Fullscreen HTML5Med Res (??MB)HTML4

JANUARY 2023 ISSN 1030-2662 01 The VERY BEST DIY Projects! 9 771030 266001 $ 50* NZ $1290 11 INC GST INC GST 30   Q Meter for Inductors 40   How to build your own mini-ITX PC 50   Raspberry Pi Pico W BackPack 58   High-Performance Active Subwoofer 80   Noughts & Crosses Playing Machine COMPUTER MEMORY THE HISTORY OF EARLY DATA STORAGE Affordable Solid State Storage Media with fast read/write speeds. 30m m IDEAL FOR GAMING, VIDEO EDITING, ETC. ULTRA SLIM, ULTRA PORTABLE STORAGE 10 0m m TYPE-C USB M.2-2280 NVMe/PCIe SSD Provides faster sustained read/write speeds compared to SATA SSDs. • Read/write up to 2500/1950Mbps Portable SSD Hard Drives FROM 6995 $ Uses flash memory compared to traditional motrised HDD. • Transfer speeds up to 440Mbps 256GB XC5930 | 512GB XC5932 Shop at Jaycar for: • SD Cards & Flash Drives • USB Memory Card Readers • USB Type-C External M.2 SATA/NVME SSD Case • USB 3.0 HDD Docking Stations 119 $ 500GB XC5920 | 1TB XC5922 FAST AND RELIABLE REPLACEMENT FOR SLOW PERFORMING HDDS PERFECT FOR PHONES, TABLETS, ACTION CAMERAS, ETC. SDXC Class 10 microSD Memory Cards Ultra-fast read write speeds for maximum performance. 16GB to 512GB XC5015-XC5020 FROM HEAD IN-STORE OR ONLINE FOR LATEST PRICING 2.5” SATAIII SSD • Reads/write up to 540/490Mbps • Compact 100Lx70Wx7Dmm 256GB XC5686 512GB XC5688 FROM 6995 $ • SATA 2.5" HDD to USB 3.0 Case • Internal 1TB 2.5" Notebook HDD • Internal 2TB 3.5" Surveillance HDD • HUGE RANGE of Harddrive Enclosures, Leads & Adaptors Explore our wide range of storage media products, in stock on our website, or at over 110 stores or 130 resellers nationwide. Prices correct at time of publication but are subject to change. Jaycar reserves the right to change prices if and when required. jaycar.com.au/mediastorage 1800 022 888 Contents Vol.36, No.01 January 2023 12 Computer Memory, Part 1 Increasingly smaller, faster and larger-capacity memory has been one of the major drivers of technological advances in computers. This two-part series covers how ephemeral data is stored and what it is stored in. By Dr David Maddison Computer technology 27 2W RF Amplifier & Wattmeter Two watts might seem like too little for an amplifier, but you might be surprised how useful this tiny device is. It’s designed and rated for VHF applications between 1-930MHz. By Allan-Linton Smith RF product review 40 How to build a Mini-ITX PC Portable computers, like laptops, are invaluable if you need to work in multiple locations. But they lack the power of a desktop machine which is why Mini-ITX is the perfect form factor for a smaller DIY computer. By Nicholas Vinen Building computers 68 Magnetic Amplification We show you how a semi-regulated voltage can be controlled by using just two separate transformers, a potentiometer and two diodes. By Fred Lever Voltage regulation feature 30 Q Meter In conjuction with an RF signal generator, this compact meter measures the quality factor (Q factor) of inductors, up to values of about 200. It is invaluable for RF filter design and component selection. By Charles Kosina Test & measurement project 50 Raspberry Pi Pico W BackPack Our upgraded Pico W BackPack now features WiFi functionality along with its 3.5-inch touchscreen. We’ve also included sample code to show how you can use HTTP, UDP and NTP with the BackPack. By Tim Blythman WiFi microcontroller project 58 Active Subwoofer, Part 1 The Active Subwoofer is a ‘no-compromise’ design. While it is designed for use with the Active Monitor Speakers from Nov-Dec 2022, it can be used in many other applications such as a high-quality home theatre system. By Phil Prosser HiFi project 80 Noughts & Crosses Machine, Pt1 This Noughts & Crosses-playing (Tic-Tac-Toe) computer uses only logic gates to play Noughts & crosses with you! It has its own case with lighting underneath the pieces, and you can even play it with a friend. By Dr Hugo Holden Game project Cover image: close-up of stacked RAM modules Page 30 Q Meter Page 40 Build your own mini-ITX PC Page 58 12-inch Driver High-Performance ACTIVE SUBWOOFER 2 Editorial Viewpoint 5 Mailbag 67 Subscriptions 76 Circuit Notebook 90 Vintage Radio 95 Product Showcase 96 Serviceman’s Log 105 Ask Silicon Chip 108 Online Shop 111 Market Centre 112 Advertising Index 112 Notes & Errata 1. Noughts & Crosses with two modules 2. MIDI Toolbox UDISCO L6 circa 1927 by Dennis Jackson 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): $65 12 issues (1 year): $120 24 issues (2 years): $230 Online subscription (Worldwide) 6 issues (6 months): $50 12 issues (1 year): $95 24 issues (2 years): $185 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: 24-26 Lilian Fowler Pl, Marrickville 2204 2 Silicon Chip Editorial Viewpoint Using DRC correctly avoids errors We commonly find mistakes in PCBs and circuits while editing, laying out or proofreading articles. That is true of both designs we generate and those that are sent to us. The frustrating thing is that while we catch and fix most errors, some slip through because we can’t find them all. The more mistakes are in the original files, the more likely one or more will get through. This is especially frustrating when the errors are things that would have easily been picked up by the error-checking features of ECAD software if used correctly. This includes things like component value mismatches between the circuit and PCB, missing tracks on the PCB, tracks shorted together, having components on the PCB connected differently than in the circuit etc. I have extensively used both Altium and EAGLE software and know that both offer a similar set of ‘design rule check’ or DRC features (I think KiCad has them too). If used and used correctly, these will pick up most errors. Whether you are designing a PCB for publication in SILICON CHIP magazine, for a business venture or just for yourself, you should take advantage of these tools. The four main steps to use DRC properly are: 1 Generate or acquire a parts library with symbols and footprints for all the components you will use in your design. 2 Draw the circuit diagram in your ECAD package (‘schematic capture’). 3 Verify that the DRC rules have been set up to suit your PCB manufacturer. 4 Run DRC and check for zero errors before submitting a PCB to a manufacturer or for publication. It would also be good if any PCB designs being published were identical to the final prototype that has been verified to work, but I realise that can’t always be the case. Using the DRC steps above should allow for minor changes between the prototype and the final version without any significant errors creeping in. I think many people do not use DRC to its full capability because extra work is involved, especially in the first two steps. Despite that, I have always done so; I feel it is worth the extra effort. You don’t want to order hundreds of PCBs only to find that you have made a silly mistake and they are unusable! Another reason some people might skip it is because there are often exceptions to rules that cause ‘violations’ that are not actual problems, such as tracks/pads close together near fine-pitch SMDs. With Altium, you can create exceptions to rules, while with EAGLE, you can ignore violations after checking them. It’s worth doing those things so you are left with no errors when your design is finished. Besides eliminating the most common errors, using these DRC features also makes it easy to rejig a layout if, for example, you need to add a few components. You can rip out (delete) some of the tracks, move components around and add some new ones; then, the software will guide you to reinstate all the removed tracks. It won’t let you get it wrong (unless you ignore its warnings!). Common errors that DRC will catch include: 䕕 unrouted nets (missing tracks/connections) 䕕 short circuits between tracks that should not join 䕕 tracks that run too close to other tracks, vias or component pads 䕕 tracks that go nowhere (and might act as antennas) 䕕 tracks too close to the edge of the PCB 䕕 high-voltage tracks too close for safety requirements 䕕 tracks that are too thin 䕕 holes that are too close together So please use DRC, especially if you plan to submit a design to us for publication. Don’t skip that essential last step of actually checking it before sending your board design off! It also helps to thoroughly inspect all the Gerber layers before committing to a design. Note that our PO Box has changed (see sidebar). The old PO Box address is valid for now but will eventually be discontinued. by Nicholas Vinen Australia's electronics magazine siliconchip.com.au Achieve a higher degree of inventory management au.mouser.com/inventory-management +852 3756-4700 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”. Time to rein in the slashed zero The slashed zero character (0) has long been used to differentiate between the numeral 0 and the capital letter O. It entered the electronic era in low-resolution displays such as 5×7 dot matrix printers and CRT monitors. For many years, this was hardwired into display components so that even if the message context was purely numeric, the slashed zero was the only way to indicate zero. But times have moved on. Display resolution has increased enormously, and programmers now have a wide selection of unambiguous fonts, including the slashed zero, when required. The prime purpose of a visual machine-human interface is to make the message as clear to read as possible. Unfortunately, a meter displaying 240 is harder to read than 240. Likewise, large format LED matrix roadside screens showing a speed limit of 80km/h are harder to read and more distracting than a straightforward 80km/h. Of course, the slashed zero remains entirely appropriate for alphanumeric strings where there may be ambiguity, such as in coding and some less-than-user-friendly password formats. Nonetheless, for consumer-facing numeric displays, the slashed zero is obsolete technology and in serious need of an upgrade. As a plea to programmers everywhere, where you can choose fonts, please render the unambiguous numeric zero in its clearest form, simply 0. Slashed zeros are so 1970s. Mark Hallinan, Murwillumbah, NSW. Servicing electronics & knowledge management I would like to commend you on your ongoing Serviceman’s Log column. I find it one of the highlights of your magazine. Although I have formal electronics training, I learned many of my diagnostic skills (especially for the old CRT televisions) from reading The Serviceman’s Log over many years of Silicon Chip & EA. The recent article about camera IR lights blinding movement detectors helped me solve a problem that has been bugging me for years. I once attended a Knowledge Management course. The basic premise was that humans transfer knowledge by storytelling, and if someone tries to tell you that knowledge management is a database, run away. They told a story about Xerox’s service department. Despite many databases and diagnostic trees, their diagnosis rates did not improve much. They began holding servicing forums where servicemen told stories about difficult faults and how they fixed them. Their diagnosis siliconchip.com.au rates immediately improved and kept improving as more stories were told. So not only is your Serviceman’s Log column very entertaining, but it’s also teaching our technical people valuable skills! Mike Hammer, Mordialloc, Vic. GPS-synchronised Analog Clock works well I have only been a member of the Silicon Chip community for a few years but I enjoy the magazine content and projects. I ordered the GPS-Synchronised Analog Clock Driver kit (siliconchip.au/Shop/20/6472) and combined it with a wall clock mechanism ordered online. I made the plywood face at our local Mens’ Shed, with a few crafty figures from the local hardware store. I attached the electronic unit to the front for people to see how the clock operates. It all works well and was a great project. Peter Pade, Chester Hill, NSW. Comments on the magazine format I recently renewed my subscription again. I like the recent changes in the format of the magazine. The recent October issue is a good example. For example, the covers seem brighter and more attention-­grabbing, and it’s nice when an article referring to a previous article, or another article in the same magazine, includes page numbers. The Contents page is a bold standout for the month’s agenda with bold page numbers. It makes referring to something quick and easy. The inclusion of Altronics and Jaycar part numbers in the parts list is also a very handy feature. Are Jaycar and/ or Altronics interested in adding a stock number to the parts list which would cover those items in the parts list with their part numbers? That would make online ordering quick and easy for readers and might also generate additional sales for them. I am not suggesting that they make a physical kit; rather, a pick order for the warehouse/store for online orders. Roy Featherstone, Woodgate, Qld. Comments: thank you for the feedback. Some of the changes you’ve noticed are due to newer and younger staff members becoming more confident as they gain Australia's electronics magazine January 2023  5 experience. We strive to keep the magazine readable, with good presentation, but without putting style over substance. Your idea about the ‘pick lists’ for retailers is an interesting one that we will bring to their attention. Starting motors with a generator I’m writing in response to the letter “Extra generator capacity needed for motor starting” in the Mailbag section of the November 2022 issue (pages 10 & 12) and your comment on it. Storm-related outages of up to four days meant that I needed to find a way to get my heat-pump hot water unit to run from our small generator. The trick is to get the generator up to its rated power with heaters first. Switch off the heaters at the same instant the heat pump is switched on. Otherwise, it takes some time for the governor to get the engine up to full power. When a motor is started, that is just too long; the generator speed drops below its power band and it can’t recover. With this method, the continuous power rating of the generator can be as little as double the rating of the motor. For small generators, ignore the rating on the box; that is the momentary peak! P.S. I used up one of your recently gifted “!”s. Philip Petschel, Kinglake, Vic. Torches article enjoyed I greatly enjoyed Dr Maddison’s article about torches in the November 2022 issue of Silicon Chip (siliconchip. au/Article/15538). It brought back some memories. In the late 1950s and early ‘60s, I used to visit my aunt, who lived about half a kilometre away in northern Finland. The trail to her place ran through a deep, dense forest of mostly fir trees. It got really dark in the late autumn evenings and nights; the sun set before 2pm. If it hadn’t snowed yet, it was overcast, or there was no moon... it was like walking in 6 Silicon Chip a cave! This is where my pocket light came in handy – absolutely essential, in fact. I still have it and I have sent some photos (shown at the bottom of this page). Years ago, every household in Finland would have had at least one of these. The torch takes a 3LR12 (or 3R12) 4.5V battery. It is made of three B cells (3×R12) in series (yes, I had never heard of B cells either). Dr Maddison’s article on batteries in the January 2022 issue (“All About Batteries”, siliconchip.au/Series/375) shows a Finnish 3LR12 battery with the RAINBOW trademark (page 17). Many other batteries also have the same logo. Surprisingly, I can’t find them via a Google search. RS Components appears to be the only local supplier of the 3LR12 batteries. I’m thinking of ordering a cheaper one from Europe for old times’ sake. If you Google “auri taskulamppu”, you will find a lot of historical information about Auri taskulamppu (“pocket lamp”) and the company GWS that manufactured them in huge quantities. Mauri Lampi, Glenroy, Vic. Comments on Torches article I’ve just got the latest issue of Silicon Chip and have been reading your article on torches. Unsurprisingly, this is extensive and very interesting! I made my first LED torch from a Dick Smith Electronics kit around 2000. A mate and I each bought one – it was a penlight-style torch that used a single AA cell, with a nice compact circuit that tripled the voltage. From memory, the white high-brightness (for its time!) LED had a forward voltage of 3.5-4V or so. I wonder now whether this project was published in Silicon Chip or Electronics Australia. Editor’s Note: It was most likely based on John Clarke’s design published in the December 2000 issue of Silicon Chip (“Build A Bright-White LED Torch”; siliconchip.au/ Article/4265). Also, it was interesting to see your mention of the PakLite. Ever since acquiring quite a few smoke-detector 9V batteries, I’ve become keen on finding devices that can use them. In addition to a 9V clip-on phone charger, I’ve found a couple of LED torches that use this method. Jaycar and Altronics sell types with six LEDs: Jaycar Cat ST3367 and Altronics Cat X0218. Another really nice (and cheap) one I’ve found is from Bunnings: siliconchip.au/link/abi6 The latter uses a COB LED setup and is switchable (I would describe the two settings as “bright” and “slightly less bright”!). I still haven’t run down the original supplied battery after several hours of use (mainly in the lower power mode), so I don’t know what the expected lifetime is; I haven’t measured the current draw. Finally, I haven’t seen any mention of COB (chip-onboard) technology in your article. It has become increasingly popular, but I’ve never really understood this LED technology, so maybe one day you (or your colleagues) could address this topic. Anyway, thanks for yet another detailed and interesting article, and all the best! Chris Naylor, Melbourne, Vic. Comment: there is good information on COB LEDs on the Digi-Key website here: siliconchip.au/link/abi7 Australia's electronics magazine siliconchip.com.au Strange Raspberry Pi Pico behaviour I’ve been having problems with the Raspberry Pi Pico Backpack (March 2022; siliconchip.au/Article/15236) I purchased not long ago, so as a man on a mission, I set out to find the problem. The symptoms were no control over the backlight when using the Arduino example BackPack code. An oscilloscope probe on the GPIO20 pin indicated an attempt to switch but with insufficient strength to drive the Mosfets. I thought I might have swapped the two Mosefts that drive the backlight LEDs, but swapping them didn’t fix it. My next idea was pulling off the Pico, which was hard as I had soldered it in. I managed to get it off using a solder sucker and hot air wand, but even with a new Pico, I had the same problem. So it was obviously something to do with the code. After much looking at the “PWM.h” file, it seemed to me that there might be a few steps in the setup code missing. So I changed the setup code in LCD.C to the following: void displaySetup() { sliceBL = pwm_gpio_to_slice_num(BLPIN); chanBL = pwm_gpio_to_channel(BLPIN); // these 2 lines of code should be added pwm_config conf = pwm_get_default_config(); pwm_init(sliceBL, &conf, false); gpio_set_function(BLPIN, GPIO_FUNC_PWM); gpio_set_pulls(BLPIN, false, true); // = >maps to % pwm_set_wrap(sliceBL, 100); // 25kHz pwm_set_clkdiv_int_frac(sliceBL, 50, 0); // 0% duty pwm_set_chan_level(sliceBL, chanBL, 0); //PWM running pwm_set_enabled(sliceBL, true); . . . It works now that the pwm_init() function is being called. I also tried the Pico W on the BackPack; it works fine with the Arduino IDE 2. Dennis Smith, Devonport, Tas. Comments: we tested the Arduino sketch downloaded directly from our website and could not reproduce your problem; the backlight slider worked as expected. We also tried the (precompiled for Arduino) UF2 file provided with the download, which also worked. We are using version 2.3.2 of the Arduino-Pico board profile; we wonder if you have a different version. It is also worth checking the performance of the audio outputs as they use PWM too. The sample Arduino sketch should produce sounds when the buttons are pressed and released. We suspect this is one of those cases where different versions of the board profile behave differently. We recommend trying version 2.3.2 of the board profile. There should be an option to choose a new version from the Board Manager dialog box in the Arduino IDE. As for the Pico W, you are right; see our article starting on page 50 of this issue. siliconchip.com.au Helping to put you in Control 1-Wire carbon dioxide sensor Monitor the fresh air level in a room or building, the TSM400-1-CP is a combined carbon dioxide and barometric pressure sensor with a 1-Wire interface. Power 4.5 to 26 VDC. SKU: TCS-016 Price: $340.95 ea Modbus carbon dioxide sensor TSM400-4-CP is a combined carbon dioxide and barometric pressure sensor with a Modbus RS485 interface. SKU: TCS-017 Price: $340.95 ea ToughSonic Chem 14 - 4.3 Meter Ultrasonic Sensor Senix new ultrasonic level sensor has been designed for demanding environments such as measurement of waste water and chemical liquids. It is built to IP68 and can be submerged. SKU: SNS-0810 Price: $1589.50 ea Din rail 4-20mA adjustable single generator Powered by 230VAC the output signal is an adjustable 4-20mA set via a front mounted potentiometer. Use for testing, VFD speed control. SKU: NTR-321 Price: $156.75 ea GBMA 0-10VDC Input 3 Digit Large Display Large three digit universal process indicator accepts 0-10VDC signal with configurable engineering units. 10cm High digits. 24V DC Powered. SKU: DBI-025 Price: $559.90 ea Climate Temperature and Humidity Sensor Wall mount RHT-Climate WM-485-LCD Temperature and Humidity Sensor with LCD display, RS485 Modbus Communications and 4 to 20mA/0-10VDC outputs. Powered by 12 to 30VDC. SKU: RHT-105 Price: $332.70 ea LabJack T7 Data Acquisition Module Is a USB/Ethernet based multifunction data acquisition and control device. It features high data acquisition rates together with a high resolution ADC. SKU: LAJ-045 Price: $902.00 ea For Wholesale prices Contact Ocean Controls Ph: (03) 9708 2390 oceancontrols.com.au Prices are subjected to change without notice. Australia's electronics magazine January 2023  7 RF interference from LED lamps is still a problem I recently bought some Philips warm white 4.6W LED lamps, GU10 style, and not for the first time, you cannot run a radio in the same room, especially on AM. Not one station could be established on a fixer set on the broadcast band, or any other, as it was drowning them all out. I have returned one LED floodlight for the same reason. Obviously, regulations have been gobbled up by privatisation and its hunger for dollars, and to hell with the consequences. I think this is like several other devices, where following up on the RFI regulations only applies to ham radio operators. One could be cynical and suggest that things are submitted with suppression and built without in production. Or perhaps the regulatory authority is just there to collect revenue like most others these days. RFI might make an interesting exposé article as it’s getting worse. Why do we have so much “switch-mode” expense incorporated in a light where a resistance and capacitor would suffice? We are probably going to FM by stealth to avoid the RFI issue. In a disaster, AM will carry further. For example, historic 2CO Corowa, when it had its ‘heritage’ Alexanderson Antenna, could be heard from Corowa NSW to Broadford near Melbourne during the day. Marcus Chick, Wangaratta, Vic. Comments: it is unfortunately quite common for LED lights to blanket the AM broadcast band with hash. Not all do it, though. You might have to try a few different brands and models until you find one that is acceptably quiet. As for why switch-mode supplies are used, we think it’s partly because the power factor of a more basic power supply would be too poor and also, such lights might not meet government efficiency requirements otherwise. Odd fault in 110dB Attenuator had unlikely cause I built the 0-110dB RF Attenuator from the July 2022 issue (siliconchip.au/Article/15385) as a companion to the AM-FM DDS Signal Generator from the May 2022 issue (siliconchip.au/Article/15306), but it would not run. It only got to the version screen and no further, plus the display had some blank sections in the lower right. I have attached a photo (shown below). Charles Kosina contacted me and generously offered to look at the Attenuator. I sent it to him, and he fixed it exceptionally quickly (a fabulous service). I built the 8 Silicon Chip Attenuator from the kit, and apparently, the regulator was the reversed connection type (as mentioned in the article). Still, I fitted it unsuspectingly on the top of the board. As the unit powered up initially, I did not consider that there was a problem with the regulator; I assumed it would not operate at all if reversed. Charles informs me that reversed, the regulator only delivered 2V, enough for the unit to start up, but the OLED stalls shortly after power-on. I cannot express my gratitude to you and Charles for your efforts! It goes to show why Silicon Chip is so highly regarded worldwide. Jim Anstey, Townsville, Qld. Comments: we were sure it was a problem with the ATmega328P chip or a faulty OLED screen. That goes to show that you have to think about all the components in the circuit when troubleshooting. On electrostatic CRTs and ion implantation Referring to Dr Hugo Holden’s article on the novel Admiral 19A1-series television described in June 2022 (siliconchip.au/Article/15354), he states that he knows of only two brands of television that employed an electrostatic deflection CRT display. The family group is a bit larger than that. The Sinclair TV80 miniature pocket TV from the mid1980s (siliconchip.au/link/abh4) employed an innovative folded flat-screen CRT that utilised electrostatic scanning in its operating principle. I recall that around the same time, Mullard in the UK announced the development of a suspiciously similar display intended for aircraft avionics. I’d like to learn more about its technical details. Does any reader know? Also, see the references for this book: siliconchip.au/link/abh5 Another point of interest: also in the June issue was Dr David Maddison’s third article on IC Fabrication Technology (siliconchip.au/Series/382). Readers may be astonished to learn that 90% of the high-precision industrial electromagnets used for ion implantation in the global IC manufacturing plants are supplied by a little high-tech company, Buckley Systems (www.buckleysystems.com/ industries), here in little old New Zealand. It’s an extraordinary level of market penetration into such a high-tech, cutting-edge industry and a real featherin-the-cap of achievement from the bottom of the world. Finally, a slightly belated obituary announcement for those readers interested. In 2020, Tim de Paravicini passed away. He was the founder and creative genius behind EARYoshino (www.earyoshino.com/), one of England’s few valve audio amplifier manufacturers. He was blessed with a natural talent for technical creativity; his late Uncle, Thomas P. de Paravicini, was a gifted mechanical design engineer working for RollsRoyce developing aero-engines during World War 2. He designed a little-known two-speed propeller airscrew reduction gearbox for R-R’s last mighty aero-engine, the sledge-hammer Eagle H46 sleeve-valve engine. Not much is known about this fascinating innovation. I would like to learn more about it and other related developments in the obscure field of multi-speed propeller gearboxes. If any reader has further information on these subjects, please feel free to contact me at pyralog<at> yahoo.co.nz Andre Rousseau, Auckland, NZ. Australia's electronics magazine siliconchip.com.au Comments: Thanks for the interesting letter, but we think you have misunderstood what Dr Hugo Holden wrote in that article. He claims that only two TV sets used the particular Faudell and White horizontal deflection circuit, not that only two CRT TV sets ever used electrostatic deflection. At the start of the article, he states: “Early CRT TV sets ... used electrostatic deflection …”. Admittedly, the sets you mentioned came much later, during the era when magnetic deflection was primarily used. Still, they had good reasons for using electrostatic deflection (mainly compactness). We mentioned the Sinclair TV80 in the article on Display Technology in the September 2022 issue on pages 22 & 23 (which you might not have seen when you wrote to us). By the time the TVs you mentioned were designed, solid state devices were much smaller and cheaper than valves had been, so there wouldn’t have been much point replacing a few of them with a large transformer, as in the Faudell and White circuit. Hybrid Bench Supply modifications I have just finished building part of Phil Prosser’s excellent project, the Dual Hybrid Power Supply (February & March 2022; siliconchip.au/Series/377). As a man of simple needs, I have only constructed a manual, mono version, which has given me an additional 5-7V at the top end, to around 32-33V total. I used a 120VA 15V + 15V AC toroid (Radio Spares RS671-9135), back-wound to slightly restrict the final AC input voltage to the regulator, producing a final 38V DC maximum input voltage to the 40V-rated pre-regulator. The original article did not nominate values for the external potentiometers connected to CON5 & CON6 for the manual version. Phil suggested 1kW pots each. For those who may be considering the manual version, I think the best option is to use two 5kW pots but reduce the highside input voltage of 5V from the LM317 (REG2) to around 2.15V on each pot input from pin 1 of CON5 & CON6. This can be achieved using 6.8kW to 8.2kW series resistors, depending on circuit tolerances. These values allow maximum rotation of the pot wipers across the entire 270° range but will not restrict/reduce the maximum voltage and amperage under control. I made one minor change in the layout, relocating the mains power switch to the rear panel so I could use the heavy-duty front panel switch as a very useful load switch. Colin O’Donnell, Adelaide, SA. Giving away unwanted magazines Regarding Bruce Dunlop’s offer of unwanted Silicon Chip magazines in the November 2022 issue (page 4), if Bruce hasn’t considered it, he might like to see if his local primary or high school would accept them as a donation to their library. The local council might even want them for their library. I am suggesting this to the Executor of my Will when the time comes, not just for my collection of Electronics Australia and Silicon Chip magazines. This also applies to the many magazines I received as a member of a food technology association when I was at uni doing food tech and for several years after whilst still a member etc. Paul Myers, Karabar, NSW. SC siliconchip.com.au Australia's electronics magazine January 2023  9 Make amazing prints with our wide range of Resin & Filament for 3D Printers 0 OV E R 2 I N S RE S CK & IN S TO G! IN COUNT FROM 3995 $ EA AVAILABLE IN 500G OR 1KG HIGH QUALITY RESIN Made specifically for colour or mono LCD/LED light source printers. Bio-Based PLA Resin Standard Resin PLA Pro Water Washable Resin • Low smell • High precision • Smooth surface • Can be drilled • Available in Red, Yellow, White, Black, Gray, Blue & Clear • High hardness & rigidity • Good moulding accuracy • Easy to print thin-walled part • Smooth printing surface • Available in Red, Yellow, White, Black, Gray, Sky Blue & Clear • Higher resolution & precision • High strength & toughness (balanced performance) • Available in Gray, Black & White • Wash print in plain water • High precision • Low viscosity • High release rate • Available in White, Clear, Black & Gray Great for making model toys 1KG TL4433-TL4439 $59.95 EA High quality resin with low viscosity 1KG TL4443-TL4449 $62.95 EA A bio-based PLA resin which has excellent resolution 1KG TL4440-TL4442 $59.95 EA Easy washup - great for large models 500G TL4450-TL4453 $39.95 EA Shop Jaycar for your 3D Printing needs: • 2 Models of Resin Printer, with over 45 types of resin • 8 Models of Filament Printer, with over 50 types of filament and counting! • Massive range of 3D Printer spare parts & accessories PLA+ 1KG 3995 5 OV E R 2 $ EA TL4454-TL4464 NT S FIL A ME & CK IN S TO G! IN COUNT Are 100% bio-degradable and FDA food safety approved. • 10 times stronger than regular PLA • Smoother prints • Low material shrinkage rate • No cracking or brittle issues • 12 colours to choose from 1KG 3995 PETG $ EXOTIC RANGE INCLUDES RAINBOW EA TL4465-TL4471 Stronger than regular PLA • Smooth & delicate prints • Low material shrinkage rate • No cracking or brittle issues • 7 colours to choose from 1KG FROM eSILK 39 $ 95 EA TL4477-TL4480 Made from a special mix of PLA+ and certain additives to give a glossy and slightly transparent appearance when printed. • Choose from Gold, Silver, Copper and Rainbow colours ABS+ 1KG 3995 $ EA TL4472-TL4475 HIGH QUALITY FILAMENT Are extremely durable, lightweight alternatives to PETG and PLA. Able to withstand heat and stress without cracking or weakening, finished prints will hide blemishes & minor print issues. • Durable and lightweight • No cracking or brittle issues • Highly resistant to chemicals • 4 colours to choose from High quality 3D filament, known for producing smoother prints & better adhesion. Our expanded range of filament colours and types includes some new exotics to help ensure we have the right filament for your next work of art. Explore our wide range of resins, filaments and accessories, in stock on our website, or at over 110 stores or 130 resellers nationwide. jaycar.com.au/3dprinting 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. > THE HISTORY OF COMPUTER MEMORY > EARLY DATA STORAGE PART 1 BY DR DAVID MADDISON One of the most critical technological advances driving the widespread adoption of computers has been smaller, faster, higher-capacity memory chips. It didn’t start with semiconductors and ICs, though; memory has been around in various forms for a long time. This two-part series will investigate how it started and grew into what we have today. “Extended ASCII” (not an official EMORY is one of the most for temporary storage, rather than ‘secM important and commonly dis- ondary memory’ used for long-term name) uses 8 bits and has 256 characters, with extended foreign language cussed elements of a computer. These storage, such as hard disks. days, computer memory size is measured in gigabytes or even terabytes. While huge & cheap memory capacities are taken for granted, early computers had tiny memories because integrated circuit technology had not yet been developed and storing even one byte was expensive and complicated. For one byte, eight zero or one values need to be stored, so something had to be duplicated eight times. Without integrated circuits, whatever was used to store that information was expensive and big. Note that there are and have been systems that use bytes with fewer or more than eight bits, but eight is the most common number. In this two-part series, we will focus mainly on ‘primary memory’, the working memory of the computer 12 Silicon Chip However, the distinction wasn’t always clear in early computers, which also lacked convenient input and output systems. Hence, we will discuss technologies like punched cards and paper tape that were used for both primary and secondary storage. Secondary storage may be the subject of another article. Bits and bytes One byte is the unit of digital information typically used to encode a character, such as the ASCII-1977 character set members, which includes the letters and numerals A-Z, a-z, 0-9, punctuation and special characters. ASCII is a 7-bit encoding scheme that represents 128 printing and non-­printing characters. Australia's electronics magazine characters, symbols, line-drawing characters etc. The exact set of symbols depends on various proprietary implementations or standards like ISO/IEC 8859. That has largely been supplanted now by Unicode (see panel). Still, with this system, one character is stored in one byte. Five bits is the minimum amount of storage necessary to represent the alphabet; however, with just five bits, all 26 letters could be represented in one case (upper or lower), but not all numbers. So most 5-bit character code sets could switch between letters (LTRS) and numbers (FIGS), allowing 60 letters, characters and other codes to be used – see Table 1. Today, a byte usually consists of eight bits represented by 0 or 1 and siliconchip.com.au in the range 0 (decimal) or 00000000 (binary) to 255 (decimal) or 11111111 (binary). However, past computers have used fewer, such as 6-bit codes to represent 64 characters. Other byte sizes are also used for addresses and number representation in modern computers in CPUs or GPUs (graphics processing units) such as 16, 32, 64, 128, 256 bits and beyond. These architectures usually still group data in multiples of 8-bit bytes. You might have noticed a discrepancy between the stated size of a disk drive and the size reported by the computer operating system. That depends upon if the size is counted in decimal or binary. One kilobyte is 1000 bytes in decimal notation or 1024 bytes (210) in binary notation, while one gigabyte is one billion bytes (1,000,000,000) in decimal notation or 1,073,741,824 bytes (230) in binary notation. This represents a difference of 7.3% for gigabytes or 10% for terabytes. It is done this way because, for a computer, indexing into a large file is much more easily done in power-oftwo chunks (like 1024) than decimal sizes like 1000. This discrepancy has resulted in new terms such as kibibyte (KiB; 1024 bytes), mebibyte (MiB; 1,048,576 bytes), gibibyte (GiB; 1,073,741,824 bytes) etc. While it might seem more confusing at the moment, the introduction of these terms is an attempt to reduce confusion about memory sizes. Memory devices The idea of using some device to input or store data or instructions of a variable nature is not new and has its origins in the form of punched paper tape or cards, as follows: 1725 Weaving looms were controlled using paper tape ‘programs’ with punched holes, a system developed by Basile Bouchon of Lyon, France. 1804 Joseph Marie Jacquard (also of Lyon) developed a loom control system using punched cards. 1832 Semyon Korsakov (St Petersburg, Russia) proposed using punched cards for information search and retrieval. 1837 Charles Babbage (London, UK) proposed using punched cards for inputting data and instructions to his never completed (by him) “Analytical Engine”, the first ‘Turing-complete’ siliconchip.com.au Table 1: patterns represented for a given number of bits. Bits Number of patterns (2bits) 1 2 (0 or 1) 2 4 (00, 01, 10 or 11) 3 8 (000, 001, 010, 011, 100, 101, 110 or 111) 4 16 (numerals 0...9 plus some punctuation) 5 32 (26 letters plus some punctuation) 6 64 (26 letters in two cases, ten digits, space & full stop) 7 128 (all ASCII characters) 8 256 (full code page or Unicode UTF-8) 16 65,536 (UTF-16) 32 4,294,967,296 (UTF-32) 64 1.84 × 1019 128 3.40 × 1038 256 1.16 × 1077 Fig.1: punched cards were used as the memory for the first ‘Turing-complete’ computer, Charles Babbage’s Analytical Engine. The smaller cards specify the mathematical operations to be performed, while the larger cards hold numerical variables. Source: https://w.wiki/5xR7 (CC-BY-2.0). computer. It was mechanical rather than electronic since electronic technology was still in its infancy. It still contained all the elements of a modern computer – see Fig.1. IBM punch(ed) cards For over half a century, the world’s most common medium for information storage was the once-ubiquitous IBM punched card. They have a fascinating and long history, but we do not have space to cover it all here, so we will just mention the highlights. Punched cards were not developed for computers, which did not yet exist, but for machines that tabulated data. The “IBM card” originated with Herman Hollerith (New York, USA) in the 1880s and 1890s, who used them in mechanical tabulating machines. These electromechanical machines were used to summarise information encoded on punched cards, such as census data (see Figs.2 & 3). Hollerith’s company eventually became part of IBM, and the machine became a core product. The tabulating machine was not a computer, but it could perform some The Unicode Standard Unicode is an international character set with 149,186 characters and symbols (as of version 15.0) in current use. Before Unicode, every different language required a distinct ‘code page’, making mixing different languages virtually impossible and leading to much confusion. Unicode solves this by bringing all the characters needed for human languages together in one set. Clearly, you can’t encode that many characters in a single byte. Therefore, in modern computer memory systems, characters are generally encoded as variable-length byte strings, providing backward compatibility with existing single-byte character sets like ASCII. There are several valid Unicode encoding schemes. Probably the most common is UTF-8, where a Unicode character that’s also part of the ASCII set is encoded as a single byte with its top bit as 0. Other characters or symbols are encoded as multiple bytes (up to four), where the first byte has its top bit set to 1. Other schemes that are part of the standard include UTF16, UTF-32 and BOM. Australia's electronics magazine January 2023  13 Fig.2: a replica of an 1890 model Hollerith punched card tabulating machine used to process data from the 1890 US Census. Source: https://w.wiki/5xR8 (CCBY-2.0). mathematical operations, group data and print results. The first IBM card had 22 columns and eight rows (punch positions); by 1900, they had 24 columns and 10 rows; and by the late 1920s, 45 columns and 12 rows. In 1928, a new version of the card was introduced with 80 columns and 10 rows – see Fig.4 (they moved to 12 rows in 1930). Those punched cards are the likely reason that early alphanumeric computer monitors had 80 columns. The cards measured 7-⅜ inches by 3-¼ inches or 187.3mm × 82.5mm. These dimensions were that of US paper currency from 1862-1923. The IBM card had many incidental uses besides computers; they were often used for taking notes and making dot points for presentations, as they fitted the inside pocket of a suit jacket. IBM was not the only manufacturer of punched cards or equipment to read and write them, but they became known by that name. People may laugh at punched cards today but, like books, if stored correctly, the data will be readable with the naked eye far into the future. However, data stored on CDs, magnetic disks and the like may deteriorate over time (disc rot) or become unreadable due to a lack of software and hardware support. Punched paper tape Fig.4: an IBM card. The data encoded is one line of a FORTRAN program: “ 12 PIFRA=(A(JB,37)-A(JB,99))/A(JB,47)    PUX 0430” Source: https://w.wiki/4icp (CC BY 2.0). Punched paper tape is conceptually similar to cards but can be kept on long rolls (sometimes formed into a loop) rather than on individual cards. It was invented in 1725 by Basile Bouchon to control looms, but that was impractical at the time. Like punched cards, punched paper tape was used for various applications in the 19th and 20th centuries, such as programmable looms, telegraphy systems, CNC machine tools and computer data input and storage from the 1940s (including military code-­breaking during WW2; see Fig.5) through to the early 1970s. Data stored on tape was also used as read-only memory (ROM) for computers. Tougher versions of the tape for industrial use were made with Mylar. Like cards, paper tape has the advantage of being able to be read by eye and is long-lasting if used and stored correctly. Paper tape was usually 0.1mm thick and either 17.5mm wide (11/16th of Australia's electronics magazine siliconchip.com.au Fig.3: a Hollerith punched card from about 1895, the predecessor of the IBM card. Source: https://w.wiki/5xR9 (public domain). 14 Silicon Chip Beyond punched cards & tape 1918 William Eccles and Frank siliconchip.com.au CAR. RET. LINE FEED LETTERS FIGURES SPACE THRU Uppercase Lowercase BELL Fig.5: paper tape as used on the WW2 Colossus Mk2 code-breaking computer in 1943. This computer had no internal memory storage (RAM), so the program tape had to be continuously read in a loop. Source: https://w.wiki/5xRA CITY an inch) for five-bit codes, or 25.4mm wide (one inch) for 6-bit or more codes. The hole spacing was 2.54mm (1/10th of an inch) in both directions. Sprocket holes were 1.2mm (0.046 inches) apart. Paper tape could store 10 characters per inch (25.4mm). A standard teletype roll was 1000 feet long (305m), so it could store up to 120kbytes, but most tapes were much shorter than that as many contemporary computers couldn’t handle that much data. Several different encoding schemes were used, starting with Baudot’s from the 1870s. It was developed for telegraphs and used five holes (five bits). In 1901, the Baudot scheme was modified to create the Murray code that included carriage return (CR) and line feed (LF) – see Fig.6. Western Union used that until the 1950s; they modified it by adding control codes, a space and a bell (BEL) symbol to ring a bell. 1924 the Western Union code was used by the International Telecommunications Union (CCIT) as the basis of the International Telegraph Alphabet No. 2 (ITA2), a version of which was adopted by the USA and called TTY. TTY was used until 1963. All of the former systems used 5-bit codes, after which 7-bit ASCII was adopted. There were also some encoding schemes that used six bits. The IBM Selective Sequence Electronic Calculator was an electromechanical machine that operated from 1948 to 1952 – see Fig.7. It used uncut IBM card stock to create tapes that were 7.375 inches (18.73cm) wide and the length of an IBM punched card (joined end-to-end). Each of the 80 columns could contain a signed 19-digit number with parity bits plus two rows for side sprockets. The tape(s) typically contained large mathematical tables; with multiple readers and up to 36 tapes, they could be searched in about one second. There were another 30 readers for program data. The rolls could be continuous or looped; a full roll weighed 400lb (181kg). About 400,000 characters could be stored on the tapes. The machine also used IBM punched cards. It gave IBM excellent publicity and was the basis for many interpretations of what a computer looked like. - ? : $ 3 ! &# 8 ( ) . , 90 14 57 ; 2 /6 " A B C D E F G H I J K L MN O P Q R S T U VW X Y Z Feed holes Paper tape showing the five-bit Baudot Code Fig.6: the five-bit code implemented on paper tape. More common characters use fewer holes. Source: https://savzen.wordpress.com/tag/baudot/ Fig.7: a retouched version of the famous photo of the IBM Selective Sequence Electronic Calculator. The 181kg paper tape rolls on the readers in the background were made of IBM card stock. Source: www. thedigitaltransformationpeople.com/channels/enabling-technologies/ mainframes-can-be-cool/ Australia's electronics magazine January 2023  15 Fig.8: circuit diagrams of the Eccles and Jordan flip-flop from their patent application. Fig.9: the magnetic drum memory from a Swedish BESK computer, with a sample of much more compact core memory of unknown capacity above it. Source: https://w.wiki/5xRB (GNU FDL). Jordan filed a patent entitled “Improvements in Ionic Relays” and received British patent 148,582 in 1920 – see Fig.8 & siliconchip.au/link/abhs While not intended for computer memory (electronic computers had not yet been invented), it was to become the basis of later computer memory. It comprised two valves (vacuum tubes) that could exist together in one of two stable states. It was originally called the Eccles– Jordan trigger circuit, the trigger circuit or a multi-vibrator, but today it is known as a flip-flop. The ability for a flip-flop to exist in either of two stable states representing a 0 or 1 is the basis of some computer memory today, such as SRAM (static random-access memory, see the 1963 entry later) and CPU registers. 1932 Austrian Gustav Tauschek invented magnetic drum memory in 1932, which became a widely used form of primary computer memory (‘RAM’) in the 1950s and 1960s. How was this device invented before the first programmable digital computer? It was initially devised to record data from punched card machines and then was adopted for early computers. Tauschek’s original device from 1932 had a capacity of 500,000 bits or 62.5kbytes. As the name implies, drum memory consists of a drum coated with magnetic material; several read and write heads are mounted along the length of the drum. Drum memory initially displaced CRT and delay line memory (see below) because it was more reliable. Magnetic core memory gradually replaced drum memory for primary storage. Keyboard drum Decimal-to-binary conversion drum Capacitor Counter drum Decimal card reader Carry-over drum Motor Drum memory was also used for secondary (semi-permanent) storage and, in this role, drums were eventually replaced by floppy disk drives starting in the early 1970s. One of the latest known uses of drum memory is in US Minuteman ICBM launch site computers (until the mid-1990s). Fig.9 shows drum memory from the 1953 Swedish BESK computer and magnetic core memory from the same machine. The capacity of neither device is known. The BESK computer was used to create the first computer animation; see the video titled “Rendering of a planned highway (1961) First realistic computer animation” at https://youtu.be/oQMD7oufO4s 1942 John Atanasoff and Clifford Berry built the little-known ABC (Atanasoff-­ B erry computer) – see Fig.10. Some argue that this machine Memory Disk (25 capacitors per side) One-cycle switch Carry-over capacitor Drive motor Base 2 card reader Base 2 output card puncher Power supply and regulator 30 add-subtract logic circuits Electrical card-punching circuits Power supply Memory-regenerating circuit Memory-regenerating circuits Add-subtract logic circuit Fig.10: an overall view of the ABC computer (left) and details of its regenerative capacitor memory unit (right) showing only one disc of 30 and one drum of two. Source: www.researchgate.net/publication/242292661 16 Silicon Chip Australia's electronics magazine siliconchip.com.au is the first automatic electronic digital computer; others dispute that because it was not programmable and was not Turing-complete. It was at least what would today be called the first ‘arithmetic logic unit’ (ALU), now built into all computers. This was the first computer to use regenerative capacitor drum memory, not to be confused with Tauschek’s drum memory mentioned above. Regenerative capacitor memory uses individual capacitors to store memory bits. They are either charged or discharged to represent a 1 or 0. Because capacitors discharge with time, they constantly need to be ‘refreshed’, much like some other forms of memory (such as DRAM, to be discussed next month). The ABC computer had two drums that stored 1500 bits each (thirty 50-bit numbers) which rotated at 60 RPM; the capacitors were refreshed on every rotation. The ABC computer (if it is accepted as such) was the first computing machine to use flip-flop memory of the type described above by Eccles and Jordan. You can see a fascinating video about how the ABC works on YouTube – “The Atanasoff-Berry Computer In Operation” – https://youtu.be/ YyxGIbtMS9E 1943 The British Colossus code-breaking computer is regarded as the world’s first programmable digital computer (see Fig.5). It was the first device universally accepted as a computer to use the flip-flop design from Eccles and Jordan. The flip-flops were implemented with vacuum tubes as transistors had not yet been invented. They were used for counting and logical operations, as the computer had no memory except the paper tape loop mentioned earlier. 1945 The first programmable general-­purpose digital computer was ENIAC, used for artillery calculations by the US military. It started with 20 words of system memory, or about 80 bytes, in the form of accumulators. Extra data was stored on IBM punched cards; a 100-word magnetic core memory unit was added in 1953. ‘Words’ are of variable size for different computers. For ENIAC, a word was ten binary-coded decimal digits in length, at a time before eight-bit bytes were standardised. Most modern computers use 16-bit (two-byte), 32-bit (four-byte) or 64-bit (eight-byte) words. siliconchip.com.au Fig.11: a 256-bit Selectron tube. Source: https://w.wiki/5xRC (GNU FDL). Cathodes Selection Bars Collector Plate Storage Eyelets Mica Backplate Writing Plate Write Pulse Reading Plate Read Pulse Faraday Cage Output Grid Signal Out Phosphor Screen Glass Plate Fig.12: how the Selectron tube worked. The arrows near the bottom indicate the secondary emission of electrons that generate a pulse indicating a one-bit. In contrast, the arrows higher up and to the right indicate no secondary emission of electrons, indicating a zero-bit. Source: https://w.wiki/5xRD The original 20-word ENIAC memory used flip-flops in the form of a pair of triode valves. Ten flip-flops were joined to form a decade ‘ring counter’, capable of storing and adding numbers. A ring counter comprises a system of flip-flops and a shift register with the output of the last flip-flop fed to the first to make a ‘ring’. A PM (p for positive and m for negative) counter circuit was also used to store the sign of the number. One PM counter and 10 ring counters made up an accumulator. 1946 Development work on the Selectron tube (Fig.11 & Fig.12) was started by Jan A. Rajchman at RCA. This vacuum tube stored digital memory data in the form of electrostatic charges, similar to the Williams-­ Kilburn tube discussed next. The original design was for 4096 bits, but that was too difficult to build, so a 256-bit form was made. The device was never a commercial success; both it and the Williams-­ Kilburn were superseded by magnetic core memory, which was more reliable, cheaper and easier to manufacture. Australia's electronics magazine The basic principle of operation is shown in Fig.12. Electrons are emitted from the heated cathodes at the top of the diagram, like an electron gun but not a point source. Each cathode is surrounded by four selection bars, two each running in one direction and two at right angles to those. The selection bars adjacent to the cathode corresponding to the selected bit are activated to address a particular bit. Electrons move from the cathode through the collector plate and toward the storage area, which consists of eyelets (like those on some shoes but much smaller) embedded in a sheet of insulating mica with a metal backing called the writing plate. The eyelets are insulated by the mica sheet but capacitively coupled to the writing plate. By pulsing (or not) the writing plate at the same time as electrons are moving toward the selected eyelet storage location (as determined by the selection bars), the eyelet can either be charged or not, thus ‘writing’ the data to be stored. If the pulse is the same potential January 2023  17 Fig.13: a Williams-Kilburn tube from an IBM 701 at the Computer History Museum in Mountain View, California, USA. Source: https://w. wiki/5xRF (CC BY-SA 3.0). Fig.14: data in the form of dots and double dots written to a WilliamsKilburn CRT memory tube. The double dots are because a second dot has been drawn as part of the erase process. Source: https://w.wiki/5xRE 18 Silicon Chip as the collector plate, electrons will pass through the collector plate and charge the eyelet (downward-facing arrows on the left of the diagram). If the potential is the same as the cathode, electrons will be blocked and not charge the eyelet. Thus, the eyelet can be in one of two states. For reading the data out, electrons from the cathodes will either pass through an eyelet or be inhibited from passing through to the reading plate, depending on its charge state. By selecting an eyelet using the selection bars and pulsing the reading plate, the signal from the output grid will indicate whether it is charged. After passing through the reading plate, electrons go through holes in a Faraday cage and strike a phosphor screen. This causes the phosphor to glow, indicating the contents of individual memory locations (the eyelets) as well as passing secondary electrons to the output grid. For more information on how the Selectron worked, see the website: www.rcaselectron.com 1946 The Williams-Kilburn tube was patented in the UK and US in late 1946, 1947 and 1949. It was the first fully electronic (and thus high-speed) memory, using a CRT (cathode ray tube) for storage. The fact that CRTs were used this way was mentioned briefly in our article on Display Technology in the September 2022 issue (page 18, middle column; siliconchip. au/Article/15458). This type of memory was first used to run a computer program in 1948. Simply put, a Williams-Kilburn tube (Fig.13) stores memory on a CRT by writing a dot pattern representing the data to be stored (Fig.14). As with any CRT, the image has a certain persistence but eventually fades away. Therefore, it must constantly be ‘refreshed’ by each bit being periodically read and re-written (similar to DRAM). A small charge of static electricity appears above each dot which fades over a fraction of a second. It is this charge that gives the tube persistent storage. So writing a ‘one’ to the display involves steering the electron beam to a specific position and delivering electrons from the gun to allow the charge to build up. To write a zero, the charge at the dot must be neutralised. This is done by drawing a second adjacent dot (or line) Australia's electronics magazine because a negative halo is generated around each dot. This eliminates the positive charge of the first dot nearby. Reading the state of a bit is done with the aid of a thin metal plate on top of the viewing screen. The electron beam is steered to that location and energised, just like writing a ‘one’. If a ‘one’ was already present, there is no change in the charge at that location, so no current flows through that metal plate. But if there was previously a ‘zero’, writing the ‘one’ will cause a detectable current to flow. The Williams-Kilburn tube was susceptible to external influences, mainly from electric fields, so frequent adjustments were required for error-free operation. Some notable uses of the tube were the IBM 701, IBM’s first electronic digital computer from 1952. It had 72 3-inch Williams-Kilburn tubes, each having a capacity of 1024 bits, giving a total memory of 2048 words, each having 36 bits. The memory could optionally be expanded to 4096 words. Another use was MANIAC I (Mathematical Analyzer Numerical Integrator and Automatic Computer Model I; Fig.15) at the Los Alamos National Laboratory, which used 40 2-inch tubes to store 1024 40-bit numbers for hydrogen bomb calculations and it became fully operational in 1952. 1947 Frederick Viehe filed for US patent 2,992,414 for magnetic core memory (Fig.16) in 1947, although it wasn’t awarded until much later, in 1961. He filed another related patent in 1962 (US3264713), awarded in 1966. Magnetic core memory was the dominant form of computer memory from about 1955 to 1975. Incredibly, Viehe was a Los Angeles pavement inspector who played with magnetics as a hobby; he was not a professional scientist or engineer. IBM eventually purchased his patents. Core memory uses tiny toroids of magnetic material wired as simple transformers. By passing a current through wires that go through the toroid, it can be magnetised in one direction or the other, thus storing a bit of information. A sense wire passing through the core detects if the toroid has changed state. Reading the data (magnetic polarity) is a destructive process, causing the bit to be set to zero. To read a bit of data, an attempt is made to flip a siliconchip.com.au bit. Nothing happens if it is a zero; if it is a one, the toroid changes polarity, inducing a pulse in the sense line. The information is retained even when the power is turned off. A piece of magnetic core memory is one of the most desirable items in any collection of electronic ephemera. They are fine examples of delicate manual construction and are almost works of art. Other claimants to this invention were An Wang (1949; US patent 2,708,722 awarded in 1955), Jan Rajchman (1950) and Jay Forrester (1951); there were many ‘intellectual property’ disputes over it. In 1964, IBM paid MIT (where Jay Forrester worked) US$13 million for his patent, a substantial amount of money at the time. Core memory eventually obtained a volumetric density of about 900 bits per litre, and the cost went down from about $1 per bit to 1c per bit. The beginning of the end for core memory was when Intel introduced the 1103 DRAM IC in 1970, costing 1c per bit. While core memory is obsolete, computer memory is sometimes still referred to as “core”. A file containing the contents of memory from when a program was running is still often referred to as a “core dump”. 1947 J. P. Eckert and J. W. Mauchly applied for US patent 2,629,827 for the mercury delay line (and other forms of delay line) in 1947, awarded in 1953. The mercury delay line is a member of various delay-line-based memory devices. Delay line memories work by sending acoustic, electrical or light pulses, representing one bit, along a path. When a pulse gets to the end of the path, it has to be refreshed by reshaping and amplifying it. It is then recirculated. Such memory is accessed by waiting for the desired bit in the ‘pulse train’ to arrive at the read mechanism at a predictable time. The memory capacity is therefore determined by the length of the mechanism, the length of pulses and the speed of sound or similar in the medium. Mercury metal, a liquid at room temperature, was a common medium used in early computers. The resulting devices had a memory capacity of a few thousand bits. J. P. Eckert originally developed mercury delay lines to reduce clutter in radar return signals during WW2. siliconchip.com.au Fig.15: the aptly-named MANIAC I computer from 1952. The boxes on top of the main structure contain two-inch Williams-Kilburn CRTs used as memory. Source: https://permalink.lanl.gov/object/tr?what=info:lanl-repo/lareport/LAUR-83-5073 Fig.16: a 64 × 64 bit (4096 bits) array of ferrite core memory from 1961. This module measures 10.8cm × 10.8cm. The inset shows a detail of the ferrite cores with two address lines per bit. Source: https://w.wiki/5xRG (CC BY 2.5). Australia's electronics magazine January 2023  19 Fig.17: the “hot box” containing mercury delay line memory used in Australia’s CSIRAC computer. It was named that way because the delay lines had to be kept at 40°C. Source: https://collections.museumsvictoria. com.au/items/406411 (CC BY 4.0). Mercury was used in the delay lines because its acoustic impedance is similar to that of piezoelectric quartz acoustic transducers, thus minimising energy loss. The speed of sound is also very high in mercury compared to certain other media, meaning there is less time to wait for a pulse to arrive. Mercury delay lines were challenging to design due to the need to ensure there were no stray reflections. They were tricky to set up and maintain as they required very tight tolerances. The UNIVAC I computer mentioned below was an early computer that used mercury delay lines. 1949 CSIRAC was Australia’s first programmable digital computer and the fifth in the world. It is the oldest preserved first-generation computer. Its primary memory was a mercury delay line with a capacity of 768 20-bit words and a supplemental disk-like device of 4096-word capacity. Some of the delay lines were 10mm in diameter, 150cm long and a pulse took 960µs to go from one end to the other (Fig.17). You can see the computer on display at Scienceworks in Melbourne: siliconchip.au/link/abe2 1949 Jay Forrester had the idea to use core memory on the US Navy Whirlwind I computer; a 1024-word core memory was installed in 1951, replacing CRT memory. 1950 A US military version of the ERA 1101 computer (later renamed the UNIVAC 1101) was the first computer to store and run programs from electronically accessible memory, as opposed to instructions that were hard-wired or read from tape or cards. The military version was known as the ERA Atlas. 1951 Magnetic tape drives on computers were first used on the UNIVAC I computer (Fig.18). The drive unit was the Remington Rand UNISERVO I (Fig.19), which used half-inch wide metal tape (12.7mm) in 1200ft (366m) lengths. The metal tape and reels weighed 25lbs (11.3kg). The tape had six data channels plus one for parity and another for timing, and had a density of 128 bits per inch. Each tape could hold 1,440,000 seven-­bit characters. Later versions of these drives used plastic Mylar tape, which became the industry standard. The IBM standard for information formatting on tape was widely adopted. You can view an original UNIVAC I promotional video titled Fig.18: a bank of reel-to-reel tape drives (background) on a UNIVAC 1108II computer from a 1965 UNIVAC sales brochure. Source: http://s3data. computerhistory.org/brochures/sperryrand.univac1108ii.1965.102646105.pdf 20 Silicon Chip Australia's electronics magazine “Remington-­Rand Presents the Univac” at https://youtu.be/j2fURxbdIZs The Autumn 1964 issue of Martins Bank (UK) magazine reported that when data from one inch of paper tape was transferred to magnetic tape, it occupied 1/80th of an inch. The same bank reported that paper tapes were used for programming branch computer terminals as late as 1981 – see siliconchip.au/link/abht 1952 The concept of ferroelectric RAM was described in Dudley Buck’s master’s thesis. Bell Telephone Laboratories conducted some experiments on the concept in 1955, but it was not commercially available until the 1980s and 1990s (which will be described in more detail next month). Ferroelectricity is a property of certain materials with an electric polarisation state that can be reversed by applying an electric field. The state is kept even without the continued application of the electric field. The two states can be used to store binary information. 1952 The IBM 726 computer was introduced. It was the first computer to use magnetic particle coated plastic tape for storage (see Fig.20 and visit siliconchip.au/link/abhu). It could read or write 12,500 characters per second and each tape had a capacity of two million characters. The tape was about half an inch (12.7mm) wide and had six data tracks and a parity track. Fig.19: a promotional image of the UNISERVO I tape drive. Source: www.computer-history.info/Page4. dir/pages/Univac.dir/images/ MagTapeDrive.jpg siliconchip.com.au The storage density was 100 bits per inch, and tapes were up to 1200ft (366m) long. 1953 The first transistor computer originated at the University of Manchester. There were several experimental designs from 1953, culminating with a commercial design in 1956 by a Manchester company with the computer called the Metrovick 950. Only a small number were built. Early transistor computers may have used valves for the clock and other functions. Possibly the first fully-­ transistor computer was the Harwell CADET from 1955, but there were several other early claimants. Philco shipped commercial transistor computers, the S-1000 and S-2000, in 1958. The RCA 501 and the IBM 7070 are also from 1958. The TRADIC (for TRAnsistor DIgital Computer or TRansistorized Airborne DIgital Computer) was an early US transistor-based computer used on the B-52 bomber. It had 684 Bell Labs Type 1734 Type A cartridge transistors and 10,358 germanium point-contact diodes. It also used one valve in the power supply. Early transistor computers used drum memory or magnetic core memory, not transistor circuits as memory elements. However, transistors were used as registers for CPUs and amplifiers for magnetic core memory. Diodes were used in arrays as a form of ROM (read-only memory). 1955 The Konrad Zuse Z22 was the first commercial computer to use magnetic core memory (14 words of 38 bits) as well as magnetic drum memory (8192 38-bit words). It also used paper tape and had 600 vacuum tubes. 1957 Bell Labs introduced Twistor memory in 1957, first used in 1965. It comprised a piece of magnetic tape wrapped around a current-carrying wire and was similar in operation to magnetic core memory. It saw limited use; however, the ideas were incorporated into bubble memory (described next month). 1958 The Ferranti-Sirius magnetostrictive delay line was introduced (see Fig.21). It used the magnetostrictive effect whereby a material changes its shape in response to a magnetic field. A long coil of magnetostrictive material was fabricated, with an electromagnet at one end that induced a torsional wave (twist) in the wire that travelled down its length. Such torsional waves were more Fig.20: an IBM 726 magnetic tape unit, as used by the IBM 701 computer system. Source: https:// johnclaudielectronics.tumblr.com/ post/42914025003/ Fig.21: a magnetostrictive delay line. Source: https://w.wiki/5xRH (CC BY-SA 3.0). siliconchip.com.au Videos on punched tape storage ● A homemade paper tape reader: “Paper tape reader demo” at https:// youtu.be/w7_9BmthB10 ● Using paper tape with an Altair 8800, a microcomputer kit sold in 1974 and the first successful PC. The computer used in the demonstration is actually a modern clone. “Altair 8800 - Video #28 - High Speed Paper Tape Reader/Punch”: https://youtu.be/wALFrUd6Ttw Electronics Australia’s EDUC-8 was published about the same time and also supported paper tape (see siliconchip.com.au/Shop/3/1816). Australia's electronics magazine resistant to noise than the compressive waves used in mercury delay lines. A typical magnetostrictive delay line in a package about 30 × 30cm could hold about 1kbit of data. They were used through the 1960s in computers, video display terminals and some calculators. 1959 US patent 3,161,861 was filed by Kenneth Olsen, awarded in 1964, concerning magnetic core memory. 1962 CRAM (Card Random-Access Memory) was introduced by NCR – see Fig.22. It used cartridges containing 256 plastic cards with magnetic coatings, which together could hold 5.5MB. The device was mechanically complex but surprisingly successful, and was an alternative to magnetic tape until being surpassed by disk drives. 1963 Robert Norman at Fairchild patented static RAM (SRAM; US patent 3,562,721). It was faster than magnetic core memory and the logic circuitry used fewer components than January 2023  21 Table 2: generations of computers and technology used Generation Technology Approximate date range 1st Valves 1940 to 1956 2nd Transistors 1956 to 1963 3rd Integrated circuits 1964 to 1971 4th Microprocessors 1971 to present 5th Artificial intelligence Present and future for other forms of memory. It was used by IBM. According to the patent, “This invention provides a new switching circuit, particularly designed for a logic memory circuit, which achieves a substantial reduction in the number of components required.” 1964 The first 64-bit SRAM was designed by John Schmidt at Fairchild. 1965 We don’t have a precise date for the introduction of rope memory (Fig.23), but we know it was used in Apollo Guidance Computers by 1965. Rope memory was a form of core memory with its physical configuration altered to be much more compact than regular core memory (due to the woven core pattern), giving the higher storage density required for spaceborne computers, but was read-only memory. It was about 18 times more compact than regular core memory. It was Fig.22: a CRAM device from an NCR product brochure. Source: http://archive.computerhistory. org/resources/text/NCR/NCR. CRAM.1960.102646240.pdf 22 Silicon Chip used not only for storing data but also computer programs. Its operation was vastly more complicated than standard core memory, with multiple wires and bits per toroid and much larger toroids. It is described in a video titled “MIT Science Reporter—Computer for Apollo (1965)” at https://youtu.be/ ndvmFlg1WmE?t=1245 The process of making rope memory for the Apollo computers can be seen from 20:45 in that video. There is also a video about restoring an Apollo guidance computer, which has more details of its operation, titled “Apollo Guidance Computer Part 14: Bringing up fixed rope memory” at https://youtu.be/2qe4W_USweE Brek Martin has made a core rope memory simulator; the first video is at https://youtu.be/c-t2qyHOs7Y 1965 The Fixed Resistor-Card Memory was an experimental form of punched card. Information was stored by severing (or not) connections to an array of resistors on a cardboard or plastic card; it could be punched on existing punch-card machines. Next month After 1965, silicon-based memory started rapidly taking over from the technologies described so far. The second and final part of this series next month will pick up where this one left off, explaining how the semiconductor revolution radically changed computer memory up to the present day. If you haven’t already seen it, in preparation for the upcoming part two, you might want to read the series of articles on IC Fabrication technology in the June, July and August 2022 issues. They tie in with the computer memory technology revolution that SC came after 1965. Fig.23: a test sample of core rope memory for the Apollo Guidance Computer. Actual production examples were much more compact than this. Source: https://w.wiki/5xRJ (CC BY-SA 3.0). Australia's electronics magazine siliconchip.com.au NEW YEAR NEW Build It Yourself Electronics Centres® Great build volume & features! GEAR K 8602 SAVE $260 639 $ Gear up the electronics workbench for 2023 with these great deals on tools, test gear & 3D printing. SAVE $350 Creality® Ender 5 Pro | Desktop 3D printer. Produce one off prototypes, replacement parts and hobby pieces with printing up to 22Wx22Dx30Hcm! The Ender 5 Pro offers workhorse 3D printing for your workshop with excellent print speed and accuracy using PLA, ABS and more. A double Y axis control system provides stability during the printing process, ensuring you get a great print every time. Q 0203A 549 $ NEW! 100MHz 2 Ch. Digital Storage Oscilloscope Iroda Mini Blow Torch ® 55.95 $ Perfect for those in R&D or servicing. • 2 channels with real-time 1GSa/s sampling. • Colour 7” TFT screen • Displays waveform plus the measured wave voltage, peak to peak plus RMS, frequency, duty cycle etc. • Realtime measurement PC software. • USB datalogging • 2 year warranty. T 1528A Need an extra laptop charger? SAVE $28 40 $ M 8868A D 2322* Build wireless charging into your desk This 65W USB-C power delivery (PD) charger offers recharging for MacBooks, Nintendo Switch and other type C equipped devices. Great for the workbench! An ultra-slim desk mount 10W wireless fast charger. Requires 60mmØ hole. Includes power adaptor & USB cable. SAVE 15% A 1300°C blow torch with adjustable gas feed for a variety of tasks such as brazing and model making. Refill with Iroda gas T 2451 $8.50. 33 Q 2118 $ SAVE 15% 38 $ T 2486 Wire Stripper & Kwik Crimper Toolbox space saver! Combines a ratchet wire stripper, cutting blade & kwik crimper (red, blue and yellow sheaths). Suits 10-24 AWG cable. NEW! SAVE 25% 49.95 $ $ NEW! 109 $ SAVE $44 145 $ M 8305 5A 60 SAVE $40 119 $ Jakemy® 60pc Tool Kit T 2192 A combined driver bit and socket set with 47 bits and 9 metric sockets. Great for odd-jobs and repairs around the house. Includes a handy magnetic latching case. T 2175A 1000V Rated Insulated Tool Kit Ideal for technicians working on mains powered equipment. Includes cutters, pliers, wire strippers, 5 screwdrivers plus a neon test driver and 2 rolls of insulation tape - all in a handy carry case. Diagnose Vehicle Battery Problems FAST This handy digital battery analyser measures cold cranking amperes (CCA), resistance and battery cell condition. Suits 12V batteries. M 8303 3A Compact 30V Lab Power Supply Great for servicing, repair and design of electronics. Low noise switchmode design. Fine & coarse voltage and current controls. 3A or 5A max models. Size: 85Wx160Hx205Dmm. Order online at altronics.com.au | Sale pricing ends January 31st Equip your maker space. SAVE $370 INTRO SPECIAL! 999 $ Hurry, only 10 available at this price. K 8604 Everything a maker space needs in one compact unit! SAVE $151 1599 $ K 8612 Creality® CP-01 3D Printer / CNC Router / Laser Engraver Creality® CR-10 Smart Pro 3D Printer The latest generation in the popular large format CR-10 FDM 3D Printer. Offers the ultimate in large volume FDM 3D printing, with all the bells and whistles you expect from a top of the line Creality printer. Includes a simply staggering number of premium features: wi-fi, bluetooth and ethernet connectivity; cloud printing via iOS/Android app, dual mode bed levelling, HD monitor camera, LED lighting, full-metal dual gear direct extruder, magnetic build platform and finally a huge 300 x 300 x400mm build volume. The ultimate do-it-all maker machine for the workbench. Create amazing prototypes and one off designs with this all in one mini home factory. Includes three interchangeable machine heads for cutting, etching and printing each with excellent accuracy. Easily assembled from flat-pack in just a few minutes. Router & engraver suitable for plastics, wood, PCBs, laminates etc. SAVE $10 39 $ Top quality, high tolerance, bubble free for great prints every time. n K 8387A Silver n K 8388A Gold n K 8389A Pink n K 8392A Green n K 8393A Yellow 120mm long life high flow fan SAVE $40 119 $ T 1295A Desktop Fume Extractor A must have for any soldering or 3D printing work space. Whisks away fumes and filters the air through active carbon filter pads (3 included). 115CFM air flow with low noise ball bearing fan. Precision Knife Set Take the hassle out of 3D printer levelling. Creality PLA Filaments The Creality CR-Touch is compatible with most FDM 3D printers and adds auto bed levelling to your printer. This means better quality prints and layer extrusion. Includes brackets for Ender/CR series printers. n K 8395A Blue n K 8396A Red n K 8397A Black n K 8398A Grey n K 8399A White K 8630 79.95 19.95 $ 29 $ Normal RRP value of tools $67.90 Includes to handles and a variety of blades (13) to suit different cutting jobs. Includes plastic carry case. T 4015A $ T 1489 SAVE 20% 50 $ T 2168A Features 1/4” and 4mm drive handles SAVE 24 % 30 $ SAVE 15% T 2162 ‘Getting Started’ Electronics Kit Never lose a tiny screw again! Great value starter set for enthusiasts and students. Includes pliers, cutters, 30W iron, solder sucker & carry case. All you need to get soldering! A 35x26cm heat resistant silicon work mat, plus a 25x20cm magnetic mat to keep screws and materials organised while you work. 69 Piece Dual Ratchet Driver Kit Superb quality ratchet driver with a wide selection of bits for most electronic jobs. Includes both a 1/4” adjustable angle (<90°) ratchet handle and a smaller 4mm ratchet handle. Order online at altronics.com.au | Sale pricing ends January 31st Gear up for 2023! 17 $ Jellyfish® Weatherproof Equipment Cases T 5049 174x108x45 Was $22.95 78 $ T 5051 302x206x162 Was $105. Top quality sealed IP67 rated cases for storing test gear, tools, cameras, drones - anything important that needs protection! Padlockable latches with perforated foam for easy customisation. Measurements are internal size. 92 $ T 5053 352x242x172 Was $125. 125 $ T 5055 412x302x182 Was $170. 150 $ T 5056 452x352x192 Was $205. 199 $ T 5066 521x292x183 Was $265. SAVE 15% 19 15.95 $ 60 $ X 0102 Makes jewellery sparkle again! Blast away dirt & grime on parts T 2865A Side Cutter T 2870A Long Nose Plier T 2860A Bull Nose Plier NEW! 5” Carbon Steel Side Cutters Tough carbon steel blades, stay sharp longer. Ideal for cutting solid core wires. 130mm. Flat Nose Carbon Steel Pliers Flat jaws (no serrations), ideal for precision electronics work. 145mm. 130 22ea 130mm Nipper Side Cutters VDE 1000V rated electrical hand tools constructed from quality drop forged steel with comfort grip insulated handles. Rust Resistant - great for the tackle box or use in moisture prone environments. 130mm. Ideal for cutting component legs on boards, small core wiring, cleaning up 3D prints etc. All heat & no flame! T 2480 Iroda® Pocket thermo-gun. Great for removing adhesives & heatshrinking. 650°C max. Refillable. Add T 2451 butane gas for $9.35. Blow torch & soldering iron in one. Totally wireless operation - No need to run extension leads! Easy to light, oneclick piezo ignition. 2 year warranty. T 2599 kit version includes hot air tip, heat deflector, additional gas cartridge, solder, sponge and hard carry case. Add T 2451 butane gas for $9.35. SAVE $20 74 $ T 2598 Iron Only SAVE $30 129 $ T 2599 Kit SAVE 24% Great for repairing modern devices! T 1300A 15 $ SAVE 24% 15 T 2356 $ SAVE 15% 30 $ T 2555 106pc Precision Driver Set T 2183 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 15% Stainless Steel Long Nose Pliers $ 27 22 1000V Rated Electrical Tools 40 $ M 8882A* $ 2 for $ Rust free stainless steel! $ SAVE 22% SAVE 22% T 2825A Iroda® 100W Portable Gas Tool SAVE $45 *Devices & charging leads not included 65 $ 19.95 SAVE 16% Ultimate benchtop charging station! • Great for families, classrooms & business. • Massive 19A charge output • QC3.0 on 2 ports • Includes adjustable dividers & power supply. 60 $ T 2741A Super comfy handles Charge 10 USB devices at once! Why pay $300 for a MaggyLamp? The inspect-a-gadget illuminated desk magnifier is an absolute bargain at under $70, we believe ours is every bit as useful. An incredible visual aid for detailed inspection and work on fine items with full clarity through the quality glass lens. Tackle complex miniature tasks with confidence! X 4204 X 4205 3+12 Dioptre 5 Dioptre $ T 2770B Superb build quality! LED Magnifier for micro tasks NEW! $ T 2748A SAVE $15 SAVE $19 Our most popular model! Clean small parts, jewellery, shaver heads, glasses and more! Shifts grease, dust and gunk from tiny crevices in just minutes using ultrasonic waves. Tank size: 155x98x52mm. SAVE 25% Say to goodbyein! eye stra Hands free, head worn magnifier. Thousands sold! Offers 1.5, 2.6 and 5.8x magnificatio with LED lamp. Requires 2xAAA batteries. Work bench must have! Rotating PCB Holder A must have for the electronics enthusiast! Work on boards up to 200 x 140mm. Metal base provides a sturdy work platform for soldering. SAVE 23% 19 $ T 1302A Micron® Solder Reel Holders Heavy weight base with solder guide. All metal construction. *Solder not included. Shop with us on eBay | www.ebay.com.au/str/altronicsaustralia Last chance Clearance! SAVE $200 Top notch outdoor sound! Active Optical HDMI Cables SAVE SAVE OVER UP TO $78 20% 699 SAVE $36 $ 99 $ Cutting edge Active Optical Cable (AOC) HDMI technology supporting 4K resolutions at longer lengths than copper cable. Plus, it’s thinner, lighter & more flexible! Why use HDMI AOC? M 8133 240V Mains Power - Anywhere, Anytime! C 0840 Powerhouse Inverter with in-built MPPT solar charge controller. ® Weatherproof Outdoor Sound Provides you with 1500W of continuous pure sine wave mains power, plus the ability to recharge your batteries via connected solar panels. In-built maximum power point trackin (MPPT) circuitry ensures maximum charge from your panels. Ideal for caravans, RVs and boats - or anywhere you need remote 240V power! Get great permanent sound outside you home with these superb ceiling speakers. Weatherproof design is suitable for boats, alfresco areas, bathrooms and under eaves. 140mm cutout. 20W 4Ω. These cables totally eliminate the need for long distance baluns, UTP conversion and boosters for HDMI signals - get full 4K <at> 60Hz over the full distance! D 0984 SAVE $19 30 $ SAVE $29 Need an extra laptop charger? 60 $ Model Length Normally NOW USB-C laptop 90W output charges any P 7427 10m $205 P 7428 12m $215 P 7429 15m $217 P 7430 20m $230 P 7432 30m $245 P 7434 50m $327 $160 $165 $170 $180 $190 $249 This 90W USB-C power delivery (PD) charger offers fast recharging for MacBooks, Nintendo Switch and other type SAVE 25% “C” devices. Plus a $ standard 2.4A USB charger output. Electret Lapel Mic Need to record high quality audio for YouTube or live demos? This 6m electret mic offers excellent audio clarity and 3.5mm TRRS or 6.35mm TS connections. S 9846 Mini Wi-Fi Cube Camera Works with Tuya Smarthome App. Internal battery - set it up anywhere! Day/night with IR. USB rechargeable. 100 mins motion activated recording time. 70 M 8994* New gear for this year... NEW! NEW! 59 $ .95 29.95 $ Q 1278B 89 $ NEW! JUMBO DIGITS! 39.95 $ X 0604C X 7100 An office clock for all! A handy office wall clock with oversized digits for easy reading up to 30m away. Provides temperature & humidity readings, plus calendar with day of the week. Size: 306 x 176 x 36mm. Free standing. No desk clamps or screws needed! Bluetooth FM Player & Hands Free Adaptor Build It Yourself Electronics Centres Sale Ends January 31st 2023 Phone: 1300 797 007 Fax: 1300 789 777 Mail Orders: mailorder<at>altronics.com.au 59 $ H 8221 Single ® Transmits bluetooth audio from your phone (music, routes phone calls etc) to your cars FM radio. Plus it’s also a QC3.0 & USB C charger. H 8223 Dual Freestanding Monitor Mounts - reclaim your desk! Durable all steel desk mount brackets for computer monitors between 17” and 32” in size, up to 9kg each. Sturdy with plenty of adjustment to suit your work space. Quick & Easy Temperature Measuring Folding waterproof spike temperature probe with bottle opener. Ideal for a wide array of uses such as monitoring liquids in labs, kitchen and BBQ use. -50°C to 300°C Western Australia Victoria » Perth: 174 Roe St 08 9428 2188 » Joondalup: 2/182 Winton Rd 08 9428 2166 » Balcatta: 7/58 Erindale Rd 08 9428 2167 » Cannington: 5/1326 Albany Hwy 08 9428 2168 » Midland: 1/212 Gt Eastern Hwy 08 9428 2169 » Myaree: 5A/116 N Lake Rd 08 9428 2170 » Springvale: 891 Princes Hwy » Airport West: 5 Dromana Ave 03 9549 2188 03 9549 2121 New South Wales » Auburn: 15 Short St 02 8748 5388 Queensland » Virginia: 1870 Sandgate Rd 07 3441 2810 South Australia » Prospect: 316 Main Nth Rd 08 8164 3466 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. © Altronics 2022. 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 0001 Find a local reseller at: altronics.com.au/storelocations/dealers/ Review by Allan Linton-Smith 2W 930MHz RF Amplifier + RF Wattmeter You might think that 2W is not much power for an amplifier, but around 65 years ago, the first artificial satellite (Sputnik) was launched with, you guessed it, a 2W RF transmitter onboard. And it generated signals that were heard around the world. So what can you do with 2W? T his handy little RF amplifier module is rated at 2W for VHF applications between 1MHz and 930MHz. It has many applications, including boosting FM radio signals in poor reception areas. The module described here was purchased under the title “RF Broadband Power Amplifier Module for Radio Transmission FM/HF/VHF 1-930MHz 2W (Version L1C)” from eBay for $21.62 (including delivery). However, several competing suppliers are now offering similar devices at even lower prices. It is suitable for all types of radio use, such as shortwave FM radio remote control, FM radio, amateur radio in the 135-175MHz or 380-470MHz bands etc. With the recommended input signal level of 0dBm, the output power is 2.0W up to 500MHz, 1.6W at 512MHz, 1.0W at 930MHz or 0.8W at 1GHz. The input and output connectors are standard SMA female RF sockets, while 12V DC power is supplied via a pair of solder pads. At 2W output, it draws around 400mA for an input power of 4.4W. The preamp stage accounts for 110mA, meaning the power amplifier consumes close to 290mA, making it about 57.5% efficient; around what you’d expect for a linear amplifier at full power. siliconchip.com.au Features The amplifier module comes fitted to quite a good heatsink which will ensure reliability for long-term use at maximum power. During prolonged testing, the measured module temperature never exceeded 40°C and was usually just warm to the touch. As you would expect from the different power figures listed above, the overall gain varies with the signal frequency. It’s around 30-34dB for a 0dBm input up to 350MHz or 20-23dB for frequencies between about 350MHz to 950MHz. It should be noted that these gains are not entirely linear; lower input signals result in higher gain figures. For example, a -46dBm 15MHz input signal gives an output of 0dBm, meaning the actual gain, in this case, is 46dB. At higher output levels, close to 2W, the resulting distortion (THD) figure is up to 20% because it is running into clipping. With that in mind, it would be wise to operate the amplifier at reduced levels to avoid radio interference from the distortion harmonics. The distortion performance before clipping is pretty good at around 1% THD, as shown in Fig.1. Circuit details The circuit of this module is shown in Fig.2. It is pretty straightforward, using an SBB2089Z IC as a preamplifier, powered by a 78L05 5V linear Fig.1: the module’s THD before clipping was measured by feeding a 15MHz signal at -46dBm to the RF input, resulting in a 0dBm output. The first harmonic (at 30MHz) can be seen here at -40dBm, along with some noise from external radio interference. The starting point for the graph is 12MHz, with the end point at 35MHz. Steps are in 10dB for level (vertical) and 2.3MHz for frequency (horizontal). Australia's electronics magazine January 2023  27 Fig.2: both the preamp and power amp chips are three-terminal devices with an input pin, an output pin and ground pin. They are fed with supply current through the output pins, via inductors which present a high impedance at the signal frequency, so they don’t attenuate the signals. regulator. This then feeds a KB042 power amplifier via a 470pF coupling capacitor. The values of both coupling capacitors (both at the input and between the preamp and amplifier) are relatively high. This is so it can accommodate frequencies down to 100kHz. The manufacturer’s recommended value for the SBB2089Z is around 8.2nF for its specified frequency range, from 50MHz to 850MHz. The data sheet for this device indicates that its gain is relatively flat over that frequency range, but I found that the gain was higher from 1MHz to about 50MHz, and lower at 930MHz than I expected. The SBB2089Z draws around 110mA from the 78L05 – more than its recommended maximum, but within its capabilities at any realistic device temperature. There is no data sheet available for the KB042 power amplifier, but we think it operates similarly to the ERA-2SM+. It is powered directly from the 12V supply via another AC-­ blocking inductor. Different versions of this board available online seem to use either 33µH or 68µH inductors in series with SMD ferrite beads. Presumably, those inductance values are not critical to its performance. Note that the +12V supply directly feeds the KB042; therefore, the applied voltage should not exceed 15V; otherwise, the KB042 may blow. It’s best to use a 12V DC regulated supply but you could probably get away with a 12V lead-acid battery. One slightly unusual feature of the circuit is the bias network from the output of the 78L05 to the input of the KB042. The 10kW/5.1kW divider generates a DC bias of about 1.7V, which is applied to the signal via a 100W resistor. Consider that many RF power amplifiers are based on Darlington transistors, and 1.7V is a little more than two base-emitter junction Fig.3: the output response plot for the 2W RF amp over 0-350MHz with a swept input signal at -30dBm. The series of dips are caused by standing waves in the measurement and not the amplifier itself. 28 Silicon Chip The 50W dummy load that was used to test the amplifier module. forward bias voltages. Either this is needed to bias the KB042’s internal transistors into their operating range or (more likely, we think) the intent is to supply additional base current to allow the amplifier to deliver more power before it runs into clipping. The PCB design is pretty much according to the manufacturer’s recommendation for the first IC and the KB042 IC is tacked on to provide the specified output power. Testing I tested this module feeding into a 50W dummy load so that I could make the measurements in dBm. The amplifier has a high internal impedance, so a suitable resistor must be used that can handle the power levels with minimal reactive impedance to maintain a constant resistance at high radio frequencies. I used a specialised resistor (EMC 5307ALN) which can handle 125W from DC-2GHz. I got this from eBay Fig.4: a similar plot to Fig.3 but for a higher frequency range, 240-960MHz. The output level starts to fall off near 400MHz. There are even more dips this time; again, they are artefacts of the measurement system, not the amp. Australia's electronics magazine siliconchip.com.au resulted in somewhat lower output levels as expected. The 2W transmitter “that changed the world” Fig.5: the RF Power Meter is a simple device but very useful nonetheless. Again, we had to trace out the circuit. We couldn’t get to the range switching components to see how they were configured, so they’re not shown here. for $23, including postage, and it performed well, hardly getting warm at 2W. I mounted this resistor inside a 4 x 4cm aluminium housing weighing 56g, although it can be bolted to a larger heatsink or fan for higher power handling. Frequency response Fig.3 shows the amp’s frequency response over 0-350MHz, while Fig.4 is a response plot over 240-960MHz. I did it in two plots since you can see the details better this way. Both were made using a tinySA spectrum analyser with the sweep generator set at -30dBm. This resulted in an output from the amp ranging from -10dBm to +4.2dBm (analyser set to max hold), which is close to the specified performance. The dips in the graph are mainly due to standing waves in the load resistor and the cable to the spectrum analyser. The actual response of the amp would be somewhat flatter than this. Surecom SW-11 RF Wattmeter I also purchased this RF wattmeter to check out the RF amp performance. Measuring just 85x50x55mm, it can handle up to 100W of RF power and has a low range of 10W. Also, it can be switched to SWR (standing wave ratio) mode for analysing antenna characteristics. Its circuit is shown in Fig.5. It is a passive device with fairly straightforward circuitry, which should be capable of accurately measuring RF power up to approximately 400MHz. However, Fig.5 does not show the range switching circuitry built into the unit because it’s virtually impossible to access without destroying the thing. I fed the output of the 2W RF amplifier to the SW-11 together with the EMC 5307ALN 50W RF dummy load. It indicated a maximum output of 2W for an input signal of -30dBm at 20MHz. Higher input levels did not increase power output, and higher frequencies We mentioned the first artificial satellite, Sputnik (1957), in the introduction because it also featured a 2W RF transmitter. For those who are interested, Fig.6 shows its circuit diagram. This was only made public in 2016 by a Russian leaker! It produced a CW signal at 20MHz using a modified Colpitts oscillator and a push-pull output stage. The valves are sub-miniature types, powered by a bank of batteries. A second transmitter was also fitted, which operated at 40MHz. The one-second “beep” was supposedly controlled by an external vibrator (likely via the “controller” input at upper right). Warning We envisage readers possibly using this amplifier module to boost received signal levels within their homes or offices. Radio amateurs could potentially use it as part of a transmitting rig. But keep in mind that unless you have some sort of radio license, transmitting at just about any frequency at 2W is illegal in Australia and New Zealand. It probably isn’t a good idea to connect an antenna to this device’s output unless you know it is legal and safe to do so. We are only aware of the exceptions outside this device’s frequency range, in the 2.4GHz & 5.8GHz bands, and only for frequency-­hopping or digSC itally modulated transmitters. Fig.6: the valve-based 2W 20MHz transmitter that flew on Sputnik, the first artificial satellite, in 1957. Compare this to Fig.2; it’s significantly more complex (although it does incorporate an oscillator) and no doubt would have cost the equivalent of many thousands of today’s dollars (rubles?). Australia's electronics magazine January 2023  29 We’ve published numerous LC meters that can measure inductance and capacitance, but you might need to know the quality factor (Q) of an inductor, not just its inductance. This Q Meter uses a straightforward circuit to measure the Q factor over a wide range, up to values of about 200. Q Meter T he history of Q Meters goes back to 1934, when Boonton developed the first Q Meter. The Q Meter is a somewhat neglected piece of test equipment these days. Hewlett Packard bought Boonton in 1959 and produced revised versions of their Q Meter. Does anyone still manufacture them? It seems not. You can find a few on the second-hand market; they fetch prices up to $3000. The HP 4342-A is an excellent unit and is a more modern version of the original Boonton design. My Q Meter design can’t come near the quality or accuracy of that HP unit. It is not designed as a laboratory instrument but will give Q measurements up to a value of about 200 with an accuracy of about 10%. Q&A So, what is Q, and why do we need to measure it? It is a measure of the dissipative characteristic of an inductor. High-Q inductors have low dissipation and are used to make finely-tuned, narrow-band circuits. Low-Q inductors have higher dissipation, resulting Fig.1: a real inductor does not just have pure inductance; it also has parasitic series resistance (Rl) and parallel capacitance (Cp). 30 Silicon Chip in wideband performance. It can be expressed as: Q = 2π × (Epk ÷ Edis) Where Epk is the peak energy stored in the inductor and Edis is the energy dissipated during each cycle. Let’s consider two passive components, an inductor and a capacitor. The reactance of the inductor is Xl = +jωL. Here, j = √-1, Xl is in ohms and ω = 2πf (f is the frequency). For example, a 10µH coil at 10MHz will have a reactance of +j628W. A capacitor has a reactance of the opposite polarity, ie, Xc = 1 ÷ −jωC. To resonate at 10MHz, the capacitor needs a reactance of −j628W, which equates to 25.3pF. But inductors and capacitors are not perfect. A practical inductor can be approximated as an ideal inductor with a series resistor. The coil By Charles Kosina winding will also add a small capacitance across the inductor, as shown in Fig.1. The capacitor is also not perfect but generally has a much smaller inherent resistance, so for this calculation, we can assume it is. The inductor’s Q is defined as Q = Xl ÷ Rl and the -3dB bandwidth of such a tuned circuit is BW = f ÷ Q. So, a tuned circuit with a 10µH coil and a Q of 100 would have a -3dB bandwidth of 100kHz at 10MHz. The Q is important if you’re trying to design something like a bandpass or notch filter. In Fig.2, we have a series tuned circuit fed by a variable frequency source with frequency f, voltage VS and source resistance Rs. At resonance, Xl = −Xc; in effect, a short circuit, so the load on the generator is Rs + Rl. By having a generator with source resistance Rs much lower than Rl, the Fig.2: we can calculate an unknown inductor’s Q (quality factor) using this circuit. It is connected in a series-tuned circuit with a capacitance, and that circuit is excited by a sinewave from a signal generator via a known source resistance. Measuring the input and output AC voltages and calculating their ratios allows us to compute the inductor Q, assuming the Q of the capacitance is high. Australia's electronics magazine siliconchip.com.au voltage measured at Vin will be close enough to VS. The current through the circuit will be Is = VS ÷ Rl. Therefore the voltage at the junction of the inductor and capacitor is Vout = Xl × Is. By measuring Vin and Vout, the Q can be calculated as Ql = Vout ÷ Vin. That assumes that the capacitance has been adjusted to achieve peak resonance with the inductance, ie, Xl = −Xc. That can be done by sweeping the capacitance until the peak Vout voltage is reached. The first design challenge is to have an extremely low generator source resistance. If we have a 10µH coil with a Q of 100, at 5MHz, the effective Rl is 3.14W (314W ÷ 100). If our source resistance is 0.1W, that will give an error of about 1%. But at 1MHz, Rl becomes 0.628W, and this error blows out to 15%. So using a higher frequency will generally result in a more accurate Q measurement. Low source resistance Boonton solved the source resistance problem by having the generator heat a thermocouple using a wire with a very low resistance, as shown in Fig.3. The voltage generated by this thermocouple was measured by a DC meter which indicated how much current was applied to a 0.02W resistor in series with the external inductor. I have a Meguro MQ-160 Q Meter, essentially a 1968 version of the original Boonton 260-A design, using such a thermocouple and resistor. No transistors in this one; it’s all valves! But for our design, a thermocouple is not practical. The HP design eliminated the thermocouple and instead used a step-down transformer. The transformer is fed by a low impedance source, as shown in Fig.4. If our source resistance is 50W, like siliconchip.com.au the output of a typical signal generator, and the turns ratio is 50:1, the effective source resistance is 0.02W (50W ÷ 502), exactly what we want. Unfortunately, it is not so simple as it implies a perfect transformer. Losses in the transformer core plus winding resistance conspire against us and push up the source resistance value. We can improve this by feeding the transformer’s primary from the output of an op amp, which has an impedance close to zero. In this case, a turns ratio of 10:1 is adequate as the resultant 100:1 impedance ratio will give an acceptable load to the op amp. This is what I have used in my design. The transformer is a ferrite toroid of 12mm outside diameter. The primary is 10 turns of enamelled wire, while the ‘one turn’ secondary is a 12mm-long tapped brass spacer through the centre of the toroid. The effective RF resistance of this spacer is extremely low, and the source resistance is then mainly a function of the ferrite material and the primary winding resistance. Table 1 – frequency versus signal source impedance/spacer Frequency Brass Steel 0.1-1MHz ~0.00W 0.02W 2MHz not tested 0.016W 5MHz 0.03W 0.13W 10MHz 0.07W 0.20W 15MHz 0.09W not tested 20MHz 0.15W 0.22W 25MHz 0.10W 0.17W The full circuit of my Q Meter is shown in Fig.5. We require a signal generator with an output of about 0dBm (1mW into 50W or 225mV RMS). You can use just about any RF signal generator. There didn’t seem to be much point in building the generator into the Q Meter since, if you’re building a Q Meter, you likely already have an RF signal generator. I’m using my AM/FM DDS Signal Generator that was described in the May 2022 issue (siliconchip.au/Article/15306). The generator feeds a sinewave into CON1, which is boosted by op amp IC2a. This is a critical item in the design, as it needs a high gain bandwidth (GBW) and slew rate, as well as the capability to drive a low impedance. The Texas Instruments OPA2677 has a GBW of 200MHz, a slew rate of 1800V/µs and can drive a 25W load, which gives us enough output voltage swing up to 25MHz. The toroidal transformer core is a critical part of the design. I tested a Fair-rite 5943000301 core which is readily available from several suppliers. I wound it with 10 turns of 0.3mm diameter enamelled copper wire. A heavier gauge (up to about 0.4mm) may be slightly better, but there has to be enough room in the centre for the spacer to pass through. I then calculated the source impedance by measuring the no-load output voltage followed by a 1W load. I did this for several frequencies, and the results are shown in Table 1. Below 1MHz, there was no measurable difference between no load and a 1W load, so the source impedance must be well below 0.01W. Core losses likely account for the higher source resistance as frequency increases, but the results are quite adequate. Brass spacers are recommended (and will be supplied in kits) due to their superior performance here, at least for the one through the toroid. Fig.3: one method of measuring Q involves current sensing via monitoring the temperature of resistance wire. It has the advantage of keeping the source impedance low, and no complicated shunt sensing circuitry is required. Fig.4: we need an RF signal source with an extremely low but known source resistance for our Q Meter. Since that is difficult to achieve by itself, feeding the signal through a low-loss stepdown transformer greatly reduces the actual source impedance, as seen by the load. Circuit description Australia's electronics magazine January 2023  31 Fig.5: eight relays switch capacitors in parallel to vary the resonant circuit capacitance from around 40pF (the stray capacitance) to 295pF. The signal from the RF generator is amplified by op amp IC2a and fed through step-down transformer T1 to the resonant circuit. The input signal level is monitored via precision rectifier IC2b while the output signal is rectified using D3 and amplified by IC3a. 32 Silicon Chip Australia's electronics magazine siliconchip.com.au The DC output of op amp IC2a is zero or very close to zero, so why do we need a 10µF capacitor in series with the transformer? As the DC resistance of the primary is a fraction of an ohm, the slightest offset voltage in the op amp output could send a high direct current through the toroidal transformer primary and overload the output. That possibility is eliminated with AC coupling. The tuning capacitor is another essential part. My Meguro has a 22-480pF variable capacitor, typical of the tuning capacitors used in valve radios. They are available on sites like eBay, but they are very large and expensive. The only easy-to-get variable capacitor is the sort with a plastic dielectric for AM radios. But once you get above the broadcast band, they are very lossy, with a poor Q, and entirely unsuitable. So instead, I designed a ‘digital capacitor’ with eight relays switching in capacitors with values in a binary sequence of 1, 2, 4, …..128pF. As these are not standard values, some are made up of two capacitors in parallel. For example, 32pF is 22pF in parallel with 10pF. Combining these allows the capacitance to be adjusted in 1pF steps from 0pF to 255pF. The measured stray capacitance due to the tracks, relays etc amounts to 40pF, so the tuning range is 40-295pF. My LC meter shows that it tracks reasonably accurately. All capacitors are not created equal, so I have used somewhat expensive high-Q RF capacitors, available from element14, Mouser, Digi-Key etc. Not all these capacitors have a close tolerance; some are ±2%, which detracts from the accuracy. So it isn’t a ‘real’ variable capacitor but it has the advantage of not needing a calibrated dial and a slow-motion vernier adjustment. Rather than measuring the very low voltage on the secondary side of the transformer, it is more practical to measure the primary side, and for the Q calculation, divide this by 10. I verified this assumption by checking that the voltage ratio corresponded to the turns ratio within measurement accuracy from 100kHz to 25MHz. A precision half-wave rectifier is formed using op amp IC2b in the classic configuration. By placing the rectifier diodes in the negative feedback network of the op amp, their forward siliconchip.com.au Australia's electronics magazine January 2023  33 rectifier feeding a high-­ impedance (10MW/1.5MW) voltage divider. The voltage drop in the diode only introduces a small error in the measurement. The voltage at the junction of this divider is buffered and amplified by IC3a, a TSV912 op amp with an extremely high input impedance – the input bias current is typically 1pA. Switch S1 changes the gain of this op amp for the low and high Q ranges, with the low range giving 8.3 times gain for Q values of up to 100. On the high range, the gain of this stage drops to 1.7 times. Power supply & control Fig.6: the PCB uses mostly SMD components for compactness, although none are particularly small. The orientations of the following components are important: all relays, ICs and diodes, plus the Arduino Nano. ZD1, IC4, CON3 and associated parts form the optional debugging interface. voltages are effectively divided by the (very high) open-loop gain of the op amp. On positive excursions of the output pin of IC2b, the 330nF capacitor at TP3 is charged up through diode D1. The extra diode, D2, is needed as without it, negative excursions would saturate the op amp and lead to slow recovery, limiting its frequency range. Both diodes are 1N5711 types for fast switching. 34 Silicon Chip The output of IC2b is amplified by IC3b, and the resulting filtered DC voltage at TP4 is about 1.9V. The secondary voltage of the transformer is typically 200mV peak-topeak or about 70mV RMS. With a Q of 100, the voltage output at the junction of the inductor and tuning capacitor would be 20V peak-to-peak or 7V RMS. That is not a suitable voltage to apply to the input of an op amp! So I used schottky diode D3 as a half-wave Australia's electronics magazine A MAX660 switched capacitor voltage inverter (IC1) provides a nominally −5V supply to the OPA2677 (IC2). This is needed for proper operation of the half-wave precision rectifier, IC2b, as the voltage at its input can swing below ground. The MAX660 is not a perfect voltage inverter, and with the current drain of the OPA2677, its output is about −3.6V, but that is adequate. The rest of the circuit operates from a regulated +5V DC fed in externally, eg, from a USB supply. An Arduino Nano module is used as the controller. This is a readily-­ available part from many suppliers at a reasonable price. Two analog inputs are used for measuring the voltages, eight digital outputs switch relays, the two I2C serial lines drive the OLED, and there are inputs for the control rotary encoder and LOW/HIGH switch sensing. The rotary encoder (EN1) is used to adjust the ‘digital capacitor’ value; its integral pushbutton switch toggles between steps of 1pF and 10pF. As usual with my designs, I have added a simplified RS-232 interface using hex schmitt-trigger inverter IC4 to aid code debugging. IC4, ZD1 and the two associated resistors can be left out unless you want to use the debugging interface. Eight 2N7002 N-channel Mosfets (Q1-Q8) drive the relay coils, while eight diodes across the relay coils (D6D13) suppress switching transients. The resonant frequency tuning is done by selecting an appropriate frequency from the external signal generator and adjusting the variable capacitance value. Ideally, the peaking should be done with an analog meter, siliconchip.com.au but I have provided an onboard LED, LED1, the brightness of which depends on the Vout voltage. It’s simple enough to adjust the capacitance to achieve maximum brightness. The third line of the OLED also shows the output voltage of IC3a, which can be used to accurately achieve resonance too. Connector CON5 drives an optional external 0-5V moving coil meter. You can add such a meter if a larger-­thanspecified enclosure is used to house the PCB. The power supply is a standard 5V USB charger. I have not included reverse polarity protection, but an offboard 1A schottky diode (eg, 1N5819) could be added in series if desired. (0.3in) pitch, then the rotary encoder, switch and LED. Use a 5mm plastic spacer for the LED, so it is flush with the back of the front panel. Wind ten turns of the specified enamelled copper wire onto the toroidal core, taking care that the turns are equally spaced around the circumference, to the extent possible, and the ends line up with the two pads marked PRIM on the PCB. Carefully attach the toroid so that it is centred on the mounting hole. Attaching the spacer to the board makes that easier. It may be anchored in place by an insulated wire across the two pads on the opposite side. It is not a shorted turn as only one side of this wire is connected to the ground plane. I recommend fitting socket strips for mounting the Arduino Nano module as they make replacing a faulty module easy (I have blown up a couple in the past!). The OLED screen also plugs into a 4-pin socket strip and is held in place by two 15mm-long M2 or M2.5 Construction The construction uses two PCBs (see Figs.6 & 7). The main one has all the electronics while the other has the screw terminals for the DUT and external capacitor. It is also used as a front panel and has a rectangular cutout for the OLED, holes for the controls and lettering. It is designed to fit in a RITEC 125 × 85 × 55mm enclosure, sold by Altronics as H0324. The top board/front panel is 98 × 76mm and fits snugly into the recess in the clear lid of the enclosure. This board could be used as a template for accurately drilling the holes in the clear lid. But other enclosures may be used as long as they have the same or slightly greater dimensions as the H0324. For those wishing to add the 0-5V moving coil meter, this requires an additional width of 45mm. A suitable 158 × 90 × 60mm enclosure is available from AliExpress suppliers at a reasonable price, but be aware that delivery can take quite a few weeks. Most components on the PCB are surface-mount types, but there are no fine-pitch ones, which simplifies construction. Solder the four SOIC chips first, then all the passives, which are mostly M2012/0805 size (2.0 × 1.2mm). The relays take a bit of care to ensure they are square on the board so that it looks neat. On the opposite side of the board are eight 1N4148 equivalent diodes; ensure they are installed with the correct polarity, with the cathode stripes to the side marked “K”. After the SMDs, add the throughhole diodes, which have a 7.6mm siliconchip.com.au Only the Arduino Nano, headers and eight diodes are on the underside of the Q Meter PCB. Note how the windings for T1 are spaced evenly around it. Australia's electronics magazine January 2023  35 Almost all the parts mount on the main PCB. The only chassismounting components are the DC input socket and optional power switch. screws through 8mm untapped spacers. Carefully slide off the plastic strip on the four pins of the OLED so that it sits lower. The board must be thoroughly cleaned with circuit board cleaner. There are high impedances throughout the circuit, and leakage through flux residue would affect its operation. So you must remove that residue. Testing Once the board has been fully assembled, cleaned and inspected, but before it is mounted in the case, attach the four 12mm spacers but not the front panel board, and connect the 5V supply. The OLED should show an initial message with the firmware version number. Using a coax cable, feed in a sinewave from a signal generator at about 1MHz. An oscilloscope probe on TP1 should show a clean sinewave, with an output of about 2V peak-to-peak. If the output of the signal generator is too high, you will get flattening on the negative half cycle. In that case, back off the level for a clean sinewave. Transfer the ‘scope probe to the top of the spacer that passes through the toroid, and the voltage should be onetenth of that measured at TP1. Measure TP4 using a DC voltmeter; you should get a reading of about 2V. Note that these values will depend on the output of the signal generator and could vary. Rotate the encoder and note that the capacitance value varies by 1pF per detent. Depending on the encoder, it might go backwards. If so, plug a 36 Silicon Chip jumper on the Arduino Nano’s programming header between pins 4 and 6; that will correct the direction. Push down the knob to change the resolution, and the capacitance should then change by 10pF per detent. By winding it fully clockwise, the maximum indicated capacitance should show as 295pF on the bottom line of the OLED, with the minimum being 40pF. Connect a 10µH moulded inductor between the two “L” spacers, using 3mm machine screws to hold it in place. Adjust the capacitance to 100pF, switch to LOW Q mode and adjust the signal generator frequency to about 5.5MHz. The LED should light up; tune the capacitance for maximum brightness. The second line of the OLED will then most likely display “TOO HIGH”. Switch to HIGH Q mode, which will dim the LED, and re-tune for maximum brightness. Depending on the inductor, a typical Q reading will be about 120. If you get a sensible reading and can peak the LED brightness by varying the capacitance, your Q Meter is most likely functioning correctly, so it can be finished. The front panel is mounted on the front of the case, and the main PCB may now be attached by the four spacers using four 8mm M3 machine screws. To improve the appearance, use black screws or spray the heads flat black. Note that the binding posts must make electrical contact with the bare pads on the front panel PCB; attach them with the supplied nuts and make sure they are making good contact. The tapped spacers connecting the two boards must also make good electrical contact at both ends. Using it The operation of the Q meter requires some initial measurements and calculations. We need to know at least the approximate inductance of the DUT. I use my LC Meter for measuring this, as described in the Fig.7: the circuitry on the front panel PCB just consists of one large track connecting the two red terminals and smaller tracks connecting the upper screws to their adjacent binding posts. It also has holes and labels for the controls and screen. Australia's electronics magazine siliconchip.com.au November 2022 issue of Silicon Chip (siliconchip.com.au/Article/15543). With the inductance known or guessed, we need to determine the frequency at which to measure the Q. That will be influenced by the inductor value and the frequency at which you want to use the inductor. Once you’ve selected a frequency, plug the values into the formula: Parts List – Q Meter Accuracy 1 RF signal generator (see May 2022; siliconchip.au/Article/15306) ● 1 RITEC 125 × 85 × 55mm plastic enclosure [Altronics H0324] ● 1 double-sided PCB coded CSE220806B, 99 × 79mm 1 double-sided PCB coded CSE220807A, 98 × 76mm, black solder mask 1 chassis-mounting SPST toggle switch with solder tabs (S1) 1 0-5V analog meter (optional) ● 1 Arduino Nano (MOD1) 1 0.96in OLED display module with I2C interface and SSD1306 controller (MOD2) [Silicon Chip SC6176 (cyan)] 8 G6K-2F-Y SPDT SMD relays (RLY1-RLY8) 1 rotary encoder with integral pushbutton (EN1) 1 knob to suit EN1 1 Fair-rite 5943000301 ferrite toroidal core, 12mm OD, 8mm ID, 5mm thick (T1) 1 30cm length of 0.25-0.4mm diameter enamelled copper wire (T1) 1 SMA edge connector (CON1) 2 2-pin polarised headers (CON2, CON5) 1 3-pin polarised header (CON3) ● ♦ 1 2.1mm or 2.5mm inner diameter chassis-mount jack socket (CON4) ● 2 red 4mm chassis-mounting banana socket/binding posts 2 black 4mm chassis-mounting banana socket/binding posts 4 M3 × 12mm brass spacers 4 M3 × 5mm nickel-plated panhead machine screws 4 M3 × 8mm nickel-plated panhead machine screws 2 M2 × 16mm machine screws and nuts 2 8mm-long untapped plastic spacers 1 5mm-long plastic LED spacer 1 20cm length of light-duty figure-8 hookup wire ● Semiconductors 1 MAX660M switched capacitor voltage inverter, SOIC-8 (IC1) 1 OPA2677 dual ultra-high GBW op amp, SOIC-8 (IC2) 1 TSV912 dual high input impedance op amp, SOIC-8 (IC3) 1 74HC14 hex inverter, SOIC-14 (IC4) ♦ 1 3mm red diffused lens LED (LED1) 8 2N7002 Mosfets, SOT-23 (Q1-Q8) 1 4.7V 400mW axial zener diode (ZD1) ● ♦ 3 1N5711 axial schottky diodes (D1-D3) 8 LL4148 75V 200mA diodes, SOD-80 (D6-D13) This meter is certainly not as accurate as the HP4342-A meter mentioned earlier. Without any standard coils of known Q, it is difficult to determine the true accuracy. But even the HP4342-A does not claim any better accuracy than ±7% for frequencies below 30MHz, and considerably worse for higher frequencies (see the PDF at siliconchip.au/link/abgn). I compared my results with the Meguro meter, but being over 50 years old, it is hardly to be trusted! Still, measurements of the same coil with the Meguro and my meter were genSC erally within 10%. Capacitors (all SMD M2012/0805 X5R or X7R) 3 10μF 16V 3 330nF 50V 10 100nF 50V RF capacitors (all ±2% 200V SMD M2012 or M1608 C0G/NP0 unless noted) 2 100pF 50V 1 10pF 1 56pF 2 8.2pF 1 27pF 1 3.9pF ±0.1pF 1 22pF 1 2.2pF ±0.1pF 1 15pF 2 1.0pF ±0.1pF Resistors (all SMD M2012/0805 1%) 1 10MW 3 3.3kW 1 1.5MW 1 1.2kW 1 12kW 1 1kW 3 18kW 1 270W 3 10kW 1 51W 4 4.7kW C = 25330 ÷ (2 × f × L) Where C is in pF, f is in MHz and L is in µH. If you get a value of C below 40pF, select a lower frequency and redo the calculation; if you get a value above 295pF, choose a higher frequency. Repeat until your calculated capacitance is in the range of 40-295pF. Set the capacitance to that value and adjust the frequency from the signal generator, or the capacitance, for resonance. The resulting Q will be shown on the second line of the OLED. If the switch is set to LOW and the Q exceeds 100, the second line will show “TOO HIGH”. In that case, switch to the HIGH position. I find that it is better to start with the switch set to LOW as it is easier to figure out if you are close to resonance. The “C” terminals allow a capacitor to be placed in parallel with the internal capacitance in case you can’t achieve resonance at a sensible frequency with the available range. So that it doesn’t detract from the Q, it should be a high-quality RF capacitor. ♦ optional components only required for debugging interface KIT (SC6585) – $100 + P&P: includes everything in the parts list that isn’t marked with a ● PCBs are also available separately siliconchip.com.au ● Kit – a kit is available with all the above parts except those marked with a red circle. Its catalog code is SC6585 and it costs $100 + P&P ($90 + P&P for active subscribers). Note that the Arduino Nano is supplied unprogrammed. The PCBs are also available separately. Australia's electronics magazine January 2023  37 IDEAL FOR STUDENT OR HOBBYIST ON A BUDGET • DATA HOLD • SQUARE WAVE OUTPUT • BACKLIGHT • AUDIBLE CONTINUITY Don't pay 2-3 times as much for similar brand name models when you don't have to. ONLY 1995 $ QM1517 HANDY FOR THE HOME TOOLBOX AUTORANGING FOR EASE OF USE • TEMPERATURE • DUTY CYCLE • NON-CONTACT VOLTAGE ONLY 5995 $ QM1323 FOR THE TECH'S TOOLBOX FOR A STUDENT OR APPRENTICE ONLY 49 $ 95 QM1321 ACCURACY OF TRUE RMS MEASUREMENT • CAPACITANCE & FREQUENCY • TRANSISTOR HFE TEST • DIODE & AUDIBLE CONTINUITY • NON-CONTACT VOLTAGE ONLY 3295 $ QM1020 EASY TO READ MEASUREMENTS • MIRROR BACKED SCALE • TRANSISTOR LEAKAGE TEST • AUDIBLE CONTINUITY Test & Measure with our GREAT RANGE of multimeters at the BEST VALUE, to suit hobbyists and professionals alike. Explore our wide range of multimeters, in stock on our website, or at over 110 stores or 130 resellers nationwide. www.jaycar.com.au/multimeters 1800 022 888 LARGE BACKLIT DISPLAY AND IP67 WATERPROOF RATED • TRUE RMS • CAPACITANCE • FREQUENCY • RELATIVE MEASUREMENT TAKE EASY ENVIRONMENTAL MEASUREMENTS • MULTIMETER FUNCTIONS • SOUND LEVEL • LIGHT LEVEL • INDOOR TEMP • HUMIDITY ONLY 99 $ 95 QM1549 WIRELESS BLUETOOTH® FEATURE FOR DATA LOGGING • AUTORANGING • TRUE RMS • 6000 COUNT • IP67 WATERPROOF ONLY 149 $ QM1594 ONLY 169 $ QM1578 Use this colour coded selection guide to pick the meter that best suits your needs. GREEN labelled product suit hobbyists and those on a budget. BLUE suit makers familiar with multimeters and want more features. For all the bells and whistles and the highest ratings, choose from the ORANGE professional range. ENTRY LEVEL * QM1500 QM1517 QM1527 MID LEVEL QM1529 QM1321 QM1020 QM1446 Display (Count) 2000 2000 2000 2000 4000 Analogue Security Category Cat II 500V Cat III 600V Cat III 500V Cat III 600V Cat III 1000V Cat II 1000V • • Autorange True RMS PROFESSIONAL QM1323 QM1552 2000 4000 2000 4000 4000 2000 4000 6000 4000 Cat III 600V Cat III 600V Cat IV 600V Cat III 600V Cat IV 600V Cat III 600V Cat IV 600V Cat IV 600V Cat III 1000V • • • • • • QM1551 QM1549 • • • • • XC5078 QM1594 QM1578 • Voltage 1000VDC/ 750VAC 500V AC/DC 500V AC/DC 600V AC/DC 1000VDC/ 750VAC 1000V AC/DC 1000VDC/ 700VAC 600V AC/DC 1000VDC/ 750VAC 600V AC/DC 1000V AC/DC 600V AC/DC 600V AC/DC 1000V AC/DC Current 10A DC 10A DC 10A DC 10A AC/DC 10A AC/DC 10A DC 10A AC/DC 10A AC/DC 10A AC/DC 10A AC/DC 10A AC/DC 200mA AC/DC 10A AC/DC 10A AC/DC Resistance 2MΩ 2MΩ 2MΩ 20MΩ 40MΩ 20MΩ 20MΩ Capacitance 100mF Frequency 10MHz Temperature Duty Cycle 20MΩ 40MΩ 200MΩ 40MΩ 40MΩ 40MΩ 60MΩ 100μF 100µF 100mF 100µF 100µF 100µF 6000µF 10MHz 10MHz 10MHz 10MHz 10MHz 10MHz 10kHz 1000°C 760°C 1000°C 760°C 750°C 760°C • Continuinty • • • • • • Relative Min/Max/Hold • Non Contact Voltage • • • $24.95 $29.95 $49.95 • • • • • • • • • • Max Hold • • • $59.95 $69.95 $69.95 IP Rated Price • • • • $19.95 *Lifetime warranty excluded on models: QM1500/QM1517/QM1527 $32.95 $54.95 4000MΩ • • IP67 $9.95 1000VDC/ 750VAC • • Max Hold • QM1493 $99.95 IP67 $89.95 $149 $169 $249 Prices correct at time of publication but are subject to change. Jaycar reserves the right to change prices if and when required. miniITX PCs desktop power in a small package Portable computers clearly have advantages and are invaluable for people who must work while moving from place to place. I prefer desktop computers for their higher performance, lower overall cost, ability to drive many monitors and expandability, but they can take up a lot of space. Enter the Mini-ITX PC: the power of a big desktop in a much smaller package. M ini-ITX PCs have been around for some time; they are popular as “home theatre PCs”, used for playing videos and music in an entertainment centre. I had a large tower computer under a desk because it wouldn’t fit on top, taking up valuable leg room and making the desk feel cramped. So I decided to build a compact PC to replace the big tower. I didn’t want to compromise on performance. After some research, I realised that just because the computer would be smaller, that didn’t mean it needed to be slow or lacking in RAM, storage or graphics processing power. There’s even room for water cooling if you go about it the right way. In building it, I learned many of the tricks to constructing a good Mini-ITX PC. I am writing this article on that 40 Silicon Chip computer which is not only smaller and faster, it’s also very quiet and reliable. After that success, I built several other Mini-ITX computers, including one to play music and videos that is dead silent, so it doesn’t compromise music listening enjoyment. We also use one in our warehouse. This article covers building one of these mini computers from scratch, explaining your options in choosing parts and some things to watch out for. I was forced to change plans when parts I bought wouldn’t fit; that is something you’ll want to avoid. ATX vs Mini-ITX Full-size desktop computers by Nicholas Vinen Australia's electronics magazine generally use ATX motherboards and power supplies. These are based on a standard published in 1995 by Intel and updated several times since. An ATX motherboard is 305 × 244mm, and an ATX power supply is generally 150 × 86 × 140mm, although the last dimension can vary up to 230mm (usually for very high power units). Those dimensions dictate the minimum practical size of a case. ATX cases are typically around 500mm tall, 500mm deep and 210mm wide, although they can be significantly larger or a little more compact. That’s a typical volume of about 50-60 litres. Photo 1 shows a direct size comparison of a typical mid-size ATX case (left) and two of the Mini-ITX cases I used (middle & right). A Mini-ITX motherboard is siliconchip.com.au considerably smaller than ATX at 170 × 170mm, and they are often teamed up with SFX power supplies that measure 100 × 125 × 63.5mm – see Photo 2. The Corsair SFX power supply shown there works really well (it’s also available in a 750W version), although it costs significantly more than the more-powerful ATX supply on the left. Mini-ITX case sizes and volume vary dramatically, from just a couple of litres at the low end (about the size of a thick notebook) to almost as large as a mid-tower ATX case. The cases I chose, shown in Photo 1, measure 325 × 166 × 310mm and 16.7 litres. I feel that is about the sweet spot, although you can go smaller if you want to. Interestingly, case design can have more to do with what fits than the size. For example, there’s nowhere to mount a water-cooling radiator in the larger case on the left in Photo 1, while the smaller ones fit a 120 × 240mm radiator nicely. Photo 1: two Fractal Design Era Mini-ITX cases (right) and a low-cost Deep Cool Tesseract ATX case on the left. I chose the Era cases mainly for their looks, as one would be visible in my living room. I discovered they are pretty good to work with, although there are more functional Mini-ITX cases for the money. Mini-ITX limitations So, what do you give up with Mini-ITX+SFX compared to ATX? Not a lot. You usually only get two RAM slots on an ITX motherboard compared to four with ATX (see Photo 3), although you can install 64GB of DDR4 RAM or 128GB of DDR5. You also get fewer expansion slots but these days, with at least two NVME slots on most boards plus high-speed onboard USB and networking, that won’t matter to most users. You also get less space in the case, although you can still usually fit several SSDs (solid-state drives) and even a traditional hard drive or two if you need them. Probably the most significant limitations are with the graphics processor unit (GPU). While you can build a Mini-ITX PC with a high-end GPU, it is not trivial to fit anything more potent than a mid-range GPU like an Nvidia RTX 3060 or an AMD RX 6600 XT. You might also have trouble powering the beefiest GPUs, as there are few SFX power supplies above 750W. You could choose a Mini-ITX case that supports standard ATX power supplies, in which case you could get a 1000W+ supply. But dumping that much heat into something the size of a shoe box might not be a great idea! The cases I used from the Fractal Design Era series (www.fractal-design. siliconchip.com.au Photo 2: a 650W ATX power supply (left) and 600W SFX power supply (right). ATX supplies up to about 1kW are available in the size shown here (or more, but they are physically longer), while SFX supplies usually top out around 750W. That’s still plenty for most builds in a small case, though! Photo 3: the Mini-ITX motherboard I used. Note how packed it is with components and connectors! The large space at the top is for the CPU and cooler. Below that is the chipset plus NVME heatsink/ fan and the expansion slot, while the RAM slots, SATA and power connectors are on the right. The I/O plate dominates the left side, the same size as for an ATX board. Australia's electronics magazine January 2023  41 com/products/cases/era/era-itx/) are currently available from various suppliers for around $250-300 (some on sale for $120 at the time of writing). Many other similar cases are available, and most of my advice applies to them too. I will mention another great case I have experience with towards the end of the article. Choosing a Mini-ITX case I chose the Era for a few reasons. One was the support for both SFX and ATX power supplies, although I learned while building them that you have much more room to breathe (and install hard drives) if you take the SFX option. Other cases that offer that choice will be similarly squashed if you go with the ATX option. As mentioned earlier, the biggest problem with SFX power supplies is the cost; you can get some great power supplies, so you aren’t compromising much in terms of performance as long as 750W is enough. The first thing you will need to do if you want to build a Mini-ITX PC is decide on what parts you want to put into it and start shopping for cases that will fit them all. The cases vary so much in size and design that finding the right one will take a while. For a start, if you want to water cool your CPU or GPU, you’ll need a case that fits a radiator or two. The largest radiator that will fit in most MiniITX cases is 280 × 140mm, although a maximum of 240 × 120mm is more common, and some will only fit 120 × 120mm or none at all. With cases that can fit a radiator, you almost certainly give up some other capability if you install one. For example, the radiator might limit the maximum length of the GPU or the ability to use an ATX power supply. So check all that carefully. I have even seen cases where the radiator interferes with tall RAM sticks! Use low-profile non-RGB RAM if possible, to ensure it will fit (see Photo 5). Note: while we don’t recommend it, you can also consider removing the Photo 4: the Scythe Shuriken 2 costs around $100 and sits only 58mm tall. To get any more compact than this, you pretty much need a passive cooler and rely on case airflow. You might also need low-profile RAM sticks; sometimes, less tall DIMMs are required to clear other things that might be in the case, depending on its exact configuration. heatsink from the RAM if you need extra space, as most RAM does not get hot enough to need the heatsink for dissipation. Once you have found a case you think is perfect, download and read its manual. There should be a section discussing what will fit and hopefully explain any such limitations. Verify that one of your parts won’t interfere with fitting another; if necessary, download their manuals too, to determine their exact dimensions. Doing this now will save you a headache later! If you are air cooling the CPU (generally the cheaper and more sensible option), check the maximum cooler height supported by your case, which is usually limited by its internal width. Low-profile CPU coolers exist to suit compact cases, but they typically have worse cooling performance and are noisier. If you’re willing to potentially sacrifice some performance, you can also look into a passive cooling setup. I couldn’t find a standard 120mm tower cooler that would fit in the Era case, but I did find a Noctua 92mm tower cooler that fits just fine and is extremely quiet, which I used in my later home theatre build (there’s a photo of it below). Another option is something like the Scythe Shuriken 2, which also has a 92mm fan but it’s horizontal, making the whole thing only 58mm tall, so it will fit in quite compact cases - see Photo 4. Check the GPU size limitations carefully. One of the nice things about the Era case is that it will fit a “2.2-slot” GPU up to 295mm long (depending on the power supply type and location). The EVGA 3070 XC3 GPU I ended up using is a 2.2-slot design that’s 285mm long, and it just fits with an SFX power supply in there. And I mean “just” – more on that later. I suspect the considerably more powerful EVGA 3080 XC3 is the same size. So you could possibly squeeze one of those into the same case, but they were unobtainium at the time I was building this computer. Other Mini-ITX cases can limit GPU width to 2.0 slots or less, so check that. You can get two NVME SSDs onto most Mini-ITX motherboards (possibly three). They are very fast and come in capacities up to about 2TB, although the 512GB and 1TB models are much better value. So unless you need more than 4TB of storage space, you don’t need any external storage. If you need external storage, check what your proposed case will fit and whether you lose any of those slots based on other things like space for a radiator or ATX power supply. Building it Generally, it’s easiest to attach the CPU, RAM and NVME SSD(s) to the motherboard before you install it in Photo 5: these DIMMs are good value for money and perform pretty well, but their shape causes many clearance problems! This is typical of RAM with RGB LEDs, as they are usually mounted in a housing at the top of the actual DIMM. RGB DIMMs look nice (assuming you can see into the case) but cause fitment problems even in full-size ATX cases. Do yourself a favour and avoid them. 42 Silicon Chip Australia's electronics magazine siliconchip.com.au What to look for when choosing a Mini-ITX case ∎ What size/volume do you want? ∎ Do you like how it looks? ∎ What size of power supply does it accept? ∎ Will it fit your proposed CPU cooler (radiator/tower cooler)? ∎ What size GPU will it fit, considering width, length and height and choice of power supply and CPU cooler? You need to have a specific GPU in mind and check its dimensions. ∎ Does it have space for extra fans for case airflow, and if so, how are they configured (intake/exhaust/etc)? ∎ If you need off-motherboard storage, how many and what size of SSDs or hard drives will it fit? ∎ What sort of front-panel I/O does it have? USB 3.0, USB4, Thunderbolt? How many ports? the case. If using an air cooler, you can probably also mount that first. This will give you an ‘assembly’ you can drop straight into the case and then wire up. After that, it’s mainly just a matter of installing the power supply, plugging in the GPU (if you aren’t using onboard graphics), plugging in any external storage, connecting any water cooling hardware, and you’re ready to fire it up. Photo 6 shows the parts I initially chose for my first Mini-ITX computer. They all fit in the case, although I discovered later that the fit of the video card/GPU (shown at the bottom of that photo) was a bit too tight. I ended up swapping it for the slightly smaller but more powerful (and unfortunately, more expensive) EVGA GPU I already mentioned. Next to the case are the boxes for the motherboard and CPU (on top). The 240mm all-in-one (AIO) CPU water cooler is to the right, with the 600W SFX power supply box on top. The 32GB of RAM (2 × 16GB) pack is in the middle, with the 1TB NVME SSD to the right and the GPU in the foreground. I got these from a few different retailers, all of which have good prices and service. There are links to their websites at the end of the article. By now, some of this hardware is no longer cutting edge, so you will probably want to do some research. However, replicating my build is still possible, and some parts are now cheaper. The mesh panel next to the case in Photo 6 came with it, to be used in place of the (elegant) solid timber lid. The timber top is not great for high-power builds as it reduces the exhaust airflow. This case is designed to intake air into the slots around the base, draw it up past the GPU/motherboard/CPU, and exhaust it out the top – a clever configuration that works pretty well. The timber top (or glass in other colour versions of the same case) will be fine for a basic ‘office’ PC. Photo 6: all the parts I used to build my first Mini-ITX PC. The only real change was the GPU (at bottom); while I got it into the case (just!), the fans rubbed on the bottom, so I ended up swapping it for a more powerful but slightly smaller EVGA brand NVIDIA RTX 3070 that just squeezed into the available space. siliconchip.com.au Australia's electronics magazine January 2023  43 CPU choice I chose a Ryzen 5600X for this computer because of its thermal design power (TDP) rating of 65W, along with very good performance. While the cooling system chosen could handle a 105W TDP CPU, that would be an extra 50W+ that the power supply would have to deliver and the cooling system would have to remove, for only a modest increase in performance. At the time (and possibly still today), AMD CPUs were considerably more efficient than comparable Intel parts, which could pull 200W or more under full load. The GPU already draws and dissipates a lot of power, so I didn’t want to overload the system with heat. You could also build a Mini-ITX PC with a newer Ryzen 7000-series or Intel 13th-gen Core series CPU if you’re willing to spend more for higher performance. The only real difference will be in how the CPU mounts on the motherboard. Assembly Photo 7: the side panel is held on by magnets and comes off easily. You need to remove the vertical brace to do most work. This brace can be used to mount two 2.5-inch SSDs or a single 3.5-inch HDD, but I didn’t need it. The two brackets at upper right are for vertically mounting ATX (upper) or SFX (lower) power supplies. Note the small inbuilt IEC mains extension cable. Photo 8: with all the brackets removed, we now have good access to the inside of the case. It comes with one rear exhaust fan, but it’s pretty small at around 80mm across, so you’ll want to keep it running at a low speed to avoid noise. You will want to install two 120mm exhaust fans at the top, by themselves or on a radiator, as they will do the bulk of the cooling. 44 Silicon Chip Australia's electronics magazine Photo 7 shows what you see when you remove the side panel of the case, which, in this case, is easy – just pull it to the side, then up. Most other cases will not have that internal side brace; I removed it early on and only refitted it after I finished installing all the parts. You can mount two 2.5inch drives (or one 3.5-inch) on it, but I didn’t need that feature. I only kept it for rigidity. Photo 8 shows the case with the brace and both power supply brackets removed – much better! The bundle of cables on the right are for the power switch and front-panel USB ports; the wires are far too long, which is quite annoying and caused slight problems that I eventually solved after some head-scratching. One difference between typical ATX and Mini-ITX cases is that, with the former, the power supply usually mounts so that its mains input socket and power switch are externally accessible via a cut-out in the rear of the case. But most Mini-ITX cases mount the power supply internally, hence the IEC mains extension cable. It plugs into the power supply, and the external power cable plugs into a socket built into the rear of the case, visible at lower left. The ATX supply’s power switch is not externally accessible, although that isn’t a big deal. siliconchip.com.au Note how the rear of the case has three large cut-outs. These allow access to the back of the motherboard when installed, which is very helpful as most current motherboards have at least one NVME storage slot on the back. They also give you access to the CPU cooler mounting screws, which often need to be removed to install a custom bracket for a high-end CPU cooler (air or water). Photo 9 shows the parts that came in the box of the 240mm AIO CPU cooler I purchased. It’s called ‘all-inone’ because the radiator, pump, tubes and water block are supplied as a single unit, pre-filled with coolant. The pump is on the hoses near the radiator; it’s more typical to find it integrated with the water block. Two low-noise fans are supplied for cooling the radiator, along with all the necessary mounting hardware and some spare coolant, for a few years down the track, when it might start getting low. I am happy with the performance of this cooling solution and would recommend it to others. It is near-silent unless you are very close to it, and it not only cools the CPU well, it also serves to exhaust hot air from the case (more on how important that is later). Still, the air cooling solution I tried Photo 9: everything shown here comes in the Pure Loop AIO package. The fans mount on the radiator using the supplied hardware in either ‘pull’ or ‘push’ configuration. Either way will work; it mainly depends on whether you want air to flow into or out of the case at that point. The pack at lower left includes thermal paste in a syringe plus mounting brackets to suit various CPU types. later is much cheaper and probably preferable for many builders. In Photo 10, I have installed the radiator in the top of the case, with the fans underneath blowing up. That was easy, as the top of the case is removable. It is at this point that I discovered it’s impractical to use an ATX power supply in this case if you have a 240mm radiator; note how the radiator and fans occupy the space at upper right, forcing the power supply bracket to be installed on one of the lower three screw holes. I don’t believe the manual or specifications sheet mentioned that. Good thing I chose an SFX power supply! Photo 10: the Pure Loop AIO, installed in the case, with the fans blowing air up and out. The tubes are a bit too long as this cooler has to work on larger ATX systems too. Photo 11: the CPU and RAM have been installed, and the fan/heatsink has been moved out of the way to reveal one of the NVME M.2 storage slots. There’s another one on the underside of the board but this one has better cooling, plus it has a direct connection to the CPU, so it should be faster. siliconchip.com.au Australia's electronics magazine January 2023  45 Photo 12: I removed the top of the case and lifted the AIO out to give myself more room to mount the motherboard. I ran the CPU power cable under the motherboard before installing it since it would be hard later. This is a modular power supply, so you plug in the DC cables as you route them, making life easier. It also reduces clutter, as you don’t need to install cables you won’t use. Photo 13: with the remaining power supply DC cables installed and the case top/AIO refitted, the case is getting more crowded. Still, assembly is almost complete. I moved the power supply bracket down to its lowest position so air could flow freely to the right-hand upper exhaust fan. Next, I inserted the CPU in the motherboard’s ZIF socket (Photo 11) and removed the NVME/chipset heatsink/ fan to reveal the upper NVME socket. By this point, I had also installed the two DDR4 DIMMs (32GB of RAM). After installing the NVME SSD and replacing the heatsink/fan, I mounted the motherboard in the case using the supplied screws (Photo 12). However, I routed the power supply wires to the CPU power input at upper left under the motherboard first. You 46 Silicon Chip can see that I have also installed the SFX power supply via the supplied bracket and plugged in the mains input extension. In Photo 13, I have installed all the remaining power supply cables, bundled them up out of the way and plugged in some of the front panel I/O cables to the motherboard. I used spare holes in the power supply bracket to attach cable ties to keep the long power supply wires in check. Another advantage of SFX supplies Australia's electronics magazine for small builds is that they usually come with shorter cables. I discovered that after building some other similar computers using ATX power supplies, the cable slack took up a lot of the spare space inside the case and routing the wiring became quite tricky! In Photo 14, I have applied a thin layer of thermal paste to the CPU (it came with the cooler). Some people like to put a blob or two in the middle and allow the pressure from the cold plate to spread it out, but I prefer this method as it guarantees even distribution. Note the brackets on either side of the CPU – they came with the AIO and are specific to it, replacing the standard AMD brackets. In Photo 15, I have clamped the AIO cold plate down on top of the CPU, installed the GPU (visible at the bottom) and strapped all the unused captive cables to other things to keep them out of the way. The goal was to keep plenty of space for cooling air to flow past the GPU and components on the motherboard. Don’t forget to remove the plastic film from the bottom of the cooler before installing it! The cooler won’t work properly and the CPU will enter thermal throttling under load. Note the slots in the bottom of the case for cool air to be drawn in (visible in the adjacent photos). There is a chamber at the bottom of the case with a perforated section between it and the GPU area to allow fresh air intake. This is where I ran into a snag: the GPU sagged a little bit under its own weight, and its fans touched that perforated section. As soon as the computer powered up, it made a horrible noise as the fans tried to rotate against the plastic. My solution was to get some small adhesive rubber pads from Bunnings and stick them between the GPU’s plastic shroud and the bottom of the case, lifting the GPU enough for the fans to spin freely. That worked, but sometimes when I switched the computer on, the fans would still strike the case. So, I replaced the 5600 XT with the EVGA XC3 RTX 3070. I chose it because it was a hair thinner than the 5600 XT while being quite a bit more powerful. I still needed those rubber pads to prevent its fans from rubbing, but that completely fixed it, and I haven’t had a problem since. siliconchip.com.au All that was left after Photo 15 was to refit the side brace, reattach the side panels, fire the system up and install the operating system. The only other catch I ran into was that some of the front panel cables had become caught between the power supply and the front of the case, causing the power button to act as if it was pressed all the time. That threw me at first, as the system refused to power up (because the button was already ‘pressed’). But once I figured it out and freed those cables, it worked perfectly and has since. Screen 1 shows the result of a benchmark I ran shortly after building the computer, showing that despite its size and lack of ‘high-end’ parts, it was still in the 91st percentile in terms of overall CPU performance at the time. Thermal tweaks Besides the fans on the GPU, which serve only to cool it (circulating and heating the air around it), there are only four fans in this system: the small one on the motherboard that cools the SSD and chipset, a small rear exhaust fan that doesn’t do much, and the two passing air over the AIO radiator. That means the AIO radiator fans also serve as ‘case fans’ to draw air up through the slots in the bottom, past the GPU, chipset and regulators on the motherboard, then through the CPU radiator and out the top. They are vital to the system operating at a reasonable temperature. The RX 5600 XT I initially chose could draw up to 150W (about three times what the CPU does under a typical load). But the RTX 3070 can consume around 250W; I even bumped it up to around 275W to improve its performance. That leads to the air drawn in getting quite warm before it each reaches the motherboard or radiator. Along with the CPU, motherboard, SSD and other parts, the system can draw 350W or more under load. Despite that, system temperatures remain reasonable, with the CPU sitting at around 50-60°C under load, the motherboard at about 40-50°C and the GPU core around 60-70°C (at an ambient temperature of 23°C). This does have the effect of the computer acting like a small (albeit relatively quiet) space heater with all that hot air being ejected from the top. But it will run like that all day. siliconchip.com.au Photo 14: this close-up view shows some details that aren’t apparent in earlier photos and also shows how I smeared thermal paste on the CPU. I could be accused of using a bit too much, but it’s better than too little! I tied cables up to keep them out of the airflow path as much as possible and that also made my life easier when I had to plug cables into the motherboard. Photo 15: assembly is basically complete. Note how I had to loop and tie the water tubes (highlighted in yellow) and how little space there is for the GPU at the bottom. Getting it in and out was not easy, especially given the multiple power cables that needed to be plugged into it (visible at lower right). Note how the GPU heatsink occupies the entire length of the case. Australia's electronics magazine January 2023  47 Photo 16: this Noctua NH-U9S is not cheap at around $120, but it is an almost ideal solution in this style of Mini-ITX case. It provides decent cooling in nearly complete silence and will fit a small case; it’s just not suitable for super-compact jobs. There was a trick to achieving that. By default, the motherboard adjusts the CPU fan speed based on CPU temperature. Because this cooling system is effective enough to keep the CPU at a low temperature even under load with a low fan speed, the fans would not speed up, even if the GPU was dumping a lot of heat into the case. While the system never ‘melted down’, it ran hotter than I liked. The solution was to go into the BIOS and get it to use the “system temperature” to control the CPU fan speed instead. That temperature is sensed at the chipset, so when the GPU starts heating the air inside the case, that also heats the chipset. The fans will quickly ramp up and provide the airflow necessary to remove the GPU heat and keep the motherboard and regulators cool. It works surprisingly well. The other tweak I made was to replace the timber top panel with the supplied mesh panel. Again, the system didn’t melt down before, but that dropped the GPU and motherboard temperatures by about 10°C and made the system run quieter under load. So I consider it a worthwhile change. a radiator at the top of the case, I installed two normal low-noise 120mm computer fans in the top panel and again set them to be controlled by the system temperature. Because those fans are much thinner than a radiator, I could raise the power supply bracket to fit a lower-cost regular ATX power supply (it’s a different computer, but Photo 17 shows the arrangement). Other Mini-ITX cases Sometime after building these systems, I needed a new computer for my home office and decided to see how powerful a computer I could jam into a Mini-ITX case. I did a lot of research using the experience gained from building in the Era cases and settled on the MetallicGear Neo Mini V2 (https://metallicgear.com/products/ Neo-Mini-V2). It is somewhat larger than the Era, although still quite a bit smaller than a Mid-ATX case, and has a tempered-­ glass side panel so you can see the innards. It was also a lot less expensive than the Era cases at $99, although it isn’t quite as stylish. I chose this case because you can mount two radiators in it, one 120 x 120mm at the rear and one 240 x 120mm on the right side (looking at the front). Other cases can mount two radiators, with one at the top, but the computer was going under a desk, and a top exhaust would cause warm air to bounce off the desk and wash over my legs (good in winter, not so much in summer). I used similar parts to the build documented here, except the GPU is a monstrous water-cooled NVIDIA RTX 3090 putting out about 400W under full load (slightly more than its stock power level). It came with an integrated 240mm radiator, which I mounted to exhaust through the side of the case, with dual 140mm intake fans at the front – see Photo 18. Air-cooled system I built a similar system for my ‘home theatre’ using the Noctua ‘tower’ air cooler with a 92mm fan shown in Photo 16. I fitted it to the CPU/motherboard combination before putting it in the case as that was easier. As all that system does is play music and videos, this handles the CPU dissipation just fine. Because that system doesn’t have 48 Silicon Chip Photo 17: you can see how much more easily the Mini-ITX board with the Shuriken 2 fits into this Era case, despite using a much larger ATX power supply. Note also the mess of cables at lower right due to ATX supplies coming with longer cables. I used a smaller and less expensive video card in this system as it is intended for basic office tasks. Australia's electronics magazine siliconchip.com.au Screen 1: the CPU temperature is a very reasonable 62°C with all six-cores loaded due to the generouslysized water cooling system. Fresh air to cool the GPU comes into the front of the case through mesh-­ covered side slots, passes through its radiator and immediately exits the case, so it can’t heat anything else up. Similarly, some of that air coming into the front goes into the rear-mounted CPU radiator and exits the case. Airflow through the case keeps the power supply and motherboard cool. The result is a very high level of performance in a small space (380 x 355 x 190mm; 25.6L). While plenty of warm air comes out, it is blown away from me; the only side-effect I notice is that the room gets warmer after a while. If you wanted to build a similar computer to sit on top of a desk, you could choose the top exhaust option using a suitable case; that might work even better. Overall, MetallicGear Mini-ITX cases (and the related brand, Phanteks) seem well-designed and well-built and are good value for money. Still, there are numerous manufacturers of good small form factor (SFF) cases, including Mini-ITX, so it’s worth doing some research and browsing before deciding which one to purchase. Unfortunately, I can no longer find this case for sale, but similar Metallic­ Gear Mini-ITX cases are available. and came up with the following estimates for a basic but decent system: $100 for the case, $280 for a Ryzen 5600G CPU with onboard graphics, $150 for a B450 motherboard, $100 for 16GB of DDR4 RAM, $150 for a 1TB SSD and $100 for a power supply. That totals almost exactly $900 for a ‘reasonable’ system without a discrete GPU. A system with a discrete GPU will be much better for tasks like playing games or 3D rendering, but you’ll need to add the cost of that GPU to the $900 base price estimate. In that case, you might want a faster CPU, better motherboard and more RAM, adding perhaps another $200 to the total cost. While you can build a Mini-ITX system for under $900, the result will be compromised in some areas (eg, a slower CPU and less memory and/or storage). But for typical use you could likely get away with a cheaper CPU. For your average user, a Mini-ITX PC will do everything a full-size desktop will do in a much smaller package. You can put together a computer like this for around $800-2000, depending on your chosen parts. I checked the cost of parts from Umart (see below) as I am writing this Photo 18: this system built in a MetallicGear Neo Mini V2 case consumes over 500W but still technically has a ‘small form factor’! While it is a bit larger than most Mini-ITX systems, it’s still about half the volume of a typical ATX case. The motherboard ends around where you can see the top of the DIMMs; the space to the right (at front of the case) is occupied by the GPU radiator; the small CPU radiator is visible at lower left. Amazingly, it runs cool due to the carefully considered layout. siliconchip.com.au Australia's electronics magazine Conclusion Remember that those prices don’t include a keyboard, mouse or monitor, although you might already have those if you upgrade from an existing system. Where to buy parts I have found the vendors below reasonably reliable and offer fast and inexpensive delivery within Australia. Compare the prices as sometimes one will offer a product cheaper than others. Umart (NE Queensland): www.umart.com.au Scorptec (Clayton, Vic with stores throughout Australia): www.scorptec.com.au mwave (Lidcombe, NSW): www.mwave.com.au You can also find good parts deals on Amazon (www.amazon.com.au) and eBay (www.ebay.com.au – caveat SC emptor!) 49 Raspberry Pi Pico W BackPack Our Raspberry Pi Pico BackPack from March 2022 has a powerful dualcore 32-bit processor, 480 × 320 pixel colour touchscreen, onboard real-time clock, SD card socket, stereo audio output and infrared receiver. Now, for only about $5 more, it has WiFi too! Project by Tim Blythman M icrocontrollers have become so easy to use, cheap and accessible for hobbyists, while chips like the ESP8266 have made it simple to use WiFi. The Raspberry Pi Foundation’s Pico W is an inexpensive, well-documented 32-bit microcontroller board with WiFi that is well-suited to being used with the LCD BackPack. We reviewed the Pico W in the November 2022 issue and found that it was mostly interchangeable with the Pico (siliconchip.au/Article/15547) but with added WiFi support. So it was only natural for us to update the Pico BackPack to include WiFi support using the Pico W. As it turns out, that was not hard to do. From launch, the Pico supported the MicroPython and C languages (using the Raspberry Pi Foundation’s Features and Specifications ∎ Includes a 3.5in LCD touch panel and a dual-core microcontroller with WiFi. ∎ Also includes all the features of the original Pico BackPack. ∎ We provide software demos and examples for the Arduino IDE, C SDK and MicroPython. ∎ Our sample code demonstrates practical uses of HTTP, UDP and NTP. 50 Silicon Chip Raspberry Pi is a trademark of the Raspberry Pi Foundation C software development kit). Arduino support in the form of the Arduino Pico board profile came soon after. The Raspberry Pi Foundation has made many inexpensive single-board computers and microcontroller boards available to the masses, even amid continuing electronics component shortages. The Pi Pico series are simple but well-thought-out boards and are attractively priced for what they offer. BackPack hardware We considered whether it was worthwhile to update the Pico BackPack PCB to complement the Pico W, but ultimately, we decided not to make any significant changes. The thing is that the Pico BackPack crams a lot of features into a small area corresponding to the size of the matching LCD touch panel. To add any features would likely mean removing some of the existing features, which we didn’t want to do. The Pico BackPack has a row of I/O pins to make external connections, so it’s easy enough to connect different hardware if necessary. Thankfully, we’d already established that the Pico W didn’t ‘break’ any existing functionality of the Pico BackPack. So the BackPack PCB remains the same for the Pico W, although we will recommend a minor Australia's electronics magazine assembly variation to enhance the WiFi capability. The Pico W BackPack The only substantial difference between the Pico BackPack and the Pico W BackPack is the replacement of the Pico module with a Pico W. All the pins on the Pico W are labelled the same as those on the Pico, so none of the signals or I/O pin breakouts need to change. Still, as we noted in our review of the Pico W, both the BackPack PCB and LCD touch panel have large solid copper areas that could impede WiFi signal propagation. Therefore, we recommend that the Pico W is mounted slightly away from the BackPack PCB to provide better clearance for its onboard WiFi antenna. We used header strips to provide this spacing. You could also use low-profile socket headers and short pin headers if you wish to make the Pico W pluggable. We tried this and found it worked well, although it was fiddly to assemble. Circuit details Fig.1 shows the circuit diagram for the Pico W BackPack. It is identical to the original Pico BackPack, with the Pico replaced by a Pico W. IRRX1 at top left allows the Pico W to receive IR signals on its GP22 digital siliconchip.com.au Fig.1: the Pico W BackPack circuit is almost identical to the Pico BackPack. It includes an IR receiver, microSD card, real-time clock, audio output and LCD touch panel. A 20-way header provides access to power and spare I/O pins for adding more features. The 1kW resistor at IRRX1’s output is not needed in most cases. input. The LCD touch panel connects to power and the SPI bus at the top, as does the microSD card socket at upper right. The two transistors on the right control the power to the LED backlight on the LCD touch panel. Below this, a DS3231 real-time clock and calendar IC connects to the I2C bus. siliconchip.com.au Finally, the components at the bottom, including the op amps, can deliver line-level audio at CON3. They connect to pins on the Pico W that generate pulse-width modulated (PWM) signals to provide synthesised analog voltages. For more details and specifics about how the various features work on the Australia's electronics magazine Pico BackPack PCB, refer to the article from March 2022 (siliconchip.au/ Article/15236). The original software to interface to the BackPack hardware was also explained in that article. Construction While that March 2022 article has more detail on assembling the PCB January 2023  51 removable. We figure it’s inexpensive enough that you are better off saving the effort and just soldering it. Software with WiFi support The release of the Pico W has allowed us to update the Pico BackPack with WiFi. It’s a powerful combination that we think will be the basis of some diverse and interesting projects. We’re providing several practical WiFi demos to make it easy to pick up and use. and fitting it to the LCD touch panel, experienced constructors should have no trouble using the overlay in Fig.2 to assemble the PCB. If you refer to that earlier article, the PCB construction is no different until you get to the Pico W module. Most IR receivers will not need the 1kW resistor; in fact, it will interfere with their weak internal pullup. Hence, it has been omitted from the overlay and is not seen in our photos. Don’t forget the cell holder on the reverse of the PCB if you are fitting the real-time clock IC. Lines separate the various sections of the board on the silkscreen. That helps you to omit some components if you wish to reclaim some I/O pins by not using those features. As we mentioned earlier, the Pico W should be spaced away from the main BackPack PCB and also kept clear of the LCD above. Thus, we have added two 20-way pin header strips to the parts list. Solder these to the BackPack PCB, with the plastic carrier sitting above. Then solder the Pico W to the top of the pin headers. The plastic carrier separates the Pico W from the BackPack PCB. Our photos show how the Pico W is spaced above the BackPack PCB by a small distance. The other option requires low-­ profile (5mm) header sockets too. Altronics Cat P5398 can be used but you will need two lengths, cutting them down to 20 pins each. 52 Silicon Chip The fiddly part is fitting the pin headers to the Pico W, as this requires removing the metal pins from their plastic carrier to minimise the height. Although the plastic carrier is only 2.5mm high, it’s enough to cause the Pico W to foul the LCD, so it must be removed. After pulling the pins out of the plastic carrier, insert them individually into the socket header entries. You can then place Pico W over the pins and solder them to it. Depending on the length of the pins, they might also need to be trimmed so that the pins do not foul the LCD screen. The only advantage of that more fiddly approach is that the Pico W is Of course, we need some sample code that uses WiFi to show off the Pico W’s new feature. Since PicoMite BASIC will not support the Pico W’s WiFi (as noted in the November review article), our software samples do not include PicoMite BASIC. Existing PicoMite BASIC programs should work fine on the Pico W, with the minor exception that the Pico W’s onboard LED is driven differently, so it can’t be controlled as it would be on a Pico. We have updated the Arduino, C SDK and MicroPython examples to add WiFi features. As we noted in our review of the Pico W, a document called “Connecting to the Internet with Raspberry Pi Pico W” explains how to do this with the C SDK and MicroPython. But that guide is quite basic; our sample code does much more. Since the updated demos are based on the earlier versions we made for the original Pico BackPack, we recommend reading the original Pico BackPack article for information on the original features. One of the great features of the Pico and the Pico W is the bootloader which implements a virtual flash drive, allowing software to be uploaded by simply copying a file to the virtual drive. The bootloader is in mask ROM in the RP2040 microcontroller that runs the Pico and Pico W. This makes it Fig.2: the lines on the overlay delineate the components that provide the different features of the Pico W BackPack. There is also a cell holder on the rear of the PCB, used by the real-time clock IC to keep time when power is not otherwise available. The Pico W is spaced above the main PCB to improve the performance of its WiFi antenna. Australia's electronics magazine siliconchip.com.au practically impossible to ‘brick’ the Pico or Pico W as the bootloader cannot be overwritten. Bootloader mode is entered by holding down the BOOTSEL button on the Pico or Pico W while powering up or resetting the chip. Since the BackPack provides a reset button, you can start the bootloader by pressing and holding BOOTSEL while pressing S1 on the BackPack. Software images for the Pico and Pico W use the UF2 file type, which is a binary format, unlike the text-based HEX files used for other chips like PIC microcontrollers. If you are simply interested in seeing what the Pico W BackPack is capable of doing, all you need to do is copy the respective UF2 file to it after putting the Pico W into bootloader mode. We’ll go into a bit more detail about the workings of the software later in this article. To simplify entering the WiFi credentials, you can set them using the virtual serial port. You will need a serial terminal program, such as Tera­ Term, minicom or the Arduino Serial Monitor, to communicate with the Pico W. You might notice that the demo .uf2 files are larger than the Pico examples due to the extra libraries needed to communicate with the WiFi chip. The WiFi chip also needs a 300kB binary ‘blob’ to work, which is bundled into the firmware images. Arduino coding The team that created the Arduino-­ Pico port for the Arduino IDE has done a good job of aligning the Pico W’s WiFi API (application programming interface) to that used by other WiFi boards, such as those based on the ESP8266 and ESP32 processors. Indeed, it is based heavily on that of the ESP8266. You might remember the D1 Mini BackPack from the October 2020 issue (siliconchip.au/Article/14599). It uses an ESP8266-based D1 Mini module to drive an LCD touch panel and has many features in common with the Pico W BackPack. We’re using version 2.5.2 of the Arduino-Pico board profile, although versions as old as 2.30 should support the Pico W. You can find more information about the board profile at https://github.com/earlephilhower/ arduino-pico siliconchip.com.au Parts List – Pico W BackPack 1 double-sided PCB coded 07101221, 99 x 55mm 1 Raspberry Pi Pico W Module (MOD1) [Altronics, Core, Digi-Key, Little Bird] 1 3.5in LCD touchscreen [Silicon Chip Shop Cat SC5062] 1 14-pin, 2.54mm pitch socket header (for LCD panel) 3 20-pin, 2.54mm pitch pin header (CON2 & to mount Pico W) 2 20-pin low-profile (5mm tall) 2.54mm pitch socket headers (optional) 2 2-pin, 2.54mm pitch pin headers with jumper shunts (JP1, JP2) 1 6mm x 6mm tactile switch (S1) 8 M3 x 6mm panhead machine screws 4 M3 x 12mm tapped spacers Semiconductors 1 IRLML2244TRPBF/SSM3J372R P-channel Mosfet, SOT-23 (Q1) 1 2N7002 N-channel Mosfet, SOT-23 (Q2) Resistors (all M3216/1206, 1%, ⅛W) 1 10kW 1 1kW Optional Components SD card 1 SMD microSD card socket (CON1) [Altronics P5717] 1 10μF 10V X7R SMD ceramic capacitor, M3216/1206 size 1 100nF 10V X7R SMD ceramic capacitor, M3216/1206 size Real time clock/calendar 1 surface-mounting CR2032 cell holder (BAT1) [BAT-HLD-001] 1 DS3231 or DS3231M in SOIC-16 (wide) or SOIC-8 package (IC1) 1 100nF 10V X7R SMD ceramic capacitor, M3216/1206 size 2 4.7kW 1% ⅛W M3216/1206 size IR receiver 1 3-pin infrared receiver (IRRX1) [Jaycar ZD1952] 1 10μF 10V X5R SMD ceramic capacitor, M3216/1206 size 1 1kW 1% ⅛W resistor M3216/1206 size (see text) 1 470W 1% ⅛W resistor M3216/1206 size 1 100W 1% ⅛W resistor M3216/1206 size Stereo audio 1 MCP6272(T)-E/SN, MCP6002(T)-I/SN or -E/SN dual op amp, SOIC-8 (IC2) 1 3-pin, 2.54mm pitch pin header (CON3) 2 1nF 25V X7R SMD ceramic capacitors, M3216/1206 size 2 100nF 10V X7R SMD ceramic capacitors, M3216/1206 size 2 10uF 10V X5R SMD ceramic capacitors, M3216/1206 size 4 100kW 1% ⅛W resistor M3216/1206 size 2 47kW 1% ⅛W resistor M3216/1206 size 2 22kW 1% ⅛W resistor M3216/1206 size 2 10kW 1% ⅛W resistor M3216/1206 size 2 100W 1% ⅛W resistor M3216/1206 size As well as adding WiFi support, we’ve updated the Arduino sample code to include an infrared receiver decoding library. In our original Pico BackPack article, we mentioned that we expected the IRRemote library to be ported to the Pico (and Pico W), which has now happened. You can find that library online at https://github.com/Arduino-Irremote/ Arduino-Irremote or it can be installed by searching for “irremote” in the Arduino Library Manager. We have also included a copy of the version Australia's electronics magazine we’ve used in the software bundle. Screen 1 shows the BackPack running our updated Arduino Pico W sample. We have added some text to the LCD panel to show the status of the WiFi hardware. Setting up the WiFi Since using the Pico W in a meaningful way requires that it connect to a WiFi network, we have added a configuration menu on the virtual serial port. We did it that way, rather than using the touchscreen, because it’s easier to January 2023  53 ► Screen 1: the Arduino demo for the Pico W has the most features, primarily due to the excellent library support the Arduino community offers. Apart from the new WiFi features, there is now also support for the IR receiver. ► Screen 2 (right): all the demos include a menu system that can be accessed from a serial terminal program. This is to simplify entering the WiFi credentials needed for the demo to work. The Arduino output is shown here. enter WiFi credentials via a computer rather than an on-screen keyboard. Screen 2 shows the menu that is presented over the serial port by the Arduino software. Items are selected by typing the number and pressing the Enter key. Items 2 and 3 will prompt for the SSID name and password, also followed by Enter. This demo can scan for WiFi networks and connect by name and password. It can also connect to a website over HTTP to retrieve data from the internet. In this case, we use ip-api. com to get some location text to display, along with a timezone offset for that location. This isn’t perfect and would probably be fooled by a VPN (virtual private network), but it will usually give the correct timezone. We think it is a simple and effective way of demonstrating the use of HTTP on the Pico W. We also use NTP (network time protocol) to provide the current time in UTC, adjusted by the timezone offset to provide accurate local time. This can then be saved to the RTC IC on the BackPack. To do all this, you would use menu items 2, 3 and 4 to connect to a WiFi network, followed by 8 to get the offset and 7 to set the RTC. You can set the offset manually using item 6 if item 8 does not work. The IRRemote library is also used to capture and decode IR signals, as displayed in the line beginning “NEC” in Screen 1. This indicates that an NEC 54 Silicon Chip code was last received and shows that code. Code differences The Arduino code for the updated Pico W Backpack differs from the earlier Pico BackPack example only in the main sketch file, plus the requirement to have the IRRemote library installed. It uses other library files that are part of the Arduino-Pico board profile, including those needed for WiFi. Those who have worked with modules based on the ESP8266 or ESP32 will be familiar with how WiFi works under the Arduino IDE; the Pico W is similar. Three library includes are used to implement the WiFi features: #include <WiFi.h> #include <WiFiUdp.h> #include <HTTPClient.h> NTP requires the UDP protocol for communication, hence its inclusion. Fetching web pages uses HTTP. Scanning for networks is done by running a single line of code, as is connecting to a network: WiFi.scanNetworks(); WiFi.begin(ssidname,ssidpass); These calls are blocking (ie, the program doesn’t proceed until the action is completed), and the latter can take up to ten seconds to run. So they may not suit all applications. The C SDK gives better access to the low-level commands and might be more suited if blocking calls are not desired. Australia's electronics magazine It is possible to use function calls from the C SDK in the Arduino IDE, but we preferred to keep the Arduino code consistent with the Arduino way of doing things. NTP is implemented as a background routine that simply needs to be started and then quietly synchronises in the background. Fetching a website using HTTP can be done in a few lines: http.begin(wificlient,URL); httpCode=http.GET(); Serial.print(“Return code:”); Serial.println(httpCode); if(httpCode == 200) { Serial.println( http.getString() ); } We got around some of the longer blocking sections by using the second processor core to do some tasks in the background without interrupting the main program flow. These can be seen in the setup1() and loop1() functions. At the time of writing, we have not seen an official Arduino board profile for the Pico W, so we were unable to try this out as we did for the Pico. But the Arduino-Pico board profile appears to be updated regularly and works well; we have no hesitation in recommending it. Using it with the C SDK Screen 3 shows the LCD panel of the BackPack loaded with the C SDK (software development kit) demo. It siliconchip.com.au as it gets updated in time. This means that the main program is not blocked from other operations while network activity occurs. Using HTTP requires several callback functions to be set, meaning that using the C SDK can seem a bit more complicated than using the Arduino IDE. Still, if you have the patience to set up and delve into the C SDK, we recommend trying it, especially if you need to get the most performance from your Pico W BackPack. MicroPython Screen 3: the C SDK demo runs fast, with good access to low-level functions. Support for protocols like NTP and HTTP is very good once you get it working. includes similar elements to the Arduino example, although the C SDK does not have library support for the IR receiver or RTC chip. There is an RTC feature in the Pico W (and Pico) that can be used by C SDK, but it doesn’t provide the battery backup timekeeping feature that chips like the DS3231 have. It needs the time to be set each time Pico W is reset. Since the Pico W uses a crystal oscillator, it should be pretty accurate once it has been set. The C SDK performs similar tasks to the Arduino demo, using a WiFi connection and NTP to update the RTC. Location and timezone data are also fetched from ip-api.com using HTTP. Several library files are needed for WiFi support. The first file is required to interface with the Infineon CYW43439 chip that provides the WiFi interface, while the others provide library support for HTTP and NTP: Pi computer, we ran it on a Windows PC using the pico-setup tool that can be found at https://github.com/ndabas/ pico-setup-windows This resulted in many minor glitches, especially as some of the commands are subtly different. If you have a Raspberry Pi computer handy, you might find it more straightforward to program the Pico W via the C SDK. Just as for the original Pico BackPack demos, the C SDK software runs very fast and some lower-level functions allow more control than we could easily achieve with the Arduino IDE. In most cases, the serial port menu is used to start an action, such as starting a network scan or connecting to a WiFi network. These do not return immediately like the Arduino equivalents. Instead, the main program monitors the status of variables like the Pico W’s IP address and displays information The MicroPython version available for the Pico W at the time of writing is tagged as ‘unstable’, although we did not have any issues using it. We have included a copy of this version with our software bundle. Note that there are different MicroPython UF2 files for the Pico and Pico W. Be sure to use the correct version. Our MicroPython demo has much the same features as the C SDK demo, as shown in Screen 4. We haven’t made any changes to the two library files (from the original Pico BackPack demo); only the “main.py” file has been updated. Like the Arduino IDE, several libraries must be imported to provide WiFi functionality: import network import urequests import ntptime We noted that the original Micro­ Python software was barely fast enough to be useful. The addition of the WiFi features does make interacting with the LCD touch panel quite slow. Still, we expect most people would not try to cram in all the features that we have. Like the Arduino code, many #include “pico/cyw43_arch.h” #include “lwip/apps/http_client.h” #include “lwip/apps/sntp.h” To properly use the C SDK with the Pico W, we had to make a few changes to the CmakeLists.txt file, especially in the target_link_libraries and add_definitions sections. Look at our sample project to see what to do before creating your own projects. While the C SDK is primarily intended to be used on a Raspberry siliconchip.com.au This shows the spacing needed to give clearance for the Pico W’s WiFi antenna. Short pin headers are the simplest way to achieve this while also keeping clear of the LCD touch panel, which is mounted above. Australia's electronics magazine January 2023  55 it uses a compiled rather than interpreted language. The C SDK was a bit more tricky to work with, but the results are fast and responsive. It also gave us much better access to low-level operations. Bluetooth will be a nice feature to have when it arrives, but as it stands, the Pico W is very useful at its current price and works very well with the BackPack hardware. Now that we have WiFi working well with the C SDK, we think the Pico W will be a good choice for future projects needing WiFi. The Arduino IDE will be a handy option when we want to quickly interface with hardware, especially if it needs library support. Availability Screen 4: the MicroPython demo has similar capabilities to that of the C SDK. It’s possible to use the drawing feature of the demo, but it is not very responsive. MicroPython routines are blocking and may not return for many seconds. The features available are much the same as the C SDK, with options to scan for networks and set the SSID name and password. You can connect, disconnect and make an HTTP request to retrieve data. Is there Bluetooth support? Since the Infineon CYW43439 WiFi chip has support for Bluetooth, many people have been left wondering whether the Pico W will be able to use Bluetooth. At the time of writing, it appears that is not the case. Instead, we are simply left with the tantalising statement from the folks at the Raspberry Pi Foundation that it “may be enabled in the future”. Summary Our demo code does many things you might typically do with a WiFi-­ capable microcontroller: connect to a network, make HTTP requests to fetch data from websites and use NTP to set the time. The Arduino IDE (using Arduino-­ Pico) and MicroPython made this very easy. We found the Arduino IDE more attractive as it has better library support, and the code runs quicker since At the time of writing, the Pico W was available from: ∎ Altronics (Z6424) siliconchip.au/link/abi5 ∎ Digi-Key Electronics (SC0918) siliconchip.au/link/abgw ∎ Core Electronics (CSE08703) siliconchip.au/link/abgx ∎ Little Bird Electronics (SC0918) siliconchip.au/link/abhj Other retailers we expect might stock the Pico W when it becomes available in volume include element14 SC and Mouser. SC6625 Kit ($85 + P&P) includes all parts in the parts list except the socket headers and DS3231 IC (the DS3231 is available separately – SC5103 or SC5779). U Cable Tester S B Test just about any USB cable! USB-A (2.0/3.2) USB-B (2.0/3.2) USB-C Mini-B Micro-B (2.0/3.2) Reports faults with individual cable ends, short circuits, open circuits, voltage drops and cable resistance etc November & December 2021 issue siliconchip.com.au/Series/374 DIY kit for $110 SC5966 – siliconchip.com.au/Shop/20/5966 Everything included except the case and batteries. Postage is $10 within Australia, see our website for overseas & express post rates 56 Silicon Chip Australia's electronics magazine siliconchip.com.au ADD MOTION DETECTION TO YOUR PROJECT PIR MOTION DETECTION MODULE ADD OBSTACLE DETECTION OR AVOIDANCE DUAL ULTRASONIC SENSOR MODULE • Adjustable delay times XC4444 $6.95 • 2 - 45cm 15° range XC4442 $8.95 Expand your projects with our extensive range of Arduino® compatible Modules, Shields & Accessories. OVER 100 TYPES TO CHOOSE FROM AT GREAT PRICES. ADDRESSABLE RGB LEDS DETECT WHEN PLANTS NEED WATERING SOIL MOISTURE SENSOR MODULE 1.3" MONOCHROME OLED DISPLAY • Analogue output XC4604 $4.95 • 128x64 Pixel XC3728 $19.95 VIEW OVER 70 ARDUINO® PROJECTS YOU CAN BUILD AT: jaycar.com.au/projects Shop at Jaycar for: • Arduino® Compatible Development Boards • Display Modules • Servos, Solenoids & Motors • Wheels & Chassis ADD AMAZING COLOUR TO YOUR NEXT PROJECT 5V LED STRIP WITH 120 ADDRESSABLE RGB LEDS HALL EFFECT SENSOR MODULE • 2m long, flexible, waterproof XC4390 $29.95 • Sense magnetic presence XC4434 $4.95 • Prototyping Hardware and Accessories • Project Enclosures • Servos & Motors • Switches & relays Explore our wide range of Arduino® compatible modules, shields and accessories, in stock on our website, or at over 110 stores or 130 resellers nationwide. Prices correct at time of publication but are subject to change. Jaycar reserves the right to change prices if and when required. jaycar.com.au/shieldsmodules 1800 022 888 High-Performance Part 1: By Phil Prosser Active Subwoofer For HiFi at Home This Subwoofer is designed to be a ‘no compromise’ approach to a sub, making it a perfect match for a high-quality home theatre system, or as part of a high-fidelity stereo system. T he Active Subwoofer uses an SB Acoustics SB34SWNRX-S75-6 346mm (12-inch) driver plus a built-in 200W class-AB amplifier module that can deliver up to 180W of continuous output power in this application. It is designed to match the Active Crossover Amplifier and Active Monitor Speakers, described over the last two issues. However, it is a very high-quality sub that you could use in any application. It will provide high power, extremely low distortion bass for the lower octaves. Subwoofers are all about moving large volumes of air. The deeper you go into bass frequencies, the more of a challenge that becomes. For true high fidelity, we want a -3dB point well below 30Hz and to achieve solid output to 20Hz. Unfortunately, we also need to consider practicalities like the physical volume required. That requires us to set aside exotic approaches such as infinite baffles or horn loading. After modelling quite a few similar drivers, I settled on the SB Acoustics SB34SWNRX-S75-6. Mounted in an 80-litre enclosure tuned to 25Hz, it gives a -3dB point at 25Hz and is only 8dB down at 20Hz in free space. This enclosure is modest for such a hefty driver and for operating to such low frequencies. I could have opted for a much larger enclosure and tuned it lower, but I feel that the increase in size and porting difficulties are not in line with most people’s needs. This is a serious subwoofer. With the amplifier running flat out, delivering 58 Silicon Chip close to 200 watts, this driver operates entirely within its linear region right down to 20Hz. I have built a lot of subs, including professional audio products, and this is an outstanding result in comparison. Driven at this power level, the Sub will produce over 110dB SPL (sound pressure level) right down to 30Hz and over 100dB SPL at 20Hz. Those figures are for free space; in the real world, there is a floor and usually a wall or two, which will increase them by up to 6dB. The fact that we are in a finite volume room means the Subwoofer basically produces a flat response to close to 20Hz. The voice coil on this driver is 75mm in diameter and 28.5mm long. That is a very long voice coil, required to achieve the linear excursion mentioned above. One consequence of this is that much of the voice coil is outside the magnetic air gap, which is 6mm high. That significantly impacts driver efficiency, which is the price we pay for achieving high output at low frequencies. It can be driven from a home theatre amplifier’s subwoofer output or an active crossover. I recommend that the Subwoofer be placed not too far from your main speakers, but somewhere that your family members will accept. If cost is no object, two subs are always better than one. I would place each Subwoofer in the general proximity of one main speaker. To be honest, though, it is not likely that a single active subwoofer will ever ‘run out of puff’. The fantastic thing about this Australia's electronics magazine Subwoofer is that the very extended frequency response does not come at the expense of power handling, and you can safely drive it at very high levels right down to 20Hz. Yes, it is a significant investment to achieve this, but in use, it is truly impressive. Vented or passive radiator I have opted to use a slot vent in our Subwoofer. Passive radiators exist that can be paired with the Subwoofer, but they are pretty expensive, and you need two of them! The port is as large as I could fit and has flared ends to minimise ‘chuffing’ at high outputs. It is made with stacked layers of MDF cut to form flares at both ends, resulting in a 48-50mm high, 180mm wide port. The vent configuration is shown in the ‘X-ray’ style overview of Fig.1, along with the amplifier and enclosure, both described below. If you are not expecting to drive the Subwoofer at high levels or very deep, a single 10cm diameter round port of 41cm length will suffice. Still, with the investment this Subwoofer represents, I feel that compromising on the port is missing the point. The amplifier The integrated amplifier takes its input from an RCA line-level input from the Active Crossover amplifier and delivers about 180W. The amplifier I used is the Ultra-LD Mk.4 module (August-October 2015; siliconchip.au/Series/289). Alternatively, you can use the Ultra-LD Mk.3 200W module (July-September 2011; siliconchip.au/Series/286) if you don’t siliconchip.com.au What is needed to build an Active Subwoofer Ultra-LD Mk.3 or Mk.4 Amplifier Mk.3 – July-September 2011; siliconchip.au/Series/286 Mk.4 – August-October 2015; siliconchip.au/Series/289 Multi-Channel Speaker Protector (4-CH) January 2022; siliconchip.au/Article/15171 Timber for the case, acoustic wadding, heatsink, wires and other miscellaneous parts (see the parts list) like working with SMDs. Both are fine performers in this role. I have designed a chassis that will suit either amplifier module as they are the same size. where the Sub is located (unless things are rattling around it). If you use it with a different home theatre system, I expect the crossover to be in the 80-150Hz region, which will The enclosure work fine. The enclosure is made from This size is at the sweet spot where 18mm-thick MDF. To provide extra a subwoofer moves from being ‘disstrength and reduce vibration, the guiseable’ in a home to something front and rear panels are double-­ you need to work to accommodate. layered, and there is a full brace in the The enclosure is rock solid and capamiddle of the enclosure. The enclo- ble of both incredible precision and sure is 560mm deep, 470mm wide earth-shattering bass. and 470mm tall. In our loudspeaker system, the Performance active Subwoofer is crossed over at Fig.2 shows the modelled (expected) 80Hz with a very steep 24dB/octave response, while Fig.3 shows the actual slope, so there is no chance of ‘hearing’ measured response. This was made outdoors, about 1.5m from a shed, with the microphone at listening height for the active monitor speakers on 0.8m stands, and at a distance of 1m from the Subwoofer. The measured -3dB point is 27Hz. The subsonic filter for the subwoofer output was active; removing that would extend the bass deeper. There Fig.1: a top-down ‘X-ray’ view of the Subwoofer complete with its integrated ‘plate amplifier’. siliconchip.com.au Australia's electronics magazine January 2023  59 Fig.2: the modelled response of the SB Acoustics SB34SWNRX-S75-6 365mm driver in an 80.5-litre enclosure with a tuning frequency of 25.03Hz. is some ripple in the response, but that is unavoidable without going to extremes. The frequency response of subwoofers is tough to measure cleanly indoors due to room resonances and the impact of floors and walls on overall gain. One measurement I took indoors is shown in Fig.4. This is a composite measurement about 20cm from the woofer and port. “Room gain” is a phenomenon where the resonance of a room increases the output from a subwoofer. This is mainly seen below the frequency at which the room’s longest dimension is half a wavelength. For a 10m-long room, that is about 17Hz. Our measured response shows greater output at low frequencies than the ThieleSmall modelling suggests we should see, almost certainly due to room gain. The Subwoofer’s impedance curve is shown in Fig.5. It is well within the handling capabilities of the Ultra-LD amplifiers we are using and low enough to get almost the full 200W available into the driver. The enclosure There are many ways you can build the enclosure. Fig.6 shows how you can cut all the panels from a single 2400 × 1200mm sheet of 18mm-thick MDF while minimising the number of cuts. I did it that way as I don’t have a table saw and wanted to get the sheet cut at the local hardware store where I purchased it. This proved very successful, and in less than 15 minutes, I had all the major panel cuts done and the panels within 1mm of the specified size. The whole lot then fit in the back of the VW Golf to get it home. The tools you will need to finish the raw panels include a router, jigsaw, cordless drill or hand tools and a lot of elbow grease. Review the drawings before you proceed; detailed views of the cut panels Fig.4: the composite response of the indoor output from the cone (dark blue) and port (red) show they combine to give the predicted response. 60 Silicon Chip Fig.3: a measurement of the Subwoofer’s response outdoors, as far away from sound-reflecting objects as was practical (excepting the ground). are shown in Figs.7-12. I used routed rebates for all panel joints that allow you to simply glue and clamp the enclosure together if you have many sash clamps. This routing can be done very simply using a jig, described below. You will also need to cut out the holes for the port and amplifier module, and rebate the driver hole. If you don’t like the idea of using a router, you could resize the panels and screw them together as butt joints. You will see in the photos that I used screws as well as rebates. That was to make assembly clear and simple for Zak, my 9-year-old helper who was over for the weekend. He really wanted to get involved and, between us, gluing and screwing the rebated panels went very well. My suggested numbered assembly steps are as follows. 1 - Purchase the MDF panel and get it cut into the main pieces. This should be a fair stack of timber. Fig.5: the impedance of the Subwoofer mounted in the enclosure before connecting the power amplifier. The peaks show that our tuning is as predicted. Australia's electronics magazine siliconchip.com.au Fig.6: these are the subwoofer panel cuts from 18mm MDF when using the recommended rebated joints. 2 - Route the panels as shown in the panel routing figures (Figs.7, 10 & 11). By screwing an off-cut of 18mm MDF to your worktop and a straightedged off-cut at 90° to it, you can make an extremely effective routing jig into which the 18mm panels fit perfectly, as shown in Photo 1. Using this jig and an end stop, there is no need for measuring and fiddling to route the brace as the rebates are all at the same depth (5mm). Similarly, you can route the rebates on the end panels using this jig to ensure everything is square. 3 - Make the driver hole. I used a circle jig made from an aluminium off-cut. I made several holes in it to get the diameter of the rebate hole and driver cut-out just right, testing with the driver to ensure they were correct. The result is shown in Photo 2. The driver rebate is 10mm to ensure the frame sits flush with the front panel. Photo 1: with a router and some MDF off-cuts, you can build a jig to make precisely aligned rebates. siliconchip.com.au Australia's electronics magazine January 2023  61 Photo 2: My home-made circle jig allowed me to create a clean circular rebate and cut out the driver hole perfectly. Photo 3: the stack of panels after the rebates and holes have been made. The vent sides are on the top of the pile (and shown below). They are made from three layers of stacked MDF glued together & sanded smooth. Fig.7: details of the rebates routed in the top and bottom panels (all 5mm deep). Other than that, they are simple rectangles of MDF. Photo 5: it’s critical to ‘dry fit’ everything together before applying glue. If you start gluing and find a problem, it will be harder to fix. 62 Silicon Chip 4 - Cut out the vent holes and holes in the brace. I used a jigsaw. 5 - Cut out the vent sides and flares, glue them together and fill and sand them smooth. I used some ‘bog’ I found in the shed; any sandable filler will work. Don’t use acrylic filler as Australia's electronics magazine it will not sand! It does not need to be super smooth, but I did want to smooth over some of my less spectacular jigsaw cuts. Assembly With the panels made, as shown in siliconchip.com.au Fig.8 (left): here’s how to make the internal brace. The sizes and shapes of the holes don’t need to match mine exactly but make them reasonably close to get the specified performance. Fig.9 (below): the rear panel is made of two pieces of MDF glued together, one slightly smaller than the other. Photo 3, it’s time to assemble them using the following steps. Fig.13 is a side ‘X-ray’ view of the Sub, which might help you understand how it all goes together. 1 - Do a ‘dry fit’, as shown in Photo 5. Take all the pieces and assemble the enclosure without glue or screws. Use masking tape to hold the panels together. You need to be sure that everything fits and that there are no unmanageable gaps. If you need to file or trim any panels, now is the time, as a good job is almost entirely in the preparation. siliconchip.com.au 2 - If you plan to use screws and glue, drill and countersink the holes to accommodate the screws. A 4mm drill is about the right size. When assembling the box, you will want to use a 3mm drill to make pilot holes for the screws in the end grains. This might seem like a large pilot hole, but the 50mm screws will be totally secure, and you will experience no splitting of the MDF. 3 - Install the rear panels. This step requires the rear exterior and interior panels to be attached to the base. First, sit the two rear panels in the rebate and Australia's electronics magazine Photo 6: installation of the rear panels. I routed straight across the bottom panel, then filled the rebate with wood filler in the port area. then dry-fit the side panels to ensure the alignment of the rear panels is good. Screw the rear interior and exterior panels together using 35mm-long 8G screws with PVA adhesive between the panels. Make sure they are held tightly together. Now align this on the base panel, ensuring the two side panels fit perfectly. Screw this to the bottom panel. 4 - Attach the sides and the port braces. To get the left side perfectly aligned, drill pilot holes for the screws in the right spots and screw and glue it in. Then fit the brace pieces so they January 2023  63 Fig.10: similar to the rear panel, the front panel is two pieces of MDF glued together. See our hints on how to make a jig to route the circular rebate and cut the hole neatly. are flush on the rear exterior panel. Make sure they are parallel inside the enclosure and secure them. Finally, install the right-hand panel. 5 - Install the internal brace and front panels. First, glue and screw down the panel that forms the top of the port. The internal brace and front panels should slide straight into place in their rebates. If not, adjust them until they are a perfect fit. Glue and screw them in. 6 - Finally, attach the top panel (Photos 7 & 8). Make sure any glue that squeezes from the joints is cleaned up as once dry, it is hard to remove. Finishing the enclosure I chose to paint the Active Subwoofer, the key steps being: 1 - Routing the corners with a 6mm radius router to make the edges smooth. 2 - Sealing the enclosures with acrylic primer applied with a roller. 3 - Sanding the enclosure lightly to get rid of any gross roughness. 4 - Filling all screw holes and end grains with filler, ensuring not to put too much. That would be a terrible mistake to make; a thick layer of filler is very hard to sand down. 5 - Sanding it smooth (Photo 9). 6 - Repeating the filling and sanding until the surface is perfect. 7 - Prime again, sand and paint (Photo 10). The subwoofer amplifier I built the Ultra-LD Mk.4 amplifier and mounted it with a suitable power supply on an aluminium plate. I chose this amplifier as it will deliver close to 180W continuous into our 6W subwoofer driver. I fabricated a bracket and panel to accommodate the amplifier and all parts to make a stand-alone module, that slips into a 220 × 170mm cut-out Fig.12: the vent is made from these pieces, but note that you should cut the six side pieces from 16mm MDF to get the required 48-50mm total thickness for three pieces, or use four cut from 18mm MDF and two from 12mm MDF (18mm × 2 + 12mm = 16mm × 3 = 48mm). Photo 7: at this point, all the panels except the top have been attached. 64 Silicon Chip Australia's electronics magazine siliconchip.com.au Photo 8: after installing the top panel, I applied clamps liberally and waited for it to dry. You can see the exit of the port and the flush fit of the brace to the top panel of the port here. Fig.11: the two side panels are identical and have a central 5mm rebate (for the interior brace) and one at each end (where the front and rear panels will join). in the Subwoofer’s rear panel. This includes the following: ■ One Ultra-LD Mk.3 (mostly through-hole) or Mk.4 (mostly SMD) amplifier module ■ The Multi-channel Speaker Protector (with one channel used) ■ A 250-300W power supply ■ Heatsinking, switching and protection Refer to the August to October 2015 issues of Silicon Chip for details on the Ultra-LD Mk.4 Amplifier (siliconchip. au/Series/289); most of the construction information is in the September 2015 issue. The Speaker Protector we’re using was described in the January 2022 issue (siliconchip.au/Article/15171). The only change from those instructions is to install just one relay on the Speaker Protector as we are running it Fig.13: an internal side view of the finished Subwoofer without the side panels. siliconchip.com.au Australia's electronics magazine Photo 9: I sanded and primed the active Subwoofer, then sanded it again and added a few filler touch-ups to make the joins perfectly smooth. Photo 10: the Active Subwoofer with the final coat of “rattle can” black paint. It’s supposed to be satin but looks a lot like gloss. January 2023  65 Parts List – Active Subwoofer 1 assembled plate amplifier – see below 1 SB Acoustics SB34SWNRX-S75-6 346mm subwoofer driver [Wagner Electronics SB34NRX75-6] 1 2400 × 1200 × 18mm sheet of MDF or similar, cut as per Fig.6 100 50mm-long 8G wood screws (optional) 16 35mm-long 8G wood screws 30 28mm-long 8G wood screws 4 100mm diameter thick stick-on felt furniture foot pads 1 3m length of 5-10mm wide soft foam sealing tape (for the driver & plate amplifier) 1 1m × 1m acoustic wadding blanket [eg, Lincraft “king size thick wadding”] 1 250mL tube of PVA glue 1 tub of sandable wood filler 1 250mL tin of acrylic primer paint 1 350g can of spray primer paint 1 350g can of spray paint (for two or more top coats) 1 small tube of thermal paste large quantity of 120, 240 & 400 grit sandpaper (available on 5m reels) Plate Amplifier 1 assembled Ultra-LD Mk.3 or Mk.4 amplifier module on 200mm-wide finned heatsink ● 1 assembled 4-way Speaker Protector with a single relay (January 2022) ● 1 40-0-40 toroidal transformer, 250VA or 300VA [Tortech 0300-2-040] 1 screw-mount IEC mains input socket with integral fuse [Altronics P8324, Jaycar PP4004] 1 yellow insulated chassis-mount RCA socket [Altronics P0219] 1 miniature 250V AC 6A illuminated DPST rocker switch with solder lugs [Altronics S3217, Jaycar SK0995] 1 3-way mains-rated terminal block strip [Altronics P2130A] 1 5A 250V slow-blow 3AG fuse [Altronics S5685, Jaycar SF2232] 1 35V 400V bridge rectifier [Altronics Z0091A, Jaycar ZR1324] 4 8000μF 80V electrolytic capacitors [Jaycar RU6710] 1 10nF 63V MKT capacitor 1 270W 10% 10W wirewound resistor [Altronics R0440, Jaycar RR3369] ● PCBs and some other parts are available from our online shop. Hardware 4 M3 × 25mm panhead machine screws 16 M3 × 16mm panhead machine screws 10 M3 × 6mm panhead machine screws 2 M3 × 6mm countersunk head machine screws 2 15mm-long M3 tapped spacers 5 M3 flat washers 25 M3 shakeproof washers 5 M3 hex nuts 1 260 × 210 × 3mm aluminium sheet 1 377 × 150 × 1.5mm aluminium sheet 1 152 × 72 × 1.5mm aluminium sheet 1 20 × 38 × 1.5mm aluminium sheet (resistor bracket) 1 90 × 70mm sheet of Presspahn or similar insulation 4 blue 6.3mm insulated female spade crimp connectors [Altronics H2006B, Jaycar PT4625] 2 3.2-4.3mm solder lugs [Altronics H1503, Jaycar HP1350] OR 2 3.7-4mm crimp eye terminal [Altronics H1520, Jaycar PT4930] Wire & Cables 1 1m length of brown 7.5A mains-rated hookup wire [Altronics W2273, Jaycar WH3050] 1 1m length of blue 7.5A mains-rated hookup wire [Altronics W2275, Jaycar WH3052] 1 10cm length of green/yellow striped 7.5A mains-rated wire (stripped from a mains cord or mains flex) 1 2m length of red heavy-duty hookup wire (0.75mm2/18AWG) [Altronics W2270/83, Jaycar WH3040/45] 1 2m length of black heavy-duty hookup wire (0.75mm2/18AWG) [Altronics W2272/84, Jaycar WH3041/46] 1 2.2m length of green heavy-duty hookup wire (0.75mm2/18AWG) [Altronics W2274/85, Jaycar WH3042/47] 1 2m length of white heavy-duty hookup wire (0.75mm2/18AWG) [Altronics W2271/81] 1 30cm length of red medium-duty hookup wire [Altronics W2260] 1 30cm length of green medium-duty hookup wire [Altronics W2263] 1 40cm length of shielded/screened audio cable [Altronics W3010, Jaycar WB1500] 66 Silicon Chip Australia's electronics magazine from ±57V rails. Using only one relay halves the dissipation in the regulator, and we only have one channel to protect. I used a 3mm-thick panel of aluminium as the main plate for the chassis. To that, I mounted a folded bracket made from 1.5mm-thick aluminium for the transformer and an L-shaped panel for the speaker protector. Next month We don’t have enough space to fit the construction details of the internal amplifier for the Active Subwoofer in this issue. All the remaining construction details will be in the final article next month, concluding the series of articles on the Active Monitor Speakers. In the meantime, if you’re keen to commence construction of the High-Performance Active Subwoofer, you can gather all the parts in the adjacent parts list. You can then assemble the subwoofer cabinet using the instructions in this article. After that, you could assemble the Ultra-LD Mk.3 or Mk.4 amplifier module using the instructions in the August 2011 or September 2015 issue of Silicon Chip, respectively (but without installing the output devices yet). It would also be a good idea to build the Four-Channel Speaker Protector module (January 2022) but leave off one of the relays and the associated driving components. We only need to protect a single channel in this application. Do not install the driver in the cabinet yet, although you can prepare to fit it. That’s because you will need to install the acoustic wadding first (to be described next month). You will also need to connect a suitable length of heavy-duty speaker cable to the driver so that it can be connected to the yetto-be-assembled amplifier module. Next month, we’ll have instructions for building the bracket that the amplifier sits on and that the mains power supply is also mounted on it. The amplifier module sits on one side of the bracket, with the speaker protector next to it. The transformer, bridge rectifier and capacitor bank mount on the other side, making for a compact integrated amplifier module. On the rear of this module, outside the subwoofer cabinet, will be the amplifier heatsink, mains input socket, power switch & RCA signal input. SC siliconchip.com.au Subscribe to DECEMBER 2022 ISSN 1030-2662 12 The VERY BEST DIY Projects! 9 771030 266001 $1150* NZ $1290 INC GST INC GST 31 | Dual-channel BreaDBoarD PSu Take the mess out of prototyp ing new designs 48 | Jaycar Qc1938 oScilloScoPe Reviewing a 2-channel, 100MHz digital scope 62 | active Monitor SPeakerS, P Building the amplifiers and art 2 Active Crossover 76 | nrF5340 Dk DeveloPMent B An ARM-based module for oarD wireless communication 81 | Digital BooSt regulator Generate an adjustable voltage while multi-tasking Australia’s top electronics magazine 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 $65 $75 $50 1 year $120 $140 $95 2 years $230 $265 $185 6 months $80 $90 1 year $145 $165 2 years $275 $310 6 months $100 $110 1 year $195 $215 2 years $380 $415 All prices are in Australian dollars (AUD). Combined subscriptions include both the printed magazine and online access. Try our Online Subscription – now with PDF downloads! James Webb Space Telescope; December 2022 Digital Boost Regulator; December 2022 30V 2A Bench Supply; October-November 2022 Dual-Channel Breadboard PSU; December 2022 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 Regulation, 1960s style: Magnetic Amplification & Voltage Regulation No transistors or even valves are needed! This article describes how a transformer’s output voltage can be controlled using another two transformers, a potentiometer and two diodes. I recently carried out many bench experiments studying the subject of “magnetic amplifiers”. I studied some fascinating textbook concepts and methods of controlling voltages using laboriously self-wound toroidal transformers. For this particular article, I will stick to a practical theme: how some standard toroidal transformers can be used to regulate DC power (without going too much into the boring parts of the theory). I wanted to use components you can buy from places like Altronics or Jaycar, so anyone interested can easily replicate the design, whether just for a lab experiment or to make a power supply. This design delivers an adjustable 10-15V DC up to 12A without transistors, chips, microprocessors or circuit boards! We are firmly transported back to the 1960s, when silicon rectifiers were just coming onto the market, radios and computers were full of valves, and a telephone was made of black Bakelite with a rotary dial. A little bit of theory The simplest technique described in textbook literature for magnetic power control is the two saturated toroid arrangement, as shown in Fig.1. By Fred Lever Here, a pair of toroidal transformers are connected to an AC supply, with each handling one half-wave, gated by diodes D1 and D2. The power passing through the load windings (Ng) can be controlled by varying the bias on the control windings (Nc). Some very interesting waveforms are generated in doing this, as shown in Fig.2. In several separate experiments, I was able to reproduce these waveforms. The change in control bias level causes a similar change to phase control using an SCR or Triac. Fig.2(e) gives a bit of a hint of this. The curve of particular interest in the practical sense is Fig.2(g). This Fig.1 (above): the basic Magnetic Amplifier circuit, from page 457 of Benedict and Weiner’s book “Industrial circuits and applications” (see References section). Fig.2 (right): the expected waveforms in a Magnetic Amplifier, from page 458 of “Industrial circuits and applications” (see References section). 68 Silicon Chip Australia's electronics magazine siliconchip.com.au Fig.3: this is the transfer function I plotted from my experimental Magnetic Amplifier. While the control voltage range spans 80mV, the Jaycar toroidal transformers are much more sensitive, needing only millivolts to control the output over a range of amps. shows the change of load current in Ng with respect to the control (bias) current, Nc. This is similar to the bias transfer curve of electronic devices used in power control such as valves, transistors and SCRs. Fig.3 plots the transfer curve I achieved in my experiments using the hand-wound transformers. The load current could be controlled from near zero to maximum over a 40mV bias range, giving a similar result to Fig.2(g). The Jaycar transformers used in the circuit described below give a similar curve but with about twice the bias range for the 12A load. The gain of the cores is in the order of 10s of amps of load divided by milliamps of control, giving an effective gain in the thousands. sense. I could control the AC supply (a-c), over a range of current (iL) from almost zero to full load through a rheostat load, R. That was good enough to demonstrate manual control of an AC level of power by adjusting the bias level (d-c). One point of interest was if the cores were unloaded (with rheostat R open-circuit), the cores would act like air-cored chokes (no flux = no inductance) and lose control, producing the full output voltage. A hint to this is shown in both Figs.2 & 3; you will note that neither transfer curve reaches complete cut-off with no load. When you examine the circuits of industrial equipment, this situation never arises as there are extra bias Warning: Mains Voltage This project involves mains voltages which can be dangerous if not handled correctly. windings that provide a no-load flux. For simplicity, there are no extra windings in my power supply. Instead, an auxiliary circuit draws a constant current from the output, so the transformers always have a load. Fig.4 shows the circuit for a bench experiment using Jaycar MT2112 12-012V toroidal transformers. This shortform circuit can be used to verify the winding connections and draw a control transfer curve like Fig.3. The output of a 20V isolating transformer is applied to diodes D1 & D2. These drive the toroidal transformer load windings (the old secondary), which are connected in parallel series. The load winding centre tap feeds bridge rectifier BR1 to provide a Taking a practical approach The explanation of precisely what is happening inside the toroidal transformers is quite long-winded. Suffice to say that the saturating effect of the diode-guided feedback causes rapid changes in core flux that produce a ‘phase angle firing’ effect, resulting in a high effective gain. There are many books and online sources that you can peruse to understand this in more detail. An excellent mathematical treatment can be found in the paper by Brayton M. Perkins titled “Design of a self saturating magnetic amplifier utilizing high frequency excitation”, University of Arizona, 1956. You can download this from http://hdl.handle. net/10150/319332 Hooking up the Jaycar transformers on a piece of timber in the basic circuit shown in Fig.1 proved that they work using this scheme in a practical Fig.4: this is about the most basic circuit you can put together to test the Magnetic Amplifier principle. Besides the three transformers, two diodes and a bridge rectifier, you just need some meters, an adjustable load and a variable voltage to act as the bias source (which can be based on a bench supply). siliconchip.com.au Australia's electronics magazine January 2023  69 Supply Specifications Size: 420 x 265 x 200mm Weight: 15kg Output: 12-15V at up to 12A, 150W maximum Voltage regulation: ±5% Output ripple: 20mV at light loads, rising to 2V Power consumption: 300W pulsating DC output. Rheostat VR2 gives a variable load, with a voltmeter and ammeter connected to it. The colours of the transformer windings are shown in Fig.4. The control windings of the toroidal transformers (the old primary) are connected in series and to a DC lab supply of about 12V for bias. This needs to supply positive and negative bias voltages to swing the load windings over the entire range. As I only have a single polarity adjustable supply, I used a wirewound 100W resistor with a centre tap, plus wirewound 300W potentiometer VR1 to form a bridge-style circuit to give both polarities. During testing, the voltages applied were in the range of −1V to +1V at up to 500mA. The load rheostat I used (VR2) was rated at 500W and could handle load voltages up to 20V with currents of up to 20A. All the meters I used are true-RMS responding. Once set up, the output levels can be plotted against the control bias voltage. I adjusted the load rheostat to get 12V across it for loads applied in 1A steps by varying VR1. This gives a plot similar to Fig.3. If you can’t control the output as expected, that suggests a connection-­ phasing problem. There are 12 connections to the toroidal transformers that all need to be made correctly, as per Fig.4; one wrong connection will result in incorrect operation. Practical power supply design Moving on to the practical power supply, the short form circuit is expanded to include components to reduce the ripple on the DC output and provide the necessary controls and protection. The whole circuit is shown in Fig.5. All the required sections of a 1960s era supply are included in the finished design, specifically: 1) an unregulated source of AC 2) a power regulating control device 3) an independent reference supply 4) an error amplifier and correction signal 5) a rectifier & ripple control device 6) stability control to provide any transient damping and correct hunting Some wonderful old textbooks exist that clearly explain some of these points and are well worth reading, such as “Industrial Electronics” by Gullicksen and Vedder (1935); see the References section at the end of the article for more. Expanding on these: For this general-purpose bench supply, isolation from the mains is required, and a transformer with a nominal output of 24V AC at 12A can provide this. This sets the limit for the maximum output current of the supply. This transformer, T1, needs to provide a minimum of 20V AC at full load to give enough headroom for the power control device to deliver 15V DC. I used a transformer rescued from a discarded 300W UPS in this supply. Other transformers can be used, either toroidal or E-core, so long as they can supply the voltage and current required. #2 The regulation control devices in this supply are a pair of Jaycar MT2112 toroidal transformers, T2 and T3. This pair can handle about 20A of load current in this circuit arrangement, having an individual secondary rating of 12A with the secondaries connected in parallel. Each transformer handles onehalf wave of the AC power as guided by bridge BR1, so they are operated well within their ratings. Using devices that you can buy off the shelf removes the frustration of sourcing toroidal cores and copper wire, and the pain of winding them. #3 The reference supply could be any circuit that provides an adjustable 10-15V DC into a nominal 100W load. I experimented with various sources such as an independent bench supply, a battery of AA cells, a magnetic saturable reactor, a zener diode supply, #1 Fig.5: this more complete Magnetic Amplifier circuit gives a practical, usable adjustable voltage source for powering various circuits and doing things like charging batteries. While it has some limitations compared to the valve-based adjustable supplies back in the day, it has a certain elegance. Its simplicity means that such a supply would have been considerably cheaper to produce. 70 Silicon Chip Australia's electronics magazine siliconchip.com.au a 7815 regulator IC based supply and an unregulated 15V DC supply using a small Jaycar MT2002 transformer and a bridge rectifier. The most practical configuration that a home builder can easily reproduce is the last one, once again using readily-available parts. The stability and accuracy of this reference largely determine the performance of the overall supply. This corresponds to the portion of Fig.5 that includes T4, BR3 and VR1. It works well enough for many practical jobs, such as testing automotive 12V parts and supervised lead-acid battery charging. Performance could be improved with extra components to stabilise the voltage across the 2200µF filter capacitor, eg, a zener diode or an integrated regulator. #4 The error amplifier circuit in this supply is the simplest kind possible. The terminal voltage of the supply is applied to one end of the toroidal transformer control windings, +SENSE, and the reference voltage connected to the other end, +REF. Any differential between the two voltage levels causes a bias to be applied to the control cores. The reference supply is made variable from 10-15V, which becomes the panel control to set the voltage. The phasing of the control windings and the connection to the external circuits is critical; only the connection as shown on the circuit will work correctly. In a steady state, the differential voltage parks the toroidal transformers at a point on the transfer curve. With any disturbance such as moving the set voltage control or a change in load impedance, the differential voltage shifts the operating point on the curve and equilibrium is restored to suit the disturbance once the system’s time constant elapses. #5 The AC-to-DC rectifier, BR2, is a straightforward rectification bridge. An LC low-pass filter is formed using inductor L1 and a large 15,000µF capacitor provide ripple control. This is a more practical solution than just using a huge capacitor bank, with the benefit of a lower phase lag effect on the transient response, which could otherwise lead to instability. The capacitor used should be a proper low-ESR power supply filter capacitor. Up to about 10A can flow through it depending on the capacitor siliconchip.com.au The internals of the finished supply – it’s bulky but simple. The front panel and base plate are Earthed for safety, while mains-rated terminal blocks are used to make the connections. Australia's electronics magazine January 2023  71 Parts List For the test rig shown in Fig.4 1 24V output mains transformer, ideally at least 300VA (T1) 2 12-0-12V toroidal mains transformers (T2, T3) [Jaycar MT2112] 2 400V 35A bridge rectifiers (D1, D2, BR1) [Jaycar ZR1324] For the complete supply shown in Fig.5 1 24V output mains transformer, ideally at least 300VA (T1) 2 12-0-12V toroidal mains transformers (T2, T3) [Jaycar MT2112] 1 15V output mains transformer, ~15VA (T4) [Jaycar MM2002] 2 400V 35A bridge rectifiers (BR1, BR2) [Jaycar ZR1324] 1 400V 6A bridge rectifier (BR3) [Jaycar ZR1360] 1 20A+ diode (D1) [Jaycar ZR1039] 1 60mm 12V DC fan (FAN1) 1 12V lamp (LAMP1) 1 ~73mH 12A choke (L1) 1 20V FSD moving coil panel voltmeter [Jaycar QP5020] 1 20A FSD moving coil panel ammeter [Jaycar QP5016] 1 2A mains circuit breaker (CB1) 1 15A mains circuit breaker (CB2) Capacitors 1 15,000μF 40V 23A electrolytic power supply filter capacitor 2 2200μF 25V electrolytic [Jaycar RE6330] Resistors 3 100W 10W 10% wirewound (paralleled to give 33W 30W) [Jaycar RR3364] 1 39W 5W 10% wirewound [Jaycar RR3264] 2 47W 1W 5% carbon film [Jaycar RR2542] 2 150W 1W 5% carbon film [Jaycar RR2554] 1 100W wirewound potentiometer (VR1) value, choke size and load. Large standard electrolytic capacitors will work but will get hot and have a shorter life than a power supply capacitor. Any capacitor that does not have screw connections is not the best permanent choice. The capacitor I used was rated at 15,000µF, 40V DC with a ripple current of 23A. The 73mH, 15A filter choke I used is not an over-the-counter item at Jaycar! A functional unit can be wound using the stack of E and I laminations from a discarded transformer. My unit had around 40mm2 of core rated at about 120W, and I crammed 100 turns of 12A wire into the window to achieve 73mH. The inductance value is not critical; the trade-off is in physical size. I would have liked at least 250mH, but that would have taken a 300W size lamination stack and 150 turns of 15A wire. What I used is good enough for the job. I stacked the laminations interleaved but not air-gapped; the iron saturates on full load, giving the operation known as a ‘swinging choke’. 72 Silicon Chip The transient response of the supply is determined by the time lags inherent in the circuit. All of the gain in the comparator section is contained within the control toroidal transformers and is just sufficient to give millivolt-­level regulation. No anti-hunt or phase lead/lag techniques external to the comparator are needed to modify the transient response for stability. In addition to points #1-#6, a few other items are needed for a practical supply, such as terminals, meters, and overload or fault shut-off protection. In the supply described here, I included CB1, a 2A AC circuit breaker on the input that doubles as a power switch; CB2, a 15A AC circuit breaker on the DC output that doubles as a load switch; a panel light (LAMP1) to indicate life; and a pair of panel meters to indicate voltage and current levels. The meters can be just about any type that is available with the ranges required. You could replace the breakers with switches and fuses of similar ratings, but breakers are easier to reset. I used junked units from old electrical switchboards. #6 Australia's electronics magazine The components can be mounted in a cabinet that could be re-purposed or made from scratch. I mounted a heatsink inside that carries parts that will get hot such as the bridge rectifiers and the 33W ballast resistor. The heatsink can be anything made of metal of about the same size used here. Fan FAN1 is mounted in the cabinet to push air in over the heatsink and through the cabinet. This could be a 12V DC powered fan, or a mains-­ powered type around 60mm rescued from another device. General operation The mains is isolated and reduced to 24V AC at no load by transformer T1. Bridge BR1 may look to be connected strangely, but functions as a diode guide to gate toroidal transformers T2 and T3 with alternate halfwaves from T1. Toroidal transformers T2 and T3 are the power control devices, with their secondary inductance varied to regulate the output voltage. The cores need to have a high inductance on low load and a falling inductance as load current increases. This is accomplished by applying a bias current to the control windings (formerly primaries) of T2 and T3. Rectifier BR2 converts the controlled AC voltage to DC with a large ripple content, which is then applied to choke L1. This choke, combined with the 15,000µF capacitor, provides a low-pass filter to remove the 100Hz ripple. It is known as a “swinging choke” since it saturates as the load increases and its inductance falls to a lower value. Diode D1 is strapped across the outgoing rail to assist CB2 to trip if a reverse polarity is applied back into the output terminals, such as an incorrectly connected battery. The 33W resistor provides a minimum load to the toroidal transformers so that with no external load, some flux is generated in the toroidal windings, and start-up inductance is assured. The panel lamp and the cooling fan are also fed from this point to add to the minimum load current, resulting in around 0.5A. The cooling fan, FAN1, is run at a reduced voltage due to its series resistor. This limits the maximum voltage applied when the supply is set to 15V, especially as it has a high load ripple. Under this condition, without siliconchip.com.au the resistor, the fan coil could experience up to 18V. The fan runs at a slow speed on a low voltage setting and speeds up in proportion to voltage setting and load, with 12V applied when there is a high-current load and the output is set to 15V. Output ripple The output ripple level varies, going up as the load rises and is predominately 100Hz. With the values of L and C used, at 100Hz, the inductive reactance of the choke is about 40W and the capacitor reactance is about 0.1W. Thus, on a low load, the 100Hz component is attenuated by a factor of about 400 (40W ÷ 0.1W). The resulting ripple is in the 10s of millivolts. As the load current rises, the choke saturates and its inductance falls. This causes the loading effect at 100Hz to reduce and, at full load, its inductance is about 10% of nominal, giving a reactance of about 4W. The ripple attenuation factor is then approximately 4W ÷ 0.1W = 40 times, giving ripple levels of volts on top of the DC. This could be reduced by using a physically larger filter choke. If the reference voltage is lost or too low (<8V), the toroidal transformers may lose control and turn fully on. After the usual mains safety checks, the first power-up of the circuit can be via a reduced supply such as from a variac or with a light bulb in series with the mains supply. Applying power, you will note that the voltmeter swings up to the set voltage, and the circuit breakers should not trip. Then the supply is ready for testing. The voltage control should swing the output voltage between about 11V and 15V. Apply a load and the voltmeter will dip, then rise back close to the set voltage. Shed the load and the voltmeter will swing high momentarily and then settle close to the set value. If a short circuit is applied, the ammeter will smack hard over past 20A and then, depending on the tripping curve of the circuit breaker, a few seconds will elapse until it trips off. Supply waveforms Noting the difference between waveforms at no load and full load can give insight into how the control scheme works. No part of this supply circuit is connected to mains Earth except for the metalwork. Thus, an oscilloscope’s ground lead and probes can be connected anywhere on the low-voltage circuitry to examine the waveforms at any point with no fear of smoking Earth leads! Scope 1 shows the AC voltage from Powering it up Apart from getting the phasing of the toroidal transformer windings correct, there are no mysteries. If the connections are incorrect, the output might be the full uncontrolled voltage, a low voltage or just not work. If the 33W ballast resistor is not fitted, the transformers will simply operate like air-cored chokes and give the full output voltage. siliconchip.com.au The supply is mounted in a wooden cabinet. The heavy electrics are also bolted to an internal Earthed steel chassis. The cabinet is screwed to this chassis and this also secures the front panel. Australia's electronics magazine January 2023  73 Scope 1: the output of the mains transformer feeding this circuit is a distorted sinewave similar to the incoming mains waveform. Scope 2: the voltage across control transformers T2 & T3 under a low load condition. The spikes are due to core magnetic hysteresis. Scope 3: the waveform delivered to bridge rectifier BR2 when the output is not drawing much current. the isolating transformer, which has minor distortion. Scope 2 & 3 show the no-load AC voltage across T2 & T3 and at the junction of T2 & T3, respectively. Scope 4 & 5 repeat this but at full load. With no load, the toroidal transformer reactance is high, and a large portion of T1’s output voltage appears across them, with the remainder fed to the output. The toroidal transformer reactances are low at full load. Only a small voltage drop remains across them; the bulk of the sinewave is transferred to rectifier BR2. Scope 6 depicts the AC voltage across BR2 (yellow) and half-wave positive rectified output (cyan) with no load, while Scope 7 repeats this for the full-load condition. Scope 8 shows the output ripple (yellow) with no load, measuring ~60mV peak-to-peak, with the voltage across BR2 in cyan, while Scope 9 shows the same but at full load, giving about 4V peak-to-peak ripple. Scope 10 shows the transient response of the supply when switching from no load to full load and back at 1s/div and 2V/div. The voltage dip and overshoot is about 8V, with a recovery time of about one second. The voltage regulation on load is within the ripple level; an average-­ reading panel meter interprets this as a fall of 0.5V, while an RMS-­responding meter interprets as a drop of 0.25V. A peak-responding meter shows a rise of 0.5V (due to the ripple), so take your pick! Even just as a lab experiment, it would be prudent to mount the heavy isolating transformer T1 and control transformers T2 & T3 on a decent base like a sheet of MDF or plywood with an Earthed aluminium sheet adhered to the top for safety. As mentioned earlier, the phase and order of winding connections is critical. It so happened that the correct order of connections on my Jaycar transformers followed the notions of ‘starts’ and ‘finishes’ of the windings in order around the cores. The heavy windings are colour-coded as to where they start (a dot symbol on the drawing) and finish (no dot). The control windings (primaries) use all blue wires but emerge in a uniform order, from start to finish. Unfortunately, this means that while you can easily figure out how to wire them correctly in series, the polarity of the bias voltage connection is not obvious. So if the circuit doesn’t work as expected, the first thing to try is swapping the polarity of the control voltage to those windings. For convenience, bridge rectifiers BR1 and BR2 can be bolted to a piece of metal acting as a heatsink. Many of the winding connections join there, as you can see in my photos. You could choose to build the ‘lab exercise’ circuit shown in Fig.4 or progress to the power supply of Fig 5. This being a mains-powered circuit, you have to be careful how you wire it up to ensure it is safe. Follow my photos and ensure all the following steps are taken: • Use 10A mains-rated wire for all the mains connections in the correct colours: green/yellow striped for Earth, brown for Active and light blue for Neutral. • Insulate all exposed points at Active or Neutral potential with heatshrink tubing or similar insulating material (don’t use electrical tape except as a temporary measure). If using crimp connectors for the mains wiring, ensure they are appropriately sized and are the insulated type (or add heatshrink tubing over the top as insulation). • The incoming Earth wire must go straight to a substantial lug making good electrical contact with the metal base plate. Other Earth wires can run from this point to any other metal panels (eg, the front panel and/ Scope 6: the AC voltage across BR2 (yellow) with a light load, plus one half of the rectified waveform (cyan), taken from the positive side of the bridge only. Scope 7: the AC voltage across BR2 (yellow) at full load, plus one half of the rectified waveform (cyan), taken from the positive side of the bridge only. Construction advice This is more of an experiment than a project. Despite that, I have included a parts list (in case you want to try the experiment yourself) and some basic guidance on how to build such a supply. 74 Silicon Chip Australia's electronics magazine siliconchip.com.au Scope 4: the voltage across control transformers T2 & T3 at full load. They no longer drop much voltage across much of the mains waveform. Scope 5: the voltage applied to bridge rectifier BR2 at full load. This looks an awful lot like a Triac phase control waveform! or lid). There is no need to make Earth connections elsewhere on the supply, except perhaps if you wish to provide a front-panel Earth binding post. • Ensure that there are no exposed mains-potential metal contact points on any mains sockets or switches (including when the switch is in either position). • Use cable ties to connect mains wires together close to any connection point. This is so that if one wire breaks loose, it is held together with the rest of the bundle and can’t move to contact any low-voltage wiring or exposed metal. • Use a mains-rated terminal block, ideally bolted to the base, to connect the incoming mains wires to the mains transformer. Place a sheet of insulating material such as Presspahn, cut larger than the terminal block, between this and the Earthed base. • Ensure proper separation between all mains wiring and all isolated, low-voltage wiring. It’s best to keep all the mains wiring in a separate chassis section, away from the rest. One thing to note in my photos is the lack of cable ties on each side of the terminal block that joins the incoming mains wires to the transformer primary (I mounted this on a bracket attached to the base to save space). I corrected this omission after taking the photos. Scope 8: the ripple at the unit’s output (yellow) at no load, with the input of the LC filter (cyan). The yellow trace is less than 50mV peak-to-peak (p-p), while the cyan waveform is ~25V p-p. Scope 9: the ripple at the unit’s output (yellow) at full load, with the input of the LC filter (cyan). The yellow trace is around 2.3V RMS, while the cyan waveform at about 30V p-p. siliconchip.com.au Two different versions For the short form circuit (Fig.4), the bridge output can be terminated in the ballast resistor. You can then connect a suitable load bank (resistors or lights) directly to the rectifier output with flying leads. Connect measuring instruments (volt/ammeters) as needed. The reference supply can be a bench supply arranged so that voltages of either polarity can be applied to the toroidal transformer control windings. That may be all that some people wish to do to experiment. There is no reason that cheaper, lower-current toroidal transformers cannot be used for such a demonstration; the main advantage of the specified transformers is that it saves a lot of time and effort compared to winding your own. However, Fig.5 can be built into a working, practical supply, as shown in my photos. I expanded the floor plan to add in the filtering components and the reference transformer, then packed Australia's electronics magazine the rest into the rear of the enclosure and the front panel. Since most people who decide to build this version will have differently-­ s ized enclosures, it’s hard to give highly detailed assembly instructions. Look at my photos, decide how you can adapt the layout to your enclosure and start mounting and wiring the bits. Just make sure you follow the safety advice above. The physical size of the enclosure will depend on the parts used. I wound up with a 400mm wide unit with 240mm of depth and 200mm of height. The enclosure I made was a composite of plywood and steel sheets. The steel sheets are all connected to Earth wires for safety. The front panel has an angled metal section to carry the meters and voltage control, also Earthed. The terminals and circuit breakers are mounted on a ply section. The floor is a plywood sheet with a metal sheet laid over it, Earthed as described above. The heavy parts are bolted to the floor, with the remainder screwed to the rear of the front panel. The front panel is mounted on hinges, has the operator controls and load terminals and swings down once released by removing the plywood cabinet. All the essential details are shown in my photos. Happy experimenting! References 1. Book: Benedict and Weiner, 1965, “Industrial circuits and applications”, Prentice Hall, NJ. 2. Paper: Brayton M Perkins, 1956, “Design of a self saturating magnetic amplifier utilizing high frequency excitation”, University of Arizona (http://hdl.handle.net/10150/319332). 3. Book: Gullicksen and Vedder, SC 1935, “Industrial Electronics”. Scope 10: the transient response from light load to full load and back. The regulation is good, but there is more ripple on the output under full load, and the response time is slow (~0.5s). January 2023  75 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. Noughts & Crosses game using just two modules In the October 2021 issue of Silicon Chip, the Australian entrepreneur Dick Smith announced a competition to develop a Noughts and Crosses machine. Dick Smith mentions that he built a machine using electromechanical components in his announcement. I’d love to be able to build something like that, possibly using modern relays. But I’d need an enormously large number of them and they cost a few dollars each. I decided that something using an Arduino or Raspberry Pi would be the obvious way to do the job in 2021. More or less randomly, I decided to follow the Arduino path. An Arduino Uno and an Adafruit TFT LCD are an obvious combination; mine is a ‘plain jane’ implementation. Each of the cells of the Noughts and Crosses grid becomes a button to be pressed to indicate the move the human wants to make. My design includes an extra rectangular button at upper left to toggle who plays X to start each game. The Adafruit TFT LCD is available in two versions: V1 and V2. Adafruit supply and support only V2, but most clones are V1. If a clone is used, it is important not to use the Adafruit support library but a different library: MCUFRIEND_kbv. The Arduino Library Manager knows of this library, 76 Silicon Chip and it is easy to install, but it is necessary to do quite a lot of careful reading of fine print to persuade it all to work. I have not debugged my program for V1 clones, only for V2 TFT LCDs. I implemented the Newell, Simon strategy as published. Although I eventually got it to work, the descriptions of it are not easy to follow. I studied at least three and became thoroughly bamboozled by all three of them. There are eight components of the Newell, Simon strategy and seven of them are simple, obvious, and easy to understand. One isn’t. The problem strategy is described as “Block Fork”, but it tries to do two or maybe three jobs at once, and all descriptions have “unless...” qualifications. Because I struggled to understand the Newell, Simon strategy, I tested my implementation often. I found mistakes often. I’m reasonably confident that the program submitted never allows a human to create a fork and consequently always forces a draw, unless it wins. An Excel file is available to confirm this. The Adafruit examples demonstrating the touchscreen don’t mention debouncing, but it is essential. Each touch sends a stream of X and Y coordinates of many touches, not just one. Simple delays don’t flush the redundant points out of the FIFO. I read Australia's electronics magazine several examples to discover how to combine an appropriate delay while flushing the buffer to avoid unexpected touch events. Although the Adafruit library includes several font sizes, they must be integers, and I found that size 1 is too small but size 2 is too big. Ideally, I’d like size √2, but it is unavailable. For most text, I’ve used size 1 even though that is a bit small. I’ve used size 2 for the game’s name at the top of the screen and size 3 for the Xs and Os played within the board cells. A box at the top left advises who is X and who is O. This box is touch sensitive. If it is touched at any time, including during a game, it will toggle who is X and O, and start a new game, aborting the game in play if necessary. If a game is won, it is traditional to draw a line through the winning row, column, or diagonal. I don’t do exactly that; instead, I change the colour of the text in the winning cells to white. Note that the computer doesn’t attempt to vary its moves. After a few games, it becomes a bit predictable. The software sketch (NaughtyCross. ino) is available for download from siliconchip.com.au/Shop/6/92, along with the Excel spreadsheet demonstrating that it will play a perfect game. Keith Anderson, Kingston, Tas ($70). siliconchip.com.au MIDI Toolbox There is a lack of tools available for MIDI hardware developers. You can use a MIDI keyboard (controller) to generate MIDI messages, but what messages are being generated? For example, you are pressing a C, but what octave is it? What channel is it using? What commands are sent when you turn a knob or press a pad? Most low-cost MIDI keyboards are designed to connect directly to computers using USB and are unsuitable for hobbyists who want to work with the old current loop MIDI system. The keyboard is unlikely to generate every command you want, as the keyboard siliconchip.com.au I was using could not switch to a different patch. Also, some keyboards are very large and will not fit on most workbenches. This project addresses those problems and includes a sequencer, providing the MIDI equivalent of a function generator. On the hardware side, the circuit acts as an adaptor that allows modern MIDI keyboards to work with the old current loop system. It includes a HobbyTronics USB Host adaptor (www.hobbytronics.co.uk/usb-hostboard-v24 – make sure you order MIDI firmware option) that powers Australia's electronics magazine a keyboard and behaves like a PC while sending the MIDI commands to a serial port. Switch S1 allows a traditional MIDI device to be connected using either the 5-pin DIN connector or the more modern 3.5mm stereo ‘TRS’ plugs. The MIDI commands are sent to 5-pin DIN and 3.5mm stereo sockets. The square wave output is handy when debugging as it avoids the need for extra hardware to generate sounds. There are also 1V/octave, trigger (1ms) and gate outputs that allow connection to traditional analog synthesiser hardware. January 2023  77 The circuit is based on the Micromite Backpack V3 with a 3.5-inch (9cm) touchscreen LCD screen (August 2019; siliconchip.au/Article/11764). The PCB and display are both available from the Silicon Chip Online Shop, as is a complete kit (siliconchip.au/ Shop/?article=11764). The logic-level MIDI signal is fed into I/O pin 22 and, after processing, comes out of I/O pin 21. It is then buffered and fed to the MIDI output connectors. Only the PIC chip, voltage regulator and associated components need to be installed on the BackPack PCB as this project uses none of the other features. The LCD panel can be set to maximum brightness. The current loop MIDI input is converted to TTL level voltages using a 6N138 optocoupler with a standard circuit. The optocoupler’s output is connected to the Micromite through switch S1, selecting between the current loop input and the USB Host. The COM1 serial output is buffered by two 74HC04 gates and connected to the MIDI output via 220W resistors. The gate and trigger outputs are buffered by 74HC04 gates to create TTL levels required by analog synth hardware. The output at pin 4 of the BackPack is buffered by a non-inverting op amp to provide the audio output. Any 5V single-supply dual op amp can be used, although the LMC6482 is suggested as it is readily available. The output at pin 26 provides a 50kHz PWM signal proportional to the MIDI note at 1V/octave. This is filtered by a 100Hz low-pass filter and buffered by an op amp to create the 1V/octave output. The circuit is powered by a 5V plug pack and draws about 250mA when powering a small MIDI keyboard. When the device starts, you are presented with a menu with four options: Analyser, MIDI Commands, Sequencer and Keyboard. Analyser is a protocol analyser that captures the raw MIDI data and decodes the message into something like “Note on C2 44”, meaning the C2 key was pressed with velocity 44. The analyser suppresses system messages, so you only see the actions you are generating. MIDI messages are passed through to the MIDI output in this mode. Press the screen to return to the menu. MIDI Commands allows you to generate arbitrary MIDI commands. Press Edit, enter the hex codes for the command and press Save, then press the button where the command should be stored. You can also assign a colour to each preset. You can set commands for all 20 presets and save them to flash by pressing Save. MIDI commands are passed through in this mode, so this feature can work as an add-on for an existing keyboard. The Keyboard mode displays one octave of a keyboard. Pressing the keys will play the notes through the MIDI output. You can use the Oct Down and Oct Up buttons to select any octave, and there is a button that allows you to choose the MIDI channel. The notes are also sent to the audio output, which generates a square wave at the correct pitch for each note. Sequencer allows up to 64 steps to be captured and will automatically loop through that sequence. There are three layers that output to three different MIDI channels, and up to three notes can be captured for each layer. The tempo is adjustable from 30 to 800 beats per minute. The sequencer ignores the velocity component of captured notes but allows you to control the velocity output as Flat, Slow or Fast. “Flat” produces all notes with a velocity of 127. “Fast” plays notes as quiet, medium and loud in groups of three, while “Slow” increases the velocity in six steps in groups of six. This is useful when testing sounds as it allows you to hear the same note at different velocities in sequence. There is also an arpeggiator that splits three-note chords into arpeggios and supports all possible orders of the notes. This works with percussion to produce interesting beat patterns. When you press Quit, the sequence is saved into flash memory. To record a sequence, press the Record button and begin playing on the keyboard. It captures approximately one note per second, then advances to the next note. You have 200ms from when the first message is received to capture the notes in a chord, and it sorts the notes from lowest to highest. This is necessary to give some room for fingering errors and get consistent sound from the arpeggiator. If you pause for one second, it will insert a pause. If the keys are held for one second, a tie will be inserted to create a longer note. Press Record again to stop recording, and it will calculate the sequence length as the longest of the three layers. Change the layer to record the next layer; the first layer will play quietly as you record the next layer. There is a Clear button to clear a layer, or you can use the touchscreen cursor controls to move to an error and begin re-recording from that point. Dan Amos, Macquarie Fields, NSW ($100). Screen 1: Analyser mode captures and decodes raw MIDI data, while MIDI Commands generates arbitrary commands via hex. Screen 2: the Sequencer screen captures up to 64 steps, which can then be looped. Screen 3: the Keyboard screen for the MIDI Toolbox; the buttons at upper left are to change the current octave. Australia's electronics magazine siliconchip.com.au 78 Silicon Chip Software functions Keep your electronics safe with our HUGE RANGE of Low Voltage Circuit Protection SAME GREAT RANGE AT SAME GREAT PRICE. ESPECIALLY HANDY IN A BOAT OR CARAVAN WHICH MAY NOT HAVE A CHASSIS GROUND TYPE SYSTEM CONTROL LIGHTING AND OTHER 12V GEAR AROUND THE BOAT & VEHICLE Blade Fuse Blocks with Bus Bar • Suits standard blade fuses • 30A per output, 100A combined • Red LED blown fuse indication • Negative bus bar 6 Way SZ2031 | 12 Way SZ2032 FROM 2995 $ PERFECT FOR HIGH END CAR AUDIO SYSTEMS, SOLAR INSTALLS AND OTHER HIGH CURRENT 12V APPLICATIONS Heavy Duty Circuit Breakers • Visible Trip / Circuit Breaker Button • Switchable / Manual Push-To-Trip Operation • Multi-wire gauge inputs and outputs 60A SZ2081 | 120A SZ2083 | 200A SZ2085 • 10A rated illuminated rocker switches • Push-to-reset 10A, 8A, 6A and 4A circuit breakers • Negative bus bar • Mount vertically or horizontally 4 Way SZ1902 6 Way SZ1903 JUST 46 $ Marine Switch Panels with Circuit Breakers FROM 4495 $ 95 EA Here's a small selection of our Fuses and Holders See our HUGE RANGE in-store or online SEE OUR STAFF IN-STORE FOR THE FUSE OR CIRCUIT BREAKER TO SUIT YOUR APPLICATION ANL In-line Fuse Holder • Suits ANL Wafer Fuses • Nickel plated • Protective cover SZ2078 JUST 14 $ 95 Gold ANL Wafer Fuses • 80A, 100A, 150A, 200A & 250A models • Screw down contacts SF1990-SF1999 JUST 7EA $ 95 Automotive Fuse Assortment • 20 x 5A, 10A, 15A, 20A, 25A & 30A ATO size fuses included • 120 pieces with storage case SF2142 Shop at Jaycar for Mains Voltage Circuit Protection: • Surge Arrestors • Transient Voltage Suppressors • Fuses and Fuse Holders • MOV & Noise Suppressors • PTC Fuses • Thermal Fuses and Switches Explore our full range of circuit protection products, in stock at over 110 stores or 130 resellers or on our website. jaycar.com.au/circuitprotection 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. JUST 2895 $ Part One by Dr Hugo Holden Play your own game of Noughts × Crosses This Noughts & Crosses-playing Computer (also known as Tic-Tac-Toe) has no processor, clock signals, registers or latches. It merely uses logic gates and a ROM that behaves as a complex array of gates. The user interface is simple but elegant: plastic chips printed with Xs and Os are placed in one of nine recesses on the game board and the Computer signals its move by lighting LEDs on the appropriate spot. I n this type of system, the only computation delay is the propagation delay of the gates or logic devices. Since this Computer responds to static logic conditions, it can never get confused or out of step with itself, or lock up. While speed is not important for this application, a low current draw is. Current consumption increases with clocking frequency in clocked systems. This makes a CMOS-based static computer extremely attractive for battery operation and low power consumption. The current consumption from the 9V DC power source varies between 75mA and 90mA, most of which is for lighting the LEDs. This design does include an oscillator, but it is not used for any computation. Instead, it behaves like a random number generator (RNG) using a ‘spinning wheel’ technique described below. Before tackling this design problem, I decided it would be good to consider how two people play the game. The person who starts the game 80 Silicon Chip first has a significant advantage over the other player, so alternating which player starts first is required to average that bias out over numerous games. That means the machine needs to be able to make the first move, and it also needs to allow the human player to start first. A predictable player might start the game the same way every time, for example, starting on the central square to gain maximum advantage. But that sort of predictable behaviour soon gets very dull, so the Computer should vary its opening strategy when it is the first player. When a human wins the game, they likely announce it with great enthusiasm, so the Computer needs a way of alerting the human player when it wins. Finally, two humans playing each other would be imperfect to the extent that sooner or later, one might make an error of judgement. This would allow the other human, who didn’t make any errors, not just to prevent the other Australia's electronics magazine human from winning (causing a draw) but to beat the other player. A proper and complete Noughts & Crosses machine would not only be unbeatable by the human player, but it should be able to win against the human at every opportunity. This requires the analysis of every possible mistake the human could make during gameplay. In light of the above features, I decided that the way to design the machine would be to initially create a two-player board game. Each player could place a disc, with an X or an O label on it, in the player area. This board would work fine even with no electric power available, and two players could enjoy the game together as usual. However, if one of the human players ‘goes missing’ and the game is powered, the machine steps in to replace one human player. It then becomes a human vs machine scenario. The machine must be able to perform the functions that a ‘flawless’ siliconchip.com.au human player would possess: never make any mistakes, never be beaten, but also win when given the opportunity. The game is configured as a player board, with X and O player pieces, but the machine only has a ‘brain’ and not eyes and arms. So the machine asks the human player to place the machine’s discs on the board by lighting the LED where the machine wants its disc placed. Note that in this design, the machine always plays as O and the human as X. The board can sense the presence of an X or an O disc on the player board. The computer ‘knows’ where on the board each is, and it ‘knows’ when it is the machine’s or the human’s turn to move. When it is the machine’s turn to make a move, the move is computed in under 200ns. After analysing the board pattern of Xs and Os, it lights LEDs on the board where it wants its O piece placed. The maximum number of X player pieces that can be applied to the player board when X starts first is five, limiting the number of O discs that can be placed to four. And that the maximum number of O discs when O starts first is limited to five, thereby limiting the number of X discs to four. This means there is either an X disc or an O disc left over, depending on whether the human (X) or the machine (O) started the game. Therefore, an extra space is provided to store the unused disc. This space also acts as a bipolar electrical switch to configure the computer circuitry for who starts first. This is not only convenient but it also avoids the need for any mechanical switches. If X (human) is to start the game first, an O disc is placed in the spare space, but if O is to start first, an X disc is placed in the extra space. This instruction is engraved onto the player board surface along with the fact that the human plays the X pieces and the machine plays the O pieces. When the game is initially powered, with no discs placed anywhere on the board, all the LEDs are lit. This represents the ‘start randomiser’ function. The LEDs on the board are actually rapidly lighting up in sequence, one at a time. This also serves as the LED test function, similar to how it was once customary to briefly light all lamps on an instrument at switch-on. siliconchip.com.au If an X disc is placed in the spare space (meaning the machine starts first), the ‘spinning wheel’ stops and locks in the first move for the Computer. The LEDs that remain lit show the random position for O’s first move. When the game’s lid (top hinged cover) is closed, it safely stores all the playing pieces (discs) inside, so they do not get lost or separated from the game. The prototype game is powered from a 9V plugpack, but since the current consumption is low, it could be powered from a 9V battery or battery pack. You can watch a short video of the machine in operation at https://youtu. be/IE9a5ZJZCgE Circuit design I was inspired by the fact that Dick Smith built a noughts & crosses machine from parts from a telephone exchange in 1958. Most likely, those parts were vintage at the time; most exchange spare parts then dated to the 1930s. I decided it should be possible to do something similar using logic gates. I’m very fond of 74-series logic gates and commonly use 74xxx (TTL) or 74LSxxx (low-power schottky logic) types. There are also CMOS versions like the 74HCTxxx series. These perform the same logic functions with lower power consumption, so I chose them for this project. I also used blue LEDs as they are very energy efficient. The circuitry for the Computer is spread across two PCBs, a ‘game board’ with all the user interface parts (Hall Effect sensors, LEDs etc) and a ‘compute board’ which has all the control circuitry. To analyse the gameplay patterns and make the correct responding move, I am using an EPROM or EEPROM. These need 18 address lines to process the player board logic. The circuit uses ‘parity’ information from the playing board to control the Computer’s action, depending on who starts first. The AT27C020 from Microchip is a suitable EPROM that comes in PLCC The underside of the Noughts & Crosses Computer has a clear acrylic lid which lets you peer through and see the two main PCBs. Australia's electronics magazine January 2023  81 82 Silicon Chip Australia's electronics magazine siliconchip.com.au Fig.1: the game board circuit is large but quite simple as it consists of repeating patterns. The nine Hall Effect sensors under the game board (HS1-HS9) each have four associated LEDs with current-limiting resistors. A pair of op amps are also assigned to each of the ten Hall Effect sensors, acting as window comparators to detect whether an X or O piece is present (or neither). siliconchip.com.au Australia's electronics magazine January 2023  83 or DIL packages. The compatible Winbond W27C020 EEPROM is readily available and low in cost. I used a similar 27C020 UV-erase EPROM in my prototype as I had one on hand, along with the required programming equipment. Fun with (weak) magnets Wanting to avoid mechanical switches and not having a microcontroller to support a touchscreen interface, I decided to use ferrite magnets from Jaycar embedded in the player pieces, along with ratiometric (analog) Hall Effect sensors for user input. The ferrite magnets (Jaycar Cat LM1616) come in a packet of 12, and I used 10, five for the Xs and five for the Os. I cut a 14.5mm diameter, 4.5mm deep hole into the bottom of each 25mm diameter, 10mm thick plastic disc. I then glued the magnets into them with 24-hour epoxy resin. I deliberately did not use powerful neodymium magnets as their magnetic fields would be too strong and could interfere with nearby pieces. They also have a habit of jumping to each other or the nearest magnetic object and magnetising it, and are good at accidentally erasing magnetic media. While I used weak magnets, their flux field is above what the Hall Effect sensors require. Fig.1 is the circuit diagram for the game board. The 3×3 grid of Hall Effect sensors labelled HS1-HS9 and the four LEDs associated with each form the game board. The tenth Hall Effect sensor, HS10, controls which player starts first, as described above. Six LM324 quad op amps (a total of 24 individual amplifiers, four unused) process the outputs of the Hall Effect sensors. With no applied magnetic field, the DC output of each Hall Effect device sits within 70mV of 2.5V. The X and O player pieces have magnets glued inside them in opposite orientations. Placing an O disc causes the output from the Hall sensor to go low (below 2V), while placing an X disc on the sensor causes the output to go high (above 3V). The op amps are used as comparators to produce a logic 1 (high level) at Data X or O when an X or O player piece is placed on the board respectively. LM324s are handy for this job because their output voltage swing, when powered from 5V, is a perfect match for TTL logic levels. The general arrangement for each sensor is shown in Fig.2; this is duplicated 10 times on the game board. The 2kW/1kW/2kW resistive string across the 5V supply generates 3V and 2V reference levels. Two of the LM324 op amps are used as a window comparator. The output of one goes high when the Hall Effect sensor is producing more than 3V and the other’s output goes high when the sensor produces less than 2V. As shown in Fig.3, four LEDs are arranged around each sensor for two Fig.2: this shows a pair of Hall Effect sensors as in Fig.1 in isolation. Two are shown as they share a single quad op amp. Whether the DATA X or DATA O line goes high when a magnet is placed near the sensor depends on that magnet’s orientation and thus the magnetic field’s polarity. The 100nF bypass capacitors were unnecessary in the prototype, so they are not shown in Fig.1. 84 Silicon Chip Australia's electronics magazine reasons. One is that the resulting symmetry gives a pleasing appearance, and the other is redundancy. If one LED fails (which can happen), the game can still function normally. Each LED in the group of 4 has its anode connected to +5V with a separate cathode resistor. The 1.6kW current-limiting resistors result in around 1mA through each when they are on, but being blue types, they are pretty bright at this low current level. The manufacturers of the Hall Effect devices recommended bypass capacitors across their supply pins, so I added these as surface-mount ceramic types between the solder pads. Still, the devices appeared to work fine without them, probably because this Computer is static, so there are no switching transients on the power rails. The game board connects to the compute board via a 40-way SIL connector (socket on the game board and header on the compute board), avoiding a large mass of wires. Hall Effect options Several different versions of the Hall Effect sensor are ‘available’. I put that in quotes because, as is typical these days, most of them are out of stock. The A1 version is the most sensitive and that is what I used. Unfortunately, it is not in stock anywhere at the time of writing this, although hopefully, that will change shortly after publication. Fig.3: the four LEDs are arranged around the periphery of each ‘well’ on the game board to produce a pleasing symmetry. Each has its own currentlimiting resistor that sets the LED current to around 1mA. That results in the blue LEDs being quite bright without using much power, even if all 36 are lit. siliconchip.com.au The A2 version is easier to get, but has half the sensitivity of the A1 version. Luckily, it’s pretty easy to compensate for this by changing the single 1kW resistor in the 2kW/1kW/2kW string to 510W. That will make the window comparator thresholds 2.25V and 2.75V instead of 2V and 3V, compensating for the reduced output swing from the Hall Effect sensors. Compute board Fig.5 is the circuit of the compute board that uses all through-hole components. No electrolytic capacitors are used in this design, only film types; this makes the circuitry long-lasting. While not shown in that diagram, there are test points for all the EEPROM address lines (A0 to A17) on the board. They were handy for confirming during development that the game board was working correctly. The player board must recognise any possible pattern or combination of human X and machine O player pieces. It must distinguish, at each of the nine locations, if an X is placed there, an O placed there, or no piece at all is placed there. To convert this information into a binary format suited to computer evaluation, a different piece placed at each location generates a binary number, as shown by the blue and black numbers in Fig.4. These values are summed to produce a single 18-bit number representing the state of the board at any given time, which is used as the EEPROM address to look up the next move. For example, if an O piece was placed in the centre (square 5), this generates a decimal value of 8192. Then if an X piece is placed on the board at location 07, this generates decimal value 64. Therefore, for this simple state of two pieces on the board, the address produced is 64 + 8192 = 8256. It is at that location in the EEPROM where O’s next move is stored; in other words, the EEPROM produces a value (by reading the memory at that address) to light the appropriate LED where the machine’s next O piece is to be placed. No player pieces being present on the board generates an address of zero, and the EEPROM has a value of 255 (0xFF hexadecimal, 11111111 binary) stored at that location. The lower four bits all being one means that there is no output from 74HCT42 decoder IC2, as it is an invalid code, so no LEDs are lit by that decoder in that state. Instead, this value triggers the initial move randomisation process. After this, if an X player piece is placed on the spare space, a randomly selected square will remain lit for the first O piece to be placed, corresponding to the count that IC7 stopped on and the four-bit value presented to IC6. When pieces are on the board and it is the machine’s turn to make a move, the game’s electronics go into ‘compute mode’. The outputs of the EEPROM are enabled by the COMPUTE control line going low, which connects to pin 22 of IC1. The EEPROM’s outputs are activated, resulting in the appropriate LEDs lighting to show where the machine wants its piece placed. The rest of the time, when the COMPUTE control line is high, the EEPROM outputs are tri-stated (open The Noughts & Crosses Computer uses a set of 10 magnetic pieces to play the game, with one piece determing which player goes first. Fig.4: each possible playing position is assigned a different power-of-two value depending on whether an X or an O is placed there. With 18 possible values, those numbers can be used to address a 218 = 256KB EPROM or EEPROM. The numbers stored at those addresses tell the machine what move to make next (1-9 to place a token in a given square or 0/255 for no move). siliconchip.com.au Australia's electronics magazine 85 circuit) and pulled up by four 10kW resistors. This means that the BCD to decimal decoder, IC2 (74HCT42), is presented with all ones (due to the pull-ups), so no LEDs are lit. 86 Silicon Chip The start randomiser Imagine a spinning wheel with the numbers 1 to 9 on it. It’s spinning so fast that it’s just a blur. If you throw a dart at it or stop it abruptly with a Australia's electronics magazine brake, one of the numbers would be selected in an apparently random manner. A similar task falls to IC7, a 74HCT161 counter. It is clocked at siliconchip.com.au Fig.5: the compute board contains the logic that decides which move to make next and whether the machine player has won. Each section described in the text is highlighted and labelled in a different colour to aid in the understanding of the overall circuit. All connections between the circuits of Fig.1 & Fig.5 are made via 40-pin SIL headers CON1 & CON2. around 3kHz by an oscillator formed by IC3a with a 5.1kW feedback resistor and 100nF capacitor. When IC7 reaches a count value of nine, output pin 11 of IC3d goes low, which loads siliconchip.com.au the counter with the value of one, so on the next clock pulse, the counter resets to one. When an X piece is placed on the spare space on the board (signifying Australia's electronics magazine machine is to start first), the oscillator output at pin 3 of IC3 is inhibited. Therefore, a count between 1 and 9 remains, lighting the LEDs on that square on the board. January 2023  87 Notice how I assigned the EEPROM outputs shown in Fig.4 values 1 to 9, not 0 to 8. This is because the O-start randomiser is generated by IC7, a 74HCT161 binary counter which needs to make nine counts, from 1 to 9, and cyclically reset to 1. But it also needs another unique state of zero, when it is reset and does not light any LEDs. The four-bit output of IC7 (Q0-Q4) goes to 4-to-10-line decoder IC6 and it, in turn, lights the appropriate LEDs via diodes D11-D19, so it doesn’t conflict with IC2’s control of the LEDs. If IC6 is presented with an input value of 0, its output lines 1 to 10 remain high, and no LEDs are lit by it. One might wonder why I did not use the count enable pins PE & TE on IC7 to inhibit its counting, rather than stopping the clock. The reason is that these inputs should not be toggled when the clock pulse is low, and there is no synchronisation between the moment when the clock is stopped by the human placing the X on the spare space and the state of the clock pulse at that moment. Fig.6: this demonstrates how the same board state can be achieved by two different games, one of which starts with the human player (X1, at top) and the other starts with the machine player (O1, at bottom). The numbers indicate the sequence of the moves, while the Xs and Os show which player places a token in which square. 88 Silicon Chip While the particular LEDs lit by the output of the randomiser, when it has stopped, remain on for the remainder of the game, the first O piece is placed over them, so they are not visible. If the human player (X) starts first, an O player piece is placed on the spare space. This causes pin 8 of IC3 to go low, clearing the 74HCT161 counter to zero, so the randomiser lights no LEDs. More LEDs are only lit after the first human X piece is placed, to satisfy the machine’s next move. Game sequencer using parity The Computer only should go into ‘compute mode’ when it is the machine’s turn to make a move (place an O piece). Data patterns generated by the encoder board, just after a selected O has been placed, have no meaning to the Computer because the human player has not placed their next X piece yet. The question of ‘when to let the computer compute’ has two answers, depending on who starts the game first. In the case of the machine starting first, there is initially the one O piece on the board (selected initially by the randomiser), then an X is placed by the human. Now there are two pieces on the board, an even number, and it is time for the machine’s next move. After the machine’s move, the total number of pieces goes to three, an odd number, and it should not act. However, if the human starts first and places their X piece, one is an odd number; in this case, it is time to compute to determine the machine’s move. Once the O is placed, the number of pieces becomes even, and now the Computer waits because it is X’s move again. X plays again, and the number goes odd, putting the Computer into compute mode. I realised that I could solve this problem using a parity IC, with a data selector on its two outputs, to select either the odd or even outputs of the parity IC to control the COMPUTE line. Nine sections from three quad 2-input OR gates (IC13, IC15 & IC16) combine the X and O lines from the game board into nine ‘piece present’ signals that then go into the parity chip, IC9 (74HCT280). Its EVEN and ODD outputs go to quad 2-input NAND gate IC5, along with the HS10x and HS10o signals that indicate which player started first. This allows IC5 to generate the Australia's electronics magazine The playing pieces only use ‘weak’ magnets. COMPUTE signal, which comes from the pin 6 output of IC5. To summarise, it depends on whether an even or odd number of pieces are on the board, and who started first. Game state ambiguity Note that in the case that O (the machine) starts first and no pieces are on the board, the COMPUTE line is low, enabling computation. However, with no pieces on the board, all the address lines are zero, and as mentioned earlier, the data at address 0 in the EEPROM is hexadecimal 0xFF. When the 74HCT42 is presented with all four bits high, it will not light any board LEDs. Something to consider is that since the ROM is programmed only to produce valid output values with valid input values that correspond to an achievable pattern of player pieces on the board, why is it necessary to have the game sequencer circuitry at all? Invalid addresses/states would/could result in an output of 0xFF and therefore, no LEDs would be lit anyway. Indeed, you would not need the sequencer circuitry if the game were designed for one of the players (human or machine) to always start first. Fig.6 shows two possible identical patterns that could be achieved through different game sequences. Therefore, these generate the same address for the EEPROM. In one case, X started first, while in the other case, O started first. On the next move, the 5th piece placed could be an X or an O, depending on who started the game. This is why the game sequencer with the parity IC was required, as it gives a different state for the COMPUTE line in these two cases, despite the EEPROM address being identical. Game sounds When the machine wins against the human player, it sounds a beep. There is an oscillator to drive a piezo buzzer in the circuit (based around 555 timer siliconchip.com.au IC4). However, in my prototype, I used a beeper with an inbuilt oscillator, element14 Cat 107-2397, so I bypassed the oscillator. No beep functions are assigned to the Human player because the human can make a sound themselves if they want. They might, especially when the machine not only prevents them from winning but it insults them further by beating them with the slightest error or lack of attention. Since the human can never beat the machine, an automatic announcement for the human player winning is not required. Therefore, it is only necessary to examine the O piece data from the game board (not the X data) for the eight possible configurations of O alignments that indicate a win. Parts List – Noughts & Crosses Computer In the second and final instalment next month, I’ll explain the gameplay strategy and how I generated the gameplay data for the EEPROM. I will then go over the PCB assembly process, case construction and putting it all together into a working game. SC 1 set of parts to make the enclosure (see below) 1 double-sided PCB coded 08111221, 138 × 166mm (‘game board’) 1 double-sided PCB coded 08111222, 138 × 124mm (‘compute board’) 1 40-pin header socket, 2.54mm pitch (CON1) 1 40-pin header, 2.54mm pitch (CON2) 1 piezo buzzer or sounder, 7.5mm lead pitch 1 32-pin DIL IC socket (optional; for IC1) 3 16-pin DIL IC sockets (optional; for IC2, IC6 & IC7) 16 14-pin DIL IC sockets (optional, for IC3, IC5, IC8-13, IC15, IC16 & IC21-26) 1 8-pin DIL IC socket (optional; for IC4) 4 10mm M3 tapped spacers 8 M3 × 6mm panhead machine screws 1 100mm length of 0.7mm diameter tinned copper wire or bell wire Semiconductors 1 W27C020 high-speed, low-power 256KB EEPROM or equivalent, DIP-32, programmed with 0811122A.bin (IC1) 2 74HCT42 BCD-to-decimal decoders, DIP-16 (IC2, IC6) 2 74HCT132 quad 2-input NAND gates, DIP-14 (IC3, IC5) 1 555 timer, DIP-8 (IC4) 1 74HCT161 4-bit presettable counter, DIP-16 (IC7) 1 74HCT30 single 8-input NAND gate, DIP-14 (IC8) 1 74HCT280 9-bit parity generator, DIP-14 (IC9) 3 74HCT10 triple 3-input NAND gates, DIP-14 (IC10-IC12) 3 74HCT32 quad 2-input OR gates, DIP-14 (IC13, IC15, IC16) 6 LM324 quad single-supply op amps, DIP-14 (IC21-IC26) 10 DRV5055A1QLPG linear hall-effect sensors, TO-92 (HS1-HS10) 1 7805 or LM2940CT-5 (see text) 5V 1A linear regulator (REG1) 1 BS270 or 2N7000 N-channel Mosfet, TO-92 (Q1) 36 blue 3mm LEDs (LED1-LED36) [Jaycar ZD0134] 1 1N5819 40V 1A schottky diode (D1) 19 1N4148 75V 250mA small signal diodes (D2-D20) Capacitors 2 1.5μF 50V MKT or multi-layer ceramic 3 1μF 50V MKT or multi-layer ceramic 3 100nF 50V MKT or multi-layer ceramic 1 10nF 63V MKT Resistors (all 1/4W 1% axial) 1 470kW 6 10kW 2 5.1kW 36 1.6kW 2 2kW 2 1kW Enclosure 1 machined and engraved lid made from 10mm-thick acrylic, 160 × 200mm 1 machined and engraved top panel (10mm-thick acrylic), 160 × 200mm 1 160 × 200mm sheet of 3mm thick smoked translucent acrylic (for base) 2 200 × 40mm sheets of 10mm thick acrylic (side panels) 2 140 × 40mm sheets of 10mm thick acrylic (front and rear panels) 1 150mm hinge 2 small lid latches/clasps 4 screw-on rubber feet 20 10mm-long countersunk hex socket cap head 4-40 UNC machine screws (for attaching the base & top panel) 10 10mm-long hex head 4-40 UNC machine screws (for clasps & feet) 8 10mm-long countersunk hex socket cap head M2 machine screws (for attaching hinge) 8 10mm-long, 4mm diameter M2-tapped metal inserts (for hinge) 4 4-40UNC hex nuts (for attaching feet) 1 chassis-mount barrel socket OR 1 9V battery holder OR 1 6 x AAA cell holder (CON3) 1 200mm length of light-duty figure-8 wire Playing pieces 10 25mm diameter, 10mm thick black plastic discs 10 weak ferrite magnets [Jaycar LM1616] siliconchip.com.au Australia's electronics magazine Power supply options I was unsure whether the circuit current draw might exceed 100mA, so I built it with a 7805 regulator. So it varies in the range of 75 to 90mA. I am running mine from a 9V 0.5A-rated Jaycar plugpack. As it turns out, it would have been OK with a smaller TO-92 package 78L05 regulator. The current briefly peaks over 100mA when the piezo beeper sounds, but the 78L05 should be able to handle that short-term demand. Due to its low power consumption, the game can be powered by a 9V alkaline battery with about a 500mAh capacity or a 9V Li-ion battery with a 1200mAh capacity. However, there is plenty of room inside the case for six AAA cells in a holder, which would significantly increase the running time over a standard 9V battery. For battery operation, it would be wise to leave the 1N5819 diode (D1) in circuit (to prevent reverse polarity mishaps). A 5V low-dropout (LDO) regulator would be better than the 7805 to get the most out of the battery life. For example, you could use an LM2940CT-5.0, which is pin-compatible and has a dropout voltage of just 110mV at 100mA, compared to around 1.5V for the 7805. Next month January 2023  89 Vintage Radio UDISCO L6 TRF Radio from 1926 or 27 By Dennis Jackson There was a least one Australian company that developed the batterypowered triode valve TRF wireless receiver of the early 1920s to its limit: UDISCO, the United Distributing Company. When advertising, UDISCO would often draw attention to the single dial control they used. According to the 1993 Electronics Australia publication “The Dawn of Australia’s Radio Broadcasting”, wireless telephony in Australia kicked off after the end of The Great War (also called World War 1) in 1918. The first direct wireless telegraphic messages between England and Australia were received on the 22nd of September 1918 by Ernest Fisk (later Sir), one of the founders of AWA. On the 13th of August 1919, Ernest Fisk demonstrated audible sound reception across Sydney without wires. Various professional experimenters and amateurs were to add their talents to the development of wireless transmission, until licensed broadcasting became available to the general public in early 1923, after the government enacted the necessary regulations. 90 Silicon Chip As one who witnessed the advent of television around 1960, I can imagine the excitement of the times. Peoples’ daily lives were taken up by manual effort, and there was little opportunity to understand the broader world. The cost of owning a wireless was far higher then compared to today. I have memories of conversations with those who lived during that era. One uncle told of his excitement when receiving his first feeble signals after months of experimenting with a simple single-valve regenerative receiving while living on the family farm. But not all were wholly in favour. A common belief was that wireless sets could distract women from their domestic duties during the day or affect peoples’ social lives in the evenings. The simple crystal wireless might Australia's electronics magazine suit a boy lying in bed listening to his favourite cowboy show, but was not of much use in a family situation. Once tuned radio frequency (TRF) sets became available, they were soon the instrument of choice. Two stages of tuned radio frequency amplification selected and amplified the station of choice, followed by a detector to separate the audio signal from its carrier frequency. Two stages of transformer-coupled audio frequency amplification were used to power a loudspeaker, giving a somewhat distorted output of less than 1W, still sufficient to amaze all listeners in an average room. UDISCO’s history UDISCO was founded in Australia in 1911, selling household goods and importing electrical components. siliconchip.com.au Fig.1: this circuit was initially traced by hand with sparse few known values filled in. As we couldn’t find a circuit diagram online for the UDISCO L6, this should be the next best thing. The company produced a wide range of sets between 1925 and 1929, ranging from kit sets sold under the brand name UMAKIT to advanced receivers like the UDISCO Super Six, a TRF set selectively tuning over six wavebands from 2000m to 20m with no gaps. Early sound technology and valveera radio, particularly from the 1920s and 1930s, have fired up my imagination as far back as I can remember. Around 30 years ago, I came by one of my more valued wireless finds at auction. The name engraved in the bottom corner of the front panel reads “UDISCO Model L6 Made in Australia by United Distributors Ltd Patent No. 20643 US No. 1610918 No.176”. The only reference I can find to my UDISCO receiver on the internet is one photo. My first task was to gain some understanding of how it worked. The circuit is more elaborate than the usual simple five-valve TRF receiver. I admit to at first being puzzled, so I began to sketch a rough circuit. Three attempts later, it began to make sense. It’s an upmarket TRF set with single-­ point cable-ganged tuning, housed in a heavy 300mm-high 819 × 380mm solid oak stained cabinet. The cabinet is meant to last for a generation or two and is typical of its time. The circuit has four RF stages with unusual choke-capacitive coupling (including the detector), plus the usual two stages of transformer-coupled audio. siliconchip.com.au The four RF coils are of the binocular type, resembling a single coil cut in half and bent over on itself, each half having an equal number of windings that are effectively wound in opposite directions. Both confined RF fields are intended to oppose each other, preventing interaction between adjacent coils and unwanted incoming signals making neutralising, shielding and angular placement unnecessary. The four sets of binocular tuning coils are mounted out of sight under one of the four sub-panels fixed to the baseboard. As can be seen from the circuit diagram (Fig.1), the first half of the first binocular tuning coil (L1) is tapped and switched to facilitate aerial matching. The binocular tuning coils seem to work, making this set very stable in operation with clear reproduction. It tunes stations without the howling or whistling common in TRF sets using simple triode valves. The choke-capacitive coupling between the four RF valves is via four large circular honeycomb-wound RF chokes mounted between the sub-­ panels and the front panel. They block RF from the B+ 90V supply and divert the RF signal through a 4nF mica capacitor to the tuned grid of the next stage. Valve lineup Most Australian-built radios from the 1920s that I have seen used Philips Bakelite-based triode valves with dumpy glass envelopes. This set initially used Philips A609 6V triode valves, designed for use with a 6V accumulator for the A filament supply. The A609 was first manufactured An example photograph of the ‘binocular’ type RF coils. Australia's electronics magazine January 2023  91 in 1926, the same year as this set. Its oxide-coated filament drew only 60mA and used the same four-pin base as the popular USA-manufactured UX201A. The UX201A had a thoriated filament drawing 250mA at 5V, making them interchangeable with some adjustment of the filament rheostats, while Philips had a sales advantage due to reduced battery drain. Mounted along the top edges of sub-panels two, three and four are six metallic tubular adjustable capacitors of a few picofarads each. My interpretation of their purpose is that C14, C16 & C18 provide a small measure of feedback to their respective valves giving some regeneration. C15, C17 & C19 also appear to be part of this network. This is just an educated guess, they could actually be for balancing out inter-electrode capacitance within the RF valves. Possibly confirming my determination, removing the adjustable sliding rods inside the insulated tubes gives a modest reduction in sound volume. The grid bias to V1, V2 & V3 is from the negative filament line via taps on tuning coils L1, L6 & L9. Controls The front panel looks uninteresting, with only two controls. There is a reduction dial for tuning and directly under it, a smaller knob for adjusting the rheostat (R1) controlling the plate-anode current to the four RF valves for volume control. R1 is bridged by capacitors marked C21 and C22, which are in series and centre tapped, going to the positive filament line. Choke L13 in the anode circuit of V4 has the primary of L14 (the first audio transformer) taken from its more positive side instead of directly from the anode, as one might imagine, but it works better that way. A long, narrow sub-panel just under the hinged lid holds additional knobs. The first on the right controls a vaned trimmer capacitor (C1) across the first of the four tuning capacitors, to adjust for any misalignment as stations are tuned across the bands. The second knob controls a rheostat (R2) in series with the A+ battery supply to adjust the valve heater current according to the battery voltage. It also affects the volume (along with the external knob mentioned earlier). Knobs three (C2), four (C3) and five (C4) perform similar functions as knob one, tuning capacitor trimmers. Knob six (C9), marked “control”, adjusts the positive feedback from the plate of V5 to the grid of V4 (the detector) to provide regeneration. Restoration Opening up the lid reveals a series of aditional knobs connected to the chassis. The knob at the top of this photo is connected to trimmer capacitor C1. 92 Silicon Chip Australia's electronics magazine At least one of the audio coupling transformers was replaced sometime during the history of this set. I also noticed that, at some point, radio-­ frequency choke L3 had been added across choke L4. It seemed unnecessary, so I removed it. I inspected the set and couldn’t spot any more apparent problems, so I decided to switch it on and see if it worked. I connected to an aerial and Earth plus my most trusted horn speaker before wiring in my special battery eliminator power supply and making voltage adjustments. As is typical, there was not even a buzz, and no amount of knob twiddling could coax this set into the faintest whisper. I should have performed a closer inspection by checking the voltages on the valve pin sockets. Using a signal tracer, I found a signal at the grid of V1 but none at the plate. Also, detector V4 lacked HT on the plate, indicating there were open-circuit anode chokes. The very fine-gauge cotton-­covered wire used in these large honeycomb-wound coils was adrift from the respective terminals. Worse, both siliconchip.com.au C1 R2 C2 C5-C8 C4 Reaction Feedback Bias C22 C21 Adjustable Capacitors L14 L15 Anode choke coils I removed the chassis from the cabinet to effect some repairs (the 9V battery was used for testing and is not part of the set). Four of the six knobs adjust trimmer capacitors across the tuning caps. wires on L4 had broken close to the coil. The outer was easy enough to pick up, but the inner close to the coil former had only a couple of millimetres of stub left. With no second chances, several careful scrapes with a razor blade exposed a streak of clean copper and I gently added a dab of solder to join another thin wire. I then added a small blob of Blu Tack to keep it rigid. But there was still no continuity. With fading hopes, I decided on a closer inspection under a large magnifying glass with good light. A tiny green spot of verdigris was visible. I poked it with a needle to expose two short, stubby wire ends, which I then bridged and set in place with another blob of Blu Tack. It then had continuity which was a considerable relief. A tiny drop of acid solder flux, probably splattering during manufacture, had corroded the wire through in subsequent years. That explained why RF choke L3 had been soldered across it. Now that I’d fixed L4, it was no longer necessary. With both anode choke coils now repaired, there should have been some siliconchip.com.au response from the horn speaker, but it is never that easy. I re-checked everything twice more; all seemed good, but there was still no response. One or more of the six A609 valves must be low on emission, so I would have to set up my Paton valve tester, which has a four-pin UX socket to suit these early triodes. All valves displayed less than 50% emission, with one being a total dud, probably resulting in this set’s retirement. These old Philips 6V triodes are very seldom for sale now. I keep a few known-good UX201A 5V triodes for replacements, so I fitted them after re-adjusting the filament ‘A’ supply. This brought forth a hint of croaky reception from the ancient horn speaker. Some careful adjustment of the single tuning control on the front panel, together with the anode voltage rheostat and then all six knobs along the under-lid sub-panel, resulted in surprisingly ample sound. A rocking armature speaker gave a less strident output; no doubt, a further improvement could be obtained by fitting a moving-coil unit through its output transformer. Australia's electronics magazine The similar AWA Radiola C54 At this stage, I remembered that I had previously purchased an AWA Radiola Battery Six model C54 from around 1928. Electronically, it is a similar set but probably as basic as a six-valve TRF wireless could be. The point of interest was that the model C54 also used four sets of binocular RF tuning coils. In that set, the more typical inductive coupling was used between all four RF stages instead of choke-­ capacitive coupling. I decided to try to get both TRF sets working so I could compare their performance. Unfortunately, both coils in each AWA audio coupling transformer had gone open circuit. Someone had worked around that by inserting the high impedance speaker in the HT circuit of the first audio valve and feeding its grid through a 100nF capacitor from the detector anode. The final audio stage had simply been disconnected, making this a five-valve set instead of the original six. I have successfully rewound open circuit windings on audio coupling transformers using very fine enamelled January 2023  93 An advert from Wireless Weekly, June 1927 showing a UDISCO Neutrola which uses a case that is very similar to the L6. copper wire (0.1mm/4-thou diameter). Patience and a gentle touch are required. Fortunately, I was able to scrounge two working replacements from the junk boxes of friends, with the originals perhaps to be rewound sometime in the future. Two of the six valves were missing. I tested the remaining four British Marconi Osram valves for emission, and three came up good. The Philips B406 appeared to be similar, and once added and everything connected and tuned in correctly, this set now gives good reception for our two main local stations. Both sets are inaudible when the aerial is removed and are free of any sign of oscillation in everyday usage, possibly due to the use of binocular tuning coils. Sensitivity is limited in TRF sets due to the low RF gain of the front-end when compared to my two superheterodynes from the same period. Still, the output volume is good considering the meagre gain of these early triode valves, particularly in the output stages. Substituting a 71A or a UX112A power output valve (both have 5V 0.25A filaments) gives a noticeable increase in audio volume. These valves are compatible with UX201A types and became available in early 1927. In conclusion, the UDISCO model L6 is a good user-friendly receiver, making up for its plainness in ornamentation by its sheer bulk, complexity and exceptional performance. SC A photograph of the AWA Radiola model C54. Like the UDISCO L6, it is also a six-valve TRF set and uses four sets of binocular tuning coils. 94 Silicon Chip Australia's electronics magazine siliconchip.com.au PRODUCT SHOWCASE element14 now shipping Rohde & Schwarz MXO 4 series oscilloscope element14 is now stocking the new Rohde & Schwarz (R&S) MXO 4 Series, the first in the next generation of oscilloscopes. This oscilloscope provides the world’s fastest real-time update rate of more than 4.5 million acquisitions per second. This rate enables development engineers to see more signal detail and infrequent events than any other oscilloscope, providing an unparalleled understanding of physical layer signals and faster testing. The integrated 12-bit ADC in the R&S MXO 4 Series has 16 times the resolution of traditional 8-bit oscilloscopes at all sample rates without any trade-offs, providing the most precise measurements. A standard acquisition memory of 400Mpts on all four channels gives the instrument up to 100 times the standard memory of comparable instruments. The MXO 4 Series also features a unique 200Gbps processing ASIC, a notable breakthrough for accelerated signal insights. The R&S MXO 4 Series features a 13.3” full-HD capacitive touchscreen and an intuitive user interface. The instrument’s small footprint, very low audible noise, VESA mounting and rackmount kit for installation makes it an ideal oscilloscope for any engineering workspace. Other features include: • 18-bit architecture • The fastest and most accurate spectrum analysis in its class • Trigger rearm time of 21ns • Digital triggering technology • Sensitive trigger of 1/10,000 div • Trigger jitter of < 1ps • Dual-path protocol analysis • Features the R&S SmartGrid user interface The new R&S MXO 4 oscilloscope series is available from element14 (https://element14.com) in the Asia-Pacific region. element14 72 Ferndell Street, Chester Hill NSW 2162 Phone: 1300 361 005 https://au.element14.com/ New ARM-based PIC with Bluetooth from Microchip Microchip’s new PIC32CX-BZ2 family includes System-on-Chip (SoC) devices as well as RF-ready modules. In addition to Bluetooth Low Energy (LE) functionality, the family includes Zigbee stacks and Over-the-Air (OTA) update capabilities. Hardware features include a 12-bit ADC, multiple timer/counters for control (TCC) channels, an on-board encryption engine, and a broad set of interfaces to touch, CAN, sensor, display and other peripherals. The family’s 1MB of flash memory supports large application codes, multi-protocol wireless stacks, and OTA updates. The PIC32CX-BZ2 MCU family simplifies development through Microchip’s MPLAB Harmony 32-bit embedded software development framework. MPLAB Code Configurator integration enables developers to quickly begin prototyping with the PIC32CX-BZ2 family using drag-and-drop auto code generation. Numerous examples are hosted on GitHub and linked through MPLAB Code Configurator and MPLAB Discover. RF design with PIC32CX-BZ2 Lowest-cost, 48-pin QFN package. SoCs is simplified with the ecosystem’s chip-down reference design packages and wireless design check services. Customers with little to no RF expertise can benefit from Microchip’s WBZ451PE & WBZ451UE modules that are pre-certified and feature an optimised on-board RF design. In addition to the MPLAB Code Configurator, the MPLAB Harmony V3 framework includes numerous other tools and an ecosystem of debuggers, programmers, virtual sniffer, and compilers. Other support includes documentation, wireless design check services, and building blocks that walk developers through all the steps involved in the development process. The PIC32CX-BZ2 family is supported by the PIC32CX-BZ2 and WBZ451 Curiosity Development Board (Part number: EV96B94A). For more details visit: siliconchip.au/link/abi8 Microchip Technology No RF experience is needed with the WBZ451 module. siliconchip.com.au Australia's electronics magazine 2355 West Chandler Blvd, Chandler Arizona 85224-6199 USA Phone: (480) 792 7200 www.microchip.com January 2023  95 SERVICEMAN’S LOG Sometimes it all just falls into place Dave Thompson Often in the service industry, we get these weird coincidences where a new appliance in need of repair comes in, then a few days later, another similar unit turns up as well. Although I have no experience repairing this type of device, I was fortunate that both had simple faults that became apparent once I dug into them. This sort of weird coincidence happened to me recently when a computer-repair client mentioned they’d just opened the packaging on one of those oil diffusers that seem to be all the rage these days. They’d purchased it a while back, but when they went to plug it in, they discovered it wasn’t working. Of course, I said I could take a look at it (it’s an electronic device, after all), though I made it clear that I’ve never opened one up before, so this would be a new experience for me. They were happy for me to crack it open and have a look, as it was now out of any warranty that might have applied, and they accepted that doing something is better than doing nothing. I understood completely because I know that many of these diffusers are not cheap; some go for hundreds of dollars, a significant outlay in anyone’s money. If I plugged in a brand-new device and it didn’t work, I’d also be more than a little miffed about it! I hadn’t even started on the repair yet when another client called and asked me if I’d ever had an oil diffuser in 96 Silicon Chip for repairs. I replied that, of course I had, before explaining to them that it had only been one day, and I hadn’t even had a chance to look at it yet! They too claimed that it had cost a pretty penny and, while it had been working fine for a while, it had started failing to stream vapour properly. In the meantime, their cat had knocked it off the table it lived on, and now it sounded like something had come adrift inside. Could I look at it? Bemused by the coincidence, I agreed to take the job on – I mean, how hard could it be to repair something as apparently simple as an oil diffuser? Before the second one arrived at the workshop, I decided to crack open the case of the first one and see what was going on. Preparing for surgery The main body of this diffuser is made of injection-­ moulded plastic and consists of three sections. The base contains the power input socket and controls. A water tank section is mounted on top of the base, while a removable funnel-shaped ‘chimney’ caps off the whole caboodle. Vapour streams from the open ‘chimney’ when the device is operating. This diffuser has other features; a digital clock and on/ off timer are included, as is one of those sound synthesisers that can simulate rain, wind, the ocean and, in this case, a forest with birdsong or a running brook or stream. A row of pushbuttons and a rotary volume control (similar to what you’d find on an old transistor radio) are set around the middle of the base part of the body. These control everything to do with the diffuser, the clock and the sound generator. Removing the funnel is simple enough – it is designed to be removable and is simply press-fitted onto the middle section (the water tank). This is how water and oils are added. The cone is then re-fitted, and the diffusing process starts. With the cone off, I could ensure that no water was trapped in the internal components. It was bone dry. The next thing was to separate the two bottom parts – this would reveal all the actual components. The two parts were fastened together with two simple PK-style screws. There were also three clips at 120° positions around the circumference of the body; these required a little careful fettling to remove. This method of clipping things together instead of screws or other fasteners is increasingly used these days to hold plastic cases together. The screws in the base are likely a Australia's electronics magazine siliconchip.com.au Items Covered This Month • Sometimes it all just falls into place • Tips for fixing an LCD TV backlight • An unfortunate series of battery chargers 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 legal backup requirement, given that there’s power floating around inside. I’m always very wary of breaking clips when disassembling devices – many modern laptops and monitors are renowned for having very flimsy (possibly intended as single-­use) clips holding everything together. There’s nothing wrong with the method itself, but on many devices, it reeks of the “no user-serviceable parts inside – not intended to be repaired” philosophy. Regardless, I got it open. As I suspected, a single PCB took up the majority of the room inside the base. A short wire lead from the tank assembly was plugged into the PCB via a connector, and this had to be removed before the two parts could be separated. I set the tank aside for now. The board was screwed to the base with the external buttons toggling standard small SMD switches positioned around the board’s edge. Power to the device is via a phone charger-style plugpack and a standard 3.5mm barrel connector at the end of a one-metre cable. The power supply connects to a socket mounted into the plastic base. The pick of the plugpack The connections to the socket were easy to get to, so my first task was to plug the barrel jack into the socket, plug in the supply and test the voltage coming from the socket. With my multimeter across the contacts, I got a reading of exactly nothing; there was no power getting to the socket. I re-checked that I had actually plugged the supply in properly, and there was voltage at the power board on my test bench (I’ve been caught before by first not ruling out the basics!). Other devices were running from the power board powered on OK, so it was time to check the power supply output. Testing barrel jacks is always a bit of an act, especially when the meter probe is too large to slide down the centre contact. I use part of an old dental pick that broke off years ago; being tapered down to a point, it is a universal fit into almost all of these smaller barrel jacks. Once in place, it is much easier to just hold the meter probes against that and the outer contact. Top tip – ask your dentist next time you visit for any old picks and probes – they are the handiest things for all electronics tinkering, especially SMD placement and other delicate work! This time, I measured 5.2V, close enough to the 5V listed on the product labels. So, power was coming out of the supply but not reaching the output of the jack socket. I flipped the whole thing over so I could eyeball the socket more closely. Looking at both the plug and the socket with my loupe, the plug looked fine, but there was something siliconchip.com.au in the socket. The plug didn’t feel secure and looked to be protruding slightly, even when pushed as far in as I could. With a light beamed into the socket, I could see what looked like a piece of plastic in the way. I used another of my handy dental picks to fish around, and the plastic moved when I touched it. It seemed to be right around the centre pin of the socket. Flipping it all back over, I used the time-honoured method of shaking things loose by holding the base in my right hand and clapping it down into the palm of my left hand, in the hope the sudden stop would dislodge the foreign object and gravity would do all the work for me. I did this several times and could see the plastic was moving. The piece came out with a bit more probing with various picks and tweezers and a few more soft taps. After plugging the power source back in – noting this time it went all the way in – I measured the same 5.2V at the socket connections. The clock display lit up a very nice blue and happily flashed 12:00, so I knew I’d found the problem and that now it was going to work. The debris prevented the power jack from going all the way in, so no contact was made. I’d save trying the diffuser part for when it was all back together. Quality assurance backfires On closer inspection under a magnifying glass, the plastic ring turned out to be the top part of the insulation ring separating the two contacts of a barrel jack plug. The plug on the supply that came with the unit was intact, so I can only assume that a QA tester used a single power source to quickly test all the diffusers coming off the production line. My guess is that they pulled that power plug out, leaving the last bit of the ring behind. They might not have even noticed it for a while, and by then, they wouldn’t know where the piece had gone. My client had drawn the short straw! While it was apart, I looked at the other components. I was most interested in the diffuser itself and had no idea how it worked until I started looking into it. I assumed heat was involved, which vaporises the oil and water mix, creating the stream of ‘steam’. Not so, or at least not in this one. Australia's electronics magazine January 2023  97 Some nebulisers operate that way, but they are usually found only in high-end medical devices. These so-called homeopathic diffusers for domestic use utilise ultrasonics; no heat is involved. An ultrasonic disc transducer is mounted in the centre at the bottom of the water reservoir. When power is applied, ultrasonic waves vaporise the oil and water in the tank and the specially shaped funnel corals it all into a nice stream of scented vapour. Safe and very clever! Editor’s note – they are basically the same design as ultrasonic humidifiers; the ‘steam’ generated is actually a cloud of tiny water droplets that quickly vaporise unless local humidity is very high. Once I had it all back together, I filled it with water to the embossed mark on the side of the reservoir and added some ‘essential oil’ I’d had stored for years. I originally used it with a different type of scented oil diffuser, which used a simple tea-light candle to heat a ceramic bowl containing the oil. With the funnel back in place, I hit the button and a fine stream of mist poured from the top of the outlet. It is surprisingly powerful, totally cool to the touch and very 98 Silicon Chip fragrant – though I think I used a few too many drops of oil. It turns out these are very efficient and only need a few drops for a full tank of water (roughly 200mL on this model), depending on the concentration of the oil and the scent itself – some scents are far stronger than others. I set the clock and messed around with the sounds and the timer function, and it all checked out OK. So, a relatively simple fix then; it would be interesting to see what was happening with this other one, though, because it had apparently been in use for quite a while but now didn’t work ‘properly’. Plus, it had been dropped. Diffuser #2: Electric Boogaloo The client brought that second diffuser in a few days later and I asked him to be more specific about how it operated before it had been dropped. He said it worked fine at first, but the output had reduced significantly of late. As it just wasn’t as good as it used to be, they had stopped using it. This model was quite a bit different than the last one; it didn’t have anything as fancy as a clock, timer and sounds, but it did have RGB lighting, controlled by a single pushbutton switch that toggled between the different colours and modes. Ominously, it rattled when I lightly shook it, so something had come adrift inside. I’d have to open it up to see what was going on. The funnel on this model also pops off easily for filling, and I could see a problem straight away; there was a small plastic coin-sized disc sitting in some sludge at the bottom of the empty water tank. I set that to one side. The whole inside of the tank, funnel and recessed ultrasonic transducer was covered in a thick film of oil residue. It did smell nice, though! I’d need to clean it out properly at some point, but in the meantime, I used a paper towel to wipe as much of it out as possible. There was still something loose in the base somewhere, so that had to come apart. This time, I encountered three ‘security screws’ holding the bottom to the tank stage. Fortunately, I now have a good collection of bits that undo these fasteners, so it only slowed me down a little. With the screws out, the two sections came apart Australia's electronics magazine siliconchip.com.au 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 169 $ TS1328 GREAT ES. FEATUR GREAT PRICE! DIGITAL MULTIMETER WITH TEMPERATURE • Autoranging • Cat III 600V • 10A AC or DC current • 40MΩ resistance • 100µF capacitance • 760°C temperature • K-type probe & case included 20MHz USB Oscilloscope • High accuracy interface • Spectrum analyser (FFT) • 48M Sa/Sec sampling rate • 20mV/div sensitivity ONLY 199 $ QC1929 HEAVY DUTY WIRE STRIPPER • Cutter, crimper & wire guide • Strips 10-24 AWG/0.13-6.0mm • Single handed operation TH1827 $39.95 QM1323 $59.95 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 $59.95 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 169 $ 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 110 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 jaycar.com.au/workbench 1800 022 888 easily. I also had to unplug the transducer lead from the small circuit board inside to separate the parts fully. I could see a fan had come loose in the base and was floating around on its wiring. It had been screwed to two plastic posts, which had come adrift; they were still screwed to the fan. I left it all as it was and placed the fan back where it should be. The broken-off parts matched up very well, so I decided to glue it back into position. Everything else in there looked OK – there really wasn’t much to see. A strip of SMD LEDs was mounted in a ringshaped moulding around the inside of the base and hardwired into the PCB. All seemed OK, so I plugged the diffuser in and tried it. The fan spooled up and the lights came on and changed colours when I repeatedly pressed the button. I’d have to plug the ultrasonic transducer back in and fill the tank with water to test it properly, but I felt sure it would work after I gave it a thorough cleaning. So I went ahead and glued the fan back in with 24-hour epoxy; I didn’t want it moving again. While that set, I took the tank and funnel assembly inside and used a good detergent to clean the inside and outside of both parts. The transducer appeared to be glued to the bottom of the tank, so I wasn’t about to disturb that, but I used an old toothbrush to give the recessed, visible part of it a very good clean. I reassembled the diffuser, filled it with water and added a few drops of oil. I put the funnel back in place and hit the button. Instead of a nice stream of vapour, I got water spitting all over the place! The sharper knives in the block among you will recall I found a small, coin-shaped plastic piece lying in the tank. On closer inspection, I could see where this baffle had broken away from the inside edges at the top of the funnel. After fiddling it back into position, I tacked it with instant glue. On testing, it worked perfectly, so I glued it properly with epoxy. Job done then, and two happy clients. An example is a Linden L55UTV17a TV I looked at. It has six strips each of 15 LEDs connected in series, as shown in the photo below. I tested these with an LED backlight strip tester, and only three of the strips were OK. I also looked at an LG 49LB650 TV. LG uses a higher current in the backlight strips, which causes them to go blue after a while. This gives a blue cast to the picture. In this case, one of the LEDs had burnt and actually damaged the strip. With LG TVs, it is best to replace all the strips when they fail; AliExpress has a good range. Newer TVs divide the backlight into sectors. The Hisense 50P7 has eight sectors, and the power board that drives the backlight is now on the serial bus, so the board is more complex. Similarly, the LG 75 NANO86 power board is also on the serial bus and drives 12 LED strips. Due to the more complex power boards, it is becoming more challenging to determine if the main board or power board is faulty. Most TV repair places deem it uneconomic to replace backlight strips due to the time involved and the risk of breaking the LCD panel. Still, it is worth a try if you are doing it to fix your own TV. Up to about a 55in (140cm) TV, you can, with care, be successful. First, from the circuit board side, very carefully disconnect the long board that connects to the LCD panel ribbon connectors. Then turn over the TV and remove the screws holding the retaining edge around the panel. Do not do this from the circuit board side, as you will probably break the panel. Remove the panel very carefully, taking care not to flex it much. Put it aside. Then take off the retaining edge around the sheets of plastic that diffuse the light. Put the sheets aside, keeping them in order. You are now at the backlights. Test them with a backlight tester (available on eBay) to determine the faulty strip(s). See if you can buy replacement strips on eBay or AliExpress. Re-test before reassembly. Tips on fixing LED TV backlighting An unfortunate series of battery chargers R. S., of Figtree Pocket, Qld has found that LCD TV LED backlighting can be troublesome. Still, if it fails, it generally can be fixed, and he has some good tips on how to do that... The change from cold cathode backlight tubes to LED strips for LCD screens was supposed to be an improvement, but they seem to be less reliable. Many newer TVs will not turn on if a backlight fault is detected. J. B., of Burpengary, Qld sent in a saga involving two battery chargers and a seemingly never-ending series of faults, trials and tribulations... The chargers in question are Truecharge 20i (TC20i) models made by Statpower (now Xantrax). They are rated at 20A 12V with three stages and can simultaneously charge two batteries of the same chemistry semi-independently. Two small slide switches select between flooded and gel, and three temperature ranges on the front face: cold, warm and hot. Charging and float voltages are listed for each range for both flooded and gel cell batteries. The charging voltage has a range of 13.8-14.8V and float 13.1-14.2V, both in 0.2V increments. Not ideal for some chemistry types. An eventual upgrade to lithium-ion will need a revisit of what to do, but I only have flooded and AGM at this present time. Two extras are available: a battery temperature sensor and a remote panel. With a temperature sensor connected, the front panel temperature switch is ignored. They connect via two RJ12 6P6C sockets. The three charging stages are the usual bulk, absorb and float. A hidden fourth stage (equalise) is accessed by holding down a small recessed button on the front face with a narrow pointed object, eg, a straightened paperclip, for two seconds. This Linden L55UTV17a TV has six strips of 15 LEDs arranged horizontally and connected in series. 100 Silicon Chip Australia's electronics magazine siliconchip.com.au There are also charging and charged LEDs and an overall current readout in 4A steps on the front face. Short-circuit and reverse polarity protection are built in, the latter via a pair of 30A blade fuses, one for each battery. I will shamefully admit I have blown more than one fuse while using these chargers. The first charger saw service in an ambulance. It was used while parked at the station to keep both the start and house batteries topped up. The house battery runs a small custom-made fridge with strict temperature control to keep drugs in. While going out to a job under lights and sirens, this particular ambulance caught fire in the electrical enclosure due to a flasher unit being under-rated for the task required. The boss was told about the flasher unit but chose to do nothing about it till this fire happened. A recall was issued so the flasher units could be replaced. Firefighters seem to completely disregard the fact that electronics don’t like being doused with water. This particular charger copped an absolute drenching. This ambulance eventually found its way back to our workshop for repair. No doubt there was a big argument over who would pay for the repairs. The whole electrical enclosure and other fire-damaged parts were removed and stored. Before it was all dumped, I managed to retrieve the charger and a 12V to 240V inverter. There was also a remote panel for the charger, but all that was left was the bare board; everything else was gone: the solder mask, the copper tracks, even the vias. The remote panel has a line of LEDs for each battery to show the battery voltage in 0.5V increments and five LEDs to show the overall current. Two LEDs also indicate the charging and charged states. The inverter was a modified square-wave type. I used it only a handful of times over 14 years until it gave out. Once home, a cursory inspection of the charger showed that the IEC inlet socket had the melted remains of the plug in it, and there was a small amount of fire damage to the aluminium cover next to the socket. A screw was also missing that held the socket to the cover. It appears that a lot of heat was next to that screw, and it melted the plastic housing so that the screw fell out. The charger is mostly made from a long U-shaped finned aluminium section that is also the heatsink. A long, thin hat shape made of sheet steel fits over the top and ends and has flanges to mount the charger vertically to a wall. It has a small circuit board to hold seven LEDs, two slide switches and a recessed push button. A ribbon cable connects this panel to the main board near the processor. Removing the top cover revealed that the inside was surprisingly clean. The only thing to do was to remove what remained of the IEC plug and give it a go. It worked straight away, much to my delight. Some six months later, I was charging a battery in the carport when a heavy downpour came through. I didn’t know at the time that the carport leaked water during heavy rain. Naturally, the charger was right under the leak, and it protested the impromptu shower by ceasing operation. It was time to remove the top again and see what damage had been done. Removal of the circuit board requires the disconnection of five clamps that hold large heat-­generating components, four screws that hold two tabs at either end, and one Earth wire to be undone from the heatsink. The board then slides out. siliconchip.com.au Australia's electronics magazine January 2023  101 Two small-signal transistors had their sides blown out, removing most of the type numbers. There were also some black marks around one of the two IRF840 Mosfets, a blown 4A fuse and a slightly blackened and cracked resistor. The first order of business was to try to get a circuit diagram. The internet revealed nothing, so I sent an email to Xantrax. Their response was to send money plus charger plus return postage. At the time, the exchange rate was not in our favour; it would have cost as much as a new one to do that particular activity. The only remaining option was to figure out what the blown parts were and hope for the best. Looking closely, I discovered that only two small-signal transistor types were used throughout the whole charger: 2N2222A and 2N2907A. There was one of each type next to the two blown ones, and the circuits appeared to be the same as both pairs drove the gate of their associated IRF840. So I felt sure I knew what to replace those transistors with. There is also a UC3845A controller chip (U1) that I felt should be replaced. There are two opto-couplers as well, but figuring out their types was an arduous process as the markings were very hard to see, even with my strongest magnifying glass. After an hour of researching possible type numbers, looking yet again using different light sources, and getting just the right angle of reflected light, I finally found both to be 4N25s. After replacing all the above and the 47W resistor plus the fuse, I fired it up only to reveal that the charger would go through its boot-up sequence but not put any current into the battery. Something wasn’t right. I spent a lot of frustrating time trying to locate the problem. Measuring everything in-circuit didn’t show anything out of order. Eventually, I gave up, put the charger away and waited for inspiration to hit. About six months later, while looking for something else, I came across the charger and pulled it out again to have another look. This time, I measured the three 0.1W 3W resistors out of circuit. One of a paralleled pair was open-circuit, which I very much later discovered is part of the current sense circuit. I found a suitable resistor in an old CRT monitor. Replacing it finally fixed the charger (again!). Buoyed by that success, I noticed a second identical charger gathering dust in the storeroom of my then-employer. I asked if I could have it as it wasn’t working. After fixing one, how hard could another be? Opening it revealed the same two burnt transistors. I replaced all the same parts except for the 47W resistor, but I did have to replace one of the 0.1W 3W resistors. Switching it on without a battery connected, it went through its usual power-up sequence, and no smoke escaped. After connecting a battery, however, it was a different story. Much fire and brimstone issued forth as soon as the startup sequence completed and current was applied to the battery. “Oh, dear!” I shouted, or perhaps a slightly less polite word to the same effect. I was now trying to do things on the cheap by leaving parts out and powering on or not replacing parts that I should have. It resulted in a growing pile of blown-up silicon, much smoke venting into the atmosphere, many sparks and damage to heavy tracks. Smarter people would know that switchmode supplies require all parts present and working, but Muggins here is a slow learner. After the fifth time, I decided to replace all the silicon parts listed above and, while I was at it, fit a socket for the controller chip. I also replaced three 15V zener diodes. After that, finally, the charger fired up properly. I then reassembled it and tested it for three months by running a 12V fridge connected to a small lead-acid battery before declaring it fixed. Unfortunately, on the first camping outing to the “outlaws” (wife’s parents) with this charger, Murphy found us overnight and hit the charger with some strange ‘stop working’ spell. “Bother!” I said quite loudly (and perhaps not so politely). Back home, investigations revealed that the startup resistor (220kW 1W) had gone high in value. I didn’t have one on hand and couldn’t find one in my pile of disassembled bits, but I made a close facsimile from two 470kW 0.5W resistors in parallel. Once again, it worked as it should. This charger subsequently travelled across some of the worst roads in Australia on various camping holidays for several years till Christmas 2011, when I went camping on the largest sand island in the world. For this trip, we bought a small 720W two-stroke generator from a large hardware chain. Its voltage is regulated by adjusting engine RPM. It was backup if the sun decided to hide during the day. It also has a dedicated 12V output for battery charging. Two 35L 12V fridges (actually one fridge, one freezer) take a heavy toll on batteries. So for this trip, I set up two batteries dedicated for both fridges and brought both chargers, figuring I could run the generator half the time. Murphy must have followed me or disguised himself as a dingo as there was very little sun to keep my solar panels busy. I was forced to use the generator. Well, things didn’t go to plan as some 20 minutes after starting the charging process, the generator suddenly started to labour. I quickly determined that the second charger had stopped working. A close-up one of the Trucharge 20i battery chargers. The main PCB suffered some water damage. 102 Silicon Chip 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 9995 $ Digital Microscope • LED illumination • Rechargeable QC3193 FULLY ADJUSTABLE POWERFUL 127MM DIA. 3-DIOPTRE LENS Clamp Mount Desktop Magnifier with LEDs • 1.75x, 2.25x & 3x magnification • 60 LEDs with high/low brightness • Mains powered FULLY ADJUSTABLE ARM ONLY 119 $ QM3554 ONLY 2995 $ 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 12 $ 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 199 $ 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 110 stores or 130 resellers nationwide. jaycar.com.au/magnify 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. I later discovered, to my disappointment, that if the separate battery charger option on this cheap generator was used, the AC voltage the generator puts out goes up to 265V AC. Oh well, at least one charger was still working. I kept a close eye on the output voltage and managed the RPM for the rest of the time on this island. Once back home, I discovered that everything inside the charger looked pristine. The generator didn’t have the wattage to blow the fuse or let the smoke out. Investigations revealed that the IRF840s were both shorted, as were the gate driver transistors and the UC3845 switchmode controller. I also decided to replace the opto-couplers and the zener diodes. I checked the resistors and the capacitors in-circuit with my multimeter and they all appeared to be OK. However, upon firing it up after replacing the semis, nothing happened whatsoever. I then spent several weeks trying to find out why and replacing many components all over the high voltage section, none of which helped. I finally made voltage comparisons with the other charger on each pin on the UC3845. All voltages were very close to each other except pin 7, which measured 10.5V. This pin is fed directly from the startup resistor and a winding on T2 via a simple regulator. It should have been above 12V. Was it a load or supply problem? Around this time, I drew up a circuit diagram to work out what was going on (reproduced below). I discovered that the two high-­ frequency transformers are identical, but only one has its feedback winding connected. I swapped U1 over, but again, it made no difference. Replacing C27, C28 and C29 made no difference. Replacing R5 and R26 again drew a blank. In desperation, I fed 12V from a small battery directly to pin 7 of U1. To my surprise, the charger fired up and proceeded to work as it should. I could remove the small 12V battery once current was supplied to the battery, and the charger would keep going. Every 15 minutes or so, the charger would stop for about five seconds to, I assume, read the battery terminal voltage before continuing to charge it. It was at this point that the charger would stop dead. Feeding 12V to pin 7 would once again bring the charger back to life. This proved that the feedback from T2 was working, but the startup resistor wasn’t supplying enough current. Or was it? Once again, I replaced R5, but it made no difference. In desperation, I started to replace the small capacitors around U1. C21 broke apart while removing it. Replacing it was the answer to all the troubles. But why? The UC3845 (IC1) has a 5V reference available at pin 8. It appears that C21 was drawing more current that the 5V reference could supply, and at startup that was keeping the supply voltage below the threshold required to start the chip. During charging, extra current from the feedback winding provided the current required. We gave away that generator and now have an inverter generator to run the chargers. Both ran flawlessly for over 10 years. The first charger recently developed a problem where it would go through its startup sequence, then reset and repeat in a continuous loop. Even activating the equalisation mode didn’t stop this behaviour. I just hoped it wasn’t the processor, so I swapped it from the second one. The problem stayed with the first charger. Looking closely, I could see a white film all around the processor but couldn’t get in there to clean it. Removing both RJ12 sockets revealed a white film under them. A good clean and reassembly was the fix. It appears that the drowning SC close to 20 years ago finally showed itself. A reproduction of the selfmade circuit diagram for the battery charger 104 Silicon Chip Australia's electronics magazine siliconchip.com.au ASK SILICON CHIP Got a technical problem? Can’t understand a piece of jargon or some technical principle? Drop us a line and we’ll answer your question. Send your email to silicon<at>siliconchip.com.au GPS Clock Driver has no USB communication I have built the New GPS Analog Clock Driver from the September 2022 issue (siliconchip.au/Article/15466). It is working correctly, with GPS, and driving the modified clock stepper as designed. However, I have hit a brick wall attempting to configure the clock driver as per the instructions on page 63. I am running Windows 10 and have PuTTY installed. I am reasonably familiar with PuTTY. When I connect the PC to the module, the LED blinks three times to indicate that it is in configuration mode. Here my problems begin! Under Device Manager, I cannot find the virtual serial port. The USB socket is OK; I am well set up for SMD-type work. The fact that the blue LED on the module flashes three times confirms the connection to the laptop according to the instructions. Can you clarify what I am doing wrong? (C. H., Deep Bay, Tas) ● Geoff Graham responds: the three flashes of the LED on the clock controller board indicates that it has detected the +5V on the USB connector, has switched to configuration mode and is trying for a USB connection. It does not mean that the connection has been made. Almost always, your symptoms indicate a bad solder joint on the USB socket. A more unlikely reason is that you have a faulty cable. Other than those, there is no other fault that makes sense. I suggest rechecking the soldering on the USB connector carefully. Sourcing double-sided tape pads I need to obtain some of the double-­ sided adhesive foam tape supplied with the September 2022 GPS Analog Clock Driver kits (SC6472). It’s used to secure the GPS module to the clock PCB. I understand it is a 3M product. Can you please tell me the part number for this item? (G. D., Bunyip, Vic) siliconchip.com.au ● 20×20mm or 20×25mm double-­ sided foam-cored tape pads are widely available and are about the right size for that job. 3M does make them, among others. You can get them from eBay, eg, www.ebay.com.au/ itm/384175409461 You can also use strips of double-­ sided tape from the hardware store. Substituting VK2828 GPS module for VX16E I am attempting to build another copy of a device I built about two years ago. It made use of a GPS module (VK16E) that was connected to a Micromite by just one wire (yellow), apart, of course, from GND and 3.3V (so three wires in all). Is the VK2828U7G5LF GPS module (Cat SC3362) a ‘plug-in’ replacement? If so, can I leave the other leads disconnected? (A. F., Salamander Bay, NSW) ● They should be compatible. It might seem that you only need to connect three wires: GND, VCC (3.3-5V) and TX. However, connecting EN and RX to VCC is a good idea. Geoff Graham found that leaving RX floating was not a good idea when designing the GPS-Synchronised Analog Clock. The PPS output can be left disconnected if you don’t need that signal. D1 Mini emulating GPS module not working I have purchased and programmed a WeMos D1 Mini WiFi module for the GPS Analog Clock Driver (November 2022; siliconchip.au/Series/391). The firmware has programmed OK as it has a tick in the green dot at the bottom left of the screen. But when I reconnect the module to the USB cable, the blue LED on the module does not come on and stay on. It flashes on and off two times. When I connect a terminal emulator, it finds the module on COM4. When I press the button on the module, it displays a lot of rubbish on the screen. The same thing happens on two WeMos D1 Mini modules. Australia's electronics magazine Do you have any idea what is happening? It seems like two faulty modules. (R. W., Mount Eliza, Vic) ● Geoff Graham responds: I have received a report from another constructor who had a problem that sounded a little like yours. He tracked it down to the regulator on the D1 Mini that could not supply the peak current demanded by the WiFi chipset when it was initialised on power-up. He had three modules that acted the same. He fixed the problem with another module from a different supplier. The only other cause I can think of is that your computer cannot supply (via USB) the peak current needed by the module. That would be easy to test, just plug the module into a different computer with a different USB cable. Editor’s note: the D1 Mini modules we supply with Analog Clock Driver kits (when that option is selected) do not seem to have this problem, as none of the customers that bought them have complained about it. Pi Pico BackPack touch sensing is not working I’ve purchased and assembled the Raspberry Pi Pico BackPack kit (SC6075, March 2022; siliconchip. au/Article/15236) and tried the precompiled examples of Arduino and MicroPython code. In both cases, the touchscreen is not working. I tested the display with another Pico board, and everything is working. I’ve checked all joints and connections and am pretty sure that the circuit board is assembled correctly. I wonder if I’m missing something, if you’ve had similar feedback from other readers, or have any advice. (R. Z., Fitzgibbon, Qld) ● Looking at your photos, it seems that you have JP2 set incorrectly, which can interfere with communication with the touch controller. You should set JP2 to open, as can be seen in the photo on page 37 of the March 2022 issue. If you can’t control the backlight, you should also check the position on January 2023  105 JP1, which appears to be different in your photos. If you have trouble with the IR receiver, remove the 1kW resistor adjacent to the IR receiver; you can see that it has been left off our prototype as well. Testing the Amplifier Clipping Indicator Firstly, thanks for a great magazine; it is always a good read. I have built a stereo version (with a single indicator LED) of the Clipping Indicator project (March 2022 issue; siliconchip.au/Article/15240). I was planning on installing it in an Ultra-LD Mk.3 200W Amplifier and wanted to know if there is a preferred test procedure for checking the operation of the clipping boards. I have it wired into the chassis and have been trying to check its operation by briefly jumpering the “amp out” connections on the clipping indicator boards to the positive and negative power rails. When jumpered to the positive rail, the indicator LED lights as I would expect. However, the LED does not light when I do the same with the negative rail. The result is the same when testing both boards. The component values are correct and in their correct orientations; can you offer any advice? (J. M., Auckland, NZ) ● Make sure you have used the correct resistor and zener diodes for the ±57V amplifier power supply as per Table 1 in the article. You can test the negative operation by connecting the “to amplifier output” to the negative supply and checking that the base of transistor Q3 is pulled to about 0.6-0.7V above the negative supply. If not, check the value of the two 100kW resistors at the collector of Q2. Reduce the value of the 100kW resistor that connects directly to Q2’s collector if Q3 is not switching on. Also, pin 2 of IC1 should be pulled below 1.5V with respect to 0V when the input goes to the negative rail. If this voltage is not low enough, it will not trigger the monostable to drive the LED. ZD4 or ZD5 may need to be a lower voltage type if the pin 2 voltage of IC1 is not dropping low enough. On further discussions with the constructor, he found that ZD3 (3.9V) was conducting at a low 2.8V preventing pin 2 of IC1 from going low enough to trigger. Using a 4.7V zener solved it. Optimising High Power Ultrasonic Cleaner After a long break from trying to get the High Power Ultrasonic Cleaner (September-October 2020; siliconchip. au/Series/350) to work correctly, I have the following results. I have added and removed turns using three different pot core formers with little difference in results. After recalibration, with a 12V supply from either a 5A plugpack or LiPo battery, I get the same results. TP1 measures 4.18V (full power). I measured the frequency with a CRO and multimeter as 97.9kHz; no resonance, of course. Interestingly, there is only 134mV across each 0.1W resistor. Is that a clue? In diagnostic mode, I measure 40.9kHz with 4.13V at TP1 and 152mV across the 0.1W resistor (the October 2020 issue suggests around 300mV). I get similar results with the original 57 turns and 18 extra turns on the transformer primary. I replaced the BC547s, suspecting they were of low quality, but it made no difference. Any GPS-Synchronised Analog Clock with long battery life ➡ Convert an ordinary wall clock into a highlyaccurate time keeping device (within seconds). ➡ Nearly eight years of battery life with a pair of C cells! ➡ Automatically adjusts for daylight saving time. ➡ Track time with a VK2828U7G5LF GPS or D1 Mini WiFi module (select one as an option with the kit; D1 Mini requires programming). ➡ Learn how to build it from the article in the September 2022 issue of Silicon Chip (siliconchip. au/Article/15466). Check out the article in the November 2022 issue for how to use the D1 Mini WiFi module with the Driver (siliconchip.au/Article/15550). Complete kit available from $55 + postage (batteries & clock not included) siliconchip.com.au/Shop/20/6472 – Catalog SC6472 106 Silicon Chip Australia's electronics magazine siliconchip.com.au suggestions would be greatly appreciated. (P. J., Aberfoyle Park, SA) ● The 40.9kHz and 4.13V at TP1 in diagnostic mode means you should have full power drive to the transducer. The voltage across the 0.1W resistors is switched at the piezo drive frequency with a 50% duty cycle. A reading of this voltage depends on a meter’s accuracy at that frequency. You would get low readings if the multimeter cannot operate accurately at that frequency. The value of the 0.1W resistors can be checked by measuring with a multimeter set to ohms with the power off. After running diagnostic mode, the settings should be stored for use each time it is powered. If this does not occur, you likely have a faulty microcontroller. We know of at least one other constructor who needed to replace the microcontroller when it did not store the values correctly. Another version of the R80 Aviation Receiver I built the R80 Aviation band receiver kit V6.2 En with the squelch modification (November 2021; siliconchip. au/Article/15101). However, I’m having squelch problems with the updated V7.1 En – Issue A kit. Any ideas? (J. E., Goonellabah, NSW) ● Andrew Woodfield replies: the V7.1 kit contains major changes from the original. These include changing to a very simple squelch, possibly to save cost. Without access to that kit, it’s not easy to derive a good solution. These receivers typically use a noise or signal-level squelch. Since the devices used (two TA2003 chips) do not provide those functions, adding an external squelch circuit is the best option. A thorough search through older amateur radio magazines and handbooks from the 1970s and 1980s, before squelch was integrated into receiver ICs, may locate a suitable circuit. You could also look at mobile transceiver mute circuits in markets where AM was used for commercial VHF mobile radio, such as the UK, Australia and New Zealand. In New Zealand, for example, Tait’s T510 VHF AM mobile from the late 80s was a very good performer in this regard, being designed at the end of the VHF AM mobile radio era. The T510’s receiver actually used a very popular Motorola MC3357 FM(!) chip with an external AM squelch and noise blanker. An extract of that part of the circuit is reproduced below. Sensing pieces on a chessboard Would it be possible to program microchips to sense chess pieces on a board? Also, can I turn on and off LED lights connected to each of the squares, so the board will know which piece is which? I am a year 12 student doing major work and want to do this, but I don’t have the programming ability. (L. P., via email) ● That is certainly possible, and similar things have been done. This Instructable has a sensor for each square and a mechanism to move the pieces automatically: siliconchip.au/ link/abi1 As you can see, it uses extra chips (as well as a microcontroller board) to handle the numerous inputs needed to check all 64 squares. Of course, you would need a way of detecting the pieces. Magnets and reed switches (as used in the Instructable) are simple and robust. The squelch portion of the Tait T512 VHF AM mobile circuit. siliconchip.com.au Australia's electronics magazine January 2023  107 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. 01/23 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 ATtiny85V-10PU ATtiny816 PIC10F202-E/OT PIC10LF322-I/OT PIC12F1572-I/SN PIC12F617-I/P Digital FX Unit (Apr21) Si473x FM/AM/SW Digital Radio (Jul21), 110dB RF Attenuator (Jul22) RGB Stackable LED Christmas Star (Nov20) Shirt Pocket Audio Oscillator (Sep20) ATtiny816 Development/Breakout Board (Jan19) Ultrabrite LED Driver (with free TC6502P095VCT IC, Sep19) Range Extender IR-to-UHF (Jan22) LED Christmas Ornaments (Nov20; versions), Nano TV Pong (Aug21) Model Railway Level Crossing (two required – $15/pair) (Jul21) Range Extender UHF-to-IR (Jan22) PIC12F617-I/SN Model Railway Carriage Lights (Nov21) PIC12F675-I/P Train Chuff Sound Generator (Oct22) PIC12F675-I/SN Tiny LED Xmas Tree (Nov19) PIC16F1455-I/P Digital Lighting Controller Slave (Dec20), Auto Train Controller (Oct22) PIC16F1455-I/SL Ol’ Timer II (Jul20), Battery Multi Logger (Feb21) PIC16LF1455-I/P New GPS-Synchronised Analog Clock (Sep22) PIC16F1459-I/P Fan Controller & Loudspeaker Protector (Feb22) Secure Remote Mains Switch Receiver (Jul22) PIC16F1459-I/SO Multimeter Calibrator (Jul22), Buck/Boost Charger Adaptor (Oct22) PIC16F15214-I/SN Improved SMD Test Tweezers (Apr22), Tiny LED Icicle (Nov22) PIC16F1705-I/P Flexible Digital Lighting Controller (Oct20) Digital Lighting Controller Translator (Dec21) PIC16F18146-I/SO Digital Boost Regulator (Dec22) PIC16LF15323-I/SL Secure Remote Mains Switch Transmitter (Jul22) W27C020 Noughts & Crosses Computer (Jan23) ATSAML10E16A-AUT PIC16F18877-I/P PIC16F18877-I/PT PIC16F88-I/P High-Current Battery Balancer (Mar21) USB Cable Tester (Nov21) Dual-Channel Breadboard PSU Display Adaptor (Dec22) Battery Charge Controller (Dec19 / Jun22) Railway Semaphore (Apr22) PIC24FJ256GA702-I/SS Wide-Range Ohmmeter (Aug22) 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) RCL Box (Jun20), Digital Lighting Controller Micromite Master (Nov20) 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 ATmega644PA-AU AM-FM DDS Signal Generator (May22) 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) PIC32MX795F512H-80I/PT Touchscreen Audio Recorder (Jun14) $25 MICROS dsPIC33FJ64MC802-E/SP dsPIC33FJ128GP306-I/PT 1.5kW Induction Motor Speed Controller (Aug13) CLASSiC DAC (Feb13) $30 MICROS PIC32MZ2048EFH064-I/PT DSP Crossover/Equaliser (May19), Low-Distortion DDS (Feb20) DIY Reflow Oven Controller (Apr20), Dual Hybrid Supply (Feb22) KITS, SPECIALISED COMPONENTS ETC RASPBERRY PI PICO W BACKPACK siliconchip.com.au/Shop/ NEW GPS(/WIFI)-SYNCHRONISED ANALOG CLOCK (SEP & NOV 22) Includes the PCB, all required onboard parts (excluding optional debug interface) and the front panel. Just add a signal source, case, power supply and wiring (see page 37, January 2023) $100.00 VGA PICOMITE KIT (CAT SC6417) (JUL 22) DUAL-CHANNEL BREADBOARD PSU MULTIMETER CALIBRATOR KIT (CAT SC6406) (JUL 22) 110dB RF ATTENUATOR SHORT-FORM KIT (CAT SC6420) (JUL 22) BUCK-BOOST LED DRIVER KIT (CAT SC6292) (JUN 22) SPECTRAL SOUND MIDI SYNTH KIT (CAT SC6261) (JUN 22) IMPROVED SMD TEST TWEEZERS KIT (CAT SC5934) (APR 22) RASPBERRY PI PICO BACKPACK KIT (CAT SC6075) (MAR 22) 500W AMPLIFIER HARD-TO-GET PARTS (CAT SC6019) (APR 22) HUMMINGBIRD AMPLIFIER (CAT SC6021) (DEC 21) SMD TRAINER COMPLETE KIT (CAT SC5260) (DEC 21) Complete kit: includes all parts in the parts list, except the DS3231 real-time clock IC (Cat SC6625; see page 56, January 2023) - DS3231 real-time clock SOIC-16 IC (Cat SC5103) - DS3231MZ real-time clock SOIC-8 IC (Cat SC5779) Q METER SHORT-FORM KIT (CAT SC6585) (JAN 23) $85.00 $7.50 $10.00 (JAN 23) (DEC 22) Power Supply kit: complete kit with a choice of red + green, yellow + cyan or orange + white knob colours (Cat SC6571; see page 38, Dec22) Display Adaptor kit: complete kit (Cat SC6572; see page 45, Dec22) DIGITAL BOOST REGULATOR KIT (CAT SC6597) (DEC 22) LC METER MK3 (NOV 22) BUCK/BOOST CHARGER ADAPTOR KIT (CAT SC6512) (OCT 22) MINI LED DRIVER (SEP 22) WiFi PROGRAMMABLE DC LOAD (SEP 22) Complete kit that also includes all optional components (see page 87, Dec22) Short Form Kit: includes the PCB and all non-optional onboard parts, except the case, front panel label and power supply (Cat SC6544) $40.00 $50.00 $30.00 $65.00 Includes everything in the parts list (see page 64) except the Buck/Boost LED Driver (see adjacent; Cat SC6292) $40.00 Complete Kit: includes everything in the parts list (Cat SC6405) Short Form Kit: includes all SMDs, the power Mosfets, four 0.02W 3W resistors and the VXO7805 regulator module (Cat SC6399) - laser-cut 3mm clear acrylic side panel (SC6514) - 3.5-inch TFT LCD touchscreen (Cat SC5062) WIDE-RANGE OHMMETER (CAT SC4663) (AUG 22) $25.00 $85.00 $7.50 $35.00 Partial Kit: includes the PCB, programmed micro, all SMDs, most semiconductors, PPS capacitors and calibration resistors $75.00 - 16x2 alphanumeric LCD with blue backlighting (Cat 5759) $10.00 GPS-Version Kit: includes everything in the parts list with the VK2828 GPS module (Cat SC6472; see Sep22 p63) $55.00 WiFi-Version Kit: includes everything in the parts list with the D1 Mini module instead (Cat SC6472; D1 Mini is supplied not programmed, see Nov22 p76) $55.00 - VK2828U7G5LF GPS module with antenna and cable (Cat SC3362) $25.00 Complete kit with everything needed to assemble the board, you just require a few external parts such as a power supply, keyboard and monitor $35.00 Complete kit with everything needed to assemble the board Includes the PCB, programmed micro, OLED and all other on-board parts Complete kit with everything needed to assemble the board Complete kit including all programmed PICs (no case or power supply) $45.00 $75.00 $80.00 $200.00 Complete kit with PCBs, all onboard parts, new microcontroller and gold-plated header pins to use for the tips. Does not include a lithium coin cell $35.00 Complete kit, includes all parts except the optional DS3231 IC $80.00 All the parts marked with a red dot in the parts list, including the 12 output transistors, driver transistors, VAS transistors, input pair (2SA1312), BAV21 & UF4003 diodes, TL431 ICs, 75pF capacitor, E96 series resistors and 10kW 1W resistor $190.00 Hard-to-get parts includes: two 0.22W 5W resistors; plus one each of an MJE15034G, MJE15035G, KSC3503DS & 220pF 250V C0G ceramic capacitor Includes PCB & all on-board components, except for a TQFP-64 footprint device *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. $15.00 $20.00 PRINTED CIRCUIT BOARDS & CASE PIECES PRINTED CIRCUIT BOARD TO SUIT PROJECT ARDUINO DCC BASE STATION NUTUBE VALVE PREAMPLIFIER TUNEABLE HF PREAMPLIFIER 4G REMOTE MONITORING STATION LOW-DISTORTION DDS (SET OF 5 BOARDS) NUTUBE GUITAR DISTORTION / OVERDRIVE PEDAL THERMAL REGULATOR INTERFACE SHIELD ↳ PELTIER DRIVER SHIELD DIY REFLOW OVEN CONTROLLER (SET OF 3 PCBS) 7-BAND MONO EQUALISER ↳ STEREO EQUALISER REFERENCE SIGNAL DISTRIBUTOR H-FIELD TRANSANALYSER CAR ALTIMETER RCL BOX RESISTOR BOARD ↳ CAPACITOR / INDUCTOR BOARD ROADIES’ TEST GENERATOR SMD VERSION ↳ THROUGH-HOLE VERSION COLOUR MAXIMITE 2 PCB (BLUE) ↳ FRONT & REAR PANELS (BLACK) OL’ TIMER II PCB (RED, BLUE OR BLACK) ↳ ACRYLIC CASE PIECES / SPACER (BLACK) IR REMOTE CONTROL ASSISTANT PCB (JAYCAR) ↳ ALTRONICS VERSION USB SUPERCODEC ↳ BALANCED ATTENUATOR SWITCHMODE 78XX REPLACEMENT WIDEBAND DIGITAL RF POWER METER ULTRASONIC CLEANER MAIN PCB ↳ FRONT PANEL NIGHT KEEPER LIGHTHOUSE SHIRT POCKET AUDIO OSCILLATOR ↳ 8-PIN ATtiny PROGRAMMING ADAPTOR D1 MINI LCD WIFI BACKPACK FLEXIBLE DIGITAL LIGHTING CONTROLLER SLAVE ↳ FRONT PANEL (BLACK) LED XMAS ORNAMENTS 30 LED STACKABLE STAR ↳ RGB VERSION (BLACK) DIGITAL LIGHTING MICROMITE MASTER ↳ CP2102 ADAPTOR BATTERY VINTAGE RADIO POWER SUPPLY DUAL BATTERY LIFESAVER DIGITAL LIGHTING CONTROLLER LED SLAVE BK1198 AM/FM/SW RADIO MINIHEART HEARTBEAT SIMULATOR I’M BUSY GO AWAY (DOOR WARNING) BATTERY MULTI LOGGER ELECTRONIC WIND CHIMES ARDUINO 0-14V POWER SUPPLY SHIELD HIGH-CURRENT BATTERY BALANCER (4-LAYERS) MINI ISOLATED SERIAL LINK REFINED FULL-WAVE MOTOR SPEED CONTROLLER DIGITAL FX UNIT PCB (POTENTIOMETER-BASED) ↳ SWITCH-BASED ARDUINO MIDI SHIELD ↳ 8X8 TACTILE PUSHBUTTON SWITCH MATRIX HYBRID LAB POWER SUPPLY CONTROL PCB ↳ REGULATOR PCB VARIAC MAINS VOLTAGE REGULATION ADVANCED GPS COMPUTER PIC PROGRAMMING HELPER 8-PIN PCB ↳ 8/14/20-PIN PCB ARCADE MINI PONG Si473x FM/AM/SW DIGITAL RADIO 20A DC MOTOR SPEED CONTROLLER MODEL RAILWAY LEVEL CROSSING COLOUR MAXIMITE 2 GEN2 (4 LAYERS) BATTERY MANAGER SWITCH MODULE ↳ I/O EXPANDER NANO TV PONG LINEAR MIDI KEYBOARD (8 KEYS) + 2 JOINERS ↳ JOINER ONLY (1pc) DATE JAN20 JAN20 JAN20 FEB20 FEB20 MAR20 MAR20 MAR20 APR20 APR20 APR20 APR20 MAY20 MAY20 JUN20 JUN20 JUN20 JUN20 JUL20 JUL20 JUL20 JUL20 JUL20 JUL20 AUG20 NOV20 AUG20 AUG20 SEP20 SEP20 SEP20 SEP20 SEP20 OCT20 OCT20 OCT20 NOV20 NOV20 NOV20 NOV20 NOV20 DEC20 DEC20 DEC20 JAN21 JAN21 JAN21 FEB21 FEB21 FEB21 MAR21 MAR21 APR21 APR21 APR21 APR21 APR21 MAY21 MAY21 MAY21 JUN21 JUN21 JUN21 JUN21 JUL21 JUL21 JUL21 AUG21 AUG21 AUG21 AUG21 AUG21 AUG21 PCB CODE Price 09207181 $5.00 01112191 $10.00 06110191 $2.50 27111191 $5.00 01106192-6 $20.00 01102201 $7.50 21109181 $5.00 21109182 $5.00 01106193/5/6 $12.50 01104201 $7.50 01104202 $7.50 CSE200103 $7.50 06102201 $10.00 05105201 $5.00 04104201 $7.50 04104202 $7.50 01005201 $2.50 01005202 $5.00 07107201 $10.00 SC5500 $10.00 19104201 $5.00 SC5448 $7.50 15005201 $5.00 15005202 $5.00 01106201 $12.50 01106202 $7.50 18105201 $2.50 04106201 $5.00 04105201 $7.50 04105202 $5.00 08110201 $5.00 01110201 $2.50 01110202 $1.50 24106121 $5.00 16110202 $20.00 16110203 $20.00 16111191-9 $3.00 16109201 $12.50 16109202 $12.50 16110201 $5.00 16110204 $2.50 11111201 $7.50 11111202 $2.50 16110205 $5.00 CSE200902A $10.00 01109201 $5.00 16112201 $2.50 11106201 $5.00 23011201 $10.00 18106201 $5.00 14102211 $12.50 24102211 $2.50 10102211 $7.50 01102211 $7.50 01102212 $7.50 23101211 $5.00 23101212 $10.00 18104211 $10.00 18104212 $7.50 10103211 $7.50 05102211 $7.50 24106211 $5.00 24106212 $7.50 08105211 $35.00 CSE210301C $7.50 11006211 $7.50 09108211 $5.00 07108211 $15.00 11104211 $5.00 11104212 $2.50 08105212 $2.50 23101213 $5.00 23101214 $1.00 For a complete list, go to siliconchip.com.au/Shop/8 PRINTED CIRCUIT BOARD TO SUIT PROJECT TOUCHSCREEN DIGITAL PREAMP ↳ RIBBON CABLE / IR ADAPTOR 2-/3-WAY ACTIVE CROSSOVER TELE-COM INTERCOM SMD TEST TWEEZERS (3 PCB SET) USB CABLE TESTER MAIN PCB ↳ FRONT PANEL (GREEN) MODEL RAILWAY CARRIAGE LIGHTS HUMMINGBIRD AMPLIFIER DIGITAL LIGHTING CONTROLLER TRANSLATOR SMD TRAINER 8-LED METRONOME 10-LED METRONOME 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) REMOTE GATE CONTROLLER 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 30V 2A BENCH SUPPLY MAIN PCB ↳ FRONT PANEL CONTROL PCB 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 DATE SEP21 SEP21 OCT21 OCT21 OCT21 NOV21 NOV21 NOV21 DEC21 DEC21 DEC21 JAN22 JAN22 JAN22 JAN22 JAN22 JAN22 FEB22 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 OCT22 OCT22 NOV22 NOV22 NOV22 NOV22 DEC22 DEC22 DEC22 DEC22 PCB CODE 01103191 01103192 01109211 12110121 04106211/2 04108211 04108212 09109211 01111211 16110206 29106211 23111211 23111212 15109211 15109212 01101221 01101222 01102221 26112211/2 11009121 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 04105221 04105222 09109221 09109222 24110222 24110225 24110223 CSE220503C CSE200603 08108221 16111192 04112221 04112222 24110224 01112221 Price $12.50 $2.50 $15.00 $30.00 $10.00 $7.50 $5.00 $2.50 $5.00 $5.00 $5.00 $5.00 $7.50 $2.50 $2.50 $7.50 $5.00 $5.00 $7.50 $20.00 $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 $7.50 $2.50 $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 PICO W BACKPACK Q METER MAIN PCB ↳ FRONT PANEL (BLACK) NOUGHTS & CROSSES COMPUTER GAME BOARD ↳ COMPUTE BOARD JAN23 JAN23 JAN23 JAN23 JAN23 07101221 CSE220701 CSE220704 08111221 08111222 $5.00 $5.00 $5.00 $12.50 $12.50 NEW PCBs We also sell the Silicon Chip PDFs on USB, RTV&H USB, Vintage Radio USB and more at siliconchip.com.au/Shop/3 Using a 3.5in touchscreen with Explore 64 (Micromite Plus) I have a few models of the Micromite BackPack and was wondering if I could use the 3.5in LCD (from a Micromite BackPack V3) with an Explore 64 (August 2016; siliconchip.au/Article/10040). This would allow me to take advantage of the MM+ programming on a larger screen, as the Backpack V3 is incapable of MM+ programming. (M. H., Quakers Hill, NSW) ● Geoff Graham responds: The ILI9488 controller in the 3.5in LCD is supported on both the standard Micromite and Micromite Plus. This is done via an embedded CFunction and works well on both chips. The CFunction is included in the Micromite firmware download under “Display Drivers”. Driving LEDs for each square would be done most easily with LEDs that include a WS2812 (or similar) driver chip. You could use an Arduino Mega (as it has nearly 70 usable I/O pins) to directly read a reed switch for each square (with magnets fitted to the underside of the pieces). For the LEDs, I would use a string of the WS2812based LEDs. Arduino boards and the IDE are good for getting started in programming microcontrollers. There are also forums that may offer suggestions. Another way to handle many inputs and outputs is to use shift register ICs. Our Stackable Christmas Tree shows how you can use shift registers to drive many LEDs (November 2018 issue; siliconchip.au/Article/11297). (a slight downgrade at 100MHz compared to 175MHz). We recommend that you simultaneously replace the KSC2690As with the complementary TTC004B so that you have matching NPN/PNP driver transistors. Both types are available and in stock in reasonable numbers at the time of writing this. As for the FJA4313/FJA4213, there is no apparent direct replacement, but they are both still currently available, and we have plenty in stock. So we will be able to continue to supply them for quite some time. They are part of our ‘hard-to-get parts’ set for the SC200, Cat SC4140 (siliconchip. au/Shop/20/4140). SC200 Amp transistors discontinued I am looking for a PCB for an old Silicon Chip amplifier project. I am interested in the TO-3 version of the SC480 but cannot locate a PCB anywhere. (M. C., Armidale, NSW) ● We don’t sell PCBs for the SC480 because it is obsolete and was replaced by the SC200 (January-March 2017; siliconchip.au/Series/308). The SC200 delivers more power than the SC480 in a smaller package that’s no more difficult to build. It has a flatter frequency response and generally lower distortion. We suggest you consider building it instead. We have PCBs for the SC200 and can also supply all the harder-to-get parts; see siliconchip.au/Shop/?article=10582 If you still want to build the SC480, we can get PCBs made for you. PCBs generally take about two to three weeks to order. I want to build the SC200 Amplifier (January-March 2017; siliconchip.au/ Series/308), but I am having trouble getting some of the transistors. The KSA1220A PNP driver transistors are obsolete and no longer available although, oddly, the complementary KSC2690A NPN transistors are still in production. Also, the FJA4313 PNP output transistors have been announced as ‘end of life’, although they are still available. Again, the NPN equivalent (FJA4213) is still an active part. It’s odd that they are discontinuing one but not the other. Regardless, what parts should I use to build an SC200 module? (E. Z., Turramurra, NSW) ● You are right that it is strange the way they are discontinuing one-half of a pair of transistors. For the KSC1220A, we recommend you use the TTA004B transistor, which is pin-compatible and has the same voltage rating, higher current rating (1.5A vs 1.2A) and a sufficiently high transition frequency 110 Silicon Chip The SC480 Amplifier Module is obsolete Sourcing parts to build the CLASSiC DAC I was wondering if it was still possible to build the CLASSiC DAC (February-May 2013; siliconchip.au/ Australia's electronics magazine Series/63). Reviewing the parts list, much of it has been EOL or out of stock (as most semiconductors are these days). That isn’t really surprising given the age of the design. Do you have plans to revisit it? To give some context, I’m looking to build this as my $50 ‘Gumtree special’ Denon AVR is flaking out (bulging PSU caps), and I want to build the SC200 Amplifier (January-March 2017; siliconchip.au/Series/308). The problem is that most of my system is digital, with the only analog component being the turntable. Without a DAC, the whole project is a non-starter. The AP5002SG is unobtanium. DigiKey has the CS8416K-CZZ, CS4398KCZZ and PLL1708DBQ available. I’m assuming the K variants are a revised model; they look the same from a quick glance. I thought the Cirrus Logic chips were unavailable; I must have glossed over the K versions. The Si4804DY and IRF7309 Mosfets are still available. I was planning to omit the USB part of the circuit by installing the 1MW pull-down resistor and leaving the rest unpopulated. I don’t need the USB input, and the chip is quite expensive for what it is. I wonder if a USB microcontroller could do the same job for less money nowadays. This leaves two remaining non-­ standard parts: the 3.3V LDO regulator and the TOSLINK sockets. I found the LT1963AEST-3.3 at Mouser. It’s pin-compatible with the specified LDO and has a 1.5A output. It looks like a promising replacement. It’s expensive for a regulator but no big deal for a one-off. Some other regulators are pin-compatible but only have 300mA current output. I don’t know if that is sufficient. I think the intended TOSLINK socket is the Altronics Z1604, which is out of stock online and in the Brisbane store. Jaycar seems to have phased their sockets out. Some sockets on AliExpress look like they may work with some creativity. Are you able to comment more on the TOSLINK socket? The remainder of the parts seem pretty conventional and shouldn’t be a problem to source. (M. T., Ferny Hills, Qld) ● We believe you can still build it. You are right to ask about the availability of the semiconductors. We have stockpiled some of the critical parts as we were planning to produce continued on page 112 siliconchip.com.au MARKET CENTRE Advertise your product or services here in Silicon Chip KIT ASSEMBLY & REPAIR 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 KEITH RIPPON KIT ASSEMBLY & REPAIR: * Australia & New Zealand; * Small production runs. Phone Keith: 0409 662 794 keith.rippon<at>gmail.com 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 FOR SALE LEDsales LEDs and accessories for the DIY enthusiast LEDs, BRAND NAME AND GENERIC LEDs. Heatsinks, LED drivers, power supplies, LED ribbon, kits, components, hardware – www.ledsales.com.au VISIT THE NEW TRONIXLABS parts clearance store for real savings on new parts at clearance prices, with flat rate express delivery Australia-wide – go to https://tronixlabs.com BUSINESS FOR SALE Well known Australian electronics company for a bargain price. GENUINE BUYERS ONLY Phone: 0410600330 FOR SALE OATLEY ELECTRONICS www.oatleyelectronics.com CHECK OUT OUR “BULK BUYS” WITH FREE POSTAGE: * 7 X 5M 12V rolls of white/green/blue LED strips for $29 * 10 X 0.5M 12V LED bars for $50 Both include postage. Add any or as many other items for no more than an extra $8 P+P. www.oatleyelectronics.com Phone: 0428600036 SILICON CHIP ASSORTED BOOKS FOR $5 EACH Electronics and other related subjects – condition varies. Some of the books may have already been sold. See all books at: siliconchip.com.au/link/aawx Email for a quote (bulk discount available), state the number directly below the photo when referring to a book: silicon<at>siliconchip.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 Glyn (02) 9939 3295 or 0431 792 293. WARNING! Silicon Chip magazine regularly describes projects which employ a mains power supply or produce high voltage. All such projects should be considered dangerous or even lethal if not used safely. Readers are warned that high voltage wiring should be carried out according to the instructions in the articles. When working on these projects use extreme care to ensure that you do not accidentally come into contact with mains AC voltages or high voltage DC. If you are not confident about working with projects employing mains voltages or other high voltages, you are advised not to attempt work on them. Silicon Chip Publications Pty Ltd disclaims any liability for damages should anyone be killed or injured while working on a project or circuit described in any issue of Silicon Chip magazine. Devices or circuits described in Silicon Chip may be covered by patents. Silicon Chip disclaims any liability for the infringement of such patents by the manufacturing or selling of any such equipment. Silicon Chip also disclaims any liability for projects which are used in such a way as to infringe relevant government regulations and by-laws. Advertisers are warned that they are responsible for the content of all advertisements and that they must conform to the Competition & Consumer Act 2010 or as subsequently amended and to any governmental regulations which are applicable. siliconchip.com.au Australia's electronics magazine January 2023  111 a new version of this project, but it will be some time before that comes to fruition. So we don’t think that should stop you from building the already-published design, which is still a great performer. Because of the shortages, we temporarily stopped selling the set of critical parts to build the CLASSiC DAC but have now reintroduced the parts set in our Online Shop (siliconchip.au/ Shop/20/1815). You would likely have trouble getting the microcontroller too, but we can supply them programmed from siliconchip.au/Shop/9/1850 Note that the SD card socket used in the original project is now unobtainable. We have redesigned the PCB to use a commonly available type from siliconchip.au/Shop/8/5655 Before ordering anything, go through the parts list and check that you can get everything we don’t sell. While you said you don’t plan to use it, the PCM2902E is available, although stocks are low (it isn’t part of our set). We have added the MIC391003.3WS LDO regulator to our parts set. While alternatives are available, it is the part tested in the design and found to work well. We chose it Advertising Index Altronics.................................23-26 Dave Thompson........................ 111 Digi-Key Electronics...................... 3 Emona Instruments.................. IBC Jaycar................. IFC, 10-11, 38-39, ................................. 57, 79, 99, 103 for its good ripple rejection. While more expensive, we agree that the LT1963AEST-3.3 should work well as an alternative. As for the TOSLINK sockets, they are definitely a problem. Unfortunately, Altronics have told us they will discontinue them, hence the lack of stock. The closest alternatives we are aware of are RS Components Cat 8051677 & 805-1680 or element14 Cat 2991612. They are all currently in stock. You will probably have to cut off the plastic posts as there are no matching holes in the PCB, but they also have solder pins to retain them, plus the rear panel, so we think they should be OK. I have a question about the headphone amplifier from the November 2005 issue of Silicon Chip (“Studio Series Stereo Headphone Amplifier”, siliconchip.au/Article/3231). That module was included in the Studio Series Preamplifier from July 2006, which I purchased as a kit from Altronics circa 2010 and constructed, except for the headphone amplifier part. Now I’m completing the headphone amplifier. The instructions supplied by Altronics say to apply heatsink compound to each output transistor and to avoid using insulating washers in mounting these to the heatsinks. However, the Altronics kit supplied Finding an article on a insulating washers for the output Frequency Switch transistors. Some time in the last 10 years or so, Which approach would you recyou published a project that used the ommend — washers or heatsink comLM2917 frequency/voltage conversion pound? (P. H., Warwick, Qld) chip as a frequency switch. Can you ● Since the heatsinks are separate tell me which issue the project was in and the transistors don’t require isoand whether the kit is still available? lation from them, the ideal mount(P. H., Gunnedah, NSW) ing method is just to use the heatsink ● You can find articles using the Word compound. That gives the best heat Search page on our site: siliconchip. transfer and will keep the transisau/Articles/WordSearch tors running at a lower temperature, Using that to search for projects which is safer. mentioning “LM2917”, you are most However, you can use insulating likely referring to the Frequency-­ washers instead. If the washers are Activated Switch For Cars (June 2007; silicone, thermal transfer compound siliconchip.au/Article/2261). It was is not required. If using mica washavailable as a Jaycar kit (KC5378) ers, thermal compound is needed on which is now discontinued. both sides of the mica sheets. InsulatThat project has been superseded ing bushes are not required in either by the Deluxe Frequency Switch (May case since the transistors do not have 2018; siliconchip.au/Article/11062), exposed metal tabs. SC Keith Rippon Kit Assembly....... 111 LD Electronics........................... 111 LEDsales................................... 111 Oatley Electronics..................... 111 Ocean Controls............................. 7 SC GPS Analog Clock............... 106 SC USB Cable Tester.................. 56 Silicon Chip Shop............ 108-109 Silicon Chip Subscriptions........ 67 The Loudspeaker Kit.com............ 9 Tronixlabs.................................. 111 Wagner Electronics................... 101 112 Silicon Chip Errata and Next Issue Microchip Technology.............OBC Mouser Electronics....................... 4 To insulate or not to insulate LC Meter Mk3, November 2022: (1) the initial release of the PCB has a short circuit between the top middle terminal of S2 and the track above it going to pin D10 of the Nano. This will not stop it from working but will reduce the accuracy of capacitor measurements above about 800pF. Run a sharp knife along the short circuit, taking care not to cut the track above, then verify that the short circuit is gone. (2) the 330pF capacitor shown in the circuit diagram and PCB overlay, and in the parts list, should be 470pF instead. Kits were correctly supplied with two 470pF capacitors and no 330pF capacitors. (3) switch S1 (not used by the provided firmware) is not connected the same way on the PCB as shown in the circuit diagram. Neither of the ‘NO’ and ‘NC’ contacts are connected to GND, and the 15kW pull-down resistor is connected between them. If constructors wish to modify the firmware to use this switch, one end would need to be connected to GND. (4) the supplied HEX file can be uploaded to the Arduino Nano using AVRDUDESS, with the Programmer set to “Arduino” and a baud rate of 57600. Next Issue: the February 2022 issue is due on sale in newsagents by Monday, January 30th. Expect postal delivery of subscription copies in Australia between January 27th and February 14th. Australia's electronics magazine siliconchip.com.au “Rigol Offer Australia’s Best Value Test Instruments” Oscilloscopes NEW 200MHz $649! New Product! Ex GST RIGOL DS-1000E Series RIGOL DS-1000Z/E - FREE OPTIONS RIGOL MSO-5000 Series 450MHz & 100MHz, 2 Ch 41GS/s Real Time Sampling 4USB Device, USB Host & PictBridge 450MHz to 100MHz, 4 Ch; 200MHz, 2CH 41GS/s Real Time Sampling 424Mpts Standard Memory Depth 470MHz to 350MHz, 2 Ch & 4Ch 48GS/s Real Time Sampling 4Up to 200Mpts Memory Depth FROM $ FROM $ FROM $ 429 ex GST 649 ex GST 1,569 ex GST Multimeters Function/Arbitrary Function Generators New Product! RIGOL DG-800 Series RIGOL DG-1000Z Series RIGOL DM-3058E 410MHz to 35MHz 41 & 2 Output Channels 416Bit, 125MS/s, 2M Memory Depth 425MHz, 30MHz & 60MHz 42 Output Channels 4160 In-Built Waveforms 45 1/2 Digit 49 Functions 4USB & RS232 FROM $ 479 FROM $ ex GST Power Supplies 725 ONLY $ ex GST Spectrum Analysers 789 ex GST Real-Time Analysers New Product! RIGOL DP-832 RIGOL DSA Series RIGOL RSA Series 4Triple Output 30V/3A & 5V/3A 4Large 3.5 inch TFT Display 4USB Device, USB Host, LAN & RS232 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 $ 749 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