Fullscreen HTML5Med Res (??MB)HTML4

The History of Last month, we described the rapid developments which took place after Silicon Chip was founded in 1987. That brought us up to 1993, by which time we were ticking along quite nicely and looking to grow the magazine as fast as we could. We even launched our Fifth Birthday Celebration in January 1993, culminating in the award of a brandnew Ford Festiva car to a lucky reader. Part 2 by Leo Simpson O ne small innovation we made around that time was the acquisition of a Polaroid scope camera. It was a DS-34 which used very fast Polaroid film and had a visor that fitted a standard oscilloscope screen (see siliconchip.au/link/abfl). All you had to do was to place the visor over the scope screen and pull the trigger. After a minute or so of film development, the result was a sharp, precisely-focused photo showing the signal traces on the screen. We used this quite frequently, to illustrate circuit operation for many of our project articles. However, digital scopes came out 66 Silicon Chip not too many years after that. It was then a simple matter to take a screen grab of whatever measurements you were doing, automatically saved in JPG (also called JPEG), PNG or TIF format, ready for inclusion in an article. So the relatively expensive Polaroid scope camera was made completely obsolete. Never mind, such is progress. I think it might still be gathering dust somewhere in the Silicon Chip workshop. Silicon Chip to be published in the USA In May 1993, there was a major business development that we had been Australia's electronics magazine working on for some time. We were very proud to announce that Silicon Chip was to be published in the USA and Canada, under licence to Gernsback Publications Inc, of New York. They were the publishers of Popular Electronics and Electronics (formerly Radio Electronics). This was a big coup for us. The arrangement was for them to initially publish four issues a year, with most of the editorial to be reproduced from the Australian issues of Silicon Chip. But soon after the agreement was made, the arrangement hit hurdles as Gernsback asked us for bromides for siliconchip.com.au their initial issue. That shocked us, as we had been producing Silicon Chip using Pagemaker for several years; it had not crossed our minds that they would still be using the old production methods. I cannot remember the details of how we solved those problems, but I do recall that they had to hastily acquire suitable computers and the necessary software. Apparently, very few magazines in the USA were using desktop publishing software at the time, and we were some way ahead of the curve. Ultimately, they only produced one issue, then decided it was all too hard. That was quite disappointing to us (apart from missing out on a revenue stream from the licensing agreement), as we knew from our experience that it took several years to establish a new magazine. In mitigation, the USA and Canada did not have the very efficient newsagency distribution scheme we have in Australia. Most large circulation American magazines were (and still are) primarily sold by subscription. Much later, around 2006, we signed another licensing agreement with Everyday Practical Electronics (EPE) magazine in the UK, now known as Practical Electronics (PE). That agreement continues today. In the meantime, our well-­appointed Mona Vale office had been a very pleasant place to work and we stayed there until January 1994. But I wanted to put the business on a more certain footing. By that time, I felt confident enough to buy into a very large industrial complex on Jubilee Avenue, in the Warriewood Valley. Above: a clipping from a local newspaper with Leo Simpson holding the new American version of Silicon Chip with the Studio Twin 50 Stereo Amplifier shown on the cover. Left: the Editorial from the first American issue of Silicon Chip. The American operation was based in Farmingdale, New York, with the publisher being Larry Steckler. Dolby Pro-Logic decoder Talking of licensing agreements, it was not long after moving to the Jubilee Avenue address that we were able to publish our Dolby Pro-Logic Surround Sound Decoder, in the December 1994 issue (siliconchip.au/Series/162). This was a big project for us, with all of the design work carried out by Technical Editor John Clarke. Significantly, it was sponsored by Jaycar Electronics, who did a lot of liaison work to get the design licensed by Dolby Laboratories. That was necessary for Jaycar to be able to obtain the Dolby chips for the subsequent kits for the project. (There was a second siliconchip.com.au Australia's electronics magazine September 2022  67 The Dolby Pro-Logic Surround Sound Decoder project was sponsored by Jaycar Electronics and our design was approved and licensed by Dolby Laboratories. version of this project several years later). The design prototypes had to be submitted to Dolby Laboratories in America to be approved and to my memory, they required several modifications before the approval was granted. It was a world-first for a technical magazine and was not repeated anywhere else in the world, as far as we know. Interestingly, there was another milestone in the same issue, with the publication of the first article in a series on Bob Young’s Radio Control unit that used surface-mount components (siliconchip.au/Series/198). Zoom magazine Throughout 1995, we featured many articles on car electronics and car modification projects, all generated by a very prolific and enthusiastic writer, Julian Edgar. The circuit designs were prepared by John Clarke. Those articles were so popular that I saw a place for a car magazine covering similar topics. And so it came to pass, with the publication of the first issue of Zoom magazine in April-May 1996. Julian Edgar produced and edited most of the editorial, and Ross Tester (who had previously worked at EA & Dick Smith Electronics) joined our staff to do all the layout and production. It was a bi-monthly magazine in full colour. Zoom was another big step forward for us. It was not only in full colour and more expensive to produce but also required much higher production standards. While we thought the first issue was pretty good, that illusion was soon shattered by Julian Edgar, who was utterly scathing in his assessment of picture quality. Well, that was pretty 68 Silicon Chip hard to swallow but we had to lift our standards substantially and quickly to meet the deadline for the next issue. Julian was a very fine photographer of cars, and he was used to seeing his photos reproduced in motoring magazines. So we, meaning Ross Tester and Greg Swain, had to learn how to get the same high-quality results from our desktop publishing equipment. It meant that we had to have our colour monitors properly calibrated and learn the subtleties of photo processing using Photoshop. With a few issues under our belts, the production standard became very good. But the magazine was not a financial success. While the circulation growth was satisfactory, we had a lot of difficulties in getting the many advertisers to pay us. They were mostly small businesses and their cash flow was often insufficient to justify their advertising commitments. Ultimately, I decided that the magazine was not financially viable for us and we sold it to a specialist publisher, Express Publications, in early 1998. Maybe I should have kept ownership of the Zoom name, though. In the light of “Zoom” meetings today, I might have become a multi-millionaire (or maybe not!). Giving up on Zoom was a setback, but one good aspect was that it meant Ross Tester could work full-time for Silicon Chip as a writer and layout artist. He would really come into his own when we went to full-colour production some years later. In addition, in about August 1997, our regular contract photographer, Glen Keep, decided to retire. So we acquired the key equipment of his studio set-up with flash gear and ‘soft Australia's electronics magazine boxes’. Ross Tester then took over all our photography, initially using his own Minolta film gear and later, Nikon digital cameras and lenses. As well as being a graphic designer, layout artist and clever advertising copywriter for many years at Dick Smith Electronics, Ross had also been a freelance wedding photographer – he was a man of many parts. His photography skills enabled us to achieve a long-term aim – high-quality, finely detailed pictures of all our electronics projects. These were so good that readers building projects could easily see the colour codes on tiny resistors, component numbers on semiconductors and so on. They could even determine if we had used a component that was not exactly the same as depicted in circuits and wiring diagrams. We had to be diligent, and readers loved it. We even tried to ensure that the colour codes on resistors in the diagrams ran the same way as in the photos, so as not to confuse our readers! The same comment applied to series connections of resistors and capacitors – ideally, they had to be in the same order on the circuit, PCB overlay diagram and in the assembled project, even if it didn’t affect circuit function. Otherwise, readers would complain that we had them back-to-front! The introduction of GenCAD The obvious next step in our continuing technology adoption was to go to CAD for our circuit diagrams and drawings, which we did in the latter half of 1995. The package chosen was an MS-DOS-based system called GenCAD, which ran quite well on the hardware of that era. It allowed us to send complete Postscript pages with everything in place to the commerical printers. A year or so later, we also moved from Windows 3.11 to Windows NT, which eliminated all those annoying operating system reboots. While GenCAD had been a great step forward for circuit diagrams, I was still dissatisfied with our wiring diagrams, particularly for large projects like stereo amplifiers, high-power inverters etc. Depicting multi-strand colour ribbon cables was a real challenge. I wanted to have the same standard as that achieved by the American Model Railroader magazine. They used to depict large model railway siliconchip.com.au layouts in full colour with detailed wiring. They would even do dioramas (ie, diagrams with a 3D perspective) of their layouts. That was far beyond the capability of GenCAD. But new software would eventually provide the answers. In 2000, we upgraded our operating systems to Windows 2000. At the same time, we ditched GenCAD and went to CorelDraw for our circuits, PCB overlays and wiring diagrams. Our draftsman devised a clever scheme of creating a component library with red bounding boxes, which all snapped into place on a grid so that everything lined up. We also developed an extensive component library which streamlined the process. Towards the end of 2003, we ditched Pagemaker (originally by Aldus, but by then owned by Adobe) and converted to Adobe InDesign. The latter was substantially more powerful and flexible, particularly when it came to type handling and special type effects. Incidentally, when we went to InDesign, the overwhelming majority of magazine producers, advertising studios and the like had standardised on Quark Express, again mainly on Mac hardware. Typical Silicon Chip – we went against the trend. Fast forward to today, and the vast majority use InDesign. After that, there were mainly just various upgrades to hardware, operating systems and the inevitable frequent software upgrades for InDesign, CorelDraw and Photoshop etc. Using that technique enabled us to provide incredibly sharp images. It would have been great when we were publishing those beautiful photos of cars in Zoom magazine. Moving to four-colour printing Initially, like the vast majority of magazines, Silicon Chip was printed with a four-colour (CMYK – Cyan▪ ▪, Magenta▪ ▪, Yellow▪ ▪ & blacK▪) cover. Still, the inside used ‘spot colour’, where certain pages could have a single second colour applied. As time went on, we printed one or two sections of the magazine in full colour, which allowed us to have photos in some articles in full colour, as well. But most sections of the magazine could still only have spot colour, which looked rather drab by comparison. The move to full four-colour printing came about due to a chance conversation between Ross Tester on a plant visit and the printer’s production manager. The bulk of their work – women’s magazines and catalogs – was printed in four-colour. The production manager was moaning that before Silicon Chip went on the press, they had to remove the C, M The cover of Zoom’s ninth issue, from August/ September 1997. Not long after its publication, in early 1998, Leo Simpson sold the magazine to Express Publications as it turned out to be too much trouble getting some advertisers to pay invoices. Unsharp masking Those software upgrades were often tiresome but they did bring production benefits. One of these was to be revealed when Ross Tester attended one of the many seminars discussing Photoshop’s latest features. It was called “unsharp masking”. While it sounds like something that would reduce photo sharpness, the process gets its name from a traditional photography darkroom technique initially developed in Germany in the 1930s. This was where a negative copy of the original photo is blurred, or “un-sharpened”, and then applied to the original image as a mask. As strange as it sounds, this blurring method actually results in a sharper image (there is a good description of the process on Wikipedia at https://w. wiki/5Vkz). siliconchip.com.au and Y stations and wash one of them down to use the special spot colour ink – then reverse the process to go back to four-colour. “Why don’t you guys print in four-colour? If you must have spot colour, you can get that from a CMYK ink mix”. We expressed the long-held belief that four-colour printing was too expensive. Up to that time, it had been, but when you took into account the press down-time, it came out line-ball. So the printers gave us a four-colour price which was very similar to the spot colour price – and Silicon Chip went all colour! Technology again came to the rescue here, with a technique known as computer-­to-plate or CTP. This digitised the plate-making process by using lasers to etch the plates directly, eliminating the expensive and cumbersome film process (one large piece of film for each CMYK colour). In addition, Kodak had developed the Photo CD process some years earlier – a cost-effective method of scanning 35mm film and placing the resulting files onto a CD. A special Kodak plug-in for Photoshop allowed the files to be retrieved from the CD and converted to JPEG files. Australia's electronics magazine September 2022  69 Pushing the boundaries of audio amplifier performance Leo Simpson operating the Audio Precision System One (bottom of stack), 1kW dummy load (above it, with a brick-wall filter in between) and digital scope (top) to test the 20W Class-A Amplifier. We still use a similar setup, albeit with an AP System Two. It was much cheaper than having colour slides digitised on a drum scanner, meaning it was cost-­effective for Silicon Chip to go to full-colour reproduction by the latter half of 1998. However, as noted above, we did not manage to incorporate full-colour circuits and wiring diagrams until several years later. The move to colour also required hardware upgrades. The Radius monochrome monitor had to be finally retired and high-end colour monitors substituted, and we invested in an expensive CMYK Postscript colour laser printer. The monitors had to be calibrated regularly so that what you saw on-screen matched the printed magazine page. State-of-the-art test equipment While we were grappling with Zoom magazine, other developments had been in train. We had spent quite a lot of money on desktop production equipment but we had also added to our laboratory equipment. In particular, we had acquired several oscilloscopes, including digital models, but we still did not have a really good distortion analyser. Those instruments we did have were quite old and certainly not state-of-the-art. That induced us to purchase the very best audio analyser available at that time, from US company Audio Precision. This represented a substantial outlay for us, but ultimately, I decided 70 Silicon Chip that the expense was justified. It would allow us to measure harmonic and intermodulation distortion down to previously unimaginable levels, as little as 0.0003% or even lower, along with commensurately low noise signal levels (to below -120dB). It brought the great time-saving ‘auto-nulling’ feature as a harmonic distortion test was run over a complete frequency sweep of the entire audio spectrum. That capability, and the ability to produce easy-to-read performance graphs of signal-to-noise ratio, frequency response and distortion curves, gave our audio designs a degree of credibility that could not be achieved in any other way. We started to feature performance graphs from this machine for audio equipment in the February 1995 issue. But the first significant design produced with the Audio Precision unit having been used as an actual design tool was the Plastic Power amplifier in the April 1996 issue (siliconchip.au/ Article/5015), shown above. This design used rugged new plastic-­ e ncapsulated power transistors from Motorola and it was an absolute joy to produce the excellent performance curves with the Audio Precision test set. The Plastic Power amplifier’s lowest distortion level was about 0.004%. It was good, but this amplifier was still far above the noise and distortion limits of the new test equipment. We were a long way above what we would achieve just two years later, in 1998. Australia's electronics magazine It was in July 1998 (siliconchip. au/Series/140) that we produced an amplifier with astonishingly low distortion, as low as 0.00025%. That’s only 2.5 parts per million! But making those extremely low harmonic distortion measurements was not solely due to the Audio Precision equipment, as we shall see. The amplifier in question was a 15W class-A module using “bog-standard” small signal transistors (BC547, 548, 556, 557 etc) and a pair of Motorola MJL21193/94 power transistors operating as current feedback pairs. The PCB was relatively unremarkable in appearance but was attached to an enormous heatsink, required to dissipate the standing quiescent power of 80W. Such high power waste is unavoidable for class-A amplifiers, which was the sole reason we had previously rejected requests from keen ‘audiophile’ readers for a high-­ performance class-A design. But we finally relented. So how did we make the measurements? Harmonic distortion measurements for hifi audio amplifiers are almost always presented as THD, meaning “total harmonic distortion”, ie, that the figure consists of the harmonic distortion plus residual noise (made explicit by writing THD+N). It is usually predominantly the various harmonics of the sinewave test signal, but there is always a noise component, including 50/100Hz hum, but mainly white noise. That wasn’t the case with the THD figures obtained from the class-A amplifier module. While the module’s absolute noise was incredibly low at -113dB (unweighted 22Hz to 22kHz; -116dB A-weighted) with respect to full power, it was still quite a significant amount of noise, often almost obliterating the harmonic components. This was clearly illustrated using a 100MHz analog oscilloscope that had on-screen measurements. We used this to show the noisy residual THD waveforms, as can be seen on page 61 of the July 1998 issue. So we knew that the actual harmonic distortion was actually much lower than the total THD figure. The question remaining was how to remove the noise to reveal the harmonic waveform. The solution was to use a technique described at about siliconchip.com.au that time in an article by noted audio designer Douglas Self in the British magazine Electronics World. It involved using a digital oscilloscope in averaging mode to remove the random noise from the low-level signals, to enable the buried harmonic content to be clearly displayed. And that allowed us to give precise estimates of the actual harmonic content. My Publisher’s Letter in the July 1998 issue has more on this topic. While kits for the design were ultimately not a big seller, the project did demonstrate what was and probably still is the “holy grail” of ultra-linear circuit design: the proverbial “straight wire with gain”. We will never quite get there, but that class-A amplifier is exceptionally close to ideal and far better than any present program source, analog or compact disc, or any audio transducer for that matter. We produced a 20W version of the class-A design in May 2007 and the following months. This had a simplified power supply, a shielded toroidal power transformer and other slight circuit changes and again resulted in some worthwhile performance improvements. Having seen what was possible with a great class-A design, we wondered what could ultimately be achieved with a really good class-AB design. Could we approach or even equal the performance of our class-A design? That was to become our benchmark. And up to that time, such a quest would have been seen as futile since class-AB amplifiers are, or were, inherently less linear. As it turned out, there were several design innovations to come which would help us in that quest. These did eventually allow us to achieve a major advance in class-AB amplifier design to go very close to class-A performance levels (and, in some ways, surpass them). But it took four attempts to get results which we think will now be almost impossible to improve significantly. The Ultra-LD series The first attempt was the Ultra-LD module presented in the March 2000 issue (siliconchip.au/Series/113), a 100W module that was essentially a refinement of the Plastic Power amplifier design featured in the April 1996 issue. The major differences were better output transistors (Motorola siliconchip.com.au MJ15030/MJ15031 and MJ1302A & MJ3281A) in compound current-­ feedback triples. Also, the input and class-A driver stages were fed with regulated supply rails to eliminate hum and noise on those rails. It was significantly better than the Plastic Power module, with lower harmonic distortion and less residual noise. But our next attempt, the Ultra-LD Mk.2 amplifier module in the August 2008 issue (siliconchip. au/Series/51), was considerably better. It had a greater power output (135W into 8W or 200W into 4W), much lower residual noise and again, much lower harmonic distortion. You will have to read the articles in the August & September 2008 issues to gain a full appreciation of all the changes we made. Briefly, they Australia's electronics magazine involved using new five-lead “ThermalTrak” power transistors which had integral power diodes for bias compensation, a modified input circuit with new low-noise transistors and significant modifications to improve the PSRR (power supply rejection ratio). That last innovation allowed us to eliminate the regulated supply rails for the input and driver stages, simplifying amplifier construction. Magnetic field cancellation This was a completely new circuit design compared to the March 2000 module, but the most significant improvement was the radically different double-sided PCB which introduced a break-through concept. The idea was to cancel the considerable magnetic fields generated by the class-B currents in the output stages, September 2022  71 The first Micromite series by Geoff Graham included two projects: the ASCII Video Terminal (at left); and the 44pin Micromite (below). which would otherwise induce distortion signals into the input stage transistors. Again, you will need to read the circuit description in the August 2008 issue to fully understand what we did. I was very proud of the magnetic field cancellation concept. It came about one day when we were trying to reduce the effects of currents in the power supply leads. The standard approach was to twist the positive, negative and 0V rail wires together and then dress them to avoid their deleterious effects on distortion performance. This process’s effect, or lack of effect, was clearly demonstrated by repeated testing with our Audio Precision test set. As we went through this futile process, I suddenly realised that it is impossible to cancel the magnetic fields generated by the positive and negative class-B currents in any amplifier. Why? Because they don’t flow at the same time! The positive rail currents are positive half-wave rectified versions of the signal waveform, while the negative rail currents are negative halfwave rectified versions of the signal. So twisting supply wires and playing with their routing was never going to work. It was utterly futile! But the new PCB did achieve magnetic field cancellation. John Clarke devised an ingenious layout for the top and bottom side power tracks. He carefully arranged the whole circuit to minimise the induction of distortion signals into the input stages and it worked brilliantly! However, a few years later, we had to revise the design again, mainly due to shortcomings in the claimed benefits of the ThermalTrak power transistors in preventing thermal drift and eliminating the need for adjustments. We presented the revised design in the July 2011 issue (siliconchip. au/Series/286). And again, the new design further improved the distortion performance. Could we do any better? At the time, I didn’t think so. Well, I was wrong. Again! In July 2015 up to the October 2015 issue (siliconchip.au/Series/289), Nicholas Vinen presented a radical re-design of the PCB using SMD transistors for the low signal level stages, SMD emitter resistors for the output transistors and a ground plane to shield the input stages. Notably, he also realised that the air-cored inductor in the output filter was generating a magnetic field that interfered constructively or destructively with the remaining magnetic field generated by the supply tracks on the PCB. This led to the idea of adjusting its orientation and value until maximum cancellation was achieved, then changing the other filter components so this did not impact the way the filter operated. That was the last change that got the distortion curve of the amplifier to track below that of the earlier 15W & 20W class-A designs. The result was quite remarkable. But none of these achievements would have been possible without our stateof-the-art test equipment. Mind you, while many of the solid-­ state amplifiers described above were undoubtedly popular, there was another design that was definitely not state-of-the-art, but it was nonetheless very popular. That was the Currawong stereo amplifier (November 2014-­January 2015; siliconchip.au/ Series/277), which was a real winner, and is still popular today. The attraction? The glowing magic of valves! I have detailed this epic quest for audio perfection because it illustrates the tireless work done by the Silicon Chip design team and many The Micromite Explore 100 was one of the more advanced Micromite-based projects (September-October 2016; siliconchip.com.au/Series/304) Australia's electronics magazine contributors over the years. The aim was to present the very best circuits we could, involving analog or digital technology, using the latest components and leading-­edge techniques. I also need to make special mention of the PIC32 microcontrollers and the Maximite (siliconchip.com.au/ Series/30) & Micromite (siliconchip. au/Series/261) series of projects developed by Geoff Graham to allow those micros to be programmed in BASIC. In terms of overall impact and popularity, these had far more impact than any of our audio projects. Indeed, a search of the internet will reveal countless mentions of Micromite, and it was all originally conceived by Geoff Graham (https:// geoffg.net/). Enter Nicholas Vinen In the latter years, Nicholas Vinen played a significant part in circuit design and most other aspects of Silicon Chip. He introduced himself sometime in 2009 and claimed that he had produced a digital-to-analog converter (DAC) that was a world-beater. Naturally, I challenged him to prove that it was as good as he said it was by putting it through a battery of tests with our Audio Precision gear while he looked on. It bombed out. That did not faze Nicholas in the least. He immediately got the message that unless you tested, tested and tested again, there was no way that you could make any changes to a design and hope for some improvement in results. He came back to our workshop quite a few times after that. He would go straight to the test equipment and run through another set of tests with the latest iteration of his design. In fact, he quickly became much more adept than I was at running the equipment. He learned very fast, the clever sod. And eventually, his DAC was a great design and we published it in the September-­November 2009 issues (siliconchip.au/Series/4). After that, we couldn’t keep him away from the place and he joined the staff in February 2010. I was very glad to welcome him on board, and his importance to the magazine grew continuously from that point until he took over Silicon Chip when I retired in July 2018. Postscripts All that remains in this story is to siliconchip.com.au The Ultra-LD Mk.4 Amplifier was the latest iteration of the Ultra-LD series. It was followed by the simpler SC200 (January 2017). briefly mention what happened to all our competitors. We started with three other electronics magazines against us in the market: Electronics Australia (EA), Electronics Today International (ETI) and Australian Electronics Monthly (AEM). Plus, we had trade and overseas magazines in the Australian market. Virtually every one of them has gone, with a few overseas exceptions. AEM dropped out relatively early, while ETI kept going until April 1990. But Electronics Australia kept going strong until 1999, finally fizzling out in January 2001. Federal Publishing then launched a hybrid publication called “EAT”. It lasted for five issues: April 2001, May 2001, June 2001, July/ August 2001 and September/October 2001. So Silicon Chip is now one of very few electronics magazines with DIY projects in the world. Funnily enough, seeing all our competitors fall by the wayside really did not give us a great deal of satisfaction. As far as we were concerned, they had ceased to be relevant years before, as the internet tidal wave rolled over everything. But there are a couple of satisfying postscripts. The first of these involved Jim Rowe. He was initially a long-time staff writer at EA from March 1960 (when it was Australia's electronics magazine still “Radio, TV & Hobbies”), becoming Technical Editor in 1965 when it was renamed to EA and Editor in April 1971. He left EA in 1979 and went to work at Dick Smith Electronics (DSE), becoming Technical Director. After Gary Johnston left DSE to start Jaycar Electronics in August 1983, Jim became marketing director of DSE but resigned shortly after, in March 1984. He then joined Federal Publishing as Managing Editor of their electronics and computer magazines (including EA, which they acquired later in the same year). In October 1985, he left Federal Publishing and worked at Applied Technologies (MicroBee) for a short time. Ultimately, he went back to run Electronics Australia after I was dismissed in early 1987. Lightning then struck again, and Jim Rowe and EA parted ways in August 1999. This was great for us. With some of us having worked with Jim Rowe in the 1960s and 70s, we knew him to be a highly qualified and extremely knowledgeable designer/writer. We invited him back, and he joined us in late 1999. That was a very significant development for the long-time staffers of EA and Silicon Chip. It meant that ...continued on page 75 September 2022  73 Leo’s early days at Electronics Australia Readers may wonder how Leo Simpson rose to the position of Managing Editor at Electronics Australia and then went on to start an entirely new magazine in competition to EA. Leo takes up the story... My first encounter with EA magazine was almost 60 years ago, involving the August 1963 issue. I was working as a clerk in my first full-time job after leaving school, at the Defaults department in the Australian Taxation Office. A fellow worker had just finished reading the issue, at that time called Radio, TV & Hobbies, and he threw it over to me, saying that I “might be interested”. That turned out to be an understatement. Until then, I had no interest or knowledge of electronics, although I had enrolled in a Science degree at the University of NSW (instead of doing a TAFE course in accountancy, the standard choice of my clerical workmates). I read that magazine from cover to cover that very day and then I read every back issue and any books that I could find on the subject. I became interested in hifi and then haunted the university library for every magazine on that topic and anything remotely related (to the detriment of my studies). In short order, I decided that I would change my degree course to Electrical Engineering at the end of the year. Also at the end of that year, I was extremely fortunate to gain a position as a cadet engineer at Ducon Condenser Pty Ltd, at their vast Leightonfield plant in Sydney’s western suburbs. I was one of only three cadet engineers taken on that year from about 600 applicants. The Ducon plant was a huge operation with over 2000 employees, making a vast range of passive electronic components such as all types of capacitors, resistors and potentiometers for Australia’s booming radio, TV and stereogram manufacturers. Ducon also made massive power engineering components for high voltage switch-yards at power stations, such as three-phase reactors weighing many tonnes. Over the next two years, I worked in most of the manufacturing and engineering departments of Ducon and enjoyed it immensely, learning a great deal. But that suddenly ceased when my university results came in, and I had failed two years running. I was out of a job, which really was 74 Silicon Chip a shock. I had no one to blame but myself since I was a hopeless student, utterly bored by the course subjects. Moving to EMI Only a few days afterwards, I started working at EMI (Electric Musical Industries, manufacturers of His Master’s Voice products) at Homebush, in Sydney’s west. Their products included TV sets, stereograms, radios and car radios. I was assistant to the Quality Control (QC) manager, Fred Stirk, and my job was to write QC procedures for all of the above products. To this end, I would spend time in all the production departments and, using specifications provided by the design engineers for the products, write the testing procedures to be used in each department and on the production lines. Because every radio, TV and stereo product was a unique design, each one had to have its own testing procedure and they would need to be modified each time there was a model or design change. While I was nominally under the supervision of Fred Stirk, I was pretty much a free agent and I was able to learn a great deal about manufacturing procedures. As well as very good design laboratories with very clever engineers, EMI had their own plating shop, transformer winding department and loudspeaker assembly (including magnetisation) department. Most punched steel chassis, PCBs and timber cabinets were outsourced, but everything else was made in-house. The labour force was predominantly female, and the production lines where the women assembled the chassis and soldered the circuitry ran like clockwork. All the supervision and testing staff were male. All the assembled TV chassis were powered and subjected to a full voltage heat soak test for several hours above the assembly lines on an elevated conveyor. Sometimes the TV sets had faults which resulted in spectacular bangs and the occasional fire. All assembled radios, TVs & stereograms had to be aligned and tested. To this end, suitable sweep alignment signals were distributed by 75W cables fed all around the factory. As well as spot frequencies for alignment of the antenna circuits on AM broadcast radios, there was a sweep frequency and marker test centred on 455kHz for IF (intermediate frequency) alignment. There was also a sweep and marker generator signal for alignment of TV IF strips and another sweep signal for alignment of ratio detector coils This photo of Leo Simpson was taken as he toured the A&R Electronics factory in Box Hill, Victoria, in 1977. He is being shown their new Arlec DMM 10, a 3-digit portable multimeter, with 7-segment red LEDs and powered by a rechargeable battery. He was a staff writer at the time (not Editor yet). The resulting article, titled “The A&R story”, started on page 20 of the March 1978 issue of EA. Australia's electronics magazine siliconchip.com.au in the 5.5MHz FM detector (for TV sound). All alignment tests were done using in-house oscilloscopes designed and manufactured by EMI with 5-inch CRT displays. That was really quite advanced for the time (the mid-1960s). Inevitably, some sets did not work properly as they came off the assembly lines. The men who fixed them became very adept at sussing out really weird faults caused by wrong value components or parts soldered to incorrect circuit points. Most products were entirely valve-based with point-to-point wiring, although there were some portable radios that used germanium transistors on PCBs. The car radios did use transistors, having just evolved to hybrid designs with transistors in the RF stages and valves in the audio output stages. HMV car radios were very good designs, with RF and audio performance far superior to any imported (mainly Japanese) designs of the time. Interestingly, there was also a large portable hybrid TV model which used the cathode voltage of the 6CM5 horizontal output valve (about 8V) to supply some of the small-­signal transistor stages. Working at EA It was mainly on the basis of my background at EMI and Ducon that I got the job at Electronics Australia magazine. I started in about March 1967 in a very junior capacity. My electronics knowledge at the time was quite sketchy, although I was very familiar with the circuitry of TV sets and radios. In most other respects, I regarded myself as a complete novice. My first project at EA was to assemble a transistor RIAA preamplifier to be installed in a valve amplifier. The circuit and PCB design came from the Technical Editor, Jim Rowe, who struck me at the time as a ‘god’ of design, having worked there for many years, producing myriad designs. After assembling it, I had to sketch out the circuit for the draftsman, Bob Flynn, and then write the article for the magazine, which would be edited by Neville Williams (another ‘god’). My next project was a rehash of an earlier valve-based stereo amplifier and was to become the Playmaster 118, with 6GW8 triode-­pentodes in the push-pull output stages. This project incorporated the previous transistor preamplifier, and it was then that I learnt about the difficulties of minimising hum in high-gain audio circuitry. siliconchip.com.au From there, I effectively had a project article published each month and I also reviewed a great many hifi stereo amplifiers, speakers, turntables, test equipment, records and books. By late 1971, I became dissatisfied with my progress at EA and realised that my chances of promotion were very limited. In May 1972, I got a job as a foreman at National Instruments Pty Ltd, at Kogarah. They made elevator control systems but their main product was jukeboxes, under license to an American manufacturer, Rowe-AMI. These were a very complex mechanical design with not much in the way of electronics, apart from the audio amplifiers. This change was a big culture shock for me. I missed the intellectual stimulus of the job and the people at EA. It was a big learning experience as I had to quickly become familiar with the mechanical complexities of the jukeboxes and, more importantly, learn about managing production staff, who were mostly women and all older than I was. I came to quite like the job, but I soon realised it was another dead end and started looking for another position. But in February 1973, I was ‘rescued’ by Neville Williams, who wanted me to come back as he had a staff vacancy. This was very opportune for me as I had become engaged to my future wife, Kerri, and we were looking to buy a house. It eventually happened with the purchase of our first home (at 74 Aubreen Street, Collaroy Plateau) in March 1973. We received the keys to our house on 16th March, the day before we were married. It would take another nine years before I was promoted to the position of Editor of Electronics Australia in March 1983. In that time, we had two daughters and had moved to a bigger house, also on Collaroy Plateau. In the meantime, I had enrolled in a Business Degree course at the New South Wales Institute of Technology and graduated in 1982. As I settled into the position of Editor, my long-time boss Neville Williams having retired in mid-1983, I had no inkling of what lay in the future, only five years ahead. Not in my wildest dreams could I have conceived of losing my treasured position as Managing Editor and then going out to start a brand new magazine with three members of my staff at the time: Greg Swain, John Clarke and Bob Flynn. Australia's electronics magazine what remained of the old EA team (ie, Leo Simpson, Greg Swain, John Clarke, Ross Tester and Jim Rowe) was together again, working on what really was “our” magazine. Right now, the only original people remaining from the EA days are John Clarke and Jim Rowe. All the rest who had connections with EA and Silicon Chip have moved on, retired or ventured up to that great hobby workshop in the sky. The final postscript involves the Electronics Australia archive. After the demise of EAT in 2001, we started getting requests from our readers wanting reprints of articles from EA and its earlier variants such as Radio, TV & Hobbies, Radio & Hobbies and before that, Wireless Weekly. We did not have the rights to do this, so I approached Federal Publishing and purchased the entire archive, with bound copies going all the way back, 100 years, to 1922. We still have regular requests for article reprints from this massive archive. We are proud to have been able to preserve it. Conclusion In writing this story, I have been very conscious that the long-term success of Silicon Chip has been due to the great teamwork of the staff over 35 years. Many people played their part, but I will single out four very special people. The first is Greg Swain, whom I have known and worked with very closely from 1973 until he retired in 2016. Second is the industrious John Clarke, who has worked with me since 1979 until I retired in 2018. He has produced a phenomenal body of work and countless ingenious designs. Third is Ross Tester, who came to work at EA in 1972 as a brash youngster whom I initially found quite annoying. He subsequently went on to work at Dick Smith Electronics and I have been friends with him now for many years. He was chaotic, creative and disorganised. He still is! To him, a tidy desk and office are anathema. He will turn his hand to anything and he helped to add life and humour to the magazine. And finally, there was Ann Morris, who provided the very special bond that held us all together from the time she started with us in 1990 to her retirement in 2020. I thank them all from the bottom of my heart. SC September 2022  75