Fullscreen HTML5Med Res (??MB)HTML4

FEBRUARY 2026 ISSN 1030-2662 02 9 771030 266001 $ 00* NZ $14 90 14 INC GST INC GST Interrnet Radio Inte Radio to r y of s i h he a Raspberry Pi-based music and audio stream player T intel the 1101 SRAM chip to beyond DCC Remote Controller run multiple trains through one DCC Base Station, with up to five Remote Controllers Contents Vol.39, No.02 February 2026 16 The History of Intel, Part 1 Intel currently makes more desktop, laptop & server CPU chips than any other company. How did Intel get into that position, what did they invent along the way, what challenges did they overcome and what about their future? By Dr David Maddison, VK3DSM Electronics feature 35 Power Electronics, Part 4 The History of Intel Part 1: page 16 Image source: Konstantin Lanzet – https://w.wiki/GVqx In this series of articles, we explore the principles of power electronics. This month, we look in detail at the deceptively simple rectifier type AC-DC converters. By Andrew Levido Electronic design Page 28 Mains Power LED Indicator 70 How to Design PCBs, Part 3 For the final article in the series, we cover advanced techniques and options that you might need to use when designing your own PCBs. We also look into what is required to get an entire PCB assembled. By Tim Blythman Making your own PCBs DCC Remote Controller 80 Tiny QR Code Reader As suggested by the name, this is a tiny module that uses a small camera to decode QR codes. The hardware is based on the same RP2040 processor from a Raspberry Pi Pico. By Tim Blythman Low-cost electronic modules 28 Mains LED Indicator LEDs are much better, and brighter, than neon lamps but need extra circuitry to run from the 230V mains. Our simple circuit lets you operate LEDs from the mains, and it’s not kept floating at a high or dangerous voltage. By John Clarke Lighting project 44 The Internet Radio, Part 1 If you’re looking for a music/streaming audio player or you have poor radio reception in your area, then this project is for you. It’s based on a Raspberry Pi 4B and can play from local files or internet streaming services. By Phil Prosser Radio/audio project 53 Mains Hum Notch Filter This Notch Filter reduces mains hum due to long unbalanced audio signal leads and nearby power wiring. It handles stereo signals and is powered by a separate 9-15V DC plugpack. By John Clarke Audio project 62 DCC Remote Controller Using this DCC Remote Controller, you can control multiple trains at the same time. Any type of DCC packet can be sent via the Controller, and you can even connect up to five of them to a single Base Station. Part 4 by Tim Blythman Model train project Page 62 2 Editorial Viewpoint 4 Mailbag 52 Subscriptions 84 Circuit Notebook 86 Serviceman’s Log 92 Online Shop 95 Vintage Radio 101 Ask Silicon Chip 103 Market Centre 104 Advertising Index 104 Notes & Errata 1. Wireless reed switch 2. Raspberry Pi reflash helper The Columbia TR-1000 portable radio by Ian Batty 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): $72.50 12 issues (1 year): $135 24 issues (2 years): $255 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 Printing and Distribution: 14 Hardner Rd, Mount Waverley VIC 3149 54 Park St, Sydney NSW 2000 2 Silicon Chip Editorial Viewpoint Will Arduino survive? When I heard that Qualcomm had acquired Arduino in October 2025, I immediately wondered whether they would ruin it. After all, what business does a large, closed company like Qualcomm have with a much smaller business developing open-source hardware and software? On the positive side, much of what Arduino has produced over the years, being open-source and widely available, can continue to exist regardless of what Qualcomm does. The IDE can be forked, and clones of the various boards can continue to be produced. On the negative side, Qualcomm is already known for being difficult to get information from, and they have made some worrying moves. For example, in November, Arduino updated their Terms of Service and privacy policy, forbidding (or attempting to forbid) users from reverse-­engineering Arduino platforms. The terms also state that they own anything users upload to their servers. Adafruit Industries, a major supplier of Arduino-compatible hardware, publicly questioned whether any of this was for the benefit of users and posted multiple critiques of the new terms, prompting others to chime in (see https://itsfoss. com/news/enshittification-of-arduino-begins). Qualcomm/Arduino replied by saying that people had interpreted some of the changes more broadly than intended. The problem is that doesn’t change what the legal text actually allows them to enforce. It’s also worth noting that, alongside the acquisition announcement, Arduino released the Uno Q, a new board with a Qualcomm chip aimed at AI applications: www.arduino.cc/product-uno-q It has interesting features, but I wonder how many people are getting tired of the ‘put AI into everything’ trend. How many hobbyists really want an AI-enabled Arduino? Time will tell. Ultimately, whether Arduino “survives” in the sense that matters: remaining relevant, open and community-driven, depends less on what Qualcomm does and more on how the maker community responds. The new Uno Q suggests a future where Arduino becomes a vehicle for Qualcomm’s ‘AI-at-the-edge’ ambitions. But the new restrictions that many see as incompatible with open-source hardware have already damaged trust among the very people who built Arduino’s reputation. The Uno Q is an interesting design. It uses a dual-processor architecture: a Qualcomm Dragonwing system-on-chip runs Linux alongside a more conventional STM32 microcontroller. The idea is that the Dragonwing performs tasks like AI models, computer vision or networking, while the microcontroller handles real-time I/O. It’s certainly an ambitious design, but also a striking departure from what Arduino boards have traditionally been. Old-school Unos were simple, inexpensive, and easy to understand; the Uno Q is closer to a hybrid between a Raspberry Pi and a microcontroller development board. Whether that added complexity will be genuinely useful to most Arduino users is still unclear. If that trust continues to erode, platforms like Raspberry Pi, ESP32 and other genuinely open alternatives could absorb much of Arduino’s user base – especially hobbyists and educators who value transparency and community support over corporate direction. Raspberry Pi in particular has already expanded into microcontrollers with the RP2040 and could easily step further into the space Arduino once owned. So the real question isn’t just whether Arduino will survive, but in what form. Qualcomm didn’t buy Arduino to shut it down, but whether it remains the approachable, open, community-powered platform it has been for the last two decades is far from certain. If it strays too far from those roots, others are ready to step in and fill the void. by Nicholas Vinen Cover image sources (Intel, left-to-right): www.cpu-zone.com/1101.htm | https://pixabay.com/photos/intel-8008-cpu-old-processor-3259173/ | https://w.wiki/GbkJ | https://w.wiki/GYK8 | www.reddit.com/r/pcmasterrace/comments/1hhug73/ 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”. Australian-made Jindivik drone from the 1950s I read the article on drones in the September 2025 issue by Dr David Maddison (siliconchip.au/Article/18847) and noticed that it did not mention the most famous Australian drone of all, the Jindivik. Maybe it did not come up when looking for information on “drones” because the word drone was not in common usage then like it is now. Considering what an Australian success story this was, it is a shame it was missed. The Jindivik was developed by the Australian Government Aircraft factories for the Brits. It was regarded as a world-beater and saw more than 20 years’ service. Some of them went to the USA and Sweden, too. Australia produced over 350 Jindiviks. They were amazing target-towing radio-controlled planes and could test 4 Silicon Chip guided weapon systems (at Woomera). They could operate at altitudes from under 100 feet to 60,000 feet and fly at 650 miles per hour (1050km/h, 560 knots)! This was in 1969. The Jindivik even featured on the front cover of Electronics Australia. Dr Hugo Holden, Buddina, Qld. Comment: thanks for the information. The Jindivik has been mentioned in past Silicon Chip articles. The focus of the September 2025 article was on developments made in the last 10 years or so; hence, there was limited detail on older drones, covering just the ‘firsts’. Capacitor leakage is not predictable I am responding to Fred Lever’s comment regarding using capacitors as a bias method in a Reinartz TRF Receiver (Mailbag, December 2025). Having worked on valve radios from just about every era, over a span of over 50 years, I note that in earlier times, the materials and methods used did lead to capacitors of dubious integrity. The majority of low-value (picofarad range) capacitors in earlier times were mica and silvered mica, the latter being the most unreliable as they often shorted out due to silver whiskers. The higher values of non-polarised capacitors tended to be paper and aluminium foil types. The unfortunate thing with these was, no matter how they tried, eventually the dielectric material absorbed enough moisture for them to become resistors. This problem continued until polymer film capacitors came about. I cannot understand why people refurbishing valve radios leave paper caps in them. They will all be out of specification and leak current like sieves. There are two tests for high-voltage capacitors in vintage radios: one is leakage, the other value. If a capacitor is leaky, I won’t bother to test its capacity; it needs to be replaced. Newer caps don’t leak. Non-polarised caps should not pass DC. In older times, in accordance with the manual for an acquired “Lafayette TE46 Capacitance – Resistance Analyser” (circa 1962), if the NP capacitor had a resistance below 50MW, it was not suitable as a decoupling cap. Below 200MW, it was considered not suitable for coupling. I have found that many of those older capacitors are lucky to get to 1MW. I usually test at either the running or maximum voltage it will be exposed to, or the rated voltage using an insulation tester, or similar device. One tester from 1938 uses a 235V DC supply and a neon lamp. The extent to which the neon lamp glows shows how bad the leakage is. It is supposed to extinguish with a good cap. In the case of coupling and a cap that goes from plate to Australia's electronics magazine siliconchip.com.au grid, this is quite reasonable. In most of the radio valves, the grid is more of a voltage control device and draws virtually no significant current. Therefore, any badly leaking capacitor in this situation will send the grid positive, turning it into a diode, possibly killing it. Tying the grid to positive was done, and the Lyric 70 series is one where a triode is used as a rectifier for around 54V for a #50 output valve’s grid. Fred’s idea that capacitors were intentionally used for grid bias is very much a hypothetical. To use a capacitor as a resistor, you would have to go through a box of duds to get the right resistance, and I see no reason for that not to result in signal and reliability issues. Looking at a UX-201A of the era, one of the things forgotten is that this is a filament tube and in many cases, the filament was polarised and part of the valve’s bias. So, using a heater tube is going to alter many parameters. Also of note is that as a detector, it has its grid leak tied to filament positive (RCA). So, long-term, I do not see a leaking cap as a stable resistor, nor capacitor, and is likely to affect the signal and the valve in a negative way. Therefore, I see it as a hypothetical only and not practical in a DC or AC situation. Marcus Chick, Wangaratta, Vic. Comment: some early radio designs from roughly 19131920 omitted an explicit grid-bias (grid-leak) resistor and would operate only because of incidental leakage paths, such as grid current, capacitor leakage, surface leakage and antenna or coil DC paths. They were sometimes unreliable for reasons that should be evident. It turns out it was quite a straightforward modification (see the photo below). I used an ESP8266 on a prototype board, which has the minimal support circuitry needed: pull-up resistors and a 3.3V low-dropout (LDO) regulator. The ESP8266 was preprogrammed with the WLED software before being mounted on the prototype board. I used double-sided foam tape to mount the ESP8266 board to the back of the ornament. Power and ground wires attach to pins 2 and 3 respectively of the ICSP socket, CON2. The 330W resistor from the PIC to the LED string was lifted and resoldered to the LED-side pad only. A wire connects the free end of the 330W resistor to GPIO4/D2 on the ESP8266. I then powered up the ornament and opened the webpage presented by WLED. In LED Preferences, I set Length to 80 and Data GPIO to 4. All the LEDs then illuminated with the warm white default colour. I was then able to play with the abundant colour and effect combinations available. A simpler way to achieve this modification might be to use one of the ESP32 mini development boards, such as the ESP32-C3 or ESP32-S3. These have an in-built USB port, so programming is as simple as connecting a USB cable from the board to a PC and using the Chrome browser, following the WLED Quick Start guide. I hope this is of interest. P.S. In the process of making this modification, I discovered a minor circuit error. Pin 2 of CON3 is actually connected to the PIC side of the 330W resistor, not the LED side as shown on the schematic. David Smith, East Melbourne, Vic. Wirelessly controlling the RGB LED Star Power Electronics articles enjoyed After reading about the RGB LED Star ornament in the December 2025 issue (siliconchip.au/Article/19372), I purchased the (mostly assembled) kit for use as a Christmas decoration. It arrived quickly, so I soldered on the additional components and it was immediately up and running. However, adjusting and using the ornament was not all that easy – adjusting tiny trimpots and buttons on each side – complicated by the intended location, where it wouldn’t be readily accessible. I have built several RGB smart lights in the past using the excellent, open-source WLED package (https://kno.wled.ge) loaded onto an ESP8266, so I looked into modifying the ornament to use a similar solution. 6 Silicon Chip The recent articles by Andrew Levido are excellent. It’s great to have a bit of engineering theory mixed with practical considerations. Paul Howson, Warwick, Qld. Salvaging Li-ion cells from dud batteries I have been fortunate to have obtained a number of failed Li-ion batteries over time. In every instance, only some of the cells were faulty. An example is a battery for a Dyson wand-type vacuum cleaner. My neighbour told me that the battery refused to charge, so she bought a new one and was going to throw the old one in the bin. Besides wanting to recycle the battery, I was curious as to why it wouldn’t charge. So I opened it and discovered that there was only one cell dead or rather near death. I could see no damage to the controlling circuit, nor to any part of the battery, and came to the conclusion that the controller had blocked charging to prevent a fire. The battery had been used for some time, but the question arose whether the battery had reached the claimed life of the cells. The other cells were in good condition, which suggested that, if the dead cell had been like the others, the battery might have lasted much longer. It took only one cell of worse quality than the others to shorten the life of the battery. Perhaps devices should be designed to operate on a range of voltages and the batteries be designed to isolate one or two dead cells to prevent premature battery disposal. I have found that if Li-ion cells are used in spring-loaded battery holders, the contact resistance can prevent the cells from being fully charged to 4.2V. The charging circuit will Australia's electronics magazine siliconchip.com.au W LI H VE IL S E Y R E NC CO M RDED IN IA G ! Get high end digital film camera features on phone or iPad! Blackmagic Camera unlocks the power of your phone or iPad by adding digital film camera controls and image processing! You can adjust settings such as frame rate, shutter angle, white balance and ISO all in a single tap. Recording to Blackmagic Cloud lets you collaborate on DaVinci Resolve projects with editors anywhere in the world, all at the same time! Live Sync to Blackmagic Cloud Cinematic Quality Images Remote Camera Control and Multiview Blackmagic Camera puts the professional features you need for feature film, If you’re positioning an iPhone in an area that’s hard to reach, or shooting with television and documentaries in your pocket. Imagine having a run and gun multiple phones using Blackmagic Camera, you can get full control using camera on hand to capture breaking news whenever it happens! Or use remote camera control! Simply set your iPhone or iPad to be the controller, Blackmagic Camera as a B Cam to capture angles that are difficult to reach with and you can change settings for all iPhones using the same Wi-Fi network. Plus traditional cameras, while still retaining control of important settings. you can view each camera’s shots in a multiview! Interactive Controls for Fast Setup Blackmagic Camera has all the controls you need to quickly setup and start shooting! The heads up display, or HUD, shows status and record parameters, histogram, focus peaking, levels, frame guides and more. You can shoot in 16:9 or vertical aspect ratios, plus you can shoot 16:9 while holding the phone vertically if you want to shoot unobtrusively. www.blackmagicdesign.com/au Blackmagic Camera records an HD proxy that uploads to Blackmagic Cloud in seconds, so your media is available back at the studio in real time. The ability to transfer media directly into the DaVinci Resolve media bin as editors are working is revolutionary and has never before been possible! Any editor working anywhere in the world will get the shots! Blackmagic Camera Free Download Learn More! stop charging when it detects a 4.2V ‘cell voltage’, but with a non-zero contact resistance, the charging circuit actually detects the cell voltage plus the contact resistance multiplied by the charging current. As a result, the cell will never be fully charged. I discovered this when my Arlec torch would indicate fully charged, but the Li-Ion cells only measured 3.96V. After some thought, I cleaned the spring contacts, the other contacts and then the torch charged correctly. I used to wonder why mobile manufacturers would gold plate the contacts on the Li-ion cells and use gold-plated spring contacts. Now I know. George Ramsay, Holland Park, Qld. Comment: the idea of designing the device to work over a range of pack voltages with the ability for it to bypass one or two dud cells is a good one and probably wouldn’t add too much cost or complexity. However, we doubt manufacturers would bother as the premature failure is seen as a benefit to them (since the customer is likely to just buy a replacement if it fails after a few years). Challenges with SMD soldering Thanks as always for a brilliant magazine! I found the Editorial Viewpoint on SMD soldering in the January 2026 edition interesting because I have also found that a 0.6mm conical tip I bought for my Weller iron is best for SMD soldering. That said, while I’ve had great success with 0805 (2 × 1.2mm) components, I struggle with 0603 (1.6 × 0.8mm) components if there are very close together, with stray bits of solder getting onto nearby pads. I also pay others to solder expensive ICs with many finepitched leads or Mosfets with exposed pads since I wrecked a TQFP-64 microcontroller and CLASSiC DAC PCB many years ago. It’s partly the cost of having to reorder the components and having to pay an excessive amount for shipping (unless you can get an order up to $60), but also the delays. I did eventually complete a CLASSiC DAC and used it for a few years. I’ve been playing around with different topologies of DC-DC converters for the last couple of years, with a longterm plan of building an SMPS to provide all the voltages required for a stereo SE valve amp. I’ve always stuck to my policy of paying for soldering the difficult components. I seriously considered building the DIY Reflow Oven (April & May 2020; siliconchip.au/Series/343) but decided to stay focused on the converter projects. Suitable ovens are more difficult to obtain now, and the $400 total project cost pays for soldering quite a few PCBs. Then I decided to build the High Bandwidth Differential Probe (February 2025; siliconchip.au/Article/17721). It’s a great project and just what I needed to look at switching waveforms. Andrew Levido made soldering the WSON12 regulator sound easy with a hot air gun, so I decided to give it a go. Sadly, when it came time to test the regulator, it didn’t work. Either the exposed pad is not connected or I overheated the chip. Fortunately, I bought two PCBs because I thought I might build two. There was a long delay in getting another regulator IC while I built up an order that qualified for free shipping. I then paid to have all three chips with exposed pads attached to the second PCB. The precision op amps are not 8 Silicon Chip Australia's electronics magazine siliconchip.com.au cheap, and I didn’t want to risk damaging them. I would really have liked to be able to order a PCB with the exposed-pad ICs attached. Maybe this project is considered sufficiently advanced that it wasn’t considered necessary. Anyway, I’m now working my way through the rest of the probe assembly. David Hanslip, Sorrento, WA. Comment: the problem with providing pre-soldered PCBs is that it removes a lot of the educational, DIY aspect of the projects. Still, we may have to do it more often now that most new parts are not being released in hand-­solderingfriendly packages. More warnings about hardware that relies on ‘the cloud’ I just wish to advise Silicon Chip readers against buying equipment that relies on a cloud-based server for its operations. I fell into the trap by purchasing a Bose Wave SoundTouch 4. It has great sound and functions, especially for internet radio. However, it requires using their servers to change connection settings and local streaming to your home devices. I found out that they were shutting down their servers by accident because I could not set the internet connection from Wi-Fi to Ethernet. Maybe the shutdown has already limited what you can and can’t do with the device. It will be a brick once you need to change these settings after a factory reset, as you won’t be able to set things up anymore. As for other cloud-based stuff like photos, please keep them on your own storage (multiple copies) as you may lose them at the whims of server providers. Another reason not to keep them in the cloud is that Apple and Microsoft are allowed to use and modify them for things like advertising without copyright protection. This is apparently specified in their EULA. Wolf-Dieter Kuenne, Bayswater, Vic. Nicholas comments: I wrote about this problem in the February 2022 Silicon Chip Editorial. Anything that relies on ‘cloud’ services will only work for a few years. The frustrating thing is that it would hardly cost these companies much to keep the services going, but they like the planned obsolescence aspect. Personally, I will not buy anything that relies on an app or cloud services. That includes vehicles. The problem is that you often have to do a lot of research to find whether a product has this built-in obsolescence or not. Using NFC IR Keyfob with the NEC protocol I’m using a TSOP36438TT infrared receiver connected to the IR pin of a Micromite Explore-64 with a Jumbo NEC-­ protocol universal remote control. Recently, I decided to try the NFC Programmable IR Keyfob (February 2025; siliconchip.au/Article/17730) to replicate some specific frequently used functions and thought that other readers may benefit from my experience with this. I found that the IR Keyfob sends the code in reverse bit order (LSB first) compared to how the Micromite receives and assembles that code. The universal remote’s power on/ off button sends a code that is received in the Micromite as decimal 72, or 01001000 binary. In reverse bit order, this becomes 00010010 or 18 decimal. It took me quite an embarrassing while to realise this! ...continued on page 15 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 10 Silicon Chip Australia's electronics magazine 0417 264 974 siliconchip.com.au GOT A BRIGHT IDEA? LET’S BUILD IT! Your next big build starts here. Jaycar’s wide range of prototyping accessories delivers great value for every idea. PB8815 HP9572 FROM 6 FROM 6 $ 25 $ 75 . Breadboard Layout Prototyping Boards 6 models available. PB8815 - PB8832 PB8820 QUICK AND EASY PROTOTYPING . Breadboard Layout Prototyping Boards 400 Hole HP9570 | 862 Hole HP9572 MAKE YOUR BREADBOARD PROTOTYPE PERMANENT WC6027 Breadboard Jumper Kit 70 Pieces PB8850 $12.50 150mm Jumper Leads MAKE YOUR OWN CIRCUIT BOARDS Blank Fibreglass Copper Sided PCBs • 4 sizes available HP9510 - HP9515 FROM $6.95 SHOP AT JAYCAR FOR: 8 x 25m Hook-Up Wire Rolls WC6024 - WC6028 FROM $7.95 26AWG WH3009 $39.95 MAKE PCBS IN 4 EASY STEPS 1. PRINT/COPY 2. IRON ON 3. PEEL OFF 4. ETCH 20 Piece Micro Drill Set • Sizes: 0.3 - 1.6mm TD2406 $15.95 • Soldering & Accessories • Components, Cables and Connectors • Magnifiers and Inspection Aids • Tools, Service Aids and Chemicals PBC Wash Defluxing Solution • 1 Litre Bottle NA1070 $15.95 $ JUST 4695 . Press 'n' Peel Film 5 sheets of 215 x 280mm transfer film with full instructions. HG9980 Explore our great range of soldering 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 All 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. HP9570 NEW GEAR JUST LANDED From everyday essentials to clever problem-solvers, here are just some of our latest additions for your workbench, toolbox or tech lifestyle. (THIS IS MY EXCITED FACE) $ ONLY ONLY 7995 299 $ TD2474 QC8724 SEE IN HARD-TOREACH PLACES RECHARGEABLE 2-way articulating camera FHD ARTICULATING VIDEOSCOPE Thumbwheel controlled articulating inspection camera with 4.3" screen. Records to 32GB card to playback later. 1080p. $ ONLY 0-150MM DIGITAL VERNIER CALIPER Stainless steel caliper with 0.01mm/0.0005" resolution. Charge via USB. Storage case included. 2000V PV 98 95 $ ONLY 229 QM1506 QC1940 POCKET SIZED FOR IN THE FIELD IDEAL FOR TESTING PV SOLAR SETUPS 250kHz bandwidth Rugged, IP67 rated case HANDHELD POCKET DIGITAL STORAGE OSCILLOSCOPE Capable of handling both periodic analog and non-periodic digital signals. Rechargeable. Bag and cables included. SOLAR PHOTOVOLTAIC MULTIMETER Suits high-voltage DC applications including PV arrays, modules, power storage & industrial equipment. True RMS. IP67. CAT III. NEED IT NOW? SCORE FREE DELIVERY ON US 1 HOUR CLICK & COLLECT With orders over $99 at jaycar.com.au Shop online & collect in-store *Conditions apply *Conditions apply BUY IT ONLINE BEFORE IT'S EVEN IN STORES! $ ONLY 89 12-IN-1 95 $ XC6014 Google Assistant integration for smart control 4K ANDROID MEDIA PLAYER WITH REMOTE Give an older TV a second life with modern smart TV features, instead of sending it to the curb. Built-in screen mirroring. $ ONE CABLE. FULL DESKTOP SETUP. 4K USB-C TO MULTI-PORT LAPTOP DOCKING STATION Turn a single USB-C port into a full desktop setup because one screen is never enough & everything stays connected. ONLY 24 119 XC5914 GIVE YOUR TV A SECOND LIFE 4K ONLY 95 $ ONLY 119 AC1820 AA0408 MAKE USE OF YOUR OLD AV GEAR PERFECT FOR STUDIOS, STAGES, & TECH BENCHES Switchable 720P or 1080P Tests continuity and grounding COMPOSITE AV TO HDMI MINI CONVERTER 1080P SCALER Upscales analogue composite video to HDMI output for modern displays. 5V 1A via USB. $ ONLY MULTI-STYLE CABLE TESTER Test a wide range of connectors, including XLR, TRS, RCA, HDMI, USB, RJ45, Speakon, Powercon, and more. DUAL-BAND 34 95 $ LT3306 WI-FI THAT REACHES YOUR WORKBENCH Built-in USB-powered Amplifier Ultra thin, lightweight design with grey fabric for a stylish look. Mount using supplied double-sided tape or push-pins. 199 YN8972 BLENDS INTO YOUR DÉCOR ULTRA THIN INDOOR UHF/VHF TV ANTENNA ONLY Base & Satellite included AX3000 WI-FI 6 DUAL-BAND MESH NETWORK KIT 2PK Eliminate Wi-Fi dead spots with fast, reliable whole home coverage, right out to the workshop where projects happen. Explore our great range of new products, in stock on our website, or at over 140 stores or 130 resellers across Australia and New Zealand. SEE WHAT DIDN’T FIT ON THE PAGE Scan the QR code or visit: jaycar.com.au 1800 022 888 | jaycar.co.nz 0800 452 922 All 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. Australia New Zealand AU: jaycar.com.au/new-products NZ: jaycar.co.nz/new-products NEED TO TOP-UP YOUR SERVICE AIDS AND ESSENTIALS? GREAT RANGE. GREAT VALUE. In-stock at your conveniently located stores nationwide. 4 2 1 3 5 BUY IN BULK & SAVE!!! 1 2 Isopropyl Alcohol 99.8% 250ml Spray NA1066 BUY 1+ $13.95 EA. BUY 4+ $12.45 EA. BUY 10+ $10.95 EA. 99.8% 300g Aerosol NA1067 BUY 1+ $15.95 EA. BUY 4+ $13.95 EA. BUY 10+ $12.45 EA. 70% 1 Litre Bottle NA1071 BUY 1+ $21.95 EA. BUY 4+ $19.45 EA. BUY 10+ $17.45 EA. Electronic Parts Cleaning Solution 1 Litre Bottle NA1070 BUY 1+ $15.95 EA. BUY 4+ $13.95 EA. BUY 10+ $12.45 EA. 3 Liquid Electrical Tape 28g Tube, Black NM2836 28g Tube, Red NM2838 BUY 1+ $25.95 EA. BUY 4+ $22.95 EA. BUY 10+ $20.45 EA. 4 5 175g Aerosols Contact Cleaner Lubricant NA1012 Electronic Cleaning Solvent NA1004 BUY 1+ $13.95 EA. BUY 4+ $12.45 EA. BUY 10+ $10.95 EA. PTFE Dry Lubricant NA1013 BUY 1+ $15.95 EA. BUY 4+ $13.95 EA. BUY 10+ $12.45 EA. J-B Weld Products J-B Weld Epoxy 28g NA1518 BUY 1+ $24.95 EA. BUY 4+ $22.45 EA. BUY 10+ $19.95 EA. SuperWeld Extreme 15g NA1539 BUY 1+ $16.50 EA. BUY 4+ $14.50 EA. WaterWeld 57g NA1532 BUY 1+ $25.95 EA. BUY 4+ $22.95 EA. BUY 10+ $20.45 EA. Shop at Jaycar for even more service aids & essentials: • Adhesives & Insulation Tapes • Solder & Soldering Aids • Wire & Heatshrink Tubing • Fasteners & Cable Ties • Ultrasonic Cleaners • Tools & Workbench Accessories 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. As a test, I programmed the left button of the keyfob with the code “N,0,18”, and sure enough, the Micromite received the desired value of 72, and switched on and off the attached device. I also successfully coded the numerical 2 and 7 buttons, with the following code translations: 2 button code 64 → 01000000 → 00000010 = 2 → “N,0,2” 7 button code 224 → 1110000 → 00000111 = 7 → “N,0,7” It is much, much easier to carry the keyfob in your pocket than the Jumbo Remote! Hope this proves helpful to someone. Ian Thompson, Duncraig, WA. Comment: we noted that Micromite behaviour in the project article. We suspect it’s the Micromite that’s reversing the bit order, since with the Keyfob’s order, the 2 button maps to code 2 and the 7 button maps to code 7. to stop publishing audio projects anytime soon, although we may occasionally take a break from them. Majestic Loudspeaker project enjoyed I built the May 2025 RGB LED Clock (siliconchip.au/ Article/18126) and wanted to update the firmware, but I only had a PICkit 3 programmer, and it will not work with MPLAB IDE/IPE 6.25. It will work under MPLAB 6.20, but then the chip itself is not available in the list of chips that can be programmed. I saw your review of the newer PICkit Basic in the September 2025 issue, and I thought this might be the way forward, so I ordered one. Lo and behold, I have been dragged screaming into the 21st century. I’m only 86, so I guess I might have a few years left to use it. Thank you for publishing a great magazine. Jack Holliday, Nathan, Qld. Comment: The PICkit Basic is a good option for programming newer PIC and AVR chips. It’s a lot faster than the PICkit 3, too. We recently used a PICkit 3 to program a PIC32MZ with 2MiB of flash memory because our other programmers were elsewhere, and it took over a minute for the programming to complete! The PICkit 4/5, SNAP and PICkit Basic can SC program such a chip in around 10 seconds. I’ve been meaning to contact you for some time to thank you for an old project that I completed about three years ago: the Majestic Speakers (June & September 2014 issues; siliconchip.au/Series/275). It ended up being a big effort, as I had to get the Jarrah veneer from a timber supplier and then cut up the panels on a table saw at my work. I’m absolutely thrilled with the sound. I’m into vintage hifi restoration, hence all the equipment in the photo above. The Jarrah veneer has had four coats of satin clear varnish spray. This is the second set of speakers that I’ve built myself. I’m planning on feeding them eventually with an Ultra Low Distortion (Ultra-LD) Mk.3 amplifier, which is in a case with large illuminated VU meters. I think a nice combination and all designed by Silicon Chip. Thanks again, and keep those audio projects coming. Alby Judge VK6ALB, Martin, WA. Comment: we still use the prototypes with an Ultra-LD Mk.1 amplifier (not the best we’ve published, but still pretty good) and we agree that they sound great! We are unlikely siliconchip.com.au Praise for Reciprocal Frequency Counter Thanks for sending a replacement OLED screen as the one I received with the Reciprocal Frequency Counter kit I ordered would not light up (from the July 2023 issue; siliconchip.au/Article/15863). I have installed the new one and am very happy with the Counter’s operation. I used the Counter for another project I am working on, the Soldersmoke Challenge Direct Conversion Receiver (40m/7MHz receiver). I use the Frequency Counter to monitor the receiver’s oscillator. It works well, and I recommend it. Robert Farrugia, Holsworthy, NSW. PICkit Basic programmer supports newer chips Australia's electronics magazine February 2026  15 Image source: https://pixabay.com/photos/intel-8008-cpu-old-processor-3259173/ T o r y of t s i h he intel Pa rt 1 b y D r D avid Mad K3D V , n o d is SM Intel is (or some would say was) one of the world’s most influential and largest manufacturer of computer chips, including microprocessors. That includes the central processing units that power a large portion of modern computers and related devices. S tarting with the world’s first microprocessor in 1971, which sparked the personal computer revolution, Intel grew to a market capitalisation of US$509 billion in 2000 ($930 billion in today’s money). Today it sits at around US$188 billion and fluctuating, while facing AI challenges, serious competition and the legacy of management deficiencies, leading to failures to innovate, among other problems. Intel is currently building new foundries in the United States but still has management challenges after a rocky few years. The founding of Intel Fairchild Semiconductor was founded in 1957 by the “traitorous eight” engineers from Shockley Labs, who were dissatisfied with the way Shockley ran it. Two of those eight were Gordon Moore (famous for Moore’s Law) and Robert Noyce, the co-inventor of the integrated circuit (see Fig.1). Moore and Noyce left Fairchild to 16 Silicon Chip found Intel on the 18th of July 1968. Another Fairchild employee, Andy Grove, also left and joined Intel on the day of its incorporation, although he was not a founder. He helped get Intel’s manufacturing operations started and move Intel’s focus from memory to CPUs in the 1980s, establishing it as the dominant player in the market. In addition, investor Arthur Rock provided US$2.5 million in funding (equivalent to US$23.3 million today or AU$35.5 million). The new company was originally proposed to be named Moore Noyce, but they decided it was best to avoid the “more noise” pun, which is understandable for an electronics company. It was named NM Electronics initially, but after a few weeks, was renamed to Intel, which is derived from “integrated electronics”. Intel was already a trademark of the hotel chain Intelco, so they also had to buy the rights to that name. Intel’s first headquarters was in Mountain View, California (it is now in Santa Clara, California). Its first Australia's electronics magazine 106 employees are shown in Fig.2 (in 1969). Noyce and Moore left Fairchild because they saw the potential of integrated circuits (ICs) and wanted to create a company centred on their research and production. For more on Fairchild and the traitorous eight, see our articles on IC Fabrication in the June-August 2022 issues (siliconchip. au/Series/382). They had become dissatisfied at Fairchild because they felt it was not reinvesting enough in research and development. They felt Fairchild wasn’t growing enough, were dissatisfied with the administrative workload, and stated that it no longer had a hands-on creative culture like it used to have. They also wanted to standardise the mass-production of ICs. Specifically, what they wanted to standardise was a manufacturing process for chips that could be widely adopted, was cost effective, scalable and could be applied to many different chip designs. siliconchip.com.au Fig.1: Andy Grove, Robert Noyce and Gordon Moore in 1978. Source: www.flickr.com/photos/8267616249 (CC BY-SA 2.0) Noyce invented the first commercially viable monolithic IC (a circuit on a single piece of silicon or other material containing all the circuit’s transistors, resistors, capacitors etc) and licensed Fairchild’s “planar process” for manufacturing it. Thus, the new company was to be based on investing extensively in research into the manufacturing of integrated circuits, with a focus on standardisation of the production processes for the monolithic ICs. Moore’s Law provided an ongoing objective for Intel to strive toward. Moore’s Law was an observation he made in 1965 that the number of components on a chip doubles roughly every two years, a compound growth rate of 41%. Moore’s Law held until roughly 2016, at which time the physical limits of component density were reached. The rapid increase in computing power continues through advanced chip packaging methods, architectures and higher clock speeds. Intel’s striving to fulfil Moore’s Law siliconchip.com.au Fig.2: a photo of Intel’s first 106 employees in 1969. Source: https://intelalumni.org/memorylane allowed for an ongoing reduction in the cost of ICs and computers to consumers. That’s because fitting more components onto one silicon chip means a more powerful device for the same cost or less. Conversely, the cost to producers, including Intel, to continue to manufacture higher and higher component densities increases as it becomes more difficult to make cheaper and faster chips. The hope is that improvements in manufacturing technology and economies of scale reduce the cost enough that chips become both more powerful and also cheaper. Intel processor history overview Intel is mostly identified with lines of microprocessors, although it has created many other products, which we will also discuss. Since Intel has produced such a wide range of processors, its history is complicated and can be hard to follow. An abbreviated timeline of Intel processor release dates is shown in Table Australia's electronics magazine 1 overleaf. Many of these will be discussed in more detail later. Understanding Intel’s history Intel has a complex history, so we have broken it up into its dominant features in every decade. The main features of each decade can be summarised as follows: 1970s invented the microprocessor almost by accident with the 4004; the 8080 derivative launched the microprocessor revolution. 1980s Intel dominated the establishment of the PC era. The IBM PC was released, using the 8088, 80286, 80386 or 80486. Along with clones, it became the dominant PC. 1990s Intel continued to dominate the PC market. Intel and Pentium became household names, helped by the “Intel Inside” advertising campaign. 2000s the NetBurst architecture ultimately failed, losing market share to AMD, which reached 25% in 2006. They clawed back some ground with the Core microarchitecture February 2026  17 diversification, but faced various challenges. 2010s stagnation, delays in the 10nm process node, mobile market failure, AMD catching up. 2020s Taiwan Semiconductor Manufacturing Company (TSMC) technologically overtook Intel. Despite this, Intel still has foundry ambitions and developed hybrid cores. Unlike TSMC, Intel is an integrated device manufacturer (IDM) that designs, manufactures and sells its own chips; Intel wants to become the TSMC of the West. The IDM 2.0 strategy of CEO Pat Gelsinger saw five nodes in four years from 2021 to 2025: Intel 7, Intel 4, Intel 3, Intel 20A and Intel 18A. Now that we’ve given a broad overview, let’s look at Intel’s history in more detail. Table 1: Intel processor families Processor family Release date 4004 1971 8086/8088 1978 80286 1982 80386 1985 80486 1989 Pentium 1993 Pentium Pro 1995 Pentium II 1997 Pentium III 1999 Pentium 4 2000 Core & Core 2 2006 Core i3/i5/i7 (1st-8th gen) 2008-2017 Core i3/i5/i7/i9 (9th-14th gen) 2018-2023 1969-1970s: starting as a memory company Core Series 1 2023 Intel began the decade as the world’s leading memory chip maker and ended it by accidentally igniting the personal computer revolution with the 4004 (1971) and then the 8080 (1974). The 4004 microprocessor was originally just a side project for calculators, but became the company’s future when dynamic random access memory (DRAM) profit margins started to collapse. Core Series 2 2024-2025 Core Series 3 Early 2026 Intel’s first products Intel’s most important early products, which established the microcomputer revolution, were based around five chips or chipsets. These were the 3101 (memory), 1101 (memory), 1103 (memory), 1702 (EPROM or erasable programmable read-only memory) and the 4004 (microprocessor) and its associated chipset. We will now describe each of these chips. 1969: Intel 3101 Intel’s first product was the 3101 Schottky TTL bipolar 64-bit static random access memory (SRAM) chip, released in April 1969. By today’s standards, it had an incredibly small storage capacity, equivalent to just eight characters (64 bits). Nevertheless, it was a remarkable achievement as the company was only established in July 1968. Due to the use of Schottky technology, it was nearly twice as fast as earlier implementations of such chips and was designed for use with computer CPUs. Even though Intel initially wanted to focus on research and development, they were incentivised to produce this chip by Honeywell’s announcement that they would purchase SRAMs from anyone who made them. This triggered a competition among memory manufacturers. Honeywell ended up not using the chips because they wanted more than 64 bits, but Intel’s achievement made it known to the world that Intel was now a serious company, no longer the underdog, and other companies became interested in the 3101. The 3101 was unsuitable for main memory, the dominant form of which at the time was magnetic core memory, which had capacities in mainframes up to around 4MiB (in the IBM 360 model 195). Still, it was suitable where high-speed memory devices were needed, such as for processor registers in minicomputers as offered by Burroughs, Xerox and Interdata. 1969: Intel 1101 Following soon after the 3101 was an even more important product, the 1101 256-bit SRAM chip (Fig.3), which was the first with two key technologies: metal oxide semiconductor (MOS) and silicon gates rather than metal. The MOS technology allowed for higher memory capacity (more memory per area of silicon) and higher chip densities. It had access times of 1.5 microseconds (1.5μs) and ran at 5V, consuming 500mW. 1970: Intel 1103 The 1103 (Fig.4) was the first commercial DRAM (dynamic random access memory) memory chip with a The difference between SRAM and DRAM SRAM is faster than DRAM while using less power, as it doesn’t need constant refreshing to maintain data, but it is more expensive and has a lower capacity per chip than DRAM. On the flip side, DRAM is cheaper and has a higher capacity per chip, but it uses more power and is slower than SRAM as it needs to be constantly refreshed. Both types of memory are volatile, meaning they lose their data when power is removed. Fig.3 (top): Intel’s first really successful product, the 1101 256-bit SRAM chip. Source: www. cpu-zone.com/1101.htm Fig.4 (bottom): Intel’s first DRAM chip, the 1103 introduced in 1970. Source: https://w.wiki/ GYXb (CC BY-SA 4.0) 18 Silicon Chip Australia's electronics magazine Fig.5: the three-transistor memory cell was invented in 1969 by William Regitz and colleagues at Honeywell. Original source: https://w.wiki/GYJp (GNU FDL v1.2) siliconchip.com.au capacity of 1024 bits or 128 extended ASCII characters. It had a sufficiently high capacity and low enough cost that it began to replace magnetic core memory. By 1972, it was outselling all other types of memory combined due to costing less and being smaller than core memory. The chip was discontinued in 1979. It was used in computers such as the HP 9800 series, Honeywell minicomputers and the PDP-11. The actual three-transistor dynamic memory cell configuration shown in Fig.5 was invented by Honeywell, who asked the fledgling Intel to manufacture it. It was later also manufactured by National Semiconductor, Signetics and Synertek. 1971: Intel 1702 The first EPROM chip was developed by Dov Frohman at Intel – see Figs.6 & 7. It had 2048 bits of memory that could be erased with UV light and rewritten electrically. It was revolutionary because, before then, “firmware”, the most basic instructions for a computer or similar device to boot, had to be in the form of hardwired logic that was difficult or impossible to change. Intel offered another cheaper version of this chip, which was ‘write once’ and could not be erased. The only differences were that it did not have an expensive transparent quartz window for UV erasure, and it came in a plastic rather than ceramic package. Today, flash memory has replaced EPROM memory for things like firmware, but the 1702 was an important development as it made prototyping new products much easier, along with allowing product updates. Fig.6: a demonstration of the 1702 chip in 1971, using its stored information to display the Intel logo on an oscilloscope. Source: https:// timeline.intel.com/1971/the-world’sfirst-eprom:-the-1702 Fig.7: the Intel 1702 had a transparent window through which the contents could be erased by UV light and then electronically rewritten. Source: https://timeline.intel.com/1971/theworld’s-first-eprom:-the-1702 1970s: the microprocessor revolution Intel’s and the world’s first microprocessor would not have happened at the time had it not been for a request from the Japanese Busicom calculator company. The Busicom calculator In 1969, Busicom asked Intel to design a set of chips for their proposed electronic calculator. At the time, calculators contained large numbers of discrete components and complex wiring, so they wanted to reduce the cost by using a dedicated chipset. The siliconchip.com.au Fig.8: a Busicom 141-PF / NCR 18-36 circuit board with chips Intel developed for it. Note the blank space for the optional 4001 ROM for the square root function. Source: Nigel Tout, http://vintagecalculators.com Busicom engineers designed a calculator that required 12 ICs and asked Intel to make these custom chips. Ted Hoff at Intel, aided by Federico Faggin and Stanley Mazor, came up with a much more elegant design needing only four chip types Australia's electronics magazine containing ROM (read-only memory), RAM (random-­access memory), a shift register and what was to become the 4004 microprocessor. These chips were developed, produced and sent to Busicom in January 1971, and they had exclusive rights to them. February 2026  19 The 4004 microprocessor was a single silicon chip that contained all the basic functional elements of a computer’s central processing unit (CPU). Until the 4004, CPUs had to be fabricated using multiple individual components at much greater cost and complexity. The resulting calculator was the Busicom 141-PF, also marketed as the NCR 18-36 (see Fig.8). An optional ROM chip was available to provide a square root function. In common with other calculators of the era, it printed the results rather than displaying them on a screen. This was an important moment in the history of calculators because, at the time, calculators had to have their functionality designed into hardware, which meant every calculator required extensive customised hardware. The new Intel microprocessor and ROM allowed new designs to be made simply by changing the programming of the microprocessor via ROM. The calculator used four 4001 ROM chips, two 4002 RAM chips, three 4003 shift registers and one 4004 microprocessor. More about this chipset later. At the same time as the Intel developments, Busicom commissioned Mostek to produce a ‘calculator on a chip’, which resulted in an even lower chip count than the Intel solution. The chip developed and released in November 1970 was the Mostek MK6010, but that’s another story. In mid-1971, Busicom asked Intel to lower the chip prices, which resulted in Intel renegotiating the contract such that Busicom gave up their exclusive rights, enabling Intel to sell the chips. Then, in November 1971, Intel announced the release of the MCS-4 chipset family based on the chips developed for Busicom. 1971: the beginning of the microprocessor revolution On the 15th of November 1971, Intel commercially released the 4004 microprocessor that they had developed for Busicom and licensed back to themselves. The Intel 4004 was a revolutionary product for the computer industry. It was designed to be affordable, easy-touse and accessible to a wide variety of computer designers. Early microprocessors such as the 4004 were not initially intended for general-purpose computing, but to run embedded systems such as calculators, cash registers, computer games, computer terminals, industrial robots, scientific instruments etc. In addition to the Busicom calculator mentioned above, it was used in Busicom automated teller and cash machines, the Intellec 4 microcomputer from Intel (Fig.9) to support software development for the 4004, a prototype pinball machine by Bally, and the Pioneer 10 spacecraft. The software to run such systems could be developed on the Intellec 4 and then permanently programmed into ROMs such as the 4001 during manufacture, or burned into EPROMs such as the 1702 (which could be erased and updated). The 4004 cost US$60 at the time, which in today’s money would be US$501 or AU$774. The MCS-4 (see Fig.9: the Intellec 4 microcomputer for software development for the 4004, available to developers only. It was programmed via front panel switches or an optional terminal interface. Source: https://w.wiki/GYJr (CC BY-SA 3.0) 20 Silicon Chip Australia's electronics magazine Fig.10) included the 4001 ROM, 4002 RAM and 4003 I/O chips that together formed the basic elements of a complete computer. The ~$750 price is similar to that of a high-end (consumer) CPU today. The 4004 contained 2300 transistors and was fabricated using a 10-micron (10μm) process. It could execute 60,000 instructions per second with a 740kHz clock speed and a 4-bit architecture. It could address 640 bytes of RAM and up to 4kiB of ROM – see Fig.11. The specifications of the MCS-4 chipset chips were: 4001 a 256 × 8-bit (256 byte) ROM. 4002 a 4 × 20 × 4-bit (40 byte) DRAM. 4003 an I/O chip with a 10-bit static shift register, serial and parallel outputs. A static shift register comprises flip-flops that store and shift binary data. 4004 the microprocessor. Using this chipset, a fully expanded 4004 system using sixteen 4001s could have 4kiB of ROM and sixteen 4002s for a total of 640 bytes of RAM, plus an unlimited number of 4003s for I/O. The most powerful 4004 system? The most powerful Intel 4004 system, called Linux/4004, was built by Dmitry Grinberg in 2024. It was created to use “ancient” 4004 hardware merged with a modern Linux operating system. It is a testament to the powerful and flexible nature of the 4004 chip, which was originally intended to power a calculator, but is not exactly practical. The system took 4.76 days to boot a stripped-down Linux kernel to the Fig.10: the Intel MCS-4 chipset. Source: https://en.wikichip.org/wiki/ File:MCS-4.jpg siliconchip.com.au Fig.11: the chip layout (a drawing, not a photograph) of the 4004 processor. Source: https://w.wiki/ GYJq (CC0 1.0) 4004 image source: https://w.wiki/GYZY 8008 image source: https://w.wiki/GYZZ i960 image source: https://w.wiki/GYK8 Fig.12: the die of the Intel 8008, their first 8-bit CPU. Source: https://x.com/ duke_cpu/status/ 1980293005644107812 Fig.13: an Intel i960 die (80960JA). Note the large cache memory banks (rectangular grids); the actual core is pretty small since it’s a RISC processor. Source: https://w.wiki/GYK9 (CC BY 3.0) siliconchip.com.au Australia's electronics magazine February 2026  21 Fig.14: an 8080 chip made by Intel. Source: https://w.wiki/GYJy (CC BY 4.0) Fig.15: the Altair 8800 computer was sold as a kit, and also has an optional 8-inch floppy drive. It popularised the use of the Intel 8080 processor. Source: https://americanhistory.si.edu/collections/ object/nmah_334396 command prompt. It could perform rudimentary mathematical fractal calculations of the Mandelbrot set. A full description of the project can be found at siliconchip.au/link/ac9t and there is a video on it at https://youtu. be/NQZZ21WZZr0 After the 4004 The success of the 4004 led to the development of the 8008 and the 8080 CPUs, which established Intel as the world leader and led to great expansion of the company in the 1970s, 1980s and 1990s. 8008 the 4004 led to the development of the 8008 in April 1972. It was the first 8-bit microprocessor and could address 16kiB of memory. It was manufactured but not designed by Intel. CTC (Computer Terminal Corporation) designed it for use in their Datapoint 2200 programmable terminal, but Intel licensed the design for use in other products. The 8008 was discontinued in 1983. Its clock speed was 500-800kHz and it used 10-micron technology, with 3500 transistors. The 8008 is most famous for being the microprocessor used in the first enthusiast personal computers: the SCELBI (US, 1974), the Micral N (France, 1973) and the MCM/70 (Canada, 1973). It was also used in the HP 2640 computer terminals. 8080 the 8080 followed in 1974 (Fig.14). It was originally conceived for embedded systems, but it was broadly adopted and remained in production until 1990. Made with a 6 micron (6μm) process node, it had a clock rate of 2-3.125MHz and was an 8-bit processor but had the ability to execute 16-bit instructions. It could address 64kiB of memory. 22 Silicon Chip A variety of support chips were available for it. It had about 6000 transistors and could execute several hundred thousand instructions per second. It was used in the first commercially successful personal computers, like the Altair 8800 (see Fig.15), and other S-100 bus systems running the CP/M operating system. 8085 the 8085, introduced in March 1976 and discontinued in 2000, was the successor to the 8080 and Intel’s last 8-bit processor. It was compatible with the 8080 but had the advantage of only needing to be supplied with one voltage, not three like the 8080, making system development simpler. It ran at a clock speed of 3MHz, 5MHz or 6MHz, used a 3 micron process node and had 6500 transistors. It was not widely adopted in microcomputers because the Zilog Z80 chip (1976-2024) was introduced, which took over much of the 8-bit market (eg, running the Osborne 1, TRS-80 and ZX Spectrum). However, the 8085 was used as a microcontroller and in video terminals like the VT-102. 8086 in 1978, Intel introduced the 8086, its first 16-bit processor with 29,000 transistors, built on a 3.5 micron process (switching to 2 microns in 1981) – see Fig.16. It extended the 8080 architecture, introduced segmented memory addressing, ran at up to 10MHz and could support 1MiB of RAM. It had a simple two-stage pipelining unit to improve performance. It laid the foundation of the x86 instruction set family of processors. This processor, along with dominance of the memory chip market, paved the way for the commercial personal computer boom. The x86 instruction set The x86 instruction set that’s still widely used today was introduced with the 8086. It became standardised with the release of the 8088 processor thanks to its use by IBM in their open PC architecture in 1981. x86 has had many updates over the years, but today’s processors can still run code that was written back in the late 1970s. This does not mean that such code will run on a modern operating system like Windows 11, but that is a restriction of Windows, not the processor itself. It is possible to boot Microsoft DOS from 1981 on a current x86 CPU. There would be problems such as a lack of USB and other driver support, and a lack of compatibility with a modern UEFI (unified extensible firmware interface) BIOS. There is a video of a system with a 2016 Intel Celeron N3450 CPU booting a 45-year-old version of DOS at https://youtu.be/BXNHHUmVZh8 (the Celeron name was generally applied to a cut-down or simplified Pentium processor). Microsoft also played a role in the standardisation of x86 by supporting a wide range of hardware that used x86. With time, new instructions have been added Fig.16: an 8086 chip in a ceramic dual-inline package (DIP). Source: https://w.wiki/GYK4 (CC BY-SA 4.0) Australia's electronics magazine siliconchip.com.au to x86, but the old ones have been kept to ensure compatibility. Intel and AMD, who both make x86-compatible processors, have formed an alliance to standardise future instructions to ensure their consistent implementation across future products from both companies. Competing instruction sets include ARM, MIPS and RISC-V. Backward compatibility is important because there are enormous amounts of commercial, financial, industrial, military, medical and domestic software written for old processors that may still be in use. Some of this software, which can be decades old, runs on DOS, including accounting software, payroll systems, programmable logic controllers, CNC machines and retro games. This is one reason that attempts to replace the x86 instruction set have not generally been successful, although ARM has made some inroads. Emulation (where software running on one processor can interpret instructions from a different set) can help to ease the transition. From 2020 to 2023, Apple moved away from the x86 architecture as they transitioned from Intel microprocessors (which they used since 2006) to their own designs based on the ARM architecture. Apple’s reasons were they wanted a common technology across all their platforms, better performance per watt and they wanted to integrate all components on a single chip (see also the section later on the stagnation of Intel’s innovation). Over the years, Intel has developed extensions to the x86 instruction set, including: ● MultiMedia eXtensions (MMX) ● the Streaming SIMD (single instruction, multiple data) Extensions, which superseded MMX: SSE, SSE2, SSE3 and SSE4 ● Advanced Vector eXtensions (AVX, AVX2 and AVX-512) ● Advanced Encryption Standard – New Instructions (AES-NI) ● Software Guard eXtensions (SGX) ● Trusted eXecution Technology (TXT) ● Transactional Synchronisation eXtensions (TSX) ● haRDware RANDom number generator (RDRAND) ● Carry-Less MULtiplication for cryptography (CLMUL) siliconchip.com.au Table 2 – Intel’s process node names (only consumer CPUs listed) Year Process Name Chips made # transistors 1972 10μm 10μm 4004 2.3k 1974 8μm 10μm 4040 3k 1976 6μm 6μm 8080 6k 1977 3μm 3μm 8085, 8086, 8088 29k 1979 2μm 2μm 80186 134k 1982 1.5μm 1.5μm 80286, 80386 275k 1987 1μm 1μm 80386, 80486 (up to 33MHz) 1.2M 1989 800nm 800nm 80486 (up to 100MHz) 1.3M 1991 600nm 600nm 80486 (100MHz), Pentium (60200MHz) 3.1M 1995 350nm 350nm Pentium (120-200MHz), Pentium MMX (166-233MHz), Pentium Pro (150-200MHz) 5.5M 1997 250nm 250nm Pentium Pro, Pentium II (233450MHz), Pentium III (450600MHz) 9.5M 1999 180nm 180nm Pentium III (500-1133MHz), Pentium 4 (NetBurst, 1.3-1.8GHz) 42M 2001 130nm 130nm Pentium III (1.0-1.4GHz), Pentium 4 (NetBurst, 1.6-3.4GHz) 125M 2003 90nm 90nm Pentium 4 (NetBurst, 2.4-3.8GHz), Pentium M 169M 2005 65nm 65nm Final Pentium 4, Core, early Core 2 Solo / Duo 291M 2007 45nm 45nm Late Core 2 Duo / Quad, Core i3/ i5/i7 (1st gen) 731M 2009 32nm 32nm Core i3/i5/i7 (1st gen refresh & 2nd gen) 1.17B 2011 22nm 22nm Core i3/i5/i7 (3rd & 4th gen) 1.4B 2014 14nm 14nm Core i3/i5/i7/i9 (5th to 9th gen) 3B 2019 10nm 10nm Core i3/i5/i7/i9 (10th & 11th gen) 4.1B 2021 10nm+ Intel 7 Core i3/i5/i7/i9 (12th & 13th gen) 21B 2023 5nm Intel 4 & 3 Core i3/i5/i7/i9 (14th gen), Core Ultra 1 30B 2024 3nm Intel 20A Core Ultra 2 ~45B 2025 2nm Intel 18A & 14A Core Ultra 3 ~80B ● x86-64, a 64-bit version of x86 that allows, among other things, access to more than 4GB of RAM (developed by AMD but also implemented by Intel) ● Advanced Performance eXtensions (APX) Process nodes Throughout Intel’s history, it was shrinking the feature size of its chips, achieving higher numbers of transistors and higher component densities. We will divert from the history for a moment to describe process nodes, an essential part of understanding subsequent processor development. Australia's electronics magazine A process node (or technology node, which means the same thing) is a term used in semiconductor manufacturing representing a specific generation of chip technology. It was traditionally named based on the size of a transistor gate, which continued to shrink while Moore’s Law still applied. As it is difficult to shrink transistors much more than they are now, the names no longer correspond to any particular physical size, and are more of a marketing term representing performance and density increases, which continue due to 3D packaging and other techniques. February 2026  23 Intel 4004 Architecture D0-D3 bidirectional Data Bus Data Bus Buffer 4 Bit internal Data Bus Temp. Register Register Multiplexer Instruction Register Stack Multiplexer Flag Flip Flops ALU Stack Pointer Instruction Decoder and Machine Cycle Encoding Index Register Select Accumulator Program Counter Level No. 1 Level No. 2 Level No. 3 Address Stack Decimal Adjust 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Scratch Pad Timing and Control ROM Control RAM Control Test Sync Clocks CM ROM CM RAM 0-3 Test Sync Ph1 Ph2 Reset Fig.17: the microarchitecture of Intel’s (and the world’s) first microprocessor, the 4004 from 1971. Source: https://w.wiki/GYJu (GNU FDL v1.2) 32 KB Instruction Cache (8 way) 128 Entry ITLB Shared Bus Interface Unit 128 Bit 32 Byte Pre-Decode, Fetch Buffer Instruction Fetch Unit 6 Instructions 18 Entry Instruction Queue Complex Decoder Microcode Simple Decoder Simple Decoder 4 µops 1 µop Simple Decoder 1 µop 1 µop Shared L2 Cache (16 way) 7+ Entry µop Buffer 4 µops Register Alias Table and Allocator 4 µops 4 µops 96 Entry Reorder Buffer (ROB) Retirement Register File (Program Visible State) 256 Entry L2 DTLB 4 µops 32 Entry Reservation Station Port 0 ALU SSE Shuffle ALU ALU 128 Bit FMUL FDIV Intel Core 2 Architecture SSE Shuffle MUL 128 Bit FADD Internal Results Bus Port 3 Port 5 Port 1 ALU Branch SSE ALU Store Address Port 4 Store Data Port 2 Load Address Memory Ordering Buffer (MOB) 128 Bit Store 128 Bit 32 KB Dual Ported Data Cache (8 way) Load 256 Bit 16 Entry DTLB Fig.18: the microarchitecture of the much more advanced Intel Core 2 processor from 2006. Source: https://w.wiki/GYJv (GNU FDL v1.2) 24 Silicon Chip Australia's electronics magazine The number of atoms across the smallest dimension of a transistor of the Intel 18A process node (representing 18Å or 1.8nm) is estimated to be 180, but because of the 3D nature of the transistor, the overall number is estimated to be thousands. This is currently the minimum number required for reliable function. That might not be improved on for a long time, if ever, for practical devices as adverse quantum mechanical effects like electron tunnelling are already a concern with the 18A process node. But technology always develops in unexpected ways... By way of comparison, the smallest process node described by Samsung is 2nm or 20Å. The distance between centres of silicon atoms in a crystal lattice is 0.235nm or 2.35Å. Commonly used terms for Intel fabrication processes are listed in Table 2. The 18A process node (1.8nm) is what Intel is focusing on for the future. It will be produced at its Arizona and Oregon foundries, which are its most advanced in the world and will lead the way to the “one trillion transistor laptop”. This process node incorporates all the above technologies and is the culmination of the so-called 5N4Y (five nodes in four years), which was former CEO Patrick Gelsinger’s turnaround strategy, announced in 2021. Gelsinger was asked to leave in December 2024 when the board felt improvements were not being made fast enough (his replacement has had some controversies). The 5N4Y plan nodes were: Intel 7 (~10nm): the first use of their Enhanced SuperFin transistors. Intel 4 (~5nm): produced with extreme ultraviolet (EUV) lithography and moving to chiplets/tiles and associated interconnect technologies, like Foveros and EMIB (more on these later). Intel 3 (~5nm): with improved performance per watt. Intel 20A: A marks the move to Angstrom-­ b ased measurements. It didn’t go into full production, but led the way to the implementation of Ribbon FETs and PowerVias in 18A (more on these later). Intel 18A: the current process node with the first processor being the Core Ultra series 3 (Panther Lake) and the second to be the Xeon 6+ (Clearwater Forest). siliconchip.com.au Microarchitectures Microarchitecture (or μarch) is the particular way a processor’s internal hardware (pipelines, execution units, caches etc) is designed and organised to implement a given instruction-set architecture (ISA) such as x86. It is typically illustrated with pipeline or block diagrams, like Figs.17 & 18. Intel re-uses microarchitectures across multiple processor generations and models. Most (but not all) major new Intel processor families introduce a new or significantly revised microarchitecture. A new microarchitecture appeared every 2-4 years, while new processor series (new brand names or model numbers) were released every 12-18 months; this was called their tick-tock model. Examples of Intel microarchitectures are shown in Table 3. Let’s now look at more recent eras of Intel products. The 1970s PC boom Intel’s processors of the 1970s had a great cultural impact and were a leap forward for microcomputing via the hobbyist PC boom of that era. They were responsible for democratising computing and sparking a global DIY computer revolution, which ultimately led to the widespread commercial development of microcomputers. As mentioned, the 8080 was released in 1974. It was the first truly affordable 8-bit CPU that a hobbyist could purchase. It cost US$360 in single units, but kit manufacturers like MITS, the creators of the Altair 8800, could get them for US$75 (equivalent to AU$757 today) in volume and sell them via mail order. The chip was small, relatively inexpensive and well-documented, so it was something hobbyists could make something with. Thus, computing moved out of the corporate lab and into garages and bedrooms. The Altair 8800 featured on the cover of Popular Electronics magazine in 1975 and, after that, 4000 were sold in weeks at US$439 (AU$4000 today) pre-assembled or US$297 (AU$2750 today) as a kit. Hobbyists saw the chip and the Altair computer that used it as a ‘blank canvas’. After seeing the magazine, Bill Gates and Paul Allen wrote Altair BASIC in 1975 as Microsoft’s (then called MicroSoft) first product. They used a PDP10 mainframe running an 8080 emulator. Gates released the source code siliconchip.com.au Table 3 – Intel microarchitectures from 1993 to the present Microarchitecture Years Processor families or brands P5 1993-1997 Pentium (60–200 MHz), Pentium MMX P6 1995-2003 Pentium Pro, Pentium II, Pentium III, Celeron (early), Pentium II Xeon, Pentium III Xeon NetBurst 2000-2007 Pentium 4, Pentium D, early Xeon Core 2006-2008 Core 2 Duo / Quad (Yonah → Penryn) Nehalem 2008-2010 Core i3/i5/i7 (1st gen) Sandy Bridge 2011-2012 Core i3/i5/i7 (2nd & 3rd gen) Ivy Bridge 2012-2013 3xxx series (22nm shrink + tweaks) Haswell → Broadwell → Skylake → … → Coffee Lake → Comet Lake → Rocket Lake 2013-2021 4th gen → 11th gen Core (various), Skylake derivatives used for six consecutive generations (2015-2021) Alder Lake (Golden Cove + 2021-2023 Gracemont cores), Raptor Lake 12th, 13th & 14th gen Core Meteor Lake 2023 Series 1, chiplet-based design Arrow Lake / Lunar Lake 2024-2025+ Series 2 (15th Gen) in April 2025 to mark Microsoft’s 50th anniversary. Steve Wozniak was also inspired by the Altair, which motivated him to design his own computer, the Apple I kit, released in July 1976. It used fewer parts than the Altair. He demonstrated it at the Homebrew Computer Club and shared the design and software for free, but the basic kit was sold for US$666.66 or AU$5800 today. It did not use an Intel processor, but a MOS 6502 instead. The Homebrew Computer Club held Silicon Valley garage meetings where hobbyists shared 8080 designs and code. Intel provided free datasheets, reference designs and even engineers who attended. Their slogan was “Build it. Share it. Improve it.” Other hobbyist computers of the 1976-1979 era were the IMSAI 8080, with the Intel 8080, and computers inspired by the 8080, like the TRS-80 (1977) that used the Zilog Z80 (which was 8080 compatible), and the Commodore PET (1977), which used the MOS 6502 like the Apple I. Intel provided open documentation for its products and encouraged chips such as the Z80 which, being compatible with the 8080, helped establish the 8080 ecosystem. This led to the dominant x86 architecture, which is still in widespread use today. Hobbyist computer magazines supported this new technology; magazines like BYTE, Creative Computing, Kilobaud Microcomputing and Dr Dobb’s Journal. During this period, there were price drops of the 8080, 8085 and 8088 chips, which led to mass adoption of microprocessors. By 1980, hundreds of thousands of hobbyists worldwide were programming in assembly language, swapping floppies and “building the future”. In 1978, Intel released the first electrically erasable programmable readonly memory (EEPROM), the 2816, which had a capacity of 16kib (2kiB). It is non-volatile, meaning it retains its memory when the power is switched off, but it can be erased and rewritten when desired without needing a UV light source, as the earlier 1702 EPROM did. It is considered a major achievement in the history of computing, allowing easy in-system reprogramming for both hobbyists and commercial users. The IBM PC is introduced In 1979, the 16-bit 8088 CPU with 29,000 transistors was introduced as a lower-cost version of the 8086 (see Fig.19). It was the heart of the original Fig.19: an original Intel 8088 processor. Source: https://w.wiki/GYJw (CC BY-SA 4.0) Australia's electronics magazine February 2026  25 IBM Personal Computer, which was released on the 12th of August 1981 (see Fig.20). Even though it was a 16-bit processor, external communications were via an 8-bit data bus for cost efficiency, but it could address 1MiB of memory with its 20-bit memory address bus. It was designed in Israel (as many of Intel’s processors have been). IBM’s decision to use the 8088 led to the standardisation of the x86 instruction set, because IBM’s open architecture approach encouraged cloning of the computer and the development of compatible expansion cards, which led to the rapid expansion of the Intel and x86 ecosystem. Also, IBM insisted on a second source for their PC chips, leading to Intel licensing their designs to AMD. AMD continues to make Intel-­ compatible CPUs to this day. It had simple pipelining in the form of a prefetch queue that read instructions from memory before they were needed. This enabled a performance increase. An 8087 mathematical coprocessor was available to complement the 8086 or 8088, which dramatically improved the speed of floating-­ point arithmetic operations. 1980s: dominating the PC era A low point of the 1980s for Intel was being forced out of the DRAM market by Japanese competition. Intel’s DRAM market share had fallen from over 80% in the 1970s to 2-3% by 1985, and they decided to withdraw from the market and fully focus on microprocessors. Intel bet everything on the x86 family. The 80386 (1985), in particular, turned the IBM PC standard into a near-monopoly and made Intel the indispensable heart of personal computers. The IBM PC and its clones dominated the PC market and cemented the legacy of the x86 instruction set that is used in almost all Intel and many competing processors (eg, from AMD) to this day. By the end of the decade, x86 processors generated almost all the company’s profit, and Intel processors dominated the PC market. Other processors they developed in this era were: iAPX 432 The iAPX 432 (1981-1985) was Intel’s ambitious but ultimately unsuccessful first attempt at a true 32-bit microprocessor. It comprised two chips (the 43201 general data processor and 43202 interface processor), was not based on the x86 architecture, and represented a radical departure from Intel’s prior designs. The 432 was designed from the ground up to support high-level languages like Ada directly in hardware, with features like object-oriented memory management, ‘garbage collection’ (a means to manage and recover unused memory) and capability-­based addressing (a memory and resources access model in which access is granted via tokens rather than raw addresses). These ideas were decades ahead of their time. This allowed modern operating systems to be implemented with significantly less code. However, technological limitations resulted in a performance roughly one-quarter that of the 80286, despite its advanced architecture. Compounding the problem, the 432 was not backward compatible with any existing Intel processor, alienating developers accustomed to the 8086/8088 ecosystem. These factors, combined with its high cost and complexity, led to its commercial failure. Fig.20: the original IBM PC from 1981, built around the Intel 8088. Source: https://w.wiki/ GYJx (CC BY-SA 3.0) 26 Australia's electronics magazine 80286 The 16-bit 80286 microprocessor was introduced in 1982 (Fig.21). It added ‘protected mode’ operation, enabling it to address up to 16MiB of memory instead of the 1MiB of the 8088, with improved multitasking capabilities compared to the ‘real mode’ limitations of earlier x86 chips. 16-bit data could be fetched in one bus cycle, while the 8088 required two bus cycles. Clock speeds up to 20MHz were supported, and the ‘286 facilitated more advanced operating systems such as IBM’s OS/2, Windows 3.0, Concurrent DOS, Minix and QNX that supported multitasking and more memory access compared to standard DOS. A disadvantage of ‘286 protected mode was that there was no way to return to real mode without a CPU reset, so standard DOS programs could not be run once the CPU was switched to protected mode. The ‘286 had simple pipelining, allowing the instruction unit, address unit, bus unit and execution unit to work concurrently to improve performance. An 80287 mathematics coprocessor was available. The ‘286 had between 120,000 and 134,000 transistors depending upon the variant, and was built using a 1500nm (1.5μm) process. The direct competitor to the ‘286 was Motorola’s 68000 (“68k”), which was used in the first Apple Macintosh, Commodore Amiga and Atari ST. It was a 32-bit processor with a 16-bit bus, but the ‘286 gave superior real-world benchmarks, and the IBM PC had an open architecture, giving it more software compatibility and therefore more popularity than the 68000. 80386 The 80386 was released in 1985, and came in two versions: the lower-­ priced SX, with a 32-bit internal architecture but a 16-bit external data bus and 24-bit memory address bus; and the DX, which was the ‘full’ version with a 32-bit external bus (Fig.22). It could support up to 4GiB of physical memory and up to 64TiB of virtual memory using advanced segmentation and paging. It was designed specifically with multitasking in mind. It had a simple six-stage instruction pipeline to allow the execution of different phases of certain instructions somewhat in parallel over multiple clock cycles, to siliconchip.com.au Fig.21: an 80286 chip. Source: https://w.wiki/ GYK6 (CC BY-SA 3.0) keep the processor busy at all times. Mathematical co-processors (80387) were available for both versions of the ‘386. It had 275,000 transistors and was built with a 1000nm (1μm) process. A special version produced for IBM, the 386SLC, had a large amount of on-chip cache, with 855,000 transistors. i960 Intel’s i960 (also known as the 80960), sold over 1988-2007, was a major shift away from the x86 architecture toward RISC (reduced instruction set computer) principles, which streamlines the instruction set, theoretically enabling faster execution – see Fig.13. It was mainly used as an embedded processor in military, industrial, and networking systems and achieved great success in niche markets such as laser printers, routers and even the F-22 Raptor stealth fighter. Intel discontinued the i960 in 2007 as part of a legal settlement with Digital Equipment Corporation (DEC) over patent disputes. In exchange, Intel gained rights to DEC’s Strong­ ARM design. 80486 The 80486 (Fig.23) was introduced in 1989. It had a built-in floating-point unit and so did not need an external coprocessor. It also had an inbuilt 8kiB cache, later increased to 16kiB in the DX4 variant, which gave it much better performance compared to the ‘386. It also had a five-stage pipelining architecture, similar to the ‘386 but with a more advanced architecture. Even though the 8088, 8086, ‘286 and ‘386 had instruction pipelining, the ‘486 was the first in which pipelining was tightly integrated. The 486SL variant was optimised for lower power consumption in laptops. It had 1.2-1.6 million transistors dependent upon variant and was only discontinued in 2007. The underside of the AMD version of the 80286, which had a higher clock frequency. Source: https://w.wiki/GYaQ Fig.22: an 80386DX chip. Source: https://w.wiki/GYK7 (CC BY-SA 3.0) The AM386 is a clone of the 80386. Source: https://w.wiki/GYaY Next month That’s all we have space for in this issue. We’ll pick up the rest of the Intel story in the March issue, at the start of the 1990s. That second article will bring us up to date, and then in the final instalment, we’ll look at the current state of microprocessor technology and how Intel plans to remain SC competitive in the future. siliconchip.com.au Fig.23: an exposed 80486 chip die. Source: https://w.wiki/GYKB (CC BY-SA 3.0) Australia's electronics magazine February 2026  27 Mains Power LED Indicator Neon lamps can run from 230V AC with a simple series resistor but they’re pretty dim and flickery. LEDs are much better, but you need this extra circuitry to run them directly from a mains supply. By Julian Edgar & John Clarke T here are many applications where you want to run LEDs from mains power; a pilot light near a switch is the most common. There are numerous simple circuits available to do this – typically they use just a resistor, diode and capacitor. However, these circuits all have a major safety problem: the LED is floating at mains potential above Earth. If it’s being run by one of those common circuits, a person touching the outside of the LED’s plastic envelope is relying on the dielectric strength (insulation) of the plastic LED envelope to avoid getting a shock. That’s why many commercial mains switches have the LEDs mounted in bezels or even hidden within the switch and not accessible. But what if you want the LED to project through the faceplate of a wall switch? No LED manufacturer specifies the dielectric strength of the LED’s plastic envelope, so there’s no guarantee of safety. Our simple and cheap circuit overcomes that problem. It allows the LED to be operated from the mains, but the LED is not floating at high and dangerous voltages. The LED’s current can also be easily set. Also, unlike many approaches that run LEDs off mains power, this circuit protects against surge over-current at switch-on, providing a long life. This project is designed for use with mains house wiring, so it must be installed by an electrician. How it came about This circuit came about because, in the house I am building, I am using a 12V system to operate ventilation hatches. These hatches are operated by a DPDT switch triggering a linear actuator. There is a 10mm high-­ intensity green LED on the wall switch plate to indicate when the ventilation hatch is open. Testing showed this approach to be very effective. The LED is visible from many metres away, and because it projects through the switch plate, it can be seen at quite acute angles. To operate different systems, I am also using mains power switches that need pilot lights. In those cases, I was originally using green neons, and the contrast with the 10mm LEDs was profound. To see if the neon was lit in bright light, you needed to peer closely at the bezel. Even in dull conditions, the neon’s brightness was borderline – only when it was quite dark was the neon brightness adequate. Furthermore, it was impossible to see the neon indicator at any angle other than with the viewer directly in front of the indicator. What was needed was a way of running the same 10mm high-intensity green LEDs as the 12V power indicators, but on mains power. How it works In the circuit diagram (Fig.1), capacitor C1 is the main voltage dropping This shows a 10mm LED being driven by the Mains Power LED Indicator. Unlike a neon indicator, it is visible at acute angles and in bright light. 28 Silicon Chip Australia's electronics magazine siliconchip.com.au The PCB needs to be mounted in an IP65 enclosure with cable glands used on the mains power and LED connections. The length of the LED leads depends on the application. Table 1 – selecting capacitor C1 component that also limits current through the LED. Its capacitance provides an impedance at the mains frequency of 50Hz that is 1 ÷ 2πfC, where f is the frequency and C is capacitance in farads. For a 100nF capacitor, this works out to 31.8kW. Ignoring the effect of the relatively small series 1kW resistor, this limits the mains current for a 230V AC supply to around 7mA. More LED current is available with a larger capacitance – refer to Table 1. The parallel 1MW resistor discharges the capacitor when the circuit is switched off. The supply after the series capacitor and 1kW resistor is full-wave rectified by bridge rectifier BR1 and filtered to a smooth voltage by the 470μF capacitor. Zener diode ZD1 limits the voltage across the capacitor to 4.7V. The 1kW 1W resistor is included since, when power is initially connected to the circuit, the mains voltage could be anywhere in the voltage swing of the 230V AC waveform (up to ±325V DC). At initial power-on, the discharged capacitor will briefly present a short circuit. So, if the voltage is high at power on, the capacitor charging current is limited via the 1kW resistor. The zener diode then conducts and prevents the voltage rising much above its clamping voltage of 4.7V. The initial surge current through the zener diode could be as high as 325mA (325V ÷ 1kW). However, this is only an instantaneous current for the zener, and it can easily withstand that briefly even though its rated maximum continuous current is 212mA for its 1W rating. The LED is protected against surge over-current by being driven via the 150W resistor across the DC supply, that in turn is limited in voltage by the zener diode. Under normal conditions, ZD1 does not conduct. This is because, even when using the largest capacitor value LED current Capacitor C1 value 1.4mA 22nF 3mA 47nF 6.4mA 100nF 9.5mA 150nF 13.7mA 220nF for C1 at 220nF, the current through the LED is 13.7mA and voltage across the 150W resistance is 2V. Adding this to the voltage across the LED (typically 1.8V) gives a value less than the zener voltage. Therefore, the zener conducts only when there is the potentially higher current flow at power-up. Without the zener diode and 150W resistor, the LED surge current would be up to 325mA. Assuming 1.8V across the LED, with this circuit, the maximum LED current is limited to 19.3mA when the zener voltage is at 4.7V. Taking this approach gives a long LED life – something simpler circuits often don’t provide. We minimise the risk of electrocution by ensuring that the LED itself is, at most, only a few volts above mains Neutral. Mains Neutral is tied to Earth on the household property, so typically, it is within a few volts of Earth. 1W resistors are used to achieve the WARNING: MAINS VOLTAGE Fig.1: capacitor C1 is the primary component that drops the mains voltage to the 1.8-3.6V needed to drive LED1. The AC is then rectified, with the surge current at switch-on limited by the 1kW resistor and zener diode ZD1. Importantly, the LED cathode is tied to mains Neutral via BR1 for safety. The small PCB (shown adjacent) uses only a handful of components and is quickly assembled. siliconchip.com.au Australia's electronics magazine This circuit operates at mains live voltages. Do not build it unless you are confident working with mains-powered circuitry. Don’t touch any part of the circuit when it is connected to mains power. Fixed wiring installation must be performed by an electrician. Fig.2: we have simulated the circuit using LTspice, and the simulation file is available to download from siliconchip.au/Shop/6/3314 required mains voltage rating, since ¼W types may have a lower voltage rating (eg, 150V). An LTspice circuit simulation for this project is available to download from siliconchip.au/Shop/6/3314 in case you want to see how it behaves and check the LED current with different capacitor values (see Fig.2). Construction We have created a small PCB, coded 10111251, that measures 38 x 56mm for this circuit. Its overlay diagram is shown in Fig.3. Building it will take just a few minutes. Fit the low-profile components first (the resistors and the zener diode), ensuring that the diode’s cathode stripe faces as shown in Fig.3. Next, install the electrolytic capacitor and bridge rectifier, with both components inserted in the correct orientation as shown. The terminal blocks can be installed next, with the wire entries facing towards the nearest edge of the PCB. Finally, mount the large X2 capacitor. The LED leads need to be connected to CON2 using mains-rated wire. Each lead needs to be insulated using Parts List – Mains LED Indicator 1 double-sided PCB coded 10111251, 38 × 56mm 1 64 × 58 × 35mm IP65 polycarbonate enclosure [Jaycar HB6120 or HB6121 (with mounting flanges)] 2 3-6.5mm cable diameter cable glands [Jaycar HP0720 (pack of 2)] 1 3-way 5.08mm spacing screw terminals (CON1) [Jaycar HM3132] 1 2-way 5.08mm spacing screw terminals (CON2) [Jaycar HM3130] 2 M3 × 5mm panhead machine screws 1 length of 7.5A mains-rated wire (for the LED wiring) Heatshrink tubing (to insulate the LED connections) Semiconductors 1 5mm or 10mm LED (LED1) 1 W04(M) 1A 400V bridge rectifier (BR1) 1 4.7V 1W zener diode (1N4732) (ZD1) [Jaycar ZR1402] Capacitors 1 470μF 16V PC (radial) electrolytic 1 X2 mains-rated capacitor (see Table 1 for suitable value) Resistors (all axial, ±5%) 1 1MW 1W 1 1kW 1W 1 150W ½W 30 Silicon Chip Australia's electronics magazine Fig.3: before starting, work out the value you need for capacitor C1. Make sure the terminal blocks and zener diodes are orientated correctly and don’t forget that the wiring for Active and Neutral at CON1 is critical. heatshrink tubing. After that, cover both LED leads with a larger diameter heatshrink tube and shrink it down. The mains connection is made to the two outer terminals of CON1. It is very important that the Neutral and Active connections to mains power are made as shown on the PCB (electricians will be used to such requirements). As it is not used, the centre contact of CON1 can be removed if you wish. A three-way terminal is used so that the Active and Neutral connections are sufficiently separated. Installation As well as being connected correctly, the incoming mains wires need to be clamped to the enclosure using a cable gland mounted on the side directly opposite CON1. The LED leads are also secured to the enclosure using a cable gland that’s mounted directly opposite CON2. The PCB is secured to the base of the enclosure using two M3 screws into the two mounting posts located down the centre line of the enclosure. If you use the corner mounting holes and standoffs instead, use nylon or polycarbonate screws to provide insulation to the outside of the box should a wire come adrift and contact one of the screws. Do not use metal screws that could conceivably become live if a mains wire comes adrift. This project is designed for use with mains house wiring, so it must be installed by an SC electrician. siliconchip.com.au Only until February 28th. altronics.com.au Drops & Deals. Handy Countdown Timer & Stopwatch Fitted with 194Wh battery bank & 150W mains inverter. Allowing you cable free power for both AC and DC appliances anywhere! Plus 2.1mm DC power & USB charging for keeping everything charged on the go. SAVE $30 269 $ Compact and lightweight 2-in-1 magnetic timer will help you manage your time with X 4015 ease - whether it’s for NEW! schools, the office or at home. M 8199B SAVE $20 129 $ Caravan Fan 360° 12V/24V DC 22 $ Portable power for any adventure .95 NEW! 149 $ Stay cool on the road or off the grid with a 360° caravan fan that’s quiet, energy-smart and built for life in caravans, campers, sheds and tiny homes. 12V/24V compatible to suit a wide range of setups. Up to 150m light beam! With handy phone holder SAVE $10 89 $ 5 Channel Mixer With Bluetooth A 2548A A perfect little mixer companion for your PC/laptop for home recording. This versatile audio mixer features two microphone inputs and three stereo audio inputs, each with 2-band EQ control. X 6100 White X 6101 Black X 0214 Powerful 6000 Lumen Torch A super bright USB C rechargeable torch designed to light up areas up to 150m away! With multiple light modes, durable anodised casing and 4000mAh power bank functionality. Rechargeable USB Sensor Light Recharge anywhere! 24% OFF! D 0514 A 2798A SAVE 33% SAVE 12% 2 for $ 28 40 22 $ $ X 2386 4W 500 Lumen Handy Magnetic Wireless Power Bank Magnetic 15W wireless charging plus USB C power delivery for fast charging. Just 120 grams! 5000mAh capacity. AM/FM Pocket Radio Features easy-to-use digital tuning, a bright LCD display, and a builtin speaker with headphone jack. Requires 2xAAA batteries. 30 $ X 2387 7W 800 Lumen LED Solar Sensor Lights Add instant security to your place with these weather resistant solar lights! Require no wiring and are IP54 rated for use outdoors plus stainless steel rust resistant hardware. 3 dusk activated lighting modes. Three colour & dimmable! A handy 40cm sensor light with in-built USB rechargeable battery. Great for wardrobes. 30s on time. Detaches for recharging. SAVE 22% 38 $ X 2396 Great night light for kids! 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 February 28th 2025. Build It Yourself Electronics Centre® Get a CLOSER view. This month 109 $ Handy workbench accessory! SAVE $20 X 0435 20 Dioptre SAVE 28% X 4200 3 Dioptre X 4201 5 Dioptre The ultimate workbench lamp. Let “gadget” be your eyes in your work space. Identify those impossible to read miniature parts without straining your eyes. 130mm glass with quality optics. Great for collectors, model makers, jewellers etc. 33 $ X 4306A NEW! 229 $ USB Clip On 5x Magnifier Lamp Digital Microscope Camera A handy Inspect-A-Gadget magnifier powered by a USB port. Provides a crisp, clear view of your workbench. 430mm long. 1.5m USB lead. Inspect the finest details with this 12MP digital microscope featuring up to 1200x magnification, a vibrant 7” HD display, LED lighting, photo/video recording, and USB PC connectivity. Goodbye eye strain! S 8741 NEW! 89.95 $ LED Magnifier for micro tasks With over 5000 units sold, our 95mm illuminated desk magnifier is an absolute bargain at under $70! Every bit as useful as expensive brand name versions which sell for up to $300! An incredible visual aid for detailed inspection and work on fine items with full clarity through the quality glass lens. Tackle complex miniature tasks with confidence. S 8742B NEW! 119 5 Metre HD Inspection Camera 1 Metre HD Inspection Camera Ideal for extended inspection needs, such as in pipes, drains and conduits. USB recharging with built in 4.3” screen for a clear view. Includes hook, magnet and mirror camera attachments. Perfect to see in tight spaces to quickly diagnose problems. This shorter length model is great for mechanics and installers. An easy way to diagnose wiring problems, blockages and leaks. $15 OFF THIS MONTH! 60 $ X 4204 3+12 Dioptre SAVE 15% X 0430A 30 65 GREAT VALUE! X 4205 5 Dioptre SAVE 33% 10 $ $ “The most useful present i’ve ever received” - Mark Edmonds Augusta, WA. Arriving early Feb! $ $ T 2555 Hands free, head worn magnifier. 8x Handheld Dual LED Magnifier Thousands sold! Offers 1.5x, 3x, 8.5x,10x magnification with LED lamp. Requires 2xAAA batteries. High quality 28 diopter 37mm glass lens aided by 2 x LEDs for illumination. 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. HOME Bargains. NEW! D 2220 299 $ $ NEW! 29.95 D 2221 NEW! 49.95 $ With outdoor sensors & smartphone app! LIVE & LOCAL WEATHER. USB C to Laptop Adaptors This fantastic weather station displays your local weather data - great for boaties & gardeners. Bright & clear base station provides readings for indoor/outdoor temperature, humidity, air pressure, rainfall, wind speed and direction. Plus handy weather trends. You can even connect it to wi-fi for monitoring readings & data with your phone. 100m sensor range. 14.95 $ each SAVE $10 39 $ A combined driver bit and socket set with 47 bits and 9 metric sockets. Great for oddjobs and repairs around the house. Includes a handy magnetic latching case. T 2192 Handy all metal desktop stands for your phone or tablet. Use as a second screen while you work. D2221 suits devices 4.7”-13”. D2220 max device size 10”. X 7063A Wireless Weather Monitoring Station. Jakemy® 60pc Tool Kit Freestanding Device Mounts 22.50 These 100W compatible USB C adaptors allow you to power your laptop from a USB C power delivery (PD) power supply. See website for compatibility list. PA1922 PD1922 PG1922 PB1922 PE1922 PH1922 PC1922 PF1922 PJ1922 $ BARGAIN! D 2326A S 9431A *phone used for illustration purposes only Wireless 15W USB Phone Fast Charging Pad NEW! 99 $ G-Sensor Wi-Fi Dash Camera Delivers fast, cable-free charging, just place your phone on top and go. Its slim, lightweight design makes it perfect for home, the office, or travel. Capture every moment of your travels with this compact dashcam with 1296P super HD recording. Easy wi-fi recording management. 32Gb SD card to suit DA0365 $41.25 T 2127 Universal Aircon Remote SAVE 19% 65 $ Repair faster with a lithium screwdriver. The Jakemy JM-Y05 is a USB rechargeable screwdriver has an adjustable torque drive for accurate driving of precision screws. Suits 4mm driver bits. 2 hrs use per charge. Two way control. Lost your aircon remote? Or has your toddler destroyed it? This replacement works with 100’s of brands. SAVE 22% 22 Handy USB Note Taker Record CD quality audio with excellent audio pick up for taking recording interviews & more. 16GB storage. SAVE $24 75 $ $ A 1014 X 0705A 1000’s sold! SAVE 18% A 1103B Send TV sound to headphones! Transmits or receives audio via Bluetooth. aptX low latency - no lip sync issues! Can be used at home or in the car. 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. 49 $ CHARGE your journey. N 1114A 100W N 1117A 200W SAVE $110 SAVE $100 339 $199 $ DC Control Box & Power Meter A complete pre-wired DC connection solution for your auxiliary battery system. Connects via 50A Anderson style plug and provides 3 Anderson style outputs, 100W USB PD charging, car accessory sockets and MORE! SAVE $60 T 5096 Compact 20A Solar Regulator Heavy Duty Solar Blankets Portable solar charging made for life off the beaten track. SAVE 29% Provide portable charging power for your campsite set up. Double stitched panels, durable webbing straps and metal hanging loops and zippered cable pocket. 200W version has folding legs which allow the panel to be used freestanding. Folds up and secures with velcro for a fast getaway! 5m Anderson cable connection. SAVE $54 $ 119 $ N 2024A 20A N 2026A 40A SAVE $10 185 Utilises MPPT circuitry to extract up to 20% additional charge from your solar panels compared with traditional regulators. Suits 12 or 24V lead acid, gel battery & lithium chemistry batteries. S 2753 55 $ N 2018 40 $ Now with LiFePo4 support Powerhouse® MPPT Lithium Ready Solar Regulator SAVE $40 199 $ Works with 12V or 24V lead acid batteries. Easy on screen monitoring. Huge range of protection features. Q 2118 SAVE $20 89 $ M 8521B SAVE 16% 49 $ Handy Battery Maintainer Diagnose Battery Problems FAST Hassle free maintenance charging for vehicle & LiFePO4 lithium batteries. 6/12V 1.5A output. This handy digital battery analyser measures cold cranking amperes (CCA), resistance and battery cell condition. Suits 12V batteries. SAVE $20 79 $ SAVE 25% S 2750 44 $ SAVE 22% 19 $ S 2752 Switch & Charge With Ease! Anderson & 4 Way Switch Panel Switch all your gear on and off. Includes; 5 x Rocker switches, 1 x 12V cig. socket, 1 x 5V USBA and 1 x 5V USB-C, voltage meter. Size: 150 x 10.5 x 60mm. Handy power control plate for cars, caravans and 4WDs. 4 x rocker switches, dual USB charger, cig socket and dual 50A Anderson style sockets. Size: 150 x 120 x 100mm. SW3244 3 Way Switch Panel 3 x 12V switches with red illumination and 15A DC breakers. IP54 rated. Size: 114W x 96H x 60Dmm. Sealed In-line Switch Box IP65 weatherproof rated DPST rocker. 16A <at> 250V AC rated. Sale Ends February 28th 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 B 0002 © Altronics 2025. E&OE. Prices stated here in 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. *Devices for illustration pursposes only. By Andrew Levido Power Electronics Part 4: AC-DC Conversion with Rectifiers Power electronics as we know it today started with the invention of the mercuryarc rectifier in 1902. Rectifier-type AC-DC converters therefore predate the DC-DC converters we have been studying so far, by many decades. This month, we will look into these deceptively simple circuits. B efore the invention of the mercury-arc rectifier, the conversion of alternating current to direct current at scale required the use of rotating machines (eg, an AC motor driving a DC generator). I think their early emergence is the reason that most power electronics textbooks and courses start with rectifiers. I have taken a different approach because I believe that rectifier circuits are more challenging to analyse than DC-DC converters. In rectifier circuits, many of the quantities are sinusoidal or partly-­ sinusoidal instead of square or triangular, as they are in DC-DC converters, so calculating averages and RMS values is more difficult. Regardless of this extra difficulty, we will take a very similar approach to analysing rectifier circuits as we did with DC-DC converters. That means we will start with the simplest possible configuration; in this case, a single-­ phase half-wave rectifier feeding a resistive load, as shown in Fig.1(a). Similar to the DC-DC converter analysis, we have a source voltage and current on the left, and a load voltage and current on the right. I’ll use the same conventions for AC, DC and average quantities as I have previously. We will also use the same average value analysis technique we learned in the first article in this series. As a reminder, average value analysis is based on the fact that under periodic steady-state conditions (PSS), the average voltage across any inductor is zero and the average current through any capacitor is zero. A single-phase half-wave rectifier quantity given by the expression vs = Vs(pk)sin(ωt). This Vs(pk) term describes the amplitude of the sinusoidal voltage, and the sin(ωt) part just describes a unit sinewave (a sinewave with an amplitude of one). The resulting voltage is shown in red on the top graph. It is also common to describe AC quantities in terms of their ‘root mean square’ or RMS value. This is calculated by squaring the signal, taking the average over one cycle, and working out the square root of the result. For sinusoidal signals like our source voltage, the RMS value is just the peak value divided by √2. The horizontal axis is also worthy of a few comments. Instead of using time as we have done so far, it is conventional to use the closely related units of phase angle (ωt, pronounced omega-t – yes, ω is the lower-case version of ω) since trigonometric functions like sine operate on angles. If ωt is an angle and t is time, then ω must be an angular velocity, expressed in radians per second. Angular velocity in this context is just another way of describing frequency. You can see from the graph of vs that the length of one full cycle is 2π radians, so it should be apparent that one cycle per second (1Hz) is equal to 2π rad/s. Don’t get hung up on this if it does not seem intuitive to you – it is not critical to understanding what follows. Given that the diode can only conduct when the voltage across it is positive, the voltage seen at the load is just the positive half-cycles of the input. Since the load is resistive, the load current will have the same shape, and the source current will be the same as the load current. The average load voltage ‹vl› is harder to calculate than for the DC-DC converter where all the waveforms were square, since we have to calculate the area under the half-sine curve and divide it by the length of one cycle (2π). This requires a fairly straightforward integration, but I will spare you the details and simply give you the result, which turns out to be ‹vl› = Vs(pk) ÷ π. Adding a filter The output voltage is not exactly smooth DC, so to improve things, we could add an LC filter just as we did for the DC-DC converter. Before we do that, I want to add just the inductor, We will initially assume the voltage source and diode are ideal components. The source voltage is an AC Fig.1(a): half-wave rectifiers are simple but have several disadvantages. Their input current has a DC component, they have a poor power factor, and they require more filtering than their full-wave counterparts. siliconchip.com.au Australia's electronics magazine February 2026  35 as in Fig.1(b), to demonstrate a point relating to switches like diodes that can’t be switched off externally. With the inductor in circuit, when the source positive half-cycle reaches zero at phase angle π, the inductor current is still flowing, so the diode must remain in conduction. If a current is flowing in the diode, the voltage across it must be zero. The voltage vx must therefore go negative, following the source voltage, until the diode current falls to zero. The load current (and source current since they are the same) is therefore ‘smeared’ past the end of the half-cycle. The average voltage ‹vx› will be lower than for the unfiltered rectifier because vx is negative for a period after the zero crossing. The average load voltage ‹vl› will be equal to ‹vx› since the average inductor voltage must be zero due to our steady-state analysis rules. If we were to try to increase the inductance to the point where the load (and therefore source) current is continuous, the diode would have to conduct continuously, forcing the voltage at vx to follow the input voltage for the whole cycle. Under these circumstances, the average voltage at vx would be zero, so the current would also be forced to zero. Fig.1(b): a series inductor provides smoothing, but introduces new problems. This circuit clearly will not work to create a smooth DC current in the load. We have already seen the solution in the DC-DC examples; introduce a freewheeling diode, as shown in Fig.1(c). We can now make the inductor as large as we like, since the inductor current now has a path to flow when the source voltage reverses. In a practical circuit, we would probably add a capacitor to the output, making an LC filter, but it can be ignored for the purpose of our analysis. After all, if the inductor is large enough to eliminate current ripple, no current will flow into or out of it. So now the voltage at vx will be a half-sinusoid, which we know from above has an average value of Vs(pk) ÷ π. Due to the periodic steady-state rules, the average load voltage will be the same. Ohm’s law dictates that the average load current must be Vs(pk) ÷ πR, and since there is no current ripple, this is also equal to the DC current, Il. The input current will therefore have a rectangular shape, with an amplitude of Il and a duty cycle of 50%, as shown in blue in the upper graph. The current though D2 (lower graph) will be similar, but out of phase by half a cycle. Commutation Fig.1(c): a second 'freewheeling' diode solves most of those problems. Fig.2: the non-zero source inductance L2 causes commutation, where both diodes conduct simultaneously for a short period, impacting power factor and regulation. 36 Silicon Chip Australia's electronics magazine We have assumed a perfect voltage source up to this point, but we know that won’t be the case in real life. If our rectifier is powered from the mains, it will see a source voltage with an inductive component. If it comes from the mains via a transformer, there will be even more inductance, due to the transformer’s leakage inductance. In fact, it is hard to think of a situation where there won’t be any source inductance. Adding source inductance to the circuit of Fig.1(b) makes no difference to its operation, because it is effectively in series with L1. However, things get more complicated if we add it to the single-phase rectifier with a freewheeling diode, as shown in Fig.2. When D1 is conducting and D2 is off, the current through L2 is the DC load current, Il, so the voltage across it is zero and vx tracks the source. However, a problem arises when the source voltage goes negative. D1 must remain conducting while current is flowing in L2, and the voltage vx cannot go negative due to D2, so the siliconchip.com.au voltage across L2 begins to rise until the energy stored in L2 is all transferred to the load via D1. Therefore, the current through D1 does not drop instantly at the end of the half cycle; instead, it tapers off, as shown in the figure. This means D2 does not take over the current instantly either, and it ramps up in a complementary manner, because the total current flowing through L1 to the load remains fixed. D1 and D2 are therefore both in conduction for a (hopefully) short period. The same thing happens in reverse when the positive half-cycle starts. The input current is now zero, so it takes a finite time for it to ramp up through L2 and D1 to the level of the load current. During this period, D2’s current ramps down to maintain the constant load current. While both diodes are conducting, the voltage at vs is held at zero, truncating the beginning of the voltage half-cycle and slightly reducing the average output voltage. This process, where current is transferred between the switches, is called commutation. The effect of commutation is to reduce the load regulation and change the shape of the source current. The effect on regulation is equivalent to adding a resistor of value XL2 ÷ 2π in series with the output. The effect on the input current is to make it more trapezoidal, meaning that the RMS source current for a given level of load current is higher with commutation than without. Power factor This segues nicely into the topic of power factor, which has become increasingly important in power electronics. All the rectifiers we have looked at so far have a source current waveform that is non-sinusoidal. This means that the apparent power entering the rectifier, calculated as the source voltage times the source current, is higher than the real power delivered to the load, calculated as the average over a cycle of the instantaneous voltage times the instantaneous current. How can this be? Consider the simple example in Fig.3. At the top, we have a sinusoidal source voltage (red trace) feeding some converter that produces a distorted (non-sinusoidal) current waveform (blue trace). siliconchip.com.au In this case, the source current iS the sum of two sinusoidal currents: iS1, at the same frequency as the source voltage, and iS2, which is of lower magnitude but twice the frequency. The lowest chart shows the instantaneous product of the source voltage vs with the fundamental component of the current iS1 (in dark green) and the product of the source voltage vs with the second harmonic component of the current iS2 (in light green). The average power available from the fundamental component of the current is positive, but the average power available from the harmonic component is zero. This turns out to be true for all harmonics. From Fourier theory, we know that any periodic current waveform can be decomposed into a series of sinusoids, including a fundamental component and its harmonics. However, if the source voltage is sinusoidal (like the mains), only the fundamental component of the current contributes any useful power to the load. This is the ‘real power’, which we designate ‹p›, in units of watts. While the harmonic components of the current do not contribute to real power, they do contribute to the RMS value of the current, and therefore to the ‘apparent’ power – the product of RMS source voltage and RMS source current. We use S to describe apparent power, which has units of VA (volt-amps), and which will always be greater than or equal to the real power. The ratio of real power to apparent power, ‹p› ÷ S, is the definition of power factor. It is a unitless quantity that varies between zero and one. A power factor of one means that the real and apparent power are equal, as would be the case for a resistive load, for example, and implies that the current is purely sinusoidal and in phase with the voltage. A power factor less than one means that the real power doing useful work is lower than the apparent power being consumed. Power factor is important because the electrical supply system has to be dimensioned for apparent power. For example, an Australian domestic power outlet is rated to deliver 230V at 10A for a nominal apparent power of 2300VA. If you applied a unity power factor load (like a resistive heater), you can expect to get 2300W of usable power from such an outlet. Australia's electronics magazine Fig.3: with a sinusoidal voltage source, only the fundamental component of the current waveform contributes to real power. The average power of any current harmonics is zero. Fig.4: if the current is sinusoidal but out of phase with the voltage, the average power available will be reduced. However, if the load produces a non-sinusoidal current, the usable power will be lower. If the power factor were 0.75, for example, you would only be able to get 1725W of useful power from the circuit, even though the RMS current sourced from the outlet would be 10A. Clearly, we can get the most out of the power distribution system by keeping the power factor high. Mains-connected power electronics has become a major contributor to poor power factor in electricity distribution systems around the world. A variety of techniques are available to February 2026  37 improve or ‘correct’ power factor. We will cover some of these circuits in the next instalment of this series. For completeness, I should point out that just eliminating current harmonics won’t get you to unity power factor if the current is out of phase with the voltage. In fact, when I studied electrical engineering (many decades ago), the only discussion of power factor related to phase. The distortion component was skipped altogether because switch-mode supplies were far less common then. Fig.4 shows why phase matters. A sinusoidal source voltage feeds a device that draws a current that is also purely sinusoidal, but slightly out of phase with the voltage. The green trace shows the instantaneous power obtained by multiplying the two. If you compare this to the green power trace in Fig.3, you will see that instead of riding on the zero line, this trace is shifted down, reducing the average power compared to the in-phase case. The vertical dashed lines show that as the phase shift increases, the zero-crossings of the voltage and current sinusoids diverge, so the power curve must drop to keep its zero crossings aligned. When either the voltage or current is zero, the instantaneous power must also be zero. If the phase shift reaches ±π/2 radians (±90°), the average power, and therefore the power factor, drops to zero. If voltage and current are pure sinusoids, cos(ø), where ø is the phase shift, is a shorthand way to calculate power factor. The power ratio equation is the one to use in power electronics as it works for both phase- and distortion-­ related power factor or any combination thereof. Full-wave rectifiers Fig.6: we typically use a capacitor filter instead of an inductor. Circuit (A) is simple and cost-effective, but has a fairly poor power factor. The half-wave rectifiers that we described above are rarely used in practice since they suffer from three major drawbacks. First, the average input current is non-zero. This means there is a DC component to this current, which will not play nicely with transformers in the source network (and can accelerate conductor corrosion in some cases). Secondly, they require large filtering components to achieve low voltage ripple because energy is supplied only during every other half-cycle. Lastly, they have a poor power factor. We can confirm this pretty easily. Consider the half-wave rectifier with freewheeling diode in Fig.1(c). We saw that the average load voltage was ‹vl› = Vs(pk) ÷ π. We know the load current Il is DC, so we can calculate the average power dissipated in the load (and therefore supplied by the source) to be ‹p› = Vs(pk)Il ÷ π. The RMS input voltage is Vs(pk) ÷ √2 and the RMS input current is √Il² ÷ 2 = Il ÷ √2, so the apparent power must be S = Vs(pk)Il ÷ 2. Dividing ‹p› by S cancels the voltage and current terms, leaving a power factor for this topology of 2 ÷ π, which is about 0.64. Not great. Fig.5 shows a full-wave bridge rectifier and its associated waveforms. You can think of this circuit as two single-phase rectifiers with freewheeling diodes – in positive half-cycles, D1 conducts and D3 is the freewheeling diode, while in negative half-cycles, D2 conducts and D4 is the freewheeling diode. The full-wave rectifier addresses all the problems of the half-wave rectifier. The input current swings positive and negative alternately, so has an average value of zero and therefore Australia's electronics magazine siliconchip.com.au Fig.5: the full-wave bridge rectifier with inductor overcomes the disadvantages of half-wave rectifiers. The input current has no DC component, energy is delivered to the load on every half-cycle, and the power factor is much better. 38 Silicon Chip no DC component. Power is supplied every half-cycle, doubling the output frequency and thus requiring less filtering to achieve a given level of voltage ripple. The power factor is also much better. The average load voltage is twice that of the half-wave rectifier, so the average power is ‹p› = 2Vs(pk)Il ÷ π. The RMS input voltage is the same, but the RMS current is now just Il, giving an apparent power of S = Vs(pk) Il ÷ √2. Dividing ‹p› by S cancels the voltage and current terms, as before, but leaves us with a power factor of 2√2 ÷ π, which is about 0.90. This is much better. Capacitive filters Of course, all of this assumes the presence of a large inductor to smooth the current, but this is not how we usually build rectifier-filter circuits. For the most part, we simply add a capacitor directly after the half- or fullbridge, as shown in Fig.6. The circuit acts like a peak detector, with the capacitor charging to Vs(pk) via the diodes at the crest of the half- or full-wave rectified waveform (shown dotted). The capacitor discharges via the load until the next peak. The voltage ripple can be (roughly) approximated by assuming that the voltage takes on a sawtooth profile (ie, it charges instantly at the crest and that the discharge is linear). For a half-wave rectifier, this gives Vl(pk-pk) = Il ÷ f C, where f is the source frequency and C is the capacitance. For a full-wave rectifier, the ripple is half of this, ie, Vl(pk-pk) = Il ÷ 2f C. This approximation is quite pessimistic for small supplies where the source impedance is relatively high, as we shall see. Fig.7: I built and simulated this simple transformer/rectifier circuit. It showed we could get a maximum power of about 14W from this 20VA transformer due to the limited power factor. This topology means that current only flows into the capacitor for a short period, resulting in a current waveform that is made up of narrow spikes of current aligned with the crests of the input voltage. The width of the current spikes is related to the ripple (the lower the ripple, the narrower the spikes), the source impedance and the capacitor’s ESR. All of this is difficult to calculate, but is a great candidate for simulation and experimentation. A practical example I had a 20VA, 240V to 12+12V toroidal transformer in my junk box, so I decided to build the simple full-bridge AC-to-DC converter shown on the right-hand side of Fig.7 to see how it performed. Notice that the transformer is specified for apparent power. The transformer windings are connected in series to get a nominal 24V RMS, which is rectified by four chunky 6A10 (6A, 1000V) diodes I had lying around, and filtered by a 1000µF 63V electrolytic capacitor. First, I measured the open-circuit voltage of the transformer (28.6V RMS), the DC resistance of the secondary windings (1.3W each), the transformer leakage inductance (250µH) and the filter capacitor’s ESR (0.06W). This allowed me to build the simulation model shown in Fig.7. The simulation results are shown in Fig.8. The average output voltage is 31.6V, with a peak-to-peak ripple of 3.4V. The average output power is therefore 16.0W. The input current is shaped as we would expect, but the input voltage shows a flattened top. This is due to the voltage drop across the source impedance when the current pulses occur, and is typical for supplies of this size. The simulator calculated the RMS input voltage and current to be 26.2V and 0.95A, respectively, for an apparent power of 24.9VA (a little higher than our transformer’s rating). The power factor is therefore 0.64. The simulation compares well with the measured results below. The average load voltage is 30.6V and there is 4.8V peak-to-peak ripple. The output power is therefore 15.3W. Fig.8: the experimental results agree fairly well with those obtained by simulation. siliconchip.com.au Australia's electronics magazine February 2026  39 These measurements were taken using a Current Probe and two Differential Probes, described in the January and February 2025 issues of this magazine, respectively, so the appropriate scaling factors need to be applied. The RMS input voltage and current are 26.0V and 0.824A for an apparent power of 21.9VA (still a touch too high for the transformer in the long term). The power factor is therefore 0.69, slightly better than the simulation, but nothing to get excited about. The important thing to note here is that the relatively low power factor puts an upper limit on the real power you can get from a given transformer. For this 20VA toroid, it is about 14W. I should also point out that this type of rectifier/filter results in fairly high 100Hz current ripple in the capacitor, which raises its internal temperature and potentially shortens its life. In this circuit, the capacitor ripple current is 0.8A RMS. Electrolytic caps usually come with a maximum 100Hz ripple current specification, so it is worth checking you are not exceeding this limit. The ripple current rating is one of the reasons you almost always see large filters made up of multiple parallel capacitors. If the capacitors are identical the ripple current rating of the bank is the sum of the ripple current rating of each capacitor. The ‘peaky’ current waveform has an impact on the output voltage you will achieve with this circuit. There are two diode drops between the peak value of the transformer secondary voltage and the voltage across the filter capacitor. These will likely be higher than the nominal 0.6-0.7V you might expect because the capacitor only charges when the current is at its peak. The diode data sheet should provide a curve called “instantaneous forward characteristic” or similar, which relates forward voltage drop to peak forward current. In the example of the 6A10 diodes I used, this curve shows a forward voltage of 0.8V at the peak current we see in the simulation, for a total drop of 1.6V. It is not at all unusual for the voltage drop to approach 2V if larger currents are involved. This drop can eat up a significant portion of the available voltage in low-voltage applications. As an aside, this is why active rectifiers are becoming more popular (see our September 2024 design; siliconchip. au/Article/16580); they involve very little voltage loss and so improve efficiency. Inrush current This topology also comes with potential inrush current concerns. When power is first applied, and the capacitors are fully discharged, the inrush current is limited only by the supply impedance and the capacitor ESR. This is rarely a problem with lowvoltage supplies fed by relatively small transformers like this one, but can be a big problem for off-line rectifier/filters and very large capacitor banks, as you might find in a high-power audio amplifier, for example. The Variable Speed Drive for Induction Motors (October & November 2024; siliconchip.au/Series/430) used a bank of five 330μF 400V capacitors to filter the rectified mains. A simulation made at the time showed that without inrush limiting circuitry, the peak inrush current would be around 200A, almost certainly tripping the supply circuit Silicon Chip kcaBBack Issues $10.00 + post $11.50 + post $12.50 + post $13.00 + post $14.00 + post January 1997 to October 2021 November 2021 to September 2023 October 2023 to September 2024 October 2024 onwards September 2025 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 40 Silicon Chip Australia's electronics magazine breaker and maybe damaging the rectifier diodes. In that case, we used a special inrush-limiting thermistor with a cold resistance of 10W to limit these peaks to less than 35A. The thermistor’s resistance drops as current passes through it, and the model we used is rated to conduct 15A continuously. These low-cost inrush limiters are available in various sizes and packages and are very common in off-line converters of all sizes. There was even one in the flyback converter we looked at last month. You can use a resistor to limit inrush if you short it out with a relay or similar after it has done its job, but you need to be sure that the resistor can withstand the short pulse of power that occurs during inrush. In the case of the Variable Speed Drive, this peak power is well over 100W for a few milliseconds. Some power resistors are specified for pulse power, but many are not, so be careful. Other topologies There are many variants of the fullwave rectifier, and a couple of the more common ones are shown in Fig.9. The load voltages shown assume the diode voltage drops are negligible. At the top is a centre-tapped variant that is a little more efficient for low-voltage supplies than the full bridge, since the current passes through only one diode instead of two. This comes at the expense of a more complex transformer, but in reality, dual secondary windings are common in small off-the-shelf transformers. The middle rectifier uses a dualwinding transformer to produce a symmetric split (±) power supply – very useful for audio amplifiers or op amp circuits. The final circuit is a full-wave voltage doubler that is effectively two halfwave rectifiers in series. During positive half-cycles, the upper capacitor is charged almost to the peak of the secondary voltage, and during negative half-cycles, the lower capacitor is charged to a similar voltage. The result is an output voltage twice what could be expected from the same transformer with a full bridge rectifier. Phase control No discussion of rectifiers would be complete without introducing phase-controlled rectifiers. This is a siliconchip.com.au Fig.10: using a thyristor in place of the diode allows the output voltage of this half-wave rectifier to be controlled. Fig.9: here are some variations on the full-wave rectifier theme that might come in handy. technique that is used less these days than it used to be, but is still relevant in industrial applications where very high currents must be controlled. The classical phase-control switch is the thyristor (sometimes called the silicon controlled rectifier, or SCR). You can think of a thyristor as a diode that will not conduct in the forward direction until the appropriate gate signal is applied. When the gate is positively biased with respect to the cathode while the anode-cathode voltage is positive, the thyristor switches on and remains on, even if the gate bias is removed, until the current drops to zero. In this sense, it is a latching device. In fact, the anode current must drop to zero for a short time (tq) for the thyristor to recover its forward blocking capability. This time is in the order of tens to hundreds of microseconds, limiting the application of thyristors to fairly low-frequency applications. Thyristors are very robust devices and are available in voltage ratings siliconchip.com.au Fig.11: the full-wave phase-controlled rectifier is capable of inversion, where energy is transferred from the load to the source. up to 6kV and current ratings up to 4kA. They have very good overload performance and, unlike most semiconductor switches, can be protected by fast fuses. A more modest example, the 800V TN1605H-8G, is rated for a current of 16A RMS (8A average) and can withstand a non-repetitive half-cycle (10ms) surge of 160A. Just as a matter of interest, this thyristor requires a gate current of 1.5mA to switch on (you would usually drive it at around 5mA to be sure), and has a tq of 25µs at 25°C, rising to 85µs at 150°C. Fig.10 shows a half-wave phase-­ controlled thyristor rectifier. In this case, the thyristor is switched on at a phase angle (sometimes called firing angle or delay angle) of θ. You can see that if θ is zero, the output will be identical to the half-wave rectifier (ie, ‹vl› = Vs(pk) ÷ π), but as the phase angle increases, the output voltage drops until it is zero when θ = π. The relationship between phase angle and average voltage is not linear, Australia's electronics magazine due to the truncated sinusoidal shape of the voltage waveform. It can be shown that the output voltage is ‹vl› = (Vs(pk) ÷ 2π) × (1 + cos[θ]). Full-wave phase-controlled rectifier The full-wave phase-controlled rectifier (Fig.11) has some very interesting properties, as we shall see. The thyristors are gated on at phase angle θ as before, with SCR1 and SCR4 on in the positive half cycle and SCR2 and SCR3 on in the negative one. Let’s have a close look at what happens over a couple of cycles. As we come to the end of a positive half-cycle (at phase angle π, let’s say), SCR1 and SCR4 are conducting, but SCR2 and SCR3 have not yet been gated on to take over the constant current through the inductor. This means the current continues to flow through SCR1 and SCR4, so they must remain conducting past phase angle π, and the voltage at vx follows the input voltage negative. February 2026  41 Fig.12: an AC-DC converter can, in theory, operate in any one of these quadrants. Inversion occurs in quadrants II and IV. Fig.14: a three-phase rectifier has excellent low output ripple and a very good power factor, thanks to energy being delivered to the load six times per cycle. a 50% duty cycle, as it was for the uncontrolled bridge rectifier in Fig.5, but now its phase can shift from being in phase with the input voltage when θ = 0 to 180° out of phase when θ = π. Let’s pause and take in what this means. On the load side, a positive DC current with a negative average voltage means negative power is ‘dissipated’ in the load! The same is true on the source side; at phase angles greater than π/2, the average input power is also negative, so power is delivered to the source. This means that for phase angles greater than π/2, this circuit will transfer energy from the load to the source. This is known as inversion, and it can be very useful in practice. All four quadrants Fig.13: a three-phase source or load can be configured in star (Y) or delta (∆), since no current flows in the Neutral wire if the phases are balanced. When SCR2 and SCR3 switch on at phase angle π + θ, the current commutates from SCR1 and SCR4 to SCR2 and SCR3, so the voltage flips positive abruptly. The same thing happens as the negative half-cycle comes to an end at phase angle 2π, although this time, it is SCR2 and SCR3 that remain on. The upshot of this is that the average voltage, ‹vx›, and hence the average load voltage, can go negative. The input current is a square wave with 42 Silicon Chip I will give a couple of examples where this can be useful, but first it is worth looking at the general case illustrated in Fig.12. Here, I have shown a generic ‘four-quadrant’ AC-DC converter powering a DC motor. The output voltage and current of this converter can each be positive or negative. This gives four possibilities, illustrated by the four quadrants in the V/I chart. The quadrants are denoted by Roman numerals counter-clockwise from the top right. In quadrants I and III (shaded green), both the voltage and the current have the same sign, so the power is positive. In quadrants II and IV, the voltage and the current are of opposite polarities, so the power is negative. I have positioned the current arrows on the motors so that they are always Australia's electronics magazine on the most positive terminal, but they are consistent with the upper diagram in that positive current flows down and negative current flows up. You can see that in quadrants I and III, the current flows into the most positive terminal, so the motor will be consuming power and driving the load (possibly in opposite directions, depending on the motor type). In quadrants II and IV, the current emerges from the most positive terminal of the motor, so the motor must be behaving as a generator and exporting power. Getting back to the full-wave phase-controlled rectifier in Fig.11, we can see that this operates in two quadrants (I and IV, because the current is always positive). If the load on this converter was a DC motor with a high-inertia mechanical load, quadrant I could be used to drive the load, and quadrant IV could provide regenerative braking. The reversed voltage applied to the motor creates a retarding torque that brings the motor and load to a stop quickly, since the energy stored in the rotating mass is transferred to the source. The motor will come to a stop much faster than it would if left to coast, losing its stored energy only to friction and windage. I have also seen this circuit used to quickly switch off (and therefore de-magnetise) a large electromagnet, moving the energy stored in the magnet’s inductance to the supply much faster than it would if we relied on the freewheeling effect of a full bridge rectifier. This allowed the electromagnet, which was picking up and moving siliconchip.com.au steel components, to release them promptly on command. We should note here that the inversion described here is not sufficient to create a DC-AC converter. The AC source must be present for this circuit to work, and power can only flow back to the source while it is present. I will cover DC-AC inverters in a future article. Three-phase systems The electromagnet example I mentioned above was actually fed from a three-phase supply. I won’t go into multi-phase rectifiers in any great detail, since they are really only used in industrial applications, but I will touch on them for completeness. A three-phase voltage source is just three sinusoidal voltages with equal amplitude and frequency, but shifted in phase by one third of a cycle. I have drawn these three sources and loads (at the top of Fig.13) in a slightly unconventional manner, but you can probably see why this arrangement is called a star or Y configuration. The centre point of the star is the Neutral connection. The phase-to-Neutral voltages are shifted from each other by 2π/3 radians (or 120°), and this brings some very useful benefits. Firstly, if the load in each phase is balanced, the sum of the three line currents is zero, and no current flows in the Neutral wire, which is why I have shown it dotted. In fact, many threephase loads such as motors don’t even have a Neutral terminal. On top of this, for balanced loads, any current harmonics that are multiples of three (the 3rd, 6th, 9th etc) also cancel to zero so, they don’t contribute to apparent power, meaning you get a power factor advantage for free. Given that the Neutral is unnecessary if the load is balanced, we could think about line-to-line voltages rather than line-to-Neutral voltages, and redraw the circuit as shown at the bottom of Fig.13. Again, I have drawn it unconventionally, but this configuration is called a delta (∆) arrangement because the elements are arranged in a triangle. You can have a star-connected load with a delta-connected source and vice versa. The line-to-line (or phase-to phase) voltages are the sum of the two relevant line-to-Neutral voltages. I will leave the maths out, but it is easy enough to siliconchip.com.au show that the line-to-line voltages are larger than the line-to-Neutral voltages by a factor of √3, and displaced from them in phase by π/6 radians (30°). The nominal line-line voltage for domestic three-phase supplies in Australia is therefore 400V, corresponding to √3 times the 230V nominal phaseto-neutral voltage. Before you rush to correct me, I am well aware the typical line-to-line voltage is closer to 415V in many areas of the country, as we continue to transition from the old 240V/415V standard to the newer 230V/400V standard. In any case, 240V/415V is within the allowable range of the 230V/400V standard and vice versa. A three-phase full-wave rectifier Fig.14 shows a three-phase fullwave rectifier with an inductor. The upper graph shows the U-phase line-Neutral input voltage and the U-phase line current, with the V-and W-phase line-Neutral voltages shown dotted. The lower graph shows the voltage at vx, with the six half-cycles that contribute to it shown dotted. The average voltage at vx and hence the average load voltage ‹vl› = 3vll(pk) ÷ π or approximately 0.96 times the peak line-line voltage. The load voltage ripple is obviously much lower than for the single-phase rectifier, since there are now six ‘pulses’ of voltage each cycle instead of two. The power factor for the three-phase rectifier is also better than the single-­ phase case, as you might expect by observing that the line current waveform looks more sinusoidal with its ‘stepped’ shape. The power factor also turns out to be equal to 3/π (0.96) for this topology, so quite close to unity. You can also use thyristors in place of the diodes to create a phase-­ controlled version of this rectifier. It behaves in much the same way as its single-phase counterparts with phase angles above π/2, producing negative output voltages. Like the single-phase case, it can operate in quadrants I and IV. That’s all for this month. Next month, we will continue to look at AC-DC converters, with a focus on power factor correction. While we are on the topic of being responsible with regard to the power grid, we will also touch on the basics of electromagnetic SC interference (EMI) filtering. Australia's electronics magazine Silicon Chip Binders REAL VALUE AT $21.50* PLUS P&P Are your copies of Silicon Chip getting damaged or dog-eared just lying around in a cupboard or on a shelf? Can you quickly find a particular issue that you need to refer to? Keep your copies safe, secure and always available with these handy binders These binders will protect your copies of S ilicon C hip . They feature heavy-board covers, hold 12 issues & will look great on your bookshelf. H 80mm internal width H Silicon Chip logo printed in goldcoloured lettering on spine & cover Silicon Chip Publications PO Box 194 Matraville NSW 2036 Order online from www. siliconchip.com.au/Shop/4 or call (02) 9939 3295 and quote your credit card number. *see website for delivery prices. February 2026  43 Internet Radio Part 1: by Phil Prosser If you have terrible radio reception in your house or shed, or have been looking for a neat computer-based music player, this project is for you. M y workshop and sound room is in a terrible location for radio reception.. Given that it is clad reception in corrugated iron, an indoor antenna was never going to work. Even a substantial outdoor antenna was not enough to overcome the poor signal level in my area, and I still get terrible radio reception. This is not such a problem when I lug my laptop out and stream music, but that is a bother. Recently, I was working in the shed, lamenting the poor reception yet again, and the irony that in 2025 I get better internet services than radio. As I trudged back inside to get the laptop, the seed of this project was planted. I had some ‘spare’ Raspberry Pi 4B boards and knew how easy they are to set up to stream my favourite stations. The problem was how to package them other than in the tiny Pi cases you can buy. Something that ran off a plugpack and connected to speakers, more-or-less standing by itself, seemed like the ideal solution. Essentially, a modern version of a “boombox”. As I sat in my armchair, wishing I was listening to the radio, my gaze fell on the 3D printer. The answer lay there. So, what were the requirements? It needed to be: • Easy to build • Not too expensive • Based on a Raspberry Pi 4B or Pi 44 Silicon Chip 5 with an audio HAT or USB audio interface • Able to drive speakers to a decent sound level • Capable of Bluetooth streaming as a bonus • Controlled using an inbuilt LCD touchscreen • Able to plug in USB storage devices easily • One plugpack to run the whole thing • No complicated mechanical work • The option of a modest battery would be a bonus Some of you will be thinking that internet streaming services are not exactly high in fidelity. Yes, the bitrate and quality of streaming services varies, and the sound quality from the Raspberry Pi headphone socket is limited. However, given the option of poor, or even absent, radio reception, I figured that average sound quality is better than nothing. Also, a secondary goal of this project is to introduce people to how easy it has become to construct a fully working software platform that can be used to explore Linux and its multimedia capabilities. It has been decades since I used Unix in anger, so I thought it was a great opportunity to brush up. As a bonus, I wound up with a working radio! Australia's electronics magazine Looking over the Altronics website, the following products caught my eye: • The Raspberry Pi 4B (we only need the 4GB version) [Z6302G] • TPA3110 2 × 30W audio amplifier with Bluetooth input [Z6409] • DC/DC converter that can deliver 5V <at> 5A from 8-32V DC [M7832] • 7-inch (178mm) LCD touchscreen with a 1024 × 600 pixel resolution [Z6516A] With these, all we really need is a housing. 3D printing seemed an obvious approach. Of course, if you are handy with timber or metal, there is no impediment to your building this from either of these materials. Note, however, if you want to use WiFi, a solid metal case will reduce your WiFi range significantly. If you really do want a metal case, then you might want to connect to your internet via the Gigabit Ethernet port, or using an external WiFi antenna. If you have ever heard someone say they will “just 3D print” something, that means someone else has done a lot of work to prepare the files they print, or they are underplaying how hard it is to design a complex 3D object. I am still learning how to use Fusion 360, and am definitely no expert. So after numerous hours with vernier calipers and the computer, and quite a lot of muttering and head scratching, we siliconchip.com.au Photo 1: if you have external speakers (or prefer to use them), you can build the Internet Radio like this, with just the centre section that houses the Pi, touchscreen, amplifier and power supply. had a first version of a console for the Internet Radio. It was not perfect, but it looks a lot like the middle section of the Internet Radio in the picture. If you choose to dip your toes into designing 3D objects, Fusion 360 is probably at the high end of packages you might choose. It is used in industry and uses the common “sketch / model” definition process, so any skill you develop on this tool is directly relevant to professional hardware engineering work (hint hint). The free version does everything we need, so this is a great place to start price-wise. A few tips from a true beginner: • Objects are created from sketches. Sketches are fundamental in this sort of CAD system, and understanding that 3D objects are created from and defined by sketches is the most important first step. • Typical operations used in this project were extrusions, cuts and fills from profiles using sketches to create and modify the bodies that make up the Internet Radio. • Sketches are defined on planes; obvious ones are the X, Y and Z planes from the origin, but you can create them on surfaces of objects, allowing you to define things like holes and the text we put on top of the radio. • You can modify bodies, for example, to place chamfers on edges; still, the critical mechanical definitions are in the sketches. • The bodies that we create from sketches remain defined by the sketch. An extrusion of a square on a sketch could create a cube or, if it is long, a bar. If we change the square on the sketch to a rectangle, the body created siliconchip.com.au by the extrusion will become rectangular. There is a lot of power in this approach, but it can take some getting used to. • There are many excellent tutorial videos on using Fusion 360. Getting things to fit and ‘click together’ does require some thought. CAD can lead to a false sense of security. For example, if you want an aperture, like our SD card hatch, to open and close, the actual aperture in the case needs to be larger than the hatch. In the CAD world, without applying design rules and checks, this works with zero tolerance. For our 3D-printed case, we added a 0.25mm gap around the hatch, as the tolerances of the print demand this. This sort of consideration needs to be applied to every surface in our design. This includes things like the LCD screen hole, plus the front and rear panels. We have an Ender 5 S1 printer that has a 220 × 220mm print area; this is the same as the very popular Ender 3, and many other 3D printers. This defined the size of the main case. It just fits the LCD screen, Raspberry Pi and amplifier, leaving room in the middle for a battery if you are creative. We used these limits in our design, and the project assumes you have this print area to work with. Options The original intent was to design a simple internet radio box that plugs into the stereo in the shed. That is exactly what the first iteration of this project was (shown in Photo 1), and it remains a perfectly valid application. It involves omitting the power Australia's electronics magazine amplifier and running the output to RCA sockets. However, once that was complete, we thought, why not make some speakers that can either sit on a shelf along with the main unit, or even attach on either side, turning the unit into a boombox? Additional speakers definitely wouldn’t fit in the print with the main console, but they could definitely be printed separately. If the speakers are to attach boombox style, the height and depth are fixed (they must match the main unit). To make them look reasonable, the width is constrained to be something similar to the height. This is small, but given this is really more about a functional radio than hifi, that’s OK. In fact, once we added some bass and treble boost in the Media Player settings, the Internet Radio’s sound is surprisingly good. But to be right up front, if you want proper hifi, you need to connect more substantial speakers. With that in mind, the Silicon Chip Internet Radio was born. You will note that we have not included an AM/ FM receiver. This might come across as ridiculous, but remember our use case is for environments lacking radio reception. If you want to swap out the Bluetooth module for AM/FM radio, the switch is there, and all you need to do is integrate the tuner and switch to it instead of Bluetooth. 3D printing and supports Now let’s get back to 3D printing. For those of you who are veteran 3D printers, we are sure you are looking at the radio and thinking, “That is a lot of printing”. That’s true, but it also makes the assembly dead easy. February 2026  45 To those experienced in the art of 3D printing, our extreme laziness on the mechanical aspects of this project has led us to designing models for which no supports are required for any of the printed parts. We hate cleaning off supports, so have spent more time designing supports out of the design than we would have spent cleaning them up. That’s great for you, since it means if you print this design, the pieces should all pop off the print bed pretty well ready to use. For the uninitiated, a 3D printer is an additive manufacturing process tool. It lays down, in our case, 0.2mm thick layers of plastic one on top of another. So what happens if you have a feature that does not start on top of an underlying part of the print? The answer is you need to add ‘supports’, which are printed with the only purpose of holding up features in the final design, but need to be broken off and cleaned up prior to using the print – see Figs.1 & 2. Even running our Ender 5 S1 moderately hard using Klipper on the Creality Sonic Pad to optimise print speed, the main case still took more than 10 hours to print, and the speakers not that much less. So with this project you trade patience, and the pleasure of seeing a whole thing come off your printer, against many hours of manual labour. On and off, this print would run over three days or nights for most people. Of course, once you start the print, there is no effort required. Like most projects, we have built more than a few prototypes. Only one problem arose, which was caused by the print coming loose from the bed. This was down to our being lazy and not cleaning the bed properly before starting the print run. We toyed with the concept of including grilles on the speakers. We don’t prefer grilles, but offer three options: no grilles, small-hole grilles and large-hole grilles – see Photo 2. The choice is yours, and they all come out in a single print. Overall design So what does our Internet Radio comprise? As shown in Fig.3, it is an aggregation of off-the-shelf modules wired together. The wiring is not complicated, but as we will describe later, running this from a single plugpack 46 Silicon Chip Fig.1: “Bob”, designed by young Zak. In an additive print process, there is nothing to support the lower extremities of the arms given that a 3D print starts at the bottom and adds layers to build it upwards. Fig.2: “Bob” as the 3D printer would need to print to provide supports to the arms. The supports can be broken off, but they leave messy bits and it is really never as neat as a clean print. does mean we need to pay attention to the ground routing to minimise noise. The user-friendliness of Linux distributions is now so high that rolling out a Raspberry Pi OS (which we will shorten to RPi OS) with inbuilt tools such as LibreOffice and the VLC media player takes only a few button clicks, and is certainly no more complicated that setting up Windows. It just works. VLC media player is ubiquitous and found on every computing platform, and also very well supported. By using VLC, we can get users up and running with some tunes in a very familiar environment, which can be a springboard for them to dip their toes into much more complex or specialised tools. Why didn’t we use a dedicated multimedia centre app? There are many Australia's electronics magazine dedicated multimedia players available, which can install on everything from a Raspberry Pi through to a full PC. We have played with most of the following, and once you are comfortable with the whole Pi bit and have the hardware running, suggest that you might consider them. The reason we did not start with one of these dedicated players is that some of the configuration is quite specific to an individual’s application, and we ran the risk of the project becoming a complicated description of how to configure one player or another. Still, you could consider using: • Moode (https://moodeaudio.org) • Volumio (https://volumio.com/ get-started) • piCorePlayer (www.picoreplayer. org) siliconchip.com.au Photo 2: we prefer to have bare speakers but you can print one of the case options with a grille if you prefer, for a bit of extra protection against curious fingers etc. Photo 3: we have included a hatch you can use to access the SD card, provided the speakers are not bolted onto the side of the case. These programs do not use the RPi OS desktop, which means that if you install them, the Raspberry Pi stops being a generic Linux machine and becomes a dedicated music player. There are some aspects that might make this very attractive to you, though; for example, some of these allow you to control your stereo from a smartphone. We will go on to describe a much more plain-vanilla RPi OS version, which we believe any DIYer should be able to get up and running. More on Linux For any of you reading this who are intimidated by the fact that this is running RPi OS Linux, we assure you that if you start with the RPi OS desktop, you will wonder what you were worried about. From there, you can read and learn a few of the command line instructions and get a feel for how it works. Oh, and have an internet radio and media player in the deal. At first glance, the RPi OS desktop is just another graphical user interface (GUI). If you compare it to Windows, many menus are in different places, but all the expected things are there. The support for this on the internet is superb. If you type a question into Google like, “How do I set up a Bluetooth mouse in Raspberry Pi OS”, you will get crisp instructions on how to do this in the GUI or at the command line. If you are new to Linux, use the GUI and ease into the command line if you need it. We will describe a pretty simple setup, but you can create a much more complex and specialised media centre setup on exactly the same hardware. siliconchip.com.au You could even have multiple different interfaces on various SD cards and swap between them. SD cards are cheap! We would love to hear from those more expert in Linux/RPi OS and the many media centre programs regarding how you set this up to be much better than our ‘minimum viable product’ offering. Initial setup The first thing you should do is get RPi OS running on your Raspberry Pi. While we have made the build easy to put together and update, it is reassuring to know that the Pi is running prior to putting everything in the case. This first requires us to populate a microSD card with the RPi OS software and plug it into the Raspberry Pi. We have made a special hatch on the side of the case so you can change the SD card without disassembling the main case once it is all built (Photo 3). However, if you have screwed the speakers to the box, then you will need to unplug and remove the Raspberry Pi to change the SD card. Power for the Raspberry Pi may come from the specialised Raspberry Pi power supply or a beefy USB-C supply. To set it up, you will also need a keyboard and mouse to plug into the USB ports, any HDMI display, and a micro HDMI to HDMI cable to connect it to the Raspberry Pi. As mentioned earlier, you also need a microSD card. There is a bewildering array of options for microSD cards; the “extreme” ones allow somewhat faster writes, but this won’t affect most users. This card also provides storage for applications and data such as music, so if you wish to store a lot of data on this card, choose a higher capacity device. Fig.4 shows the minimum configuration to get things running. We will keep these setup instructions brief, as there are plenty of tutorials on loading RPi OS on the web. 1. Download the “Raspberry Pi Imager” from www.raspberrypi.com/ software (it is free and just works). 2. Run it. If you are on Windows, you will see a security pop-up; click “allow the app to make changes”. 3. Insert your microSD card into an adaptor to allow you to plug it into your computer. A simple USB to microSD adaptor works fine (some computers, especially notebooks, have integrated adaptors). Fig.3: the block diagram for the Internet Radio. Australia's electronics magazine February 2026  47 Fig.4: the minimum configuration to get a Raspberry Pi running. This can be lashed together on your desk; once everything is set up, you can switch to using the touchscreen and a small wireless keyboard and mouse. 48 Silicon Chip Australia's electronics magazine 4. Click “Choose Device”. Select the Pi board you’re using; we used a Raspberry Pi 4B. 5. Click “Choose OS”. We suggest that you select “Raspberry Pi OS (Other)” for the operating system, then scroll down and select “Raspberry Pi OS Full (64 bit)”, which will install a whole range of applications and tools – see Screens 1 & 2 opposite. If instead you choose the vanilla “Raspberry Pi OS (64 Bit)”, it omits a lot of very handy utilities and tools. 6. On the “Would you like to apply OS Customisation Settings?”, click “Edit Settings” and enter the following (this is not essential, but does mean your SD card is pre-loaded with this detail making setup easier): a A host name that is simple and you will remember. We used “TGMRadio”. b Untick “Set Username and Password”. We left the password blank, as this device is in our locked shed. You might consider this a risk, so we leave this choice up to you. c Tick “Configure Wireless LAN”. In SSID, put in the SSID of the WIFI network you want the Pi to use. Type your WiFi password into the provided box. d Click “SAVE”. 7. You will now be back at the screen with “Would you like to apply OS Customisation Settings?”. Click “Yes”. a You will get a screen saying, “All existing data (on your SD card) will be erased, Are you sure you want to continue”. b Click “Yes”. 8. Wait until the data is written and checked. 9. Remove the microSD card. Now let’s run through the initial boot and getting it all running. Initial boot: 1. Install the microSD card into your Raspberry Pi; connect a keyboard, monitor and mouse and apply power. You can’t do the initial setup using a Bluetooth keyboard and mouse, although these are OK once RPi OS is configured. The operating system initially looks for them on USB rather than Bluetooth. 2. Upon booting, you will be asked for your country and time zone. Put this data in. 3. Then create a username and password if you want to. Keep this as something you won’t forget. siliconchip.com.au 4. If you didn’t set the WiFi SSID and network password in the Raspberry Pi Imager tool, enter them now. 5. Click OK to let the system update itself from the Raspberry Pi servers. This might take a few minutes. 6. Once everything is up to date, click Restart. The system will reboot straight to the desktop. 7. If you have a Bluetooth keyboard and mouse, now is the time to pair them. The Bluetooth menu is at the top right of the screen; click this and follow the prompts to pair your devices. You can now dispense with the wired devices you used for setup. 8. Send sound to the AV jack by right-clicking on the speaker symbol at the top right of the screen and selecting AV Jack. Once you have RPi OS or your favourite application loaded on the SD card, you will be able to update and load music and applications via your WiFi (or wired Ethernet) connection. RPi OS updates itself over the internet, so long-term support for the operating system will be fine. At this point, we can start assembling the case. Screen 1: we recommend that you install a full Raspberry Pi OS; choose “Raspberry Pi OS (Other)”. Overall build and assembly Printing the parts is not at all hard, but will take a while. We used the following settings: • 10% fill • 1.6mm wall thickness • No supports • No build plate adhesion • Speed will be specific to your printer; we were running around 180mm/s • Material: PLA (or whatever plastic you are using) The overall system comprises the following parts, which are shown in Table 1. We used about one reel of filament in total. We suggest you have two on hand as you always run out at exactly the wrong time. With Klipper and our selected print speeds, we saw the print times reduced by around 40%. Of course, the print time will vary from printer to printer. The last two files listed are the plain speaker with a grille built into the print. We don’t think it’s essential, but you might prefer this. It will definitely give your printer a workout. You need to print one each of the files, except that you either print “Internet Radio Final V1.0 - Speaker x.stl” or “Internet Radio Final Speaker siliconchip.com.au Screen 2: next, select “Raspberry Pi OS (Full)”, which will install many useful programs alongside the operating system. Table 1 – Part name Filament weight Est. print time Internet Radio Final V1.0 - Case Handle.stl 35g 2 hours Internet Radio Final V1.0 - Case Rear Panel.stl 69g 4 hours Internet Radio Final V1.0 - Case SD Hatch.stl 2g 8 minutes Internet Radio Final V1.0 – Case.stl 290g 16 hours Internet Radio Final V1.0 - Speaker 1/2 Rear Panel.stl 62g each 3 hours Internet Radio Final V1.0 - Speaker 1/2.stl 255g each 15 hours Internet Radio Final Speaker 1 With Grille.stl 266g 18 hours Internet Radio Final Speaker 2 With Grille.stl 266g 18 hours Internet Radio Final Speaker 1/2 With Grille Large Holes.stl 280g each 18 hours Australia's electronics magazine February 2026  49 Fig.5: the ground circuit from the plugpack input to the audio output jack on the Raspberry Pi is far from clean, so some creative ground wire routing is required. Screen 3: click Yes here to customise the operating system configuration. Fig.6: this is how we will wire everything up once they come together in the case. Details will be in the second and final part of this series next month. x With Grille.stl”, not both (where x is 1 for the left speaker or 2 for the right). We have tested the speaker prints with grilles, but all our work was with the plain speakers without grilles. Aside from the investment in time, the case should be pretty straightforward. As you go, check that the parts actually go together. They did on our multiple prints, but that is using a sample set of one printer. Our printer is not modified or special, so we expect most people will achieve similar results. We have used moderately generous margins and expect that most printers will replicate the end result we achieved. There should be minimal post-­ processing required. Still, if you were to fill, sand and paint this, you could 50 Silicon Chip achieve a real retro ‘silver’ boombox outcome. As well as STL files, the download package contains the Fusion 360 files so that you can modify them. We apologise that our novice approach to the design is indeed naïve. We make it available for what it is worth. Wiring it up We really wanted to use a single power supply for this, which simplifies its use. This also leaves open the possibility of running this from a 3.8Ah or similar LiFePO4 battery. A challenge created by using a single power supply with a buck regulator deriving 5V DC for the Raspberry Pi is noise. By powering the amplifier and the Raspberry Pi from the plugpack, the circuit from the power pack Australia's electronics magazine to the Raspberry Pi ground has noise induced on it, as shown in Fig.5. It might seem that this is fussing over things, but our initial approach with wiring was to hook everything together using the input socket as the star ground point. We were really surprised at the level of noise that resulted. The easy way to eliminate this noise is to power the Raspberry Pi from a separate isolated power supply, which is an option you might consider. If you power the Raspberry Pi from its own plugpack (omitting the DC/DC converter) and power the amplifier from its own plugpack, all the noise problems go away, but you now need two plugpacks to power the system. The alternative is to follow our guide to move the amplifier’s ground siliconchip.com.au Parts List – Internet Radio Screen 4: fill in your preferred configuration on this screen. reference to the Raspberry Pi’s GND output, which helps considerably. It is not perfect, but for a ‘medium-fi’ internet radio, it does the job. To achieve this, we connect the ground for the amplifier to the ground of the 3.5mm audio plug that goes into the Raspberry Pi, and run a ground wire from the 3.5mm connector back to the power supply input. The resulting configuration is shown in Fig.6. 1 Raspberry Pi 4B 4GB [Altronics Z6302G] OR 1 Raspberry Pi 5 4GB [Altronics Z6302J] AND 1 Raspberry Pi audio adaptor (untested) [Altronics D0290] 1 7-inch (178mm) LCD touchscreen with 1024 × 600 resolution [Altronics Z6516A] 1 32GB+ microSD card [Altronics DA0329] 1 microSD card adaptor (required if your computer has no microSD/SD card interface) [Altronics D0433A] 1 8-32V to 5V 5A USB-C DC-DC converter [Altronics M7832] 1 TPA3110 2 × 30W stereo audio amplifier with Bluetooth [Altronics Z6409] OR 1 TPA3110 2 × 30W stereo audio amplifier [Altronics Z6407] 1 15mm diameter knob to suit spline shaft [Altronics H6540] 1 18V DC 2.8A plugpack [Altronics M8951] 2 SPDT solder tail miniature toggle switches [Altronics S1310] 1 2200μF 35V 18mm diameter electrolytic capacitor [Altronics R6207 or R5206] 2 100mm loudspeaker drivers (optional) [Altronics C0635] 1 wireless USB keyboard [J.Burrows KB210 Wireless Keyboard from Officeworks] 1 wireless USB mouse 1 HDMI to HDMI cable (included with LCD touchscreen) 1 micro HDMI to HDMI adaptor (for secondary display) [Altronics P1925] 1 micro Type-B USB to USB Type-A cable (included with LCD touchscreen) 1 piece of acoustic speaker wadding (optional) [eg, open-cell foam from packing] Hardware & connectors 1 2.1mm inner diameter chassis-mount barrel socket [Altronics P0622] 2 2-way vertical polarised headers [Altronics P5492] 5 2-way polarised header plugs and pins [5 × Altronics P5472 + 10 × Altronics P5470A] 1 3.5mm stereo jack plug [Altronics P0030] 2 4mm red captive head binding posts [Altronics P9252] 2 4mm black captive head binding posts [Altronics P9254] 1 HDMI socket to micro HDMI plug adaptor [Altronics P7374A or P1925] 2 right-angle HDMI adaptor [Altronics P7371A] 1 2m length of red light-duty hookup wire [Altronics W2250] 1 2m length of black light-duty hookup wire [Altronics W2251] 1 1m length of green light-duty hookup wire [Altronics W2255] 22 9mm-long Jiffy box self-tapping screws [Altronics H1139 – pack of 25] 22 M3 flat washers [Altronics H3180 – pack of 25] 2 M4 × 16-20mm panhead machine screws [Altronics H3320A – pack of 25] 2 M4 flat washers [Altronics H3385 – pack of 25] 2 M4 hex nuts [Altronics H3380 – pack of 25] 1 200mm length of 4mm diameter heatshrink tubing 1 200mm length of 3mm diameter heatshrink tubing 1 200mm length of 2mm diameter heatshrink tubing 10 100mm-long, 2.5mm-wide cable ties [Altronics H4031A] 12 12mm round adhesive slim rubber feet (optional) [Altronics H0896 – packet of 4] Next month If you’re building the Internet Radio, you can start printing the case pieces in preparation for next month’s follow-up article. It will have the details on wiring up the modules, mounting them in the case, finishing the software setup and getting the SC Radio up and running. The finished Internet Radio has a handy integrated carrying handle. The volume knob is on the top. siliconchip.com.au Australia's electronics magazine February 2026  51 Subscribe to JANUARY 2026 ISSN 1030-2662 01 The VERY BEST DIY Projects ! 9 771030 266001 $14 00* NZ $14 90 Digital Command Control INC GST DCC Base Station to provide power and data to mod el railway tracks Acoustic Imaging using a camera and micropho ne array to ‘see with sound’ Power Electronics, Part 3 the properties of transformers and inductors, and their use in power converters Weatherproof Touch Switch a simple switch with no moving parts that can be used outdoors Earth Radio, Part 2 hear solar and atmospheric disturba nces with this receiver Australia’s top electronics magazine Remote Speaker Switch 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. switch up to six pairs of speakers connected to a single amplifier 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 Length Print Combined Online 6 months $72.50 $82.50 $52.50 1 year $135 $155 $100 2 years $255 $290 $190 6 months $85 $95 1 year $160 $180 2 years $300 $335 6 months $105 $115 1 year $200 $220 All prices are in Australian dollars (AUD). Combined subscriptions include both the printed magazine and online access. 2 years $390 $425 Prices are valid for the month of issue. Try our Online Subscription – now with PDF downloads! DCC Base Station; January 2026 Humanoid Robots; November-December 2025 An online issue is perfect for those who don’t want too much clutter around the house and is the same price worldwide. Issues can be viewed online, or downloaded as a PDF. To start your subscription go to siliconchip.com.au/Shop/Subscribe INC GST Stereo signal processing Deep adjustable notch Q and frequency adjustments Silent power-on and power-off Flexible power requirements Power supply: 9-15V DC <at> <100mA Notch Frequency: 50Hz or 60Hz Notch adjustment: ±2% (±1Hz <at> 50Hz) Q adjustment: 7-14.5, with 10.5 recommended Notch depth: typically >30dB Mains Hum Notch Filter Project by John Clarke Long unbalanced audio signal leads can pick up significant mains hum. This stereo mains hum notch filter can help to reduce it to inaudible levels. SC7598 Kit ($50 + postage): includes the PCB and all onboard parts. You just need to add the case and power supply. W hen using long audio leads between a signal source and amplifier, most of the time, there will be mains induction in the leads from nearby power wiring. If balanced leads are used, the amount of mains signal pickup will be the same in the twisted pair signal wires and can be cancelled out at the receiving end. But with unbalanced leads, the hum pickup remains. Proper grounding methods and using balanced leads with unbalanced-­ tobalanced and balanced-to-­unbalanced converters at each end will prevent hum in most cases. Sometimes, full galvanic isolation between sections is necessary due to different signal grounds and can be accomplished using audio isolation transformers. These techniques are fully detailed at siliconchip.au/link/ac9f siliconchip.com.au So, with the correct Earthing methods and use of balanced leads and isolation transformers where necessary, you shouldn’t have any hum. But what if, for example, you are living in a house that is already wired up with audio leads and there is hum? The best method to remove it is to rewire it using balanced leads with converters at the signal source and receiver ends, but that may not be practical. That leaves the possibility of removing the hum with an audio filter. The effect on audio frequency response will be minimal, provided that the filter produces a deep and narrow notch that centres around the mains frequency. Note that the filter will only work if the mains hum is due to pickup in the interconnecting leads. It won’t necessarily help if the hum is caused by Australia's electronics magazine ground loops or the power amplifier is producing the hum. Our Notch Filter has a stereo input and output with a mains notch filter between them to reduce the hum signal component dramatically. The Notch Filter is connected using RCA leads at the power amplifier input, so that the signal passes through the filter before being applied to the power amplifier. It is powered from a DC plugpack that provides 9-15V DC. The power requirements are modest; a 100mA plugpack is more than adequate. The filter is housed in a small instrument-­style enclosure with the left and right channel RCA inputs at the rear and the outputs on the front. The DC input socket and power-on indicator are also at the front. The Notch Filter operates silently when power is switched on and off February 2026  53 0dB -6dB -12dB -18dB -24dB -30dB Q=7 | VR1, VR5 anti-clockwise Q=10.5 | VR1, VR5 centred Q=14.5 | VR1, VR5 clockwise -36dB -42dB -48dB -54dB -60dB 44Hz 46Hz 48Hz 50Hz 52Hz 54Hz 56Hz 58Hz 60Hz Fig.1: a simulation of the Fliege filter showing how the width and depth of the notch varies as the Q factor is adjusted using a variable resistance. using reed relays to keep the signal isolated from the circuitry until voltages stabilise. When switching it on, the relays remain off for about five seconds before being energised, preventing DC voltage swings at the output. At switch-off, the relays open immediately, preventing DC shifts in the audio output as the power decays. Filtering is achieved using what is called a Fliege notch filter. This has the advantage of being adjustable in frequency over a small range using a single trimpot. This simple frequency adjustment is not possible with a Twin-T filter. Both an active Twin-T and Fliege filter can be adjusted for filter Q with a single potentiometer. For more information on such filters, see www.ti.com/lit/pdf/slyt235 Fig.1 shows the simulated response of the notch filter. The notch is at 50Hz; three Q values are shown, covering the adjustment range of our filter. The higher Q values have a narrower notch and so less of the audio band is affected. A Q of around 10 usually provides the best compromise, allowing a small amount of variation in the mains frequency while maintaining a good notch depth. This filter could be used for an alternative purpose, such as nulling out mains control tones that may encroach into your audio signal. These tones are superimposed on the mains and are used to control things like street lights and off-peak loads. The tones may enter the audio pathways within your preamplifier. Typically, mains control signals are at 492Hz, 750Hz and 1050Hz. Changing the filter components can provide a notch at any one of those frequencies. Performance Fig.2 shows the frequency response of the unit, which is quite flat except for the obvious notch centred around 50Hz. The response is +0,-1dB from 20Hz to 20kHz except between 45Hz and 55Hz (5Hz either side of the notch frequency on our prototype). If you look only at the response above the Fig.2 (left): except for the notch region, the circuit’s frequency response is flat within +0,-1dB from 20Hz to 20kHz. It’s only down by 1dB <at> 20Hz and is otherwise ruler-flat from 100Hz to 20kHz. Fig.3 (right): a close-up of the 40-60Hz region in Fig.2, showing the notch in more detail. We didn’t quite tune ours to exactly 50Hz but then it’s unrealistic to assume every constructor will tune it perfectly. It still has good attenuation at exactly 50Hz, showing why you don’t necessarily want to set the Q factor to maximum. Also, the mains frequency drifts a little over the course of a day. 54 Silicon Chip Australia's electronics magazine siliconchip.com.au is quite usable even with lower-level signals such as ‘line level’ (~775mV RMS), where the THD+N reading is around 0.004%. There are a couple of odd steps/ shelves in Fig.5, we suspect due to resonance effects from the notch filter (and possibly quirks of our test equipment). These deviations are not so large as to be concerning. Circuit details The rear of the Notch Filter case features the input connectors and little else. notch, it’s flat within a fraction of a decibel. Fig.3 shows a ‘zoomed in’ view of the 40-60Hz region so you can see the notch. This is with a Q set to the recommended value of around 10. You can see the depth is around 27.5dB; it could be set deeper with a higher Q value, but then the sides would be steeper, and it would need to be set more precisely. You can see that we set ours to around 49.8Hz, so even though it isn’t perfectly accurate, the attenuation is still over 20dB at 50Hz. Distortion performance is good, with measurements at a variety of signal levels shown in Fig.4. 2.3V RMS was chosen as it’s a typical value that you would get from a CD, DVD or Bluray player and with that signal level, it gives a THD+N reading below 0.002% across much of the audible frequency range. Performance at 2V and 1V is slightly worse because the signal is closer to the noise floor. The signal-to-noise ratio (SNR) with a 2.3V RMS signal is about 100dB. Fig.5 shows how the THD+N at 1kHz varies with the signal level. As you’d expect, the THD+N figure goes down as the signal level goes up due to the improving SNR. With a THD+N figure better than 0.01% for signals of ~350mV RMS and above, this unit The circuit of the Notch Filter is shown in Fig.6. It comprises nine op amps and a 555 timer IC. Eight of the op amps are within two quad op amp ICs. Some op amps provide buffering, some active filtering and another provides a low-impedance half supply. The timer is used to provide a delayed signal switch-on at power-up. The signal common throughout most of the circuit is set at half supply (~4.5-7.5V) so the signal can swing symmetrically within the supply rails. Having a positive ground reference means that we can use a single supply rail (a negative rail is not required), which can be provided by a standard DC plugpack. The half-supply rail is derived using two 10kW resistors connected across the main supply, resulting in a nominal 6V level when a 12V DC supply is used. This is then decoupled with a 100μF capacitor and buffered by IC3 Fig.4 (left): the distortion of this circuit is generally pretty low (the spike around 50Hz, in the notch, is to be expected). The best performance is at 2.2-3V, which is exactly what many DACs and CD/DVD/Blu-ray players will deliver. Fig.5 (right): at 1kHz, the THD+N figure is below 0.01% for all signal levels from 350mV RMS up to 3V RMS. siliconchip.com.au Australia's electronics magazine February 2026  55 to provide a low-impedance reference voltage from its output. The common reference half-supply voltage from pin 6 of IC3 is used in the filter circuitry for both channels. Only the left channel circuitry is shown on the diagram, with the right channel being identical except for the component labelling; the designators for the other channel are shown in brackets. The signal for the left channel comes via CON1 and is biased to 0V by a 100kW resistor. This discharges any AC coupling capacitor that could be in the signal line before CON1, and makes the signal swing about ground. The ground connection for CON1 is via a 10W resistor to reduce possible Earth loop currents between interconnecting leads. The ferrite bead (FB1) and 150W resistor provide high-­ frequency attenuation of radio signals that could otherwise be picked up and accidentally demodulated to produce spurious audio signals. Following the 150W stopper resistor, the signal is AC-coupled to the non-­ inverting input of IC1a. This input is biased at the half supply via a 100kW resistor. The output from IC1a’s pin 1 drives the notch filter that comprises op amps IC1d, IC1c and IC1b, several resistors and the two 47nF capacitors, Cx and Cy. The Rx and Ry resistances are formed using either VR3 and VR4 or the fixed-value resistors, R3a/R3b and R4a/R4b. Assuming Cx = Cy and Rx = Ry, the notch filter frequency is 1 ÷ 2πRxCx. For a 50Hz notch and 47nF capacitors, Rx and Ry should both be 67.7255kW. This resistance is made up using a 62kW and 5.6kW resistor in series, or using VR3 and VR4 set to this resistance. For a 60Hz notch, the resistances are different, as shown on the circuit diagram. The resistance values are suitable when the 47nF capacitors are actually within ±1% of 47nF (about 46.5~47.5nF). If you don’t use 1% capacitors, their values could differ by 5%. The capacitors need to be chosen so that they are within 1% of each other, but not necessarily within 1% of 47nF. VR3 and VR4 are then adjusted to set the notch to the correct frequency. How these are adjusted is described towards the end of the article. Once the notch filter is adjusted correctly, a small frequency adjustment is also available using VR2. This provides a frequency trim to get the best null from the notch filter. The adjustment uses feedback from the notch output at pin 14 of IC1d back to the filter components. VR2 adjusts the signal level difference between the filter input and output, with the 22kW resistors setting the frequency range adjustment limits. The circuit only works with a frequency adjustment over a small range, so the notch depth remains relatively unchanged over the adjustment range. The filter Q is adjusted with VR1. This sets the narrowness of the notch. The higher the Q value, the narrower the frequency range over which the notch will attenuate the signal. A narrower notch will affect the audio signal less, but allows for less variation in the signal frequency you want to Fig.6: only the left channel is shown here; the right channel is identical, with the corresponding designators shown in brackets. The signal chain includes a buffer (IC1a), half-supply generator (IC3), the Fliege notch filter (IC1b/c/d), output isolation reed relay (RLY1), a timer to drive the relay (IC4) and a regulator to power the relay (REG1). 56 Silicon Chip Australia's electronics magazine siliconchip.com.au remove before it will go outside the notch region. Typically, a Q of around 10 is a good compromise. VR1 allows a Q adjustment of between about 7 and 14.5, with 10.5 being at the centre position. For typical mains frequency excursions from 49.75 to 50.25Hz, the notch filter provides a minimum attenuation of 16dB for a Q of 14.5 and 23dB for a Q of 7. The attenuation in the centre of the notch stays constant over any of the Q settings. Two series 10kW resistors reduce the filter signal level input by half. Fliege filters typically then apply this signal to the other side of the Cx capacitor, and the resistance values are much higher than the 10kW values we use. To adjust the Q, both these resistors need to be changed to alter the overall resistance, but they must also maintain the same ratio. The Q value is the parallel resistance of the two divider resistors divided by Rx. So, for a Q of 10, the divider resistors need to be 20 times larger than Rx. Our Fliege filter has a modification where the half-signal level of the filter input is applied to the input of buffer IC1b. Since the divider ratio is maintained at ½, the Q is adjusted using a single resistance change following the buffer output. For this arrangement, the Q is calculated as the RQ value over the Rx value. Following the filter at pin 14 of IC1d, the signal is AC-coupled using a 1μF capacitor. The 100kW resistor to ground makes the output signal swing around 0V. The RLY1 contact connects the signal to the output via a 150W stopper resistor. This prevents IC1d from oscillating should capacitive loads be connected to CON3. Power Power for the circuit is supplied via CON5 with a 9-15V DC supply from a DC plugpack. Reverse polarity protection is provided by diode D4, and the supply is then filtered by a 470μF 16V capacitor. This voltage, labelled V+, powers all the op amps. REG1 is a 5V regulator to supply the 555 timer, IC4. REG1 includes a 100μF capacitor at its output to prevent regulator oscillation and improve transient response. The 555 could run off V+ but then its output would need to be regulated to 5V to drive the relay coils, so it’s easier to just regulate the voltage applied to IC4. siliconchip.com.au Parts List – Mains Hum Notch Filter 1 double-sided, plated-through PCB coded 01003261, 129 × 101.5mm 1 140 × 110 × 35mm plastic instrument enclosure [Jaycar HB5970, Altronics H0472] 2 red PCB-mounting RCA sockets (CON2, CON4) [Altronics P0145A] 2 white or black PCB-mounting RCA sockets (CON1, CON3) [Altronics P0147A] 1 PCB-mounting barrel socket (CON5) [Jaycar PS0520, Altronics P0621A] 2 small ferrite beads (FB1, FB2) [Jaycar LF1250, Altronics L5250A] 2 SPST 5V reed relays (RLY1, RLY2) [Jaycar SY4036] 2 500kW top-adjust, single-turn trimpots (VR1, VR5) 2 1kW top-adjust single-turn trimpots (VR2, VR6) 2 14-pin DIL IC sockets 2 8-pin DIL IC sockets 4 No.4 × 6mm self-tapping or M3 × 5mm panhead machine screws Semiconductors 2 TL074 quad JFET-input op amps, DIP-14 (IC1, IC2) 1 TL071 single JFET-input op amp, DIP-8 (IC3) 1 555 timer, DIP-8 (IC4) 1 78L05 5V 100mA linear regulator, TO-92 (REG1) 1 BC337 45V 0.8A NPN transistor, TO-92 (Q1) 3 1N4148 75V 200mA signal diodes (D1-D3) 1 1N4004 400V 1A diode (D4) 1 3mm or 5mm LED (LED1) Capacitors 1 470μF 16V PC electrolytic 2 1μF 16V PC electrolytic 5 100μF 16V PC electrolytic 2 220nF MKT polyester 2 10μF 16V PC electrolytic 3 100nF MKT polyester Resistors (all ¼W axial ±1%) 1 470kW 7 10kW 2 470kW (for 50Hz notch) 1 4.7kW 2 390kW (for 60Hz notch) 1 620W 8 100kW 4 150W 4 22kW 2 10W Extra parts for the 1% capacitor version 4 47nF ±1% polypropylene capacitors [RS Components 166-6465] 4 62kW ±1% ¼W axial resistors (R3/4/7/8a) for 50Hz 4 5.6kW ±1% ¼W axial resistors (R3/4/7/8b) for 50Hz 4 56kW ±1% ¼W axial resistors (R3/4/7/8a) for 60Hz 4 430W ±1% ¼W axial resistors (R3/4/7/8b) for 60Hz Extra parts for the 5% capacitor version 4 47nF ±5% MKT polyester capacitors with closely matched values 4 100kW top-adjust multi-turn trimpots (VR3, VR4, VR7, VR8) Power indicator LED1 is supplied via a 620W series resistor and provides a consistent light output regardless of the input supply voltage, provided this is between 9V and 15V, sufficient to keep REG1 in regulation. Several additional capacitors are used to bypass the supply for the four ICs. Relay operation The two relays (RLY1 and RLY2) switch the output signals to prevent thumps (large voltage excursions) at power-up and power-down. IC4 delays relay switch-on after power up to allow everything to stabilise first. IC4 is connected as a monostable timer, with the pin 3 output high (5V) Australia's electronics magazine at power-up. This is because the pin 2 (trigger) input is lower than 1/3 of the supply voltage due to the 10μF capacitor being initially discharged. The output at pin 3 stays high until the 10μF capacitor voltage at pins 2 and 6 rises to above 2/3 of the supply voltage, whereupon the pin 6 (trigger) input signals the pin 3 output to go low. This time period is around five seconds due to the time constant of the 470kW resistor charging the 10μF capacitor. At this point, the relays switch on due to IC4’s pin 3 output going low, while transistor Q1 is switched on due to the incoming supply voltage being applied to its base via a resistive divider. Q1’s collector February 2026  57 is connected to the 5V supply, so any voltage 0.7V above this will cause base current flow, switching Q1 on. So the relays are energised a few seconds after power is applied. When the incoming voltage drops, Q1 loses its base current and so disconnects power from the relay coils. This therefore disconnects the output signals immediately from CON3 and CON4. Diodes D1 and D2 across the relay coils clamp the reverse voltage developed when the relays are switched off, and this charges the 100nF capacitor. The diodes prevent excess voltage from damaging Q1. Diode D3 is used for reverse-polarity protection since this part of the circuit is powered from before diode D4. That diode also prevents the 470μF filter capacitor from holding up Q1’s base at switch-off. Capacitor selection As mentioned, the 47nF capacitors for the notch filter need to be selected so that the values are within ±1% of each other. Typically, if you buy standard ±5% capacitors on a bandolier (paper/cardboard tape), the adjacent components will have a similar value. We found that four capacitors of the same marked value in a row wouldn’t necessarily be within ±1% of 47nF, but whatever value they were, three would be within ±1% of each other. You may need to get more than four capacitors so that at least four will be of a similar value. That’s a lot cheaper than purchasing 1% capacitors, although 1% capacitors can be used if you want. If you have a capacitance meter, the values can be measured and compared. Alternatively, if you have an oscilloscope or frequency meter, the capacitors can be tested using a standard astable oscillator made with a 555 or 7555 timer. The frequency of oscillation will be inversely proportional to the capacitance of the timing element. Fig.7 shows the circuitry required. Using 10kW for RA and RB, the oscillation frequency would be around 1023Hz for a 47nF capacitor. Note that the oscillator does not allow you to accurately determine the exact capacitance value. However, it is suitable for comparing the values of several capacitors as long as you make the measurements at around the same time, so there are no ambient temperature change effects affecting the readings. Select capacitors that run at the same frequency to within 1%. A 1% variation in capacitor value will mean that the oscillator frequencies will be within about 10Hz using this circuit. Construction The Audio Notch Filter is built using a double-sided, plated-through PCB coded 01003261 that measures 129 × 101.5mm. It is housed in a plastic instrument enclosure measuring 140 × 110 × 35mm. All the parts are through-hole types that mount on the top side of the circuit board. Some resistor values depend on whether you are setting the notch filter to 50Hz or 60Hz. The resistors that vary are R1, R2, R3a, R3b, R4a, R4b, R7a, R7b, R8a and R8b. There are two options when building the notch filter. One is to use ±1% 47nF capacitors and fixed 1% resistors for R3a, R3b, R4a, R4b, R7a, R7b, R8a and R8b. Alternatively, use similar-­ value 47nF capacitors and adjustable resistors (trimpots) VR3 and VR4 for the left channel and VR7 and VR8 for the right channel. This allows for trimming of the notch frequency. 47nF ±1% capacitors are hard to find and expensive, so our kit includes the trimpots. If using fixed resistors, their values are shown on the circuit diagram for 50Hz and 60Hz notch frequencies. Do not use both the trimpots and fixed resistors. Follow the overlay diagram (Fig.8) and begin construction by installing the resistors and four diodes. Check the value of each resistor before installation by measuring with a multimeter (they have colour-coded stripes but it can be hard to distinguish some colours). Fig.7: a simple circuit for an oscillator that produces a signal frequency proportional to the capacitor under test. Fig.8: follow this overlay diagram to assemble the PCB. This shows all the fixed resistors and trimpots fitted, but you should either install VR3/4/7/8 or R(3/4/7/8)(a/b), not both sets. Take care with orientations of the ICs, diodes, LED, trimpots and electrolytic capacitors. 58 Silicon Chip siliconchip.com.au Diodes D1, D2 and D3 are small glass-encapsulated 1N4148 types, while D4 is a larger, plastic-cased 1N4004 diode. Ensure these are all fitted with the orientations shown in the overlay diagram and PCB screen-­ printing. Mount ferrite beads FB1 and FB2 using resistor off-cut wires fed through the centre hole and then bent to insert into the PCB holes. Now install the sockets for the four ICs, taking care to orientate them correctly, with the notches facing as shown. Next are RCA sockets CON1 to CON4 and the DC socket, CON5. We used white for the left channel and black for the right channel. Red sockets should be used for the right channel sockets, not black as in the photos, as this is the standard colour for the right channel. However, at the time we purchased these, the red sockets were out of stock at Altronics and Jaycar only sells the black type. Trimpots VR1/VR5 (500kW) and VR2/VR6 (1kW) can be installed now. If VR3, VR4 and VR7 and VR8 are being used, the adjustment screws need to be orientated as shown. That’s so the resistance changes with clockwise direction as indicated on the circuit. For the 500kW and 1kW trimpots, be sure to place the correct value in each position. The trimpots will have printed codes, but you can also check the value by measuring the resistance between the outer two leads. Transistor Q1 and the 5V regulator (REG1) can be mounted now, taking care to orientate these correctly. They are in the same TO-92 package, so check the correct one is placed in each position before soldering. Relays RLY1 and RLY2 can be installed now as well. The capacitors are next. Electrolytic types need to be orientated with the correct polarity; the longer lead goes into the pad marked +, with the striped (negative) side of the can near the opposite pad. The MKT and ceramic types can be installed either way around. LED1 sits horizontally with the leads bent at 90°. Position the LED so that the top of the lens dome is 12mm in front of the PCB edge, and the centre of the LED is 5mm above the top surface of the PCB. When bending the leads, make sure the anode and cathode leads will go into the correct pads on the PCB (the longer anode lead goes to the pad marked ‘A’). siliconchip.com.au Removing mains harmonics In many areas, the mains voltage is not a reasonably shaped sinewave. Typically, the waveform is distorted and has a flattened top, as shown at the top of Fig.a. This shows a typical mains voltage waveform. The flattened top is mainly due to industrial and household appliance power supplies that draw power from the peaks of the mains waveform. Nonlinear loads will also cause flat-topping. Note the difference in shape between the measured mains waveform in yellow and the cyan sinewave trace. So, while the main fundamental mains frequency is 50Hz (or 60Hz in some other countries), the distorted waveshape means that the waveform includes harmonics of that frequency. These are primarily the odd harmonics (3rd, 5th, 7th etc), which are at 150Hz, 250Hz, 350Hz etc for a 50Hz mains (or 180Hz, 300Hz, 420Hz etc for 60Hz mains). Fig.b shows the frequency components and levels that are present in the distorted mains signal. The horizontal axis is 50Hz per division, while the vertical axis is 10dB per division. The fundamental at 50Hz is followed by harmonics at 150Hz, 250Hz, 350Hz and 450Hz. The third and fifth harmonics (150Hz and 250Hz) are only about 26dB below the 50Hz fundamental. You may need to notch out these harmonic frequencies as well as the fundamental if they are intrusive. This can be done with more notch filter circuits, connected in series, with one set for the fundamental (50Hz or 60Hz) and further notch filters tuned to the harmonic frequencies. If building such a system, the two relays and 555 timer and associated circuitry (such as Q1, D1-D3 etc) are only required in the final filter circuit, to disconnect the output during power-up and power-down. The power supply can be paralleled from one notch unit to the other, provided the plugpack can supply the extra current. Wire links would need to replace the relay contact positions on the PCB. These could all be installed in the same, larger box with fixed wiring from the output of one stage to the input of another. The filter component values to change to notch different harmonics are listed in Tables 1 & 2. We show the capacitor and resistor values for the various fundamental and harmonic notch components. Resistors R1 and R2 are unchanged at 470kW regardless of the notch frequency. Freq. Cx & Cy R*a R*b Freq. Cx & Cy VR3/4/7/8 Initial 50Hz 47nF ±1% 62kW 5.6kW 50Hz 47nF 100kW 67.73kW 150Hz 15nF ±1% 68kW 2.7kW 150Hz 15nF 100kW 70.74kW 250Hz 10nF ±1% 62kW 1.6kW 250Hz 10nF 100kW 63.66kW 60Hz 47nF ±1% 56kW 430W 60Hz 47nF 100kW 56.43kW 180Hz 15nF ±1% 56kW 3.0kW 180Hz 15nF 100kW 58.95kW 300Hz 10nF ±1% 51kW 2.0kW 300Hz 10nF 100kW 53.05kW Table 1: fixed components for mains harmonics Table 2: adjustable components for mains harmonics Fig.a: while the mains waveform is theoretically a sinewave (and is produced as a sinewave by the steam turbine alternators in large-scale power plants), by the time it reaches you, it will usually be flat-topped like this (top yellow trace). Compare it shape to the pure sinewave in cyan below. Fig.b: the spectrum of the mains waveform shows the 0dB fundamental at 50Hz with a series of harmonics at lower levels: the third (150Hz, -26dB), fifth (250Hz, -28.5dB), seventh (350Hz, -44dB) etc. Australia's electronics magazine February 2026  59 The ICs can now be inserted into their sockets, making sure that the pin 1 dot or notch is near the notch on the socket in each case. Also ensure that the leads don’t fold under the body during insertion, instead going into the holes on the socket. Panel cutouts The required holes in the panel pieces are as specified in Fig.9. It shows the positions of the holes for the LED (3mm diameter), RCA sockets (9mm diameter) and the DC socket (12mm diameter). Fig.10 shows the panel labels. You can download these as a PDF from siliconchip.au/Shop/11/3584, print them onto vinyl labels (or similar), ready to attach to the panels. Holes can be cut out with a sharp craft knife. For more information on making panels, see siliconchip.au/Help/FrontPanels Once the panels are completed, place the front and rear panels onto the RCA and other protruding components and slide the panels with the PCB into the baseplate of the enclosure. Secure the PCB to the enclosure base with No.4 self-tapping screws (short M3 machine screws can be used; the threads will self-tap the plastic posts). Setting it up SC7598 Kit ($50 + postage): includes the PCB and all onboard parts. You just need to separately purchase the case (shown above) and power supply. Initially, set VR1, VR2, VR5, VR6 at their mid positions. If using ±1% capacitors and fixed resistors, then skip to the section titled “Adjustments”. Adjust VR3, VR4, VR7 and VR8 to 67.73kW for a 50Hz notch or 56.43kW for a 60Hz notch. You can measure this resistance using a multimeter across the test points: TP3a/b for VR3, TP4a/b for VR4, TP7a/b for VR7 and TP8a/b for VR8. Connect a 9-15V DC plugpack and check that LED1 lights with the power switch on. Disconnect the power and insert IC1, IC2, IC3 and IC4 into their sockets. Be sure to orientate each correctly; IC4 is the 555. Apply power and measure the supply current, which should be less than 50mA. If your 47nF capacitors are all outside the 1% tolerance of 47nF (below 46.5nF or above 47.5nF), then VR3, VR4 and VR7 and VR8 will require trimming for best nulling of the mains frequency. You can use a signal Australia's electronics magazine siliconchip.com.au The completed circuit board housed in the case, with the lid off. This prototype used trimpots and ±5% capacitors rather than fixed resistors and ±1% capacitors. 60 Silicon Chip Fig.9: drill the holes in the front and rear panels as shown here. Fig.10: the panel labels for the front and rear of the device. The holes are drawn undersized here to allow for slight misalignment; use the holes in the panels as a guide to cut them out after attaching the label. generator set at 50Hz (or 60Hz) with a level of 1V RMS or similar. Alternatively, without a signal generator that is accurate enough, you may need to feed a signal with mains hum into the input, listen to the output and adjust the trimpots to minimise the audible hum. An alternative approach is to attenuate the output of a low-voltage AC plugpack (eg, 9V AC) with a resistive divider, say 100kΩ and 1kΩ. Connect the centre of the divider to one of the inputs and the other end of the 1kΩ resistor to the RCA shell/ground. You can use an oscilloscope or audio millivoltmeter to monitor the signal at the CON3 output, or an amplifier and headphones, earbuds or a loudspeaker to listen to it instead. If using an amplifier, make sure the volume control is turned down to minimum initially, then turn it up slowly when you apply power until you can hear the hum signal to avoid overload. Adjust VR3 and VR4 by small amounts each (either way) to minimise the mains hum in the left channel. siliconchip.com.au Similarly, adjust VR7 and VR8 in the right channel to minimise hum. Try to maintain the same value for each trimpot. Adjustments Adjust VR1 and VR5 to set the Q; higher settings will give a deeper notch but with less allowance for mains frequency variations. You could adjust the Q while monitoring the actual signal you want to remove hum from, allowing you to select the minimum setting that removes audible hum so as to avoid affecting ~50Hz bass in the actual audio signal too much. VR2 and VR6 are for the frequency adjustment for the left and right channels. These allow the notch frequency to be trimmed, they also affect the notch width. The frequency adjustment will be most useful when you are using the ±1% capacitors with fixed resistors. It is usually easier to adjust the frequency when the Q is set to a low value first (VR1 and VR5 set clockwise) before adjusting the Q higher as you Australia's electronics magazine further adjust the notch frequency. A midpoint setting for VR1 and VR7 (a Q of around 10.5) gives a good compromise between notch depth and a wide enough notch to allow for slight SC mains frequency variations. Mains Power-Up Sequencer February-March 2024 Hard-To-Get Parts SC6871: $95 siliconchip.au/Series/412 The critical components required to build the Sequencer such as the PCB, micro etc. Other components need to be sourced separately. February 2026  61 By Tim Blythman Decoder Base Station Using DCC Remote Controller DCC Booster Digital Command Control is a great way to run multiple trains on a layout at the same time. Our DCC Base Station allows control of five locomotives, but there is only the option to directly drive one at a time. The DCC Remote Controller allows more trains to be controlled at the same time, and Image source: www.pexels.com/photo/miniature-train-in-a-garden-9018266/ multiple can be connected to one Base Station! DCC Remote Controller T he previous articles in this series have included the designs for adding DCC (Digital Command Control) to a model railway. The first part, a DCC Decoder, constitutes the electronics that is fitted to rolling stock such as locomotives and self-powered railcars. A decoder uses the electrical DCC signal from the track to control the motor in a locomotive. It can also control lights and accessories, if fitted to the locomotive. The DCC signal provides both power and control commands. The signal is generated by a DCC Base Station and our design was presented last month. It has an LCD touchscreen for user input and status display. It needs only a low-voltage DC supply, typically 12V, to operate. The Base Station offers a main track output for running trains on a layout and a programming track output that can be used to configure decoders through configuration variable (CV) programming. In addition to the two constructional articles, we also ran a feature article about getting started with DCC, including what CVs should be programmed and how to choose the necessary values. That feature focused on using our Decoder and Base Station, but we expect that it will be helpful to anyone starting out with DCC. DCC Remote Controller Many commercial DCC systems offer so-called ‘throttles’, which are units that can plug into a base station There are a handful of regular SMD parts on the PCB, plus some larger ones, such as the RJ45 sockets, the OLED screen and tactile switches. The LED is a throughhole part that’s surfacemounted. Note the extra wire securing OLED display MOD1. 62 Silicon Chip Australia's electronics magazine to expand its capabilities, allowing the control of a locomotive through inputs such as buttons, knobs and switches. Our Controller is in this vein. We have designed it to be simple and inexpensive, so it is not onerous to add more than one. The interface we have designed is both simple and powerful. It uses a straightforward serial data protocol to transmit data. It can be used to send any type of DCC packet directly to the track, meaning that its capabilities are not restricted, even if the DCC standards were to change. The protocol can also be used to send operating commands to the Base Station, including the ability to switch the track power off and on. There is scope to add new and different commands, if necessary. The Base Station firmware was developed with such a Controller in mind, so it does not need a software update to work with this design. Interface We noted in the project article that the Pico 2 microcontroller pins chosen for the extension interface (on CON5 and CON6) of the Base Station are capable of either I2C or UART (asynchronous serial) mode operation. Ultimately, we have chosen to use serial communications, mostly siliconchip.com.au Features & Specifications 🛤 Potentiometer speed knob and six tactile pushbuttons for control 🛤 Multiple Controllers can be daisy-chained 🛤 Status LED and compact OLED display 🛤 Can select decoder addresses independently of the Base Station 🛤 Compact design fits in a UB5 Jiffy box 🛤 Uses common Ethernet (Cat 5/6) cables for connection 🛤 Control pages for three locomotives on each Controller 🛤 Power provided from Base Station; a separate power supply not needed 🛤 Only 12mA current draw per Controller DCC PROJECT KITS DCC Decoder, December 2025 (SC7524, $25) includes everything in the parts list DCC Base Station, January 2026 (SC7539, $90) includes everything in the parts list, except for the case, power supply, glue, CON4 & CON5 headers DCC Remote Controller (SC7552, $35) includes all required parts, except for the UB5 case and wire/cable because programming microcontrollers to work as I2C slaves can be fraught with difficulties. I 2C is also designed for short-­ distance communications within a single PCB; it is short for “inter-­ integrated circuit” after all. Serial data has been proven to work over longer distances, and we are using a low rate of 9600 baud. This rate also means that the time to send a packet on our bus is about the same time as it takes for the Base Station to send it to the track. Fig.11 shows the arrangement of the wiring between a Base Station and a pair of daisy-chained Remote Controllers. The design permits more Remote Controllers to be connected at the right-hand end in the same fashion. The number of connected units is not subject to any hard limit, but will depend mostly on factors such as bus latency and traffic. The general idea is that the bus will allow communications from the Base Station to any connected Controllers (from the upstream end to the downstream end), while also allowing communication from any Controller back to the Base Station in the upstream direction. The first important point is that, like DCC, this data is divided up into siliconchip.com.au packets of various types. One packet type is used to transmit a DCC packet on the rails, so these packets need to encapsulate binary data. We’ll describe the packet format in more detail in the Firmware section. Secondly, the microcontroller we are using has two UART peripherals, with the left-hand end of the Packet Processor using one UART to communicate with the next device upstream. The other UART communicates downstream. We can easily handle the cross-traffic in software. Thus, the Packet Processor is mainly concerned with handling the daisy-­ chain of data lines by moving data between the two UARTs as needed. The green lines in Fig.11 indicate packets sent from the Base Station. At each Controller, the packets are copied, with one copy kept for local processing, and the other sent downstream to the next Controller. The packets can also be modified before being sent out. For example, one packet type passes an index value along the chain. Each Controller adds 1 to the index as it is processed, so each Controller knows its position in the chain. The blue lines indicate traffic heading towards the Base Station. Here, the Packet Processor is tasked with queuing the packets that are sent from this Controller alongside other packets coming from other Controllers further downstream. Each Packet Processor keeps a buffer of incomplete packets, and only forwards a packet once it is received successfully. Since each packet includes a data checksum, corrupted packets are rejected before they reach the Base Station. The queuing process adds a small amount of latency, typically 20ms per Controller in one direction, but that is not much more than the time taken to receive and then retransmit each packet. It is a cooperative system, so any Controller that does not promptly forward the packets that it receives will lock up the bus. Of course, such problems are possible with other systems. For example, an I2C device can lock up its entire bus by simply holding either of its connected lines (SDA or SCL) low. On the other hand, the system is simple and expandable. Controllers can forward packets even if they don’t understand them. Just about any microcontroller with one UART peripheral can be used to put data Fig.11: what appears to be a single bus is actually separate devices that receive, process and then add or retransmit data. Each leg is logically separate; the system is designed so that each Controller acts as a bus repeater. Australia's electronics magazine February 2026  63 Table 2 – SLIP encoding Packet content Serial line data End of packet marker 0xC0 0xC0 0xDB 0xDC 0xDB 0xDB 0xDD All other bytes unchanged Fig.12: microcontroller IC1 takes the user inputs from switches S1-S6 and potentiometer VR1 and produces commands to send to the Base Station. It also moves data as needed between the downstream (CON2 and CON4) and upstream (CON1 and CON3) legs of the bus, and shows information via LED1 and OLED display MOD1. onto the bus, as long as it does not have any other devices further downstream of it. Circuit details Fig.12 shows the circuit of each Remote Controller. IC1 is a 20-pin PIC16F18146 microcontroller with the standard 100nF bypass capacitor and 10kW pullup on its MCLR line. During operation, 3.3V power is available at CON1-CON4; typically, it will be supplied from CON1 or CON3, since these will be facing the Base Station at the upstream end. CON5 is a header to allow in-circuit serial programming (ICSP) of IC1, with power, ground and three of IC1’s other pins connected as needed for this purpose. To simplify development, pins 18 and 19 are dedicated to programming and not used for anything else. CON1 is a four-way header and would be expected to connect to CON5 on the Base Station. CON3 is an RJ45 socket that can connect (via a Cat 5 or similar Ethernet cable) to a matching RJ45 socket (CON6) on the Base Station. Similarly, CON2 and CON4 would connect to either CON1 64 Silicon Chip or CON3 of a subsequent downstream Controller in a chain. We used the RJ45 sockets and Cat 5 cables for practically all of our prototypes. The cables must be wired ‘straight through’; pin 1 to pin 1, pin 2 to pin 2 and so on. So-called crossover cables will not work. We also built a prototype Dual Controller using two Controller PCBs fitted into a 3D-printed case. To connect these two PCBs, we soldered insulated wires directly to the pads of CON1 and CON2 where the two boards abut. Two I/O pins (TX1/RX1) connect to the communication lines on CON1/ CON3, while another pair (TX2/RX2) connects to the downstream CON2/ CON4. All of these lines are provided with 2.2kW pullup resistors to ensure that the lines are in a known state, even if nothing else is connected. The bulk of the remaining circuitry forms the user interface for the Controller. Six tactile pushbuttons connect between various I/O pins on IC1 and circuit ground. Internal pullups on these pins allow the state to be determined by the micro. Australia's electronics magazine 10kW variable resistor (potentiometer) VR1 is used primarily as an analog speed control, so it is wired as a divider across the 3.3V supply. Its wiper is connected to a 100nF capacitor via a 10kW resistor to filter out noise and provide a low source impedance for the analog-to-digital converter pin that is used to read its position. OLED module MOD1 connects to a further two pins on IC1. These provide a ‘bit-banged’ I2C interface to display a small amount of text for the user. Since there are two more I/O pins free, we have connected them across bi-colour LED1 and its series resistor. This is another indication we can provide the user. The circuit diagram shows two options for this LED; either a standard two-lead through-hole part or a four-lead reverse-mount SMD device can be fitted. The latter is wired in inverse parallel to provide the same function as the two-lead device. Firmware The firmware running on microcontroller IC1 must perform the packet processing mentioned earlier, as well as receive user input and display that on siliconchip.com.au the OLED and LED. It must also generate packets based on the user input and forward them to the packet processor. Plain serial data does not have a native packet marker. If we were interested in sending only ASCII text over the link, we could use one of the ASCII control codes (hexadecimal 0x00 to 0x1F) as a packet marker. But we want to send binary data, so we need a different technique. SLIP (serial line internet protocol, also known as RFC 1055) is a protocol from the 1980s designed to encapsulate internet protocol (IP) packets for transmission over serial connections. This encoding uses one byte code (0xC0) as an end-of-packet marker. The only place this code can appear is at the end of a packet. If this byte needs to occur inside the packet data, a two-byte sequence is used instead. Thus, a few single-­ byte values are encoded as two-byte sequences while travelling on the serial data line. Table 2 shows the encoding scheme. To this, we add a simple packet type marker byte as the first byte of each packet, and a checksum byte for error checking as the last byte of each packet as it exists in memory (and not the data on the serial line). The checksum system is the same as used for DCC; it is simply the XOR of all the other bytes. This means that if you calculate the checksum of the packet, including the checksum byte, it should result in zero (0x00). In our firmware, this means that a single function can be used to both generate and validate the checksum value. Since the checksum scheme is the same as DCC, and DCC packets must have a checksum of zero, the checksum for a packet carrying a DCC packet is the same as the packet type marker byte. For our scheme, we have chosen code point 0x42, which is the same as ASCII ‘B’. Fig.13 shows the encoding for a typical DCC packet. When the Base Station sees a packet containing a DCC packet, it sends it to the track output, if possible. The upshot of this is that it is quite simple to create a device to generate data that the Base Station can understand. You could build your own Controller variant using this protocol. There are commands that can instruct the Base Station to switch the main track power off or on, and Table 3 below lists the supported packet types. siliconchip.com.au Fig.13: the encoding of a DCC packet (with ‘B’ marker) onto the bus requires adding the marker and checksum bytes in memory and then translating the byte sequences as they are sent out on the wire. Other packet types (see Table 3) have different marker bytes, allowing the recipient to identify their purpose. Fig.14: this simple design uses the PCB as the front panel, so all components and traces are relegated to one side. A handful of components that would normally be mounted in through-hole fashion are instead treated as SMDs, using pieces of wire as necessary. The first overlay shows the Controller as built from the kit. The second overlay uses the alternative components listed in the parts list, with four-way headers for CON1/CON2 and a throughhole LED1. Australia's electronics magazine February 2026  65 The DCC Remote Controller can connect to our DCC Base Station via Cat 5/6 cables. Multiple Controllers can be added & each provides the ability to control up to three locomotives. The Controller stores the states of up to three locomotives, and they can be separately updated by rotating VR1 or operating the switches. A priority system sends out packets more frequently when data is changing, making the best use of the bus. We will provide more detail on the user interface once assembly is complete. Assembly This is an SMD design, so you will need tweezers, flux, a magnifier and so forth. None of the parts are too small, so it should not be too difficult (see Fig.14). Start by fitting the SMD parts on the black PCB, which is coded 09111245 and measures 83 × 53 × 0.8mm. This includes IC1, the two capacitors and seven resistors. Tack one lead of each, check the part is aligned and then solder the remainder of the pins. Now is a good time to fit the LED, whether you are using the surface-­ mounting version or not. If you are using the through-hole part, bend the leads over by 180° so that they reach the adjacent pads and allow the lens to shine through the hole in the PCB solder mask. For the through-hole part, check the data sheet or test its polarity to determine the lead that is the anode for the red element and cathode for the green element. This lead should be placed so that it is closest to the potentiometer. If you are using the SMD part, be sure to align the dot on the part with the silkscreen marking. Clean up any excess flux and allow the PCB to dry. There are now enough components fitted to allow IC1 to be programmed if necessary. If you have purchased a kit or IC from the Silicon Chip Online Shop, Table 3 – DCC Remote Controller Packet types before encoding Type Marker byte Notes DCC Packet ‘B’ (0x42) Sent by the Controller to the Base Station (see Fig.13). The packet content is a ‘B’ followed by the DCC data bytes, including a checksum, followed by a ‘B’ as the packet checksum. Host query ‘C’ (0x43) Sent by the Base Station, with an index that is noted by each Controller and incremented by one when the packet is sent to the next Controller. The first Controller sees 0x43, 0x01, 0x42 and sends 0x43, 0x02, 0x41 to the second. The third sees 0x43, 0x03, 0x40 and sends the fourth 0x43, 0x04, 0x47 etc. Host reply ‘D’ (0x44) When a Host query is received, the Controller replies in the format 0x44, nn, dd, cc. Here, nn is the index from the Host query, dd is an arbitrary ID byte and cc is the checksum. For the ID byte, the Controller generates a fixed but random value from its internal MUI (Microchip Unique Identifier). This allows the Base Station to know how many Controllers are connected and to calculate the bus latency by measuring how long the Host reply took from each Controller. System ‘I’ (0x49) Currently supported commands can be used to control the main track power. The sequence 0x49, 0x00, 0x49 will switch the track power off and 0x49, 0x01, 0x48 will switch it on. 66 Silicon Chip Australia's electronics magazine then there will be no need for programming. If programming is needed, solder a five-way header strip vertically to CON5 so that the pins point directly up. This header can be left in place, since it should not foul the wall of the enclosure. Connect a PICkit 5, PICkit 4, Snap or PICkit Basic programmer and use the MPLAB IPE to program and verify the 0911124C.HEX file into IC1. Solder the RJ45 connectors next, if you are using them. We found it easiest to tack the larger pads in place and then solder the smaller leads. We didn’t need to use much extra flux because the pads are much larger than most surface-mounting parts, and the solder we used has a flux core. Next, fit the OLED screen. This is done similarly to other projects where we have used a PCB as a front panel, including the USB-C Power Monitor (August & September 2025 issues; siliconchip.au/Series/445), which used the same screen module. Take four lengths of wire around 15mm long and make a right-angle bend about 3mm from one end. Solder the L-shaped wires to the pads of MOD1 on the PCB so that the long legs are pointing upwards. Remove the protective film from the OLED module and thread it over the wires. Push it down flat against the PCB and gently adjust it so that it is square within the marked silkscreen outline, then solder each wire to its pad and trim the excess. The longer pads towards the other end of the OLED can be used to affix a piece of wire to physically secure the screen module better. You can see how we have done this in the photos of our prototype. Next, solder the tactile switches. siliconchip.com.au We fitted two Controllers to a 3D printed case, making for a compact unit that can directly control two locomotives simultaneously. See the photo below for how we wired the two Controllers together. Tack one lead of each and carefully adjust them so the actuator is centred on the hole through the PCB. The 0.8mm-thick PCB allows the actuator to protrude slightly, so make sure that the tops are even, too. When they are all aligned neatly, solder the remaining leads. Like the RJ45 connectors, we didn’t need extra flux to do this. Next, thread potentiometer VR1’s shaft through the PCB and secure it from the outside with the washer and nut. It should line up squarely with the silkscreen outline. Solder short lengths of wire between the pads on the PCB and the leads of VR1. At this stage, you should be able to connect CON3 to a Base Station using a Cat 5/6 cable. Power on the Base Station and you should see a display like Screen 1. This is a good indication that everything is working as expected. CON3 (or CON1) should always connect to the cable going to the Base Station, with CON4 (or CON2) used to connect more Remote Controller(s) if required. This ensures the serial data travels in the correct direction. In use Controller protocol. You can connect to a computer via the USB socket on the Pico 2 on the Base Station, and use a serial terminal program to view status information about the connected Controllers. To allow one potentiometer to control multiple decoders, we need a way of switching control without immediately relying on the position of the potentiometer, at least until we can be sure that it does not conflict with the last setting made for that decoder. This is the <SP indicator visible on Screen 1. It is an instruction to rotate the pot anti-clockwise until it matches the last speed setting used. At startup, that position corresponds to zero speed. You might also see SP>, indicating that the pot should be rotated clockwise until it matches the speed setting. You’ll see this indication pretty much every time you change screens. Screen 2 shows the controls for the first (1:) slot on the Controller. No locomotive address is selected, so --- is shown on the first line. The HOLD > is an instruction that helps the user to set an address. Fig.15: two simple symmetrical slots for the RJ45 sockets are all that are needed to fit the Controller to its UB5 case. One edge of each socket is right on the centreline of the box. All dimensions shown are in millimetres. There is no need to update the Base Station firmware, since our original release incorporated support for the Screen 1: the initial screen. The “<SP” means to turn the pot anti-clockwise. Screen 2: with the pot rotated, it now shows HOLD > to help set an address. siliconchip.com.au Inside the-3D printed case, we soldered wires directly between CON1 on one board and CON2 on the next. Note that the wires need to cross over. Australia's electronics magazine February 2026  67 The left-hand end of the Controller is the upstream end and connects via the Cat 5 cable to the Base Station. Further Controllers are added in similar fashion. The second line shows the direction and speed, which can be toggled with the REV and FOR buttons (S1 and S2). These can also be used to set the speed to zero without adjusting the potentiometer. The three dots (...) show the function outputs, which are all off at startup. The 1 at lower right is the host index received from a Base Station, so no two Controllers should show the same value. If this is showing --- for more than five seconds, the Base Station might not be communicating due to a problem with the connectors or wiring. Hold the SEL > (S3) button for a second and release it; this will take you to Screen 3 to select an address. You can set a short address by pressing F0 and adjusting the potentiometer until the address is shown. The address shown at upper right will be activated when SEL > is pressed again. Pressing F1 will allow the top two digits of a long address to be set, and F2 sets the lower two digits. All long addresses are shown with five digits on all screens. You can also use the REV button to cancel address selection. Screen 4 shows a short address set for slot 2, with the <SP indicating that the potentiometer needs to be adjusted. The LED will be off; it will switch on when the position is correct. Generally, red means stopped and green means a speed greater than zero. You can see that the lighting function control F0 (shown as L) is on, as is F1, while F2 is off. Pressing the FOR or REV buttons will toggle direction and speed control. You can perform low-speed shunting by leaving the potentiometer set and toggling the direction and speed with just the FOR and REV buttons. 68 Silicon Chip A display of --- means that the speed is toggled off (set to zero); pressing FOR or REV will change direction or activate the speed set by the potentiometer. If a number is shown, that speed is being actively sent to the selected address. Pressing SEL again will show slot 3, as in Screen 5. Here, a locomotive with short address 19 is operating normally in the forward direction at speed step 22 with its F0 (headlight) output activated. The LED will be lit green. One more press of SEL will bring up the system page (Screen 6). Pressing FOR or REV will send a signal back to the Base Station to switch the main track power on or off, the LED will light up green or red to show the command that has been sent, and a message should appear on the display. Pressing F0 will save the currently selected locomotive addresses, so they will be available after a power cycle. Finally, the operation of the F1 and F2 buttons can be switched between toggle and momentary action (for control of function outputs) by pressing the respective button. The next press of the SEL button will return to slot 1, as in Screen 1 or Screen 2. Summary Since the Base Station is simply passing packets from the Controllers Parts List – DCC Remote Controller 1 double-sided black PCB coded 09111245, 83 × 53mm (0.8mm thick) 1 UB5 Jiffy box [Altronics H0205, Jaycar HB6015, Bud Industries CU-1941] 2 four-way right-angle 0.1in (2.54mm) pitch locking headers (CON1, CON2; optional) 2 SMD RJ45 sockets (CON3, CON4) [DigiKey 4414-3253-0007-02CT-ND] 1 five-way 0.1in pitch header strip (CON5; optional, for ICSP) 1 0.91-inch 128×32 pixel I2C OLED module (MOD1) 6 reverse-mount SMD tactile switches (S1-S6) [Adafruit 5410] 1 10kW linear 9mm horizontal potentiometer (VR1) [Jaycar RP8510] 1 knob to suit VR1 [Jaycar HK7734] 1 10cm length of lead offcuts or similar solid uninsulated wire 1 Cat 5/5E/6 ‘straight through’ cable OR wire to suit CON1/CON2/bare solder pads if using those Semiconductors 1 PIC16F18146-I/SO 8-bit microcontroller programmed with 0911124C.HEX, wide SOIC-20 (IC1) 1 red/green reverse-mount SMD LED (LED1) [Kingbright AAA3528SURKCGKC09] OR 1 3mm bicolour red/green LED (LED1) Capacitors 2 100nF M3216/1206 X7R 50V SMD ceramic capacitors Resistors (all SMD M3216/1206 size, 1% ⅛W SMD) 2 10kW 4 2.2kW 1 1kW Australia's electronics magazine siliconchip.com.au The first article in this DCC series was the Decoder shown here. The Base Station followed that. directly to the track, it’s advisable to set CV11 (packet timeout) on all locomotives. That way, if there is a communication problem, such as a Controller being inadvertently unplugged, the locomotives will stop after the timeout instead of running away out of control. A one-second timeout should be sufficient, but you can try a higher value if the locomotives appear to be stopping unexpectedly. For our Decoder, the CV11 value is measured in seconds, so a value of 1 should work for most small layouts. The Controller sends out packets for each address every 400ms, at most. If the controls are changing, then packets can be sent out as close as 100ms apart for each active address. Remember that each slot will continue to send commands for the last speed selected, whether the slot screen is visible or not. There are no interlocks against changing a slot’s address without setting the speed to zero. Again, the CV11 timeout is expected to perform the safeguard role. We measured each Controller’s current draw at between 8mA and 12mA. The switch-mode regulator on the Pico 2 can source 800mA, so the current consumption of the Controllers should not dictate how many can be connected. Each Controller adds around 20ms of latency to each packet. We ran some tests with five Controllers connected and did not think there was a noticeable delay, even from the most remote unit, although this will depend on how much bus traffic is present. This Controller design rounds out our suite of DCC equipment to include most of the things you might need to add DCC to a small layout running up to about 10 trains at a time. Having said that, it's possible we'll expand the system in future. The protocol used for communication between the Controllers and Base Station is simple but powerful, so it could be used to add more custom SC features to the DCC system. Screen 3: selecting the address of the locomotive to control. Screen 4: short address 3 has been selected in slot 2. Screen 5: the loco is going forward at speed 22 with F0 active. Screen 6: the system page lets you turn the track power on/off and more. siliconchip.com.au Screen 7: the terminal output from a Base Station with two Controllers connected. The Host Check is sent every five seconds, and the times shown are for a return trip (host query and host reply). Typical working latencies are half the figures shown for regular packets, such as commands from the controllers. Australia's electronics magazine February 2026  69 HOW TO DESIGN Printed Circuit Boards Part 3 by Tim Blythman Getting PCBs made is quite cheap these days, and as we have explained in the first two parts of this series, EDA (electronic design automation) software is powerful and easy to use. This final article in the series looks at some of the advanced options and techniques that you might use to design your own PCBs. We’ll also cover what’s required to get entire PCB assemblies made. I n the first part of this series on How to Design Printed Circuit Boards, we described the basics of setting up symbol and footprint libraries to streamline the PCB design process in Altium Designer (most other EDA packages have similar workflows). We also explained how a manufacturer takes the Gerber files and turns them into a completed PCB. In Part 2, we walked through the steps of laying out a schematic (circuit diagram) and then transferring that to the PCB editor to allow components, traces and other features to be arranged to complete the board. We offered a few tips and tricks along the way. The most recent article finished with instructions on how to use exported “Gerber” files to order from a manufacturer like PCBWay. The number of options is incredible, as you 70 Silicon Chip would have seen in Fig.20. While the defaults are suitable for a vast majority of designs, we will next delve into some of the more interesting and useful options. This article will then investigate some of the requirements for PCB designs that involve high voltages, high currents or high-speed signals. That will include how to approach these concepts at the design stage, and how some of the specialised PCB purchasing options can address concerns relating to advanced designs. As you would have seen from project articles such as the RP2350B Computer (November 2025; siliconchip. au/Article/19220) and the RGB LED Star (December 2025; siliconchip. au/­Article/19372), it is now possible (easy, even) to design and order complete, custom PCB assemblies (PCBAs). Australia's electronics magazine A PCBA is simply a PCB that has been fully or partially populated with components. In cases like the RP2350B Computer, that means that you could receive a practically completed project; perhaps needing little more than a case. So this article will also discuss what is needed to design and order a PCBA. More PCB options Some of these options are fairly obvious, while others are a bit obscure, and their cost can vary markedly. Fortunately, manufacturers like PCBWay automatically update their pricing based on selected board options, so you can easily see what specific combinations of options might cost. In Fig.20, you can also see the small “?” icons that provide further detail on how some of these options work. We’ll discuss some of the more interesting options below. The board type option allows the PCB to be manufactured in larger panels consisting of more than one board. There is little advantage in this if you are ordering just a few PCBs. The larger panels are easier to handle if there are further automated processing steps that need to happen, such as being fitted with components to create a PCBA. These panels make for easier processing in the pick-andplace machine and reflow oven. When you upload Gerber files at the start of the ordering process, the size will be automatically detected, but you can manually enter a figure to see how the cost changes for different sizes. Keep in mind that the size really refers to a rectangle that contains the entire PCB shape, so an unusual shape might benefit from being rotated to minimise its dimensions. A good example of how this works is the RGB LED Star. Hanging in its obvious orientation (with the long arms vertical and horizontal), a PCB manufacturer would measure it as 240mm × 240mm. By rotating the design by 45° within Altium Designer, this is reduced to 170mm × 170mm, which ends up being much cheaper to manufacture. You can see this in Fig.21. For a prototype, you might only need a single board, but five is the usual minimum order quantity (MOQ). It just isn’t worthwhile for the manufacturer to make fewer than that. Five small double-sided boards can siliconchip.com.au be surprisingly cheap to order (a few dollars plus postage). Advanced options We are now getting into some of the more advanced (which can mean expensive) options. Multi-layer boards (with more than two layers) have certainly become cheaper, and will often be necessary for high-speed designs. Four-layer boards are commonplace. Manufacturers no longer offer discounts for single-sided designs except in huge quantities; two layers is generally the minimum practical number. The material option refers to the substrate; FR-4 glass-epoxy laminate (fibreglass) is widely used and well characterised, so it is easily the cheapest. Aluminium-cored PCBs are not too expensive for single-layer designs, and would be chosen for their improved thermal conductivity over FR-4 in high-power designs like LED lamps. However, they can be difficult to solder by hand; a reflow process is generally required. PCBWay offers flexible PCBs (www. pcbway.com/flexible.aspx), these are reasonably priced for small designs; we used a flexible PCB as a slim interboard connector in the USB-C Power Monitor (August & September 2025, siliconchip.au/Series/445). So they are worth considering where you need a board or cable that can bend. Some simulation features in Altium Designer can depend on the dielectric characteristics of the substrate, so if you are planning to use a different substrate, be sure to update the Layer Stackup to suit. The impedance of differential pairs also depends on the substrate characteristics, so you will need to check this if you are routing high-speed differential pairs. A typical PCB is 1.6mm thick, but for a two-layer board, you can reduce the thickness to 1.2mm, 1.0mm or 0.8mm without increasing the cost. Thinner boards are available but are more expensive and less robust. For panels, a slimmer PCB will often be more elegant. Some components, like the USB plugs that we used in the USB-C Power Monitor, require a specific board thickness, so you might find that your components dictate this option. The default options for minimum hole size, track width and spacing should be fine for most hand-soldered siliconchip.com.au Fig.20: PCBWay offers many options for its PCB manufacturing service; there are other tabs offering advanced options and flexible PCBs as well. To see the full range of options, visit www.pcbway.com/orderonline.aspx Fig.21: our RGB Star looks best hanging vertically, but designing it like this (or at least rotating it before fabrication) allows it to be made much more cheaply. Australia's electronics magazine February 2026  71 Fig.22: this ruler is actually a PCB that has been designed by PCBWay to show off their multi-coloured PCB printing capabilities. designs. If tighter tolerances are needed, they may be available for an extra charge, since the processes need to be more exacting. If your design uses BGA-packaged chips or other finepitch parts, you might need to check these parameters when setting up your design rules. In our experience, the different solder mask and silkscreen colours do not add extra cost, but anything different to white silkscreen printing on a green solder mask will take longer to produce. So we generally stick to that unless there is a good reason to use something different, such as using a black solder mask for panels so that they match the rest of the enclosure. The PCB colour can also be chosen for aesthetic purposes, such as the red PCB used in the FlexiDice (November 2024; siliconchip.au/Article/17022). Note that while black and white PCBs look nice in certain applications, it can be hard to see the tracks under those solder mask colours, which may make debugging harder. Multi-colour printing Multi-colour printing on PCBs has recently become available; the printing applies to the silkscreen layer. This process uses UV-reactive inks that are similar to those used for traditional silkscreens and solder masks, except they are capable of reproducing a full range of colours. Fig.22 shows a sample of a PCB that has been produced using this process. The process is analogous to CMYK printing on white paper, so a white solder mask is required as the base to give the best results. PCBWay provides a guide to their process at siliconchip. au/link/ac9g Since the Gerber format has no way to handle this colour information, the process involves creating image files (JPG etc) for the top, bottom or both layers. A third image can be provided as a reference to show how the images should be aligned to the PCB. with a suitable receptacle. Various types of computer cards are probably the best-known examples; hence, they are also known as card-edge connectors. Modern PCI Express cards still use the same principle as the original IBM PC from the 1980s. Figs.23 & 24 show a typical edge connector and a matching receptacle. While they look deceptively simple, edge connectors require extra PCB processing steps for correct operation. They should have a hard gold plating to give the necessary durability to the contact surfaces. The edge should be given a bevel to ease its insertion into the connector; all these steps add extra cost. Surface finish There is also the option to choose a surface finish for the exposed copper on the board; that is, the copper that is not covered by solder mask, which mainly means component pads. We mentioned some of the options in the Part 1 panel on the PCB manufacturing process. These finishes are intended to protect the pads from corrosion until they are soldered to. For cost reasons, we practically always choose the HASL (hot air solder level) process; this coats the copper with a thin layer of solder. Interestingly, the process for flexible PCBs requires a gold finish such as ENIG, since the tin-based solder used in the HASL process does not handle flexing well. Other options include OSP, which stands for organic solderability preservative, a coating that is dissolved during the soldering process. ENEPIG adds a durable palladium An edge connector is made of traces on the PCB that end in fingers that mate Fig.23: this PCIe receptacle is typical of the type that allows an edge connector to plug in. Source: Mouser 571-5-1734857-5 72 Australia's electronics magazine Edge connectors Silicon Chip layer between the nickel and gold of the ENIG process. The silver and tin immersion finishes use a chemical (non-electrolytic) plating process to add thin layers of their respective metals to the copper for protection. These are not as resistant to oxidation as HASL, but this is not a concern where the boards are populated soon after manufacture, such as when you’re using a PCBA service. Plugged vias are more expensive than plain vias. In this case, the empty space of the via hole is filled with resin to provide a flat surface at each end. This is only necessary in cases such as where there is a via in a pad and the PCB is assembled with a reflow process, although it also reduces the chance of via corrosion later, especially for larger vias that can’t be tented. You can designate uncovered (untented) vias by having openings in the silkscreen, but the best practice is usually to leave them covered, since that will leave them less exposed to oxidation or inadvertent contact. Production code You might have seen that our PCBs have a code printed on their silkscreen layer that does not match the eightdigit PCB code that is printed elsewhere. This is a tracking code used by the PCB manufacturer during the production and is selected at the “Remove product No.” option. This is needed because many PCB orders are combined into a much larger panel during production. When Fig.24: an edge connector has goldplated fingers to mate with the connector shown in Fig.23. Source: PCBWay – siliconchip.au/link/ac9j siliconchip.com.au the panel is separated, the individual PCBs need to be identified and sorted. Removing the tracking code entirely will cost extra, since the PCBs need to be identified another way. It’s also possible to add specific text (eg, “WayWayWay” for PCBWay) to one of the silkscreen layers to mark a desired location for the code. This means that the marker text will be replaced by the tracking code in the finished PCB. The above covers many options, many more than we have ever used. For the curious, there is also an advanced tab, with even more options! High-current designs Last month, we noted that many simple designs can be completed without worrying about requirements related to high currents, high voltages, high-speed signals or RF. The main option in PCB manufacture that relates to high-current design is copper thickness. The standard copper thickness on FR-4 PCBs is one ounce per square foot, which you will see quoted as “1oz copper”. Based on the density of copper metal, this is nominally 0.035mm (35 microns or 35μm) thick. You might choose thicker copper to reduce resistance in a high-­current or high-power design; the aim is to reduce dissipation through ohmic (resistive) losses in the traces. We have used 2oz (70μm) copper in a handful of high-current designs, most recently the Ideal Diode Bridge Rectifiers (December 2023; siliconchip.au/ Article/16043). Much heavier copper layers are possible; Fig.25 shows an example of a PCB with 20oz (0.7mm-thick) traces! The thickness is made by plating extra copper onto the existing copper, which means that extra copper must also be etched away in places. The PCB Assembly Pitfalls While it’s certainly tempting to get someone else to assemble boards for you, the process is not without its hazards. Two problems we’ve experienced so far are: #1 Defective parts: prototypes of the Pico 2 Computer (April 2025 issue; siliconchip.au/Article/17939) worked fine. The ‘production’ batch of boards unfortunately didn’t due to a different batch of CH334F USB hub ICs being used, which were faulty. Luckily we just needed to remove two resistors from the board, bypassing the faulty function, allowing the boards to work. But the chips could easily have had a flaw that wasn’t fixable without replacing them, and they’re QFN chips – not easy to replace! #2 Incorrect assembly: we quadruple-checked the orientation of the small yellow SMA tantalum capacitor shown in the photo below before ordering the boards. On receiving them, when power was applied, too much current was drawn. We realised that the tantalum capacitors had been installed backwards. The right-hand photo shows the preview on the JLCPCB website. When we queried it, they told us that the preview is not 100% accurate and that we need to request to be sent images to check before manufacturing starts. Again, this was fixable, but time-consuming. Still, we think they should have alerted us that the manufacturing plan differed from the preview. The yellow/orange SMA tantalum 22μF capacitor shown in the left-hand photo was installed backwards compared to the adjacent preview image. deeper etching requires tighter controls to achieve the same outcome as 1oz copper. The etched copper also adds to the amount of dissolved copper that must be handled as a waste stream of the process. For these reasons, it’s often cheaper and quicker to design wider traces with 1oz copper in mind. Also Fig.25: this board for a Formula E electric race car costs over $2000. It has extremely thick tracks for high current handling and spacing for voltage separation. Source: PCBWay siliconchip.au/ link/ac9k siliconchip.com.au Australia's electronics magazine consider that at higher frequencies, the skin effect makes thicker traces less effective. The copper layers can also be enhanced with manual post-processing. For the Versatile Battery Checker (May 2025; siliconchip.au/Article/18121), we removed the solder mask above some of the high current traces, allowing them to be supplemented by adding solder during the construction phase. This is a trick that many manufacturers use as it’s cheap if done sparingly. Design rules review Now we will look more closely at some factors that might complicate designs involving high currents, high voltages, high-speed signals or RF. It’s a good idea to have experience with these sorts of concepts before attempting to design PCBs with them. February 2026  73 In these cases, there are design rules that can be applied to ensure that the necessary requirements are met. The design rules won’t guarantee perfect results, especially when the PCB exists in a real world with unpredictable external conditions, but they will help. For high-current designs, the trace width is typically the most critical parameter. Copper has a finite resistivity, typically given as 1.7×10-8Wm at room temperature. The units of Wm mean that you can get a resistance (in ohms) by multiplying by the length and dividing by the cross-sectional area. On a 1oz PCB, this means that a trace 1m long and 1mm wide has a resistance of around 0.5W. That on its own does not tell you how wide a trace should be, so the IPC-2221 standard has been developed to formalise good practice. Altium Designer has a built-in resistance calculation tool in its PCB editor as well as an online guide and calculator for this aspect of IPC-2221 at siliconchip. au/link/ac9h These calculations are based on the expected rise above ambient temperature due to ohmic heating, and are simplified with a number of assumptions; for example, the ability for internal layers (on a multi-layer PCB) to shed heat is much reduced compared to external layers. A good working figure is a 10°C rise, and even then, IPC-2221 is considered quite conservative, since it does not take into account other nearby traces and copper areas. IPC-2152 is another standard that considers even more factors. Thus, it’s a good idea to set up a design rule that ensures that all the traces are wide enough for the current they will carry. Since you don’t need all traces to be subject to the same width rules, Altium Designer also includes the concept of net classes to selectively apply different design rules. We can also use net classes in high-voltage, high-speed and RF design. Net classes While it is possible to create a net class in the PCB Editor, it’s best to do so from within the Schematic Editor. Here, the nets correspond to wire objects, so we simply need a way of marking each wire object with its desired net classes. This is done by placing a Parameter Set object (Place → Directives → Parameter Set). The Parameter Set object can be used to set much more than just net class. It is attached to the wire and needs to have a net class added. The net class name is set with a string (such as “POWER”), and its label can be set so that its purpose on the schematic is clear. The Parameter Set object can now be copied and pasted as needed to add other wires to the same net class. Fig.26 shows a design with several POWER net class objects. The net classes are carried through with the nets into the PCB design (when Update PCB Document is performed); thus, the traces for those nets will also belong to the net class. The next part of using net classes is to create custom rules that apply to them, such as a minimum trace width rule for current handling. Fig.27 shows the updated design rules in this case. Fig.26 (below): adding a Parameter Set object allows wires (and thus the resulting nets and also the traces in the final PCB) to be assigned to a net class to allow specific design rules to be applied. Fig.27 (left): this custom rule applies to members of the net class and enforces a minimum width. Fig.28 (lower left): during routing, a trace is flagged if it does not meet the width specification for its net class. 74 Silicon Chip Australia's electronics magazine siliconchip.com.au Fig.29: the Layer Stackup Manager is used to enter the properties of the PCB stackup, such as layer thicknesses and dielectric properties. Among other things, this allows an impedance profile to be created. Fig.30: the impedance profile is used as the basis of a design rule to enforce the trace width and spacing to maintain the impedance of the differential pair. A second routing width rule (that we have called Width_POWER) has been added. It is applied to the POWER net class by using the dropdown menus to select the correct object matching criteria. Its priority has been set to overrule the default rule (when it is applicable) and the minimum width increased to an appropriate value. Fig.28 shows the result of this rule being applied to a trace in that net class. When the trace is reduced below the minimum width, it is flagged as a design rule violation. Another, thinner trace is not flagged, since it is not a member of the POWER net class. High voltages The most obvious design rule for high-voltage design is clearance, which is the spacing between traces on the same layer. Altium Designer can also apply a design rule for creepage, which tests the distance between traces along the board surface and can take account paths through holes, cutouts and even around the edge of the board. The way to enforce clearance for high-voltage traces is to use the Parameter Set method to create a high-voltage net class and then create an appropriate design rule invoking that net class. Since clearance and creepage rules involve two traces, there are two dropdown menu options to be selected. One should be the relevant net class, while the other should be “All” to ensure that clearance and creepage are maintained to all other copper. It’s possible to set a net to be part of multiple classes if needed. siliconchip.com.au Creepage is also affected by the substrate thickness, so the Layer Stackup becomes important, since it will dictate the board thickness and thus the length of the creepage path. The PCB thickness is used in the IPC-2152 PCB trace width calculations. It is critical in high-speed design, especially since dielectric characteristics will affect signal propagation. High-speed signals High-speed and RF PCB design is a very broad topic. There isn’t necessarily a fixed point at which a PCB becomes high-speed; it is related to when the traces behave more like transmission lines than simple wires, so concepts like trace impedance become important. It’s imperative to use the Layer Stack Manager (under the Design menu in the PCB Editor) to make sure the settings match the intended PCB manufacturing process and materials if high-speed signals are involved. Fig.29 shows the Layer Stack Manager with the Impedance tab opened. With an impedance profile set, it becomes available as a design rule, and can be applied to traces in the same fashion we have discussed for other net classes. Single conductor Australia's electronics magazine and differential pair (Fig.30) impedance profiles can be set. Altium Designer can also provide calculations and simulations, so it’s possible to check and validate a design after it has been routed and before it is manufactured. PCB design is an iterative process, so don’t be surprised if you need to go back at some point and rework your layout. One important factor in high-speed design is that if you have multiple related signals (eg, a parallel memory bus or a differential pair), the track lengths should be as close to identical as possible so the signals arrive at the same time. Altium and other ECAD packages provide tools to help ensure this is the case. Minimising magnetic loops (eg, through the use of a ground plane) is also important, as is considering the effect of crosstalk between adjacent or nearby high-speed conductors. PCB assembly Some PCB manufacturers now offer PCB assembly (PCBA) services. This involves having the PCB made, then populated with components. We have done this now for a handful of projects where it would be difficult to hand-­ solder the necessary components, such as the QFN-80 package RP2350B chip. February 2026  75 Since JLCPCB was quick to offer the RP2350B chips, we used their PCBA service for two RP2350B-based projects. We also used them for the RGB LED Star, since we were familiar with their requirements and process. Fig.31 shows the Star assembly that we received from JLCPCB. Different PCBA manufacturers offer different ranges and sources of components. So we suggest picking a company before performing schematic capture, as you will need to know what components and variants are available in sufficient quantities before commencing layout of your design. JLC’s low-cost service is well-suited to simple designs, while PCBWay offers considerably more flexibility, so they are generally recommended for assembling more advanced designs. For example, JLC doesn’t offer blind or buried vias, which are required for many PCB designs that include BGA (ball grid array) package parts. Overview The process we’ll describe for designing and ordering PCBAs applies to JLCPCB’s service. It should be fairly similar for other manufacturers like PCBWay, but we recommend checking their specific requirements before starting a design. In addition to the Gerbers needed for making the PCB, you’ll need a bill of materials (BOM) and a component placement list (CPL) files. The latter might also be known as a ‘pick-andplace’ file; it is mainly a list of the components and their locations and orientations on the board. Both of these are simply spreadsheet files in Microsoft Excel (XLSX) format. Other spreadsheet formats, such as comma separated value (CSV), are also supported, so you can view and edit them using free software such as LibreOffice (which also supports the XLS/XLSX file formats). Altium Designer can export these files, but there is specific information that needs to be entered to ensure that the correct data is available. This includes things like component part numbers and suppliers, which will be specific to a PCBA manufacturer. The PCB ordering process happens as usual and is followed by an option to enable PCB assembly. This step will require the BOM and CPL files to be uploaded. Then there are selections related to the assembly process that will need to be made. Let’s start by looking at what needs to happen in Altium Designer. In Part 1, we provided a panel detailing how PCBWay takes the Gerber files and turns them into a PCB. The panel opposite describes how the BOM and CPL files are used to assemble the PCB and components into a PCBA. Schematic capture changes During the schematic capture, each component needs to have information added to indicate its supplier and part number. There are added as Parameters in the component properties, as seen in Fig.32. The Supplier and Supplier Part fields are required, but we have added the other fields for completeness. LCSC (www.lcsc.com) is a sister company of JLCPCB, and the part numbers are the same as JLCPCB’s (https://jlcpcb.com/ parts). It’s possible to source parts from other distributors, although we have not needed to do this. These parameters will be carried over if the parts are copied and pasted during schematic capture. Where possible, use the Basic parts type. Extended parts are more expensive to use, since they will need to be manually loaded into the pick-and-place machines before they can be installed on the PCB. You can filter by type in JLCPCB’s parts search. For example, this means that it’s considerably cheaper to use M2012/0805size passives or smaller, as they are Basic parts, while M3216/1206-size parts are mostly Extended. Remember, you don’t need to solder these parts – they will be doing it for you! Of course, you want to make sure that the parts have ample stock; we would expect that the Basic parts would be maintained in stock, since they are always loaded in the pick and place machines. (JLCPCB lets you preorder parts to ensure they’re in stock when you’re ready for assembly, but we won’t explain that process here.) Broadly speaking, the design will be cheaper to manufacture if you can minimise the number of different part numbers that are used, since there will be fewer parts that need to be loaded into the pick and place machines, and you will get better quantity discounts. It will also be less work to source substitutes if needed. This is just a small part of the larger field known as design for manufacture (DFM). Fig.31: RGB LED Stars are received by us attached to PCB rails that have fiducial (locating) marks to assist their processing during assembly. Fig.32: adding these parameters to each component during schematic capture ensures they are linked to the correct inventory part for the assembly stage. 76 Australia's electronics magazine siliconchip.com.au The PCB assembly process We explained in Part 1 of this series (December 2025 issue; siliconchip.au/ Article/19373) how Gerber files are turned into a PCB. Now, we will look at the processes involved with populating that PCB with parts as might be done by a typical PCBA provider. For boards with just surface-mounted parts, there are four main steps. First, the boards have solder paste applied to the pads where needed. Then the components are placed onto the PCB by pick-and-place machines. The components are soldered by passing the board through a reflow oven, after which a final inspection occurs. Through-hole parts are often still manually fitted and soldered, although some can be placed by machine, with the board being soldered by a wave-­ soldering process that rides the board over a bath of molten solder. We’ll focus on the surface-mounting process, since we expect most readers will be interested in that aspect. You can see from Fig.31 in the main article that the PCBs for our RGB LED Star are fitted with rails along the edges. These rails have markers so that the various processes work to the same alignment. The rails also make it easier for the boards to be transported through and along the steps in the process. Solder paste To apply the solder, a laser-cut stainless steel stencil is produced. The thickness of the steel, combined with the size of the holes, determines how much solder is applied. It is applied with a squeegee that forces the solder down onto the PCB through the holes in the board. PCBWay uses an automated camera-­ based inspection process to verify the process. Differently coloured lights are shone from different angles to allow the height and location of the applied solder paste to be checked. The YouTube video at https://youtu. be/24ehoo6RX8w shows a tour of PCBWay’s assembly factory in Shenzhen, Fig.c: the boards enter the reflow oven for soldering. Source: https://youtu. be/24ehoo6RX8w siliconchip.com.au Fig.a: solder paste application using an automated stencilling machine. Source: https://youtu.be/24ehoo6RX8w Fig.b: components are picked up from the reel at the front & placed on the PCB. Source: https://youtu.be/24ehoo6RX8w China by Scotty of Strange Parts. Fig.a is a still from this video and shows the automated stencil applying the solder paste to a board. from April and May 2020 implements this same process (siliconchip.au/ Series/343). Since the factory is more like an assembly line, the reflow oven is a long machine, with the temperature profile being achieved by different temperature zones along the machine’s length. Fig.c shows the boards entering the reflow oven. The solder paste is a suspension of small balls of solder in flux paste, so when the appropriate temperature is reached, the solder melts and the flux is activated, soldering the component to its pads. Pick-and-place The BOM file is used to determine which parts are loaded into the pick-andplace machines. Fig.b shows one of the machines in operation. In the processing line shown in the video, the board actually passes through three pick-and-place machines in succession. The machines take components one at a time from a reel using a small vacuum head. They then place them on the board, where they loosely adhere to the solder paste. The machines in PCBWay’s factory are also fitted with cameras. One camera is used to register the markers to know where the board is. Another camera observes each component after it is picked up, and the computer can determine how much the part needs to be moved or rotated to get it in the correct position. There is another inspection stage after this; an operator can move any components that are not where they should be before the next stage. Reflow The reflow soldering process demands an exacting temperature profile to achieve optimal results. The temperature is slowly ramped up to the target and is then held for a time before being allowed to decrease. Our Reflow Oven Controller Fig.d: automated inspection uses coloured lights to highlight defects. Source: https:// youtu.be/24ehoo6RX8w Australia's electronics magazine Inspection The completed board is inspected with a similar camera to that used for the solder paste. Fig.d shows a view from the computer that processes the inspection. Differently coloured lights are projected at different angles and strike the components and solder fillets in distinctive patterns. The patterns are compared to a board that has been manually inspected and validated. If necessary, components are marked for rework, which is done manually. BGA (ball grid array) chips don’t have any visible pins, since they are all under the body of the part. These can be inspected by an X-ray machine. Summary These are just the main steps involved in PCB assembly. Double-sided boards can be made with these processes, but usually require the components on one side to be secured with glue, so that the board can be inverted to process the other side. There are optional post-processing steps that can be done, such as programming, functional testing and conformal coating. But it’s incredible to think that it’s now possible to design your own project and have it be fully assembled and delivered to your door at a price that hobbyists can afford! February 2026  77 PCB export Fig.33: after the CPL file is exported here, it may need some editing to ensure that it conforms to the format expected by JLCPCB. The new parameters are carried over to the PCB layout stage, and can be viewed there, but there isn’t anything else that needs to be done during layout until the design is finalised and exported for manufacture. After exporting the Gerber files in the usual fashion, use File → Assembly Outputs → Generate pick and place files. Fig.33 shows this screen. Ensure Metric units and Show Units are selected and export to CSV format. Find the Project Outputs subdirectory where your project is saved. You will see a CSV file that you can open in LibreOffice Calc, Microsoft Excel or similar. The first 12 or so rows are not useful to us, so delete them, moving the column headings up to the first row. Next, we need to change some column names as they are not what JLCPCB is expecting. Change the “Center-­X(mm)” heading to “Mid X” and the “Center-Y(mm)” heading to “Mid Y”, then save it as an XLS file. This will be your CPL (component placement list) file. To generate the BOM, click Reports → Bill of Materials. On the right side of the dialog that appears, under Properties, click Columns and then make sure your parameter columns are visible (click the grey eyes to turn them white). Go back to the General tab and under File Format, select “Generic XLS”, then click the Export button at lower right. Manufacturing Figs.34 & 35: the RP2350B Development Board uses tiny SMD passives and a QFN chip. It would be quite difficult to hand-solder, so it’s handy to be able to get this board fully assembled. It is a simple design with components on one side. Thus, it qualifies for the Economic manufacture option. Let’s work through the ordering process using the files for the RP2350B Development Board. The board is shown in Fig.34. You can follow along by downloading the required files from siliconchip.au/Shop/10/2832 Start by uploading the Gerber file (with the ZIP extension) as you would for any other PCB design. Validate that the Gerber is correct and make any selections as necessary for the PCB. Scroll down the page and turn on the switch for PCB Assembly, which will pop out some related options, which you can see in Fig.35. Economic PCB assembly is possible for this board, since it is an uncomplicated design with components on just one side, and the remaining options can be left as their defaults (you may want to select the Board Cleaning option to remove residue as it costs Australia's electronics magazine siliconchip.com.au 78 Silicon Chip Fig.36: on this page, you can opt to leave components off or select substitutes if your preferred part is unavailable. Fig.37: the Component Placements page allows the position and orientation of the components to be checked & adjusted if needed. Note the purple dot indicating pin 1 on the polarised components (but you can’t always rely on this). little). There is a comprehensive list of the different assembly types at siliconchip.au/link/ac9i Interestingly, we had to use the Bake Components option for the RGB LED Stars, since the WS2812B RGB LEDs are highly susceptible to absorbing moisture. This can lead to the evocatively named ‘popcorning failure’ when the parts are heated during reflow soldering. We also had to select the Standard PCBA type for the RGB LED Stars, since these PCBs have components on both sides. Click Next to proceed; the next page is simply a PCB viewer, so you can click Next again if the PCB looks correct. This page allows you to upload the BOM and CPL files, after which you should click Process BOM & CPL, which leads to the screen seen in Fig.36. This page allows you to check and confirm that the listed components are able to be matched. If they are not, you can use the search button to find an alternative. Any parts that do not have a blue tick in the Select column will not be fitted, so you can use this page to deselect any parts you don’t want fitted. The Lib Type column shows that the Basic parts are mostly passive components with common values. After clicking Next, you might see a warning about using a non-standard power supply configuration for the RP2350 IC; this is fine to click through, since this is a proven design. Our RP2350B Development Board article explains the configuration. 🔍 siliconchip.com.au Fig.37 shows a simulated view of the board with all the components in place. Here, you can check and edit the orientations and locations of the components. You can see that polarised components have a purple dot marking pin 1. You can match this to the pin 1 silkscreen marker to confirm the orientation. If anything is wrong, there are buttons to move and rotate the parts. You can also click on the image or list to select and highlight certain parts before editing them. If there are problems, it is a good idea to go back to your design and edit the components to ensure that future designs do not have such problems. Click Next when you have checked all the components on this page. Fig.38 shows the final breakdown of the costs for board manufacture and assembly (in USD). There is an item for a stencil, but it’s interesting to note that you do not need to provide paste mask files (for the stencils). The paste masks and stencils are generated by JLCPCB. The components are the largest cost, but the fee for using extended components does make up nearly 1/3 of the total. To complete the order, select Save To Cart and complete the order as you would for any other online shop. As you would have seen from the RGB LED Star, there is no requirement that all parts be fitted. In the same vein, it’s not necessary to have all boards assembled either. You could order five boards and only have two boards assembled (the minimum number), which would save on parts and Australia's electronics magazine assembly costs if the design is only at the prototype stage. Summary PCB (and PCBA) design is a broad field, and we cannot hope to cover all the factors that influence the journey from concept to completed project. We hope that the information we have provided in this series is helpful in producing your design. If in doubt, simply try making your own PCBs if you have not done so already! The Altium Academy YouTube channel has numerous tutorials on PCB design using Altium Designer (www. youtube.com/<at>AltiumAcademy). SC Fig.38: the final cost breakdown shows how much of the total is due to the use of Extended components. So it’s a good idea to use Basic parts if possible. February 2026  79 Using Electronic Modules with Tim Blythman Actual Size Tiny QR Code Reader Combining a camera with a microcontroller opens up many possibilities, but typically adds the requirement to process vast volumes of data. The Tiny Code Reader is a fairly inexpensive module that includes a camera and can decode QR codes, making it quite useful indeed. T his tiny module is available from Mouser and DigiKey for around $15 and we thought that it would be worth trying out; that’s a good price for a module that can read QR codes. If you want to learn more about QR codes, see our panel overleaf. The Tiny Code Reader has a straightforward interface, with example software for numerous languages and processors. We didn’t see any PicoMite code, so we’ve written a BASIC program that allows the PicoMite to interact with the Reader. The Reader is produced by a firm called Useful Sensors, based in the USA. They specialise in AI-powered technology; some of their other products include speech-to-text and translation features. It is very small, measuring about 16 × 19mm and about 8mm thick overall. The lead photos show the front (featuring the camera lens) and rear. Pin headers are not supplied, so we fitted those ourselves. The hardware appears to be similar (electrically) to a Raspberry Pi Pico module. It is based on an RP2040 processor, and you can see the flash memory chip and oscillator on the small PCB. It appears to be a closedsource design, and we did not find any circuit diagrams or the like at www.­ usefulsensors.com The camera module is glued in place and attaches via a slim mezzanine connector. That and an RGB LED are about the only parts that would not be found on an RP2040 microcontroller board such as the Pico. The RGB LED is on the same side as the camera lens. The main external interface is a fiveway 0.1in/2.54mm pitch header that breaks out an I2C interface along with power. We used the pin headers during our testing but there is also a four-way 1mm-pitch JST connector that provides a so-called ‘Qwiic’ I2C interface. The Qwiic interface was developed by SparkFun but is now used on many different development boards. There is more information available on it at www.sparkfun.com/qwiic Tiny Code Reader The Tiny Code Reader has a microcontroller that reads and decodes image data from a camera sensor. It can communicate via an I2C interface and has an RGB LED that flashes to report its status. During normal operation, the LED flashes blue, turning green when a valid QR code is detected. If it shows red, an error has occurred. The entire device operates at 3.3V, which simplifies the circuit, since no regulator is needed for the 3.3V microcontroller. ▶ Fig.1: the wiring is straightforward; the connections shown here will work with our sample code. We didn’t need to fit any external pullup resistors during our tests. The module is shown larger than life here for clarity. Fig.2: the approximate ranges at which the Tiny Code Reader could decode a 62mm-wide QR code. It has much the same vertical range as horizontal range. The user guide suggests a distance of 100mm should work, and we were able to achieve this with a smaller QR code spanning a 20° field in the camera’s vision. 80 Silicon Chip Australia's electronics magazine siliconchip.com.au A guide can be found at https://github. com/usefulsensors/tiny_code_reader_ docs There are links to numerous code examples on this page; we will also discuss our code (Arduino and Pico­ Mite) shortly. There is also a data sheet, found at https://usfl.ink/tcr_ds This indicates that the maximum operating current is 40mA. Our unit ran very close to 37mA whether the LED was on or off. The wiring connections we used during our tests are shown in Fig.1. Note that the INT pin has no function on this module. Since the module flashes its LED green when it detects a QR code, we found it easy to check its operation. Once we had our software loaded, everything worked as expected, printing the detected codes on a serial monitor program. Later, we will study its range and field of view. We found it was a bit tricky to aim the device since there is no viewfinder. It would have been more useful to have the LED on the opposite side of the board so it is visible when you are facing the QR code. Interface The interface is quite simple. It uses 7-bit address 12 (0xC) and will respond to all reads with a simple data structure up to 256 bytes long. The first two bytes report the length of the detected QR code (zero if not detected) and the remaining bytes are the contents of the code. There is also a write command that can be used to disable or enable the status LED. As we mention in the panel on QR codes, they can encode data much longer than 256 bytes. The Tiny Code Reader is only recommended to work with codes up to 40 bytes, although we were able to successfully read a 177-byte code. The Tiny Code Reader performs a scan every 200ms, since that is how long it takes to process each image. It can work with a 400kHz I2C bus, but even with a 100kHz I2C bus, reading 256 bytes will only take around 23ms, so the reader is not limited by the bus speed. Some QR codes can encode non-­ ASCII data, such as numeric data or Japanese kanji symbols; it appears that these encodings are not supported by the Tiny Code Reader. As you can see, the interface is quite simple, so we recommend that you have a look siliconchip.com.au Start The Tiny QR Code Reader should be at address 0x0C Found 7 bit address: 12 (0xC) 8 bit write address: 24 (0x18) 8 bit read address: 25 (0x19) Done. Found 1 device(s). T=2349 Code detected: test Code detected: test Screen 1: the output from our Code detected: Arduino test sketch includes an I2C test scan to confirm that communication No code detected. with the Reader is working. No code detected. at some of the code examples if you want to learn more. Code examples We have created code examples for the Pico microcontroller in both the Arduino and PicoMite BASIC languages. There are compiled (UF2) files for directly programming the Pico; these work with the wiring shown in Fig.1. Screen 1 shows the output of the serial port from when the Arduino program starts and runs. Initially, it performs an I2C device scan to allow you to check that the Tiny Code Reader is correctly wired. It then reports any codes it sees and their contents. The output is updated every two seconds unless the content changes, in which case it is updated immediately. The PicoMite BASIC program works similarly, although it shows a different style of I2C scan. Both programs allow you to switch the status LED off or on by sending 0 or 1 to the serial port. Other notes There is a nominally 2.2mm diameter mounting hole in one corner of the PCB near the headers. The Reader is quite small, but if you are able to use the Qwiic connector, it can be made even smaller by snapping off a portion of the PCB. That would include the mounting hole, so it may not suit all situations. The documentation is quite firm on the Reader only being suitable for 3.3V logic levels. We still expect it would work fine with a 5V microcontroller, as long as the power and I2C lines are limited to 3.3V, since most 5V micros will accept anything above about 3.0V as a high level. Just be sure not to apply 5V pullups to the I2C lines. We tried reading linear (1D) barcodes, but it seems that the Tiny Code Australia's electronics magazine Reader does not support any of the common linear barcodes. We plan to review a 1D/2D barcode reader module in the near future. If the Tiny Code Reader could be expanded to handle linear barcodes, we think it could be much more versatile. Given that linear barcodes are simpler, we expect they would be easier to decode. On that note, Useful Sensors points out that they do not provide support for reprogramming the firmware on the Reader. The RP2040 chip uses an external, unencrypted flash memory chip, and the Reader has about 10 test points exposed. So we think it wouldn’t be too hard for someone to extract the firmware if they really wanted to. Abilities For these tests, we printed out some short QR codes on white copy paper. We found this to give better results than the same code on a computer monitor; we suspect that the refresh rate of the monitor might be causing artefacts in what the camera sees. The codes we used were the smallest version and can hold up to 19 ASCII characters. The printed codes were 62mm wide and tall. We used normal office lighting and rigged up the Tiny Code Reader on the workbench with some rulers to measure the ranges over which it could read our codes. So, our conditions were fairly optimal without needing extreme measures. Fig.2 shows the regions over which we could perform successful reads. The spans shown are in the horizontal plane, but we found the vertical spans to be much the same. The functional span (of the camera’s field of view) varies between 33° close up and 13° at a distance. At 150mm, the 62mm code covers 22° of the sensor’s field, while it covers only 4° at 900mm. February 2026  81 QR Codes QR codes were invented in 1994, and QR stands for “quick response”. QR codes were developed in Japan by Denso Wave, originally as an improvement on linear barcodes used to track automobile parts. Denso Wave maintains the website at www.qrcode.com/en/ These applications previously used codes similar to the EAN and UPC barcodes used in retail environments to identify units of stock. Like linear barcodes, QR codes are a pattern of light and dark shapes that encode data. The design of linear barcodes is in turn inspired by Morse code. Other 2D barcode types also exist. The EAN (European article number) linear barcode can encode 13 numeric digits, equivalent to 43 bits of data. The simplest QR code can hold 152 bits, while there are versions that can encode up to 23kbits (2.9 kilobytes) of data. While linear barcodes have error detection, QR codes support multiple levels of error correction and can be decoded even when some symbols are completely missing. Crucial for their popularity, Denso Wave has made the specifications for standard QR codes publicly available, so it is possible for anyone to create and decode QR codes. Note that some of their specialised codes are still protected by patents, though. Despite having a logo covering some of its modules, this QR code can still be scanned and will provide a link to the Silicon Chip website. Structure The figure below shows the layout of a QR code. The black or white squares are known as modules, and the smallest QR codes measure 21×21 modules; this is known as version 1. Each version adds four modules in each direction, up to 177 × 177 modules for version 40. A reader uses the quiet zone to establish the rough framing of a QR code, then detects the position patterns to determine the exact location and orientation of the code. The alignment and timing patterns provide enough information to determine the location and thus value of each module. Once the module data has been extracted, the format and version information is decoded, which dictates how the remaining data is decoded. It includes redundancy in the form of error detection and correction codes, to allow data to be successfully recovered even if the code is somewhat corrupted. For example, a version 1 code, which can carry up to 152 bits of useful information, has about 200 modules available for data and error correction after the necessary patterns have been counted. There are also different ‘levels’, which allow more data to be encoded with greater redundancy. At the highest level, up to 30% corruption will still allow the data to be recovered. The redundant data uses Reed-Solomon coding, which is also used on compact discs. The format information is used to decode the modules. A mode marker embedded in the data can be used to select between different types of encoding, such as ASCII (byte) data and the Kanji encoding noted earlier. The encoding process also involves interleaving the data, which means shuffling bits around such that a localised ‘burst’ error is easier to detect and correct. This technique is also used on compact discs. Encoding also involves a so-called masking step. The masks are known patterns that are used to modify the image to make it less likely to have artefacts that are difficult to decode, such as areas of a single colour or an uneven count of dark and light modules. The decoding step involves reversing the interleaving and masking processes. All these steps may seem complex, but they make QR codes quite robust. They will work with just about any two colours that can be distinguished by a camera. It’s even possible to create a customised code by deliberately corrupting a QR code and replacing some of the modules with a logo or similar, since the error correction can handle the missing data. There are numerous online QR code generators, although we would be dubious about entering any sensitive information into an untrusted website. Denso Wave provides QR code software at www.denso-wave.com/en/adcd/product/software/ We also found an Arduino library by Richard Moore that can generate QR codes. The example 1. Version information sketch prints a code to the serial monitor using block characters. It can be found by searching for 2. Format information QRCode in the Library Manager or downloaded 3. Data and error correction keys from https://github.com/ricmoo/qrcode/ We tried using it with the Tiny Code Reader 4. Required patterns decoding the codes that the library created and 4.1. Position it worked well enough. Fitting an Arduino board with a display and Tiny Code Reader could be a way to have slow but simple bidirectional com4.2. Alignment munication! 4.3. Timing 5. Quiet zone 82 Silicon Chip While QR codes may appear to be a random assortment of black and white squares, they are actually highly structured and robust. Source: https://w.wiki/BRVs Australia's electronics magazine siliconchip.com.au There is clearly an interplay of factors such as focus and resolution at play. For example, we were also able to read a 31mm-wide code at a distance of 90mm from the sensor; in this case, the code covers 20° of the sensor’s field. The data sheet states that a distance of 10-15cm is best for the camera’s focus. Uses While we thought that the Tiny Code Reader sounded like a novel and interesting device, we weren’t sure exactly what uses it might have. The Useful Sensors documentation does offer one suggestion: as a way to provision WiFi network information to a microcontroller. There are specific code formats intended to carry WiFi network information (SSID, password, encryption type etc), so this seems straightforward enough. It’s probably not practical for a one-off setup, but if a device is expected to connect to multiple different networks, it is quite an elegant method. We have seen smartphones that can display a QR code for this purpose. Similar situations, where a microcontroller needs a small amount of data for an initialisation or occasional configuration, would be well-suited to using a QR code. If an application already requires an I2C bus, no extra I/O pins are needed. While the Reader hardware might end up a bit more expensive than, say, a small display and some buttons, it could simplify the software if the QR code data can be structured to avoid the need to program a complicated user interface. In this regard, it has parallels to the way we used the NFC chip in the IR Remote Control Keyfob (February 2025; siliconchip.au/Article/17730). The bottom of the PCB has an RP2040 processor, flash memory chip and crystal oscillator. The white connector is a JST header that’s compatible with SparkFun’s Qwiic connector system. The Tiny Code Reader is compact and uncomplicated. The top side shown here includes the camera, while the small brownish part is an RGB status LED. In it, the NFC chip is used to provide a one-off configuration of the codes that the Keyfob is programmed to transmit. We also found a YouTube video about a robotics project that uses the Tiny Code Reader to detect QR codes as fiducial (location) markers and allow the robot to know its position. The robot’s work area is populated with small QR codes that hold (x,y) coordinate pairs. The video is at https:// youtu.be/UL-vF4JaKqQ The presenter of the video discusses his experiences implementing Tiny Code Reader in his project. He also mentions the need to put an LED on the robot to illuminate the QR codes. Watching this video made us think that access to some of the Reader’s metadata might also be useful; unfortunately, there is no way to access it. Metadata is simply data relating to other data. For a QR code reader, the firmware would likely have access to data about where the code is within the camera’s field of view and how many pixels it spans. This information could be used to determine the code’s position in space relative to the camera, which would be handy to know in a robotics application. Summary We found the Tiny Code Reader to be a straightforward device that was easy to use and program. The lack of a screen can make aiming the camera a bit tricky, but the LED meant that it was simple to confirm that a code had been read. It feels like a niche device, with limited practical applications, but is a fairly inexpensive unit for what it is capable of doing. The Tiny Code Reader is available from Mouser (485-5744) and DigiKey SC (1528-5744-ND). Raspberry Pi Pico W BackPack The new Raspberry Pi Pico W provides WiFi functionality, adding to the long list of features. This easy-to-build device includes a 3.5-inch touchscreen LCD and is programmable in BASIC, C or MicroPython, making it a good general-purpose controller. This kit comes with everything needed to build a Pico W BackPack module, including components for the optional microSD card, IR receiver and stereo audio output. $85 + Postage ∎ Complete Kit (SC6625) siliconchip.com.au/Shop/20/6625 The circuit and assembly instructions were published in the January 2023 issue: siliconchip.au/Article/15616 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. Wireless Reed Switch This circuit was designed to monitor the opening and closing of a remotely located door, where it was impractical to wire a reed switch into the door and there was no wired power source near the door. The solution was to interface a reed switch to a readily available 433MHz two-button key fob. The interface is a PICAXE microcontroller that emulates the momentary pressing of the keyfob buttons in response to the closing and opening of the door. Keyfob remotes generally have a small button battery or two that can power the fob for a year or more, depending on usage and the quality of the battery. The two-button fob chosen has a small EV1527 encoder chip that generates a fixed (non-rolling) code sequence corresponding to the button pressed. This code is fed to the 433MHz transmitter section of the fob and then received and decoded by a two-­ channel 433MHz receiver at the other end of the radio link. It ultimately 84 Silicon Chip controls a set of contacts on a small onboard relay. The fob and receiver are available in Australia from eBay for under $11, including postage (www.ebay.com. au/itm/155694654180). The package includes a simple procedure to mate the fob to the receiver and to set the receiver into ‘latching’ mode, where Button A on the fob latches the relay and Button B unlatches the relay. To ensure the longest possible battery life, the EV1527 chip in the fob is not continuously powered. Rather, when a button on the fob is pressed, a small PNP transistor is biased on via an external 1kW resistor and the internal pull-down resistors inside the EV1527. This allows the transmission of the code sequence before the button on the fob is released and power to the chip is removed. The PICAXE08M2 has the task of converting the reed switch opening and closing actions into simulated one-second button presses on the fob. The reed switch used is a normally Australia's electronics magazine closed type that opens in the presence of a magnetic field. With the monitored door being closed most of the time, there is no power applied to the PICAXE via 1N4148 diode D1. When the door is opened, the reed switch closes and the PICAXE quickly powers up. Its first task is to bring pin C.4 high, which switches on N-channel Mosfet Q1, which then switches on P-channel Mosfet Q2, which supplies ‘alternative’ power to the PICAXE. This keeps the PICAXE powered up when the door is closed and the reed switch opens, as there is still the Button B/door closed code transmission to be sent. The next thing the PICAXE does is generate the Button A/door open pulse that the receiver uses to latch the small onboard relay. Subsequently, when the door is closed again, the PICAXE senses the reed switch opening via input pin C.3, which is pulled low by a 1MW resistor. It then generates the one-second siliconchip.com.au Button B pulse that the receiver uses to unlatch the relay. After a short delay, the PICAXE then switches itself off by returning its C.4 pin low. The Button A and Button B signals are from PICAXE pins C.1 and C.2, respectively. They are fed to the fob via N-channel & P-channel Mosfet pairs Q3/Q4 and Q5/Q6. The sources of Q4 and Q6 are joined to form the fob button ‘common’, while the drains connect to the specific Button A and Button B inputs to the EV1527. This arrangement preserves the normal power-saving features of the fob and obviates the need to remove components or further modify the fob PCB. It’s important that the miniature pushbutton switches on the fob PCB are left operational, since they are used in the initial receiver setup procedure to determine which button performs the latching function (Button A) and which one perform the unlatching function (Button B). There are five soldered connections to be made to the fob PCB: GND, +Vbat, Button A, Button B and Button common. Each solder point is easy to discern and access. In this circuit, the PICAXE consumes a modest 700μA when powered. This mostly corresponds to when the door is open, so the average battery drain is very small. However, to further extend battery life, if the door is left open for 120 seconds or more, the PICAXE goes into a SLEEP mode, which further reduces power consumption to 60μA. The PICAXE wakes up briefly every minute or so to check the status of the door. If the door is still open, it repeats the SLEEP cycle. If the door has been closed, it generates the Button B/door closed pulse, then switches itself off. Again, depending on door usage and the quality of the batteries used, it is easy to imagine three good-quality AAA batteries lasting well over a year. The PICAXE program, “wireless Circuit Ideas Wanted We pay for your interesting original circuits. We can pay you by electronic funds transfer, credit or direct to your PayPal account. Email your circuit and descriptive text to editor<at> siliconchip.com.au reed switch.bas” can be downloaded from siliconchip.au/Shop/6/3569 Finally, the whole circuit, including the three AAA alkaline cells and holder, fob PCB, PICAXE circuitry and reed switch fit in a small UB5 Jiffy box that can easily be mounted close to the door using adhesive Velcro strips. This makes it easy to remove the box to replace the batteries when the time comes. I used a clear Jiffy box so I could see the flashing blue light from the small LED on the fob PCB whenever the door is opened or closed. David Worboys, Baulkham Hills, NSW ($75). Raspberry Pi Reflash Helper I recently purchased some RP2350B Dev Boards (August 2025; siliconchip.au/Article/18635) from Silicon Chip and am having fun with them. I also use RP2040-based boards for many projects, particularly the RP2040-Zero. This circuit allows one to upgrade the flash of the RP2040 chip on the Pico module after the project is boxed and the BOOTSEL pushbutton is inaccessible, as long as the USB port is exposed. I was forever taking apart cases to upgrade the Pico code, so I thought there had to be a better way! This setup has been great when developing code without having to push buttons or open boxes! It’s based on an Atmel ATtiny85 microcontroller. The signal to pin 3 of IC1 can come from any free GPIO pin on the RP2040 set as an output. When the RP2040 pulls this line low (via GP25 in this case), that signals the ATtiny85 to run the reset sequence to put the Pico into bootloader mode. So any program can have a flash upgrade option. I have tested this while programming the RP2040 with both the Arduino IDE and using MMBasic; the software includes a siliconchip.com.au demo Picomite UF2 image with MM­Basic 5.08, a program and the options set up. It’s intended for testing the circuit (siliconchip.au/ Shop/6/3566). From the MMBasic command line, you can invoke the bootloader mode by setting the GPIO states. Bootloader mode is triggered by bringing TP5 (GP25) high for at least one second. If the Pico is unprogrammed, it will be high-­impedance and so pulled up by the 10kW pull-up resistors, automatically triggering bootloader mode. To avoid “bricking” the device if you accidentally write the wrong UF2 file to the Pico, or a buggy one that can’t trigger the update conditions, I added a ‘heartbeat’ failsafe. The firmware in the ATtiny85 chip needs to see TP5 (GP25) go low periodically (eg, once per second). If this stops, it will Australia's electronics magazine automatically put the Pico into bootloader mode. That’s the same pin that drives the Pico’s onboard LED. A small amount of extra current flows through the added resistors, causing the Pico’s LED to be dimly lit all the time. So make sure your program toggles GP25 low/high at least once per second to avoid the fail-safe (‘watchdog’) timer triggering a firmware update. James Langdon, Kalgoorlie, WA ($75). February 2026  85 SERVICEMAN’S LOG Closed for Christmas! Dave’s on an early holiday, so this month we’re instead featuring some of our contributor’s items, starting with a weather station that was fixed up by Bryce Templeton. Despite being a solar-powered unit, it still needed non-rechargeable lithium cells. I bought a new weather station about three years ago to replace an old one that had fallen to bits. I purchased it online as, at the time, store-bought units did not offer what I thought was a handy option: solar cells to power the outdoor section. That would save me having to pull the unit down to change batteries. I should have done more homework, as when it arrived, I found that it still required batteries and I was warned in the instructions not to use rechargeable cells. In fact, they recommended using non-rechargeable lithium cells. I didn’t have any lithium cells on hand and, as I was keen to get it going, I used ordinary alkaline AA cells, which worked fine. The theory of the solar cells was that the unit will run on solar if it is available; otherwise, it is powered by the batteries. This results in the batteries needing to be changed about every six months. This was the situation for more than a year, when I noticed that a section of the indoor display was blank. This was the section that shows the intensity of the sunlight in W/m2, and the UV index. So the next time I had to change batteries, I decided to investigate this problem. An examination of the device showed that while it was well made mechanically, electronically, it was a different story. Getting it to connect to the home WiFi had been an arduous task, and it never managed to send anything to Weathercloud. Items Covered This Month • Fiddling with a finicky solar weather station • Bruce Pierson’s troubles: lights, fans and angle grinders • Repairing a foldback monitor speaker • The bargain bin 65-inch TV 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 86 Silicon Chip Anyway, I took the covers off and discovered that the sunlight sensor is a tiny disc-shaped PCB containing an unmarked IC encased in clear material. It lives in a tower with a small window on the top of the unit. Examination didn’t show anything unusual, so I did a quick re-solder of the ribbon cable that connects to the main board, which of course did nothing. Alongside the sunlight sensor is a bubble level, apparently so that the unit can be mounted truly level, which is important for tipping-bucket rain gauges. Unfortunately, in most installations it can only be seen from a helicopter or drone. I decided to try to get a replacement sunlight sensor and sent off an email to the firm I had purchased it from. After several very confusing emails, in which they never said if they actually sold parts or if they had this part, they came up with the best ‘catch-22’ I have heard: even if they did have the part, I would not be able to buy it from them unless I had previously bought one. At this point, I gave up on the sunlight sensor. However, on this battery change, I decided to try the lithium cells they recommended. It seems that alkaline cells have problems if the temperature drops below zero, whereas lithium cells will operate to -20°C. There is not much chance of even 0°C here, but I thought they might last longer, so at great expense, I purchased a pack of four. I installed three of these, and all was well. I was very surprised about a week later to find the outdoor unit not transmitting. Down it came again, and I found the 4.5V battery reduced to about 1.5V. Measuring each cell revealed that one cell was reading about 0.5V in reverse! I decided that I must have gotten a dud cell and replaced it with the remaining new cell, and we were away again. Australia's electronics magazine siliconchip.com.au The weather station, both mounted and lying around (shown left), and the output data provided on the internal display. But not for long; a week later, again no transmission. Again, down came the unit, and again, I found one cell with reversed voltage. Closer investigation this time revealed that with no batteries inserted, there was about 2V at the battery terminals. The penny dropped; I took the unit out into the sun, and the voltage at the empty battery terminals shot up to about 7V! After tracing the circuit as best I could, I came to the conclusion that the solar was an add-on, as there was no circuitry on the main PCB to do with it. It seems that the solar cells are just paralleled with the battery, using the battery as a voltage regulator to prevent the voltage from going too high in strong sunlight. Apparently, alkaline cells tolerate this treatment, but lithium cells strongly rebel. This would indicate that although the instruction book recommends lithium, it had never been tried in practice, or more likely, the book was written before the solar addition came about. The solution? Well, it’s currently running with no problems on alkaline cells. A proper solution would be to use some sort of charge regulator and fit rechargeable cells. Editor’s note: maybe this is a case of confusion between lithium and lithium-ion cells, as noted in my July 2025 editorial? Still, charging lithium-ion cells without current or voltage limiting is a bad idea! Bryce Templeton, Mudgeeraba, Qld. Bruce Pierson’s troubles around the globe My wife asked me to replace a light globe as it was not working. I checked our box of spare globes, but I did not have that wattage, only the next wattage up. I decided to use a globe from the lounge room to replace the failed globe and put the higher wattage globe in the lounge room. But when I put the new globe in the lounge room light, it did not work. It was brand new, so I was not impressed. On inspection, I could immediately see why it didn’t work. One of the contacts on the base was completely missing. I wondered if I could repair it by using a contact siliconchip.com.au from the failed globe. I managed to prise out the contact from the failed globe with the point of a knife, so it looked possible. I tried to push the contact into the hole in the base of the new globe, but it kept popping out. I could see that there was a springy wire in the hole that was causing this. It seemed that the wire should be on the side of the hole and not in the middle, so it was a manufacturing fault. This also explained why the original contact had come out. I managed to bend the wire into the correct position using a small flat-bladed screwdriver, and that enabled me to push the contact in and it remained in place very firmly. I put the globe into the light and it worked. Of course, I could have taken it back for a replacement, but I presume it would have ended up in landfill despite such a simple fault. That would have taken up more of my time, too. This was an unusual situation that I have not encountered previously. It must have been missed in the quality check, or maybe the contact fell out during transporting. It’s always pleasing to rescue these devices. A wall-mounted fan repair Our son had been using a Heller wall-mounted fan. My wife asked me if I could clean it and put it away, as our son no longer needed it. I asked her if it still worked, and she told me he’d said that it worked the last time he used it. I started by unclipping the front guard and removing it. Then I unscrewed the blade retaining nut, which is a left-hand thread. That enabled me to remove the blade for cleaning. Then I thought I had better test the fan before cleaning to ensure it did work. There was no use in cleaning it if it didn’t work. This is an electronically controlled fan, and sometimes such types decide to stop working for some reason. In this case, when I pressed the start button, nothing happened. I felt the blade spindle, and I could tell that the motor was trying to turn it, but it was not succeeding. This indicated that the bushes had run dry and seized, which is a Australia's electronics magazine February 2026  87 cosmetic condition, but they are a lot more reliable and longer-lasting than newly purchased fans, which often fail when they are just out of warranty. A bit of work, some lubrication and a good clean, restored this fan to good working order again. The photo opposite shows the fan after repair and cleaning. Following with a Bosch 9-inch angle grinder common thing to happen with fans, so I switched it off and unplugged it. I then refitted the blade so I could turn the fan, finding that it was very difficult to turn. I removed the blade again, then unscrewed the nut holding on the back guard and took the fan out to my workshop. With a #2 Phillips screwdriver, I removed the four screws holding on the front plate that the back guard is attached to, then I removed the rubber plug on the back of the motor guard and unscrewed the single screw and removed the motor guard. I refitted the blade so the fan could be turned over by hand. Next, I put a few drops of engine oil on the front bush and turned the fan multiple times. It was still difficult to turn, so I added a few drops of engine oil on the back bush and continued turning the fan by hand. Repeating this process several times eventually freed up the bushes, and the rotor spun freely, so I removed the fan blade again. After wiping up the excess oil, I plugged the fan in to test it, and it worked nicely, so I refitted the back motor guard and the front plate after cleaning them. It is better to use engine oil when servicing fans, rather than machine oil, which is too light for this purpose. I have used this process many times on various fans (including exhaust fans) with good results. With the fan working again, I finished the cleaning job. I used a brush to clean the front and back guards, then a damp cloth to wipe the blade clean. I dried it with a dry cloth. This blade cleaned up easily, as the fan was relatively new. I reassembled the fan and gave it a good test run on all three speeds. As it was now working correctly, I put a cover on it and put it away for future use. This particular fan has a bracket that is screwed to the wall, and the fan sits on that bracket, so it isn’t really portable. When cleaning fan blades, I’ve found in some cases that it is sometimes necessary to use a brush and soapy water if the dirt is really stuck to the blade; still, they mostly come clean with a damp cloth. The hardest fans to clean are used fans we pick up at the Tip Shop. These old fans are usually not in very good 88 Silicon Chip I can’t remember where I got this 9in Bosch angle grinder, but I’ve never used it. I’m used to 100mm and 125mm angle grinders; this one is much scarier at 230mm. Still, I needed to cut some concrete, and the smaller grinders just would not cut deeply enough. I managed to find a 230mm diamond blade on eBay for $27.20 (they are normally over $100!). The grinder did not come with a tool for replacing the blade. I found the correct tool on eBay, but it was $30, so I decided to make one. I got a section of power pole bracing that I’d picked up at the tip shop and cut it to length. I then drilled three holes in it, and I welded a pin in the two smaller holes. It was good enough to remove the grinding disc that came on the tool and fit the diamond disc. I also fitted the side handle to the grinder for added safety. The grinder would not start if the disc was in the vertical position, but it would start with it horizontal and kept running when turned vertical. After cutting the concrete, I checked the cable and brushes; they were all good. Later, when I needed to do some more concrete cutting, the grinder no longer worked. I removed the cover and checked the switch with my multimeter. The switch was open circuit with the trigger held in, so I would have to replace it. I suspect that the switch had just worn out. I found a switch listed for this model, but it was over $30. It was a little different from the original but looked like it should fit. I then changed my search criteria and found the same switch listed for a slightly different model grinder for $17.50. Having eventually received it, I compared it with the old switch. There were some differences, but the new switch looked like it should fit in the case the same way as the old Australia's electronics magazine siliconchip.com.au The repaired fan (left) and Bosch angle grinder (right). one. The main differences were the shape of the trigger and the X2 capacitor being in a different place. It fit nicely into the handle section of the grinder, which comes off the main body after removing four screws. However, when I tried to plug the internal plug on the main part of the grinder into the switch, it would not go over the pins. I then realised that the pins on the end of the switch were closer together on the new switch than on the old switch. I’m not sure if this was because the new switch was for a later version of the grinder, or because I had ordered a switch for a slightly different model. In any case, after making some modifications to the plug and the switch, they went together. Thankfully, the only real difference was the size of the blank section in the middle of the plug that spaced the terminals apart, so it was easy to modify. I cut out the middle spacing section with a utility knife, which left me with two separate insulated plugs. On my first attempt, I found that I could not get the plugs onto the pins, as the switch had a ridge in the middle of where the plugs plug in, which the original switch did not have. I used a utility knife to remove this ridge, then it all went together. It was quite a nightmare getting the grinder back together because of the way the two handle halves went over the switch and over the main body of the grinder. It took a lot of trial and error to get everything lined up, and the two handle sections correctly positioned on the main body of the grinder. Then it was just a matter of installing the four screws and the repair was complete. A quick search revealed that it would cost $450-500 to buy a new grinder like this. Because I got it for free and only spent $17.20 for a new switch and a bit of time, I ended up with a good quality Bosch angle grinder for a fraction of the cost of a new one. I will repair anything that I can get parts for! Bruce Pierson, Dundathu, Qld. Repairing a “VoiceSolo” foldback monitor speaker The TC Helicon VoiceSolo foldback monitor is a self-­ powered speaker designed to be used with a microphone stand, with the mic boom attached to the top of the monitor. I was recently presented with one of a set of four that was described as “dead” by the users. siliconchip.com.au Australia's electronics magazine February 2026  89 The VoiceSolo preamplifier (left) and power supply (right), with the failed electrolytic capacitors circled in red. The monitor case is of diecast aluminium, with the front assembly containing the speaker and input controls, secured to the main enclosure by four screws. Applying power to the monitor and connecting an input signal confirmed no power indicator LED and no sound. Opening up the monitor revealed a loudspeaker and four circuit boards, one attached to the front assembly and three within the rear enclosure. The power amplifier is a ‘BASH’ amplifier design where a Class-AB bridge amplifier module is supplied with a main DC supply that is modulated by a secondary switch-mode variable voltage supply tracking the amplifier audio input. This particular design has a 200W power supply and amplifier module built on two boards by Indigo Canada. A check with a multimeter at the main switch-mode supply PCB confirmed a steady +60V DC main supply, but none of the four low voltage rails, ±24V and ±15V, were present. Visual inspection revealed a ¼W resistor burned to a crisp and a 100μF 25V electrolytic ruptured. This power supply derives its low-voltage rails from an additional secondary winding on the main switching transformer. A group of diodes and electrolytic capacitors create unregulated positive and negative DC rails, which are then fed through a 7824 linear regulator and a discrete transistor regulator circuit to deliver ±24V rails for the BASH amplifier control circuits. A pair of 7815/7915 linear regulators supply the mixer/ preamplifier circuits. The ¼W resistor is connected between the bottom of the secondary winding and GND, acting as a fuse. A circuit diagram could not be found with extensive internet searches, so I needed a working monitor for reference. This duly arrived after a week, revealing the resistor to be 100W. I replaced the two failed parts and reassembled the monitor, unsure whether the fix would work. 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. 90 Silicon Chip I couldn’t spot any other components visually damaged; I needed to test the low voltage rails before doing anything more. Reconnecting the loom and powering on the speaker, I was greeted with a green power light on the front panel, so I grabbed the multimeter to check the rails. But before I could take any readings, the resistor began emitting magic smoke and failed again. So, we have a burned-out resistor and a ruptured capacitor on the output of the 7915 regulator. What was causing this? My best guess was a failed 7915 regulator, as it was feeding the failed electrolytic. Since it is secured to the heatsink with a shared mount to the other three regulators and a TIP30C power transistor that delivered the -24V rail, I decided to test the lot. I removed the four from the circuit board and tested the regulators on a breadboard. All were good, including the 7915. So I reassembled the power supply board and replaced the resistor again. This time, I was wiser and powered up the switch-mode supply board on its own on the bench. This time, the resistor didn’t fail, and all four low-voltage rails were within spec. So the root cause was elsewhere. The preamplifier, mixer and tone control board were mounted to the front panel in an assembly comprising the combined vent/carry handle. Dismantling the assembly revealed a board with eight NE5532 op amps, 13 electrolytic capacitors and a mix of SMD components. I was about to begin meter checking each of the NE5532 op amps when I noticed a very slightly bulging electrolytic hiding among a cluster of four identical ones. A closer look revealed that this was another 100μF 25V capacitor. In fact, all the electrolytics on the board were 100μF 25V, of the same make and type. I removed the bulging one and tested it on the component tester. Its value had risen to over 150μF, and it had a high ESR reading as well. The VoiceSolo speaker and I/O box. Australia's electronics magazine siliconchip.com.au Given that the four DC regulator circuits were working correctly, I began to suspect the electrolytic capacitors themselves. In for a penny, in for a pound, and out with the vacuum desoldering gun. I removed and tested all 13 of the capacitors. Only five tested good! Some of the capacitors were functioning as supply bypasses, while others were for audio coupling. I decided to replace all the 100μF capacitors in the monitor with good-quality low-­ leakage types. For the ‘acid test’, I reconnected the wiring harness and gingerly reached for the power switch. On powerup, it was a bit anti-climatic. No magic smoke, all supply rails within spec, green power LED on, good to go! I reassembled the monitor and gave it a thorough bench test, playing Sting at a modestly loud level. With so many failed capacitors, I began to wonder if this might be a manufacturing problem. Would it happen again soon with the other three monitors? I checked the spare that was used to identify the 100W resistor. In this one, the resistor hadn’t failed, but sure enough, there were another eight faulty electrolytics. In the end, I dismantled and replaced the capacitors in all four of these monitors. Testing revealed the majority in each to be on their way out, with strange capacitance and high ESR readings. In all, I replaced 60 capacitors. Ray Ellison, Dover Gardens, SA. The $19.00 65-inch television set Dave Thompson’s article in the August 2025 issue about repairing discarded devices struck a chord with me (siliconchip.au/Article/18644). Over the years, while walking around the streets of McCrae, I’ve rescued many perfectly good items discarded by their owners and left on the nature strip for council collection. The list includes a Jensen X-125 subwoofer, a 150mm reflector astronomical telescope, sundry computers and laptops, to name just a few. However, my most rewarding nature strip pickup was a large-screen Sony TV. It was buried under a pile of old plastic chairs, a mattress and other paraphernalia, and barely recognisable as a TV. I ventured onto the premises and asked the owner if it was a TV, and why he was throwing it out. He said it had simply stopped working, and as an expert IT consultant, he had concluded that “it was a transformer failure” and not worth repairing. I was welcome to take it since it would reduce the volume of his discards, and hence his fee to the council. siliconchip.com.au The set was a 65-inch Sony Bravia KD-65X7000E of about 2018 vintage. It was awkward to load into the car – the thing was simply enormous and quite heavy – but we made it home and successfully unloaded the device into the shack for further inspection. Fearful of flexing the set too much, I cautiously removed its rear cover to be confronted by three PCBs: an RF board, a motherboard and a power supply assembly. They were absolutely dwarfed by the screen itself, and I wondered how such a small set of electronics could drive such a monster screen. A quick check revealed that the power supply was not working, further confirmed by two blackened diodes, which had obviously ‘released their smoke’. Without really checking these diodes, I reckoned a couple of 400V 3A general-­ purpose silicon devices from my parts bin would make suitable replacements, so I quickly substituted a pair of new diodes. Switching on the TV produced screen images for about 20 seconds until my replacement diodes got very hot, also lost their smoke, then the set died again. Sony, in their quest to minimise power consumption, had specified high-speed schottky diodes for their power supplies in the KD-65X7000Es, which meant my substitutes were not suitable. So I bought a strip of 10 schottky diodes from Amazon for about $20 and replaced the two faulty ones, as well as the other two that made up the bridge rectifying circuit. This time, the set sprang into life and continued to operate satisfactorily. The new diodes were barely warm. The 4K picture was crystal clear, with no screen defects, and great audio – so I had acquired a marvellous 65-inch TV for about $4, plus another $15 for a new remote control! Unfortunately, the owner had discarded the mounting hardware for the set, so I had a bit more work to do to make a frame to support the set from my old audio cabinet. The attached photo shows the setup, with the Jensen X-125 at lower left. Happy days. SC Rob Fincher, McCrae, Vic. Australia's electronics magazine February 2026  91 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. 02/26 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 PIC12F617-I/P 110dB RF Attenuator (Jul22), Basic RF Signal Generator (Jun23) 2m VHF CW/FM Test Generator (Oct23) Active Mains Soft Starter (Feb23), Model Railway Uncoupler (Jul23) Battery-Powered Model Railway Transmitter (Jan25) 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) Dual Train Controller (Transmitter / TH Receiver, Oct25) PIC16F1455-I/SL Battery Multi Logger (Feb21), USB-C Serial Adaptor (Jun24) Battery-Powered Model Railway SMD Receiver (Jan25) USB Programmable Frequency Divider (Feb25) Dual Train Controller (SMD Receiver, Oct25) 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) Vacuum Controller (Oct25) 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) PIC16F18126-I/SL DCC Decoder (Dec25), RGB LED Star (Dec25) PIC16F18146-I/SO Versatile Battery Checker (May25), RGB LED ‘Analog’ Clock (May25) USB-C Power Monitor (Aug25), DCC Remote Controller (Feb26) PIC16LF15323-I/SL Remote Mains Switch (TX, Jul22), Secure Remote Switch (TX, Dec23) STM32G030K6T6 Variable Speed Drive Mk2 (Nov24) PIC16F1847-I/P PIC16F18877-I/PT 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 PIC32MX270F256D-50I/PT Wii Nunchuk RGB Light Driver (Mar24) AM-FM DDS Signal Generator (May22) Power LCR Meter (Mar25) Digital Preamplifier (Oct25) $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 siliconchip.com.au/Shop/ DCC REMOTE CONTROLLER KIT (SC7552) (FEB 26) MIC THE MOUSE KIT (SC7508) (AUG 25) MAINS HUM NOTCH FILTER (SC7598) (FEB 26) USB-C POWER MONITOR KIT (SC7489) (AUG 25) DCC BASE STATION KIT (SC7539) (JAN 26) 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) PICO/2/COMPUTER (SC7468) (APR 25) 433MHz TRANSMITTER KIT (SC7430) (APR 25) ROTATING LIGHT FOR MODELS KIT (APR 25) PICO 2 AUDIO ANALYSER SHORT-FORM KIT (SC6772) (MAR 25) USB PROGRAMMABLE FREQUENCY DIVIDER (SC6959) (FEB 25) NFC PROGRAMMABLE IR KEYFOB (SC7421) (FEB 25) Includes all required parts, except for the case and wire/cable (see p63, Feb26) $35.00 Includes everything except for the case and power supply (see p53, Feb26) $50.00 Includes everything but the plastic case, power supply and some optional parts. The Pico 2 is supplied but not programmed (see p39, Jan26) $90.00 RGB LED STAR KIT (SC7535) Includes the mostly-assembled board and all non-optional components except the power supply (see p43, Dec25) EARTH RADIO KIT (SC7582) Includes everything to build the radio itself except the case and battery, plus the plug for the antenna (see p65, Dec25) (DEC 25) $80.00 (DEC 25) $55.00 DCC DECODER KIT (SC7524) (DEC 25) RP2350B COMPUTER (NOV 25) Includes everything in the parts list (see p73, Dec25) Assembled Board: a fully-assembled PCB with all non-optional components, front and rear panels are sold separately below (SC7531; see p28, Nov25) - front & rear panels (SC7532) - 8MiB APS6404L-3SQR-SN PSRAM SOIC-8 IC (SC7530) DUAL TRAIN CONTROLLER MICROCONTROLLERS (OCT 25) PICKIT BASIC POWER BREAKOUT KIT (SC7512) (SEP 25) - PIC16F1455-I/P programmed with 0911024D.HEX (Transmitter) - PIC16F1455-I/P programmed with 0911024S(or T).HEX (Receiver, TH) - PIC16F1455-I/SL programmed with 0911024S(or T).HEX (Receiver, SMD) firmware ending with “S.HEX” is for train 1, while “T.HEX” is for train 2 Includes all parts except the jumper wire and glue (see p39, Sep25) RP2350B DEVELOPMENT BOARD (AUG 25) 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) $25.00 $90.00 $7.50 $5.00 $10.00 $10.00 $10.00 Includes all parts except a CR2032 cell (see p64, Aug25) 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) $60.00 $20.00 $65.00 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 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 Includes an assembled PCB, separate Raspberry Pi Pico 2 and front/rear panels $120.00 Includes the PCB and all onboard parts (see p75, Apr25) $20.00 Complete kit which includes the PCB and all onboard components (see p60, Apr25): - SMD LEDs (SC7462) $20.00 - Through-hole LEDs (SC7463) $20.00 The Pico Audio Analyser kit from Nov23, but with an unprogrammed Pico 2 $20.00 $37.50 Complete kit: includes all components (see p85, Feb25) $30.00 Complete kit: includes all required items, except the cell (see p67, Feb25) $1.00ea $5.00 *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. $50.00 $60.00 $25.00 PRINTED CIRCUIT BOARDS & CASE PIECES PRINTED CIRCUIT BOARD TO SUIT PROJECT 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 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 DATE 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 MAR24 APR24 APR24 APR24 MAY24 MAY24 MAY24 JUN24 JUN24 JUN24 JUN24 JUN24 JUL24 JUL24 JUL24 AUG24 AUG24 AUG24 AUG24 PCB CODE 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 16103241 08101241 08104241 07102241 04104241 04112231 10104241 SC6963 08106241 08106242 08106243 24106241 CSE240203A CSE240204A 11104241 23106241 23106242 08103241 08103242 Price $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 $20.00 $15.00 $10.00 $5.00 $10.00 $2.50 $5.00 $10.00 $2.50 $2.50 $2.50 $2.50 $5.00 $5.00 $15.00 $10.00 $12.50 $2.50 $2.50 For a complete list, go to siliconchip.com.au/Shop/8 PRINTED CIRCUIT BOARD TO SUIT PROJECT 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 5MHZ 40A CURRENT PROBE (BLACK) BATTERY MODEL RAILWAY TRANSMITTER ↳ THROUGH-HOLE (TH) RECEIVER ↳ SMD RECEIVER ↳ CHARGER 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 DUCTED HEAT TRANSFER CONTROLLER ↳ TEMPERATURE SENSOR ADAPTOR ↳ CONTROL PANEL MIC THE MOUSE (PCB SET, WHITE) USB-C POWER MONITOR (PCB SET, INCLUDES FFC) HOME AUTOMATION SATELLITE PICKIT BASIC POWER BREAKOUT DUAL TRAIN CONTROLLER TRANSMITTER DIGITAL PREAMPLIFIER MAIN PCB (4 LAYERS) ↳ FRONT PANEL CONTROL ↳ POWER SUPPLY VACUUM CONTROLLER MAIN PCB ↳ BLAST GATE ADAPTOR POWER RAIL PROBE RGB LED STAR EARTH RADIO DCC DECODER DCC BASE STATION MAIN PCB ↳ FRONT PANEL REMOTE SPEAKER SWITCH ↳ CONTROL PANEL DATE 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 AUG25 AUG25 AUG25 AUG25 AUG25 SEP25 SEP25 OCT25 OCT25 OCT25 OCT25 OCT25 OCT25 NOV25 DEC25 DEC25 DEC25 JAN26 JAN26 JAN26 JAN26 PCB CODE Price 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 9049-01 $5.00 09110241 $2.50 09110242 $2.50 09110243 $2.50 09110244 $2.50 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 17101251 $10.00 17101252 $2.50 17101253 $2.50 SC7528 $7.50 SC7527 $7.50 15104251 $3.50 18106251 $2.00 09110245 $3.00 01107251 $30.00 01107252 $2.50 01107253 $7.50 10109251 $10.00 10109252 $2.50 P9058-1-C $5.00 16112251 $12.50 06110251 $5.00 09111241 $2.50 09111243 $5.00 09111244 $5.00 01106251 $5.00 01106252 $2.50 DCC REMOTE CONTROLLER MAINS HUM NOTCH FILTER MAINS LED INDICATOR FEB26 FEB26 FEB26 09111245 01003261 10111251 NEW PCBs $5.00 $7.50 $2.50 We also sell the Silicon Chip PDFs on USB, RTV&H USB, Vintage Radio USB and more at siliconchip.com.au/Shop/3 SOnline ilicon Chip Shop Kits, parts and much more www.siliconchip.com.au/Shop/ Rotating Lights April 2025 USB-C Power Monitor August-September 2025 USB Power Adaptors May 2025 SMD LED Complete Kit SC7462: $20 TH LED Complete Kit SC7463: $20 Short-Form Kit SC7489: $60 siliconchip.au/Article/17930 siliconchip.au/Series/445 siliconchip.au/Article/18112 This kit includes everything needed to build the Rotating Light for Models, except for a power supply and wire. This kit includes all non-optional parts, except the case, lithium-ion cell and glue. It does include the FFC (flat flexible cable) PCB. 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. Compact HiFi Headphone Amplifier Complete Kit SC6885: $70 Complete Kit with choice of USB socket SC7433: $10 PICKit Basic Power Breakout Board September 2025 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. Mic the Mouse Complete Kit SC7508: $37.50 August 2025 siliconchip.au/Article/18637 It includes everything needed to build one Mic the Mouse, except for solder, glue and a CR2032 cell. Complete Kit SC7512: $20 siliconchip.au/Article/18850 Includes the PCB, all onboard parts and a length of clear heatshrink tubing. Jumper wire and glue is 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 Columbia TR-1000 Six-Transistor Portable Radio This attractive set uses a pretty standard (for the time) six-transistor circuit. It has a few quirks, though, such as a relatively low maximum audio output power, unusual transistor bypassing and a slightly odd audio feedback configuration. By Ian Batty T he Regency TR-1, released in 1954 (see April 2013; siliconchip.au/ Article/3761), was the first practical transistor radio made in significant numbers. So Columbia’s 1957 release of the TR-1000 was well into the heyday of the transistor radio. It seemed everybody in electronics was offering a transistor radio. While most buyers would not know much about how the radios worked, they probably assumed that six transistors were better than four, but maybe thought that seven would break the bank. So, six it was. But manufacturers would still need a way to make their offering stand out. Another small black brick (like the TR-1) was not going to do it. Some 50 years earlier, Kodak’s Brownie camera had introduced affordable photography, initially to children. By the 1940s, small, affordable cameras were ubiquitous. They also represented ‘go-anywhere’ convenience, and makers of portable radios took notice. It seems Columbia intended to cash in on the camera vibe, putting its TR-1000 into a handy leather case. Radiomuseum lists 423 offerings from Columbia, but only five transistor sets. The TR-1000 was made by Roland Radio Corporation, with the only real difference from their model 71-483 being the type of output transistors. The TR-1000 thus was an early ‘badge-engineered’ design adopted by Columbia. The TR-1000 circuit (Fig.1 overleaf) looks pretty much like any set of the era. It has a converter with emitter feedback (TR1), two intermediate frequency (IF) stages (TR2 & TR3), a diode for demodulation and automatic gain control (AGC; D1), an audio driver (TR4) and a push-pull Class-B output (TR5 & TR6). It uses single-point grounding for the IF stages, with the base and collector circuits bypassed to the emitter rather than to chassis ground. Ensuring that the base, emitter and transformer cold ends share the same AC reference point prevents emitter degeneration and eliminates signal loss or unintended feedback, without requiring a high-value emitter bypass With the right-hand knob removed, you can see the concentric shafts underneath. The D-shaped shaft in the middle is the drive shaft. The outer sleeve rotates more slowly. siliconchip.com.au Australia's electronics magazine February 2026  95 capacitor. This technique reduces the component count and is most commonly used at HF to UHF, where every component’s lead inductance must be at a minimum. Eliminating the emitter bypass removes the possibility of its lead inductance affecting the circuit. Controls The tuning gang has trimmer capacitors at both ends. While that’s unusual, it does leave one pair easily accessible, ensuring that the antenna/local oscillator (LO) trimmers are accessible and that the top end can be aligned without needing special tools. The tuning dial drives the tuning capacitor via a planetary reduction gear for precise tuning. The photo on page 95 shows the bright metal driveshaft, concentric to the tuning gang shaft, on the right. The frequency indicator disc fits the tuning gang shaft and is viewed through the transparent, knurled tuning knob. A few resistors in the set (R2: 33kW, R5: 2.2kW and R16: 33kW) are ±5% types. It’s not clear why just these three, especially R5 (the decoupling resistor for the converter), have a tighter than typical specification for the time (±10% was more common, and ±20% was not unheard of). It’s especially odd given the very wide spread of transistor parameters at this early stage of development. The audio section uses negative feedback via 33kW resistor R16, but 96 Silicon Chip this behaved in a peculiar way during testing. More on this later. The SAMS circuit shows an earphone socket, with the usual cutting-­ off of signal to the speaker when a plug is inserted. My set lacked this refinement, with a pop rivet filling the hole in the case. The SAMS circuit was not super-legible, and features some oddities, especially with capacitor numbering and notation. Capacitors C1 through C4 are electrolytics. It’s customary to begin component numbering at upper left, yet C1 is the second from the right, then comes C2, well to the left as the AGC audio filter for the first IF amplifier. The paper/ceramic capacitors (C5 through C20) are numbered according to their location, but the IF bypass capacitors (C11a, C11b, C13a & C13b) are multi-capacitor assemblies. Paper/ceramic capacitor values are given in picofarads, thus SAMS’s C12 is 10000pF (10nF), but C14 is “.05”, presumably meaning 0.05μF (50nF). Circuit description The Photofact circuit’s component numbering is peculiar. Capacitors C1 to C4 appear to the right and centre. The first fixed capacitor in the diagram was then C5. I have renumbered all components to conform to accepted drawing practice. As is common with a circuit containing only PNP transistors, the battery supply (9V) is positive to ground, Australia's electronics magazine making all circuit voltages negative. Converter TR1 uses collector-­emitter feedback, continuing the design used in the first “trannie”, Regency’s TR-1. The tuning gang, with its cut plate design guaranteeing accurate tracking between the oscillator frequency and the tuned signal, eliminates the need for a padder capacitor. The low forward bias supplied by R2/R1, in combination with the high emitter resistor R3, is only about 0.1V. This confirms the converter is operating in the Class-C mode that is vital to the conversion process. In Class C, the transistor is conducting for less than 180° of the signal cycle, compared to close to 180° for Class B, more than 180° for Class AB and 360° for Class A. By biasing the transistor so that it only conducts in short pulses, its nonlinear behaviour is emphasised. The short conduction bursts act like a ‘sampling’ of the LO and RF signals, which naturally generates the frequency products. We don’t want faithful amplification of either input on its own; we want the intermodulation products, including the downmixed IF signal that’s later extracted. TR1, a 2N411, feeds the tapped, tuned primary of the first IF transformer, L3. Its untuned, untapped secondary feeds first IF amplifier TR2, a 2N409. This stage works with minimal bias, supplied via R6 (100kW), for a collector current of around 0.7mA. This siliconchip.com.au allows the demodulator’s filtered DC output to control TR2’s gain via the AGC function. Capacitor C10 (3μF) bypasses the audio component of the demodulator’s output, while dual capacitor C11a/ C11b bypasses the collector circuit and the base circuit directly to TR2’s emitter. As noted above, this single-point technique is more effective than the usual bypassing directly to ground. TR2 is neutralised by feedback from the second IF transformer via 4.7pF capacitor C12. TR2 drives the tapped, tuned primary of the second IF transformer, L4, with the signal from its untuned, untapped secondary passing to the second IF amplifier, TR3. This stage (also a 2N409), as in most six-transistor sets, works with fixed bias and is also neutralised by 4.7pF capacitor C14. TR3 feeds the tuned, tapped primary of the third IF transformer, L5. Its untuned, untapped secondary feeds demodulator/AGC diode D1, a 1N60. All three front-end transistors are alloyed-junction germanium types. These use the same construction as the Philips/Mullard OC44/OC45, but with lower frequency specifications. The audio signal is developed across 5kW volume control R14, and the IF component is filtered out by 20nF capacitor C15. The AGC signal is fed back to the base of the first IF stage (TR2) via 3.3kW resistor R9. The audio signal then goes to the base of the audio driver transistor, TR4 (a 2N405), which uses ‘combination bias’ – a resistive divider at the base, plus an emitter resistor for stabilisation. There’s a top-cut capacitor (C19, 2nF) from TR4’s collector to ground, reducing noise and making the ultimate sound less shrill. TR4 drives the interstage/phase-splitting transformer, T1. Signals from T1’s secondaries feed the bases of the output transistors, TR5/TR6, both 2N407s. These drive the speaker transformer, T2, which then drives the speaker. Capacitor C20 (50nF), placed across T2 primary, adds further ‘top cut’. These days we would use complementary output transistors (PNP and NPN), but in the 1950s, only germanium PNP types were readily available. Early germanium NPN devices did exist, but they were generally inferior in performance. As a result, the preferred arrangement was a phase-splitter transformer driving two identical PNP output transistors in push-pull. Feedback is taken from T2’s secondary and fed, via 39kW resistor R21, to the base of audio driver TR4. Transistor specifications Apart from special types, it’s rare to see valves with a maximum frequency rating. The 6BE6 miniature pentagrid, common in broadcast radios, has been used in FM receivers in the 88-108MHz band. The miniature triode-­pentodes 6U8/6BL8 worked as converters in VHF-band TV tuners. Yet the TR-1000’s germanium converter and IF amplifier transistors would struggle to operate into the middle of the shortwave band, as would the OC44/45 types we’re more familiar with. Philips’ introduction of the alloy-­ diffused OC169~OC171 family offered receiver operation up to 50MHz. The alloy-diffused technology matured with the AF186, able to operate up to 820MHz. The ‘all-diffused’ Mesa and planar transistors that succeeded them easily exceeded 1000MHz (1GHz). But even within manufacturing technologies, maximum operating frequencies vary widely, so we need an explicit ‘frequency rating’ for a transistor. Transistors are specified for high-­ frequency operation in several ways, Fig.1: the TR-1000 circuit includes six alloyed-junction PNP germanium transistors and one point-contact germanium diode. TR1 is the mixer/oscillator with emitter feedback, TR2 & TR3 are the IF gain stages and TR4 is the audio preamplifier which drives phase-splitter transformer T1. The audio output pair, TR5/TR6, drives the loudspeaker or earphones via matching transformer T2. The components have been renumbered to conform to accepted drawing practices. siliconchip.com.au Australia's electronics magazine February 2026  97 often depending on their manufacturer. The most useful specification is the transition frequency, ft. This is calculated by plotting common-­emitter current gain (hfe, beta or β) against frequency. The point where β drops to unity is the transition frequency. Two other specifications exist: the β cutoff frequency (fβ), where common-­ emitter current gain falls to 70% of its mid-band value, and the alpha cutoff frequency (fα), where the common-­ base current gain (hfb, α) falls to 70% – see Fig.2. The transition frequency is the most useful. In practice, the common base fα figure is close to ft. Thus, a common-­ base circuit will operate satisfactorily up to ft. For the common-emitter configuration, say we use a transistor with ft = 1GHz (1000MHz) and hfe = 50. It will have a common-emitter gain of around 1.0 at 1000MHz (ie, at ft), but a gain of around 50 at 20MHz and any frequency below that. This raises a confusing question. The high-gain audio BC109 (β Fig.2: a plot of transistor current gain (common base & common emitter) versus frequency. = 240~900) has ft = 350MHz, while the low-gain BF115 RF amplifier (β = 45~165) has ft = 230MHz. Why bother with the BF115? For a BC109 with ft = 350MHz and hfe = 900, its fβ is just 360kHz (350MHz ÷ 900) – its gain will progressively drop with increasing operating frequency above that. For a BF115 (ft = 230MHz, β = 165), the fall begins at 1.4MHz. While these are clearly the worst cases, the best cases put the BC109 at 1.4MHz, and the BF115 at 5MHz, before their common-­ emitter gain starts to drop off. There’s another reason for preferring the BF115. As explained in The History of Transistors, Part 2 (April 2022; siliconchip.au/Article/15272), an internal resistance exists within the base region. This intrinsic resistance, rbb’, acts as does any resistance: it is a source of electrical noise according to the Stefan-Boltzmann Law. A high rbb’ acts as significant noise source within the transistor. In contrast, a low rbb’ will result in a lessnoisy transistor, and RF transistors – with their relatively low current gains – usually have low values of rbb’ when compared to low-frequency (‘audio’) types. So while a BC109 could replace a BF115 RF amplifier, the result would be a markedly higher noise figure. If you’ve ever seen an audio preamplifier with an RF transistor in the first stage, the explanation should now be clear. The first stage in any processing system typically sets the noise performance for the entire equipment. Low Output Transformer 1st Audio 3rd IFT Driver Transformer Outputs 2nd IF 2nd IFT 1st IFT Oscillator Coil 1st IF Antenna Tuner Oscillator Tuner Converter Ferrite Rod Antenna 98 Silicon Chip Australia's electronics magazine siliconchip.com.au noise is preferable to high gain, as gain can readily be made up following the input stage. Assembly It’s built on a metal chassis with point-to-point hand wiring. The transistors, with their leads offset to conform to the TO-40 packages, are all socketed. The components are packed tightly, with three of the electrolytic capacitors on top of the chassis, connecting via sleeved wires to the underside through holes. Some components in the converter section are placed on an ‘outrigger’ tag strip. It isn’t as packed as Astor’s Mickey OZ valve mantel set, but make sure you have some patience in reserve if you tackle a TR-1000. Cleaning the set up I bought this set on eBay quite a few years back. It’s one of those situations we probably all experience: a rush of blood to the head, purchase, delivery, storage on the ‘get to it one day’ shelf. It was complete, and in good condition, only missing the leather carry strap. The original battery connector had been removed, and a 9V ‘snap’ connector added. Turning it on gave some results, but its performance was pretty poor, failing to give more than about 40mW of audio with a strong signal. The AGC bypass capacitor, C10, was suspect. I’ve had trouble with electros in this position before. With bias resistor R6 at 100kW, even a few microamps of leakage in C10 would upset the AGC action. I replaced it, but to no effect. So I connected my audio generator to the top of the volume pot. Try as I might, this radio refused to give more than around 40mW before clipping. Time for some maths. From a 9V supply, you’ll get a maximum of 18V peak-to-peak across output transformer T2’s primary. That’s an RMS value of about 6.4V. Now, with T2’s primary at 840W, I get a maximum possible 48mW of audio power. So it seems the radio was OK – it’s a good example of ‘know your beast’. That is, don’t assume that every radio works and performs the same as every other one. I checked the SAMS literature, but there was no confirmation of my calculation. Given the maximum of 40mW at clipping, I did all sensitivity measurements at 20mW output. It aligned OK, with the problem being that the IF slugs are only accessible from the underside, and they are partly obscured by components. The slugs have very small slots, maybe 5mm or less, so care is needed when adjusting them. The audio feedback was a puzzle, as it didn’t seem to work. After a bit of faffing about, I discovered that putting a 5kW resistor in series with my audio generator allowed the feedback to take effect. This circuit, unusually, uses shunt voltage feedback via 33kW resistor R16 to the base of TR4, but relies on feeding an input of some 2kW impedance. My audio generator’s output impedance of only 50W was defeating the feedback by putting TR4’s base pretty The top (left) and bottom (right) of the Columbia TR1000 chassis. We haven’t marked it on the photos, but the demodulator diode (D1) can be seen in the right-hand photo at upper right (it has a glass body with a light blue cathode stripe). siliconchip.com.au Australia's electronics magazine February 2026  99 much at AC ground, in terms of impedance. So this is another case of ‘know your beast’. With a 5kW resistor in series with the audio generator, I simply set the audio generator to give 20mW of output, then used a millivoltmeter to measure the actual signal at the base of TR4. level rise; that’s good for a single-stage AGC system. Total harmonic distortion (THD) measured about 4% at 20mW output, and the maximum output was 40mW at clipping, with 10% THD. At 10mW, THD was still 3%. The output waveform appeared asymmetrical, with one half-cycle reduced in amplitude. How good is it? To try to explain this asymmetFor this first generation of six-­ ric waveform, I tested the gain of the transistor sets, it’s pretty good aside 2N407 output transistors, TR5 and from the low maximum audio output. TR6, and got very different gain (β) For 20mW output, it needed readings of 60 and 140, respectively. 350μV/m at 600kHz and 170μV/m This explained the asymmetry. at 1400kHz, with signal+noise:noise The alloyed-junction 2N406~8 ratios better than 20dB. Scaling up to (audio) and 2N409~12 (converter/IF) 50mW out, this is equivalent to around types were released in two different 550μV/m and 270μV/m. Raytheon’s packages: the offset-lead TO-40 and contemporary T-2500, using seven the triangular layout TO-1. Searchtransistors, was only about four times ing my junk box unearthed TO-1 ver(at 600kHz) and twice (at 1400kHz) as sions of these transistors, some of sensitive. which tested well. Substitution did Its RF bandwidth measured as give improved audio performance, ±1.5kHz for -3dB and ±23kHz for and roughly doubled the sensitivity. -60dB. The audio response from Ultimately, though, I left the TO-40 antenna to speaker measured as package transistors in place. Many 150~2600Hz at -3dB; from volume other radios of the era use the eloncontrol to speaker, it was 115~8000Hz. gated cases from the Regency TR-1/ The AGC action showed a 37dB sig- grown-junction era, or shiny cylinnal increase for a 6dB output audio drical ‘bullet’ cases. I reckon the black Versatile TO-40s make the TR-1000 – should you ever see inside one – distinctive. Special handling All knobs, and the frequency indicator disc, come off with finger pressure. Be aware that the knurled tuning knob fits the small, inner bright metal shaft that drives the planetary reduction, and that the station indicator disc fits the larger, outer brass shaft. The frequency indicator shaft is not keyed, so you will probably need to carefully twist it to give a correct frequency indication. Conclusion This set is unusual enough to belong in any collection of transistor radios from the 1950s. From that time of rapidly evolving designs, Columbia’s TR-1000 is a ‘must have’. Radiomuseum has useful information on this set at www.radiomuseum. org/r/cbs_columb_tr_1000.html The circuit appears in SAMS Photofact Folder 5, set 405, and is of better quality. I could not find it in any free online catalog. The SAMS website charges US$15 (+post) for the service SC sheets: www.samswebsite.com 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 100 Silicon Chip Australia's electronics magazine siliconchip.com.au 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 Arduino IDE needs upgrading I’m having a bit of trouble compiling the “Solar_diverter_HWS_2regs.ino” sketch for the Hot Water System Solar Diverter (June & July 2025; siliconchip. au/Series/440). “ESP8266Ping.h” does not exist in the ESP8266-ping library by Alessio Loencini. I’ve tried all available versions of the library. Looking at the library source, the header is called Pinger.h, but using this, I get the following compilation error (C. W., Sawyers Valley, WA): Arduino: 1.8.19 (Windows 10), Board: “Generic ESP8266 Module”: F:\Backup\ InstalledProgs\Arduino\ SolarDiverter\Solar_diverter_ HWS_2regs\Solar_diverter_ HWS_2regs.ino: In function ‘void loop()’: Solar_diverter_HWS_2regs:365:6: error: ‘Ping’ was not declared in this scope: if(Ping.ping(inverter)){ ● We don’t think the software is compatible with such an old version of the Arduino IDE, v1.8.19, from late 2021. We used Arduino IDE version 2.3.4. We suggest you upgrade to a newer version of the IDE. Alternatively, you may be able to download it from www.arduinolibraries.info/ libraries/esp8266-ping and import it through the Library Manager. By the way, the Pushingbox mail notifications service stopped working sometime after the article was published. That part is not essential, but if we find a good alternative, we’ll modify the code to use it. Power supply causes motorboating I have built the Surf Sound Simulator (November 2024; siliconchip.au/ Article/17018). I was very impressed with the quality of the PCB. Some of the components were not that easy to locate though in the UK, though. siliconchip.com.au The board starts okay with a wave sound, but after a few seconds, starts motorboating. The LED pulses and speeds up when the motorboating starts. I notice that there is a patch lead on the underside of the board shown in the article, although there is no mention of this in the text. Is this the cause of my problems? Failing that, I am at a bit of a loss how to diagnose the fault, save that I suspect a cap is charging up and triggering instability in IC2. The voltages on IC1 and IC2 are correct, as are (I believe) all the components. I have also checked for soldering problems. (R. T., Hove, UK) ● The wire under the PCB was only required for our prototype; the tracks were fixed on the final PCB so no modifications are required. We mentioned this in the photo caption, although perhaps it wasn’t 100% clear as it only mentioned some parts being added, and didn’t specifically address wires. The problem is almost certainly due to the plugpack you are using being unable to supply the current when the wave sound reaches a crescendo. Check that the supply rails are maintained without any voltage loss when the motorboating starts. Otherwise, the volume is set too high, causing feedback and motorboating. The circuitry around the power supply with 470μF capacitors and the isolation diode (D8) is there to prevent supply loading from causing motorboating. The 470μF capacitors were sufficient for our prototype. If one of these is faulty, it could cause such a problem. However, most likely switching to a supply that can deliver more current will resolve your problem. Reducing Surf Sound Simulator gain I built the Surf Sound Simulator (November 2024 issue; siliconchip. au/Article/17018) and am about to embark upon a second. It works and Australia's electronics magazine sounds realistic, but the sound volume is excessive. If the adjustment is turned up beyond about 1/3, the sound breaks up, ceases to be realistic and is, in any case, too loud. I have experimented and have found that reducing the 270kW feedback resistor between pins 7 and 6 of IC2b to 47kW gives a more reasonable level. Has anyone else reported this? (J. H, West Sussex, UK) ● We haven’t had the same problem with the volume, but your solution to change the resistance of the feedback resistor for IC2b is suitable. This would allow a wider volume control range to achieve the sound level you require. The higher volume in your prototype may be due to the loudspeaker being more efficient than the one we used in our prototype. Remote control interference query For some years now, I have been plagued with a problem in that during daylight hours, my driveway gate remote control will only work when operated within about 5cm from the receiver antenna. At the same time, both my car remotes become ineffective, and I have to lock or unlock both vehicles with a key, which in both cases results in the car alarm going off. I recall reading a letter from a reader concerning a similar problem to mine some time ago, but I cannot remember any solution given. It may be a coincidence, but these problems started when I had a rooftop solar system and inverter installed. Any information about possible fixes would be greatly appreciated. (I. H., Glossodia, NSW) ● Something nearby is likely continually transmitting on 330MHz and/ or 433MHz. To find it, you will need some sort of RF signal strength meter, direction finder or spectrum analyser. A cheap spectrum analyser with an antenna would be the best option. Walk around to see which direction February 2026  101 the strength increases, then keep going until you find the peak, and you’ll be near the transmitter. If it’s your inverter, there may be something wrong with it or its installation, as it should not be radiating that much interference. A spectrum analyser would help verify that; if it’s the inverter, the interference will be strongest when close to it. It could also be something like a solar-powered weather station. We’re pretty sure the letter you’re referring to is the one starting on page 82 of the December 2023 issue titled “The source of the interference”. That reader used an SDR dongle to track it down; they can be made to operate as a spectrum analysers with the right software and are considerably cheaper than a proper spectrum analyser. be a problem; did I install it the wrong way around? (B. P., Scottsdale, Tas.) ● Based on the photo you sent us, the shunt monitor IC is correctly orientated. As stated in the second article in the October issue, the pin 1 orientation marker on the INA282 can be a dot on the top face, a notch at the pin 1 end of the device, or a chamfer along the pin 1-4 edge of the package. For your device, it is the chamfer along the side. Regarding the 100W resistor burning out, check that the voltmeter is connected correctly and that IC1 and zener diode ZD1 are correctly orientated. That supplies the voltmeter, IC1 and ZD1. If you can’t find a problem with any of those, the alternative is that there could be a direct short circuit on the PCB somewhere (eg, component pads bridged with solder). Burnt resistor is not due to rotated IC Wrong PIC used for Tiny Xmas Tree I have just built your 30V 2A Mk2 Bench Supply (September & October 2023; siliconchip.au/Series/403). I have a short circuit somewhere, causing a 100W resistor to let the smoke out and the 500mA fuse in the rear power connector to also go pop. Trying to find said fault, I noticed that the INA282 shunt monitor IC has no dot and was wondering if this could I built the Tiny Xmas Tree (Nov 2020; siliconchip.au/Article/14636) and tried to program a blank PIC12F1572 chip soldered to the PCB with the HEX file coded 1611119A. I’m using a PICkit 2 programmer, which correctly identified the PIC12F1572 chip. Importing the hex file and writing to the chip brings up a warning that no configuration bits are set in the code. Amplifier quiescent current takes a while to stabilise I have a question regarding setting the bias on the Hummingbird Amplifier (December 2021; siliconchip.au/Article/15126). It seems to take a long time for the bias voltage to settle. I initially set the bias to obtain 10mV. After finishing the modules and switching them back on, I rechecked the bias. This measurement showed the bias to be way too low, at something like 3-4mV. So I reset the bias measuring at this point to 10mV. It seems to take probably five minutes for the bias to increase to the required 10mV. Does this seem correct to you? Otherwise the modules are performing quite nicely. (D. J., via email) ● It certainly can take a while for the bias of a power amplifier to settle. Importantly, we usually set the bias while the amplifier is warm and has been running for a while. It’s normal that it can be a bit under-biased at first switch-on, taking several minutes (or longer) to reach the desired bias level and stabilise. There are several time constants in the system. For example, the bias transistors will heat/cool at a different rate than the driver or output transistors. And the power dissipation of the output devices can take a while to heat up the substantial mass of metal in the heatsinks, the case etc to the steady-state value. So conditions will continue to change for some time after power-up until everything has mostly settled at its final temperature. In most Class-AB amplifiers during use, there are small, unavoidable shifts in the bias point as the overall dissipation changes over time, ambient temperature changes over time etc. The main things are that it’s varying around an average that’s close to the target, and it isn’t suffering from thermal runaway, where positive feedback causes the quiescent current to keep rising as it warms up until failure. Given the fact that some of the larger transistors are not on the main heatsink, the Hummingbird amplifier will not have as good thermal stability as an amplifier where all the transistors are on the heatsink. Still, it is stable enough, as has been provenSthrough extensive testing. 102 ilicon Chip Australia's electronics magazine Can you help resolve this problem? (G. F., London, UK) ● We hooked up a PIC12F1572 to a SNAP Programmer using MPLAB X IPE 6.25 and got the same message about the Config Bits not being set. The problem appears to be that the original Tree project from November 2019 used a PIC12F675 chip, so the 1611119A. HEX file is intended to be used with a PIC12F675. The November 2020 Ornaments article (which includes the Tree) switched to the newer PIC12F1572, so the file to use is “12F1572_16111191.hex” in the November 2020 download package. You can find the PIC12F1572 files in the “Software for 12F1572 variants” sub-folder. You’re lucky that you noticed that warning; it only comes about because the PIC12F675 puts the Config Bits at a different location to the PIC12F1572. If you had gone ahead and programmed it, it likely wouldn’t have worked, and we think it would have been a tricky error to track down. One SC200 amplifier module went bang In 2021, I built four SC200 Amplifiers (January-March 2017; siliconchip. au/Series/308), all of which had been working fine until last night. I heard a rustling sound coming from one of my speakers and determined it was one of the amp modules. I switched the amp off and took off the lid, then I switched it back on, only to have Q14 on one module burst into fire with lots of smoke! I’m at a bit of a loss to know the cause of this after so much time in use. It hasn’t done a lot of damage; Q13, Q14 and Q16 failed, plus the two 0.1W resistors associated with Q13 and Q14, and fuse F2. The driver transistor Q12 also blew up. I thought I’d look around the internet to see if there had been any updates to this design. There are a couple of threads where people are discussing adding Miller capacitors to stop oscillation. I was wondering if this is legitimately helpful. I’d like to know what caused the failure in the first place. Have you had any reports of failures like this? Having replaced the blown parts, I’m having a lot more trouble getting the bias voltage to settle on a specific 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 Dual Mini LED Dice August 2024 SMD LED Complete Kit SC6961: $17.50 TH LED Complete Kit SC6849: $17.50 siliconchip.au/Article/16418 Includes either 3mm through-hole or 1206sized SMD LEDs. Choice of either white or black PCB. CR2032 coin cell not included. 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 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 February 2026  103 voltage than I seem to remember. I’m not just talking about it moving about a bit with temperature; if I just tweak the 25-turn trimpot the tiniest bit, the voltage can jump 2mV up or down. After going back and forth every half hour or so today, I’ve fluked it and got it sitting at 4.6mV, which is OK, I guess. Could the trimpot be damaged by the DC component flowing through it? Seeing that this amp destroyed itself after a lot of use, and now this bias thing, I’m wondering if the two are connected. (T. B., Bumberrah, Vic.) ● It doesn’t seem common for SC200 amplifier modules to blow. From the feedback we’ve received, they are pretty reliable. There can be many causes of such a failure, including a faulty component or a solder joint that went bad (possibly due to thermal cycling). Having said that, the adjustment shouldn’t be that sensitive. We think the quiescent current adjustment Advertising Index Altronics.................................31-34 Blackmagic Design....................... 7 Dave Thompson........................ 103 DigiKey Electronics..................OBC Emona Instruments.................. IBC Jaycar............................. IFC, 11-14 Keith Rippon Kit Assembly....... 103 Lazer Security........................... 103 LD Electronics........................... 103 LEDsales................................... 103 Microchip Technology.................. 5 Mouser Electronics....................... 3 OurPCB Australia........................ 10 PCBWay......................................... 9 PMD Way................................... 103 SC Dual Mini LED Dice.............. 103 SC Battery Checker................... 100 SC Mains Sequencer.................. 61 SC Pico W BackPack.................. 83 Silicon Chip Back Issues........... 40 Silicon Chip Binders.................. 43 Silicon Chip Subscriptions........ 52 Silicon Chip Shop.................92-94 The Loudspeaker Kit.com............ 8 Wagner Electronics..................... 89 104 Silicon Chip trimpot has gone ‘scratchy’ for one reason or another (perhaps internal corrosion or it just wasn’t made properly in the first place). That certainly could explain why it blew up. If the bias was jumping around during operation, it could have gone into thermal runaway. We suggest you replace the trimpot. If you want to, you could make it a bit more robust by shunting the trimpot with a fixed resistor so that it’s operating near the bottom of its range. That way, even if the trimpot goes open circuit, the bias will only go up a little. Speed controller fails after motor stall I built two 230V/10A Speed Controllers for Universal Motors (Feb & Mar 2014; siliconchip.au/Series/195) from Jaycar KC5526 kits. I have been using them on my porting bench for quite some time. Unfortunately, both failed in the same way. The motor stalled at quite low RPM due to the cutter getting stuck, the fuse blew, but when I replaced it, the unit still didn’t work. Do you know which part most likely failed so I can replace it? I absolutely love the ability of these speed controllers to give good torque at low RPM. (S. G., Narre Warren, Vic) ● Most probably IGBT Q1 (STGW40N120KD) has gone open circuit. We can supply a small set of parts (siliconchip.com.au/Shop/20/2614) that includes the IGBT, diode, driver IC and NTC thermistor. It would be worth replacing the IGBT, diode and driver IC in case any of the others were also damaged. Jaycar has discontinued its kits for this project, but we can supply the PCB, programmed microcontroller and the set of parts mentioned above. They’re all listed at siliconchip.au/ Shop/?article=6120 Our latest mains motor speed controller is the Refined Full-Wave Motor Speed Controller (April 2021 issue; siliconchip.au/Article/14814). CDI with a wasted spark ignition system I’d like to use either the High-­ Performance CDI Ignition from September 1997 or December 2014 on my late 1970s/early 1980s Suzuki 4-­cylinder, 4-stroke motorbike engines. That vintage of Japanese engines had two coils for the four cylinders; one would fire cylinders 1 & 4 simultaneously, while the other fires cylinders 2 & 3 simultaneously, 180° of crankshaft rotation later. On the 1970s vintage machines, the 12V signal went direct from the crank sensor (points or hall effect) to the coils, whereas on the 1980s machines, a CDI ‘igniter’ unit was used (between the crank sensor and coils) to boost the spark energy. Can either of your CDI multi-spark units be adapted to work in this situation? It would require two inputs (from the Hall effect crank rotation sensors) and two outputs to each coil. (P. H., Seattle, WA, USA) ● You would need to build two of either version of the CDI unit, one to drive the coil for cylinders 1 & 4 and the other for cylinders 2 & 3. It may work if you just build a single high-voltage supply in one of the units and supply this voltage to the other unit as well. Then use the two separate trigger circuits for driving the SC individual coils. Errata and on-sale date for the next issue RGB LED Star Ornament, December 2025: in the circuit diagram, pin 2 of CON3 should connect to pin 11 of IC1 before the 330W resistor, rather than after. Power Electronics part 2, December 2025: in Fig.4 on page 32, the labels Zr and Zc are swapped in the high-pass filter circuit. Digital Preamplifier, October 2025: in Fig.5 on page 36, pin 3 of the ADAU1467CORE BOARD connector should be labelled ADC_BCLK to match pin 73 of IC18. Also, in the pinout for Q1, the pins should be labelled (left-toright) B, C & E and the tab is C. On p39, in Fig.8, pins 13 & 15 of IC15 go to IC17, not IC12. Finally, the designator CON8 for the 8×2-pin header that’s the alternative to CON19 is not to be confused for the RCA output connectors, CON8.1-CON8.4. Next Issue: the March 2026 issue is due on sale in newsagents by Tuesday, February 23rd. Expect postal delivery of subscription copies in Australia between February 20th and March 11th. Australia's electronics magazine siliconchip.com.au “Rigol Offer Australia’s Best Value Test Instruments” New 2026 Products Oscilloscopes New 12Bit Scopes New Up To 13GHz RIGOL DS-1000Z/E - FREE OPTIONS RIGOL DHO/MHO Series RIGOL DSO-8000/A 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 4600MHz to 13GHz, 4Ch 410GS/s to 40GS/s Real Time Sampling 4Up to 4Gpts Memory Depth FROM $ 649 FROM $ ex GST 684 FROM $ ex GST 13,191 Multimeters Function/Arbitrary Function Generators 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 604 FROM $ ex GST Spectrum Analysers 632 ex GST Real-Time Analysers New Up To 26.5GHz 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 26.5Hz 4Modes: Real Time, Swept, VSA, EMI, VNA 4Optional Tracking Generator ONLY $ 799 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