how
to make a 9kHz whistle filter coil and
a 15.625kHz whistle filter coil? (W. O.,
Miller, NSW)
● The 9kHz notch filter would be for
an AM receiver, while the 15.625kHz
filter would be to remove the horizontal scan frequency from an analog CRT
TV. The components used can include
an inductance, a capacitance and a
resistance, or simply resistors and
capacitors, with or without op amps.
For passive filters, go to siliconchip.
au/link/ac79 and scroll down to RLC
notch filter. That will give you the
details and equation to relate the notch
frequency to the component values.
You can work out the details for
winding an air-cored inductor, if
required, using the online calculator
at siliconchip.au/link/ac77
Or, for a ferrite-cored inductor, use
the online calculator at siliconchip.au/
link/ac78 (you will need to know the
AL value of the ferrite core).
Active filters can be easier to build
than passive filters using an inductor, since these can just use resistors
and capacitors, as described on Rod
Elliott’s website at https://sound-au.
SC
com/articles/notch-filters.htm
siliconchip.com.au
“Rigol Offer Australia’s Best
Value Test Instruments”
New
2024
Products
Oscilloscopes
New
12Bit
Scopes
RIGOL DS-1000Z/E - FREE OPTIONS
RIGOL DHO Series
RIGOL MSO-5000 Series
450MHz to 200MHz, 2/4 Ch
41GS/s Real Time Sampling
424Mpts Standard Memory Depth
470MHz to 800MHz, 2/4 Ch
412Bit Vertical Resolution
4Ultra Low Noise Floor
470MHz to 350MHz, 2 Ch & 4Ch
48GS/s Real Time Sampling
4Up to 200Mpts Memory Depth
FROM $
499
FROM $
ex GST
659
FROM $
ex GST
1,489
Multimeters
Function/Arbitrary Function Generators
New
Product
New
Product
RIGOL DG-800/900 Pro Series
RIGOL DG-1000Z Series
RIGOL DM-858/E
425MHz to 200MHz, 1/2 Ch
416Bit, Up to 1.25GS/s
47” Colour Touch Screen
425MHz, 30MHz & 60MHz
42 Output Channels
4160 In-Built Waveforms
45 1/2 Digits
47” Colour Touch Screen
4USB & LAN
FROM $
713
FROM $
ex GST
Power Supplies
ex GST
725
FROM $
ex GST
Spectrum Analysers
689
ex GST
Real-Time Analysers
New
Product
RIGOL DP-932E
RIGOL DSA Series
RIGOL RSA Series
4Triple Output 2 x 32V/3A & 6V/3A
43 Electrically Isolated Channels
4Internal Series/Parallel Operation
4500MHz to 7.5GHz
4RBW settable down to 10 Hz
4Optional Tracking Generator
41.5GHz to 6.5GHz
4Modes: Real Time, Swept, VSA & EMI
4Optional Tracking Generator
ONLY $
849
FROM $
ex GST
1,321
FROM $
ex GST
3,210
ex GST
Buy on-line at www.emona.com.au/rigol
Sydney
Tel 02 9519 3933
Fax 02 9550 1378
Melbourne
Tel 03 9889 0427
Fax 03 9889 0715
email testinst<at>emona.com.au
Brisbane
Tel 07 3392 7170
Fax 07 3848 9046
Adelaide
Tel 08 8363 5733
Fax 08 83635799
Perth
Tel 08 9361 4200
Fax 08 9361 4300
web www.emona.com.au
EMONA
|