Fullscreen HTML5Med Res (??MB)HTML4

JULY 2024 ISSN 1030-2662 07 9 771030 266001 $ 50* NZ $1390 12 INC GST INC GST 180-230V DC Motor Speed Controller Repairable & Open-Source Electr nics 3 P1 Workman 1kW loudspeaker ary ivers Ann th 0 1 Refresh your workbench with our GREAT RANGE of essentials at the BEST VALUE. Here's just a small selection of our best selling workbench essentials to suit hobbyists and professionals alike. ALL THE REGULAR OSCILLOSCOPE FUNCTIONS IN A SMALL FORM FACTOR 2 CHANNELS SuperPro Gas Soldering Tool Kit SOLDER ANYTHING, ANYWHERE! DURABLE CASE WITH EXTRA TIP STORAGE Ideal for soldering, plastic cutting, heat shrinking, etc. • Includes two double flat tips, hot air blow, hot knife & hot air deflector tips • Up to 580°C temperature range • Up to 120 minutes run time ONLY 209 $ TS1328 GREAT ES. FEATUR GREAT PRICE! 20MHz USB Oscilloscope • High accuracy interface • Spectrum analyser (FFT) • 48M Sa/Sec sampling rate • 20mV/div sensitivity QC1929 DIGITAL MULTIMETER WITH TEMPERATURE HEAVY DUTY WIRE STRIPPER QM1323 $64.95 NOW $24.95 HALF PRICE • Autoranging • Cat III 600V • 10A AC or DC current • 40MΩ resistance • 100µF capacitance • 760°C temperature • K-type probe & case included ONLY 249 $ • Cutter, crimper & wire guide • Strips 10-24 AWG/0.13-6.0mm • Single handed operation TH1827 WAS $49.95 (Valid from 10.07.2024 - 21.07.2024) VOLTAGE AND CURRENT DISPLAY CONSTANT CURRENT & VOLTAGE IN A SLIMLINE FORM FACTOR PERFECT FOR COMPACT WORKSPACES ILLUMINATED DESKTOP MAGNIFIER • 100mm 3-dioptre glass lens • 30 bright LEDs • Mains powered QM3552 $86.95 Slimline Lab Power Supply • 0-16VDC <at> 0-5A (max.) 0-27VDC <at> 0-3A (max.) 0-36VDC <at> 0-2.2A (max.) • Up to 80W max. • Just 300L x 138H x 53Wmm 219 $ MP3842 Shop at Jaycar for your workbench essentials: • Soldering irons & accessories • Tools and service aids • Tool & storage cases • Fasteners and adhesives • Sprays and aerosols • Test equipment • 3D printers & accessories • Lab power supplies Explore our wide range of workbench essentials, in stock at over 115 stores and 130 resellers or on our website. Prices correct at time of publication but are subject to change. Jaycar reserves the right to change prices if and when required. ONLY www.jaycar.com.au 1800 022 888 Contents Vol.37, No.07 July 2024 14 Repairable Electronics Modern electronics tend to be hard to repair, either because parts are difficult to source or due to the design. Bucking this trend are devices made to be easy to fix, including modular parts and open-source hardware. By Dr David Maddison, VK3DSM Repairs & open-sourcing 39 The Raspberry Pi 5 Released last September, the Raspberry Pi 5 is their newest single-board computer. Nearly every aspect of the Pi 5 is a dramatic improvement, running at least twice as fast as the Raspberry Pi 4B. By Tim Blythman SBC review 58 Adding solar charging to a van We added solar panels (plus an inverter) to the roof of a Renault Kangoo ZE electric van. With the panels we have as much as an extra 50km of driving a day just from solar energy! By Roderick Boswell Electric vehicles 26 Automatic LQ Meter Page 26 Automatic LQ Meter Raspberry Pi 5 Review: Page 39 Adding solar charging to an electric van This all-in-one design combines two instruments into one, measuring inductance from 0.1μH to 999μH and Q (quality factor) from 10 to 300 with a test frequency from 100kHz to 90MHz. By Charles Kosina Test instrument project 44 180-230V DC Motor Speed Controller This Speed Controller is intended for high-voltage DC motors like those in lathes, treadmills, conveyor belts and more. It is rated for motors from 1A to 10A and has zero to full speed control plus speed regulation. By John Clarke Motor speed control project 54 New use for Mains Sequencer With some wiring adjustments and software updates, our Mains Power-Up Sequencer from February & March 2024 can be repurposed to cycle power to multiple devices for use with inverters & generators. By John Clarke Mains power control project 64 Jaycar-sponsored Mini Projects Mimick a lava lamp with this month’s first Mini Project. Or instead build a digital compass powered via a battery bank. Each project is designed so that anyone can build it. By Tim Blythman Mini projects 72 Workman 1kW Loudspeaker These genuine 1000W speakers deliver a tremendous amount of sound using a tool chest you can buy from Bunnings as the enclosure. The durable enclosure keeps it safe and makes it easy to transport. By Allan Linton-Smith Loudspeaker project Page 58 2 Editorial Viewpoint 5 Mailbag 80 Vintage Radio 86 Serviceman’s Log 94 Circuit Notebook 97 Subscriptions 98 Online Shop 100 Ask Silicon Chip 103 Market Centre 104 Advertising Index 104 Notes & Errata One-valve superhet radio by Fred Lever 1. JFET-based guitar preamp 2. Push-pull PWM Mosfet driver SILICON SILIC CHIP www.siliconchip.com.au Publisher/Editor Nicholas Vinen Technical Editor John Clarke – B.E.(Elec.) Technical Staff Jim Rowe – B.A., B.Sc. Bao Smith – B.Sc. Tim Blythman – B.E., B.Sc. Advertising Enquiries (02) 9939 3295 adverts<at>siliconchip.com.au Regular Contributors Allan Linton-Smith Dave Thompson David Maddison – B.App.Sc. (Hons 1), PhD, Grad.Dip.Entr.Innov. Geoff Graham Associate Professor Graham Parslow Dr Hugo Holden – B.H.B, MB.ChB., FRANZCO Ian Batty – M.Ed. Phil Prosser – B.Sc., B.E.(Elec.) Cartoonist Louis Decrevel loueee.com Founding Editor (retired) Leo Simpson – B.Bus., FAICD Silicon Chip is published 12 times a year by Silicon Chip Publications Pty Ltd. ACN 626 922 870. ABN 20 880 526 923. All material is copyright ©. No part of this publication may be reproduced without the written consent of the publisher. Subscription rates (Australia only) 6 issues (6 months): $70 12 issues (1 year): $127.50 24 issues (2 years): $240 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 Editorial office: Unit 1 (up ramp), 234 Harbord Rd, Brookvale, NSW 2100. Postal address: PO Box 194, Matraville, NSW 2036. Phone: (02) 9939 3295. ISSN: 1030-2662 Printing and Distribution: Editorial Viewpoint Jamieson ‘Jim’ Rowe is retiring By the time you read this, Jim will be just about finishing up his last work for Silicon Chip. He is undoubtedly keenly looking forward to a well-earned retirement starting in just a few days! I have had the pleasure of working with him since I started at Silicon Chip in early 2010. Of course, I read some of his articles before that. Those of you who started with Radio & Hobbies may remember him from as far back as March 1960, when he first appeared on the list of Technical Staff at that magazine. So he has been involved in electronics journalism for nearly 65 years at this stage – I guess you’d have to call that a pretty good run! Jim was the editor of Electronics Australia (the more recent name for Radio & Hobbies) for a time. He came to work for Silicon Chip soon after leaving that position. He has been writing articles and drawing virtually all of our diagrams since early 2001. If you’re interested, you can read the full story of his career in the July 2023 issue, when he wrote an article titled “Electronics in Australia; Jim Rowe’s time at RTV&H and Electronics Australia” (siliconchip. au/Article/15862). He famously developed the EDUC-8 Microcomputer; its design was published in EA from August 1974 to August 1975. It was the first kit computer design published in Australia and the second in the world (by one month). I would usually be concerned about the departure of someone with Jim’s experience. However, luckily, we have Jim Rowe in the Electronics someone else lined up who should be Australia office, 1989. able to pick up the work and deliver the sort of quality diagrams our readers are accustomed to. He happened to be retiring from his full-time job right around the same time Jim decided to call it quits, a truly fortuitous coincidence. We still have a few articles Jim has written that we haven’t had space to publish yet, so if you enjoy his articles, you will have a few more to look forward to over the coming months. I expect they will all be published by the end of this year. While he has designed projects in the past, lately, he has mainly been concentrating on reviewing various electronic modules. We plan to continue publishing that type of article, although, with Jim’s departure, they may not be as frequent as before. My thanks go to Jim for all the assistance he has provided us over the last 25-odd years. I hope he has some hobbies that he can enjoy in all his free time! by Nicholas Vinen 24-26 Lilian Fowler Pl, Marrickville 2204 2 Silicon Chip Australia's electronics magazine siliconchip.com.au The key to unrestricted access Explore millions of components for your next design Although your admission is barred to this historic imperial shrine, we welcome you to peruse millions of electronic components, from well over a thousand leading brands engineers know and trust. You don’t have to be royalty to venture inside and browse around. 03-9253-9999 australia<at>mouser.com 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”. Updates to Si4730/Si4732 based radio project Reader Andrzej Szupiluk built my SiLabs FM/AM/ SW Digital Radio design (July 2021 issue; siliconchip.au/­ Article/14926) using the Si4732 IC option and ran into some problems. I suspect most people didn’t bother with shortwave support and just used the Si4730 IC, so they would not have encountered them. One problem he reported is that, after switching from FM to AM and then back to FM, he has to rotate the volume control or reset the unit to get the sound back. I was able to fix that problem with a new version of the firmware (9.1), which is now available to download: siliconchip.au/ Shop/6/5873 There was another problem that I couldn’t fix: a very quiet “tic tic” sound at about half-second intervals on FM only. It does not occur with the Si4730. I have been unable to find the cause, although I designed another more complex receiver sometime later, which also resolved SSB, using the Si4732 that does not have this problem. One difference is that it used Arduino C code for control. I note that others have observed the “tic tic” problem as there are various sites on the internet that discuss it without any real solutions. Charles Kosina, Mooroolbark, Vic. Rest in peace, John Hill I have some bad news. My father, John Hill, has died. He wrote Silicon Chip’s Vintage Radio column from March 1988 to June 1998. He passed away at 7am on 22/05/2024 at Maryborough Hospital in Victoria after a short illness. I know that he got a lot of joy writing his Vintage Radio articles. Ian Hill, via email. from the web, so it can currently emulate a range of different CPU cores like the Z80, 6800, 6502 etc; ironically, not the SCMP yet; I will have to do that myself. I am building it mostly just to torture my children, who think they can use a computer. Still, it occurs to me that if I finish it, you might be interested in publishing it, to torture a much wider range of people! Alan Williams, Old Noarlunga, SA. Hints for improving digital TV reception I just read the letter to Mailbag by Bruce Pierson in Dundathu, Qld, titled “Frustration Over Bad TV Reception”. Over a decade ago, Silicon Chip ran a brilliant series clarifying the introduction of digital TV and everything to do with antennas, installations and the digital cliff. I think it ran for two issues. Editor’s Note: “How To Get Into Digital TV”, March & April 2008, by Alan Hughes – siliconchip.au/Series/49 I used those articles to successfully re-cable and install a new antenna on my own home (which is still working well). I also did the same for a many-roomed hotel and an aged-care facility. I suggest that you refer the reader to those back issues. Also, I have fixed similar faults to those the reader is experiencing by replacing RG59 coax with RG6 (with multi-shielding), fitting proper F-type connectors using a suitable crimping tool and replacing the wall plates with F-type connectors. Your articles explain why proper shielding is critical. Reader developing Pico-powered Miniscamp I fell about laughing when I saw the story of the guy restoring the Miniscamp computer from Electronics Australia (Serviceman’s Log, May 2024). Mine is in significantly worse repair, and looking at it now makes me wonder how anything I did as a kid ever worked. I took a different approach to the problem and started making a new one, as shown in the adjacent photo. The footprint on the left of the PCB is for a Raspberry Pi Pico. There are a few other enhancements: the Neopixels will give colour LEDs, the LM386 drives a speaker and there is some I2C NVRAM for program storage. The RUN/ HALT switch now has a third position, ‘MODE’, to set how it will treat the NVRAM. I haven’t had time to work on it lately, so the software is nowhere near finished. I borrowed some core libraries siliconchip.com.au Australia's electronics magazine July 2024  5 I had one job where a tree between the antenna and the transmitter created problems – we moved the aerial, so it was not necessary to remove the tree. Very long runs also require larger diameter coax. There is a possibility that equipment operating during the day may be interfering with the reader’s signal, and simply replacing the coax with multi-shielded RG6 will restore 24-hour reception. I hope this helps. Digital TV is not as forgiving as analog TV was. Jacob Westerhoff, via email. European countries switching to HD TV only Netherlands, Germany, Austria, France and Spain are transmitting only high-definition TV using DVB-T2 so that they can have more HD programs per broadcaster. They do not transmit programs using standard definition (non-HD) anymore. Finland, a country of 5.6 million people, will be switching off DVB-T in SD next year, leaving DVB-T2 in HD only. France and Spain are already transmitting Ultra High Definition (UHD) or 4K on DVB-T2. The amount of data required for UHD is much higher, so HEVC compression is required to fit the signal into a TV or internet channel. HEVC also works on HD signals, making a UHD program and multiple HD programs possible on one transmission channel. These broadcasters are taking on the video-on-demand companies! What are we doing? Australia transmits all TV using DVB-T. We tested DVB-T2 in 2019! Electronics retailers in Australia sell set-top boxes for DVB-T2/MPEG4 for $60. They won’t receive UHD or HEVC video compression, though, which is what the streaming companies use. Alan Hughes, Hamersley, WA. I’m still using floppy disks! Referring to the May 2024 mailbag item titled “25-yearold disks are still operational”, I was heartened to learn that I’m in good company. I also continue to use 5¼-inch floppies, many of which have survived from around 1982. That’s 42 years! I also use 3½-inch floppies of similar age, USB flash drives, CDs, DVDs, hard disks and SSDs, so I can’t be accused of being entirely a dinosaur or troglodyte. My reason for persisting with floppies is that they are so darn reliable and are ideal for use with any form of document that is regularly updated (such as rolling tallies, tax calculation records etc). I put their reliability down to the low storage density and perhaps the large physical size of each area devoted to recording each one or zero digit on the disk. Because I religiously back up each file, the worst that can happen is that 1.2MB of data needs to be replaced from the backup. By contrast, the loss of data on other forms of storage media usually runs to many files and gigabytes of data. All that being said, it goes hard against my grain to bin anything that continues to function or can be repaired. Floppy disks and drives are no exception. George Clauscen, East Oakleigh, Vic. More on converting DD floppies to HD I noticed that on p22 of February’s issue, the information in the inserted panel on “floppy disk hacks” was incomplete, possibly for space reasons. 6 Silicon Chip Something that I think is relevant is that ‘converting’ a disk from 360kB to 1.2MB or from 720kB to 1.44MB was fraught with danger to your data. Some disks could be successfully converted in this way, but not all, and sometimes, an apparent success would turn out to be a failure. The problem relates to the method used to increase the capacity of the disks. Because the bits of data were closer together to squeeze more data into the same space, the magnetic coating on a true high-density (HD) disk was manufactured to be less affected by magnetic fields from adjacent bits, which in turn meant that the write head needed to produce a stronger magnetic field to write to high-density disks. While in practice, many double-density (DD) disks seemed to work just fine as high-density disks, there was always a danger that either the intensity of the signal from writing would interfere with adjacent data, or that these disks would lose their data when stored for a long time due to the adjacent magnetic fields interfering with each other. It may also be worthwhile to mention a pair of programs commonly referred to as FDFORMAT, which allowed a high-density 1.44MB disk to be formatted to around 1.72MB through several techniques (and a similarly proportioned increase for 1.2MB disks). Some formats required the accompanying ‘terminate and stay resident’ (TSR) program FDREAD, but other formats were compatible with the default PC format. I found the information about core rope memory interesting too, although from my understanding of how it works, you could have as many bits per core as the core had space for address wires. This was not limited to 192 locations, as stated in the article, although 192 may have been the largest CRM ever made due to the physical size becoming too cumbersome. I find some old technologies quite interesting, and the ingenious methods that were found to ‘get more out of less’ are an important lesson in persistence and in how you sometimes need to ‘think sideways’ to get the job done. Jonathan Waller, Bairnsdale, Vic. Comment: FDFORMAT was a very handy utility. While we don’t recall converting DD disks to HD, we converted many single-sided floppies to double-sided using a notch punch without any problems. Odd problems with Advanced Test Tweezers The Advanced SMD Test Tweezers (February & March 2023; siliconchip.au/Series/396) I built worked but with much worse accuracy than expected based on the specifications in the article. I microscope visually re-checked all the solder joints for open or short circuits, but no faults were evident, and all joints looked well-soldered. When preparing for reflow soldering, I found I had been a little over-zealous with paste quantities – so, just in case, I got to work with solder wick – although nothing abnormal showed up. I had also previously removed a few solder bridges on the microcontroller, so I added a little solder to all the pins just to be sure. While more easily accessible, I re-measured the 1kW, 10kW, and 100kW resistors – in a couple of cases, I made a one least significant digit change to the calibration. Most notably, I removed the out-of-spec Vcap capacitor (it measured 8.5μF instead of 10μF ±10%). I had a few Australia's electronics magazine siliconchip.com.au Introducing ATEM Mini Pro The compact television studio that lets you create presentation videos and live streams! Now you don’t need to use a webcam for important presentations or workshops. ATEM Mini is a tiny video switcher that’s similar to the professional gear broadcasters use to create television shows! Simply plug in multiple cameras and a computer for your slides, then cut between them at the push of a button! It even has a built in streaming engine for live streaming to YouTube! Live Stream to a Global Audience! Easy to Learn and Use! Includes Free ATEM Software Control Panel There’s never been a solution that’s professional but also easy to use. Simply press ATEM Mini is a full broadcast television switcher, so it has hidden power that’s any of the input buttons on the front panel to cut between video sources. You can unlocked using the free ATEM Software Control app. This means if you want to select from exciting transitions such as dissolve, or more dramatic effects such go further, you can start using features such as chroma keying for green screens, as dip to color, DVE squeeze and DVE push. You can even add a DVE for picture media players for graphics and the multiview for monitoring all cameras on a in picture effects with customized graphics. single monitor. There’s even a professional audio mixer! Use Any Software that Supports a USB Webcam! You can use any video software with ATEM Mini Pro because the USB connection will emulate a webcam! That guarantees full compatibility with any video software and in full resolution 1080HD quality. Imagine giving a presentation on your latest research from a laboratory to software such as Zoom, Microsoft Teams, ATEM Mini Pro has a built in hardware streaming engine for live streaming to a global audience! That means you can live stream lectures or educational workshops direct to scientists all over the world in better video quality with smoother motion. Streaming uses the Ethernet connection to the internet, or you can even connect a smartphone to use mobile data! ATEM Mini Pro $495 Skype or WebEx! www.blackmagicdesign.com/au Learn More! 1206-sized 10V 10uF capacitors and found I could (just) fit it to the 0805-size pad provided. I read in the microcontroller specification sheet that Vcap is required to stabilise the core 1.8V voltage regulator. I reasoned that this capacitor may be a critical item in terms of its stability. On re-assembling the Tweezers, I found that the reading stabilities and accuracies were vastly improved, although not perfect. I could calibrate the band-gap voltage reference much better (with less severe fluttering around the estimated voltage). However, I found that using the (measured) coin cell as a reference did not work very well at all. Instead, I compared the METER reading of a known 6.500V reference and then scaled the band-gap reading according to that ratio. Eureka! – accuracies improved further – I set a 20.00V supply up, and the METER reading was +20.0V or -20.0V depending on the orientation. Now the probes read 0.00V open circuit, flashing occasionally to -0.01V, and the reading remains basically the same when closing the pins. However, whilst stability has improved greatly, reading a measured 9μF capacitor flickers between about 7.13μF and 8.11μF, a 10-20% error. A 469.6W resistor reads as 456W, a 3% error. A 999W resistor reads as 973W, just under 3% error. A 99.4kW resistor measures as 99.2kW, so the high end works better, as does the reading of a 101nF cap at 105nF. As an afterthought, I measured another cap from the batch from which I selected the 1206 10μF capacitor using the March 2004 ESR Meter Mk-II (March 2004). After zeroing out this meter, I read an ESR of 0.16W (if the reading is accurate on a ceramic capacitor). This figure appears quite low but is perhaps an order or two of magnitude higher than recommended in the data sheet. I don’t know the ESR meter measurement frequency or the microcontroller’s regulator details. Unfortunately, I accidentally lost the 0805 10μF capacitor when trying to remove it from the tip of the iron – I would have liked to know the ESR of that part. I suspect my residual inaccuracies and poor stability on some readings may result from a sub-optimal Vcap selection. My old Tweezers have considerably better accuracy on all resistor and capacitance readings. Ian Thompson, Duncraig, WA. Comments: it is good that changing the capacitor has improved its operation. We don’t recall running into problems like this with any of our prototypes, nor have any been reported by other constructors to date despite many kits being built. We have heard of 16-bit and 32-bit PICs failing to operate due to the Vcap filter capacitor, even when it is within the specified ranges of capacitance and ESR. The PIC24FJ256GA705 family data sheet specifies a maximum ESR of 5W and says that even tantalum capacitors (which generally have a much higher ESR than ceramics) are suitable, so your capacitor should be well within specifications. Increasing the capacitance can help, so using a 22μF 10V M2012/0805 capacitor like the Samsung CL21A226MPQNNNE may provide some benefits. The 10μF M2012/0805 ceramic capacitor we supply in kits is the Samsung CL21A106KOQNNNE purchased from a major supplier; it should be between 9μF and 11μF. The fact that yours apparently wasn’t is quite concerning; it may have been a dud. The graph at lower left shows the ESR specification for that part, so it should be very suitable for filtering Vcap, assuming it meets its specs. We don’t think there should be a direct correlation between the Vcap capacitor quality and the analog readings. However, jitter in the microcontroller’s clocks from a poorly regulated Vcore rail could affect just about anything. The ADC’s accuracy will depend on equal timing in its steps as it performs its conversion, which could affect the stability of those readings. We wonder if extra capacitance on the main (3V) supply rail could help. If you have another similar 10μF capacitor, you could try adding it across pins 2 and 3 of CON1. A large parts collection to give away At the age of 88, I’ve been reading your magazine and its predecessors for about 70 years and have built a good many projects. Radio & Hobbies’ first FM tuner still works with the original valves, and my main audio amp – a much later design – is one of yours. However, these days, most articles in the magazine, apart from the ‘vintage’ articles, go over my head. Through the decades, I’ve hoarded a lot of spare parts and devices: valve sockets, transformers, heatsinks, diodes, LEDs (handfuls), transistors, plugpacks, low-voltage motors, resistors, capacitors etc. I’d be happy to give away most of them to any genuine hobbyist(s) prepared to drive to Dora Creek, on the NSW Central Coast, about an hour on the freeway from Sydney. After leaving me a few representative bits and pieces, they’d be welcome to most of the collection, including the magazines from the 1950s on (though with some sizeable gaps in the early years). For further info, phone me on (02) 4973 4544 after 10am. Brian Wallace, Dora Creek, NSW. More praise for RTV&H oscilloscope design Ian Batty’s article on the cathode-ray oscilloscope (CRO) designed by Jamieson ‘Jim’ Rowe in Radio, TV & Hobbies has brought back a host of memories (May 2024 issue; siliconchip.au/Article/16259). I built one of these units in 1964, when I was 18 years old. I bought the complete set of parts from a Sergeant who had been posted to a desk in Canberra. He didn’t think he would have the time or facilities to further this project. From memory, I paid £50 for it. 8 Silicon Chip Australia's electronics magazine siliconchip.com.au Make amazing projects with our microcontrollers & mini computers. We have an incredible line-up of micros for beginners, hobbyists and professionals. EXPANDABLE WITH SHIELDS, SENSORS & MODULES BEST SELLER MORE PROCESSING POWER & MEMORY, WI-FI, BLUETOOTH AND MUCH MORE A GREAT WAY TO LEARN HOW MICRO'S WORK MICRO:BIT V2 ENTRY-LEVEL MICROCONTROLLER ONLY 59 $ 95 XC4324 • EASY TO USE DRAG-AND-DROP CODING • DISPLAY AND SENSORS ONBOARD • BATTERIES INCLUDED 4495 $ ARDUINO®-COMPATIBLE UNO R3 MICROCONTROLLER ONLY XC4410 • 14 DIGITAL & 6 ANALOG I/O PINS • COMPATIBLE WITH ARDUINO® SHIELDS • USB OR DC POWERED HIGHER PROCESSING POWER, SUITABLE FOR MORE CHALLENGING TASKS ONLY 5195 $ XC9211 ARDUINO® UNO WIFI R4 MICROCONTROLLER • WI-FI 4 & BLUETOOTH® 5 • BRIGHT 12X8 RED LED MATRIX • HID (HUMAN INTERFACE DEVICE) SUPPORT • 12-BIT DAC, CAN BUS AND OP AMP LARGER COMMUNITY, GREAT FOR EARLY ENTRANTS SAVE $20 WAS $119 (Valid from 10.07.2024 - 21.07.2024) ROCK 4C PLUS 4GB MICROPROCESSOR NOW 99 $ HIGH-PERFORMING SINGLE-BOARD COMPUTER XC9300 • DUAL HDMI (4KP60 & 2KP60) OUTPUTS • USB 2.0 & USB 3.0 PORTS • GIGABIT ETHERNET & WI-FI 5 WITH EXTERNAL ANTENNA • BLUETOOTH® 5 • 4GB RAM Explore our wide range of development boards, shields and modules, POWERFUL SINGLE-BOARD COMPUTER • DUAL HDMI (2 X 4KP60) OUTPUTS • USB 2.0 & USB 3.0 PORTS • GIGABIT ETHERNET & 2.4GHZ/5GHZ WI-FI • BLUETOOTH® 5 • 4GB RAM in stock on our website, or at over 115 stores or 130 resellers nationwide. jaycar.com.au Stock may be limited and offer available only whilst stock lasts. No rain checks. Savings on Original RRP (ORRP). ONLY 129 $ RASPBERRY PI 4B MICROPROCESSOR XC9100 1800 022 888 I had never built such an advanced piece of equipment. The most complex construction I had finished was a six-valve superhet wireless for my Dad. I looked at the complexity and, with all the bravado of youth, I set out to build it. At the time, I was in the Air Force at Townsville, with 10 Squadron. I had started as an Air Force Radio Apprentice (where I met Ian Batty for the first time) and was working in the Radio Laboratory on the base. I approached the Warrant Officer who ran the Lab, and in outlining my proposed project, I asked if I could borrow one of the spare benches after hours to build my CRO. He was most impressed with what I proposed to do and gave me his blessing. I started construction. Like eating an elephant, it started with the first bite. I installed the power supply circuitry first and ensured all voltages were correct. Then, it was on to the timebase and signal circuitry. As I placed each component in its position, I circled the component on the circuit diagram. Likewise, I put two diagonal lines through every connecting wire on the circuit diagram when it was in place. I completed the unit in two-and-a-half weeks. I fired up the CRO without valves first and looked for signs of distress. None appeared. I set about calibrating the unit as set out in the RTV&H article. After some twiddling, everything seemed to be okay. I had a working CRO! My project had aroused quite some interest. The Lab had the use of three or four Tektronix CROs for servicing various pieces of equipment. The Warrant Officer asked me to leave it in the Lab. Several senior members of the Lab staff asked to play with it, and several of them said it was much more intuitive to use than the very complicated Tektronix CROs available to us. High praise indeed! The CRO was used by many of the staff, but it went with me when I was posted out. I was on leave in my hometown of Blackall, central western Queensland. I showed it to the local electrical/radio repair man in town, and he asked to buy it for £250. That was a lot of money in those days! So I sold it. I have kicked myself ever since. I wish I still had it. It was a spectacular piece of 1960s design. It is 60+ years old and still a valuable service unit. It took a longish time to build it, but it was not difficult if you took it section by section. Today, if I found one, I would buy it, for sure. Philip Fitzherbert, Mentone, Vic. Cause of Multi-Spark CDI failure I am a long-retired RF engineer who stopped subscribing to the magazine for budgetary reasons. Probably my last project was to build your CDI unit (from December 2014 & January 2015; siliconchip.au/Series/279) for my classic MG Midget. All went together very nicely, using the best-quality new parts. Bench testing at 300V went well. I then connected its output to a vehicle coil, with the secondary connected to a spark plug. I set the input pulse generator (HP 3311A function generator) for about 10 pulses per second, which I don’t think was excessive. The unit started up nicely and gave good sparks across the 25-thou gap (0.635mm). It ran for about 30 seconds, then it stopped. I traced the fault to a failed Mosfet driver, IC3 (L6571). 10 Silicon Chip Australia's electronics magazine siliconchip.com.au Measuring tools for now and the future DIGITAL READOUT 7” Colour LCD Screen Colour Display Multiple Pre-Set Colours ZERO Programmable Up To 3 Axis One Touch Axis Zero Keys SCAN HERE FOR MORE INFORMATION Multi Language Menu 2-Year Warranty 352 (Q8500) $ 120mm Compact Linear Scale - MX-500-120/5U Touch Point Sensor TPS-20 NEW RELEASE Dial Bore Gauge 34-226 Outside Caliper Gauge 33-239 • Compact Scale • Glass scale with 5µm resolution • 3m connection cable • Accuracy within 0.005mm • Ø10mm hardened & ground ball end • LED Light & Beeper Sensor • 50-160mm range • 0.01mm accuracy • 150mm readable depth • Self aligning mechanism • 12 interchangeable anvils • LCD display Metric/Imperial conversion • Accuracy 0.03mm • Auto power off 198 (Q8510) $ 93.50 (M690) $ 209 (Q226) $ $ 363 (Q239) Digital Caliper - M740 Digital Indicator - 34-2205 Metric Outside Micrometers - 20-111 • 3 Modes of measurement IP54 • Absolute & incremental functionality • 4-way measuring • 12.5mm/0.5" range • Zero setting at any position • Metric/Imperial system • 55mm dial face • Data output interface • 3 piece set 0-75mm range • 1 x 0-25mm • 1 x 25-50mm • 1 x 50-75mm • Carbide tipped anvils • 0.01mm accuracy 44 (M740) $ 132 (Q2205) $ 165 (Q111) View and purchase these items online: www.machineryhouse.com.au/SIC2406 SYDNEY BRISBANE MELBOURNE (03) 9212 4422 (08) 9373 9999 1/2 Windsor Rd, Northmead 625 Boundary Rd, Coopers Plains 4 Abbotts Rd, Dandenong 11 Valentine St, (02) 9890 9111 (07) 3715 2200 Specifications and prices are subject to change without notification. All prices include GST and vild until 27.07.24 PERTH Kewdale ADELAIDEY OPENING JUL 05_SC_270624 $ The voltage on pin 1 was down to about 2V. I reckon the internal zener diode failed; it should be 15.6V. I couldn’t see why the IC would have failed, so I purchased a replacement. The same thing happened again, so now I am foxed. The data sheet implies this device should be pretty robust. The MOV should protect against anything nasty coming back from the ignition coil (is it enough? I don’t know). Meanwhile, the inverter circuit keeps on plugging away, happily producing 300V DC with 168mA current drawn from the 13V lab supply. The device originally listed in the parts list was the L6571AD, which is no longer manufactured. The L6571BD is still in production, and the only difference is a dead time of 0.72μs instead of 1.25μs. The only thing I can think of is that the coil is faulty. David Allen, Tauranga, New Zealand. Comment: We agree that the coil seems faulty. It must be internally arcing over, causing very high voltages to feed back to the driver and damaging the L6571. We suggested you stick with the L6571AD because that is what the original design called for and what we tested it with. We suspect the L6571BD’s dead time is too short and will cause the Mosfets Q3 & Q4 to burn out even with a good coil. You may not be seeing that on your bench tests because your mains power supply can’t deliver as much current as a large lead-acid battery or vehicle alternator can. While the L6571AD is no longer manufactured, some new-old stock (NOS) is still available online. For example, see www.aliexpress.com/item/1005005084636324.html The true cost of energy Coal, oil and gas are fossilised solar energy from millions of years ago. Hydro, wave and wind effects come from current solar radiation. Nuclear, geothermal and tidal energy come from the effects of the Big Bang. They are all free and there to be exploited by all life forms on our planet Earth that are able. On the 2nd of May 2024, the CIS released a report (siliconchip.au/link/abwc) that suggests the CSIRO has cherry-picked data to give a false impression of the financial viability of renewable energy systems as advice to the Australian Federal Government. While the CIS report makes good points, both the CIS and CSIRO base their analysis on the reasoning of economics, more precisely carbon economics, with little regard to the performance of renewables compared to the established fossil fuel system. The latter currently enables eight billion and counting people to live on this planet. After over thirty years of R&D and billions of dollars in subsidies, renewable energy system technology has yet to reach a stage where it can power a heavy industrial society reliably without being supported significantly by fossil fuel or nuclear energy. I refer to the letters in the September 2023 issue by Keith Anderson, George Ramsay and Dr Ken Moxham. These letters relate to the August 2023 article on differing electricity prices and the April 2023 editorial based on a Dick Smith statement on energy, a Dick Smith letter on the practicability of pumped hydro and my own letter in the July 2023 issue titled “Honesty in energy generation costs”. I believe Dr Moxham’s letter hit the correct note in that the tariff prices have no resemblance to the actual cost of generation. My letter in the July 2023 issue was centred 12 Silicon Chip around my undergraduate major paper, which illustrated that exact reality. As stated in my letter, I basically abandoned my paper as invalid by drawing no conclusion. I went further to state that we cannot evaluate the true cost of energy to society in monetary terms. I suggested that a fully science-based method could be possible, though. It is despatchable, technology-derived energy that drives our industrial society. Similarly, bioenergy powers people and all life on Earth; plants use energy from the photosynthesis of solar radiation. Therefore, the market-born monetary value placed on both bioenergy (food) and despatchable energy is only a method of energy allocation within our society. Subsidies reallocate energy for specifically politically desired outcomes. It is invalid data for determining the true value of energy supporting our society. In effect, we need to change from thinking in terms of $ per energy unit to energy per energy unit for an accurate analysis of our energy systems. There has been very little mention of energy conservation in Australia since 2008. The then-PM Kevin Rudd’s federal government created an intensive home ceiling insulation initiative as an economic stimulus to counter the global financial crisis at that time. It is an example where energy physics, economics and politics were in direct confrontation. The ongoing energy savings just from that short, intensive insulation program were considerable. The savings to consumers translated to a significant reduction in income for the power generation entities. If the program had been continued, it could have been financially catastrophic for power-generating companies. It also opens the question: would subsidy money be more effectively spent thermally upgrading infrastructure than trying to change the energy generation systems? Our expanding economic system has depended on an increase in available energy. Our population now depends on vast amounts of energy from our power systems embedded in machinery and industrial chemical products used to boost agricultural production massively. If the switch to renewables significantly impacts that, it will severely affect our society. Australia’s energy system is on the verge of mayhem. Our leading political decision-makers are currently in a state of chaotic naivety. Irrational economic reports overrule common sense. Ideologues ignore those who have a deeper understanding of energy and energy systems. Dick Smith has just released a very precise question-­ and-answer video on YouTube pleading for the lifting of the ban on nuclear energy. This excellent video expresses many of my above points more simply and clearly (https:// youtu.be/-Rm3Zfwd6dk). Silicon Chip readers with similar reasoned understanding should, in the public interest, also lobby their local political representatives. As I write this, the NSW government has come into line with most other states by extending the life of a large coalfired power station based on AEMO advice. Maybe there is still a spark of reality in the Australian energy system. These types of moves have already been made by the UK and European countries, including speeding up the construction of nuclear plants to compensate for the frailty of renewable generators. Kelvin Jones, Tasmania. SC Australia's electronics magazine siliconchip.com.au siliconchip.com.au Australia's electronics magazine July 2024  13 Repairable & Open-Source Electr nics By Dr David Maddison, VK3DSM Compared to older devices, anything with modern electronics tends to be challenging to repair. Replacement parts can be difficult to get, firmware may be unavailable and sometimes devices are designed to prevent part swapping! Bucking that trend are devices intended to be easy to repair, often by the user, including modular electronics and even open-source Image source: https://github.com/FrameworkComputer/Framework-Laptop-13 – CC-BY-4.0 hardware. T his article will cover two related topics: electronics designed to be easily repairable/upgradeable, and open-source electronics. They are related because open-source electronic devices are, by their nature, repairable and upgradeable. That’s because all the documentation, like circuit diagrams, PCB layouts, part lists, part specifications and mechanical drawings are made public. Open source is a software and hardware design model for producing software and/or hardware with an open, flexible, future-proof design that is frequently free or low in cost. 14 Silicon Chip Older devices tended to be much more repairable than modern ones. They had to be, to some extent, because they were less reliable. For example, valve radios were generally designed to be repairable, as were early transistor radios. A modern radio is more reliable and cheaper but probably tricky (if not impossible) to fix if it goes wrong. In today’s society, replacing a device is often considered cheaper than repairing it (although that is usually not true; it’s more due to laziness). Some modern devices such as laptops and phones, including famous Australia's electronics magazine brand ones, are purposefully made difficult to repair by methods like manufacturer part serialisation or restrictions on the availability of spare parts, meaning that a device often needs to be discarded just because of a tiny fault. Open-source and repairable devices attempt to address these and other deficiencies. A device doesn’t need to be open source to be repairable, but if it is open source, that means at least you will have access to all the information required for repair. It may even mean you can fabricate replacement parts if they are no longer available. siliconchip.com.au Open source Open source originated as a software design model, which these days is called free and open-source software (FOSS). With FOSS, the source code is made publicly available so anyone can inspect or modify it. It is (generally!) developed with a spirit of community cooperation and accessibility because it is free of charge (although donations are often welcome). While there is a lot of closed-source software, much of which we rely on, it has several disadvantages. One is that no one except the manufacturer knows exactly what the code does. The original programmers might have retired, so nobody might know what’s in it! That means many bugs and security problems can be lurking within. Of course, FOSS software can also have bugs and security problems, but generally, they are more readily found (by examining the source code). Theoretically, anyone can fix them, even if the original authors are no longer working on the project. FOSS’s advantages include being available at no cost, with decent privacy and security due to its open nature. Disadvantages include little-­ to-no technical support (although some projects provide free or paid support), and no guaranteed development timelines or updates (with some exceptions, eg, Ubuntu Linux releases major updates every six months). Another motivation for open-source software is that some people don’t want the uncertainty of commercial products. There have been instances where they were disabled or made useless after a certain date, had unexpected price jumps, failed to support older versions or were given no support for newer operating systems. Take as an example the (formerly?) popular computer virtualisation software VMWare. They were probably the biggest vendor in their market, but after being purchased by Broadcom in late 2023, they jacked up the licensing costs so much that many customers jumped ship or are looking to move away from their platform ASAP. Many of their (possibly former) customers have learned a costly lesson about trusting software vendors. A further advantage of FOSS is that obsolete hardware is often supported. For example, some versions of Linux can still run on a 386 processor (released in October 1985). siliconchip.com.au What is and what isn’t open hardware? There is a DIN standard that itself comprises free and open source documents (unlike most standards) to strictly define the meaning of open hardware. It is called “Open Hardware Standard – Requirements for technical documentation and community-based assessment”, and you can download it from https://gitlab.com/OSEGermany/OHS-3105 It comprises DIN SPEC 3105-1 (“Requirements for technical documentation”) and DIN SPEC 3105-2 (“Community-based assessment”). DIN is the German ISO (International Standards Organisation) member body, the Deutsches Institut für Normung (‘German Institute for Standardisation’). Well-known examples of FOSS projects include Linux, LibreOffice, Open­ Office, Mozilla Firefox and Thunderbird, Audacity audio editing software, GIMP image manipulation software and the VLC Media Player. Opensource software can be especially valuable for individuals or organisations on a budget. Open-source hardware Recently, the FOSS concept has been extended to hardware. OpenSource Hardware (OSH) or Free and Open-Source Hardware (FOSH) can include electronics, computers, mechatronics, 3D printers, silicon chip (integrated circuit) designs, radios, appliances, vehicles and many other devices. It may be in the form of non-electronic hardware components or electronic assemblies. With open-source hardware, there is usually some type of digital representation of parts that can be reproduced. For example, PCB CAD files, 3D printing files or other types of CAD files (eg, AutoCAD). That means anybody can build, repair, modify or improve these devices, or contribute to their development. Open-source hardware can keep old computers or gaming consoles usable, can be used to upgrade old cars or even new ones, or make new parts that would otherwise be unavailable due to obsolescence or because manufacturers are no longer interested in supplying them (or never were). An open-source solution is generally more repairable than closedsource equivalents and may be more economical. In the case of non-opensource (closed-source) hardware, there is typically no guarantee of spare parts availability or upgradeability into the future unless mandated by legislation (and even then, you may be out of luck). Australia's electronics magazine Some well-known examples of FOSH are Arduino, Raspberry Pi Pico, ArduPilot and Micro:bit. FOSS and FOSH have evolved to embody a set of principles known as “the open source way”: transparency, collaboration, release early and often, inclusive meritocracy and community (https://opensource.com/opensource-way). Some ideas are successful, while others are not. In researching this article, we encountered numerous opensource projects that started with great hopes but failed for various reasons. Others are success stories. Also, some designs started as opensource but later became closed-source, such as the Luka EV (mwmotors.cz/ luka-ev). Like any human endeavour, it may be that certain ‘personalities’ dominate a project, and if they lose interest, the project could fail. The more people involved in an opensource project, the less likely that is to happen. Smartphones Repairing modern phones can be very difficult or even impossible. For a start, they are often glued together. The parts can also be ‘serialised’, meaning the software will refuse to work with replacement parts, even those identical to the ones originally in it (eg, swapped from another identical phone), as shown in the video at https://youtu.be/FY7DtKMBxBw Fairphone www.fairphone.com Like many smartphones, the Fairphone is based on the open-source Android software. Its hardware is not open-source, but the phone is highly modular, and parts can be replaced or upgraded (see July 2024  15 Fig.1: the components of the modular Fairphone 5 smartphone. Source: www. flickr.com/photos/fairphone/53152347626 Figs.2 & 3: the Framework 16 laptop. Swapping expansion bays takes just a few seconds, making it an easy upgrade. It’s even possible to use the GPU module at home but switch to the smaller and lighter version of the laptop for travel! Other parts of the laptop are easy to replace such as the battery, display, internal SSD, speaker and more. 16 Silicon Chip Australia's electronics magazine Fig.1). Fairphone also guarantees compatibility with five Android version upgrades, meaning 8-10 years of updates. Many other phones also have replaceable parts; however, the Fairphone is designed to be easy for the user to disassemble and repair (the phone is not glued shut). You can even replace the battery easily! They sell many spare parts at reasonable prices, and it comes with a five-year warranty. Fairphones can run various Android operating system versions and forks, including CalyxOS, DivestOS, /e/, iodeOS, LineageOS and Ubuntu Touch. For more information, see our article on Privacy Phones (June 2024; siliconchip.au/Article/16280). Fairphone also make headphones and earbuds, which are also designed to be repairable; see https://shop.­ fairphone.com/audio For further details, you can watch the videos titled “The easiest camera repair ever? Fairphone 5” at https:// youtu.be/69-I46FSB98 and “Replacing the Display | Fairphone 5” at https:// youtu.be/CTlUOw1b5wo Based on reviews we have read and seen of the Fairphone 5, besides a few glitches, it seems like a pretty good smartphone. It is a little chunkier and more expensive than other phones with similar specifications, but not by a huge margin. The processor, cameras and OLED screen get pretty good scores, and the battery life is good, even though the battery is easily swappable with no tools! Despite the removable battery, it is still rated IP55 for water resistance. You can read a review at www.wired. com/review/fairphone-5/ Besides the battery, parts on the phone you can swap (and get replacements for) are the screen, cameras (either separately or as a module), speaker, USB connector, back cover and earpiece. Fairphone does not sell their products directly to Australia but you can get them through resellers, including on Amazon. They are currently selling the Fairphone 5 for $1449 including GST, while the Fairphone 4 is somewhat less expensive at $1086. Reports are that they work fine on Australian networks, although one user said that the dual SIM feature did not work here. If you want to buy a smartphone that’s easy to fix should something go siliconchip.com.au wrong, iFixit gives all sorts of smartphones repairability ratings at www. ifixit.com/repairability/smartphone-­ repairability-scores You won’t be shocked to find that the Fairphone 5 got their highest score, with the Nokia G22 being the second most favourable. Repairable Computers The Framework Laptop https://frame.work/au/en The main components for this modular laptop are replaceable and upgradeable (see the lead photo). As a result, it is highly repairable. The company is a prominent supporter of the ‘Right to Repair’ movement (see the lead photo and Figs.2 & 3). They sell two main models, the Framework 13 and Framework 16, where the number is the screen’s diagonal size in inches (and thus roughly corresponds to the device’s overall size). The Framework 16 is the latest model and introduces important and unique new features, such as a pluggable GPU which plugs in at the back of the laptop and sits under the display. The Framework 13 compact laptop has the option of either an Intel 13th Gen processor (previous versions had 11th or 12th Gen processors) or an AMD Ryzen 7040 series CPU, with the option of six cores at up to 4.9GHz or eight cores at up to 5.1GHz. The larger Framework 16 comes with an AMD Ryzen processor with 8 cores, 16 threads and 24MB of onboard cache memory running at a maximum of either 5.1GHz or 5.2GHz. You can use the Framework 16 without the GPU module, driving the screen and/ or an external display using its built-in Radeon 780M graphics support. Adding the graphics module, which slots between the main body and screen, makes the device slightly larger and heavier but adds an AMD Radeon RX 7700S graphics process with 8GB of onboard RAM. The GPU draws up to 100W and has two inbuilt cooling fans to handle the resulting heat. It’s handy that you could purchase and add it after owning the laptop for some time, if you later find you need it. Another interesting feature of the Framework computers is the pluggable I/O. Rather than having a fixed set of ports (say, one HDMI video port and three USB ports), the devices have four (Framework 13) or six (Framework 16) expansion slots into which a variety of different I/O ports and other devices can be inserted – see Fig.4. Available modules include USB-A, USB-C, SD card, microSD card, analog audio, SSD storage, HDMI, Display­ Port, and Ethernet, so you can really customise their devices. Third-party vendors also produce different accessories. Fig.4: there are three plug-in module bays on either side of the Framework 16 (and two on either side of the Framework 13). They use a USB-C interface internally and support external ports like USB-C, USB-A, HDMI, DisplayPort, Ethernet and more. SSD storage expansion modules are also available. siliconchip.com.au Australia's electronics magazine Because Framework computers are so modular, it’s relatively easy for the user to replace the battery, keyboard, trackpad and even the screen or motherboard. In addition to enabling repair, if they release a new laptop with the latest CPU and RAM technology, you can swap the motherboard out to upgrade it without replacing the whole device. For the Framework 16, three keyboards are available: the regular type, one with RGB lighting and one with clear keys (also with RGB lighting). You can customise it when you purchase the device or swap it for another later. It is also possible to add a numeric keypad next to the keyboard, or place a white LED matrix module on either side of the keyboard. You can also easily change the LCD bezel colour. Say you have a Framework 13 laptop, and you upgrade the motherboard. What do you do with the old one? Cooler Master makes a case that you can use to turn it into a new, standalone computer – see siliconchip.au/ link/abvp Framework laptops are competitive in performance with many ‘regular’, less repairable laptops, although the cost is somewhat higher for comparable systems. However, that higher upfront cost may be mitigated by the Framework laptops lasting longer due to the ability to repair and upgrade them. That should also lead to less waste to dispose of, as only broken modules need to be discarded, rather than the whole thing (if it couldn’t be repaired). It could also be argued that the flexibility provided by the modular design is a helpful feature worth paying for. There are no resellers of Framework laptops in Australia that we are aware of but you can order them directly from their website at https://frame.work/au/ en The prices are in AUD (check the upper-right corner of the website) and include GST and delivery. Due to high demand for the Framework 16, it could be a couple of months between placing an order and receiving the laptop. Framework 13 models appear to be in stock at the time of writing. Prices start at $1689 for the Framework 13 prebuilt with Windows, although we think 8GB of RAM is too little, so realistically you would need to spend $2359 for the ‘Performance’ July 2024  17 version (16GB RAM + 512GB storage) or $2679 for the ‘Professional’ version (32GB RAM + 1TB storage), which also have better processors. The base model of the Framework 16 costs $2819 prebuilt with Windows installed and comes with sufficient RAM (16GB) and 512GB of storage. If you don’t want to fork out for a Framework laptop, check iFixit’s repairability ratings at siliconchip.au/ link/abx6 and decide based on that. Unsurprisingly, they give the Framework 16 a 10/10 score. MNT laptops https://mntre.com MNT makes the Reform laptop, Pocket Reform, Reform, Reform Keyboard and Reform Camera (see Fig.5). These modular products use open hardware with open-source software. The main repository for the Reform laptop is at https://source.mnt.re/ reform/reform Due to the device being highly modular and using standard parts (such as user-replaceable 18650 cells for the battery), rather than everything being on one circuit board, the computer is quite large and somewhat more expensive than an equivalent non-­modular laptop. For more details, see the video titled “This laptop was made to be hacked!” at https://youtu. be/_DA0Jr4WH-4 One Laptop per Child (OLPC) https://laptop.org Also known as the “$100 laptop” (Fig.6), it was an initiative started in 2005 by a foundation to build an inexpensive and robust laptop for educational purposes. The software used was open-source, including the Sugar operating environment (www.­ sugarlabs.org), designed for interactive learning by children, which was used on some models. Sugar is still available and will run on a variety of platforms. Unfortunately, while the various computer products were good and did sell, they could never meet the targeted price points, and the foundation closed in 2014. For more on this, see the video titled “XO-1: The $100 laptop (which cost $200)” at https:// youtu.be/zZ7qkZkp57c Raspberry Pi www.raspberrypi.org The Raspberry Pi single-board computer (SBC) runs the open-source Linux operating system. However, the hardware is proprietary, as the Raspberry Pi Foundation earns income from the sale of the boards. One variation has the Raspberry Pi and other components built into a 3D-printed open-source case to make a laptop – see Fig.7. Valve’s Steam Deck www.steamdeck.com The Steam Deck is a versatile handheld gaming computer. It uses the Steam­OS distribution, which is based on Linux and was developed by Valve, the maker of the Steam Deck. SteamOS is open-source but has some closed components. Fig.5: an MNT reform laptop with the lid open. Source: www.omgubuntu. co.uk/2020/01/mnt-reform-open-source-laptop 18 Silicon Chip Australia's electronics magazine Fig.7: a Raspberry Pi based laptop that you can build using the files at www.thingiverse.com/thing:3134603 The Steam Deck is modular and repairable by the user, with spare parts available from iFixit (australia.ifixit. com/collections/steam-deck-parts). Simputer https://w.wiki/A2Df The Simputer was an Indian project to design an open-source hardware Linux-based handheld computer (like an early tablet computer) as an alternative to personal computers. The project started in 2002 and ended in about 2006. They only sold about 4000 units, much lower than the goal of 50,000. The project’s failure seems to be due to the product being introduced before there was sufficient demand. For more details, see Fig.6: the OLPC XO-1 was intended as an inexpensive and robust laptop for educational purposes. Source: https://w.wiki/A4Tt siliconchip.com.au Fig.8: the BigFDM, a large opensource hardware 3D printer. Fig.9: an ECU from a Ferrari 360, made by Bosch using a hybrid construction technique on a ceramic substrate. There is no PCB; devices are connected by thin bond wires. Source: https://youtu.be/tEBe6QWTk9U?t=777s the YouTube video at https://youtu.be/ QbDLG2EoGCw intended as a self-­replicating machine. However, the project was discontinued in 2016 due to the large number of commercial 3D printers that had entered the market. Lenovo Thinkpad www.lenovo.com/au/en/c/laptops/thinkpad The Thinkpad line of laptops has a sizeable following online for their ease-of-repair & durability (www. thinkwiki.org/wiki/ThinkWiki). Some of the Thinkpad models, such as the T430, are highly modifiable with the ability to change the screen or even the CPU. There is also custom BIOS software that can be flashed to allow for extra functionality. Open-source 3D printers Open-source 3D printing began in 2005 with the RepRap initiative. It was BigFDM https://github.com/fab-machines/BigFDM The BigFDM is an open-source large-scale 3D printer with an 800 × 800 × 900mm printing area – see Fig.8 and https://github.com/fab-machines/ BigFDM Prusa Research models www.prusa3d.com Prusa Research has a variety of open-source models and aims “for our printers to remain moddable, easily repairable, and produce amazing Repairing ‘non-repairable’ items In cases where you have an electronic module that is an expensive ‘throwaway’ item, some companies are set up to repair them. That is especially helpful if the original part is no longer available, as is becoming more common these days, although it can also be a lot cheaper than buying a replacement. It’s also vital if the module is ‘paired’ with the rest of the device or vehicle, so a replacement won’t necessarily work. One such company the author has used is www.modulerepair.com.au However, numerous other companies would offer similar services, perhaps specialising in particular kinds of modules (air conditioner controllers, TV parts, automotive modules etc). Besides modules, often, if something breaks down, it is possible to fix it yourself, even if circuit diagrams and other resources are not readily available. One of the first places many people look at for how to repair a closed-source device is in a YouTube video. You can also try a web search to find information on repairing a specific model or type of device. If you’re lucky, you could find information on a previous repair to a similar device in our “Serviceman’s Log” column! Another place to look is the website www.ifixit.com, which has free repair guides. They also sell specialised repair tools and spare parts. siliconchip.com.au Australia's electronics magazine prints even decades after their initial release”. These models can be seen on their website at siliconchip.au/ link/abvf with links to software and printable files. Models listed there include the Prusa SL1, SL1S Speed, MK2S, MK3S+, MINI, XL and MK4. The file downloads include models for the 3D-printed parts of those printers, firmware, circuit diagrams, PCB designs, parts lists and the mechanical details of other bits of hardware. Modifiable Vehicles Engine control units (ECUs) can be difficult to repair, and replacements are not always available, especially for cars built in small numbers or when they have a widespread defect and all the replacements have already been used up. Some ECUs used a hybrid construction technique (see Fig.9) without a circuit board, making component-­level repair very difficult. One solution is to replace the original ECU with a third-party version that’s either designed as a drop-in replacement or designed to be adapted to many vehicles. Such ECUs can even be used to upgrade an older car with an analog computer or a mechanical system like points and a distributor. Companies like Haltech (based in Sydney) make and sell such ECUs, but there are also open-source designs. Open-source ECUs include: • rusEFI (https://rusefi.com) • Speeduino (https://speeduino. com/home) • OpenECU (https://openecu.com/ product/openecu) July 2024  19 • FreeEMS (http://freeems.org) For more on ECUs, see our articles on Automotive Electronics in the December 2020 and January 2021 issues (siliconchip.au/Series/353). ECUs are not the only electronic modules used in cars. For example, Open Source Car Control (OSCC) is a set of “software and hardware designs that enable computer control of modern cars to facilitate the development of autonomous vehicle technology” (https://github.com/PolySync/oscc). Android Automotive https://built-in.google/cars Android Automotive, not to be confused with Android Auto, is an opensource version of the Android operating system developed by Google and Intel in collaboration with manufacturers such as Audi, BMW, Ford, General Motors, Honda, Porsche, Renault/ Nissan/Mitsubishi, Volkswagen Group and Volvo. It is embedded in the car, rather than running from the driver’s smartphone like the Android Auto App. Various manufacturers are offering it in their vehicles now, with many more coming next year. However, we urge caution as many car manufacturers have been caught violating owners’ privacy through in-car cameras, GPS tracking, phone contact synchronisation and other methods. DriveKit Fig.11 and https://polysync-xrcc. squarespace.com/drivekit Electric vehicle (EV) open-source hardware and software has also been developed, including Open Inverter, a project sharing information about how to reuse components from commercial EVs using open-source controllers (hardware and software) for EV conversions (https://openinverter. org/wiki/Main_Page). There is also an open-source inverter design to control commercial EV motors, which can be purchased prebuilt, or you can download the firmware source code, binaries, diagrams and various tools from https:// github.com/jsphuebner/ OpenEVSE (Electric Vehicle Supply Equipment) www.openevse.com Open Source Electric Vehicle Charging Station is an open-source charger for electric vehicles; see Fig.12 & https://github.com/OpenEVSE The chargers can be purchased from https://shop.openenergymonitor.com/ evse and there is a construction guide at siliconchip.au/link/abvq Automotive Grade Linux www.automotivelinux.org An open-source project by car manufacturers, suppliers and technology companies to develop Linux-based software for the “connected car”. They hope that this open platform will become an industry standard. https://docs.drivequant.com DriveKit is a commercial vehicle control module that uses OSCC to support ‘drive-by-wire’ control of a motor vehicle, for “full control of steering, brake, throttle, and gear selection for advanced testing and development”. It works with the Kia Niro hybrid and Kia Soul EV, among others. See Fig.10, Toyota The eCorolla was an open-source electric vehicle conversion for a Toyota Corolla; see https://jww.fi/home Ford Ford has open-sourced aspects of their digital instrument cluster and related software; see: siliconchip.au/ link/abvr Porsche Unlike some companies who fight the open-source movement, Porsche embraces it. They state, “By using open source software, Porsche is able to shorten development cycles, reduce costs, promote innovation and talent and improve software quality” (see siliconchip.au/link/abve). Mercedes-Benz Mercedes-Benz is a rare example of a manufacturer with an excellent track record of supplying parts, even for older models. They attempt to maintain a supply of all parts for their classic cars, so there is less need for third parties to step in and make parts that are no longer available, at least for now. Hopefully, that will extend to electronic modules when more modern cars become ‘classics’ – siliconchip. au/link/abvs Also, a Mercedes-Benz owner made an open-source enhancement for displaying data and controlling some aspects of a W211/219/209/203 series vehicle – see his post at siliconchip. au/link/abvl Open-source Tractors The John Deere tractor company is frequently cited as a key example and motivation behind the Right to Repair movement, which we covered in the June 2021 issue (siliconchip. au/­Article/14881). Only their official dealers have access to proprietary software, parts and tools. Not only does that allow them to charge pretty much what they want for repair services, but dealers can be Figs.10 & 11: an EV control module that uses open-source software. Source: https://polysync-xrcc. squarespace.com/drivekit and www. researchgate.net/figure/PolysyncDrive-Kit-with-all-of-the-componentslisted-by-name-that-are-needed-for_ fig2_363024960 20 Silicon Chip Australia's electronics magazine siliconchip.com.au Fig.14 (left): the Tabby EVO open electric vehicle platform. Source: www.openmotors.co/product/tabbyevo/ too busy to make repairs promptly. Since they can’t always visit a farm to repair critical equipment, farmers must pay large sums to transport the equipment to the dealer. So, there is a great desire to find alternative ways to repair those tractors. Tractor ‘hackers’ are decoding and then open-sourcing aspects of the John Deere CAN Bus signals using software called PolyCAN, which was developed to do this. PolyCAN can both decode and send signals from and to the tractor computers. See the video titled “PolyCAN Demo | Manipulating the RPM gauge on a John Deere Tractor” at https:// youtu.be/oqHf6C9QBmY and https:// tractorhacking.github.io We mentioned in the previous article on the Right to Repair that older tractors have been gaining popularity due to their ease of repair. ‘Basic’ tractors from overseas are also quite popular because they are not ‘locked down’. However, several open-source Fig.15 (right): the Oggún II tractor. Source: https:// ronnietractors.com or repairable tractor designs have either been released or are in testing to try to help farmers. One of these is the LifeTrac (see siliconchip.au/link/abvt). To explain their motivation, they write, “Industrial tractors are being designed increasingly for planned obsolescence with 10 year lifespans, and the user typically cannot service their own tractor due to complexity of design.” The design has even been investigated for use as a Mars Rover, as described at siliconchip.au/link/abvg This vehicle appears to be under development, which has possibly stalled. Still, it gives an idea of the sort of things that can be done with opensource concepts. Open Motors TABBY EV www.openmotors.co An open electric vehicle car platform that includes the motor, drivetrain and running gear but not the bodywork (see Fig.14). The platform can be purchased, or you can build your own from downloadable plans. They have a four-seat version at siliconchip.au/link/abvh and a twoseat version at siliconchip.au/link/ abvi As they are open-source designs using readily obtainable parts, including standard batteries, the result is highly repairable and upgradeable. Before building one, you would need to check the legality of using them on public roads in your country or region; they are legal in the USA and Europe. This type of vehicle would typically come under a ‘kit car’ exemption but would still need to pass various checks. You can see videos on these vehicles at https://vimeo.com/157998468 and https://vimeo.com/113110682 Many commercial EVs are written off by insurance companies even after minor accidents due to concerns about possible damage to the expensive custom battery pack. Often, it isn’t Fig.12: parts that can be used to build the OpenEVSE EV charger. Fig.13: a prototype of the Acorn precision farming rover. Source: https://youtu.be/fFhTPHlPAAk siliconchip.com.au Australia's electronics magazine July 2024  21 possible to properly assess the damage due to the ‘all-in-one’ nature of the pack. Using multiple standard battery packs could therefore be a good idea. Oggún Tractor https://ronnietractors.com/oggun-tractor The Oggún Tractor (Fig.15) claims to be an open-source design, although the drivetrain is not fully open-source. Nevertheless, it mainly uses off-theshelf parts and is an attempt at a lowcost, repairable tractor that might be suitable for smaller farms. For more details, see the article at siliconchip. au/link/abvj Acorn https://github.com/Twisted-Fields Fig.16: the AgOpenGPS unit steers the tractor using 3D-printed gears attached to the steering wheel. Source: AgOpenGPS – siliconchip. au/link/abw1 Acorn is an open-source, precision farming rover to perform tasks such as planting seeds, destroying weeds, monitoring plant health and other tasks – see Fig.13. AgOpenGPS https://discourse.agopengps.com An open-source GPS guidance software and hardware for tractors that allows them to perform many tasks, including automatic steering for precision ploughing and planting – see Fig.16. Aviation systems ArduPilot https://ardupilot.org Fig.17: two configurations of the ArduPilot controller with different connectors. Source: Fruugo – siliconchip.au/link/abw0 ArduPilot is an autopilot system supporting autonomous multi-copters, traditional helicopters, fixed-wing aircraft, boats, submarines, rovers and others (see Fig.17). It initially used Arduino processors but now supports many other hardware platforms. The ArduPilot code of conduct prohibits utilising the device in crewed vehicles or weapons. titled “Arduino EFIS. Part 1” at https:// youtu.be/emqc_vi7-Rg MakerPlane https://makerplane.org MakerPlane is an open-source aviation community developing opensource plans, avionics and building a community of similar-minded people. See Fig.20 and the video titled “MakerPlane Overview | An Open-Source Aviation Community” at https://youtu. be/XFis22qoJ5c OpenVario www.openvario.org OpenVario is an open-source flight computer – see Fig.21. Stratux https://stratux.me This open-source software is for building an ADS-B receiver (Automatic Dependent Surveillance-Broadcast for weather and air traffic data) using a Raspberry Pi, a radio module, a GPS module, a case and other commercially available parts. It can be connected to a smartphone, tablet or EFB (Electronic Flight Bag) to receive ADS-B data without paying a subscription. SUAVE https://suave.stanford.edu SUAVE is an open-source “aircraft design environment built with the ability to analyze and optimize both conventional and unconventional designs”. XCSoar www.xcsoar.org Open-source software for gliders Avare www.apps4av.com Avare is moving-map software for Android devices. It is compatible with Stratux (see below). As it uses FAA data, it may only be usable in the USA, with some unofficial support in Canada and the EU. Experimental Avionics https://experimentalavionics.com Figs.18 & 19: an EFIS display unit from Experimental Electronics. Source: https://experimentalavionics. com/efis-display-unit/ 22 Silicon Chip A website devoted to open-source avionics for experimental aviation, mostly based on Arduino devices. One example is an Electronic Flight Instrument System (EFIS) display, as shown in Figs.18 & 19. Information is received from the aircraft CAN bus and Arduino sensors. For more information, see the video Australia's electronics magazine Fig.20: the MakerPlane pyEFIS 2.0 beta software, electronic flight... siliconchip.com.au Fig.22: the HackRF circuit board, an open-source hardware SDR that operates from 1MHz to 6GHz. Source: https://github.com/greatscottgadgets/ hackrf?tab=readme-ov-file Fig.23: the open Module 17 implements the M17 digital radio mode in hardware. Source: https:// github.com/M17-Project/Module_17 that runs on Android, Kobo (eReader), Windows and Linux. and receive on frequencies from 1MHz to 6GHz – see Fig.22. 9600 baud serial communications is required. Radio & radio software M17 Meshtastic Codec 2 An open-source software-defined radio (SDR) platform that can transmit M17 is a project that develops hardware and software for the M17 amateur radio in digital mode. TYT model MD-380, MD-390 and MD-UV380 transceivers can be reflashed with open-source firmware to support this digital mode. It can also be used on just about any modern amateur radio that connects to a computer. An open-source hardware modem board called “Module 17” has been developed to perform the encoding in hardware rather than software (see Fig.23). A transceiver that supports Meshtastic is an open-source project that utilises the license-free LoRa mesh radio protocol to send messages over kilometres or tens of kilometres without connecting to any infrastructure, such as phone towers. It works by ‘meshing’ with other similar devices if available; the more devices are present, the longer the potential range. While LoRa boards are proprietary, the Meshtastic software is open source. There are videos about using Meshtastic devices titled “The Ultimate Meshtastic Device – Long ...information system software written in Python. Fig.21: the OpenVario open-source flight computer. Source: www.openvario. org/doku.php www.rowetel.com/?page_id=452 An open-source speech codec software for amateur radio and other digital voice applications. It is used by FreeDV and M17. FreeDV https://freedv.org FreeDV is open-source software for digital voice on HF amateur radio. HackRF https://greatscottgadgets.com/hackrf siliconchip.com.au https://m17project.org Australia's electronics magazine https://meshtastic.org/docs/introduction July 2024  23 Range Comms” at https://youtu.be/ knyg6EEiGOo and “Getting Started with Meshtastic – Devices” at https:// youtu.be/DUz6cVSaSl4 Note that you need devices that operate in suitable frequency ranges for your location, as the available frequency bands vary by country. Quansheng UV-K5 http://en.qsfj.com/products/3002 The UV-K5 radio can be reflashed with open-source firmware to dramatically improve its capabilities (see siliconchip.au/link/abvn). It has been described as “The Most Hackable Handheld Ham Radio Yet” by IEEE Spectrum (siliconchip.au/link/abvw). An amateur radio license is required to transmit using this radio – see our article on getting one in the April 2024 issue (siliconchip.au/Article/16206). uSDX https://github.com/threeme3/usdx uSDX is an open-source Class-E driven amateur transceiver. Appliances and other devices There are various possibilities for interested parties to develop opensource refrigerator designs; check out siliconchip.au/link/abvx Open-source medical ventilators were developed during the COVID-19 pandemic when there was expected to be a shortage of ventilators. We already covered this topic in an article from the June 2020 issue (siliconchip.au/ Article/14459). Open Source Washing Machine siliconchip.au/link/abvk There was an attempt in 2008 to develop an open-source washing machine for use in ‘third world’ countries. It was called OSWASH or the Open Source Washing Machine Project. It was to use recycled parts and a Freeduino as a controller. Unfortunately, it never seems to have developed beyond an idea. Open Source Scan Converter (OSSC) https://retrorgb.link/ossc OSSC helps keep classic video games running (see Fig.24). This is an example of open-source products keeping older devices running. It is “designed primarily for connecting retro video game consoles and home computers to modern displays”. There is a video on it titled “OSSC: Getting Started and Taking The Next Steps” at https://youtu.be/vHqT1God9vk The reasons that proper scan converters are needed, rather than using 24 Silicon Chip Fig.24: the Open Source Scan Converter, ManuFerHi version, for connecting older devices like gaming consoles to modern TVs. Source: https://github.com/ ManuFerHi/OSSC analog inputs on modern TVs, are explained in the video titled “Why Retro Consoles Need A Scaler” at https://youtu.be/TdfFnR-hOK8 In summary, modern TVs have poor scan conversion hardware/software, and the lag on many modern TVs is way too high for playing video games. Open-source integrated circuits (ICs) Even ICs (silicon chips) can be made open-source. The first opensource commercially available chip was released earlier this year. OpenTitan https://opentitan.org OpenTitan (https://opentitan.org) is a type of security chip known as a root of trust (RoT) component. Being opensource, the internal code is verifiable for authenticity and can be examined by anyone for weaknesses. The OpenTitan project was initiated by Google in 2018 and led by not-forprofit company lowRISC with participating companies including Winbond, Nuvoton, zeroRISC, Rivos, Western Digital, Seagate, ETH Zurich and G+D Mobile Security. The objective is to use the chip to develop trustworthy and secure platforms. For more information, see the video titled “How the Silicon Commons, developed through Open­Titan, is revolutionizing chip design” at https://youtu.be/4YfCDnpYm1Y RISC-V https://riscv.org RISC-V is an open-source and royalty-­free standardised instruction set for CPUs. Individual chip designs based on RISC-V might be commercial Australia's electronics magazine or open source. You can see a photo of a prototype RISC-V chip in Fig.25. RISC stands for ‘reduced instruction set computer’. The main advantages of RISC chips are that they are easier to implement and can be made quite power-efficient. Bitlog (siliconchip.au/link/abw2) created an open-source RISC-V bit-­ serial CPU called “SERV”, with a focus on being as minimal as possible (the world’s smallest implementation), not as fast as possible. Its source files are at https://github.com/olofk/serv One advantage of its small size is that many cores can fit on one piece of silicon. There is a video about SERV at siliconchip.au/link/abvm If you want to try a RISC-V-based computer, you can get the BeagleBoard BeagleV-Ahead small-board computer (SBC) from https://au.element14. com/4205457 for around $220. It has a 64-bit, 1.2GHz quad-core Xuantie C910 processor, 4GB of RAM, 16GB of flash, a GPU, USB3, WiFi and Ethernet. The C910 processor is an opensource design; you can download its Verilog source code and simulation files from https://github.com/T-headSemi/openc910 OpenROAD https://theopenroadproject.org OpenROAD is open-source software that allows designers to perform all steps of silicon design, from a Register Transfer Level (RTL) description (a high-level description of the chip’s functionality) to the final Graphic Data System (GDS) file. The GDS file represents the complete layout of the chip, including details of physical layers, shapes, and interconnections. siliconchip.com.au Fig.25: a RISC-V prototype chip. Source: www.flickr.com/photos/ dcoetzee/8694597164 OpenROAD works with various commercial and open-source process design kits (PDKs). PDKs are used to design, model and verify the fabrication process before the design is committed to hardware in a silicon foundry. Available open-source PDKs and their feature size capability include GF180 (180nm), SKY130 (130nm), Nangate45 (45nm) and ASAP7 (Predictive FinFET 7nm). Miscellaneous Linux is open-source software, not hardware, but we mention it here because so much open-source hardware relies on it. That includes all Android devices and many smallboard computers (SBCs), like the Raspberry Pi 5 and Rock 4C+, as well as devices controlled by an SBC. Linux is an operating system for personal computers, servers, embedded computers and many other devices. Besides being free and usable as a substitute for Windows or MacOS, Linux can also be used on old and otherwise obsolete computers; it doesn’t need the latest hardware like Windows. It can be entirely usable on modest hardware. There are versions of Linux such as gray386linux (https://github.com/ marmolak/gray386linux) that will run on an ancient 386 computer or from a floppy disk (eg, FLOPPINUX – see siliconchip.au/link/abvy). But Linux isn’t just for old computers; it can run on the latest desktop and portable computers and is even used by most modern supercomputers, customised by the manufacturers. You might have a perfectly good siliconchip.com.au Fig.26: the Gazebo software for simulating robotics. Source: https://github.com/gazebosim Windows 10 computer, but many Windows 10 computers can’t run Windows 11, so what will you do when Windows 10 support ends in October 2025? Many people have said they will switch to Linux or already have. See the YouTube video titled “Windows Just Did What? | Time to Start Switching to Linux” at https://youtu. be/NohhYEO8jaM Linux can also be used to boot a computer from a USB flash drive if the computer is otherwise unbootable, to recover a corrupted installation, or just to try out using Linux. Unlike early versions of Linux, which were for “geeks only”, modern versions are much more user-friendly and can be operated without specialist knowledge. The large variety of Linux “distributions” (versions) is listed at https://w.wiki/32za Also see the video titled “Top 5 Linux Distros For Older Hardware” at https://youtu.be/qUpdHF69BQY We like Ubuntu, especially for its long-term support versions, but there are plenty of other good distributions. ELISA (https://elisa.tech) stands for Enabling Linux In Safety Applications. Its aim is “to make it easier for companies to build and certify Linux-based safety-critical applications – systems whose failure could result in loss of human life, significant property damage or environmental damage”. OpenSCAD (https://github.com/ openscad/openscad) is free and opensource software for creating three-­ dimensional objects, typically for 3D printing. Thingiverse (www.thingiverse.com) is a repository of over one million 3D Australia's electronics magazine printer files, all free and open-source hardware designs. The website is free to use once you set up an account. ROS (Robot Operating System; https://ros.org) is a set of software frameworks for developing robot software; Gazebo Simulator (https://­ gazebosim.org/home) is a robot simulator, while Open-RMF (www.openrmf.org) enables interoperability and sharing of spaces between different fleets of robots and building infrastructure – see Fig.26. More information ● You can take failed devices to a Repair Café or become a volunteer: www.repaircafe.org/en/visit ● List of open-source hardware repos: https://github.com/topics/opensource-hardware ● Major open-source software repositories include: ¬ https://github.com ¬ https://code.google.com ¬ https://sourceforge.net ¬ www.apache.org ● 3D printing files (not all free): ¬ www.thingiverse.com ¬ www.printables.com ¬ https://cults3d.com/en ¬ www.myminifactory.com ¬ https://pinshape.com ¬ www.redpah.com ¬ www.youmagine.com ● Learn to code for free at www. freecodecamp.org ● Journal of Open Hardware: https://openhardware.metajnl.com ● “Open-Source Electronics Platforms: Development and Applications” book (2019): siliconchip.au/ link/abvz SC July 2024  25 Project By Charles Kosina Automatic LQ Meter inductance / Quality Besides adding the ability to measure inductance, so you don’t need a separate LC meter, one of the big advantages of this new design is that it has an onboard signal generator, so you no longer need two instruments to make a Q measurement. Also, its operation is entirely automatic, whereas the previous design required fiddling with knobs and a specific procedure to make the measurement. Much of the circuitry is similar to the older Q-meter design. Still, while I was adding the new features, I took the opportunity to optimise and simplify it without sacrificing any performance. As I mentioned in my previous article, there appear to be no manufacturers of Q meters any more, and the scarce second-hand ones from the likes of Hewlett-Packard fetch quite large sums. I saw one recently selling on eBay for US$2400. This one costs a small fraction of that to build! Basic operation A Q Meter is an indispensable tool for anyone contemplating RF design. My previous design in the January 2023 issue (siliconchip.au/ Article/15613) works well but has limitations; it needs an external signal generator with a well-defined output level. This new design is two instruments in one, measuring inductance from 0.1 to 999μH and Q from 10 to 300 with a test frequency from 100kHz to 90MHz! Features & Specifications ● Measures inductance (L) and quality factor (Q) over five frequency ranges ● Inductance (L) range: 0.1-999μH with 100nH resolution ● Quality factor (Q) range: 10 to 300 ● Test frequency range: 100kHz to 90MHz ● Resonant capacitance options: 18pF, 51pF, 118pF, 238pF or 488pF ● Power supply: battery (3 x AA) or 5V DC <at> 200mA T he January 2023 article explains what an inductor’s quality factor (Q) means and goes into the theory of Q measurement. In brief, an inductor with a low Q has more inherent damping, so it forms a filter with a broader response and a lower peak. In 26 Silicon Chip contrast, a high-Q inductor will make a filter with a narrow (more selective) response and a higher peak. So you need to know the Q of the inductors in your filters, at the frequency they will operate, if you want to model their response accurately. Australia's electronics magazine Briefly, we can determine both the inductance and Q by exciting a resonant LC network containing the unknown inductor and a known capacitance at a controlled frequency. There will be a peak in the amplitude of the resonance at a particular frequency. The relevant formula is: f = 1 ÷ (2π × √LC) Since we know f and C, we can rearrange it to solve for L, giving us: L = 1 ÷ C(2πf )2 f is the resonant frequency, so we can sweep the oscillator and find the point at which the amplitude is at a maximum, then plug that into the formula. Changing C will shift the resonant frequency but should give us the same inductance result. That is necessary so that small and large inductance values can be measured at a reasonable frequency (within the device’s operating range). As for the Q factor, once we’ve found the peak, we can also measure the amplitude of resonance. The ratio between that and the excitation amplitude will give us our Q measurement, as we shall explain in a little more detail later. Design decisions My first decision was how to generate the test signal over the required range. My first idea was to use a DDS siliconchip.com.au chip such as the AD9851. However, with a clock frequency of 180MHz, the Nyquist limit is 90MHz, so 70MHz is about the highest frequency it can practically generate. Also, it’s a relatively expensive chip or module. Another regular contributor to Silicon Chip, Andrew Woodfield, suggested using the Silicon Labs Si5351 clock generator. I have used this chip in other applications, and it is extremely versatile, going up to 200MHz and beyond. These are available as readymade modules with 25MHz crystals at a very low cost from AliExpress and other suppliers. Its frequency is set by loading many registers over an I2C serial bus. That makes it easy for me to use a microcontroller to perform a continuous frequency scan. The output of the Si5351 chip is buffered by a high-speed op amp, the OPA2677, configured with a unity gain. This has a gain bandwidth (GBW) of over 200MHz, so it will have a reasonably flat output to at least 90MHz. As with the previous design, the output of the OPA2677 feeds a toroidal transformer with a 10:1 turns ratio, the secondary being a threaded standoff passing through the middle. This gives an extremely low source impedance to drive the series-tuned LC circuit, typically 0.02W. The voltage on the secondary is about 0.25V peak-to-peak. The catch is that the output is not a sinewave but more like a square wave. Instead of just one frequency, we have the Fourier expansion with an infinite number of odd harmonics: sin(ω) + sin(3ω)÷3 + sin(5ω)÷5 + sin(7ω)÷7 + sin(9ω)÷9 + sin(11ω)÷11 … Where ω is 2π times the frequency. It’s an infinite series, but in practice, the higher harmonics are filtered out by the bandwidth-limited circuitry. Consider that the resonant frequency of inductor and capacitor (LC) circuit may be 15MHz. If we drive it with a 5MHz square wave, the third harmonic will resonate and give us a false reading. Fortunately, this problem is easy to overcome. Instead of scanning upwards in frequency, we scan downwards from the highest frequency. As long as the highest frequency is above the resonant point of the tuned circuit, the scan will find the primary resonance frequency on the way down. siliconchip.com.au When starting up to Automatic LQ Meter, the screen should display a message similar to the one shown. The lead image (opposite) shows the Meter measuring an air coil. For example, say we have an airwound inductor of 6µH and a test capacitance of 118pF. The resonant frequency is 5.88MHz. If we set our starting frequency at 30MHz and scan down, no other resonances will be found until we reach 5.88MHz, as the first significant harmonic, the third, will only occur with a test signal of 1.96MHz (5.88MHz ÷ 3). Given a close-to-zero source impedance, the Q value is obtained from the equation Q = Vout ÷ Vin, where Vin is the voltage from the transformer, and Vout is the voltage at the junction of the inductor and capacitor. For a maximum Q reading of 300 and a test signal of 250mV peak-topeak, Vout would be 75V peak-topeak. We need to measure the input and output voltages accurately, but it’s impractical to measure Vin accurately on the transformer’s secondary. However, we know the voltage on the primary is ten times that. My testing shows that the voltage ratio is close to 10:1 over the entire frequency range. Accuracy Measuring Q accurately is not easy. The error budget includes several parameters, including the source impedance of the signal generator. While it is low, it is non-zero. RF voltage measurements are subject to errors and the peak frequency found may be slightly off. The stray capacitance on the circuit board may not exhibit a high enough Q, which will decrease the measured value slightly. Australia's electronics magazine I compared the readings with that of my Meguro Q meter, and they were generally within 10%. Inductance measurements are likely to be within 5%. However, even the HP 4342A laboratory instrument can’t guarantee a particularly high accuracy; it has a tolerance of ±7% on Q values up to 300. Circuit description The resulting circuit is shown in Fig.1. MOD1 is the test signal generator and its output is buffered by IC1a and AC-coupled to transformer T1. The DUT (inductor) is connected across CON3 & CON4. It forms a resonant circuit in combination with one of the 33pF, 100pF, 220pF and 470pF capacitors switched in or out of the circuit by relays RLY1-RLY4 plus the stray PCB capacitance of around 18pF (or just the stray capacitance if RLY1RLY4 are all off). A half-wave precision rectifier built around the other half of the OPA2677 (IC1b) measures the amplitude of the Vtest signal (at pin 1 of IC1a). The output of this rectifier is the DC peak and IC5b buffers that voltage. The gain of this buffer stage is set to 1.25, compensating for a slight amplitude reduction due to the rectifier. The DC voltage feeds the ADC7 input on the Arduino Nano module for measurement using its internal ADC (analog-­to-digital converter). At the same time, schottky diode D7 half-wave rectifies the voltage at the junction of the DUT and the July 2024  27 Fig.1: the test square wave is generated by MOD1, buffered by IC1a and transformed by T1 before being applied to the resonant circuit comprising the DUT and some combination of the 33pF, 100pF, 220pF & 470pF capacitors switched by RLY1-RLY4. The test and resonant voltages are rectified and measured by the Arduino Nano. By knowing the peak resonance frequency, capacitance and those voltages, both the inductance and Q factor can be calculated. 28 Silicon Chip Australia's electronics magazine siliconchip.com.au ADC7 inputs are converted to an integral number from 0 to 1023 (210 − 1). The firmware calculation is simple: multiply the ADC6 value by 11 to recover Vout and divide by the ADC7 value (Vin). But what if we have a coil with a Q of only 10? Vout ÷ 11 would be only 0.225V, or an ADC count of 46, and the broad resonance peak may not be picked up accurately. For low Q values, we increase the gain of IC5a from unity to four times by switching in a 33kW resistor from pin 2 to ground using N-channel Mosfet Q1. This will give an output voltage of 0.9V in this example, or 184 counts, which can be measured far more accurately. Resonant capacitance In my original Q meter, I had eight capacitors switched by relays to select a value from 40pF to about 290pF with 1pF steps to move the frequency of the resonance peak. That was overkill, so I reduced it to a choice of only five values in this design. The stray capacitance of the circuit is around 18pF, setting the minimum value. Why relays and not solid-state switching? To eliminate errors, the capacitance must have a very high Q, preferably ten times that of the highest Q coil. The relay contacts in series with the capacitors have very little effect on the overall Q. The capacitors must be RF types with a 1% tolerance; the values are 33pF, 100pF, 220pF and 470pF, adding to the 18pF of stray capacitance. Power supply and control capacitance (Vout), converting it to a DC voltage by charging a 100pF capacitor. A precision rectifier is unnecessary because the voltage here is much higher; a small voltage drop will not cause a significant error. Applying a maximum of 37.5V DC to an op amp would destroy it, so we have an 11:1 voltage divider made siliconchip.com.au from 10MW and 1MW resistors. This limits the output to 3.4V, which is a good safety margin. This divided voltage has a high source impedance, so IC5a buffers it before feeding it to the ADC6 (A6) analog input of the Nano. The Nano’s ADC has a resolution of ten bits, so the voltages at the ADC6 and Australia's electronics magazine Because op amp IC1a needs to drive the primary of T1 with a signal that swings above and below ground, its negative supply rail needs to be below 0V. We generate an approximately -4V supply rail from the +5V rail using IC7, a MAX660 switched capacitor voltage inverter in a fairly standard configuration. The +5V rail is generated from a three-cell battery (at least 3V) by an MCP1661 switch-mode boost converter (REG1), again in a configuration pretty much straight out of the data sheet. This allows us to power the circuit with three AA or AAA cells (depending on how long we want them to last). The Nano can monitor the raw battery voltage via its ADC3 (A3) analog input. Alternatively, 5V DC can be fed in from a USB supply, such as a phone July 2024  29 charger, via CON5. In this case, REG1 will only operate to overcome the forward voltage of diode D8. If you use rechargeable cells (eg, NiMH), they will also be trickle-charged when external DC power is applied via R1. The current drain in operation is about 200mA, so a decent set of AAs (alkaline or NiMH) should last for around ten hours of use. That might not seem very long, but this type of instrument is generally only used for a few minutes at a time, so the battery life should be OK unless you’re using it constantly. If battery operation is not needed, the MCP1661, the 4.7µH inductor and diode D8 may be omitted. Just put shorting links across the inductor and diode pads. The rest of the circuit is pretty standard. The Arduino Nano has just enough I/O pins for the task. The LCD module is the standard 2x16 alphanumeric type available from multiple sources; the version with a blue backlight is recommended. The four relays that switch the RF capacitors are selected by a 74AC139 multiplexer that will power the coil of just one relay at a time. The current sink capability of the 74AC139 is quite adequate for the relays used. Diodes across the relays absorb switching transients. Fig.2: this shows how voltage samples are taken at various widely-spaced frequencies until nearing the peak, at which point the unit switches to much smaller frequency steps. It’s important to accurately find the peak frequency for precise measurements. resonance, this will be zero or close to zero. There may be a bit of noise, so the algorithm ignores anything less than an ADC count of 5. The frequency steps far from resonance are at broad logarithmic intervals. That means that each step is the current frequency divided by a number. The logarithmic step size arrived at by experimentation is f ÷ 200. For example, at 10MHz, the next step size would be 50kHz (10MHz ÷ 200), making the next frequency 9.95MHz (10MHz − 50kHz). The next step size would be 49.75kHz (9.95MHz ÷ 200) and so on. When the measured voltage is 50 counts or greater on the ADC (about 250mV), we are on the rising side of the resonance curve, so we switch to a much smaller step size of f ÷ 4000. At each step, we measure the voltage and remember the highest voltage and the frequency at which it was found. If the voltage is lower than the highest seen so far, we increment a trailing-­ edge number instead. When the trailing-edge number reaches five, we have passed the peak, so scanning stops. The highest stored voltage and frequency are then used to calculate the Q factor and the inductance. This is illustrated in Fig.2, where each point on the resonance curve is shown. The peak will be sharp for high-Q circuits, so the sampling steps must be close together to avoid missing the peak. During scanning, we switch to the low-Q setting by turning on Mosfet Q1 to increase the op amp’s gain. This means that we will detect the rising slope sooner. If left on this setting, a high-Q coil could saturate the op amp output. To avoid that, we monitor the ADC count for Vout. If this exceeds 900, we switch Q1 off, reducing the measured Vout by a factor of four. As with the previous Q meter design, the brightness of LED1 is proportional to Vout. Because the algorithm takes the scan just past the peak, the LED will increase in brightness, dim slightly, then jump back to the highest brightness as we go back and re-measure the peak value. Measuring RF voltages with great accuracy is not easy. Once the peak frequency is reached, both Vout and Vin are sampled 16 times, and the readings are averaged. That helps to remove random noise. Australia's electronics magazine siliconchip.com.au 30 Silicon Chip Rotary encoder ENC1 is a standard type with a 20mm-long shaft; 27kW pull-up resistors are used for the three switch contacts, with 100nF capacitors for debouncing on two of them. Note that we have two capacitors on the INT0 line. One is located next to the encoder, but some noise spikes must have been getting into that line, making the frequency and capacitance settings erratic. A second capacitor right next to the Nano pin fixed the problem. Two starting parameters can be set. The first is the top frequency, which can be set from 2MHz to 90MHz, while the other is the capacitance value to resonate with the inductor. Three-­ position switch S2 selects the setup mode. Up sets the top frequency, down sets the capacitor value and middle waits for the start switch (S3). These additional switches also have pull-up resistors: 4.7kW for S3 and 27kW for S2. S2 feeds either 5V, 2.5V or 0V to the ADC2 (A2) pin of the Nano depending on its position, so an analog voltage measurement is used to determine its position. Finding the resonance peak To find the peak voltage of the tuned circuit, we start at a high frequency and, at each step down, measure the voltage Vout. When far from I originally had some concerns about the accuracy of meausrements due to the square wave shape. Is the rectified input voltage Vin different between a sinewave and a square wave? To test this, I used my previous Q meter and fed it with a sinewave and square wave generators. Over a frequency range of 1-10MHz, there was no significant difference in the measured Q. Construction Most components mount on a double-sided circuit board coded CSE240203A that measures 138 × 75.5mm. The two modules, the Arduino Nano and the Si5351a clock generator board, are on the back of the PCB; almost all the remaining components are on the front. Start by soldering in all the discrete resistors and capacitors in the locations shown in Fig.3, the PCB overlay diagram. As SMD capacitors do not have any markings, take care that the correct ones are soldered in. I use ceramic capacitors throughout, so like the resistors, their polarity does not matter. Fit the SMD diodes next, all of which are polarised; their cathode stripes must be orientated as shown in Fig.3. The polarity of the surface mount diodes can be hard to see, so if you are unsure, test them with a multimeter. Follow on by soldering the five integrated circuits, including REG4. None of them are particularly fine-pitch parts. Make sure that pin 1 is orientated correctly in each case, as fixing that after you’ve soldered all the pins is a chore! The relays and 1N5711 axial diodes should be mounted next. Like the ICs, the relays must be orientated correctly. After that, solder the sole transistor (Q1) in place. Fit the 4.7μH inductor next; the SMD type is preferable for slightly higher efficiency. It’s a good idea to clean the PCB to remove flux residue before mounting the through-hole components, as it’s easier at this stage. It’s also a good idea to inspect all the SMD solder joints, especially for the ICs, before moving on, as it will be easier to fix any problems now. Winding the transformer Wind ten turns of the specified enamelled copper wire onto the toroidal core (I used 0.4mm diameter wire but 0.25mm is OK), taking care that the turns are equally spaced around the circumference, to the extent possible, Fig.3: most components mount on the top side of the PCB, with just the Arduino Nano, the Si5350a clock generator module and one or two headers on the underside. A large proportion of the parts are SMDs although they are almost all quite large and easy enough to work with. During assembly, take care with the orientations of the diodes, ICs and relays. The top overlay diagram is the front of the PCB, while the bottom diagram is the back. The pads for one 100nF capacitor were accidentally left off the PCB, so it can be soldered like this (using a throughhole cap makes it easier). siliconchip.com.au Australia's electronics magazine July 2024  31 and that the ends line up with the two small pads on the PCB (one of which is attached to the large central hole). Scrape the enamel off the ends of the wires, and tin them so they can be soldered to the PCB. Make sure it is centred correctly so that the spacer can pass through the middle. Once it is in place, gently feed one of the brass spacers through the hole in the middle of the toroidal core and feed in a bright metal M3 machine screw through the back of the PCB to attach it firmly (it needs to make good electrical contact). Attach the other brass spacer similarly to the hole just below the toroidal core and to the right of diode D7. Now it’s time to mount the various through-hole parts except the LCD, LED and modules. When fitting pushbutton switch S3, ensure that the NC contact goes towards the bottom of the board. Check which outer pin is connected to the middle pin with a continuity meter when the button is not being pushed; that is the NC contact. Also take care that the switches and encoder are exactly at right angles to the board so that they fit through the front panel neatly. The best way to do this is to solder just one pin on each, then adjust their orientation so the front panel fits over them. Once you are happy with that, solder the remaining pins. For the LED, insert its leads through the 8mm spacer before soldering it to the board. Its longer (anode) lead goes to the left, next to the adjacent resistor. The flat side of the lens should face to the right. Before mounting the LCD screen, the Arduino Nano and Si5351 modules must be attached to the back. You could use socket strips to mount them, but it is not essential. In each case, if the module didn’t come with a header soldered to it, fit one now. Finally, attach the LCD module on the front with 10mm M3 screws, hex nuts and 3mm spacers. The Si5351 module is also held in place with M2/ M2.5 screws and 3mm spacers. After cleaning the circuit board again, inspect all soldered joints and touch up any problems. The photographs show a prototype version of the board; the revised one has a few changes. Several components were not required and were removed from the artwork, while others were added. Programming the Nano Before the LQ Meter can be tested, the ATmega238 microcontroller on the Arduino Nano module must be programmed. The modules generally come preprogrammed with a bootloader, with the correct fuse settings and a 16MHz onboard crystal, so you just need to load the LQ Meter specific firmware. How you do that depends on what equipment you have. The simplest way is to plug the Nano into your computer using a suitable USB cable and upload the HEX file using free Windows software called AVRDUDESS (download from siliconchip.au/link/ aaxh or use the command-line version, avrdude, if you’re running Linux or macOS). Download the firmware from our website at siliconchip.au/Shop/6/416 then unzip it and extract the HEX file. Run AVRDUDESS and set the programmer to Arduino, select the Nano’s USB serial port, a baud rate of 115,200 or 57,600 (depending on your Nano) and click “Detect”. If it doesn’t find the chip, adjust the settings and try again. Once it does, go to the Flash window, open the HEX file for this project and click the program button. You should get a confirmation message, and that’s it – the Nano is ready to go. Initial Testing Note that the LCD screen is soldered to the PCB, as there isn’t enough clearance to mount it on a socket. 32 Silicon Chip Australia's electronics magazine Don’t install the board in the enclosure yet. With the Nano programmed, a battery or external power supply can be connected to the board. Leave the power switch off and briefly connect a multimeter on its high current range across the power switch. Around 200mA should flow. A much higher current than that could indicate a short on the board. If all is well, proceed to the next stage. siliconchip.com.au Switch it on and adjust potentiometer VR1 until the LCD screen image is legible. Switch it off and on again; the splash screen will show the version number and the battery voltage. After a couple of seconds, the following screen shows the capacitor value and top frequency. To adjust these, use the centre toggle switch and set the values with the encoder. Once the values have been set, press the encoder switch to store the values in EEPROM, which are read on the next power-up. It’s possible that the encoder will work backwards. This depends on the specifics of your encoder and is quite unpredictable. If that happens, switch off the power, hold down the encoder switch and switch the power back on. The display will show “Toggling Direction”. The direction bit is stored in EEPROM and will give correct operation from then on. Parts List – Automatic LQ Meter Use the front panel PCB as a template for drilling holes in the front panel of the enclosure; Fig.4 shows the hole sizes. The panel is a snug fit in the detent, which makes for accurate drilling. Note that the spacers have clearance holes in the case so that they contact the pads on the back of the front panel. With the front panel in the enclosure slot, attach the red and black terminal posts. Two nuts are used on the posts, one on the outside of the panel to make good contact with the pad, the other on the inside with the washer. Tighten them well to maintain a low resistance. The circuit board can then be slotted in and attached by two black 8mm-long M3 machine screws and the nuts on the switches. Tighten the inside nuts on the switches right down for a correct fit. Push the knob onto the encoder shaft, and the unit is nearly complete. All that remains is to mount the battery holder and DC socket (for external power or battery charging) in the base of the case and wire them up. Drill a hole in the side for the DC socket (if you’re using it) and mount it. Make sure it won’t foul the PCB or battery holder once it has been installed. Attach the battery holder to the base using double-sided tape, then solder the 47W axial resistor between the DC socket’s positive terminal and the battery holder’s positive wire. Solder the 1 double-sided PCB coded CSE240203A, 138 × 75.5 × 1.6mm 1 double-sided front-panel PCB coded CSE240204A, black solder mask, 138.5 × 76 × 1mm 1 165 × 85 × 55mm IP65 sealed ABS enclosure with clear lid [Altronics H0326] 1 Si5351A clock generator module (MOD1) 1 Arduino Nano (MOD2) 1 16×2 alphanumeric LCD with blue backlight (LCD1) [Silicon Chip SC5759] 4 HFD4/5 subminiature DIP signal relays (RLY1-RLY4) [AliExpress] 1 Fair-rite 5943000301 ferrite toroid (T1) [element14 2948713] 1 30cm length of 0.25-0.4mm diameter enamelled copper wire (T1) 1 4.7μH M3216/1206 SMD inductor or axial RF inductor (L1) [Murata LQM31PN4R7M00L] 1 rotary encoder with integral switch and 20mm-long shaft (ENC1) [Silicon Chip SC5601] 1 knob to suit ENC1 1 SPDT miniature two-position toggle switch with solder tags (S1) [Altronics S1310] 1 SPDT miniature centre-off latching toggle switch with solder tags (S2) [Altronics S1330] 1 SPDT miniature momentary pushbutton switch with solder tags (S3) [Altronics S1391] 1 10kW top-adjust multi-turn trimpot (VR1) 1 3 × AA side-by-side battery holder with flying leads (BAT1) 1 2-pin vertical polarised header with matching plug and pins (CON1) [Jaycar HM3412 + HM3402; Altronics P5492 + P5472 + 2 × P5470A] 1 4mm red binding post (CON3) 1 4mm black binding post (CON4) 1 panel-mount DC barrel socket (CON5) [Jaycar PS0522] Semiconductors 1 OPA2677 dual 250MHz op amp, SOIC-8 (IC1) 1 MAX660M switched capacitor voltage inverter, SOIC-8 (IC2) 1 74AC139 dual two-to-four decoder/multiplexer, SOIC-16 (IC3) 1 MCP1661T-E/OT boost regulator, SOT-23-5 (REG1) 1 TSV912(A)ID dual rail-to-rail output op amp, SOIC-8 (IC5) 1 2N7002 N-channel Mosfet, SOT-23 (Q1) 1 3mm red LED (LED1) 3 1N5711 RF schottky diodes, DO-35 (D1, D2, D7) 4 1N4148WS SMD signal diodes, SOD-323 (D3-D6) 1 MBR0540 50V 0.5A SMD schottky diode, SOD-123 (D8) Capacitors (all SMD M2012/0805 50V X7R 10% ceramic unless noted) 2 100μF M3216/1206 6.3V X5R 3 10μF 6.3V X5R 1 330nF 10 100nF 1 470pF NP0/C0G RF (high-Q) 1% 1 220pF NP0/C0G RF (high-Q) 1% 2 100pF NP0/C0G 100V RF (high-Q) 1% [DigiKey KGQ21HCG2D101FT; Mouser 581-KGQ21HCG2A101FT; element14 1856269] 1 33pF NP0/C0G 250V RF (high-Q) 1% [Johanson 251R14S330JV4T] Resistors (all SMD M2012/0805 1% unless noted) 1 10MW 1 120kW 5 27kW 2 1kW 1 220W 1 1MW 2 100kW 1 470W 1 180W 2 390kW 1 33kW 1 4.7kW 1 270W 1 47W 1/4W axial (R1) Hardware 2 M3 × 16mm brass hex spacers 6 3mm ID 3mm-long untapped spacers 4 M3 × 10mm blackened panhead machine screws and hex nuts 2 M3 × 8mm blackened panhead machine screws 2 M3 × 8mm nickel-plated or stainless steel panhead machine screws 2 M2 × 10mm panhead machine screws and hex nuts 1 8mm-long LED spacer 1 double-sided foam-core tape pad approximately 40 × 60mm (for battery holder) 2 100mm lengths of light-duty or medium-duty hookup wire (red & black) Extra parts for optional debugging interface 1 3-pin polarised header (CON2) 2 2N7002 N-channel Mosfets, SOT-23 (Q2 & Q3) 2 4.7kW SMD resistors, M2012/0805 1% 1 1kW SMD resistor, M2012/0805 1% siliconchip.com.au Australia's electronics Automatic LQ Metermagazine Kits (SC6939, $100 + postage) July 2024  33 Final assembly Includes everything in the parts list except the case & optional debugging parts. The Automatic LQ Meter measuring a moulded inductor. You can rerun the test with different resonant capacitance values to get measurements at various frequencies. battery negative wire to the DC socket’s ground tab. You can find the positive tab on the DC socket using a continuity tester touching the central pin in the socket. It will make a sound when the other lead touches the correct tab. The ground tab is trickier since many sockets incorporate a ground switch; make sure a plug is inserted in the socket (but no power is applied) and check for continuity with the outer barrel of the plug and one of the tabs. All that remains is to crimp (and possibly solder) two lengths of lightduty hookup wire into the polarised header plug and solder them in parallel with the battery leads. Make sure that when it’s plugged into the polarised header (CON1) on the PCB, ground goes to the bottom terminal and the positive supply to the upper terminal that connects to switch S1. There is no reverse polarity protection on the PCB, so if you get this wrong, smoke will escape! Double-­ check that you got it right when the wires are connected by the PCB by verifying continuity from the battery’s ground lead to one of the screw holes on the PCB and the outer barrel of the DC socket. Using it Using the LQ meter is straightforward. Just connect the unknown inductor and press START. If you have no idea what the inductance is, set the frequency to the highest (90MHz) and the capacitor value to 51pF. It will take a few seconds to run its scan and display the Q and inductance values. If you have a rough idea of the inductance, a lower top frequency will make the scanning faster. The calculation is according to the equation: f = √25330 ÷ LC ... where f is the frequency in MHz, L is the inductance in µH and C is the capacitance in pF. The constant 25330 takes into account those units, plus the various gain or attenuation factors in the circuitry, as well as the ADC range. The inductance of air-cored inductors will not vary much with frequency. However, the permeability of ferrite or iron cores varies with frequency, so you will get different values over the frequency range. The five-capacitance range of this unit is comparable to the variable capacitor in Q meters of the past. SC Fig.4: use the front panel PCB as a template to drill holes in the front panel; they should be close to the positions shown here. Once they have been located with a pilot drill, enlarge them to the sizes shown here. 34 Silicon Chip Australia's electronics magazine siliconchip.com.au Winter SAVERS Build It Yourself Electronics Centres® K 8600A 369 $ Winter! Stay inside and build this . Sale prices end July 31st BONUS! 1kg roll of black filament valued at $49.95 (K8397A) Great all rounder! Top buy for students & makers! Q 1089 SAVE $15 NEW! 50 79 $ $ Creality® Ender 3 V3 SE 3D Printer The Ender 3 is a compact 3D printer offering excellent print quality with a build volume of 22Wx22Dx25Hcm and is compatible with ABS, PLA and TPU filaments. Supplied mostly assembled and can be up and running within an hour. T 2090 5 Piece Plier & Cutter Set Multi-Display Auto Bargain 40W Soldering Station Ranging DMM. The perfect balance of value for money and features for beginners or cash strapped students and enthusiasts. Slim, lightweight handle with tip cleaning sponge and iron safety holder. Full range of spare tips also available. T 1463 SAVE 15% 50 $ Featuring a large reverse backlit display with two readout areas this compact meter offers an amazing array of test functions with AC/DC current & voltage, resistance, temperature, capacitance, frequency and more! Repair faster with a lithium screwdriver. The 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. Ultimate Helping Hands This space efficient work station for your benchtop has a sturdy metal base and packs in loads of features, including 4 flexible clamp arms, solder reel holder, dry tip cleaner, container of flux AND a magnifier lamp. SAVE $19.95 60 $ T 2127 A must have for any electronics enthusiast. Includes: • Side cutters. • Flat long needle nose pliers. • Flat bent needle nose pliers. • Long nose pliers/cutters. • Bull nose pliers/cutters SAVE 25% 22 $ T 2758A Tungsten Carbide PCB Drill Set A set of 10 PCB drill bits in a handy plastic carry case. Sizes: 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.1, 1.2mm. SAVE 16% T 2329 14 $ Includes storage case. T 1489 SAVE 24% 15 $ USB Dual LED Head Torch Weather resistant, USB rechargeable, & 120 lumens for JUST $15! Why pay $50 or more? Your local electronics store since 1976. | www.altronics.com.au Precision Knife Set Includes to handles and a variety of blades (13) to suit different cutting jobs. Great for hobbies! SAVE 24% X 0203 15 $ . P U R E W PO AMAZING VALUE! corner! Spring is just around the SAVE $54 175 $ 799 $ SL4576BT N 2087 20A Powerhouse® Solar DC-DC Battery Charger This dual input design connects to a solar panel and your cars alternator (12 or 24V) to provide charging for secondary batteries such as those used in campers, caravans and trades service vans/ trailers. Suitable for Lead Acid, AGM and Lithium Fe PO4 batteries. More BT batteries arriving this month! NEW Bluetooth equipped LiFePO4 batteries. Latest generation remote power batteries - now with Bluetooth monitoring for added convenience. Never run short of power by keeping an eye on your usage. LiFePO4 rechargeables offer longer service life & weigh HALF as much as SLA batteries. LiFePO4 also provide more usable life per cycle, allowing for longer run times by holding a higher voltage until capacity is almost exhausted. 5 year warranty. SAVE $20 Size: 360W x 185H x 138Dmm. SAVE $40 119 $ SAVE 15% 75 $ Q 0594 50A T 5099 199 $ Connect-It-All DC Power Hub A complete central power hub for your 4WD, caravan or camping remote power set up. 6 contura style illuminated rocker switches, 3 car accessory sockets (12V), USB charger, volt meter, protected by a 6 way weatherproof fuse block. N 2024A 20A SAVE $30 SAVE $54 129 $ 185 $ Q 0596 350A* N 2026A 40A The Ultimate Battery Fuel Gauge. Accurately measures battery voltage, current, power, real capacity and remaining run time of your connected battery (suits any chemistry, 8-120V). Includes shunt with 2m cable. 1% accuracy. *350A shunt pictured. Powerhouse® MPPT Lithium Ready Solar Regulator 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 2694 A central power hub for your vehicle! SAVE $20 129 $ Compact DC Power Hub & Isolator M 8521A SAVE 16% 29.95 $ Designed to manage power in your 12V or 24V vehicle. Measuring just 160x160x80mm, this box is packed with connections, including 50A Anderson style inputs and outputs, 60W USB charger, 2 x car accessory sockets. SAVE 33% 10 $ /m W 4200 Red W 4202 Black SAVE $20 49 $ Mains power from your cup holder. M 8070A This power inverter simply plugs into a cigarette lighter 12V socket & provides 240V power! 150W rated (450W surge). SAVE 25% 44 $ Home Battery Charger & Maintainer Offers hassle free maintenance charging for 6 & 12V lead acid batteries. Ideal for protecting seldom used vehicles from battery discharge. Croc clip or ring terminals. 600mA output. 3 Way Switch Panel S 2750 3 x 12V switches with red illumination and 15A DC breakers. CLEARANCE! 4 Gauge Power Cable Ideal for high current DC power distributiuon. 146A <at> 300VDC rated. Your one-stop electronics shop since 1976. | Order online at altronics.com.au Save on useful . S T E G D A G SAVE $26 Boost your o wireless audi range to 80m! 99 $ A 1107A SAVE $20 159 $ A handy benchtop claner! SAVE $20 119 $ With LED torch X 0103A Long Range Bluetooth® Audio Transceiver Transmit or receive Bluetooth 5.0 audio across distances up to 80m! Fitted with digital S/PDIF input and output for connection to the latest hi-fi equipment. Uses low latency technology so theres no lip sync issues! Powered by USB. Includes 3.5mm, S/PDIF, USB & RCA cables. D 0521 Blast away grime on jewellery, glasses and parts! This 60W ultrasonic cleaner uses water and household detergent, coupled with ultrasonic waves to clean jewellery, small parts, DVDs etc, without damage - no solvents required. Stainless steel 180x80x60mm tank. 110W Laptop Battery Bank Surprisingly compact battery bank with USB power delivery over 100W for large laptops! Features 2 PD Outputs and 2 QC3.0 outputs for simulataneous charging. Total capacity 30,000mAh. SAVE 10% 49 $ SAVE $50 S 9843B D 2321 Stay charged. Stay on time! 149 $ A stylish bedside or desktop alarm clock with in-built 15W wireless charging for your phone & FM radio. Display also shows calendar & temperature. A USB output is provided for recharging a secondary device such as your watch. S 9446D SAVE $20 Also includes magnetic balljoint bracket. Cable Free Wi-Fi Surveillance This handy 1080p camera can be installed just about anywhere indoors or out and has an in-built battery so you don’t need to run any cables! Offers 4-6 months of motion detect recording. When it’s flat, just take it off the wall & recharge via USB. Suits sheltered outdoor use. 79 Best seller! $ T 2183 Covert 1080p CCTV Recorder SAVE 22% Great for monitoring in remote locations, temporary CCTV monitoring etc. Runs off batteries, so its quick & easy to set up anywhere. Requires 8xAA batteries & 32GB SD card. 27 $ Jakemy® 106 Piece Precision Driver Set An affordable do-it-all servicing set with 92 4mm chrome vanadium bits, flexible extension bar, tweezers & magnetiser ring. For repairing phones, laptops & more! SAVE 19% 16 $ 1000’s sold! T 2266 Stay safe on the open road! Q 1300 SAVE 22% T 2480 40 $ All heat & no flame! Iroda® Pocket thermo-gun. Great for removing adhesives & heatshrinking. 650°C max. Refillable. Add butane gas for $9.50 (T2451). SAVE 22% 22 $ A must have travel accessory! A 1014 Universal Aircon Remote Lost your aircon remote? Or has your enthusiastic toddler destroyed it? This replacement works with hundreds of aircons. Handy Luggage Scales Don’t get caught over the limit at the airport! Measures up to 50kg. SAVE $20 155 $ Tyre Pressure Monitor System This solar powered TPMS unit sits on your dash and provides wireless monitoring of your tyre pressures. Provides high/ low pressure alarms, leak detection and temperature monitoring. Optional signal booster Q 1302 $95. Order online at altronics.com.au | Sale pricing ends July 31st. . T I E K A M 195 $ Z 6302K 8GB 145 BONUS! Gear for DIY Makers... $ 2 x 18650 batteries valued at $39.90. (S 4736A) Z 6302J 4GB 19.95 $ Z 6454 99.95 $ *BBC micro:bit not included Raspberry Pi Wi-Fi Pico ® The Raspberry Pi Pico W is perfect for IoT and wireless projects. An affordable 32-bit microcontroller with on board Wi-Fi. Z 0003 The latest generation Pi is here! With 2-3x the speed of the previous generation Pi. Grab yours for high powered single board desktop computing, IoT projects and more. SAVE 20% SAVE 20% 15 11 15 $ .95 K 9642 $ .95 K 9645 90° 310pc Jumper Header Kit Single row header connectors. Includes male & female pin headers, plus 2.54mm housings. 3mm and 5mm LEDs in green, red, blue, yellow and white. 300pcs. Arcade Switches SA0910-15 STEM bot is an easy to program 2 wheel obstacle avoidance and line tracking robot. Coding your program is easy using the standard BBC Micro:bit software. Simple construction with easy to folow instructions. Can also be controlled via Bluetooth. Ages 8+ SAVE 20% LED Assortment Pack SPST type. Red, Black, Green, Blue or Yellow. Raspberry Pi® 5 Build & code your own robot. Z 6424 NEW! 8 $ .95 15.95 $ 2.54mm Connection Kit 90° or straight boxed 2.54mm PCB connectors and plugs in 2, 3, 4 and 5 way. Plus crimp pins. 150pcs total. 60mm wheels with multi directional mecanum action. The classic atmega328p board! Includes headers. 9 K 9643 90° K 9641 Straight 90° boxed 1.25mm PCB connectors and plugs in 2, 3, 4 and 5 way. Plus crimp pins. 150pcs total. Arduino Pro Mini $ .95 $ 1.25mm Connection Kit Mecanum Wheels J 0115 Left J 0116 Right SAVE 20% .95 Pin to Croc Cables SAVE 38% 10pc jumper pin to croc clip cable set. 15 $ NEW! P 1024A 11 $ .95 Z 6222A Z 6338X Z 6381 SAVE 30% SAVE 24% 16 $ ea $ ESP8266 Development Board With Wi-Fi for plug and play connected projects. GPIO breakout pins, USB-serial interface & preflashed NodeMCU in one small package! 6 SAVE 17% Z 6422 2 Channel Relay Board 10A relays with 5V DC coil. Can be controlled by R-Pi, Arduino etc. Build It Yourself Electronics Centres® Sale Ends July 31st 2024 Find a local reseller at: altronics.com.au/storelocations/dealers/ Z 6339 BARGAIN! 5 $ ea 19 $ DC-DC Boost Module Allows a 3-34V DC input to be boosted up between 4-35VDC. 2A rated. Input/ output voltage display. DC-DC Converter Module Allows a 3.8-32VDC input voltage to be converted to 1.25-35V. Easy to inline connection. 2A continuous. Mail Orders: mailorder<at>altronics.com.au Victoria Western Australia » Springvale: 891 Princes Hwy » Airport West: 5 Dromana Ave 03 9549 2188 03 9549 2121 » Auburn: 15 Short St 02 8748 5388 » Virginia: 1870 Sandgate Rd 07 3441 2810 » Prospect: 316 Main Nth Rd 08 8164 3466 » Perth: 174 Roe St » Joondalup: 2/182 Winton Rd » Balcatta: 7/58 Erindale Rd » Cannington: 5/1326 Albany Hwy » Midland: 1/212 Gt Eastern Hwy » Myaree: 5A/116 N Lake Rd 08 9428 2188 08 9428 2166 08 9428 2167 08 9428 2168 08 9428 2169 08 9428 2170 New South Wales Queensland South Australia © Altronics 2024. E&OE. Prices stated herein are only valid until date shown or until stocks run out. Prices include GST and exclude freight and insurance. See latest catalogue for freight rates. *All smartphone devices pictured in this catalogue are for illustration purposes only. Not included with product. B 0007 Please Note: Resellers have to pay the cost of freight & insurance. Therefore the range of stocked products & prices charged by individual resellers may vary from our catalogue. Review by Tim Blythman Raspberry Pi 5 Originally designed as a cheap computer for use in education, Raspberry Pi single-board computers (SBCs) have been used in a vast range of applications. It’s just on five years since the release of the Raspberry Pi 4, and we finally managed to get a Raspberry Pi 5 to test and review. S ince 2012, we have seen the release of a new Raspberry Pi SBC (single-board computer) every year or so. However, there was quite a gap between the Raspberry Pi 4 and the Raspberry Pi 5, which wasn’t helped by the component shortages of the last few years. In 2021, the Raspberry Pi Foundation released the Pico microcontroller board, based on the RP2040 ARM microcontroller, followed by a Pico W variant with WiFi and Bluetooth capabilities. The inexpensive Picos have been embraced by the Arduino, Micropython and Micromite communities. We reviewed the Pico in December 2021 (siliconchip.au/Article/15125). siliconchip.com.au We have used it in numerous projects because of its low price and ease of use. The documentation for the Raspberry Pi Pico is written with the intention of using a Raspberry Pi computer as the development machine. With this in mind and many recent SBCs being touted as replacements for desktop machines, we’ll consider the Pi 5’s suitability for this task. 2021 also saw the release of the Raspberry Pi Zero 2 W, the most recent iteration of the compact Zero form factor SBCs and the first Zero with a 64-bit processor. It is based on the processor from early versions of the Raspberry Pi 3 but uses a system-in-­ package (SIP) known as the RP3A0. Australia's electronics magazine This combines the processor and RAM into the space-saving package needed to create a Zero board. The fact that the Raspberry Pi Foundation is now producing its own silicon (both for the Pi Zero 2 W and the Picos) is a notable advance. The Raspberry Pi 5 also includes an RP1 I/O controller, another of their products. We’ll delve into the RP1 and other Raspberry Pi 5 features shortly. The Pi 5 The Raspberry Pi 5 was released in September 2023, with the 4GB RAM variant being available first. There is also a version with 8GB of RAM. Interestingly, the Pi 5 drops the Model B July 2024  39 Table 1 – comparison between the ROCK 4C+, Raspberry Pi 4B & Pi 5 ROCK 4C Plus Raspberry Pi 4B Raspberry Pi 5 RockChip RK3399T (6 cores) Dual 1.5GHz ARM-Cortex A72 + Quad 1.0GHz ARM-Cortex A53 1MB + 512KB L2 caches BCM2711 (4 cores) Quad 1.8GHz ARM-Cortex A72 1MB L2 cache BCM2712 (4 cores) Quad 2.4GHz ARM-Cortex A76 512kB L2 cache per core 2M L3 shared cache Processor (CPU) 600MHz Mali T860MP4, four shaders, 256KB L2 cache 500MHz VideoCore 6, 1MB L2 cache shared with CPU cores 800MHz VideoCore 7, 2MB cache GPU two micro-HDMI, up to 4K + 2K (60Hz with one or both) two micro-HDMI, up to 4K + 4K (60Hz with one or 30Hz for both) 2 micro-HDMI, up to 4K+4K (60Hz with one or both) Display output HD stereo, up to 24bit/96kHz Stereo, PWM-based None Audio output 4GB 1GB, 2GB, 4GB or 8GB 4GB or 8GB RAM 5V/3A, USB-C or pin header 5V/3A, USB-C or pin header 5V/5A, USB-C or pin header Power req. 2× USB2, 2× USB3 2× USB2, 2× USB3 2× USB2, 2× USB3 USB 1× Gigabit 1× Gigabit 1× Gigabit Ethernet 802.11 b/g/n/ac (WiFi 5) Bluetooth 5.0 u.FL antenna 802.11 b/g/n/ac (WiFi 5) Bluetooth 5.0 PCB antenna 802.11 b/g/n/ac (WiFi 5) Bluetooth 5.0 PCB antenna Wireless 40-pin header: 1× PWM 2× SPI channels 2× I2C channels 1× ADC (analog) channel 40-pin header: 4× PWM 2× SPI channels 2× I2C channels 40-pin header: 4× PWM 2× SPI 2× I2C I/O suffix used for previous models. Given that there was no Model A for the Pi 4, it makes sense that the designations have been streamlined. We are reviewing the 4GB Pi 5 board. Table 1 shows a comparison between the Pi 4B, Pi 5 and the ROCK 4C+ SBC that we reviewed in April 2024 (siliconchip.au/Article/16210). The latter is roughly on par with the Pi 4B, although it includes a few nice features that the Pi 4B lacks. On the other hand, the Raspberry Pi machines have better software support and a larger community. Unsurprisingly, the newer Pi features a faster processor than the 4B. Most benchmarks indicate that the Pi 5 runs at least twice as fast as the Pi 4B. It is an ARM Cortex A76 in the form of a Broadcom BCM2712, which implements the ARMv8.2-A 64-bit instruction set. Not only is the processor faster, but the microSD card interface on the Pi 5 is capable of running twice as fast as that on the Pi 4B, and the Ethernet interface also transfers data faster. The GPU in the Pi 5 can also drive two 4K displays at 60Hz, compared to the Pi 4B, which can only drive one 4K display at 60Hz. The main compromises are the power and cooling requirements, with 40 Silicon Chip the Pi 5 now specifying a 5V 5A (25W) supply over the Pi 4B’s 5V 3A (15W) supply. Our Pi 5 happily booted up with the 3A supply we had been using for our Pi 4B and ROCK 4C+, although it showed a message that ‘power to the peripherals will be restricted.’ Screen 1 shows the initial desktop with this message. An official 27W Raspberry Pi power supply offers USB-C PD (power delivery), including 9V, 12V and 15V output voltages. Curiously, the output specified for use with the Pi 5 is 5.1V. Hardware Photos 1 & 2 are close-ups of the front and back of the Pi 5 with various features marked out. The overall layout is much the same as earlier models, although it is different enough that it will not fit in cases designed for earlier models. There is little of interest on the back except the microSD card socket. The main layout difference from the Pi 4B is the transposition of the USB and Ethernet connectors. The mounting holes and GPIO headers are in the same locations, and the other main external features are in much the same, if not identical, positions. Like the Pi 4B, the USB connector for power input is a USB-C type, and adjacent are two micro-HDMI (HDMI Australia's electronics magazine type D) sockets to allow dual monitor connections. The Pi 3B and earlier models have a single full-size HDMI socket and one micro-USB socket. You’ll need a cable with a microHDMI plug rather than an adaptor to use both HDMI sockets since the adaptor will likely foul the USB-C socket. Our basic single-monitor setup worked using the HDMI socket (with an adaptor), HDMI1, further from the USB-C socket. The top of the Pi 5 looks quite sparse; many of the passive components are on the back of the board. The main processor is the larger chip with a metal shield (we attached an aluminium finned heatsink to it, visible in the photos); the rectangular chip next to it is the RAM. The second shielded package is the radio module, providing WiFi and Bluetooth connectivity. The RP1 ‘southbridge’ I/O controller is the large chip with the Raspberry Pi logo near the USB sockets. This is one of the ICs the Raspberry Pi Foundation has designed and produced. The RP1 connects to the processor via a fourlane PCIe 2.0 interface. Bundling many of the I/O functions into a single chip allows substantial performance improvements for the Pi 5 over the Pi 4B. The RP1 even handles GPIO functions on the 40-pin header siliconchip.com.au GPIO Header RAM Chip RP1 Chip Fan Connector WiFi Module 2x USB2 PCB Antenna ARM Processor 2x USB3 PCIe Power Switch Status LED Ethernet USB-C (Power) PoE HAT Header RTC Battery 2x HDMI Composite Video 2x MIPI CSI/DSI Connector Photo 1: the Raspberry Pi 5 is the same size and shape as its predecessors, but the connectors have been slightly rearranged, so it requires a different case. The official case includes a small fan that provides much-needed cooling. The same GPIO pinout applies as the previous Pis, so most existing HATs should work with the latest Pi. The supplied RAM is indicated with a component fitted to the MEMORY box. and has been designed to provide the same I/O functions as the Pi 4B. The RP1 provides Gigabit Ethernet, two USB 3 interfaces, two USB 2 interfaces and two MIPI transceivers for cameras/displays on the J3 and J4 CSI/DSI connectors. The RP1 also includes the versatile PIO (programmable input-output) peripheral and an ADC (analog-to-digital converter). These latter two features are not used on the Pi 5. The RP1 relieves the main processor of most of the peripheral duties. There is more information on the RP1 at siliconchip.au/link/abvc The Pi 5 dispenses with the 3.5mm TRRS socket used for audio and composite video in earlier versions. Instead, video is available from a dedicated two-pin header (marked as VID next to HDMI1). Two of the GPIO pins on the 40-pin header can produce PWM-based audio, although this does not appear to be enabled by default. The top of the board also has a four-pin PoE (Power-over-Ethernet) header for connecting to a PoE HAT. HAT (hardware attached on top) is the Raspberry Pi terminology for a shield or daughterboard. The top of the Pi 5 also breaks out a four-pin polarised header (J17) for a fan. An active cooler is available to siliconchip.com.au suit the Pi 5, which can connect to J17. The active cooler mounts to two holes adjacent to the four main mounting holes. The official case for the Pi 5 also incorporates a fan that can be powered from J17. A three-pin polarised header (J16), labelled UART, can be used for diagnostics. The Renesas DA9091 PMIC (power management integrated circuit) is near the USB-C socket. It incorporates a real-time clock (RTC) feature that utilises an optional battery connected to the nearby J5 polarised header. The two-pin header pads marked J2 next to J5 are connected in parallel with a momentary pushbutton (marked PSW) used as a power switch. It is adjacent to a bicolour LED labelled STAT. J20 is a flexible flat cable (FFC) connector marked as PCIe that breaks out a single PCI Express 2.0 lane. It is Photo 2: the underside of the Pi 5 is populated mainly by passive components. Australia's electronics magazine July 2024  41 expected that future HAT designs will use this interface, and it is suggested that this will be most commonly used for connecting an NVMe solid-state drive (SSD) for storage. The back of the board is mainly populated with passive components and the microSD card socket that holds the operating system. There are also options to configure a boot EEPROM to allow booting from a USB storage device or an NVMe SSD. Setting it up Like just about every other SBC, the Pi 5 typically uses a microSD card for the operating system and user files. Hence, installation involves transferring a disk image to the card using another computer. The Raspberry Pi Foundation provides the Raspberry Pi OS, which is based on Debian Linux. Operating system downloads can be found at siliconchip.au/link/abvd and that page indicates which versions are compatible with which Pi boards. There are bundles pre-loaded with different programs. We used the latest version (v5.2, March 2024), which includes all the recommended software. This download comes to around 3GB and expands to a 15GB file. A 32GB card is recommended. We previously used WinDiskImageWriter to transfer the image files to the microSD card, but this time, we tried Raspberry Pi Imager, which has been available since 2020. This, as well as other software, can be downloaded from www. raspberrypi.com/software Screen 2 shows the Imager program. It can automatically download card images as well as write previously downloaded files. Imager can also configure the image with settings like WiFi, country and SSH, allowing the Pi to operate in headless mode (without a keyboard, mouse or monitor). Writing the file and verifying the image took about half an hour; the verification is a nice touch. The Imager is a good way to see what other software is available. It lists media player and emulation images, among others. Even if you don’t have a Pi, we suggest downloading Imager to see what other people are doing with their Pi. Once the image is transferred, the Pi 5 is booted by installing the card, connecting the monitor, keyboard, and mouse, then plugging in the power supply. The first boot sets up a few things and performs a system update. Once everything was set up and the update completed, the Pi 5 responded quickly. A reboot took about 15 seconds, comparable to modern computers fitted with SSDs. Using it The Raspberry Pi Foundation does a good job of making their software easy to use; the mix is much the same as earlier distributions. Educational programs like Scratch, Mathematica and Wolfram are included, as is Thonny (an integrated development environment [IDE] for the Python programming language). All these programs would be familiar to seasoned Pi users. We then looked for programs that would be useful in a typical office environment. The LibreOffice suite (including word processor and spreadsheet) was installed, as were the Chromium and Firefox web browsers. Many of the included programs may not be familiar if you have previously only used Windows or macOS. However, they will be known to those familiar with open-source alternatives to proprietary programs. Even the open-source KiCad EDA (electronics design automation) suite is installed. The Arduino IDE is not installed by default, but it and many others can be added through the Preferences → Add/Remove Programs dialog box. Using the Arduino IDE on the Pi 5 was practically the same as on the Windows machines we are used to. Some programs we use, like Altium Designer, are only available for Windows operating systems. Although the MPLAB X IDE is available for Linux (and Raspberry Pi OS is a Linux variant), currently, it only works on x86 and x64 processors and not ARM processors. We were able to program a Pico from the Pi 5 from a command line interface with relative ease. So, a good proportion (but not all) of the programs we use daily are available or easy enough to install on the Pi 5. ARM processors are becoming more common on portable and desktop computers, such as M2-based Mac computers or Microsoft Surface devices with an SQ2 processor. We expect support for ARM processors to grow steadily; hopefully, that will translate to better software options for computers like the Raspberry Pi. Screen 1: the initial desktop after setting up the Pi 5; it looks much the same as previous versions. The messages at top right indicate that it has connected to a preconfigured WiFi network and that the connected power supply cannot provide the 5A needed for full functionality. 42 Silicon Chip Australia's electronics magazine siliconchip.com.au Still, the appearance and functionality are similar. The Desktop software lacks broad hardware support, so we couldn’t fully use the PC’s features. In particular, WiFi would not work, so we had to devise an alternative way to connect to the internet using a USB dongle. If you have an old PC, Raspberry Pi Desktop could be an easy way to try out the Raspberry Pi OS. Be aware that the flash drive and your PC’s hard drive could be erased if you do that. Conclusion Screen 2: the Raspberry Pi Imager is a helpful tool for setting up the microSD card and seeing what other disk images are available. Initially, we ran this on a Windows computer but it comes preinstalled on the Pi. The performance of the Pi 5 was generally quite good, and the system seemed responsive. The processor gets very hot, though; too hot to touch, so one of the cooling options would be beneficial. Raspberry Pi Desktop An interesting footnote we found on the www.raspberrypi.com/­software/ operating-systems page is Raspberry Pi Desktop. It’s an operating system image for PC and Mac computers (those with x86 or x64 processors) that provides a Linux environment similar to that found on the Raspberry Pi boards. We loaded this onto a USB flash drive with the Rufus program (https:// rufus.ie/en/), a utility that can be used to create bootable flash drives. We plugged the drive into an older PC and booted it up. The flash drive can install the Raspberry Pi Desktop operating system to the hard drive (so you don’t need to boot from the flash drive). Alternatively, you can run it directly from the flash drive. Screen 3 shows the desktop environment and program installation. The Raspberry Pi Desktop is based on Debian 11, an older version than the Debian 12 used in current versions of Raspberry Pi OS (for the Pi SBCs). With ARM chips gaining a foothold in the market traditionally held by x86 and x64 processors, software availability for computers like the Raspberry Pi can only grow. The Raspberry Pi Foundation is now producing some of its own chips; that’s a promising sign, and we look forward to their future developments. While it’s still no match for most PCs, the Raspberry Pi 5 works well enough to do many of the daily tasks that the average person needs. Various programs are still unavailable for ARM Linux, so a Windows PC will remain our tool of choice for the foreseeable future. Still, the Pi 5 makes a great second machine and is well-priced as an educational computer for children. It’s also an excellent way to try out Linux if you haven’t done so already. The Raspberry Pi 5 and its accessories are available from Altronics (Z6302J for the 4GB version and Z6302K for the 8GB version), as well SC as Mouser and DigiKey. Screen 3: Raspberry Pi Desktop is a version of the Raspberry Pi OS for x86 and x64 computers. It is a good way to try out the Raspberry Pi environment, although the hardware support is not as good as on the Pi boards (or your average PC Linux distribution). The latest version of Raspberry Pi Desktop is also a couple of years old now. siliconchip.com.au Australia's electronics magazine July 2024  43 180-230V DC Moto Controls 180-230V DC motors rated from 1A to 10A (¼HP to 2.5HP) Controlled by four common op amp ICs with one opto-coupler and three linear regulators Zero to full speed control Safe startup procedure Emergency cut-out switch facility Automatic over-current switch-off Optional reversing switch capability PWM, Live and Power indicator LEDs Rugged diecast aluminium enclosure Current and back-EMF monitoring for speed regulation under load Initial setup adjustments can be done with a low-voltage supply DC motors from 180-230V DC power various pieces of equipment; they are particularly common in treadmills. Often, these motors are removed from the treadmill, possibly because the speed controller has failed. Such motors can be reused for other purposes, such as adding computer control to a lathe. Many of these motors are sold via the internet on sites such as eBay, often inexpensively. The type of DC motor we are referring to here typically has permanent magnets in the stator and field coils for the rotor. The electrical connection to the field coils is made via a commutator and brushes. These motors can be powered from the mains using a full-wave bridge rectifier to convert the 230 AC voltage from the mains to a pulsating DC voltage, where the voltage rises and falls in a sinewave shape over each half of the mains waveform. The resulting average voltage is close to 230V DC. Warning: Mains Voltage This Speed Controller operates directly from the 230V AC mains supply; contact with any live component is potentially lethal. Do not build it unless you are experienced working with mains voltages. 44 Silicon Chip Scope 1 shows the resulting ‘DC’ waveform that is applied to the motor for it to run at full speed. If you want to slow the motor down, you need a speed controller. Our DC Motor Speed Controller provides the same 100Hz pulsating DC mains rectified voltage as a full-wave rectifier, but it adds speed control by switching this waveform on and off more rapidly, at around 900Hz. The averaged DC voltage is thus multiplied by the proportion of time it is switched on. This type of drive is called pulsewidth modulation (PWM), where voltage is applied to the load in a series of pulses with a duty cycle (0-100%) that can be adjusted to control the average voltage applied to the motor. With the duty cycle set at 100%, the motor is driven with the full 100Hz full-wave rectified mains voltage. As the duty cycle reduces, so does the average voltage applied to the motor. A 50% duty cycle reduces the average voltage by one-half. Scope 2 shows the PWM-chopped full-wave rectified mains with it on about 60% of the time (a 60% duty cycle). The resulting averaged voltage should be about 60% of the full waveform shown in Scope 1. The mean measured value of 113V is 58% of the 195V reading from Scope 1. Australia's electronics magazine We published a 230V/10A Speed Controller for Universal Motors (February & March 2014; siliconchip.au/ Series/195) that could be used to drive a 180V DC motor. However, as that controller was intended for use with universal motors, it lacks some desirable features for use with DC motors. Our new design By designing a controller purely for DC motors, we can provide the best control of this type of motor. For example, the Universal Motor Controller mentioned earlier has feedback to maintain motor speed under load, but it only monitors the motor current. That type of feedback control does not monitor motor speed and relies on the motor current being indicative of motor speed and load. One problem with that is that, under load, the motor speed drops and it draws more current. This extra current causes the controller to increase the PWM duty cycle to increase the motor speed. That increases the current further, so the controller increases the duty cycle further. The process can quickly become unstable, possibly producing bursts of voltage to the motor. That is especially likely to happen if too much motor speed correction is applied. siliconchip.com.au or Speed Controller High-voltage DC motors are commonly used in lathes, consumer-grade treadmills, industrial conveyor belts and similar equipment. This Speed Controller can control such a motor over a wide range of speeds, from very slow to full speed. A constant speed is maintained even with a varying load due to motor-generator voltage (back-EMF) sensing and current feedback circuitry. Part 1 by John Clarke In this new design, we also monitor the motor speed using back-EMF (electromotive force). The motor generates this back-EMF from its rotation and it directly indicates the motor speed – see Scope 3. The generated voltage will drop when the motor is loaded, since it will slow down. Increasing the duty cycle of voltage applied to the motor can compensate for this reduction in motor speed. This is a negative feedback control, so it is far more stable than feedback based purely on motor current. Compare Scope 3, with the motor driven by switched PWM, to Scope 2. You can see that in Scope 3, the motor load also generates a voltage (backEMF) during the PWM off-times. This is seen as the spikes below the baseline indicated by the small arrow and “1” on the left. The motor is generating spikes of around -85V. Applying a combination of both current and back-EMF feedback control, with a measured amount of each, provides excellent speed control without the risk of speed instability. Incidentally, we don’t use backEMF speed control for a universal motor controller because that voltage is essentially non-existent. The generated voltage relies on having a magnetised core. Since the universal motor has windings for both the armature and stator, there is little remnant magnetism when the windings are not powered. Hence, little to no voltage is produced. However, with a DC motor, the stator can comprise permanent magnets or powered windings, so a voltage is induced in the rotor as it spins. Scope 1: this is the pulsating ‘DC’ waveform that is applied to the motor for it to run at full speed, created by rectifying the 230V AC mains. Note that the mean (average) voltage is lower than the RMS. This waveform (like the others) was captured using an isolated differential probe. Scope 2: the full-wave rectified mains being switched on and off with a duty cycle of about 60% (60% on, 40% off). As expected, the resulting average voltage is about 60% of the full waveform in Scope 1. Scope 3: the motor voltage when driven by the switched PWM version of the full-wave rectified mains waveform. You can see how the motor load generates negative voltage spikes of about -85V (back-EMF) during the PWM off periods while the motor is not being driven. siliconchip.com.au Australia's electronics magazine Design inspiration We based our new controller on the features of a small lathe controller (the Sieg C1 micro lathe) that was July 2024  45 extensively documented by our frequent contributor, Dr Hugo Holden. Hugo provides considerable detail on the operation of that circuit, as well as a description of how it uses op amps, in a PDF on his website at siliconchip. au/link/abmn While the Sieg controller is for small motors rated up to 1A, we have designed ours to handle any high-­ voltage DC motor from 1A to 10A. The Sieg motor controller includes several safety features we also incorporated into our new design. One such feature is the facility for a safety/emergency stop switch, where the motor does not start when the switch is open and will shut off if it opens during operation. This feature does not need to be used; it can be bypassed with a wire loop if not required. However, it’s a great idea to include such a switch for a lathe or any other piece of industrial equipment. A lathe can be set up so it will not run unless a safety shield is in place to protect against flying debris. In other cases, the safety switch can be a large, red emergency stop button (eg, Jaycar SP0786). You could even have both simply by wiring them in series. Whenever one opens, the motor will stop. There are other conditions under which the motor won’t run for safety. For example, when the controller is initially powered up or if the motor becomes overloaded during operation. The motor cannot start or be restarted unless the speed control is brought back to the full anti-clockwise position before being advanced to start the motor running. That assumes the safety switch is closed and there is no motor overload. The Sieg controller does this with a switch incorporated within the speed potentiometer. That is not terribly unusual as many vintage radios and some music instrument amplifiers have a power switch within the volume potentiometer, where rotating the potentiometer fully anti-clockwise opens the switch to disconnect power. However, in the Sieg controller and our design, the switch must be closed when the potentiometer is fully anti-clockwise and open when it is advanced clockwise. That is the reverse of a radio/amplifier potentiometer, making it a rather unique and difficult part to obtain. Even if we decided to use that style of potentiometer, with (say) a relay to reverse the switching sense, it still would not be ideal. That’s because volume control potentiometers have a logarithmic resistance change over their rotation, while we need a linear response. Therefore, we use a standard potentiometer that does not have a switch incorporated and instead provide the switching feature using a comparator and relay. That avoids having to source a special type of potentiometer. The comparator monitors the potentiometer’s wiper voltage and switches a relay according to the potentiometer position. How it works The basic block diagram is shown in Fig.1. The circuitry is based around a full-wave bridge rectifier that provides the 100Hz half-sinewave ‘DC’ voltage from the mains. The IGBT that switches power on and off to the motor is connected in series with the motor, between the DC terminals of the bridge rectifier. The IGBT’s gate is driven by PWM circuitry that controls the motor speed. The duty cycle of the PWM signal depends on the position of the speed potentiometer (VR1), the motor current (detected by sense resistors) and the motor with both back-EMF from its negative terminal. The PWM signal is generated by comparing voltage Vo (derived from the speed potentiometer voltage and the current and voltage feedback) to a sawtooth waveform. Vo comes from the output of op amp IC2c, while the Fig.1: the voltage from speed control pot VR1 (at left) is buffered and mixed with the motor current and speed feedback signals, then fed to comparator IC4a. Comparing that DC voltage to a sawtooth waveform produces a PWM output with a duty cycle proportional to the control voltage. That goes to the IGBT, which switches voltage to the motor. The other components are associated with the over-current shutdown, safety switch and power-up inhibit functions. 46 Silicon Chip Australia's electronics magazine siliconchip.com.au sawtooth waveform is from the oscillator based on IC4b. IC4a compares the sawtooth waveform to Vo and generates the variable duty-cycle PWM output. The manual speed control voltage from potentiometer VR1 is buffered via IC2a, which applies a DC voltage to the N1 node, at the input to IC2b. Voltages from the feedback loop outputs, from IC3b and IC2d, are also applied to N1. These provide the speed (Vs) and torque (Vt) signals, which are derived from the motor back-EMF and the motor current, respectively. For torque feedback, the motor current is determined by the voltage across the current sense resistance in series with the motor’s positive terminal. The resulting voltage is amplified by IC1b, which has its gain set by trimpot VR2. The gain is adjusted according to the motor current rating; more gain is used for low-current motors and less for higher-current motors. The resulting voltage is applied to op amp IC2d, with an offset voltage set via trimpot VR3. The motor’s back-EMF voltage is amplified by IC3c and is offset using trimpot VR7 before being applied to op amp IC3b. IC3c provides a low-pass filter with a roll-off point of 1.6Hz to remove the 100Hz ripple. The torque voltage from IC2d (Vt) and the speed signal from IC3b (Vs) are summed with the speed control signal at N1, while a separate torque signal is applied to the N2 junction of the main servo amplifiers, IC2b and IC2c, via VR4. This provides feedback adjustment, allowing for the best speed control as the motor is loaded. For current overload detection, the motor current is amplified by op amp IC3d and compared against a threshold by comparator IC3a (it’s actually an op amp but used in open-loop mode, so it acts as a comparator). If the current limit is exceeded, the motor is switched off by the shutdown relay, RLY2, which switches off RLY3. That disconnects power from the motor. These relays return to their normal positions when the overload is cleared and the speed control is rotated fully anti-clockwise. IC1a detects when the speed control pot is set at zero. It compares the buffered speed control voltage from IC2a against a 119mV reference. The Relay (RLY1) is powered, opening the connected contacts, when the voltage siliconchip.com.au This is the motor we used for testing, You can find lots of similar second-hand and new motors on eBay. from the speed control is higher than 119mV. The open relay contacts prevent the motor from starting, although it will continue to run if it is already running. Other relays are connected in series with the coil of the one that powers the motor, so if any of them open, that will stop the motor from spinning. That includes if an overload is detected or the emergency/safety switch is open. For the motor to run, the speed potentiometer must be set almost fully anti-clockwise (switching RLY1 off), then rotated more clockwise to start the motor. The circuit has three main supply rails: a ±12V split supply and a +15V supply. The +12V supply connects to the positive output of the full-wave bridge rectifier, so the whole circuit operates at mains potential. Level shifting between IC4a’s output and the gate of IGBT Q1 is via an opto-coupled Mosfet/IGBT gate driver (IC5). Circuit details The full circuit is shown in Fig.2. It comprises five ICs, an IGBT, several diodes, two transistors, three relays plus numerous capacitors and resistors. A significant component for PWM drive generation is op amp IC4b, wired as a sawtooth oscillator (near the bottom of the diagram). This is the second amplifier in dual LM833 op amp IC4. It is powered from the ±12V supply. A bias voltage is generated using two 100kW resistors connected in series across the ±12V supply. The centre connection of this voltage divider is Australia's electronics magazine connected to the non-inverting input, pin 5 of IC4b. The voltage at pin 5 would be 0V, except for the fact that there is also a 47kW resistor between pin 5 and the op amp output, pin 7. To calculate the voltage at pin 5, the two 100kW resistors connected across the ±12V supply can be considered a 50kW resistor between pin 5 and 0V. That leaves a 47kW/50kW voltage divider, with the voltage at the end of the 47kW resistor shifting between about 10.9V when pin 7 is high and -10.9V when it is low. So when the op amp output is high, the 47kW resistor pulls pin 5 to around 5.6V, and when the op amp output is low, it is around -5.6V. The output oscillates between the high and low states due to the 10nF capacitor at the inverting input (pin 6) and the charge and discharge resistances between pins 6 and 7. When power is first applied, the 10nF capacitor is discharged, so pin 6 is near -10.9V. Pin 5 is at a higher voltage than that, so pin 7 goes high. Pin 5 then sits at around 5.6V. The 10nF capacitor charges via the 1kW resistor and diode D4. As soon as the capacitor at pin 6 charges just beyond the pin 5 voltage, pin 7 goes low, to -10.9V, since the inverting input voltage is above the non-inverting input. Pin 5 is then at -5.6V, and the capacitor discharges via the 91kW resistor. Diode D4 is reverse-biased and does not take part in the discharge cycle. Once the capacitor voltage exceeds -5.6V, the pin 7 output goes high again, and the process repeats. So the 10nF capacitor is charged quickly via the 1kW resistor and July 2024  47 diode D4, then discharged much more slowly via the 91kW resistor. The resulting waveform is described as a sawtooth shape, rising quickly and falling more slowly. The waveform ranges from about +5.6V to -5.6V at about 900Hz. 48 Silicon Chip The sawtooth waveform at pin 6 of IC4b goes to the inverting (pin 2) input of IC4a, which compares it against the Vo voltage from IC2c (at its non-inverting pin 3 input). The output of IC4a goes high (to around 10.9V) when the voltage from the IC4a oscillator is Australia's electronics magazine lower than the feedback voltage; its output is low (-10.9V) otherwise. IC4a’s output is therefore a rectangular waveform with a higher duty cycle (higher for longer) when the Vo voltage is higher. IC4a’s output drives optically-coupled Mosfet/IGBT driver siliconchip.com.au Fig.2: virtually all the op amp based control circuitry is on the left half of the diagram; 12 op amp stages are used, inside four ICs, two of which operate as comparators. The IGBT and its driving circuitry are at lower middle, with the current measurement shunts and relays that switch power to the motor above that. The linear power supply is at upper right. IC5. It level-shifts IC4a’s square-wave output to a voltage suitable for driving the gate of Q1. Scopes 4 & 5 show the sawtooth waveform and Q1’s PWM gate-drive signal at duty cycles of about 10% and 90%. The top trace (yellow) is the PWM signal, while the lower cyan trace is the sawtooth waveform, which oscillates between ±5.6V. The yellow horizontal dotted line (representing Vo) shows how the PWM output is high when the sawtooth waveform is below that level. Scope 6 shows the PWM drive to Q1’s gate. The rise and fall times of the waveform are 1.37μs and 1.2μs, respectively, with the gate voltage siliconchip.com.au reaching 14.8V. We want the switching time to be short to minimise heating in Q1 during partial conduction periods, and the gate drive needs to be within its ±25V rating while being high enough to fully switch it on, which occurs at around 15V. The 15V supply for driving the gate comes from the positive output of the bridge rectifier via diode D2 and four series/parallel 22kW 1W resistors. This provides an average of 5mA to IC5 and the 15V zener diode (ZD2). That supply is smoothed to a DC voltage by a 100μF capacitor with parallel 1μF and 100nF ceramic capacitors so that IC5 can receive bursts of current for driving Q1’s gate when needed. Australia's electronics magazine LED1 is also driven by this 15V supply, so it indicates when power is being fed to the circuit from the positive side of the bridge rectifier. When lit, the entire circuit is at mains potential. IC5 has an internal LED between pins 1 and 3, and its output goes high when that LED is powered via the 620W resistor from the output of IC4a and diode D3. That diode prevents a negative voltage from being applied to IC5’s internal LED when IC4a’s output goes negative. The 620W series resistor limits the LED current to around 10mA. Q1’s gate is driven via a 75W rate-limiting gate resistor. July 2024  49 The IGBT is protected from overvoltage at switch-off by transient voltage suppressor TVS1, which conducts if Q1’s collector goes above 400V, causing the gate to be pulled high. That switches on Q1 to shunt the excess voltage. The 10W series resistor limits the current into protective 15V zener diode ZD1, which prevents the gate voltage from going beyond Q1’s maximum limits. Apart from TVS1, any high voltage spikes are also coupled via diode D1 into the 47nF capacitor connected across the + and – terminals of BR1. The capacitor absorbs some voltage transients. A snubber between the IGBT’s collector and emitter terminals prevents switch-off oscillations. It comprises a 47nF capacitor and two paralleled 470W 5W resistors. Inductor L1, in series with the motor, limits the current rise rate when the motor is switched on to protect the IGBT from excessive surge current. It also helps to filter the 900Hz switching drive for the motor, reducing electromagnetic interference (EMI). The motor current is monitored via a set of 0.022W shunt resistors. The four 0.022W 3W resistors are connected in series/parallel, yielding a 0.022W 12W resistance. They connect between the bridge rectifier positive terminal and L1. This shunt produces a voltage at the lower end that is below the +12V supply, in proportion to the current flow. The motor voltage is monitored via a 220W 1W resistor from the motor’s negative terminal. This voltage is also negative with respect to the +12V supply and becomes more negative with increased back-EMF voltage. The circuitry for monitoring these feedback voltages will be described separately. Current feedback IC1b amplifies the voltage across the current measurement shunt resistors. Its gain can be adjusted between 1.5 times (with trimpot VR2 at minimum resistance) and 13.1 times, when the resistance between pins 6 and 7 of IC1b is 5.22kW. This allows the circuit to work with motors rated between 1A and 10A with the correct overload threshold. Note how IC1 has a 15V positive supply rather than 12V. This means the positive supply for IC1 is 3V above the 12V that the shunt resistor is referenced to. That way, the op amp output can reach 12V when there is no voltage across the shunt. Even though the op amp is a rail-torail type, where the input and output voltages can be up to the supply rails, there will be some differences due to the input offset voltage of the op amp and the fact that the output can only reach within a few millivolts of its supply rails. So, with IC1’s positive supply above 12V, the op amp has the headroom to handle 12V signals. The output voltage from IC1b is amplified and inverted by op amp IC2d, which has a gain of -6.8, determined by the 10kW input resistor and the 68kW feedback resistor. This amplifier also acts as an integrator, filtering out the PWM signal and the pulsating DC due to the 220nF capacitor connected across the 68kW resistor. The resulting low-pass filter has a roll-off point at about 10.6Hz, well below the 100Hz of the rectified mains and the 900Hz PWM signal. IC2d’s output is applied to the N1 node via a 33kW resistor. Scope 4 & 5: these captures show the sawtooth waveform and the Q1’s PWM gate drive signal for duty cycles of about 10% and 90%. The top trace (yellow) is the PWM voltage, while the lower cyan trace is the sawtooth waveform with a range of ±5.6V. The dotted yellow horizontal line represents Vo; when the sawtooth waveform is below it, the PWM output is high. 50 Silicon Chip Australia's electronics magazine By the way, all inverting amplifiers in the motor controller circuit include a resistor from the non-inverting input of the op amp to the 0V rail. These are to equalise the two input impedances so that any input currents will balance out. This minimises offset voltages due to the parasitic input currents. Note how op amp IC2d also connects to the -12V supply via an 8.2kW resistor and trimpot VR3. This preset trimpot adjusts the voltage offset at the Vt test point (IC2d’s output). It is adjusted to provide the correct operation of the PWM by keeping voltage within range of the voltage swing from the sawtooth oscillator, IC4b. The output voltage of IC1b is also applied to the current overload circuitry that comprises IC3d and IC3a; more on that later. Motor speed feedback The motor back-EMF voltage is monitored via a 220kW 1W resistor from the motor’s negative terminal. This voltage is attenuated and offset toward the +12V rail by the connected 8.2kW resistor. The resulting voltage is applied to an inverting and integrating buffer, IC3c, via a 100kW resistor to its inverting input, pin 9. Speed trimpot VR7, in series with a 100kW resistor, provides level-shifting of the voltage from IC3c. As mentioned earlier, IC3c provides low-pass filtering with a corner frequency of around 16Hz due to the 1μF capacitor in parallel with the 100kW feedback resistor. IC3b provides voltage inversion with a gain of -2 before applying the speed feedback voltage to the N1 node via a 10kW resistor. Motor speed adjustment VR1 is used to adjust the motor Scope 6: a zoomed-in view of the PWM drive waveform at Q1’s gate. The rise and fall times are 1.37µs and 1.2µs, respectively, with the gate voltage ranging from 0V when the IGBT is off to 14.8V when on. siliconchip.com.au speed manually. It is connected in series with a 620W resistor across the 12V supply. The 620W resistor is included so that the voltage from the potentiometer’s wiper ranges from 0V to 10.6V. That provides a suitable range to feed to op amp IC2a, which is not a rail-to-rail type. The potentiometer wiper charges and discharges a 100μF capacitor via one of two separate paths. When rotated clockwise to increase the voltage, the capacitor is charged via the 10kW resistor, so it takes about one second for the capacitor to fully charge and signal full motor speed. When the potentiometer is wound anti-clockwise to reduce the speed, the voltage is more quickly decreased by discharging the capacitor via diode D5 and its series 100W resistor. This allows the motor to be stopped quickly if necessary. IC2a’s output feeds the N1 node via a 6.8kW resistor. The current and voltage feedback signals, plus the speed control potentiometer signals, are all summed at the N1 node. This results in a summed output from the IC2b mixer of the three sets of voltages for the current and voltage feedback signals, plus the speed control potentiometer. A second node (N2) mixes the IC2b output (via a 10kW resistor) and the torque from the IC2d output (via trimpot VR4 and its series 10kW resistor). The final summation is performed by IC2c, producing the Vo output signal that’s applied to the PWM comparator, IC4a. The IC4a comparator has a small amount of hysteresis so it does not oscillate when the non-inverting input voltage is close to the sawtooth oscillator waveform voltage. The 100W and 1MW resistors at pin 3 cause the pin 3 voltage to shift slightly when IC4a’s output state changes, preventing the two voltages at pins 2 and 3 from remaining at the same level for long. A voltage clamp comprising zener diode ZD3 and diode D10 at the pin 3 input to IC4a limits the voltage to one diode drop below -8.2V. This clamping prevents the input from going below the op amp’s input voltage range. Without the clamp, if the voltage went below -9V, the op amp output would swing high instead of staying low due to a phase reversal internal to the op amp. siliconchip.com.au The Speed Controller fits neatly into an aluminium enclosure. The black ‘wires’ are actually fibre-optic light pipes for the LEDs. The full adjustment range of VR1 suits 230V DC motors. For motors rated to a lower voltage, like 180V, you can simply operate VR1 over the lower 80% of its range. If you need to prevent more than an average of 180V from being applied to the motor, you can increase the 620W resistor in series with VR1 to 1.6kW. Current overload detection The motor-current-derived voltage from IC1b is applied to the current overload circuitry comprising op amp IC3d and comparator IC3a (another op amp used as a comparator). IC3d amplifies the voltage from IC1b with a gain of -4.68 times (220kW/47kW), with its output voltage level-shifted by trimpot VR5 that’s connected to the -12V supply by way of a 24kW resistor. Australia's electronics magazine IC3a compares the output voltage from IC3d against a reference voltage set by VR6. VR6 is connected as an adjustable divider across the ±12V rails with a 12kW padder resistor and sets the current overload trip level. If the motor current is high enough to produce a voltage from IC3d’s output above the overload level, the comparator output will go high, switching on transistor Q3 and consequently, relay RLY2. A 100μF capacitor holds this pin 2 input at -12V for a few seconds at power-up. The initial low voltage ensures the comparator output at pin 1 of IC3a is high at power up, so RLY2 switches on and opens its NC contact. That ensures RLY3 is not on during power-up, so the motor cannot run immediately. July 2024  51 When transistor Q3 is switched on, we ensure it’s on long enough to activate RLY2 due to the 100μF capacitor that’s initially charged via diode D9 from IC3a’s output. When there is an overcurrent condition and RLY2 is powered, it disconnects power to RLY3’s coil. RLY3 is the high-current relay that connects power to the motor by joining the M+ terminal to inductor L1. If an overcurrent condition triggers the relays, that will quickly cease as the motor will no longer be powered. RLY2’s contacts will close again, so RLY3 can be powered once more, and the motor can be restarted. However, power for RLY3’s coil comes via RLY1’s contacts, and RLY1’s contacts are open unless the speed potentiometer is fully anti-clockwise. So, the speed potentiometer must be returned to the fully-off position before RLY1’s contacts close. RLY3 is then powered to provide voltage to the motor once the speed potentiometer is rotated clockwise. Parts List – 180-230V DC Motor Speed Controller We described earlier how IC1a and RLY1 provide the ‘switched potentiometer’ action we need from a regular potentiometer, but here are more details on how that section works. IC1a acts as a Schmitt-trigger comparator, monitoring the speed potentiometer voltage after buffer IC2a. It compares that voltage to a 119mV reference from a 100kW/1kW voltage divider across the 12V supply. When IC2a’s output is below this 119mV reference, the output of IC1a is low, so RLY1 is not powered. When IC2a’s output is above 119mV, IC1a’s output goes high and drives transistor Q2 via its 1kW base resistor. IC1a includes hysteresis so the output does not oscillate at the 119mV threshold. IC1a is powered from the 15V supply, with a 1MW feedback resistor, so this hysteresis is around 15mV. In more detail, when IC1a’s output is low, its pin 3 input is pulled lower than IC2a’s output due to the 1MW/1kW voltage divider. When IC1a’s output goes, pin 3 is pulled about 15mV higher, so the output from IC2a needs to drop a further 15mV before IC1a’s output will go low again. When Q2 is on and RLY1 is powered, its normally closed contacts open, disconnecting RLY3’s 12V coil 1 double-sided plated-through PCB coded 11104241, 201 × 134mm 1 diecast aluminium enclosure measuring 222 × 146 × 55mm [Altronics H0429, Jaycar HB5050] – the Altronics case requires additional parts for PCB mounting: 4 M3 × 6mm tapped standoffs and 8 M3 × 6mm panhead machine screws 2 10A, 12V DC coil SPDT PCB-mounting relays (RLY1, RLY2) [Altronics S4160C, Jaycar SY4066] 1 30A, 24V DC coil DPDT panel-mount relay (RLY3) [Jaycar SY4041, Hongfa HF92F-024D-2C21S or unbranded FRA8PC-S2] 1 PCB-mounting 15V + 15V 7VA mains transformer (T1) [Altronics M7164 or M7124A] 2 33 × 19.8 × 11.1mm powdered iron toroidal cores (L1) [Altronics L4534A] 1 8-way 300V 15A PCB-mount barrier terminals (CON1) [Altronics P2108] 1 vertical-mount 300V 15A 3-way pluggable header with screw terminals, 5.08mm spacing (CON2) [Altronics P2573 + P2513, Jaycar HM3113 + HM3123] 1 vertical-mount 300V 15A 2-way pluggable header with screw terminals, 5.08mm spacing (CON3) [Altronics P2572 + P2512, Jaycar HM3112 + HM3122] 3 PCB-mount 5mm pitch 6.3mm male spade connectors (CON5-CON7) [Altronics H2094, Jaycar PT4914] 3 6.3mm fully-insulated female crimp spade connectors [Altronics H1842, Jaycar PT4625] 1 IEC panel-mount mains connector with integral fuse (CON10) [Altronics P8324, Jaycar PP4004] 1 M205 230VAC fast-blow fuse (F1) (with a current rating to suit the motor) 1 10A mains IEC lead 1 10A side-entry chassis-mount GPO socket [Altronics P8241, Jaycar PS4094] 1 ALPHA 24mm 5kW single gang linear potentiometer, 500V DC rating (VR1) [Altronics R2203] 3 5kW top adjust trimpots (VR2-VR4) [Jaycar RT4648, Altronics R2380A] 3 50kW top adjust trimpots (VR5-VR7) [Jaycar RT4654, Altronics R2386A] 1 knob to suit VR1 2 14-pin DIL IC sockets (optional) 2 8-pin DIL IC sockets (optional) 3 100mm-long 3mm LED fibre-optic light transporters (optional) [Jaycar HP1193 (pack of three)] 4 yellow 5mm inner diameter crimp eyelets for 4-6mm diameter wire [Altronics H2061B, Jaycar PT4714] Hardware & cables 2 M4 × 10 panhead machine screws (for Earth-to-chassis connections) 2 4mm inner diameter star washers 2 M4 hex nuts 2 M3.5 × 6mm panhead machine screws (PCB to Altronics enclosure) 1 M3 × 12mm panhead machine screw (for Q1) 4 M3 × 10mm panhead machine screws (for BR1, D1 & RLY3) 2 M3 × 10mm countersunk head machine screws (for IEC connector) 3 M3 × 6mm panhead machine screws (for REG1-REG3) 3 3mm inner diameter washers (for D1 and RLY3) 8 M3 hex nuts 1 TOP3 insulating washer 1 500mm length of 1.25mm diameter enamelled copper wire 1 500mm length of 10A green/yellow striped (for Earth) mains-rated wire 1 500mm length of 10A brown (for Active) mains-rated wire 1 500mm length of 10A blue (for Neutral) mains-rated wire 1 450mm x 8mm plastic cable tie (for T1) 1 250mm x 4.8mm plastic cable tie (for L1) 12 100mm x 3.6mm plastic cable ties 1 120mm length of black 5mm diameter heatshrink tubing 1 60mm length of red 5mm diameter heatshrink tubing 52 Australia's electronics magazine Restart switch Silicon Chip siliconchip.com.au 1 15mm length of blue 5mm diameter heatshrink tubing 1 15mm length of yellow 5mm diameter heatshrink tubing Semiconductors 1 LMC6482AIN dual CMOS op amp, DIP-8 (IC1) [Jaycar ZL3482] 2 LM324AN quad op amps, DIP-14 (IC2, IC3) [Altronics Z2524, Jaycar ZL3324] 1 LM833 dual op amp, DIP-8 (IC4) [Altronics Z2598, Jaycar ZL3833] 1 TLP5701 optically-isolated Mosfet driver, SMD-6 (IC5) [element14 3872508 or 2768341] 1 7812 +12V 1A linear regulator, TO-220 (REG1) 1 7815 +15V 1A linear regulator, TO-220 (REG2) 1 7912 -12V 1A linear regulator, TO-220 (REG3) 1 STGW40M-120DF3 1.2kV 80A IGBT, TO-247 (Q1) [element14 2470028] 2 BC337 NPN transistors, TO-92 (Q2, Q3) 4 3mm or 5mm high brightness red LEDs (LED1, LED3-LED5) 1 3mm or 5mm high-brightness green LED (LED2) 1 PB5006 600V 45A SIL bridge rectifier (BR1) [element14 3774973] 1 W04 1A 400V bridge rectifier (BR2) 1 RURG3060 600V 30A fast diode (D1) [element14 2495903] 4 1N4004 400V 1A diodes (D2, D6-D8) 5 1N4148 75V 200mA signal diodes (D3-D5, D9, D10) 2 15V 1W zener diodes (ZD1, ZD2) [1N4744A] 1 8.2V 1W zener diode (ZD3) [1N4738A] 1 P4KE400CA bi-directional TVS diode (TVS1) [Jaycar ZR1164] 3mm diameter required if light transporters are used Capacitors 2 470μF 25V PC electrolytic 1 100μF 25V PC electrolytic 4 100μF 16V PC electrolytic 4 10μF 25V PC electrolytic 1 1μF 63V or 100V MKT polyester 1 1μF 50V multi-layer or monolithic ceramic 1 220nF 63V or 100V MKT polyester 7 100nF 63V or 100V MKT polyester 1 100nF X7R multi-layer or monolithic ceramic 2 47nF 630V pulse double-metallised polypropylene (Kemet R76PI24705050J) [element14 3649826] 1 10nF 63V or 100V MKT polyester Resistors (all axial ¼W 1% unless noted) 2 1MW 2 20kW 4 1kW 1 220kW 1 12kW 2 620W 1 220kW 1W (5% OK) 8 10kW 2 470W 5W (5% OK) 8 100kW 1 10kW (SMD 1206-size) 1 430W 1 91kW 2 8.2kW 1 220W 1 68kW 2 6.8kW 4 100W 2 47kW 4 4.7kW 1 75W 3 33kW 1 4.3kW 1 10W 1 24kW 1 3.3kW 4 22kW 1W (5% OK) 2 2.2kW 2-4 0.022W 3W 1% SMD M6332/2512-size (TE Connectivity TLRP3A30CR022FTE) [element14 3828731] ● 0-2 0.05W 3W 1% SMD M6332/2512-size (TE Connectivity TLRP3A30CR050FTE) [element14 3828740] ● ● see Table 1 next month for quantities (4 × 0.22W is sufficient for all motors) 🔹 🔹 🔹 power unless RLY3 is already on and its contacts are closed. This is because a set of RLY3’s contacts are in parallel with RLY1’s contacts (the points labelled ‘a’ and ‘b’ in Fig.2). The only way to restore power to the motor via the RLY3 contacts is to return speed potentiometer VR1 to its fully anti-clockwise position. In this case, RLY1’s contacts close and +12V is reconnected to RLY3’s coil. The safety switch connection between pins 7 and 8 of CON1 can also stop the motor and prevent it from restarting until the speed pot is returned to the anti-clockwise position. An open safety switch disconnects power to RLY3’s coil, immediately removing power to the motor. Setting it up safely You might be wondering about the purpose of the CON5 & CON7 terminals near CON1 on the circuit diagram, shown joined by a wire bridge. This allows you to disconnect the +12V supply from the positive terminal of the bridge rectifier when making adjustments. Also, because the mains supply to the active side of the bridge rectifier (BR1) and transformer T1 are via separate terminals on CON1, BR1 can be left disconnected during initial setup and testing. With BR1 disconnected, the motor can’t run, and much of the circuit is essentially isolated from the mains Active. This allows you to adjust some of the trimpots and monitor the voltages in the circuit more safely. The circuit is still connected to mains Neutral via the bridge rectifier, though. So, during setup, it is essential to check that the mains Neutral is close to the Earth voltage. Even though some adjustments can be made with the mains Active isolated, some trimpots must be adjusted while the circuit is at mains potential. We will describe how to do this safely in the setup and testing section of the article next month. It involves using a high-voltage-insulated screwdriver with a multimeter and probes rated for use at mains voltages. Next month Construction, testing and setup details for the 180-230V DC Motor Speed Controller will be in a follow-up article next month. SC siliconchip.com.au Australia's electronics magazine July 2024  53 Project by John Clarke Repurposing the Mains Power-up Sequencer Generators and inverters are not always powerful enough to run more than one high-current appliance at a time. For example, if you have more than one refrigerator or a separate freezer and fridge and want to run them off-grid, they may need to run at different times. The Mains Power-Up Sequencer from February & March 2024 can be programmed to do that automatically. T he Mains Power-Up Sequencer from the February & March 2024 issues (siliconchip.com.au/Series/412) was intended for powering up appliances in sequence with brief delays in between to avoid overloading a circuit breaker at switch-on. However, the fact that each outlet is controlled independently by a microcontroller means that the way each outlet is controlled can be changed with new software. We hadn’t considered this second application until a reader wrote to us. His letter was published in the Ask Silicon Chip section of the April issue. In part, he wrote: Say a business has several fridges/ freezers to run from a small emergency power source. It would be very useful to be able to sequence the output to several loads for varying periods, like 15 or 20 minutes, making it unnecessary to manually switch loads to avoid overloading a generator or inverter. A shortlist of features » Powers on two to four mains outlets individually in a rotating sequence » Adjustable powered-on period of eight seconds to 30 minutes » Optional daisy-chain connection for up to four more outlets (up to eight total) » ‘Phantom Appliance’ load detection option (for up to four outlets) 54 Silicon Chip Happily, we can satisfy this request. The re-purposed version of the project mainly requires the microcontroller software to be changed, plus some minor wiring adjustments. Three options Three new versions of the Sequencer are described here, all using the same revised software. The first is called the Primary unit (see Fig.1). It operates with a rotating sequence, switching on Outlet 1 for a period, then switching it off before switching on Outlet 2 for the same period. This sequence continues for all outlets, and when Outlet 4 switches off, the sequence repeats. The power-on period is adjustable from eight seconds to 30 minutes. The eight-second period is mainly useful for testing the unit to see if it works without waiting too long. As with the original Sequencer, you can build it to have fewer than four outlets. If only two or three are required, it will return to Outlet 1 after Outlet 2 or Outlet 3 switches off. The second version operates similarly to the Primary version but includes current detection. When an outlet is first powered, it monitors the current drawn. If an appliance draws power, the outlet stays powered. The outlet switches off after the timeout period, or earlier if the appliance draws less than 35W. We call this the Phantom Appliance Australia's electronics magazine Detection (PAD) mode, where only the outlets that have an appliance connected (or are ready to run in the case of a fridge or freezer) will be included in the sequence. This mode can be useful for powering refrigerators and freezers because they don’t run constantly. Powering an outlet for an appliance that is not doing anything useful wastes time, since it could power the next fridge or freezer instead. Also, the fridge or freezer may finish running its compressor before the timeout expiry. In this case, the PAD unit will move on to power the next appliance early. This mode is also useful where you have the four outlets on the Sequencer, but you may sometimes only use it for two or three appliances. The Sequencer will skip over the unused outlets, and you won’t have to manually change the configuration to set the number of outlets used. It also gives you the flexibility to switch one or more loads off when you want them to be skipped. Daisy-chain mode The third configuration, Daisy Chain, can give you more than four outlets (up to eight). Daisy-chaining is impractical for PAD units; only the Primary unit can be daisy-chained. That is because the AN4 input of microcontroller IC9 used to enable daisy chaining is also used for current detection siliconchip.com.au the code. The software checks the RA3 and AN4 inputs to the microcontroller at power-up. If the RA3 input is low (near 0V), the software runs for a PAD unit, while if RA3 is high (5.1V), either the Primary or Daisy Chain code runs. To discern between these two options, it checks the AN4 input. If the voltage is low, the software determines it is a Daisy Chain unit. If the AN4 input is above the threshold voltage for mains voltage detection, it runs the code for a Primary unit. It is important to build the Sequencer according to the build details for the version you are making so the software runs correctly. Building it Fig.1: when used as a Primary unit, it continually sequences through up to four outlets, switching them on for a fixed time in turn. The PAD unit is similar, except it will only switch on outlets with an appliance connected and drawing at least 35W. Otherwise, after a 1s delay, it skips that outlet. on a PAD unit, and it can’t perform both jobs simultaneously. Fig.2 shows how a Daisy Chain unit is connected to the Primary unit. The Daisy Chain unit monitors the last outlet from the Primary unit via its Mains Detect Input. Its Outlet 1 is powered after the last outlet from the primary unit (shown as Outlet 4) powers on and then off. The Daisy Chain unit then powers each outlet on and off in sequence, stopping after the last outlet. When used in this mode, the Primary unit powers each outlet on and off in sequence, but after powering Outlet 4 off, there is a delay before powering Outlet 1 again. That gives the Daisy Chain unit time to run its complete sequence. We call this delay the return delay, and it is set so that the Daisy Chain unit finishes its entire cycle before the Primary unit starts the cycle again. The return delay can be selected as between one to four times the usual delay period that is set with VR1. That allows you to build the Daisy Chain unit with between one and four additional outlets, with the delay multiplier on the Primary unit set to match the number of outlets on the Daisy Chain unit. Version selection The three versions use the same software but run different sections of This article mainly describes the changes required for the new functions, so for the full PCB assembly instructions, you will need to refer to the articles in the February & March 2024 issues (siliconchip.au/ Series/412). Those articles describe various build options. You can build a unit with between one and four mains outlets (see Table 1), and the optional Current Detection and Mains Voltage Detection circuitry may need to be included. With the new software, switches S1-S3 provide functions different from the original Mains Power-Up Sequencer, as shown in Table 2 and Table 3. VR1 is now only used to adjust the power-on period for each outlet. The wiring and PCB changes for all three versions are shown in Fig.3. In all cases, the two connections at CON7 are bridged using 10A mains Fig.2: for more than four outlets (up to eight), you can connect a Daisy Chain unit to a Primary unit, as shown here. The Daisy Chain unit is triggered when the last Primary outlet switches off; the Primary unit waits for the Daisy Chain unit to finish before restarting the sequence. siliconchip.com.au Australia's electronics magazine July 2024  55 Photos of the completed Mains Power-Up Sequencer before of any of the modifications in this article have been made. Changes to the hardware are minimal. wire. Current transformer T1 is only used for the PAD unit, with the mains Active wire passing through T1’s core. The snubber components for the OUT1 circuitry across TRIAC1 are a 10nF X2-rated capacitor for C1 and a 330W 1W resistor for R1. Do not use the alternative 220nF X2 rated capacitor and 470W 1W resistor values mentioned in the original articles. Microcontroller IC9 must be programmed with the revised software, coded 1010823M.hex. You can download the HEX file and assembly language source code (siliconchip.au/ Shop/6/358) and program the chip using a PIC programmer. Or you can purchase a programmed microcontroller from the Silicon Chip website. The above components and wiring changes are common to all the revised versions, but specific modifications are Table 1 – setting the number of outlets (for all units) # outlets RA1 (pin 18) RA0 (pin 19) 4 (default) 0V (PCB bottom layer) 0V (PCB bottom layer) 3 0V (PCB bottom layer) 5.1V (PCB top layer) 2 5.1V (PCB top layer) 0V (PCB bottom layer) 5.1V (PCB top layer) 5.1V (PCB top layer) 1 56 Silicon Chip required for each version, as described below. Primary unit For the Primary unit, the current and voltage detection sections are left unpopulated. The mains Active wire shown going through the current transformer for the original Sequencer instead goes directly to CON6. Place a wire link between pins 4 and 5 of the pads for IC11 so that the Sequencer will run the Primary unit version of the code at start-up. You can select the number of active outlets by making the linking options as shown in Table 1. Set VR1 for the required on-period of each outlet. Fully clockwise sets a 30-minute timeout; a mid setting is about 15 minutes. If the Primary unit is not being used with a Daisy Chain unit, set switch S3’s Table 2 – return delay setting Return delay S3 position No Left (open) Yes Right (closed) Table 3 – return delay multiplier (for daisy-chained Primary unit) Multi. S1 position S2 position ×4 Left (open) Left (open) Right (closed) Left (open) ×2 Left (open) Right (closed) ×3 Right Right ×1 (closed) (closed) Australia's electronics magazine lever to the left, so there is no return delay (see Table 2). The switch positions for S1 and S2 do not matter for this version. If the Primary unit is used with a Daisy Chain unit, set switch S3’s lever right so there is a return delay (see Table 2). The return delay setting is made using switches S1 & S2, as shown in Table 3; select the ×4, ×3, ×2 or ×1 delay multiplier to match the number of outlets used on the Daisy Chain unit. Daisy Chain unit The Daisy Chain unit requires the voltage detection circuitry to be installed, with no wire link between pins 4 & 5 of IC11’s pads. You can select the number of outlets installed in the Daisy Chain unit as per Table 1. Set VR1 for the required power-on period of each outlet, but ensure it is slightly less than the period set for the Primary unit. Set switch S3 for the Daisy Chain version with the lever to the left so there is no return delay (see Table 2). The switch lever positions for S1 and S2 do not matter for this version. Phantom Appliance Detect (PAD) unit The PAD version requires the current detection circuitry to be installed, with the Active wire from CON5 looping through current transformer T1 before terminating at CON6. You will also need to connect a wire between the 0V test point and the bottom of the 10kW resistor that connects siliconchip.com.au Fig.3: besides reprogramming IC9 with the new software, just a few changes are required to the hardware. Add one of the wire links shown in red if building the Primary or PAD version. The mains Active wire only goes through T1 for the PAD version; otherwise, it connects directly to CON6. (through tracks on the PCB) to the pin 4 RA3 input of IC9, as shown above. This informs the software that the unit is the PAD version. With the revised software, the RA3 pin is set as a digital input rather than as a master clear (MCLR) reset line, as it was in the original version of the Sequencer. Put switch S3’s lever left so there is no return delay (see Table 2). The switch lever positions for S1 and S2 do not matter for this unit. Testing As per the original Mains Power-Up Sequencer articles mentioned, all wiring and adjustments must be made with the input mains power disconnected since the circuitry is live when plugged in. Also, any adjustments of the period using VR1 are only detected at power up, so there is no benefit to siliconchip.com.au adjusting VR1 with the power on. So, each time you want to make an adjustment, ensure that the power is off before opening the lid of the enclosure. Replace the enclosure lid before reapplying power. If VR1 is set to its minimum fully anti-clockwise position, the period for each outlet will be short, at eight seconds. That makes monitoring and checking its operation easier. You can see the sequencing occur as the indicator LEDs light up for each outlet. For the PAD unit, the output LED indicator for each channel will only light when an appliance that draws power (at least 35W) is detected. That is because only the Triac for each channel is switched on initially, while the current drawn by the appliance is first detected, and the LED indicators only show the relay status. Australia's electronics magazine Using the Triac to apply voltage initially saves the relay from operating if there is no current drawn by the appliance, extending the relay life considerably. For the PAD sequence, you can check each outlet by connecting a load that will draw 35W or more, such as a 40W 230V AC halogen lamp. Cycling through outlets that do not have a load connected takes one secSC ond per outlet. Warning: Mains Voltage All circuitry within the Mains Sequencer operates at Line (mains) voltages. It would be an electrocution hazard if built incorrectly or used with the lid open. Only build this if you are fully experienced in building mains projects. July 2024  57 Adding automatic solar charging to an electric van By Roderick Boswell How far can an EV travel without having to visit a charger? We added solar panels to the roof of a Renault Kangoo ZE van, plus an onboard inverter. This gives us up to 18,000km a year of driving at no further cost! H aving built the solar van, we’ve achieved up to 50km of driving per day using just the solar panels. Multiply that by the number of days in a year to get the 18,000km figure, although that assumes nice sunny weather year-round, which is perhaps a little unrealistic. Still, sitting in the van watching the onboard batteries charge at 50A for the first time, it certainly was pleasing to realise that it was working as intended. There are surprisingly few reports of this having been done, so we thought we would create a company, “Solely Solar”, based on the concepts of autonomy and freedom. In this article, I will describe how the decisions were made, what we purchased, how we configured and tested it, the integration of the solar 58 Silicon Chip system into the van and the on-road tests. The solar panels Photovoltaic (PV) solar panels have been slowly improving over the past few years. It is possible to purchase single crystal silicon arrays with passivated emitter rear cells (PERC), which were invented by a team at UNSW in Sydney. They are cut in half to reduce the resistance and hence losses. These cells have an efficiency of around 22%, so with full sun delivering 1kW per square metre, you can obtain 220W from a 1m2 panel. So, off I went to eBay to check prices; I found a real Aladdin’s cave of solar treasures. Having purchased a few, I quickly discovered that the power they Australia's electronics magazine could produce was often overstated by as much as 100%. Unless you like opening protracted disputes with eBay (which I did to see how the system works; it does, sort of), be aware that the only reliable indicator of the potential power of the panel is its area. I learned this by spending money and testing the product, an easy task with a multimeter that can measure up to 20A. The two main parameters to measure are the open circuit voltage (Voc), which increases with the area of the panel, and the short-circuit current (Isc), which manufacturers try to keep as low as possible to reduce Joule heating (I2R). For example, I tested a 2m2 solar panel with a Voc of around 50V and an Isc of around 10A. Of course, multiplying those two siliconchip.com.au figures won’t tell you exactly how much the panel will produce since they are measured under different conditions. Still, it gives you a way to estimate the power and compare different panels. We decided to use Longi 510W panels that measured 2093 × 1134mm and weighed 25.3kg since they just fitted onto the roof of the Kangoo. Interestingly, some tests showed around 550W being produced per panel. There is an efficiency temperature coefficient of -0.25%/°C, with the stated performance being at 25°C. So, on a cold morning, with a temperature around 0°C, the panels will be 5% more efficient. Conversely, of course, during the afternoon in summer, the air temperature may be 40°C, and the panels will be so hot you cannot touch them, leading to a performance reduction of up to 10%. The van There are currently several very expensive electric vans on the market. Still, a couple of years ago, the only real option was the Renault Kangoo Zero Emission, although BYD slipped around 50 T3 electric vans into Australia as they were mucking about with distributors. I decided on a 2019 Kangoo ZE that had been used to drive from the Blue Mountains to Sydney every day and back, which had travelled around 85,000km. The Kangoo has a Mennekes Type 2 7kW onboard charger that requires a Type 2 to Type 2 cable, or a destination charging cable with a Type 2 on one end and a regular 10A 230V mains plug at the other. Two of us would have to drive to Sydney from Canberra, pick the van up, and drive it back. Since the top of the CCS (Combined Charging System) plug is a Mennekes Type 2 plug, we purchased a Type 2 to Type 2 cable. We made an unpleasant discovery when we stopped at a commercial charging station at Sutton Forest on the way back to Canberra. The top Mennekes socket of the CCS charging station was not connected! As night was about to fall, we swiftly returned to Canberra in the other car, leaving the Kangoo in the parking area adjacent to the chargers. Rats! After some research, we found that the commercial CCS charger providers wanted a fast turnover so their chargers only provided DC fast charging. The siliconchip.com.au The inside of the Renault Kangoo ZE van with some basic wiring for the solar panels. The onboard inverter and the other electronics required for the solar panels are stored in the large timber cabinet on the side that doubles as a kitchen. poor old Mennekes is generally limited to 7kW, resulting in a wait of several hours. The company supplying the chargers evidently didn’t want anyone sitting on their charger for that long, so they removed the Type 2 option. The following morning, we returned much wiser and drove the Kangoo to a local winery that had a couple of Type 2 chargers. We plugged in and then discovered that you have to download Australia's electronics magazine the charger supplier’s app on your phone to arrange payment before you can start charging. After doing that, it was finally charging, and we had four hours to kill. We had lunch at the winery and drove around the area, which was really quite beautiful, and got back to discover that a watched charger never boils. Eventually, we were back on the road again, popping into the Goulburn July 2024  59 Workers’ Club later for supper and another couple of hours of Type 2 charging. That got us home. At home, we used the cable with the Type 2 connector on one end and 3-pin mains plug on the other to recharge the van overnight. The dash instruments show the instantaneous kWh/100km figure, estimated range, distance travelled and instantaneous power usage. It also has a ‘fuel gauge’ that correlates more-or-less with the battery state of charge (SoC). At 1/8 SoC remaining, you touch the red line and a warning light suggests you look for a charger, as there is only about 30km of range remaining. A double red line follows at 1/16, and another light appears that the manual explains is to warn that you are about to go into ‘limp home’ mode. I checked this scheme out, down to 1/16th full, and all worked as expected. A few tests showed that the charger is about 90% efficient, with 10% lost between the mains and the van battery. The majority of the losses are from the inverter built into the Kangoo. I conducted a sequence of tests on range and efficiency at different speeds, with the main result being that the battery still had about 30kWh left of the original 33kWh. Not too bad after 85,000km! The best efficiency of 14-15kWh/100km was at 50-60km/h. It read about 17.5kWh/100km at 80km/h and over 25kWh/100km at 110km/h. It is interesting to get used to nearly one-foot driving with the regenerative braking. I performed another test in hilly terrain, taking the van to the Picadilly Circus pass through the Brindabellas, a voyage about 50km long and 1000m vertical. At the top of the mountain, the consumption had increased to 22kWh/100km, but on returning home, it had dropped back to around 15kWh/100km, having regained most of the energy used to ascend. This was with careful driving, trying to keep the efficiency indicator out of the red, even if it meant going at only 30km/hr on the steep parts of the ascent. The regenerative braking certainly is effective. To sum up the efficiency/range tests, keeping to 50km/h, I got a range of about 220km, but at 80km/hr, it drops to around 150km. These results agree with the USA Electric Vehicle Design Base (EVDC) range estimate for the Kangoo ZE of 160km. There is real optimism in Europe with the New European Driving Cycle (NEDC) that claims a range of 270km, while the Americans take a more realistic view. Attaching the solar panels According to the Australian Design Rules (ADR) for loads carried on The inside of the timber cabinet, which contains the Victron MPPT solar chargers, circuit breakers, busbars, battery charger etc, as shown in Fig.1 opposite. The eight batteries sit under a piece of wood on which the main circuit breaker is fitted. vehicles using public roads, an overhang of 1200mm without flags is acceptable both front and rear. For side protrusions, 150mm on each side is allowed beyond the vehicle’s width. This meant the Longi 510W panels were a good fit, so we decided on having three lengthwise on the roof, with the first in line with the top of the windscreen and about 500mm of overhang at the rear. MORID Pty Ltd did the design using the roof rack attachment points (three on each side). The main challenge was the roof loading rating of the Kangoo, which is just 100kg. Having three 25.3kg panels means that the whole roof rack structure could weigh only 24.1kg and had to be strong enough to hold the panels. A plastic 3D printer was employed to print the fittings for the prototype. We then attached them to the roof to verify their stability, size and appropriateness. Having passed this first hurdle, the design was sent off to PCBWay for machining out of aluminium. These guys are really good and they have never disappointed us. The six adaptors were finished and sent to us. Perfect! Assembling the panels into an aluminium frame, drilling holes in the roof and attaching them to the van took some time. We just managed to get it a few millimetres under the protrusion rules. We were then faced with the one task we had been avoiding: drilling holes in the roof to get the cables from the panels into the van so they could be connected to the interior electronics. We took the plunge, drilled the holes in the roof (sorry Mr Renault) and fitted the grommets. As usual, after the cables were slipped through and the connectors attached, we found that we had forgotten to slip a clip on the connector; oh dear! We had to desolder the connectors, attach the wayward clip, then resolder the connectors. The Maximum Power Point Trackers (MPPTs) The solar panels do not charge the Kangoo’s battery directly, as the onboard charger does not support charging from low-voltage DC. Instead, the solar panels charge a secondary 24V battery that we installed (more on that later), and that battery runs an inverter that feeds the onboard EV charger – see Fig.1. 60 Silicon Chip Australia's electronics magazine siliconchip.com.au Our solar panels put out about 50V and 10A, and we need to charge a 24V battery, so a conversion is necessary, conserving as much power as possible. By chopping the input voltage at around 100kHz, small inductors (or coils) and an electronic circuit called a buck converter can reduce the voltage without wasting too much power. As a result, the output current is higher than the input current. If we are charging the 24V battery at, say, 27V, the charge current would be 18A minus the losses from the buck converter, which are only around 5% nowadays (ie, 95% conversion efficiency). MPPT is needed to get the most power from the panels, as the voltage/current curve has a peak that moves depending on the ambient conditions. We want to manage the panel voltage to keep it at that point while doing the buck conversion. The MPPT chargers also continuously monitor the battery SoC to provide the correct charging profile. A few years ago, such circuits comprised lots of individual components and were pretty expensive, but now a single chip can carry out most of the operations and the price of MPPT chargers has fallen dramatically. It pays to shop around! The secondary battery Once again, this was a learning experience. For safety, we decided on Lithium Iron Phosphate (LiFePO4) cells since they are less likely to fail than Lithium Manganese Nickel cells (and if they do fail, it’s usually less spectacular). However, they have a lower energy density. The next choice was the voltage. Using a 12V DC battery would require high currents and hence significant Joule losses, so we went for 24V. Should we use a series/parallel arrangement of 12V batteries or build our own 24V system from 3.2V prismatic cells? If the latter, we would need a battery management system (BMS) to balance the voltages of all the cells and prevent overcharging and overdischarging. I tried both approaches and started by purchasing four 12V 135Ah batteries. These were bought at different times during the COVID-19 years, and we soon discovered that we needed to get a balancing system, so we purchased that as well. It worked, but it was a clunky solution, so off to AliExpress to purchase eight 3.2V 320Ah PWOD prismatic cells and a 24V BMS. These took some months to arrive, and we eagerly assembled them with the BMS attached to a 20A charger and waited until the BMS cut out. We then discharged it into a bathroom heater via an inverter and surprise, we only got 275Ah. We charged it again and only got 275Ah the second time, so what should we do? Messaging the PWOD AI was highly frustrating, as it was impossible to have a coherent discussion. They finally offered $26 off the next purchase, or we could send them back at our expense. Sigh. So we swallowed the bitter pill and realised how the price could be so low – caveat emptor. We would have to make do with 6.7kWh of stored energy, 14% lower than expected. The inverter There are a great many DC/AC inverters on the market. The first one I bought was from Victron and it is installed in our solar off-grid shed. Actually, I did buy a few smaller inverters before that for use in the car Fig.1: each solar panel has its own MPPT battery charger to maximise charging efficiency. The battery management system ensures the cells remain in balance and are not overcharged or overdischarged. siliconchip.com.au Australia's electronics magazine July 2024  61 and for camping, but nothing in the kW range that we were investigating. Pretty much all the inverters now use chopped voltages rather than heavy transformers, making them quite compact. We needed 3kW continuous and 6kW peak (for a few seconds), with a charger, and we got those capabilities for well under $1000. However, we soon realised that the inverter’s internal charger could only draw a maximum of 2300W from the wall (230V <at> 10A). Since it was to be used for charging the Kangoo, we purchased a second inverter rated at 4kW continuous and 8kW peak for around the same price. We tested the batteries and inverter before installation to verify that everything was operating as expected. With everything installed in the van and the solar panels connected, the isolating circuit breakers were flipped on and, hooray, the Bluetooth app on our Android phones showed the voltage, current and power being delivered by each of the three panels. The BMS showed the battery charging. Charging the Car The last step was to charge the car with solar panels. On the first try, using the 24V battery, inverter and 10A home charger, the car refused to charge. The charger was blinking; after reading the manual, we realised that the error was related the Earth connector on the socket. Most inverters have Earth wiring, however, most of the time, it is floating. To solve this problem, we connected the Earth and Neutral inside the inverter and the car started charging. Using only the 24V battery for charging the car, the inverter would draw around 90A and could add 6.5kWh (~40km range) to the car. Using a fully-charged 24V LFP battery and solar panels on a sunny day at the same time, the solar panels provide around 50A and the battery around 40A, adding up to 16kWh (~100km range) to the car in one day. You can’t do that two days in a row, though, as the secondary battery would be discharged at the end of the first cycle, and it needs to be charged initially to provide so much energy to the EV battery. be ideal for camping. Since micro-­ campers are popular, we took the van to Kata Camperbox in Sydney to do their magic. As can be seen from the photos, the fittings are all real timber, and the result is a true beauty to behold. There is a pull-out kitchen, a slideout fridge that runs from the 24V battery and a space large enough for two electric bikes that can also be used as a sleeping space. It is about the same size as a business-class bed on an aircraft. To get an idea if everything would work, we took a camping/cycling trip to the Orroral Valley campsite that had recently reopened after the bushfires. This 55km trip from Canberra was successful; all the subsystems worked, and nothing fell off the van. My electric bike fitted in the van OK. However, I must say that I am not a great fan of sleeping in enclosed spaces, even those at the pointy end of an aircraft. I know; first-world problems! Planning for a trip In Australia, the Bureau of Statistics has determined that the average passenger vehicle travels a smidgen over 10,000km a year, an average of a bit under 30km per day, well within the parameters of our Solely Solar van. The van would have to be left out in the sun all day; still, rooftop parking is generally the last to fill up, so perhaps that is not too much of a drawback. So what sort of a trip could be made with our little Solely Solar Renault Kangoo ZE, just relying on solar generation of electricity? The Kangoo has 30kWh in its primary battery and 6.7kWh in the secondary battery. Assuming you have solar panels on the roof of your house connected to a home battery, it is simple to just charge the Kangoo at home without paying for grid energy. You could have a separate solar system for charging the car, but then you could argue that you are losing around 10¢/kWh by not selling any excess power back to the grid power supplier. Then again, nothing is completely free. However, if you would like to travel further than a few tens of kilometres (eg, to work and back), you need to do a bit of planning, especially if you want to get back in under a week. And there are limitations on how the remote charging is carried out. As mentioned, the solar panels alone cannot supply the full power necessary for charging via the inverter; they need to be supplemented with power from the solar batteries. Arriving at a campsite in the evening, the solar battery will generally be full, allowing the 6.7kWh (ignoring losses) to be transferred to the traction battery in a few hours while it is dark. The next morning, the panels will start charging the empty solar battery and will absorb around 4.5kWh by midday, at which time the inverter can be brought into play, allowing both the solar electrons and the secondary battery electrons to flow into the traction battery for the daylight that remains. Given a sunny summer day, the solar panels would provide around 9kWh, so the Kangoo would need about 3½ days to fully recharge if exhausted. So, with judicious planning and good weather, you could take a long weekend and travel within a radius of around 180km from your house and return, paying virtually nothing for the trip. Not too SC shabby! The camper conversion Our group had some discussions and decided that the Kangoo would 62 Silicon Chip The van with solar panels being used for camping for the first time. Removing the bike frees up space to sleep inside. Australia's electronics magazine siliconchip.com.au QM1493 Specialty meters combined with multimeter functions. Valid from 10.07.2024 - 21.07.2024 ONLY 329 $ SAVE $20 TAKE EASY ENVIRONMENTAL MEASUREMENTS • MULTIMETER FUNCTIONS • SOUND LEVEL • LIGHT LEVEL • INDOOR TEMP • HUMIDITY HIGH VOLTAGE INSULATION TESTING "MEGGER" • MULTIMETER FUNCTIONS • DIGITAL DISPLAY • ANALOGUE BARGRAPH • DATAHOLD NOW TEST WIRING INSULATION 99 $ ONLY 179 $ QM1594 TEST ALMOST ANYTHING! QM1632 CONTACTLESS HIGH CURRENT MEASUREMENTS • MULTIMETER FUNCTIONS • TRUE RMS • AUTORANGING • CAPACITANCE • NON-CONTACT VOLTAGE WAS $119 MEASURE HIGH CURRENT ALL MODELS FEATURE: • AUTORANGING • AUDIBLE CONTINUITY • MAX / DATA HOLD DETECT OPEN, SHORT OR MISS-WIRED LAN CABLES • MULTIMETER FUNCTIONS • PINOUT INDICATOR ONLY 119 $ XC5078 GREAT FOR I.T. TECHNICIANS Multi-function Meters Saves you money and provides the convenience to carry just one tester in your toolbox. Specialty Function Display (Count) QM1632 QM1493 XC5078 QM1594 Clamp Meter up to 600A AC/DC Insulation Test up to 4000MΩ LAN Cable Test with pinout indicator Sound, Light, Humidity & Temp 4000 4000 2000 4000 Security Category Cat III 600V Cat III 1000V Cat III 600V/Cat II 1000V Cat IV 600V/Cat III 1000V Voltage 600V AC/DC 750V AC / 1000V DC 600V AC / 600V DC 600V AC / 600V DC 40MΩ 4000MΩ 20MΩ True RMS • Current 600A AC/DC Capacitance 100mF Resistance Frequency • 200mA AC/DC 10A AC/DC 10MHz Temperature 1000°C Non Contact Voltage • Relative Measurement • 40MΩ 100µF 10MHz 750°C • • • Explore our great range of multimeters, in stock on our website, or at over 115 stores or 134 resellers nationwide. • www.jaycar.com.au 1800 022 888 Prices correct at time of publication but are subject to change. Jaycar reserves the right to change prices if and when required. Mini Projects #002 – by Tim Blythman SILICON CHIP Lava Lamp Display Lava lamps have always invoked a fascination due to the seemingly infinite patterns that they produce. The Lava Lamp Display is a simple Arduino project that emulates a lava lamp, creating a soothing view that doubles as a groovy night light. T he lava lamp was invented in 1963 and consists of a glass bulb containing a mixture of liquids like oil and water. An incandescent bulb in the base heats the contents, and the different components swirl around due to their changing densities and surface tensions. The liquids are often coloured and the random, slow movements of their contents can be captivating and hypnotic. And bizarre as it may sound, lava lamps are even used as a source of random numbers for encryption. Companies like Cloudflare use them as part of their encryption process (see https://youtu.be/1cUUfMeOijg). Our Lava Lamp Display is a simulation of a lava lamp, using software to imitate the physics. We can’t simulate things down to the atomic level with an 8-bit processor, but we can create something that looks and behaves similarly. Our Display isn’t actually random, but it looks like it is. The photos shows how the completed Lava Lamp Display uses an 8×5 LED matrix shield mounted on an Arduino Uno board to provide the processing power. Simulation The simulation involves several ‘blobs’. Each has a ‘temperature’ and position within the display. They are analogous to the balls of oil that break off and travel around a Lava Lamp. The temperature determines whether or not the blob rises or falls, mimicking its density changing. The position affects the temperature; when the blob is near the bottom, the temperature increases, as though the blob is being heated. Near the top, the temperature falls, as though by radiating heat to the surroundings. This feedback sets the scene for the constantly changing movement of the blobs. To avoid the blobs overlapping and disappearing, the simulation prevents a blob from moving on top of another. Assembly of the Lava Lamp Display just involves plugging the LED matrix shield into an Arduino Uno (shown above). The blobs in the Lamp drift around like those in a lava lamp. The software can be modified to alter the colour or behaviour if desired. 64 Silicon Chip Australia's electronics magazine siliconchip.com.au Lava lamps are produced in a variety of colours, and they produce unique and constantly changing patterns. Source: https://w.wiki/9TUn (CCA 2.0). To prevent a deadlock, blocked blobs occasionally move in a random direction. This randomness comes from a pseudo-random number generator. The blobs’ colours also change depending on their temperature, adding further variety to the display. The result is a fairly convincing simulation of a lava lamp. Hardware and assembly The construction phase of this project simply involves plugging the XC3730 shield into an Arduino Uno board. The XC3730 LED Matrix Shield uses so-called ‘intelligent’ RGB LEDs. We described how these LEDs work in an article on page 85 of the January 2020 issue of Silicon Chip magazine (siliconchip.au/Article/12228). In summary, we can drive all 40 RGB LEDs on the shield using just one digital output on the Uno. Since the LEDs are already attached to the shield, assembly is simple: plug the XC3730 LED Matrix Shield into the Uno and connect the USB cable between the Uno and a computer. Programming the Arduino You will need to install the Arduino IDE software plus some custom libraries. Adafruit’s NeoMatrix library is responsible for driving the display. It can be installed (along with its other dependent libraries) by searching for “neomatrix” in the Library Manager – look for the version by Adafruit. siliconchip.com.au Download and unzip the software package for this project, which is available from siliconchip.au/Shop/6/396 Next, open the XC3730_LAVA_ LAMP_COLOURS sketch, select the correct board type and serial port from the menus, then upload it to the Uno. Arduino boards like the Leonardo should also work, but we haven’t tested that. If all is well, you should see a display similar to that seen in our photo. There isn’t much that can go wrong; it should just work. A video of it can also be found at siliconchip.au/link/abu8 Software details The software has been written to be configurable, so there are some #defines and variables that you can change to customise your Lava Lamp Display. Remember to upload your sketch again after any changes so that they can take effect. The BACK_COLOUR #define sets the background colour; the default is a dim blue. Changing the number in the line matrix.setBrightness(6) will alter the display intensity. We have set it quite low so that the Lava Lamp Display is suitable as a night light or for nighttime mood lighting. The colour of the blobs is set by the tempColour[] array, based on the blobs’ temperatures. The default is quite subtle; you can try uncommenting one line at a time to see different schemes we have tried, or you can make your own. To speed up or slow down the Display, you can change the delay() function call within the loop() function. A higher value will result in a slower update rate. You can also change the number of blobs with the BLOB_ COUNT #define. The heatMap[] array dictates how the temperature changes based on position. The updateBlob() function encapsulates the physics of how each blob behaves based on its temperature. For more advanced constructors, modifying the code can produce some significant changes to the simulation. All of these changes will have very subtly different effects on the model’s behaviour and lead to so-called emergent behaviour, where a simple set of rules can result in complex outcomes. Another example of emergent behaviour is a set of mathematical rules called Conway’s Game of Life. You can see examples of this at https://w.wiki/3TKJ We have also written an implementation of this scheme in a sketch called XC3730_CONWAY, which you can try out by uploading it to the Lava Lamp Display hardware. It is included in the same download package. There is an array you can use to set the initial conditions, after which you can see how the state evolves. Each LED is either lit or not; its state in the next phase of the sequence depends only on it and its immediate neighbours. The rules are pretty simple, but the animations generated almost look like they are alive, hence the name. Conclusion The Lava Lamp Display takes a simple simulation of lava lamp physics and turns it into a unique and mesmerising display that can be used as a night light or simply for amusement. It shows how simple rules can combine SC to create complex behaviour. Parts List – Lava Lamp Display (JMP002) 1 Arduino Uno microcontroller module [Jaycar XC4410] 1 8×5 RGB LED Matrix Shield [Jaycar XC3730] 1 USB-A to USB-B cable [Jaycar WC7705 or similar] Australia's electronics magazine July 2024  65 Don't pay 2-3 times as much for similar brand name models when you don't have to. IDEAL STARTER STATION IDEAL HOBBYIST ENTRY LEVEL STATION ONLY 5495 $ TS1610 LIGHTWEIGHT, EXCEPTIONALLY DELICATE • 10 WATT • ROTARY TEMPERATURE CONTROL DIAL TS1620 LIGHTWEIGHT IRON WITH ADJUSTABLE TEMPERATURE • 48 WATT • SLIMLINE DESIGN GREAT FOR ENTHUSIAST'S WEEKEND PROJECTS GREAT FOR EVERYDAY ELECTRONICS ENTHUSIASTS ONLY ONLY 149 $ ONLY 8995 $ TS1564 229 $ TS1640 OUR MOST POPULAR STATION FOR HOBBYISTS • 48 WATT • ANALOGUE TEMP ADJUSTMENT Explore our great range of soldering stations, in stock on our website, or at over 115 stores or 130 resellers nationwide. RELIABLE OPERATION WITH EXCELLENT TEMPERATURE STABILITY • 60 WATT • DIGITAL TEMP ADJUSTMENT • ESD SAFE • INCLUDES FULL SET OF SPARES INCLUDING REPLACEABLE PENCIL www.jaycar.com.au 1800 022 888 Soldering Stations Soldering made easy with our BEST RANGE of soldering stations at the BEST VALUE, to suit hobbyists and professionals alike. SOLDER OR DESOLDER SURFACE MOUNT COMPONENTS COMPLETE SOLDER/DESOLDER STATION • 60 WATT IRON • 300W HOT AIR PUMP • RAPID TEMP RECOVERY • DUAL DIGITAL DISPLAY • ADJUSTABLE TEMPERATURE • ESD SAFE ONLY 379 $ TS1648 Use this colour coded selection guide to pick the soldering stationthat best suits your needs. GREEN labelled products suit hobbyists and those on a budget. BLUE suit makers who use a soldering station regularly and need ESD protection. For advanced hobbyists or technicians, choose from the ORANGE professional range. ENTRY LEVEL MID LEVEL PROFESSIONAL TS1610 TS1620 TS1564 TS1640 TS1648 Key Feature Compact Design Slimline Ceramic Element Digital Display Soldering & Hot Air Power (Watts) 10W 48W 48W 60W 300W Temp. Range 100-450°C 150-450°C 150-450°C 160-480°C 50-480°C Soldering 100-500°C Hot Air Display Digital Digital ESD Safe • • $229 $379 Price $54.95 $89.95 $149 *Temperature rating is set by the soldering iron tip. ESD means Electro Static Discharge Shop Jaycar for your soldering essentials: • Soldering stations • Electric handheld irons • Gas powered irons • Classic 60/40, lead-free, silver & paste solder options • Multiple desolder braid and tools • Wide range of stands, cleaners and PCB holders Prices correct at time of publication but are subject to change. Jaycar reserves the right to change prices if and when required. Mini Projects #008 – by Tim Blythman SILICON CHIP Digital Compass With this digital compass design, you can add an easy-to-read compass display to your off-road vehicle or build a handheld digital compass powered by a battery bank. It uses two modules and three pushbuttons, so it’s easy and quick to build. T his Digital Compass uses the Jaycar XC4496 Compass Magnetometer Module to measure the local magnetic field and determine magnetic north. We then use Jaycar’s XC3715 Quad 14 Segment Alphanumeric Display Module to show the compass bearing. To make the final result neat, we mounted the main components on an XC4482 Prototyping Shield. The shield has space to spare, so we included three pushbuttons as user controls. It is driven by an XC4430 Leonardo Main Board. The magnetometer module is a handy device based on the HMC5883 integrated circuit (or a similar chip). We described how this type of module works in Silicon Chip (November 2018; siliconchip.au/Article/11310). It has an I2C serial interface, so it is easy to connect to a microcontroller with just two wires. Rather than simply giving a compass heading, these modules measure the magnetic field in three different axes, allowing the direction of the magnetic field to be derived using trigonometric calculations. Our Modules series of articles also covered the type of display module we are using in January of this year (siliconchip.au/Article/16092). The display module is also controlled using an I2C interface. Magnetic declination Magnetic declination is a term used to describe the deviation of the magnetic north from true north. This occurs because the magnetic poles are not exactly at the geographic poles. The Wikipedia page about Magnetic declination (https://w.wiki/9doF) has maps showing how this changes over Fig.1: using two modules that connect with an I2C interface makes this a straightforward project. We have seen some variants of the compass module, so make sure you connect to the correct pins, as they could be in a different order. 68 Silicon Chip Australia's electronics magazine siliconchip.com.au These are the main parts we used for this project. You’ll also need some insulated wire to hook everything up. Depending on what headers are provided with the modules, you might also need some header pins to mount them. both space and time (see overleaf). That means the Digital Compass needs a correction factor to give accurate readings, and that factor will depend on your approximate location. The Compass will display positive declination values as E (east) and negative values as W (west), as is the convention. The Compass shows declination values to one degree of precision, but note that it will jump around by a degree or two in normal use. The easiest way to get a usable value is to perform a web search for magnetic declination with the name of your nearest city. To a rough approximation, the east coast of Australia is currently at around +10° (10°E) magnetic declination, with the west coast close to 0° magnetic declination. Circuit details Fig.1 shows the wiring diagram of the Digital Compass. The two critical modules, the magnetic sensor and 14-segment LED display, are supplied with power and connected to the microcontroller via a common I2C bus. The three switches are also connected to the controlling Arduino. The chip on the compass module runs at 3.3V and the module has an onboard voltage regulator. This means that it expects the I2C bus to be at 3.3V. Due to how I2C works, the 5V Leonardo can interface to a 3.3V I2C bus, so there is no problem with the difference between the two in voltages. siliconchip.com.au The display module runs at 5V but has a separate pin for setting the voltage on its I2C pullups; we connect that to 3.3V to maintain compatibility with the compass module. The processor provides the three tactile switches with pullup currents to hold the connected pins at 5V most of the time. It detects that they are pressed when the circuit is closed to ground, pulling those pins to 0V. This circuit could easily be wired up with jumper wires on a breadboard. If you’d like something a bit neater or more permanent, you can follow our instructions for assembling the parts onto a prototyping shield. Construction Before soldering, note the pin markings on the underside of the modules. Refer to Fig.1, but remember that you could be working from the opposite side of the devices. Start by soldering the display module to the prototyping shield. We aligned the module’s edge with the shield’s edge on one side, then used the topmost row of holes to retain the most space within the shield. Space the module vertically away from the shield to avoid short circuits. The compass module solders to the row of pads intended to accept a DIP IC. Note how it very slightly overlaps the display module’s PCB. Because of the height of the 14-segment displays, it does not protrude. There appear to be a few variants of this module; we have used the larger version, but the smaller variant should fit just as well. Next, solder the tactile switches in the space below the display. Ensure none of the leads are touching those from the other switches. The wiring is done underneath the shield to maintain a good appearance. The wire colours we have used are much the same as in Fig.1, although we used yellow wires for the 5V connections to help them stand out from the red shield colour. The colours are merely a guide to assist assembly; you don’t have to use the same ones. Parts List – Digital Compass (JMP008) 1 Arduino Leonardo main board [Jaycar XC4430] 1 prototyping shield [Jaycar XC4482] 1 digital compass module [Jaycar XC4496] 1 quad 14-segment display module [Jaycar XC3715] 3 two-pin tactile switches [Jaycar SP0611] 1 micro-USB cable to suit Leonardo assorted insulated wire straight pin headers (check what is supplied with the modules) Australia's electronics magazine July 2024  69 US/UK World Magnetic Model - Epoch 2020.0 US/UK World Magnetic Model - Epoch 2020.0 Main Field Declination (D) 180° 180° 135°W 135°W 90°W 90°W 45°W 45°W -90 90°E 90°E 70 -20 -80 10 0 -3 0 135°E 135°E 180° 180° 90 k j -20 0 50 10 40 -40 75°N 30 0 -2 -10 20 80 60 -60 75°N 75°N 45°E 45°E -5 -7 -50 0° 0° 0 0 Main Field Declination (D) 0 20 60°N 60°N 60°N 0 -10 10 45°N 45°N 45°N 0 -1 10 30°N 30°N 30°N 0 15°N 15°N 15°N 0° -2 0 0°0° 10 15°S 15°S 15°S 10 30°S 30°S 30°S 20 -30 45°S 45°S 20 -1 0 -2 45°S 0 30 0 -90 40 0 10 -40 20 60°S 60°S -70 50 k j 80 60°S 60 70 30 90 40 75°S 75°S 50 -10 75°S 135°W Main Field Declination (D) k j Position of Dip Poles Miller Cylindrical ProjectionDeclination (D) Main Field Contour interval: 2 degrees Positive (east) Negative (west) Zero (agonic) line Blackout Zones Miller Cylindrical Projection Contour interval: 2 degrees Horizontal Field (H) Strength: 0-2000 nT (Unreliable Zone) 2000-6000 nT (Caution Zone) 90°W 45°W Positive (East) Negative (West) Zero (Agonic) Line 0° -80 -60 90 180° -70 -50 80 -40 70 -30 -20 60 45°E 90°E 135°E 180° Blackout Zones | Horizontal Field (H) Strength Map developed by NOAA/NCEI and CIRES https://ngdc.noaa.gov/geomag/WMM 0–200nT (Unreliable Zone) Published December 2019 Position of Dip Poles 2000–6000nT (Caution Zone) This map shows the magnetic declination across the world in 2020. It changes over time, but the values shown here for Australia and NZ are accurate enough for most contemporary uses of the Digital Compass. Locations near the poles drift more quickly. Map developed by NOAA/NCEI and CIRES https://ngdc.noaa.gov/geomag/WMM (published December 2019). Source: https://w.wiki/9fV6 The blob of solder under the compass module is a 5V connection you can follow back via the PCB tracks. We also made some of the ground connections via PCB tracks. Start by soldering the ground connections as shown in the photo of the shield’s underside. Follow with the blue wires to the switches. Each switch should now have a blue wire at one end and a black wire at the other. Next, solder the 5V wire (and 5V blob) and one 3.3V wire. Then there are two SDA wires and two SCL wires for the I2C bus. Slot the prototyping shield onto the Leonardo, being careful to align all the pins correctly. We are using the SDA and SCL pins near D13, so this should also work with an 70 Silicon Chip Arduino Uno board, although we have not tested it. The software operation is quite straightforward. The Leonardo’s processor reads data from the compass module, calculates a compass heading, then displays that on the 14-segment LED module. We have bundled everything into a single sketch folder, including a basic library for the display module (the file is named XC3715.h) and a library for the compass sensor. The latter comprises the HMC5883L_Simple files from the same library (from James Sleeman) that Jim Rowe used in his 2018 modules article. To program the Arduino, download Screen 1: the default display shows a heading in degrees and updates about twice a second. S2 and S3 adjust the brightness. Screen 2: pressing S1 toggles to an alternative display showing a cardinal compass heading and an arrow pointing north. Software Australia's electronics magazine siliconchip.com.au Silicon Chip PDFs on USB The colours here mostly match Fig.1, except we used yellow for 5V so it stands out from the red shield board. A blob of solder feeds 5V to the compass module (circled in yellow). Some of the 5V and ground connections are made on the shield, too. Using a prototyping shield makes this a compact and tidy project, although you could also try it out on a breadboard with jumper wires. the sketch from siliconchip.au/ Shop/6/430 and unzip it, then open it in the Arduino IDE. Select the Leonardo board option and its serial port via the menus, then upload it. The serial monitor will report some debugging data once that process has finished. After a second or so, you should see a display in degrees (see Screen 1). The displayed bearing should increase if you turn the Digital Compass clockwise and decrease if you turn it anti-clockwise. If you don’t get that reading, check the wiring to the display. If the bearing does not change, you might have a problem with the wiring to the compass module. The serial monitor will also display the bearing, so you can check that the compass module is working, even if the display is not. The default is to display a bearing in degrees, but pressing S1 (the leftmost button) will change to displaying a cardinal (N, NE, E, SE, S, SW, W or NW) compass point, which you can see in Screen 2. There is also a (somewhat squashed) arrow that will point north on the right-hand side of the display. Pressing S2 or S3 will adjust the brightness; the chip on the display module provides 16 steps. If you hold S1, the magnetic declination is shown, and pressing S2 or S3 while S1 is held will adjust it, as shown in Screen 3. You could either use a declination value from a web search or, if you know where north is, you could point the Compass north and manually trim the declination until the Compass reads 0°. After 10 seconds, you might see SAVE flash up on the display (Screen 4). That means the current settings have been saved to EEPROM and will be retained if the Compass is turned off. The settings are reloaded when it SC is restarted. ¯ A treasure trove of Silicon Chip magazines on a 32GB custom-made USB. ¯ Each USB is filled with a set of issues as PDFs – fully searchable and with a separate index – you just need a PDF viewer. ¯ Ordering the USB also provides you with download access for the relevant PDFs, once your order has been processed ¯ 10% off your order (not including postage cost) if you are currently subscribed to the magazine. ¯ Receive an extra discount If you already own digital copies of the magazine (in the block you are ordering). EACH BLOCK OF ISSUES COSTS $100 NOVEMBER 1987 – DECEMBER 1994 JANUARY 1995 – DECEMBER 1999 JANUARY 2000 – DECEMBER 2004 JANUARY 2005 – DECEMBER 2009 JANUARY 2010 – DECEMBER 2014 JANUARY 2015 – DECEMBER 2019 Screen 3: holding S1 allows the magnetic declination to be set. It defaults to 0° and can be set from 99°W to 99°E (−99° to +99°). siliconchip.com.au Screen 4: within 10 seconds of making a change, the Compass will save the settings to non-volatile EEPROM and show this message. Australia's electronics magazine OR PAY $500 FOR ALL SIX (+ POST) WWW.SILICONCHIP.COM. AU/SHOP/DIGITAL_PDFS July 2024  71 By Allan Linton-Smith Workman 1000W loudspeaker T his design follows on from our 1000W IRAUDAMP9-based amplifier, published in the October & November 2023 issues (siliconchip.au/Series/405). Finally, you can build a speaker that the amplifier can actually drive to its full potential! This quality loudspeaker can safely handle 1000W RMS for extended periods. The speaker is housed in a sturdy US-built DeWalt transportable 233L toolbox (DWST38000) that measures 99×59×62cm. That makes it light, portable, rugged and very easy to build, requiring only minor modifications to the toolbox/case as purchased. The DeWalt toolbox is available pretty much worldwide! The result is a portable but powerful speaker with many applications. As for the drivers, it uses the 8W, 15-inch (381mm) Celestion FTR154080FD (or FTR15-4080HDX) woofer rated at 1000W coupled with an 8W, 1-inch (25mm) Celestion compression tweeter (CDX1-1745) rated at 75W, attached to a Celestion “No Bell” horn. When set up correctly, these drivers can easily handle a combined 1000W for up to two hours. The only catch is that our 1kW IRAUDAMP9-based power amplifier can only deliver its full-rated power into 2W. Our Class-D amplifier will drive one of these Workman speakers at 400W before clipping. That might seem low, but the speaker is very efficient at 97dB at 1W/1m (96dB for the HDX driver), so it will still be incredibly loud at that power level! If you need to drive this loudspeaker at the full 1000W, you could build two of our Class-D 1kW amplifiers and drive it in bridge mode. Each amp will ‘see’ a ~4W load, and they can each deliver 575W into 4W (or 500W with lower distortion), so they achieve the full 1kW configured like that. We published an amplifier bridge adaptor in the May 2019 issue (siliconchip.au/ Article/11626). Design considerations 72 This seriously powerful and efficient full-range loudspeaker can deliver a tremendous amount of sound, and it doesn’t sound half bad, either. It can be used for public address, DJ and music applications (if you happen to own a stadium!). I decided that a PA speaker needed to be light, portable and ideally transportable by the average person. I have no trouble loading it into and out of my Hyundai hatchback by myself, so I consider that goal to be met. The DeWalt box is light, has built-in wheels, is very tough and is water resistant, with an IP65 rating. Australia's electronics magazine siliconchip.com.au Silicon Chip The prototype crossover was smaller than the final one. Either way, there is plenty of space left in the box. The IP65 rating is ruined by our installation of the speakers in the box but, with speaker drivers installed, if the box is covered by a large plastic bag (eg, a garbage bag), it should survive a shower during transportation. It could even be used with a bag over it, although the sound quality may suffer a bit! The box is rated for a maximum load of 70kg. The woofer, tweeter and crossover together weigh about 13kg, and there is still plenty of room inside, so you could even use the box to transport stuff like cables, microphones and so on (although you’d want to be careful they wouldn’t move during transport and possibly damage something inside). If you’re careful, you could fit a big amplifier inside the box (even our big 1000W amplifier would fit), and together with a preamplifier, it could become a very good mobile PA system. Just be careful you don’t move it in such a way that any large, heavy items inside will shift around! The finished speaker can sit on the siliconchip.com.au ground or be suspended via chains or wires through its two strong steel vertical handles or the telescopic carry handle at the top. That could be very useful at outdoor functions, theatres, discos, churches or other public areas. You can padlock the box shut at a venue so nobody is tempted to poke around inside. I designed the loudspeaker using the box as a sealed enclosure, mainly to simplify construction over a more complicated ported design. That also makes sense because the woofer has a VAS of 140L (111L for the HDX version), so it is not a problem to run it in a sealed 233L box. In a sealed box, the woofer had a measured resonance of 40Hz, only marginally higher than its 38Hz free-air resonance. The priority was to create a design that’s really easy to build, even if you only have rudimentary woodworking and soldering experience. You can make this over a weekend for around $1k (about $1 per watt)! That might seem expensive, but try pricing a commercial speaker that can actually handle 1kW RMS. Many claim “1000W” but would melt in short order at that power level! Much of the cost is in the case and the woofer, two areas where you can’t really cut corners. Performance The overall performance of this system relies on the incredible power-­ handling ability of the Celestion woofer combined with the superb quality of the Celestion tweeter. The tweeter is ‘only’ rated at 75W. However, it is incredibly efficient, so we can heavily attenuate the signal going to it and still get a good bass/treble balance while keeping it within its ratings. Not many single-speaker designs can handle this power level; remember that power-handling claims are commonly exaggerated. If you look at the Celestion woofer’s construction, it is a bit of a beast, with massive coils, magnets, and heatsinks that allow it to cope with that much power. You also have to consider efficiency – it’s no good having a really powerful speaker if you don’t get much sound out of it. This woofer’s 96-97dB <at> 1W/1m rating is excellent, and it means you will get a truly deafening sound level at 1000W (just what rock fans need!). Australia's electronics magazine Tweeter Specifications ● Diameter: 120mm ● Depth: 56mm ● Weight: 3kg ● Power rating: 75W RMS (tested for two hours) ● Nominal impedance: 8Ω ● Frequency range: 1.2-20kHz ● Efficiency: 110dB <at> 1W/1m ● Recommended minimum crossover frequency: 2.2kHz (12dB/octave) ● Voice coil: 44mm diameter edgewound copper-clad aluminium ● Magnet: ferrite ● Diaphragm: PETP film ● Throat exit: 25.4mm Woofer Specifications ● Diameter: 381mm ● Depth: 170mm ● Weight: 9.5kg ● Power rating: 1000W RMS (tested for two hours) ● Nominal impedance: 8Ω ● Frequency range: 35-2500Hz ● Efficiency: 97dB <at> 1W/1m ● Voice coil: 100mm diameter, 22mm wide round copper ● Magnet: ferrite (3.1kg) ● Chassis: cast aluminium ● Former: glass fibre ● Cone: glass-loaded paper with weather-resistant impregnation ● Surround: cloth-sealed ● Suspension: double ● Xmax: 6mm ● VAS: 140L July 2024  73 Fig.1: the overall frequency response of the loudspeaker (mauve) is reasonably flat, within about ±5.5dB of the average over the whole range. The cyan and red traces show the contributions from the tweeter and woofer separately. Fig.2: the distortion levels are better than expected for a loudspeaker of this type, remaining below 2% from 50Hz to 20kHz. The measurement bandwidth is 20kHz, so the low distortion from 1.5kHz to 20kHz mainly represents noise (most likely from cabinet resonances). Fig.3: the harmonic distortion at 1W without noise is much lower than the THD+N shown in Fig.2. Odd harmonics sound bad but are very low in comparison with even harmonics; the second and fourth harmonics are the same note as the fundamental but at higher octaves, so they are in tune with it. 74 Silicon Chip Australia's electronics magazine High power handling is also helpful for situations where a lot of bass, mid-range or treble boost is applied because the speaker will have a fair bit of ‘headroom’. In movies, for example, the sound can have a huge dynamic range; an explosion can follow a whisper. You don’t want your speakers clipping when that explosion happens. Frequency response The frequency response of a loudspeaker is important; it is arguably the single most important factor determining whether it sounds good or not. The response should be as flat as possible. It’s essential to avoid peaks that could exceed its maximum power limit when running near the limit. Peaks can also sound bad and possibly even damage ears at high sound pressure levels. Dips are also best avoided as they create ‘dead zones’ where specific frequencies seem missing from the sound. For example, notes running up and down a scale can seem to disappear at a particular point if there is a significant dip in the frequency response. In the past, many of our readers have used cheaper drivers than those we recommend. That can sometimes work well, but other times, the design really relies on a specific driver. In this case, the driver’s 1kW rating is quite unusual, so we have not tested any alternatives. We couldn’t find many that were genuinely capable of handling 1kW! A quick check on the internet showed that most 15-inch speaker drivers can only handle 100-300W maximum; even if you find one that claims to handle 1kW, you will need to check that it complies with the AES standards. The woofer has a really nice response from 50Hz to 1kHz and is excellent for the human voice and woodwind instruments. However, it really shines with guitars, especially in heavy metal music, which Celestion is famous for. Fig.1 shows the responses taken with a microphone placed in front of the woofer and tweeter and one between the two. The reference 0dB level is set to 100dB sound pressure level (SPL). The overall combined response is relatively flat, within ±5.5dB over most of the range, with no harsh peaks. The response around 300Hz is critical for siliconchip.com.au Fig.4: the spectral plot of a 1W 47Hz signal at an SPL of 92dB. The first harmonic at 94Hz is 48.7dB lower (0.32%) than the signal tone; the 2kHz peak represents a THD of 0.02%. Above 10kHz, the THD contribution drops to 0.002%. vocals, while the response around 2kHz is important for electric guitars. Distortion levels The measured total harmonic distortion plus noise (THD+N) levels were better than expected for a speaker housed in a plastic box, staying below 2% from 50Hz to 20kHz – see Fig.2. Note that the measurement bandwidth is 20kHz, which is why the distortion level drops so much above 1.5kHz, as many of the harmonics above that fall above the audible (and measured) frequency range. We also measured distortion only (total harmonic distortion minus the noise) and compared the contribution of the even and odd harmonics at 1W, as shown in Fig.3. Odd harmonics are generally considered to sound bad, so it’s good that they are pretty low compared to even harmonics in this design. THD by itself is always lower than THD+N. This speaker’s harmonics are quite low, indicating good overall clarity. High THD figures usually result in muddy sound. For completeness, we also plotted the spectrum of the distortion components for a 1W 47Hz signal at a sound pressure level of 92dB, shown in Fig.4. Impedance While the speaker’s nominal impedance is 8W, like both drivers, as with The crossover With one tweeter and one woofer (ie, a two-way design), we can get away with the simple first-order crossover circuit shown in Fig.6. You may think that the 2.2μF value of the series capacitor for the tweeter is low, but Celestion recommends a 12dB-per-octave Fig.5: the minimum loudspeaker impedance is 5.4W at 2.5kHz. Across the rest of the range, it stays above 8W except for a brief dip to 7.6W at 160Hz. As a result, virtually any power amplifier should be able to drive this speaker. The prototype had the crossover capacitor connected directly to the tweeter, along with an experimental inductor. Now the wiring connects the tweeter to the crossover PCB. siliconchip.com.au most speakers, it varies quite a bit with frequency, (see Fig.5). The measured impedance shows two peaks, one at 40Hz (the woofer resonance) and one at 1.1kHz (tweeter resonance). 2.5kHz is the crossover point, and the lowest impedance value measured was 5.4W, which should not be a problem for most amplifiers. Australia's electronics magazine July 2024  75 Fig.6: the crossover circuit is dead simple, using just a series inductor to cut off high frequencies to the woofer and a series capacitor so low frequencies do not reach the tweeter. The two 20W resistors account for the higher tweeter sensitivity compared to the woofer and also protect the tweeter from being over-driven. roll-off with a 2.2kHz cut-off frequency. A single-order crossover only rolls off at 6dB per octave, so our capacitor achieves the required low-frequency attenuation by having a higher cut-off frequency without the problems that come with a much more complicated and expensive crossover. The first-order crossover used is naturally designed to handle high power levels. Two high-power resistors in series with the tweeter reduces its level by around 15dB. The tweeter has a sensitivity of 110dB per watt at one metre, but the woofer is rated at 96-97dB/watt at one metre. So we need to attenuate the tweeter by 13-15dB to match the levels, depending on the exact sensitivities of your drivers. This is good because, as mentioned earlier, the tweeter can only handle 75W maximum. Simulation shows that for an average output power of 1000W, a 40W series resistance would dissipate 138.6W and deliver only 28.2W to the tweeter, ensuring it does not burn out. This may seem like overkill, but I see a lot of burnt-out tweeters in PA speakers. To build the crossover, we used the same two-way crossover PCB that we designed for the Majestic loudspeakers from June 2014 (coded 01205141), replacing the 4.7μF capacitor with a 2.2μF cap and replacing the 2.7mH choke with a 1.5mH coil with extra thick wire so it can handle the power. The Majestic used a few onboard 5W and 10W resistors for tweeter attenuation, but there’s no way they would handle over 100W. Instead, we make up a 40W 400W resistor from two 20W 200W ceramic ‘rheostat’ resistors mounted beside the PCB and wired to it. These are connected in series. They have a slider arrangement that allows you to vary the resistance. You can reduce it slightly if you want more treble. In our tests, we set the resistance in series with the tweeter to 32W to attenuate the tweeter by 14dB (27W gave 13dB attenuation and 37W gave 15dB). Setting it below 20W is not recommended, as you risk exceeding the tweeter’s maximum power rating. The shelving circuit included in the crossover for the Majestic speaker to boost high frequencies is unnecessary because our 1kW amplifier has a 20kHz ‘lift’ that is common with most Class-D amplifiers. We mounted the whole crossover assembly on a 420×320mm piece of plywood and connected the wires to the tweeter and woofer using springmounted connectors. You could use less expensive (and probably more reliable) eyelet lugs if you want to. The binding posts I used come standard with Celestion woofers and I really love them! My back gets stiff when I bend over to hook up everything in this deep cabinet and the spring posts save heaps of time fiddling around with nuts and bolts. I bought 25 pairs from AliExpress for around $50 and they simply bolt to the PCB pads. If using them, you will need six for this project; you can also get them from eBay for about 10$ per pair, including delivery (search for “spring loaded binding post” or try www.ebay.com. au/itm/134778989440). Construction This project requires minimal construction. All you need to do is cut holes for the drivers and connector socket, solder up the crossover, mount it, and wire it up. That’s it! You need to make a couple of modifications to the case first, shown in Fig.7. Start by cutting two small pieces of timber to block off the 24×15mm deep reinforcing channels in the plastic to ensure an airtight fit for the drivers before cutting the holes. These The final crossover arrangment. It mounts on to a piece of timber, which can then be secured to the interior of the enclosure. Wire up the resistors as per Fig.8, not the way shown in this photo 76 Silicon Chip Australia's electronics magazine siliconchip.com.au Fig.7: cut the holes in the plastic case as shown here. Use a jigsaw for the larger driver holes and a hole saw or stepped drill bit for the smaller hole in the side for the connector. The timber pieces shown are used for reinforcement and to help seal the enclosure. pieces can be cut from a length of 30×18mm pine using a plane or saw. Glue them into the case using contact cement and allow a few hours for it to set before cutting the holes. Once you have done that, mark out the circles using a compass with a light-­ coloured pencil or chinagraph pencil, then use a jigsaw and slowly cut the plastic and the timber to the specified diameter. Next, make a 24mm hole in one side of the box for the Speakon speaker socket. The side is good because the box will still lie on its back for transport or whenever the lid is opened, and the connectors remain protected. You also need to modify the clamps that secure the lid. These will rattle unless you glue some rubber to them, as shown in the photo below. The clamps should then clamp everything firmly into place. Mount the inductor on the PCB using a ~25mm M4 machine screw, washer and nut. Use Loctite so that vibration won’t shake it loose (do that for all the screws used in this project). Now solder the capacitor, inductor and resistor wires to the PCB, as shown in Fig.8, and attach sufficient lengths of wire to reach the woofer, tweeter and the terminals of the input socket. Make the wires long enough to allow the lid (with the drivers attached) to open while the crossover is still attached to the inside of the case. Once you have all the holes prepared, mount the crossover at the bottom of the box with tapped spacers, machine screws and washers, then prepare the drivers by sticking a felt Left: the holes for the woofer and tweeter don’t need to be the neatest cuts, as they are covered by the drivers. The timber fills the channels that runs behind it. Right: the clamps that secure the lid need to be modified by gluing some rubber to the top of the clamp. This stops the clamps from rattling. siliconchip.com.au Australia's electronics magazine July 2024  77 or rubber material around the edge so they will give an airtight seal between the plastic box and the speakers when mounted. Screw the speakers to the box with suitable wood screws. You can use machine screws, nuts and washers, but it’s a bit fiddly, and you will need some blue Loctite to prevent them from working loose from the enormous vibrations they will experience. Now take the wires for the woofer, strip the insulation off by about 1cm at the ends and insert them into the spring clips (or crimp them to the eyelets and attach them to the PCB). Attach crimp spade connectors to the tweeter wires and push them onto the tabs on the tweeter. Ensure you get the polarity correct; the positive wire goes to the red dot on the tweeter. Finally, solder the wires to the Speakon chassis socket by poking some wire from the crossover into the terminals, then solder them. Attach it to the case using 3mm machine screws, nuts and washers. Check all your wiring thoroughly, then attach a reasonable length of speaker wire to the Speakon plug, ensuring the numbers on the plug match those on the socket. Before connecting your speaker to an amplifier, check the DC resistance across those terminals to verify that there are no shorts. You should measure around 5W. If all is well, power up the amp, feed in a signal and slowly wind up the volume. Be aware that the woofer cone has a very stiff suspension to enable it to handle 1000W, so it may need a few Fig.8: only two components are mounted on the circuit board: the 1.5mH air-cored inductor and the 2.2μF capacitor. The two series 20W rheostats are wired between the 2.2μF capacitor and the tweeter’s positive terminal. The resistors are screwed together at both ends for physical stability, but on the left end, the two resistors are insulated from each other and the blue wire only connects to the one shown closer to the PCB. 78 Silicon Chip Australia's electronics magazine siliconchip.com.au hours to ‘break in’ before it reaches its desired bass response. You might notice that we have not mentioned or used acoustic wadding in this speaker. It probably would sound better with wadding, but we wanted to keep the cabinet empty to make working on it easier and so it can be used to store items like leads and microphones, as mentioned earlier. If you don’t plan to keep anything in the cabinet, it would be worthwhile loosely stuffing it with acoustic wadding, which is available inexpensively on eBay. It is easy to remove later if you need access to the wiring or crossover. Listening tests Too often, PA sound systems at music venues are sub-par. The Editor recently complained to me that he was at a music performance at a Sydney stadium and could hardly even figure out what song they were playing, despite being familiar with the performer’s work. It just sounded like square waves! That is usually the result of poorly set up audio systems that have been taken way past acceptable limits. So we wanted to make sure this speaker actually sounds decent. It’s no good if it can deliver lots of sound if it’s just noise. We tested the speaker with a few different genres of music and made some adjustments, reducing the treble a little as it seemed ‘too bright’. This can be adjusted using the slider on the ceramic resistor in series with the tweeter by loosening its nut and re-positioning it. It would be good to measure the resistance before and after adjustment so you know what you’ve done. Ours was initially set at 32W but later adjusted to about 35W. We used the FD driver, so if you’re using the HDX, it may need to be set a little higher. As mentioned earlier, don’t go below 20W. Vocals shone in our tests, and the sound was very clear for a PA speaker. Some heavy metal we tried sounded really good! Even though this is a mono arrangement, every instrument could be clearly identified. The lead guitar really ripped in typical legendary Celestion fashion. Running the speaker at 300W barely troubled it, and the office vibrated, literally rattling filing cabinets and anything else that wasn’t screwed down! So, if that’s your cup of tea, this speaker is for you! SC siliconchip.com.au Parts List – Workman Loudspeaker 1 DeWalt DWST38000 99×59×62cm (240L) tool chest [Bunnings 0154687] 1 Celestion FTR15-4080FD or FTR15-4080HDX 15-inch (381mm) 1kW 8W woofer [eBay 232329975153 or 144393382135] 1 Celestion CDX1-1745 120mm 75W 8W tweeter [eBay 234994171597] 1 Celestion T5134 “No Bell” horn [eBay 325534095052] 1 2-Way Passive Crossover PCB (01205141) [Silicon Chip SC2734] 1 1.5mH air-cored crossover inductor with 1.5mm diameter copper wire [eBay 386228967177] 1 2.2μF 250V metallised polypropylene crossover capacitor [Jaycar RY6952] 2 20W 200W variable ceramic resistors/rheostats [eBay 225854220831] 1 Speakon panel-mount socket [Jaycar PS1082, Altronics P0792] 1 Speakon line plug or Speakon cable [Altronics P0795] 3 pairs of spring-loaded binding posts (optional, to connect wires to crossover board) [eBay 392075305616 or AliExpress 4000282183682] Hardware & cable 1 4m length of ‘jumbo’ (~3mm2) figure-8 speaker cable [Jaycar WB1732] 1 pair of red/black gold spade crimp lugs [Jaycar PT4568] 10 yellow 5.3mm crimp eyelet lugs 2 M6 × 55-60mm external hex head machine screws and flat washers 12 M5 × 16mm panhead machine screws, flat washers and nuts 1 M4 × 25mm panhead machine screw, washer and nut 4 M3 × 16mm panhead machine screws, washers and nuts 2 M3 × 16mm countersunk head machine screws, washers and nuts 8 M3 × 6mm panhead machine screws 4 M3 × 10mm tapped spacers 18 8G × 18-25mm wood screws 1 1m length of 30 × 18mm pine 1 420 × 320mm sheet of thin plywood or MDF 1 5m length of 16×10mm D-shaped self-adhesive weather stripping [Bunnings 0011953] 1 5m length of 5mm-wide by 6mm-thick self-adhesive brush-type strip or equivalent (‘door/window seal’) [eBay 274371043462] 1 small tube of blue Loctite or equivalent thread locker With the drivers and crossover fitted, the box can still be opened easily and there is plenty of room for storage, or acoustic wadding if you want to improve the sound quality. Access is also good for maintenance. Australia's electronics magazine July 2024  79 Vintage Radio Experimental one-valve superhet radio By Fred Lever For commercial superhets, specific dual valves were designed to combine the functions. For example: • An RF amplifier integrated with the local oscillator (a ‘converter’ like the 6AN7) • An IF amplifier with diodes for detection and AGC (eg, the 6N8) • An audio preamplifier and output driver in the same envelope (an ‘output valve’ like the 6GV8) That gave the designers some scope for clever circuit arrangements. The 6Y9 was one of the last twin pentodes of the valve era and was used in TV sets. It seemed like an excellent valve to build the single-valve superhet radio. Concept and techniques I have built many superhets with traditional circuit techniques, using three or more valves. However, I was intrigued when I saw a suggestion that the 6Y9 dual valve for TV sets could perform the active functions required to make a complete superhet radio from antenna to speaker. I googled “one valve superhet” and, sure enough, many people have been there before me. However, each arrangement I found did not quite meet the requirements of a practical home radio set, or they used an uncommon valve. I won’t go into much superhet theory here as it has been covered extensively in these pages. A study of Wikipedia’s “Superheterodyne receiver” entry (https://w.wiki/8DYV) will fill a reader in on the concepts and explain some of the acronyms used. 80 Silicon Chip The basic principle is to mix the incoming signal with a signal say 455kHz above or below it, then filter out everything except the 455kHz component from the mixer. After that, we can amplify and demodulate that much lower (and fixed) frequency signal. A superhet AM radio can be easily built using three valves: an RF amplifier/mixer/oscillator, an IF amplifier/ detector and an audio amplifier/loudspeaker driver. That is about six functions jammed into those three valves. Australia's electronics magazine I took up the challenge, starting with a draft circuit originated by Ian Robertson. The resulting radio, described in this article, nearly met all the criteria. I mainly used junk-box parts and modified the theoretical circuit to suit the parts I had. This completed radio sits on a shelf and, with an indoor aerial wire, produces a couple of watts of sound through a five-inch (~127mm) speaker and tunes in all the local AM radio stations. The circuit, shown in Fig.1, uses every technique possible to provide the functions mentioned above from the single valve, including autodyne, reflexing, neutralisation and negative feedback. Negative feedback is a commonly used technique these days, but the others may not be that well known. Reflexing is a method of passing the radio signal multiple times through one valve at different frequencies. In this case, the RF amplification, local oscillator and mixing are handled in the first valve section, while the IF, AGC and audio functions in the second. Neutralisation is a form of positive (regenerative) feedback that cancels out unwanted, inherent negative siliconchip.com.au feedback to get more gain from a valve or transistor. ‘Autodyne’ is a very old superheterodyne single-valve technique used in the 1930s, subsequently displaced by the dual-purpose converter valves. Essentially the incoming signal is fed to a valve set up to oscillate at a different frequency, so it acts as both the oscillator and mixer. The theory behind these techniques can be studied by consulting the textbooks of the era, such as the Radiotron Designer’s Handbook. Did I cheat? I cheated a little bit in some people’s eyes by including some solid-state diodes in the circuit. I elected to use diodes for the power supply and the detector functions. The main components in the rest of the set are from the 1960s era or modern equivalents. Still, I think I got away with it because it’s still true to say that the only active devices in the circuit are within that sole valve envelope. The diodes (bridge rectifier) in the power supply are only needed because it’s a mains-powered set; had I elected to make it battery-powered, they could have been eliminated. That leaves the detector diode (D5) as the only part that might have needed another envelope back in the valve era, although other types of rectifiers were available back then, like selenium rectifiers. I used two 1960s commercial IF transformers but scramble-wound the tuning coils on repurposed coil formers. Other parts came from my junk box or the Jaycar stock bin. I certainly used new capacitors and resistors! Practical difficulties The aim of any radio set is to gather radio waves at microvolt (μV) levels out of the air, then select and amplify the signals in a particular frequency range to drive a loudspeaker coil with a few volts at audio frequencies. That implies a level of voltage amplification of thousands of times or more. That amplification is usually spread over a chain of tuned circuits, with amplifying valves interposed at strategic points to keep boosting the signal level. The standard practice is to keep each circuit input wiring well away from the output wiring, to minimise the chance of uncontrolled feedback turning into instability. However, in this set, we surround one valve with those series of tuned circuits, but keep feeding signals back into the same valve position for another trip! It is a fact that, by necessity, the input and output signals of each ‘stage’ are in close proximity. Circuit details We have two pentode sections, V1A and V1B. V1A combines the signals from the aerial coil/transformer and the oscillator coil/transformer. The aerial coil is connected to the control grid input at pin 1, while the oscillator coil is connected to the cathode at pin 2. The valve output at pin 4 has two loads stacked in series. The first intermediate frequency transformer (IFT1) load is tuned to respond only to 455kHz, while the second load, the oscillator coil, only responds to oscillator frequencies (around 1-2MHz). Fig.1: this radio circuit I developed utilises the first half of the dual pentode, V1A, as an Autodyne mixer/oscillator, while V1B is reflexed to act as an IF amplifier as well as an audio signal amplifier to drive the speaker transformer. The only ‘cheat’ is silicon diode D5 as the detector. siliconchip.com.au Australia's electronics magazine July 2024  81 V1A receives a tuned carrier signal from the aerial coil into pin 1, which appears on the plate at pin 4. The plate is also connected to the oscillator coil, which is phased as a positive feedback and is resonant. Feedback goes to the pentode cathode at pin 2. That input signal change accelerates the feedback through the oscillator coil, and the valve bursts into oscillation at the frequency determined by the resonance of the oscillator coil with its tuning capacitors. That oscillator signal also appears at plate pin 4. The plate circuit has a combination of station carrier sine waves and oscillator sine waves, the differences between those two, plus any modulation present. The signal thus looks like an unresolved blur on an oscilloscope, but by sweeping slowly, you can get an idea of the multiple RF waves with an audio modulation sitting on top. Once past the 455kHz trap, the IFT signal Scope 1: the yellow trace is the 455kHz IF signal, while the cyan trace shows the recovered 440Hz audio modulation. resolves a bit better. In Scope 1, the yellow trace is the 455kHz IF signal modulated at 440Hz (the blue signal). Consider a tuned signal carrier at 1MHz being fed into pin 1. An amplified version of this signal appears at the plate, pin 4. The oscillator coil is also connected to the plate through the IFT1 primary. As the oscillator coil acts as a feed-forward from the output (plate) to the input pin 2 (cathode), the circuit oscillates at around 1455kHz, which also appears at the plate, pin 4. There is a difference (beat) frequency of 455kHz (1455kHz – 1000kHz). As IFT1 is a 455kHz resonant trap, any other frequency at the plate of the valve is rejected, and only the 455kHz ‘beat’ modulated by the original audio program content gets through. It therefore arrives at the input control grid of the second section, at pin 8. IF and AF amplifiers Scope 2: the signal delivered to the speaker without the gimmick capacitor; it is distorted and full of RF due to the second high-gain stage oscillating uncontrollably. Scope 3: with the gimmick capacitor added, a couple of picofarads of extra Miller capacitance have increased stability to the point where the set is only oscillating at the desired frequency (455kHz above the tuned frequency), and the detected audio signal is clean. 82 Silicon Chip Australia's electronics magazine The second pentode, V1B, also has two loads stacked in its output plate at pin 10. The top load is a second 455kHz IFT that passes only 455kHz signals and ignores anything else. The amplified 455kHz signal from pin 10 is trapped by IFT2 and passed to the 1N4148 detector diode, D5. The conducting action of the diode clamps the positive half-cycle of the 455kHz carrier, leaving the negative half-cycle of the carrier wave and the audio-frequency (AF) modulation. That signal half-cycle passes through a low-pass RC filter (100kW/270pF) into a 1MW load. The filter removes intermediate frequency 455kHz signals but not the AF modulation nor the negative DC component. The negative DC level is fed via a 1MW isolating resistor to the AGC line that goes back to the input control grid of V1A at pin 2. This acts as a level control, reducing the set’s gain for stronger stations. The audio modulation is fed forward to the pentode grid input at pin 8 via the volume control, VR1, and IFT1’s primary. This time, V1B amplifies the AF signal (at the same time it is amplifying the IF signal!), and that appears at the plate output, pin 10. This AF signal is ignored by the top load IFT2 (acting like a small RF choke only) and develops across the output transformer’s primary. It matches the low impedance of the speaker (4W) to the high impedance of siliconchip.com.au Photo 1: the routing of the wiring under the chassis is critical since so many different signals meet at the valve base. The ‘gimmick capacitor’ formed by the green and black wires twisted together at lower middle provides a bit of extra feedback to the second stage (V1B) so it doesn’t burst into oscillation. the pentode plate (~10kW), and the AF signal is fed to the speaker. That is the basics of the circuit, where V1A amplifies frequencies that are pretty close together, while V1B handles signals that differ significantly in frequency. The gain of the first section is very low; certainly less than 10 times. The rest of the gain is in the second section, where near-heroic measures have to be implemented to keep the gain high and the stage stable. That is where the neutralisation comes in. Stability and neutralisation Overall stability with fair performance was first reached by a combination of shielding and bypassing. Then, when it became unstable with more gain, I implemented the magic neutralisation by deliberately bringing some output and input leads together to remove the instability. The latter technique was new to me and seemed like witchcraft until I studied relevant technical texts. They described what happens when a careful portion of the output energy is fed back to the input, with the promise that the stage gain could be raised siliconchip.com.au without instability. I did not believe it until I had the screaming unstable IF/ AF reflexed stage go quiet and docile simply by twisting two wires together to form a very small amount of capacitance from output to input! Editor’s note: “Neutralisation” refers to adding positive feedback around an amplifying device to cancel out its inherent negative feedback due to Miller capacitance, thus enhancing its bandwidth. While the added ‘gimmick’ capacitor in this case is similar to a neutralising capacitor, its purpose is slightly different. Here, due to reflexing, the Miller capacitance couples signals between the two IF transformers, one connected to pin 8 and one to pin 10. As they are both resonant at 455kHz, feedback can lead to unwanted oscillation. The gimmick capacitor reduces that coupling by partially cancelling the Miller capacitance, increasing stability. Normally, neutralisation would reduce stability due to the added positive feedback. In Fig.1, the neutralisation is shown diagrammatically by the wire connecting to pin 10 of valve V1B being capacitively coupled to the wire connecting Australia's electronics magazine to the volume control, VR1. They are the green and black wires that run up the middle of the chassis in Photo 1. In a typical set, the green wire would be kept short and well away from any valves, and thoroughly shielded to prevent unwanted coupling! A long run of a sensitive input wire inside the chassis over the valves can provoke the amplifying stage into regenerative instability, particularly in this case where both IF and AF signals are being handled. Scope 2 shows the signal going to the speaker without the wires twisted together, while Scope 3 shows the same waveform with them in close proximity, achieving stability. The result was an epiphany to me, having been brought up in the school of keeping output leads well away from input leads. The wild oscillations began to clear up as the wires were brought adjacent, with several twists being enough to remove all bad behaviour. If too many turns were made, the instability reappeared, there being a “Goldilocks” amount. Other stability components Some negative feedback is implemented for audio-frequency signals to July 2024  83 Photo 2: you can see the internal structure of the 6Y9 dual pentode in this photo. The right-hand quarter or so is the first pentode, V1A; the power pentode, V1B, occupies a much larger portion of the structure. Photo 3: the top side of the finished chassis. The HT is pretty low at 175V, generated from a 140V winding on the transformer, as the valve’s maximum anode voltage rating is 190V. roll off the supersonic response. A feed is taken from the speaker to the bottom end of the volume control potentiometer. The 5.6kW resistor in series with V1B’s grid and the capacitors bypassing the cathode resistor, all mounted directly on the valve socket, also improve stability by attenuating signals above the intermediate frequency. The circuit notes components that had to be mounted directly at the socket for maximum stability with asterisks. The process of achieving stable running was actually a long journey and hard fought. can be operated in an autodyne oscillator/mixer configuration, combining the tuning coil circuits. The power section can be employed in the reflex configuration, combining the IF amplifier and the AF amplifier/AF output. The key to its success is the colossal gain of the power section. Even though it is not being used in the intended application, which was for TV video amplification and CRT driving up to 5MHz, the gain and bandwidth are well-suited for use at 455kHz (IF) and 2MHz (upper end of the oscillator range). Without a separate triode to act as a local oscillator, the pentode V1A must be arranged as an autodyne converter. This is the weakest part of the set as the RF gain in this section, by virtue of the dual use, is relatively low. I was not successful in using an internal ferrite loop stick or loop antenna with this front end, so I settled on using conventional tuning and oscillator coils. Without the gain of a loop stick antenna, the set needs an external wire antenna to give good reception. The 6Y9 valve While the 6Y9 is a dual pentode, its two pentodes are quite different. The first section is a medium-gain signal amplifier, making it suitable for mixer/ oscillator duty. The second is a highgain power amplifier that can drive the speaker transformer. In Photo 2, the signal section at the right of the picture uses about 25% of the structure, while the power section is the remainder. The base of the valve has 10 pins that allow the electrodes of each valve to be accessed while keeping them separate. Because of this, the signal section 84 Silicon Chip Finishing the set I had to put a bit of ingenuity into obtaining or making the parts. I made Australia's electronics magazine the chassis from pieces of scrap sheet metal bent and pop-riveted together. I drilled holes where I thought parts should go, plus more, just in case. The front panel was part of a base plate from something with vent slots spaced just right to bolt the speaker onto and let the sound out. The larger parts you can see in Photo 3 are a motley crew of new, old and modified devices. The speaker transformer is a Jaycar MM1900, using the 0.5W tap. The power transformer is a Jaycar MM2011 rewound with 140V and 6.3V AC secondaries. The speaker is a Jaycar AS3008 4W unit. The tuning gang is a dual 500pF unit from my junk boxes. The tuning dial is a reduction type, also from the junk box. The IFTs were both from my junk box as well. IFT1 was from an Astor chassis and is marked 7872, while IFT2 was made by HMV and is marked 906 0062. I verified that both resonated at 455kHz before using them. I used these types as they came from valve radio chassis, so they should be happy with valve currents and voltages. The larger HMV unit for IFT2 has quite thick wire in it; I was mindful siliconchip.com.au Photo 4: the finished radio fitted into its case. IFT1 is on the right, while the beefy coil for IFT2 is in the middle. of the plate current of the 6Y9 possibly frying any miniature IFT. The important thing with IFT1 is that the primary winding impedance does not inhibit the oscillator frequency feedback. With some later experience making other autodyne sets, I feel that any valve-type IFT with ferrite adjusting cores and large resonating capacitors will work well. Under the chassis, the rest of the parts (except for the tuning coils) are what you have in stock or buy from Jaycar etc. I selected the components with reference to an article called “Radio Therapy” from Radio and Hobbies, November 1943 that gives a run down on autodyne radio sets and suitable parts. With its 140V AC HT winding, the power transformer output 175V from a bridge rectifier. I was mindful of the manufacturer’s maximum rating of 190V for the 6Y9, as well as advice from TV-era service techs that exceeding that voltage can cause valve failure. a shielding plate between them to remove weird whistles due to field interference. I rewound the coils several times during development, so like the rest of the set, they look a little messy with taps and bits of tape hanging out. Both primaries eventually tracked the necessary frequency ranges to suit the 500pF gang I used. I moved the secondaries several times to change the amount of coupling. The aerial coil resonates from 600kHz to 1800kHz, while the oscillator coil resonates from 955kHz to 2255kHz. The oscillator coil primary has a 430pF padder in series with the gang to make the ratio of frequency change nearer to 2:1, to suit the aerial coil ratio of 3:1. By fitting trimmers to the gangs and ferrite cores in the coils, I was able to tweak the tuning to get good tracking, and a near-constant 455kHz difference beat to feed the 455kHz IFTs. The tuning coils At this point, I had a chassis that worked as a usable radio. Still, to make the set truly practical, there had to be some sort of cabinet to house the chassis. I cobbled the tuning coils together from discarded plastic formers with the original windings removed. They are mounted side-by-side with siliconchip.com.au The cabinet Australia's electronics magazine I simply ran a tape measure around the chassis and, with scraps of Bunnings 5-ply, concocted a “kennel” cabinet for the set to live in. I nailed the bits of ply together and also Aqua glued them. Once set, I sprayed the wood with enough coats of waterbased white paint until it looked shiny. Conclusion The experience of making this single-­ valve autodyne-mixer practical receiver opened my eyes to the technology of the era. The process took several workshop months and resulted in many pages of tests and experiments, far too long to reproduce in this magazine. From the lessons learned from this project, I have made several other Autodyne radio sets with 1960s miniature and 1940s octal metal valves. The latter are the most well-­ developed as my understanding of the techniques improved. This process also answered the query: why did the autodyne die? For more details on this project, see my Vintage Radio forum posts at https://vintage-radio.com.au/default. asp?f=12&th=130 SC July 2024  85 SERVICEMAN’S LOG Computer abuse Dave Thompson As most of you probably know, I don’t write much about my day job, ‘fixing’ computers. Not because you wouldn’t be able to handle the dramatic highs and lows of such a high-octane job; it is completely the opposite. You’d be bored to tears for the vast majority of it. Still, now and then, something comes through the workshop that leaves me speechless! N owadays, most 10-year-old kids can fix the things that usually go wrong with computers. Modern software and children now handle what was once a very specialised field. Most of the time, it is mundane, with the odd curly problem thrown in to make things a tiny bit more interesting. Of course, those rare wins can be very satisfying, even though they are few and far between. More often than not, they are tempered by having to break the bad news to someone who has lost their data, or simply their machine, because it has reached the end of its useful life. It doesn’t help that most computers are now designed to last for a couple of years, then die, with all their memories lost, like tears in rain (cue the heart-rending speech from Blade Runner). Most tablets and laptops these days work fine for a while, then suddenly quit, or in the case of laptops (pun intended!), they physically break because the chassis and frames are made so thin in an effort to ‘add lightness’. With many machines, simply opening and closing them a few times a day – what I’d call ‘normal use’ for a laptop – will soon break them. The other issue with many portable computers is the power socket. On tablets, this is typically a micro-USB or a USB-C connector. On laptops, it will often be something proprietary, especially in the case of Dell machines, or something more familiar like that used in most Acers. Over the years, I have repaired hundreds of power sockets because they are so easily broken. We’ve likely all picked up a device by mistake that is still plugged in and put strain on the cable, or tripped over the cable laying across the floor to the nearest power point and sent the machine flying! The resulting leverage on some of the connections is considerable due to the size of the plugs and leads, and it is no wonder that sometimes the sockets get torn from the motherboard. However, that was not the real problem with a machine that came into my workshop recently. I repaired the power 86 Silicon Chip socket on this one a year ago, but it came adrift again. The client swears it hadn’t been dropped or mistreated, but as their dad literally had to bolt the screen back on, I had some questions. I don’t think I have seen a laptop in such a poor state, and I’ve seen a few! Young people these days... An adolescent owned this machine, and in my experience, they don’t tend to look after their tech (or other possessions) very well. The number of units that come through the workshop that are scarred from schoolyard use has increased over the years. Many schools now demand that students have tablets or laptops, often dictating which make and model they should bring to school. Usually, this is an iPad or iMac, among the most expensive tech you can buy. One wonders if the schools get a kickback on these sales, as they used to do when Apples were first introduced into schools in the 1980s. When I was at school, not quite the chalkboard days, we had to have a certain number and style of books. I remember standing in line waiting with a list of requirements as we all filed past stacks of exercise books of all descriptions. This, of course, carried on into universities, where students must buy reference books, usually written by professors there, to pass their classes. That’s a rort in itself, but a subject for another forum! So now most schools tell students they must have this Australia's electronics magazine siliconchip.com.au Items Covered This Month • Stress testing your electronics • Using a chlorinator for rust removal • Repairing a 27A switch-mode battery charger • Fixing an LG42LD460 TV power supply unit • Repairing WiFi-controlled LEDs 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 or that tablet or laptop. While Chromebooks are popular in some schools (and are at least reasonably cheap to replace when they are inevitably broken), most of the more prestigious schools require expensive makes and models. That is all well and good, but give a 13-year-old kid a tablet or laptop, and we know they are going to chuck that thing around, no matter how much it costs. I’ve seen many almost-new devices with broken screens, dented corners, drinks spilled down the keyboards, the usual stuff. This one was slightly different because I had already repaired the power socket once. If we do walk away with it plugged in when the lead runs out, the strain on the socket can wreck it. Whether it’s a micro-USB or USB-C charging port, as used in many phones and tablets, or one of the bigger sockets used on laptops, the result is often the same: breaking the socket, or some of it, or tearing it off the PCB inside the machine. These sockets are often held onto the board only by the soldered joints and perhaps a few Earthing tags. As those in phones and tablets are usually surface-mounted, there really isn’t much holding them to the board. Solder is not glue; it is inherently weak. If we’re lucky, four tiny fingers of metal extending from the socket case go through four corresponding holes in the ground plane on the PCB. Still, many of these sockets are just spot-glued with some type of component cement before being soldered in a bath or reflow oven. There is virtually no strength in that glue or connections, and in the scenario of walking off with it plugged in, often the socket is torn from the PCB. In many cases, a few of the PCB tracks go along with it. Getting inside the device Over the years, I have replaced hundreds of these sockets in phones, tablets and laptops. The main difficulty is that modern devices no longer have a back that can be popped off to reveal the screws that must be removed to access the ‘logic’ board. Almost all phones and tablets now require going in through the screen, which automatically makes them infeasible to repair, even if you have a hotplate for loosening the glue on the screen. Most screens are now so thin that even a slight twist will crack them, and new screens are often prohibitively expensive, if you can even get one! At least most modern laptops are easy enough to get apart, held together by about a dozen same-sized screws and some easily ‘poppable’ plastic clips. The whole bottom siliconchip.com.au (or top) typically comes away with just a few ribbon cables connecting the keyboard or touchpad to free up. It is nothing like olden-day devices that had tons of hidden screws of all weird sizes, which were often buried under the keyboard and even under the rubber bumper ‘feet’ on the bottom, which had to be pried loose to access the screws. Of course, they had to be stuck back down with new tape when it was all reassembled. The worst of those I can recall had a long, thin strip of rubber running the length of the machine, acting as a single long ‘foot’, which invariably tore when trying to remove it. Fun and games for computer techs back then! Returning to the laptop in question So, back to this one on the bench. As mentioned, I have repaired its power socket before. While not old, the computer has had a very tough life. Usually, when I repair one of these sockets, the repair lasts for the life of the machine because I use epoxy resin to bolster the strength of the socket once I’m sure it is all reconnected and working correctly. One has to be careful doing this because there is not a lot of room in these things, and it pays to make sure the top (or bottom) cover will fit on once a fillet of rock-hard glue is added around the socket! I don’t know how this current damage was done – the kid assured the parents that she’d not been rough with it, but the fact remained that the socket and motherboard were once again rendered asunder. I could feel it floating around in there when I probed it with one of my dental picks. Opening it up was interesting, because most of the plastic that made up two of the corners, where the hinges for the screen usually live, was simply not there. The screen itself had a diagonal crack in the top-left corner. I don’t think I could damage a machine that badly if I tried! The screen frame had popped open, and the clips that are usually there were missing in action, possibly floating around the back of the panel somewhere. The most obvious hint that something was amiss is that the husband had broken out his DIY skills: he had drilled out and held the righthand hinge and screen together with two nuts and bolts, which he likely got from the local hardware store. I thought I had pretty much seen everything over the past 25 years of doing this job, but I haven’t seen a post-­industrial repair like this before! Australia's electronics magazine July 2024  87 That said, the repair he had made was quite sound. I have no idea how he missed the LCD panel while drilling the two holes; it was likely pure luck. If he’d touched even the frame of it with the drill bit, it likely would have put yet another crack in the screen, and if he had gone through one of the ribbon cables or connectors, it would have been game over, man, game over! In the above photo, you can also see the black-and-white WiFi antenna wires, which are ultra-fine shielded cables that usually wrap around the edges of the panel, ending up at the top, on either side of the camera. In this case, the black one should be wrapped around the left-hand side of the panel, instead it is hanging in the breeze through the broken hinge section on the right side. Both of these wires start on a removable/replaceable module on the motherboard and are routed through channels designed for them, then pass through the lid hinge’s normally enclosed cavities into the frame around the screen. Not any more; on this side, the black lead has been bundled with the white lead, and the white wire has been severed almost at the hinge area; it simply pokes out into space. I imagine this would compromise the WiFi performance, but I have seen machines perform quite well before with those wires wholly disconnected from the module. That can happen when a tech has neglected to replace the tiny coax plugs at the ends of these cables onto their tiny sockets. On older machines, to get the motherboard out, I often had to take the WiFi module out, as it was fouling on the chassis, and in some cases formed part of the motherboard support structure due to it being screwed into place. Those antenna leads must be routed into their channels and reconnected during reassembly. If the user is close enough to the router, it would still work, but the operating range would be much shorter. It certainly isn’t ideal to have one poking out the side of the machine! So, the hinges are compromised on both sides, although only the right side has been fixed with bolts from the shed. The left hinge floats in the broken panel frame; the little metal inserts in the plastic turrets that usually support the hinge screws have all broken away and are sitting on the screws still attached to the hinge. This was looking increasingly pointless to repair, and I hadn’t even gotten to the motherboard yet. 88 Silicon Chip Disassembling the case wasn’t as difficult as I thought it might be because most of the screws weren’t holding onto anything anyway! Most of the clips that usually kept things nicely flush had sheared off, so it was simply a matter of getting a spudger (in this case, a guitar pick I use to open cases) and going around the edges to see what would come apart. A few of the screws still bit into their threads and held the case together, so I removed those. The bottom came away cleanly and I could clearly see the power socket floating loose in the recess that usually helped locate it. The plastic surround that helped support it had broken away last time, and I could see that some of the epoxy resin was still where I’d put it, except that the socket had broken out of it. There must have been a lot of force because that stuff is pretty tough, especially when it has lots of nooks, crannies and surface area to grab hold of. Given that the plug on the power supply was bent at a crazy angle, someone had really yanked on it to cause this much damage! I fished the socket out and could see it had torn the connections from the back of the socket. It had also taken some of the corresponding PCB tracks off the motherboard; there was nothing left there to solder anything to. With the previous repair, I could at least resolder the socket before testing and gluing it in place, but now, a new motherboard was the only real option. The glue I had added previously had also torn some tracks as part of it broke away, making even more of a mess. Maybe it wasn’t such a good idea after all, gluing it on, but I have done dozens of repairs like this over the years, and to the best of my knowledge, all those repairs lasted for the laptops’ natural lives. This one is an anomaly; the sort of damage I’m looking at can only be caused by gross mishandling or perhaps being thrown in a tantrum or similar. I didn’t ask; all I was told was that the daughter assured her dad she hadn’t dropped it. Maybe she was using it as a cricket bat! I don’t know, but I’ve seen less damage from machines that have fallen down the stairs or have been run over. I went through the motions of trying to source a ‘new’ motherboard, but the best I could do was a ‘refurbished’ one out of China. The last one I bought from sources there Australia's electronics magazine siliconchip.com.au didn’t work on arrival, and I took a big hit on that, so I wasn’t keen. Given that the rest of it was in such dire shape, I made the call to the dad and told him the bad news. He was philosophical about it, likely envisaging having to buy a new device for his daughter so that the same thing could happen to it. Parenting is tough, I guess! I put it back together as best I could, and they duly came and picked it up, the daughter looking suitably sad and sorry. I gave them the good news that her data was all there, so I could easily transfer it when they got a new one. I didn’t have the heart to charge them after returning the wreckage. Repurposing a pool chlorinator I spotted a couple of pool chlorinators on a recent visit to the local tip shop. One was pretty beaten up, but the other was in good condition, apart from having the power cable cut off. I decided to grab the better one of the two, as I’ve been using one of these units for my electrolytic rust removal bin for several years, and it would be handy to have a spare. When we got home, I put it away and didn’t think any more about it. About a week later, my pool chlorinator stopped working. On disconnecting it, I noticed that one of the front panel LEDs flashed. Investigating further, I found that the 35A bridge rectifier had failed. I have spare rectifiers, but I decided to check out the “new” one to see if it worked. If it did, I’d put it into service instead of repairing the one I had been using. I could see why the power cable had been cut off; the outlet socket on the bottom of the unit was broken, exposing live terminals. The first thing I did was to remove the outlet socket and rewire the unit to bypass it and the timer, which I didn’t need anyway. I then fitted a new power cable and plugged it in. The ‘No Water Flow’ LED lit up, indicating that the unit most likely worked. The unit has two heavy output leads, one for positive and one for negative, plus a thin wire for water flow monitoring. I connected the thin wire to the negative terminal, and the unit then started pulsating. That was unusual, but it was likely because it didn’t have a load. siliconchip.com.au I connected it to the rust removal bin, the pulsating stopped and the Chlorine Output meter showed that the unit was working. It could be adjusted by turning the output control. The old chlorinator has a 9V 20A transformer, which I could always use for something else. The replacement unit has a 9V 27.8A transformer and two bridge rectifiers instead of the one in the old unit. So this unit is more powerful than the old one. I made two plates, then pop riveted one over the hole where the timer had been and the other over the hole where the outlet socket had been. That made it ready for use. I have been using electrolytic rust removal for several years. It is very handy for cleaning up rusty tools and other items but only works on ferrous metals. Once, I picked up a 6-inch vise at the tip shop that was totally seized. After three days of treatment, I got it apart, and another three days later, it was cleaned up and I could paint it and use it. B. P., Dundathu, Qld Editor’s note: for information on the process, see our article in October 2014 (siliconchip.au/Article/8041). Australia's electronics magazine July 2024  89 Switch-mode battery charger repair This Innovative Energies SR750-24 charger converts its 230V AC input to 27.5V DC at up to 27A. The charger had not been used for some years, so its failure was not a great surprise. I agreed to see if it could be repaired, but an online search showed that it was a legacy item and quite a few years old, so spare parts supply could be difficult to obtain. With some degree of trepidation, I took possession of the charger, which was not easy to carry as it weighed 5kg and measured 350 × 200 × 75mm. Removing six small screws allowed the cover to be lifted off and the reason for the weight became apparent – it was fitted with two massive internal aluminium heatsinks and four large coils (chokes and/or transformers). The probable main reason for the failure was also immediately obvious: two 0.68µF 275V AC WIMA capacitors had ‘spilt their guts’ of the insulating oil contained within, which was spread all over the main PCB in their immediate area. I considered that the oil could contain nasties, so I kept my bare hands away from it. This PCB area included several small common components and ICs. Later investigation revealed this circuitry was in the charger voltage and/or current control. The first problem was removing the main PCB from its case to access these capacitors’ solder side. That turned out to be quite simple, as the mains input wired connections and the output DC connections are easily pulled off PCB-mounted spade connectors. Then, after undoing eight small screws, the PCB lifts straight out. This revealed that each of these capacitors was wired in series with the incoming mains Line and Neutral conductors, coming before several mains filter/noise suppression components. This also indicated that these capacitors were being used as simple AC voltage-dropping devices, so their capacitance values would not be particularly critical. As the series mains input 10A fuse was not blown, other major components had probably not failed. The high-­ voltage stress placed on these aged capacitors when the unit was switched on could explain their failure. The next step was to procure and fit replacement capacitors of similar size and specification to the original failed capacitors. The closest available value stocked by Jaycar was 1µF, so I fitted them temporarily (they have a 50Hz reactance value of 3.2kW versus 4.7kW for 680nF). I thought that was close enough for a go/no-go test. I cleaned the spilt oil off the PCB as much as possible before soldering in the new capacitors and reconnecting everything. While standing well clear, I switched the mains power on. This smoke test was successful as there was no smoke, no component appeared to overheating, the Charger Load display came on and the DC output measured about 27V. I subsequently connected the charger to a suitable 24V lead-acid battery and it charged the battery in the expected manner. After a further soak-test period, it was still operating. So, another electronic unit that would have cost a considerable sum to purchase in its day was saved from the scrap heap by about $5 of replacement parts. G. C., Wellington, NZ LG 42-inch LCD TV repair After a successful service of 15 years, my LG 42LD460 LCD TV failed. During the last cricket World Cup, hosted by India, the TV served flawlessly! However, I have been unable to switch it on for the last three weeks. The LG TV expert came and opened the back cover deftly. After a few tests, plugging in and out a few cables here and there, he Left: the failed SR750-24 battery charger. Right: the 42-inch LG LCD TV 90 Silicon Chip Australia's electronics magazine siliconchip.com.au joyfully declared that the power supply unit was beyond repair. When I asked for a replacement, he chuckled and, after a few calculations, revealed that it would cost me about INR 11000 ($200) for a replacement board and would take about a month to get. He said spending so much on repairing a 15-year-old TV was not a good idea when a new and better TV only cost about twice as much. After he left, I brought down the TV from the wall, opened the back cover and checked the power supply. I had to remove 16 screws to remove the back cover and found two boards, the TV motherboard on the left and the power supply on the right. The power supply has four connectors: the mains input, a ribbon cable to the motherboard and red and black wires supplying the backlight. While all the other cables came out easily, the ribbon cable gave me some trouble. I spread the locking fangs with a small screwdriver and it came out. The power supply board has a surface-mount fuse, which I knew the LG expert checked. I checked the first bridge rectifier of four diodes and it was OK. All the electrolytic capacitors seemed fine, with no bulging, and I didn’t see any blackened resistors. The rectified DC finally comes to a six-lead STR-W6053N IC, which converts it to lower DC voltages for the motherboard. The IC’s data sheet reveals that there is a feedback signal that comes from the TV motherboard through lots of circuits. If the motherboard does not produce the requisite feedback signal, the PSU will not work. Since the repairman said the power supply was dead, I removed it and took it to a local TV repairman. He told me that the TV is required; he can’t do anything without it. I asked him if he could visit my house to check the TV in situ as shifting a large TV is difficult! His answer was no. I searched the internet for TV replacement parts and found many shops that deal in TV spares, but couldn’t find my model listed. Frustrated, I rang one shop, and they quoted me around INR 4500 ($82) for a new power supply. I immediately placed an order and crossed my fingers! The power supply arrived four days later, neatly packed in a box. I opened the box and fitted the module into the TV. Nothing happened! It was still dead. Frustrated, I called the shop again and told them that the module seemed defective as the power supply was still not coming up. The man on the other end of the phone had a good knowledge of TVs. He asked me all the details of the fault: how it started, what the branded TV expert opined and my amateur repair process. He then told me that, in all probability, the TV motherboard was defective, not the power supply. He also informed me that he had seen many similar cases of incorrect diagnoses by the branded experts. He told me that I could return the power supply module and get a refund or buy a TV motherboard. If the power supply turned out to be OK, I could return it for a refund. Therefore, I end up coughing out another INR 5000 ($91) for a motherboard card. The card arrived in three days from Hyderabad. The motherboard card replacement is a little more delicate as it involves the removal of two flat ribbon cables, but they came out pretty easily (just open the lock...), and the new card was in place quickly. And the TV woke up after a slumber of 29 days! Before boxing up the TV, I tried siliconchip.com.au Australia's electronics magazine July 2024  91 the old power supply, and it also turned out to be working! So the TV motherboard card was the culprit all along. The power supply card has been returned for refund and the TV is still working well. At one point, I almost made up my mind to get a new TV as my wife was missing her daily serial shows. Only my persistent mind kept the hope of a successful repair alive. TV companies will always lure you into making new purchases by telling you that parts for older models are unavailable etc. But don’t get fooled. The old parts are available all over the internet. Search for them or even call them. They will certainly help you out because they also want to sell their goods! B. S., Kolkata, India Repairing a WiFi-controlled mains LED light I purchased several WiFi-­enabled ‘smart’ ceiling lights from Bunnings a while ago. A nifty app allowed me to change the brightness and colour temperature using a smartphone. While that was great in theory, we found that we typically set the colour and brightness once and then, around 99% of the time, simply used the light switch to turn them on or off when needed, as these lights were in our dining room and TV area, not in a bedroom. However, these clever lights proved highly unreliable. Two out of four failed and, as is the norm these days, it happened just after the warranty expired. Initially, they only partially failed, refusing to respond to WiFi commands but still turning on via the light switch at a usable brightness. So, we tolerated this unfortunate failure and continued using them as ‘dumb’ lights. Regrettably, after several months, both lights failed entirely within a few weeks of each other. Not wanting to spend a significant amount on a new light that would likely be just as unreliable, I decided to investigate whether they could be fixed. Removing the first light, I found that all the electronics were on a large PCB, which appeared to be made of ceramic or a similar material. Whatever it was made of, it was clearly designed to dissipate the heat produced by LEDs, and it did so very effectively. Due to its excellent heat conductivity, I had to use a much larger soldering iron tip than usual. I also noticed that the light’s white metal outer case acted as a heatsink, a necessary feature since excess heat is the leading cause of premature LED failure. 92 Silicon Chip Tracing the circuit proved extremely difficult due to the opaque ceramic substrate. However, the basic design quickly became apparent. The mains was converted to a current-limited 24V DC using a WS-LV24-G24L-L LED driver module. Next was a tiny SMD 5V switching regulator powering a TYWE3S 16-pin ESP8266 WiFi controller module, along with an unmarked microcontroller. Finally, a couple of Mosfets were used to drive the two banks of LEDs (daylight [5000K] and warm white [3000K]), with each Mosfet controlling a single bank. These two banks allowed the app to set the light’s colour temperature by adjusting the power ratio of the two LED banks. A multimeter confirmed that the 5V power rail was OK, measuring correctly at both the MPU and WiFi modules. Next, I measured each SMD resistor, and they were all within spec. I then unsoldered and tested each Mosfet, and they were also both OK. The fault must be in either the PCB tracks, the MPU or the WiFi module. However, it was virtually impossible to check the PCB without a circuit diagram due to its optically opaque substrate. It appeared impossible to replace the unmarked microcontroller without a part number or, more importantly, a way to obtain the firmware. The last suspect was the ESP8266 WiFi module, but a Google search showed only one hit for it worldwide, and it was nearly as expensive as a new ceiling light. So, it appeared that replacing any of these parts wasn’t a viable option. After some thought, I wondered if it was necessary for the WiFi feature to work. I could convert the lights to ‘dumb’ lights with just one or two resistors! As I prefer daylight white LED lighting, I only needed to solder one 5W, 10W resistor between the +24V supply and the daylight white LED bank, bypassing the Mosfet. If I preferred a more ‘warm’ light, I could have used two resistors, one for each LED bank, although the resistances might have needed to be increased to get the right final brightness. While not the most elegant solution, it converted the non-functional ‘smart’ lamps into functional ‘dumb’ lights controlled by a light switch. While it would have been more elegant to repair the WiFi functionality, this makeshift solution got both lights working for less than $2, which was much better than spending $150 to replace both with new smart LED light fixtures. SC G. C., Cameron Park, NSW Australia's electronics magazine siliconchip.com.au Make building or servicing easier with our Magnifiers & Inspection Aids 4.3" OLED GREAT FOR TECHNICIANS OR ADVANCED HOBBYISTS 600X ZOOM ONLY 139 $ POWERFUL 127MM DIA. 3-DIOPTRE LENS Digital Microscope • LED illumination • Rechargeable QC3193 FULLY ADJUSTABLE Clamp Mount Desktop Magnifier FULLY ADJUSTABLE ARM with LEDs • 1.75x, 2.25x & 3x magnification • 60 LEDs with high/low brightness • Mains powered QM3554 ONLY 139 $ ONLY 3695 $ RECORD & SNAPSHOT FEATURE FOR A BETTER VIEW LED Headband Magnifier • 1.5x, 3x, 8.5x 10x magnification • Can be worn over eye glasses LARGE 4.3" COLOUR LCD QM3511 720P WITH ILLUMINATION LED ILLUMINATION ONLY 13 $ 95 Handheld Magnifier • 3x magnification • Lightweight, just 200g Inspection Camera • 3x magnification • 3 x probe attachments included • Add an SD card to record vision or snapshots QC8718 QM3535 ONLY 269 $ Shop at Jaycar for: • Eye Magnifier • Handheld Magnifier • Headband Magnifier • Desktop Magnifiers • Inspection Cameras • Digital Microscope Explore our wide range of magnifiers & inspection aids, in stock on our website, or at over 115 stores or 134 resellers nationwide. www.jaycar.com.au 1800 022 888 Prices correct at time of publication but are subject to change. Jaycar reserves the right to change prices if and when required. 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. JFET-based Guitar Preamplifier There has been much debate regarding the relative merits of valve-based and solid-state guitar amplifiers. While modern solid-state amplifiers can obtain low levels of noise and distortion, this is in some respects contrary to the requirements of a guitar amplifier, where its tone is an integral component of the instrument’s sound. The guitar amplifiers I have built in recent years attempt to bridge this divide by using solid-state components in combination with junction field-effect transistors (JFETs) in lieu of valves. They have some similar characteristics, such as high input impedances and voltage control. The transfer function is a major difference between the two that significantly impacts the tone of a given amplifier. Valves have a ‘threehalves’ transfer function; the slope of their input-to-output function is x1.5, whereas FETs have an x2 transfer function. This difference has a significant effect on the harmonics that are generated by each device. Valves tend to produce predominantly even harmonics, while FETs generate odd harmonics in their ‘native’ form. Even harmonics are generally preferred as they are effectively the same notes at higher octaves (the second harmonic is one octave up, while the fourth harmonic is two octaves up). Odd harmonics do not usually have a musical relationship to the fundamental, producing a harsher sound. A few years back, I stumbled upon an interesting document written by Dimitri Danyuk that was presented at an Audio Engineering Society conference in 2004 (see siliconchip.au/ link/abwd). He explains how FETs can be configured to emulate a valve amplifier stage by precisely controlling their bias points and with local negative feedback created by an unbypassed source resistor. That creates the ‘three-halves’ transfer function associated with triodes using a FET. The main problem with this is that the parameter spread of FETs is quite broad, so the circuit needs to be tuned to each individual device. For each device, we need to measure the gate-to-source cut-off voltage (Vgs) and the (saturated) drain current (Idss) at 0V Vgs. I found that even devices from the same production run can vary widely, often between double and half of the nominal figure specified in the data sheets. Using the information provided in the Danyuk paper, the two parameters discussed above and the intended supply voltage (Vcc), we can calculate the required source (Rs) and drain (Rd) resistances: Rs = 0.83 × (Vgs ÷ Idss) Rd = 0.9 × (Vcc – (2 × Vgs)) ÷ Idss The units are the standard volts, ohms and amps (not mA). There is also an excellent online calculator available at siliconchip.au/ link/aakp – together with lots of other useful information. Silicon Chip kcaBBack Issues $10.00 + post January 1997 to October 2021 $11.50 + post November 2021 to September 2023 $12.50 + post October 2023 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 94 Silicon Chip Australia's electronics magazine For example, if we have a 2N5457 with a Vgs of 0.8V and an Idss of 3.5mA, we can calculate Rs = 0.83 × (0.8 ÷ 0.0035) = 190W (use the nearest standard value of 180W). Assuming a 20V supply created from a 24V source with a low-pass filter for this stage, we can then calculate Rd = 0.9 × ([20 – {2 × 0.8}] ÷ 0.0035) = 4731W (the nearest preferred value is 4.7kW). Using this technique, one can calculate resistor values for multiple FET stages. One critical factor to remember is that FETs have a lower input drive capability than valves. For example, the 12AX7 first amplifier stage in a tube amp can handle around ±2V before any significant distortion occurs, whereas a FET with a Vgs of (say) 0.5V can only handle a quarter of that. You can compensate for this by attenuating the incoming signals. The accompanying circuit shows how JFETs configured in that way can be arranged to form a useful guitar preamp. It broadly follows the general arrangement of most mid-level guitar amps, including a scaled version of the classic FMV tone control circuit, which can be omitted and bypassed if desired. It has conventional ‘HI/LO’ input options and provides a line-level output suitable for connection to an external amplifier. Its operation is as follows. The incoming signal is via CON1 or CON2, providing a choice of high impedance (at a lower level) or low impedance (at a higher level). Input levels can be critical, as mentioned above. The 47pF capacitor provides some RF filtering, while ZD1 & ZD2 protect against voltage spikes. The 10nF capacitor couples the signal to the first JFET and was deliberately chosen to limit the -3dB low-frequency response to around 20Hz, in conjunction with the 1.5MW resistor. The 4.7kW and 180W resistors connected to Q1 are the all-important siliconchip.com.au source and drain resistors, calculated from the preceding formulas (for one particular 2N5457) with a 20V supply. The 100μF capacitor and potentiometer VR1 provide an additional 5dB or so of gain when VR1’s resistance is set to zero, but it also bypasses the source resistor, effectively converting Q1’s transfer function back to x2. This has been found to be useful by some who want to alternate between the ‘warmer’ sound and something with a little more ‘crunch’ (odd harmonic distortion). For the best control range, the pot should be around 10 times the value of the source resistor. The signal is then coupled to the input of Q2 via a 100kW potentiometer, effectively acting as a volume control or, if there is a downstream control, as a gain control. The resistor values for this stage are calculated as above but allow for different FET characteristics and a higher supply voltage as less decoupling is used. A 2N5458 has been chosen for Q2, which has a higher Vgs and thus is capable of handling the higher-level signal amplified by Q1. However, it should also be capable of being deliberately driven into distortion if that is what the player wants. Q3 forms a source-follower to siliconchip.com.au provide a low-impedance output. The FET source and drain resistance values can be calculated as before, but the ‘drain’ resistor goes in the source, in series with Rd, and the gate bias resistor connects to the junction of the two, as shown. While a JFET is used in this stage, a simple BJT emitter-follower could also be used. The remaining blocks in the circuit diagram (green, yellow and blue) are optional. The first one is an FMV-style tone control. I have scaled the values down by 20 times because I don’t like using high (megohm) resistance values that are prone to picking up stray noise, as well as generating their own. The tone control section should be driven by a low-impedance source, as provided by Q3. The output of the tone control section needs to be loaded by a high impedance. That is provided by the Boost Amplifier section, which compensates for signal attenuation by the Tone Stage. If the extra gain is not required, this section may be omitted as the following Output Amplifier stage has JFET Q5 configured as a source-follower identically to Q3. The signal from Q5 is AC-coupled to the output connectors to remove Australia's electronics magazine the DC bias, while VR6 provides a final volume control for output CON4; CON3 is always fed with the unattenuated signal. Op amp voltage-follower buffers could be used in place of Q3 and Q5 if desired. A TL071 or similar would do the job, remembering that a single-­ ended DC supply powers the circuit. As noted earlier, some component values will need to be selected according to the exact JFETs being used. They can also be tweaked to achieve the desired tone. Of the amps I have built recently, I can’t recall two of them being exactly the same! What about an EQ, a power amplifier and a power supply to run it all? Most of the amps I have built include a PA stage (or the ability to connect to one), ranging from 5W to 500W. There are many options there. One of the more popular versions I built uses power Mosfets and includes an output transformer, like a valve amplifier, which makes a big difference as it introduces a lot of distortion. Still, that is a story for another day. As for the EQ, I will present a suitable circuit in a future Circuit Notebook entry. Graham Bowman, Duncraig, WA. ($120) July 2024  95 Op amp based push-pull PWM Mosfet driver You might have seen my Circuit Notebook entry on the replacement switchmode power supply for a BWD oscilloscope (May 2024; siliconchip. au/Article/16258). It used a TL494 controller IC with an external Mosfet driver IC, plus plenty of passive components, to drive a pair of power Mosfets. They were connected in a push-pull configuration to the primary of a high-frequency power transformer. After designing that, I thought there had to be a better (simpler) way to do it, so I developed the circuits shown here. The circuit snippet at the top uses a dual op amp to generate a triangular waveform at point A. This is then fed to a dual high-speed comparator, shown below, which produces the alternating pulses to drive two Mosfet gates (PHASE 1 & PHASE 2). A Mosfet driver would be needed at higher frequencies to provide sufficient current to rapidly charge and discharge the Mosfet gate capacitances for efficient switching. Still, this circuit uses fewer and less specialised components than the TL494-based circuit I presented previously. VR2 adjusts the duty cycle while VR1 is used to symmetrically centre the signals for maximum dead band. The dead band (non-overlapping) time is set by the values of the 4.7kW resistors; if changed to, say, 10kW each, the dead band would be larger. In that case, VR1 might not be required. I tested the circuit with a 12V supply but it also works at 5V. Scope 1 shows the voltages at points A (yellow) and Scope 1 Circuit Ideas Wanted 96 Silicon Chip B (cyan). Scope 2 shows the PHASE 1 and PHASE 2 outputs at a high duty cycle setting, while Scope 3 is the same display but with a lower duty cycle setting. For use in a switchmode controller type circuit, potentiometer VR2 would be replaced with a control signal that Scope 2 varies between Vcc ÷ 2 and Vcc, fed to one of the then-free comparator inputs. Another op amp would be needed to invert that signal (keeping it between Vcc ÷ 2 and Vcc), with its output going to the remaining comparator input. Mauri Lampi, Glenroy, Vic. ($100) Scope 3 Got an interesting original circuit that you have cleverly devised? We will pay good money to feature it in Circuit Notebook. We can pay you by electronic funds transfer, cheque or direct to your PayPal account. Or you can use the funds to purchase anything from the SILICON CHIP Online Store, including PCBs and components, back issues, subscriptions or whatever. Email your circuit and descriptive text to editor<at>siliconchip.com.au Australia's electronics magazine siliconchip.com.au Subscribe to JUNE 2024 ISSN 1030-2662 06 9 771030 266001 $12 50* NZ $13 90 INC GST INC GST ESR Test Tweezers Rosehill Gardens Sydney 19-20 USB-C Serial Adaptor June 2024 DC Supply Protectors Australia’s top electronics magazine Privacy cy Phones How important is your priv acy when online? 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. 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 $70 $80 $52.50 1 year $127.50 $147.50 $100 2 years $240 $275 $190 6 months $82.50 $92.50 1 year $150 $170 2 years $285 $320 6 months $100 $110 1 year $195 $215 All prices are in Australian dollars (AUD). Combined subscriptions include both the printed magazine and online access. 2 years $380 $415 Prices are valid for month of issue. Try our Online Subscription – now with PDF downloads! Privacy Phones; June 2024 ESR Test Tweezers; June 2024 Compact Frequency Divider; May 2024 The Pico Gamer; April 2024 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 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. 07/24 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 24LC32A-I/SN ATmega328P ATmega328P-AUR ATtiny45-20PU ATtiny85V-10PU PIC10LF322-I/OT PIC12F1572-I/SN PIC12F617-I/P PIC12F617-I/SN PIC12F675-I/P PIC16F1455-I/P Digital FX Unit (Apr21) 110dB RF Attenuator (Jul22), Basic RF Signal Generator (Jun23) RGB Stackable LED Christmas Star (Nov20) 2m VHF CW/FM Test Generator (Oct23) Shirt Pocket Audio Oscillator (Sep20) Range Extender IR-to-UHF (Jan22) LED Christmas Ornaments (Nov20; versions), Nano TV Pong (Aug21) Active Mains Soft Starter (Feb23), Model Railway Uncoupler (Jul23) Model Railway Carriage Lights (Nov21) Train Chuff Sound Generator (Oct22) Auto Train Controller (Oct22), GPS Disciplined Oscillator (May23) Railway Points Controller Transmitter / Receiver (2 versions; Feb24) PIC16F1455-I/SL Battery Multi Logger (Feb21), USB-C Serial Adaptor (Jun24) PIC16LF1455-I/P New GPS-Synchronised Analog Clock (Sep22) PIC16F1459-I/P Cooling Fan Controller (Feb22), Remote Mains Switch (RX, Jul22) K-Type Thermostat (Nov23), Secure Remote Switch (RX, Dec23) Mains Power-Up Sequencer (Feb24 | repurposed firmware Jul24) PIC16F1459-I/SO Multimeter Calibrator (Jul22), Buck/Boost Charger Adaptor (Oct22) PIC16F15214-I/SN Tiny LED Icicle (Nov22), Digital Volume Control Pot (SMD; Mar23) Silicon Chirp Cricket (Apr23) PIC16F15214-I/P Digital Volume Control Pot (through-hole; Mar23) PIC16F15224-I/SL Multi-Channel Volume Control (OLED Module; Dec23) PIC16F1705-I/P Digital Lighting Controller Translator (Dec21) PIC16F18146-I/SO Volume Control (Control Module, Dec23), Coin Cell Emulator (Dec23) PIC16LF15323-I/SL Remote Mains Switch (TX, Jul22), Secure Remote Switch (TX, Dec23) W27C020 Noughts & Crosses Computer (Jan23) ATSAML10E16A-AUT PIC16F18877-I/P PIC16F18877-I/PT High-Current Battery Balancer (Mar21) USB Cable Tester (Nov21) 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) PIC32MM0256GPM028-I/SS Super Digital Sound Effects (Aug18) 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) $20 MICROS ATmega32U4 ATmega644PA-AU Wii Nunchuk RGB Light Driver (Mar24) AM-FM DDS Signal Generator (May22) $25 MICROS dsPIC33FJ64MC802-E/SP 1.5kW Induction Motor Speed Controller (Aug13) PIC32MX470F512H-I/PT Stereo Echo/Reverb (Feb 14), Digital Effects Unit (Oct14) 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 AUTOMATIC LQ METER KIT (SC6939) (JUL 24) ESR TEST TWEEZERS COMPLETE KIT (SC6952) (JUN 24) DC SUPPLY PROTECTOR (JUN 24) Includes everything except the case & debugging interface (see p33, July24) - Rotary encoder with integral pushbutton (available separately, SC5601) Includes all parts and OLED, except the coin cell and optional header - 0.96in white OLED with SSD1306 controller (also sold separately, SC6936) All kits come with the PCB and all onboard components (see page 81, June24) - Adjustable SMD kit (SC6948) - Adjustable TH kit (SC6949) - Fixed TH kit – ZD3 & R1-R7 vary so are not included (SC6950) USB-C SERIAL ADAPTOR COMPLETE KIT (SC6652) (JUN 24) WIFI DDS FUNCTION GENERATOR (MAY 24) Includes the PCB, programmed micro and all other required parts Short-form kit: includes everything except the case, USB cable, power supply, labels and optional stand. The included Pico W is not programmed (SC6942) - Optional laser-cut acrylic stand pieces (SC6932) - 3.5in LCD touchscreen: also available separately (SC5062) 10MHz to 1MHz / 1Hz FREQUENCY DIVIDER (SC6881) (MAY 24) PICO GAMER KITS (APR 24) Complete kit: Includes the PCB and everything that mounts to it, including the 49.9Ω and 75Ω resistors (see page 38, May24) $100.00 $3.00 $50.00 $10.00 $17.50 $22.50 $20.00 $20.00 $95.00 $7.50 $35.00 $40.00 - SC6911: everything except the case & battery; RP2040+ is pre-programmed - SC6912: the SC6911 kit, plus the LEDO 6060 resin case - SC6913: the SC6911 kit, plus a dark grey/black resin case - 3.2in LCD touchscreen: also available separately (SC6910) ESP-32CAM BACKPACK KIT (SC6886) (APR 24) PICO DIGITAL VIDEO TERMINAL (SC6917) (MAR 24) Includes everything to build the BackPack, except the ESP32-CAM module - 3.5in LCD touchscreen: also available separately (SC5062) $85.00 $125.00 $140.00 $30.00 $42.50 $35.00 Short-form kit: includes everything except the case; choice of front panel PCB for Altronics H0190 or H0191. Picos are not programmed (see page 46, Mar24) $65.00 siliconchip.com.au/Shop/ MAINS POWER-UP SEQUENCER (FEB 24) MICROPHONE PREAMPLIFIER KIT (SC6784) (FEB 24) USB TO PS/2 KEYBOARD & MOUSE ADAPTOR (JAN 24) COIN CELL EMULATOR (SC6823) (DEC 23) MULTI-CHANNEL VOLUME CONTROL (DEC 23) SECURE REMOTE SWITCH (DEC 23) IDEAL DIODE BRIDGE RECTIFIER (DEC 23) Hard-to-get parts: includes the PCB, programmed micro, all other semiconductors and the Fresnel lens bezels (SC6871) $95.00 Current detection add-on: includes the AC-1010 current transformer, (P)4KE15CA TVS and MCP6272-E/P op amp (SC6902) $20.00 Includes the standard PCB (01110231) plus all onboard parts, as well as the switches and mounting hardware. All that’s needed is a case, XLR connectors, bezel LED and wiring (see page 35, Feb24) - VGA PicoMite Version Kit: see page 52, January 2024 (SC6861) - ps2x2pico Version Kit: see page 52, January 2024 (SC6864) - 6-pin mini-DIN to mini-DIN cable, ~1m long. Two cables are required if adapting both the keyboard and mouse (SC6869) - Receiver short-form kit: see page 43, December 2023 (SC6835) - Discrete transmitter complete kit: see page 43, December 2023 (SC6836) - Module transmitter short-form kit: see page 43, December 2023 (SC6837) - 28mm square spade: see page 35, December 2023 (SC6850) - 21mm square pin: see page 35, December 2023 (SC6851) - 5mm pitch SIL: see page 35, December 2023 (SC6852) - Mini SOT-23: see page 35, December 2023 (SC6853) - D2PAK SMD: see page 35, December 2023 (SC6854) - TO-220 through-hole: see page 35, December 2023 (SC6855) *Prices valid for month of magazine issue only. All prices in Australian dollars and include GST where applicable. # Overseas? Place an order on our website for a quote. $30.00 $32.50 $10.00 - Kit: Contains all parts and the optional 5-pin header (see page 77, Dec23) - 1.3in blue OLED (SC5026) - Control Module kit: see page 68, December 2023 (SC6793) - Volume Module kit: see page 69, December 2023 (SC6794) - OLED Module kit: see page 69, December 2023 (SC6795) - 0.96in SSD1306 cyan OLED (SC6176) $70.00 $30.00 $15.00 $50.00 $55.00 $25.00 $10.00 $35.00 $20.00 $15.00 $30.00 $30.00 $30.00 $25.00 $35.00 $45.00 PRINTED CIRCUIT BOARDS & CASE PIECES PRINTED CIRCUIT BOARD TO SUIT PROJECT REMOTE CONTROL RANGE EXTENDER UHF-TO-IR ↳ IR-TO-UHF 6-CHANNEL LOUDSPEAKER PROTECTOR ↳ 4-CHANNEL FAN CONTROLLER & LOUDSPEAKER PROTECTOR SOLID STATE TESLA COIL (SET OF 2 PCBs) DUAL HYBRID POWER SUPPLY SET (2 REGULATORS) ↳ REGULATOR ↳ FRONT PANEL ↳ CPU ↳ LCD ADAPTOR ↳ ACRYLIC LCD BEZEL RASPBERRY PI PICO BACKPACK AMPLIFIER CLIPPING DETECTOR CAPACITOR DISCHARGE WELDER POWER SUPPLY ↳ CONTROL PCB ↳ ENERGY STORAGE MODULE (ESM) PCB 500W AMPLIFIER MODEL RAILWAY SEMAPHORE CONTROL PCB ↳ SIGNAL FLAG (RED) AM-FM DDS SIGNAL GENERATOR SLOT MACHINE HIGH-POWER BUCK-BOOST LED DRIVER ARDUINO PROGRAMMABLE LOAD SPECTRAL SOUND MIDI SYNTHESISER REV. UNIVERSAL BATTERY CHARGE CONTROLLER VGA PICOMITE SECURE REMOTE MAINS SWITCH RECEIVER ↳ TRANSMITTER (1.0MM THICKNESS) MULTIMETER CALIBRATOR 110dB RF ATTENUATOR WIDE-RANGE OHMMETER WiFi PROGRAMMABLE DC LOAD MAIN PCB ↳ DAUGHTER BOARD ↳ CONTROL BOARD MINI LED DRIVER NEW GPS-SYNCHRONISED ANALOG CLOCK BUCK/BOOST CHARGER ADAPTOR AUTO TRAIN CONTROLLER ↳ TRAIN CHUFF SOUND GENERATOR PIC16F18xxx BREAKOUT BOARD (DIP-VERSION) ↳ SOIC-VERSION AVR64DD32 BREAKOUT BOARD LC METER MK3 ↳ ADAPTOR BOARD DC TRANSIENT SUPPLY FILTER TINY LED ICICLE (WHITE) DUAL-CHANNEL BREADBOARD PSU ↳ DISPLAY BOARD DIGITAL BOOST REGULATOR ACTIVE MONITOR SPEAKERS POWER SUPPLY PICO W BACKPACK Q METER MAIN PCB ↳ FRONT PANEL (BLACK) NOUGHTS & CROSSES COMPUTER GAME BOARD ↳ COMPUTE BOARD ACTIVE MAINS SOFT STARTER ADVANCED SMD TEST TWEEZERS SET DIGITAL VOLUME CONTROL POT (SMD VERSION) ↳ THROUGH-HOLE VERSION MODEL RAILWAY TURNTABLE CONTROL PCB ↳ CONTACT PCB (GOLD-PLATED) WIDEBAND FUEL MIXTURE DISPLAY (BLUE) TEST BENCH SWISS ARMY KNIFE (BLUE) SILICON CHIRP CRICKET GPS DISCIPLINED OSCILLATOR SONGBIRD (RED, GREEN, PURPLE or YELLOW) DUAL RF AMPLIFIER (GREEN or BLUE) LOUDSPEAKER TESTING JIG BASIC RF SIGNAL GENERATOR (AD9834) ↳ FRONT PANEL V6295 VIBRATOR REPLACEMENT PCB SET DYNAMIC RFID / NFC TAG (SMALL, PURPLE) DATE JAN22 JAN22 JAN22 JAN22 FEB22 FEB22 FEB22 FEB22 FEB22 FEB22 FEB22 FEB22 MAR22 MAR22 MAR22 MAR22 MAR22 APR22 APR22 APR22 MAY22 MAY22 JUN22 JUN22 JUN22 JUN22 JUL22 JUL22 JUL22 JUL22 JUL22 AUG22 SEP22 SEP22 SEP22 SEP22 SEP22 OCT22 OCT22 OCT22 OCT22 OCT22 OCT22 NOV22 NOV22 NOV22 NOV22 DEC22 DEC22 DEC22 DEC22 JAN23 JAN23 JAN23 JAN23 JAN23 FEB23 FEB23 MAR23 MAR23 MAR23 MAR23 APR23 APR23 APR23 MAY23 MAY23 MAY23 JUN23 JUN23 JUN23 JUN23 JUL23 PCB CODE 15109211 15109212 01101221 01101222 01102221 SC6244 SC6204 18107211 18107212 01106193 01106196 SC6309 07101221 01112211 29103221 29103222 29103223 01107021 09103221 09103222 CSE211002 08105221 16103221 04105221 01106221 04107192 07107221 10109211 10109212 04107221 CSE211003 04109221 04108221 04108222 18104212 16106221 19109221 14108221 09109221 09109222 24110222 24110225 24110223 CSE220503C CSE200603 08108221 16111192 04112221 04112222 24110224 01112221 07101221 CSE220701 CSE220704 08111221 08111222 10110221 SC6658 01101231 01101232 09103231 09103232 05104231 04110221 08101231 04103231 08103231 CSE220602A 04106231 CSE221001 CSE220902B 18105231/2 06101231 Price $2.50 $2.50 $7.50 $5.00 $5.00 $7.50 $25.00 $7.50 $2.50 $5.00 $2.50 $5.00 $5.00 $2.50 $5.00 $5.00 $5.00 $25.00 $2.50 $2.50 $7.50 $5.00 $5.00 $5.00 $7.50 $7.50 $5.00 $7.50 $2.50 $5.00 $5.00 $7.50 $7.50 $5.00 $10.00 $2.50 $5.00 $5.00 $2.50 $2.50 $2.50 $2.50 $2.50 $7.50 $2.50 $5.00 $2.50 $5.00 $5.00 $5.00 $10.00 $5.00 $5.00 $5.00 $12.50 $12.50 $10.00 $10.00 $2.50 $5.00 $5.00 $10.00 $10.00 $10.00 $5.00 $5.00 $4.00 $2.50 $12.50 $5.00 $5.00 $5.00 $1.50 For a complete list, go to siliconchip.com.au/Shop/8 PRINTED CIRCUIT BOARD TO SUIT PROJECT ↳ 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) PICO AUDIO ANALYSER (BLACK) 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) WII NUNCHUK RGB LIGHT DRIVER (BLACK) ARDUINO FOR ARDUINIANS (PACK OF SIX PCBS) ↳ PROJECT 27 PCB 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) DATE 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 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 PCB CODE 06101232 CSE230101C CSE230102 04105231 09105231 18106231 04106181 04106182 15110231 01108231 01108232 01109231 24105231 04105223 04105222 04107222 06107231 24108231 24108232 24108233 24108234 04108231/2 04107231 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 16103241 SC6903 SC6904 08101241 08104241 07102241 04104241 04112231 10104241 SC6963 08106241 08106242 08106243 24106241 AUTOMATIC LQ METER AUTOMATIC LQ METER FRONT PANEL (BLACK) 180-230V DC MOTOR SPEED CONTROLLER 2-WAY PASSIVE CROSSOVER (BLUE, 2MM THICK) * MAINS POWER-UP SEQUENCER * * pre-existing PCBs reused for projects in this issue JUL24 JUL24 JUL24 SEP15 FEB24 CSE240203A $5.00 CSE240204A $5.00 11104241 $15.00 01205141 $20.00 10108231 $15.00 NEW PCBs Price $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 $5.00 $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 $20.00 $7.50 $15.00 $10.00 $5.00 $10.00 $2.50 $5.00 $10.00 $2.50 $2.50 $2.50 $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 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 Multi-GHz RF amplifier design wanted I totally enjoyed your Formula One power unit write up in the May 2024 issue (siliconchip.au/Article/16249). Fancy that, hybrids! Can you advise of any Silicon Chip or earlier kits or projects involving RF amplifiers in the tens of gigahertz frequency bands? I am trying to repair a 15GHz RF amplifier and being able to look at similar designs might help me. I am convinced electronic enthusiast magazines like Silicon Chip are the best source for inspiration and resource for learning electronics. (M. G., Hornsby, NSW) ● We don’t think we have published any RF amplifiers that can handle 1GHz or above. Some radio operators transmit in the GHz, 10s of GHz and even 100s of GHz ranges but the transmission ranges are relatively short, so longer wavelengths tend to be more popular (this was described in our article on Amateur Radio in the April 2024 issue, siliconchip.au/Article/16206). We have published a few articles on multi-GHz signal generators, frequency counters and dividers but those probably won’t help you. We note that RF amplifier modules are available up to about 8GHz, but they get pretty expensive above that. We can also find some Mosfets suitable for use in RF amplifiers up to about 16GHz at reasonable prices. The price seems to rise exponentially above that, while the availability falls. Thanks for the feedback; we try to provide good resources for learning about electronics although we know that some of our projects could be a bit overwhelming for beginners. Our new Mini Projects series should provide something for them to start with. Powering the SMD LED Audio Level Meter I could not resist building the Touchscreen Digital Preamp (September & October 2021; siliconchip. 100 Silicon Chip au/Series/370) and the 100dB Stereo Audio Level Meter (June & July 2016; siliconchip.au/Series/301) into one case. It’s going to be so cool! I’m using a Jaycar MT-2086 15-015V centre-tapped mains transformer to power the preamp. I assume it is not wise to tap into the standard 12V power supply section of the Preamp to power the Level Meter. Should I connect another bridge rectifier to one of the 15V windings, with a filter capacitor and 12V regulator to supply the Level Meter? Would that work? Can I connect the 0V line of the Level Meter to the 0V line of the Preamp PCB? I’m thinking about Earth reference level shifting etc. I could replace REG1 (78L12) on the Preamp PCB with a 7812, possibly mounted off-board on a heatsink. Both the Preamp & Level Meter would then share the +12V rail. Would that induce noise? Should I add a dummy load resistor? I’m torn as to what I should do. Thanks for your help! (N. G., Blue Haven, NSW) ● To avoid switching noise entering the preamplifier, it is best to use a separate isolated 12V DC supply for the LED level meter. This supply can simply be a 12V DC plugpack or internal switch-mode power supply with a rating of at least 150mA. The Level Meter ground wire should connect to the preamplifier output signal ground. Safety of circuits that run at mains potential I finished building the Refined Motor Speed Controller (April 2021; siliconchip.au/Article/14814) but it does not work. This is not surprising as I found that the 5V bus is shorted to PCB ground. I’ll find out why and fix it when I disassemble the device. However, fault-finding caused me to look more closely at the circuit. I noticed that the design has the entire PCB floating at mains voltage. The speed and gain potentiometer terminals are uninsulated and only a few Australia's electronics magazine millimetres from the Earthed lid. Are these pots rated for mains voltage? I always use an opto-coupler to isolate a PCB circuit from mains voltage in my projects. For example, a H11AA1 or similar opto-coupler could provide the zero-crossing signal here. Of course, an isolated 5V supply would be needed and would require more space. Also, I do not see how a zero-­ crossing signal is created and applied to pin 5 of the PIC. Please advise how this is done so that I can check it with my DSO. I am keen to get this project working as it will make my electric drill more flexible and more useful. Thank you for your time. (G. C., Montville, Qld) ● Yes, the project has all the circuitry operating at mains potential, and that is mentioned in the article. For the majority of the circuit to run at low voltage and be isolated, there would have to be a lot more circuitry, significantly increasing the size, complexity and price to build it. Even then, the Triac, snubber and mains wiring would still be at mains potential. We have specified 24mm potentiometers, which we have used in our mains-powered designs published since the 1970s without problems. The clearance between the potentiometer terminals and the enclosure lid is sufficient to prevent arc-over to the Earthed lid. The entire enclosure is Earthed and the circuit is only unsafe if powered up with the enclosure open. That’s true of just about any other mains-operated circuit, especially ones that control the mains waveform, as this one does. Our mains control projects do tend to have circuitry that is at mains potential to keep them simple and robust, so we warn constructors o be aware of the dangers. We recommend to ensure power is disconnected before opening the enclosure and to make sure that the enclosure is closed before connecting power. As for the zero-voltage crossing detection, since the circuit is floating siliconchip.com.au at mains Active potential, the Neutral wire provides the voltage differential for the zero voltage crossing detection at the GP2 input. Using a headphone amp to buffer radio output For many years, I have been using an old car radio (with a defunct output stage) as an audio signal source on my workbench. I feed its line output to a benchtop speaker through L + R combining and isolating resistors, then to a transformer feeding a single speaker to provide low-level continuous music in my workshop (I have the radio running 24/7). At the same time, I have the line output as a ‘wander lead’ with RCA plugs on my bench, which I use to connect to various amplifiers etc that I happen to be working on at the time. Unfortunately, the car radio has some failings, mainly that the audio output (which isn’t at line level anyway) comes from after the volume control and equaliser, which isn’t ideal. I decided I needed something better than my old setup, so I recently bought a cheap FM tuner I plan to use as my radio signal source. I am looking for a buffer amplifier design that I can use to connect an FM tuner to the outside world of my bench and my bench speaker. I thought I would use two headphone amplifiers: one to drive my bench speaker and the other one to connect to the wander lead. The idea of using headphone amplifiers came to mind because a low signal source output impedance would be ideal. I plan to add separate tone control preamps ahead of the headphone amps to provide independent tone control to either of the headphone amps. I have searched your index for suitable amplifiers/tone control amplifiers but came up with thousands of hits. Do you know of any ideal designs your magazine has described that I could make up to give me the above options? It seems likely that I would need to combine designs. (P. W., Auckland, NZ) ● You could combine our Ultra Low Noise Remote Controlled Stereo Preamp (March & April 2019; siliconchip. au/Series/333) with the Studio Series Headphone Amplifier (November 2005; siliconchip.au/Article/3231). They are both stereo units, so you could use the left and right channels separately for the FM receiver and wander lead. The stereo preamplifier tone control potentiometers will need to be wired to single-gang pots if you want separate control of each channel. The remote-controlled volume section won’t be required, so you can use a regular pot instead of a motorised one. Both circuits are powered from ±15V supplies that could be produced using our Universal 4-Output Voltage Regulator (May 2015; siliconchip.au/ Article/8562). The headphone amplifiers utilise BD139 and BD140 transistors in a class-AB arrangement. PCBs for all these projects are available in our Online Shop. Switching an Ethernet connection on and off I have a teenager that keeps on “forgetting” the time he is supposed to be off the ‘net. I tried using the parental controls in the router’s firmware to cut his access outside of set times. This has proved very difficult to do, because these settings don’t work the way they The Pico Gamer A PicoMite powered ‘retro’ game console packed with nine games including three inspired by Pac-Man, Space Invaders and Tetris. With its inbuilt rechargeable battery and colour 3.2-inch LCD screen, it will keep you entertained for many hours. SC6912 | $125 + post | complete kit with white resin case* Other Items for this project SC6911 | $85 + post | complete kit without any case* SC6913 | $140 + post | complete kit with a dark grey resin case* * LiPo battery is not included SC6909 | $10 + post | Pico Gamer PCB* See the article in the April 2024 issue for more details: siliconchip.au/Article/16207 siliconchip.com.au Australia's electronics magazine July 2024  101 should and some routers are just plain confusing (I have tried several). If I build a timer using an Arduino Uno board, can I switch the network data lines (through the actual network cable) via transistors or should I use relays to interrupt the data lines? I am unsure about using transistors to cut the data lines, as I don’t know if transistors would affect the transmitted data by introducing signal degradation or interference or if they will withstand being on for 16 hours a day. Would the relay affect the transmitted data in a similar way? Lastly, would one of your timer projects be able to control such a circuit without introducing interference to the network data lines? (D. S., Maryborough, Qld) ● We think you would be better off using ‘Parental Control’ software rather than trying to interfere with the Ethernet cable signalling. Several suitable pieces of software are reviewed at siliconchip.au/link/abw9 If you must do it with hardware, relay contacts would disconnect data lines more effectively than transistor switching. You would need to use small ‘telecom’ relays and keep all the wiring short or it would almost certainly introduce signal degradation. Using a slower Ethernet speed (eg, 100Mbit rather than Gigabit) would help preserve the signal integrity. Any timer that can switch relays on and off could be used, or a remote controlled relay. There are commercial internet kill switches available. Some use a toggle switch. Which speed controller to use with belt sander I have a small bench belt & disc sanding unit that I want to add a speed Designing loading coils for antenna impedance matching I wonder if your expert staff or any Silicon Chip reader can assist me. I have several 27MHz CB radios lying around, which I would like to press into service for communications between the house and my work shed. As the distance is short, I would like to use a simple one-metre whip antenna with a matching circuit. The idea is to use a plain VHF-type whip aerial, as they are not much affected by the harsh weather conditions here. How do I go about winding or obtaining an inductor that I can use to match such antennas to the CB radios? I’m sure there are other readers wondering what to do with these sets, as they are currently not working well due to high sunspot activity. (W. S., Broken Hill, NSW) ● The loading coil for a whip antenna can be calculated by first finding the required inductance using the following online calculator: siliconchip.au/link/abu9 You can then design the inductor using the following calculator and wind it with enamelled copper wire on a PVC plastic conduit: siliconchip.au/link/abua control to. I need to slow it down when sanding some types of timber so it doesn’t burn the timber (I suspect its speeds are geared toward metal finishing). The motor details are: • Power: 375W • Voltage: 240V AC • Frequency: 50Hz • Current: 1.7A • Poles: 4 • Phases: 1 • RPM: 1420 • Class-B capacitor start What speed controller project do you recommend I build? (R. M., Curlewis, Vic) ● That is a capacitor-start type induction motor. The best speed control would be with a variable-­ frequency drive (VFD) rather than just voltage control. We previously published an Induction Motor Speed Controller (IMSC) in the April & May 2012 issues (siliconchip.au/Series/25), although it might be considered overkill for this application. Also, capacitor-­start motors need their wiring modified for use with it. You can obtain limited speed control using a phase controller that reduces the average voltage applied to the motor. We published a suitable phase controller, the Refined FullWave Motor Speed Controller (April 2021; siliconchip.au/Article/14814). The PCB for that project and a set of hard-to-get parts are available from our Online Shop (siliconchip.au/ Shop/?article=14814). We are considering a new IMSC design that might be cheaper to build than the 2012 version (possibly with other advantages), but we can’t say at this stage whether or when it will be published. If you are making the original IMSC, note that there were significant updates to the design in the December 2012 (siliconchip.au/Article/469) and August 2013 issues (siliconchip.au/ Article/4219). Some of the parts are becoming hard to obtain although, as far as we know, they are all still available. SC200 amplifier assembly questions I am currently assembling your SC200 Audio Amplifier modules 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. 102 Silicon Chip Australia's electronics magazine 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 FOR SALE LEDs and accessories for the DIY enthusiast LEDs, BRAND NAME AND GENERIC LEDs. Heatsinks, LED drivers, power supplies, LED ribbon, kits, components, hardware. For a full list of the parts we sell, 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 For Quality That Counts... QUALITY COMPONENTS AT GREAT PRICES. Check out the latest deals this month. SMD parts and more. Go to www.lazer.com.au 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 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. (January-March 2017; siliconchip.au/ Series/308) using Altronics K5157 kits. I have a few questions that would help to clarify assembly in my head. 1. The heatsink drilling details show the central KSC2690A/TTC004B transistor 5mm lower than the two other transistors on either side. That appears different from the photos of the finished module, where all three middle transistors are mounted at the same height (30mm from the bottom of the heatsink). Should they all be level at 30mm from the bottom of the heatsink? 2. I am going to mount the clipping LEDs on the front panel. Can I use any 5mm red LEDs for this? I have some bezels that fit 5mm LEDs handy. Or is there a particular LED I should use? 3. The transformers I am going to use are Ferguson PF4361/1 200VA siliconchip.com.au 35-0-35/15-0-15 transformers from a very old ETI Series 5000 kit. I have built two power supplies and will use one transformer for each SC200 Amplifier board and power supply board. Based on your previous emails, I think these should be okay to use. My speakers are 4W Magneplanars. 4. Under “Final Assembly”, regarding mounting the transistors to the heatsink, it says, “Once the soldering is completed, trim the leads and remove the two supports near the heatsink, as these are no longer required; the transistors should be mounted to the chassis via the heatsink only, otherwise, thermal cycling could crack their solder joints.” I take it that means to remove the two plastic PCB standoffs near TP1 & TP2 but retain the other two standoffs at the front/input edge of the amp Australia's electronics magazine board. Will they still allow thermal cycling to occur safely? Thanks for all your support so far with my project. (S. W., via email) ● Our answers to your questions: 1. The heatsink drilling has been purposefully kept compatible with the Ultra-LD Mk.2, Mk3 & Mk.4 amplifiers, which used larger transistors except the middle one, even though the extra height is not needed for those two transistors on the SC200. That was to allow existing heatsinks to be reused. You could certainly drill them all at the same (lower) height, as we did on our prototype. 2. Most LEDs should be suitable, but we suggest you test how bright they are at 1mA by powering them from a DC supply with a fixed resistor, eg, a 3kW series resistor for a 5V supply or a 10kW series resistor for a 12V supply. July 2024  103 If they light up reasonably well at that current, it should be fine to use them without modifying the circuit. If they are too dim, you can change two resistors on the SC200 modules to deliver more LED current. One is 33kW and connects to Q17’s emitter and D6’s anode, while the other is 100kW and connects between Q19’s collector and D5’s anode. Halving their values will double the current delivered to the LED. If you reduce the 100kW resistor below 47kW, it should be changed to a half-watt type. 3. Check the supply outputs; we expect they will be close to ±50V unloaded. That should actually provide some benefits for driving the 4W speakers with the SC200 modules (mainly increased efficiency). That reduced supply voltage will not reduce the continuous power at all and should barely affect ‘music power’. The only effect the lower supply voltage will have on circuit operation is less current through Q6 (about 87% of the design current), but we don’t think that will be a problem. You could change the two 6.8kW resistors at the Advertising Index Altronics.................................35-38 Blackmagic Design....................... 7 Dave Thompson........................ 103 DigiKey Electronics....................... 3 Emona Instruments.................. IBC Hare & Forbes............................. 11 Jaycar..............IFC, 9, 63, 66-67, 93 Keith Rippon Kit Assembly....... 103 Lazer Security........................... 103 LD Electronics........................... 103 collector of Q6 to 6.2kW or 5.6kW each to compensate for that, but that should not be necessary. 4. That’s right; you don’t want to rigidly attach the end with the transistors as it will move up and down slightly with thermal cycling. There’s enough board/lead flex that it’s OK to attach the other end to the chassis, though. and the Isolating High Voltage Probe for Oscilloscopes (January 2015; siliconchip.au/Article/8244). There are also plenty of good commercial offerings, including the relatively low-cost Pintek DP-25. Making & using a mains isolation transformer I want to build another two Ultra-LD Mk3 amplifiers. I have the PCBs, but I am having trouble sourcing some of the transistors that are no longer being manufactured. What can I use instead of the 2SA970, 2SA1837 and 2SC4793? I think I can still get the other transistors. I really appreciate your advice. (D. S., Caringbah, NSW) ● This is very frustrating because the 2SA1837 and 2SC4793 were the transistors we chose to replace the obsolete BF469 and BF470. They have also been discontinued in just a few years, which shows how difficult it is to create a design that can be built years after publication. Also, there isn’t any direct substitute for the 2SA970; all similar lownoise transistors that are available only come in SMD packages (and many of those have also been discontinued lately). It will take us some time to find alternative transistors that are also current and available (if they exist). In the meantime, you can still get the 2SA970, 2SA1837 and 2SC4793 parts. The 2SA970 and 2SA1837 transistors are available from Futurlec (www. futurlec.com). All three transistors are also available from several suppliers on eBay and AliExpress (AliExpress 32882321479 & 32491223824). We haven’t tested those, but the reviews are mostly positive. SC I want to build a 60VA isolation transformer for my digital oscilloscope. What VA rating transformers would I need? (R. M., Melville, WA) ● It is not safe to isolate the supply to your oscilloscope so that it is floating and not Earthed. The equipment you intend to monitor with your oscilloscope should be isolated instead. The easiest way to make a mains isolation transformer is to use two identical mains transformers and connect the secondaries together. The mains supply is applied to one of the transformers, and the output transformer primary (230V AC side) supplies the DUT. For more information, see the YouTube video at https://youtu.be/ WKFFEsIh9Gw The VA rating for the transformers must be sufficient to supply the equipment you wish to measure using the oscilloscope. Alternatively, you can use an isolated or differential oscilloscope probe and keep everything Earthed. The probe’s frequency response needs to suit what you intend to measure using the oscilloscope. We published two projects that might be suitable: the Wideband Active Differential Oscilloscope Probe (September 2014; siliconchip.au/Article/7995) Microchip Technology.............OBC Mouser Electronics....................... 4 PCBWay....................................... 13 PMD Way................................... 103 Silicon Chip Back Issues........... 94 Silicon Chip PDFs on USB......... 71 Silicon Chip Pico Gamer......... 101 Silicon Chip Shop.................98-99 Silicon Chip Subscriptions........ 97 The Loudspeaker Kit.com.......... 10 Wagner Electronics..................... 91 104 Silicon Chip Errata and Sale Date for the Next Issue LEDsales................................... 103 Obtaining transistors for Ultra-LD Mk.3 amp DC Supply Protectors, June 2024: the two through-hole versions specify a maximum current of 7A but the SPP15P10PL-H P-channel Mosfets specified can only handle about 2.5A without heatsinking. A logic-level P-channel Mosfet with a lower on-resistance like the IPP80P03P4L-07 (SC6043) can be used instead without heatsinking (now included in both kits). Fan Speed Controller Mk2, May 2024: in the left-hand column of text on p73, the reference to diode D1 in the fifth paragraph should be to D2 instead. Touchscreen Appliance Energy Meter, August-October 2016: the circuit diagram on pages 30 & 31 of the August 2016 issue had previously been updated in the online version to use the safer ACS718 IC rather than the original ACS712. There were some errors in this updated circuit, mainly with the pin numbers for the Vcc (pin 10) and GND (pins 11, 13, 14 & 15) connections. They have now been fixed in the version on our website. Next Issue: the August 2024 issue is due on sale in newsagents by Monday, July 29th. Expect postal delivery of subscription copies in Australia between July 26th and August 14th. Australia's electronics magazine siliconchip.com.au “Rigol Offer Australia’s Best Value Test Instruments” New 2024 Products Oscilloscopes New 12Bit Scopes RIGOL DS-1000Z/E - FREE OPTIONS RIGOL DHO Series RIGOL MSO-5000 Series 450MHz to 200MHz, 2/4 Ch 41GS/s Real Time Sampling 424Mpts Standard Memory Depth 470MHz to 800MHz, 2/4 Ch 412Bit Vertical Resolution 4Ultra Low Noise Floor 470MHz to 350MHz, 2 Ch & 4Ch 48GS/s Real Time Sampling 4Up to 200Mpts Memory Depth FROM $ 499 FROM $ ex GST 659 FROM $ ex GST 1,489 Multimeters Function/Arbitrary Function Generators New Product New Product RIGOL DG-800/900 Pro Series RIGOL DG-1000Z Series RIGOL DM-858/E 425MHz to 200MHz, 1/2 Ch 416Bit, Up to 1.25GS/s 47” Colour Touch Screen 425MHz, 30MHz & 60MHz 42 Output Channels 4160 In-Built Waveforms 45 1/2 Digits 47” Colour Touch Screen 4USB & LAN FROM $ 713 FROM $ ex GST Power Supplies ex GST 725 FROM $ ex GST Spectrum Analysers 689 ex GST Real-Time Analysers New Product RIGOL DP-932E RIGOL DSA Series RIGOL RSA Series 4Triple Output 2 x 32V/3A & 6V/3A 43 Electrically Isolated Channels 4Internal Series/Parallel Operation 4500MHz to 7.5GHz 4RBW settable down to 10 Hz 4Optional Tracking Generator 41.5GHz to 6.5GHz 4Modes: Real Time, Swept, VSA & EMI 4Optional Tracking Generator ONLY $ 849 FROM $ ex GST 1,321 FROM $ ex GST 3,210 ex GST Buy on-line at www.emona.com.au/rigol Sydney Tel 02 9519 3933 Fax 02 9550 1378 Melbourne Tel 03 9889 0427 Fax 03 9889 0715 email testinst<at>emona.com.au Brisbane Tel 07 3392 7170 Fax 07 3848 9046 Adelaide Tel 08 8363 5733 Fax 08 83635799 Perth Tel 08 9361 4200 Fax 08 9361 4300 web www.emona.com.au EMONA