Silicon ChipServicing in the Wild West of Central Europe - August 2023 SILICON CHIP
  1. Outer Front Cover
  2. Contents
  3. Publisher's Letter: High inflation and price changes
  4. Feature: High-Altitude Aerial Platforms by Dr David Maddison
  5. Subscriptions
  6. Project: The WebMite by Geoff Graham
  7. Project: Watering System Controller by Geoff Graham
  8. Feature: The Electrical Grid by Brandon Speedie
  9. Project: Arduino-based LC & ESR Meter by Steve Matthysen
  10. Feature: RadioFest 2023 by Kevin Poulter
  11. Project: Calibrated Measurement Mic by Phil Prosser
  12. Feature: An interview with DigiKey by Silicon Chip / Tony Ng
  13. Serviceman's Log: Servicing in the Wild West of Central Europe by Dave Thompson
  14. Vintage Radio: Replacing Vibrators, Pt3 by Dr Hugo Holden
  15. PartShop
  16. Market Centre
  17. Advertising Index
  18. Outer Back Cover

This is only a preview of the August 2023 issue of Silicon Chip.

You can view 39 of the 112 pages in the full issue, including the advertisments.

For full access, purchase the issue for $10.00 or subscribe for access to the latest issues.

Items relevant to "The WebMite":
  • WebMite firmware, user manual, fonts etc (Software, Free)
Articles in this series:
  • The WebMite (August 2023)
  • Watering System Controller (August 2023)
  • The WebMite (August 2023)
  • Watering System Controller (August 2023)
Items relevant to "Watering System Controller":
  • WebMite firmware, user manual, fonts etc (Software, Free)
  • Watering System Controller PCB [15110231] (AUD $12.50)
  • Watering System Controller firmware and web server files (Software, Free)
  • Watering System Controller PCB pattern (PDF download) [15110231] (Free)
Articles in this series:
  • The WebMite (August 2023)
  • Watering System Controller (August 2023)
  • The WebMite (August 2023)
  • Watering System Controller (August 2023)
Items relevant to "Arduino-based LC & ESR Meter":
  • ESR Meter add-on PCB for Digital LC Meter [04106183] (AUD $5.00)
  • Combined LC/ESR Meter PCB [04106182] (AUD $7.50)
  • Pair of PCB-mounting right-angle banana sockets (red/black) (Component, AUD $6.00)
  • 1nF ±1% polypropylene (MKP) or C0G/NP0 ceramic capacitor (Component, AUD $2.50)
  • 20x4 Alphanumeric serial (I²C) LCD module with blue backlight (Component, AUD $15.00)
  • Firmware for the Arduino-based LC and ESR Meter (Software, Free)
  • Arduino LC/ESR Meter PCB patterns (PDF download) [04106181/2] (Free)
  • Arduino ESR Meter table and baseplate template (Panel Artwork, Free)
Items relevant to "Calibrated Measurement Mic":
  • Calibrated Measurement Microphone PCB (SMD version) [01108231] (AUD $2.50)
  • Calibrated Measurement Microphone PCB (TH version) [01108232] (AUD $2.50)
  • Short-form kit for the Calibrated Microphone (SMD version) (Component, AUD $22.50)
  • Short-form kit for the Calibrated Microphone (TH version) (Component, AUD $25.00)
  • Simulation and calculation files for the Calibrated Measurement Microphone (Software, Free)
  • Calibrated Measurement Microphone PCB patterns (PDF download) [01108231/2] (Free)
  • Calibrated mic capsule set - Panasonic WM61A lot 4A14 (Component, AUD $12.50)
  • Calibrated mic capsule set - JLI61A lot 3 (Component, AUD $12.50)
  • Calibrated mic capsule set - JLI60A V02 (Component, AUD $12.50)
  • Calibrated mic capsule set - CMC6027 (Component, AUD $12.50)
  • Calibrated mic capsule set - CMC2742 (Component, AUD $12.50)
Articles in this series:
  • Calibrated Measurement Mic (August 2023)
  • Calibrated Measurement Mic (August 2023)
  • Reference MEMS Microphones (April 2024)
  • Reference MEMS Microphones (April 2024)
Articles in this series:
  • Servicing Vibrators, Pt1 (June 2023)
  • Servicing Vibrators, Pt1 (June 2023)
  • Replacing Vibrators, Pt2 (July 2023)
  • Replacing Vibrators, Pt2 (July 2023)
  • Replacing Vibrators, Pt3 (August 2023)
  • Replacing Vibrators, Pt3 (August 2023)

Purchase a printed copy of this issue for $11.50.

SERVICEMAN’S LOG The Wild West of Central Europe Dave Thompson Let’s face it, when the serviceman goes on holiday, all he can really expect is to do less servicing than usual – not none! This time was no different, but luckily, I had some seat-of-their-pants, wing-and-a-prayer helpers that did most of the heavy lifting for me. What’s the old saying? All good things must come to an end. Nothing is truer than the end of an extended holiday. Seven weeks is a long time to be out of the loop, and the older I get, coming home and returning to my old routine is becoming an increasingly strange experience. The only benefit of the long flights home is that they give pause to reflect on what was an amazing experience of seeing people, palaces, castles, cathedrals and art. I’m not into art by any means (unless we’re talking about a Fender Stratocaster), but I did see five original van Goghs, dozens of Klimts, a few Monets, a Rembrandt, a Sargent and a Whistler. It is hard even for a philistine like me not to be moved by seeing them. I have hundreds of photos to sort through, but of course, photos don’t show what I actually saw; at least, my photos don’t. I’m very much a ‘point-and-shoot’ photographer. I can also reminisce on a few occasions where my serviceman’s skills were needed (beyond those already covered last month!). In a household where the guy who used to do all the regular maintenance has long since passed on, things gradually decline to the point of needing some TLC. Many widows there rely on others with DIY skills, or they call in ‘a man from the village’ to do it as paid work. Some of these guys are capable, while some aren’t, which makes for interesting repairs. Evolving infrastructure As mentioned previously, the power to one of the apartments my mother-in-law and her late husband built years ago up around the coast (that we 82 Silicon Chip were soon to be staying in) had gone out, so I needed to look into that. I protested that I didn’t know much about the local system and regulations, but my protests fell on deaf ears. My brother-in-law said that nobody cared about that stuff anyway, and we could do pretty much what we wanted, as long as nobody got hurt! From some of the installations I saw later, I believed him! This particular town is essentially a centuries-old fishing village with some newer holiday homes tacked onto each end of it around the coastline. The infrastructure is creaky at best, with power, sewers and plumbing an afterthought and sometimes crudely implemented. Given the country’s typical cold-war era love of concrete, there’s a lot of it about. It isn’t always nicely-done concrete either, as we know it at least; much of it looks poorly mixed, cracked, crumbly and roughly applied. I even took some photos of ‘patches’ some local had done; while I’m no expert, it appeared to me to have been mixed with a garden fork and applied with a mop. The relevance here is that the concrete foundations for this place and the two garages were poured in the 1970s. Sewers, water pipes and power cables were simply pushed through plastic conduits embedded in the concrete. The water supply initially was a rainwater catchment system with the water collected and stored in a large concrete cistern built into the house. A pump ran automatically when a tap was turned on, pumping the water through a filter system and stopping when the tap was off. A separate mains-powered boiler mounted above a sink or bath (that looks similar to an old Zip, if anyone still remembers them) still provides hot water to most Croatian homes today. If the stored water ran out, residents would chip in to pay for a water truck to come over the hill from a bigger town to top up their cisterns. It stayed this way from the 1970s until around 2007, when a water pipe was installed from that town over the hill, at the residents’ expense. I remember helping clean and overhaul that water pump a few times on my early visits, and I have also helped replace several old boilers with much more modern, efficient ones. So, everything is buried in concrete these days, and in the newer areas, this makes for a neat and tidy system. However, given that the country and this area, in particular, is about as earthquake-prone as my hometown (we had a decent quake while there), things tend to sink or shift over the years, which can cause problems. Australia's electronics magazine siliconchip.com.au Items Covered This Month • • • Servicing in the Wild West of Central Europe Repairing a Lenovo laptop A sticky situation with Reveal 6D monitor speakers 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 Most residents regularly have to deal with problems like water leaks, broken sewer pipes and, of course, power supply problems. Power distribution The mains power to each property comes from now-­ buried cables that run under the access road leading to the beach houses. For those interested, Croatia has a standard mains voltage of 230V at 50Hz. They use the typical continental European two-pronged (round pin) plugs and mains sockets, types C and F. Without a meter, I couldn’t confirm those voltage figures, but I assumed it would be like many countries, where the voltage fluctuates slightly. There is a junction box at each driveway, similar to our water mains shutoff valve boxes, with a closeable metal cover that splits the feed off to another switchboard-type setup. This is usually in a garage, carport or other structure close to the road. The cable coming in might be run up the outside of the wall or buried in a conduit beneath the surface. If multiple properties are sharing the driveway (very common due to the housing density there now), the mains feeds to those homes are added in new conduits dug into existing concrete paths and structures. Sometimes, the property switchboard junction box is utilised for these feeds, and sometimes the feed is taken from the buried box at the roadside. How it is installed depends on who does the hook-up. The switchboards near the road are similar to those used everywhere and are either old and open to the elements (not great right next to the sea!) or more modern and enclosed in a covered, grounded (and usually rusted) metal box. Pole fuses and breakers are common; sometimes, there is a mixture of both. Each apartment also has its own smaller switchboard, where the outside mains feed terminates, and this is where I started my troubleshooting process. Oh, to have a multimeter! I physically checked the fuses and breakers – they all looked sound. The mains switch itself seemed to be working, or at least toggling correctly. There was just no power. My next stop was down to the breaker box where the mains entered the property. There was power here because I could hear it, and the pole fuses arced a little when I removed and replaced them. I also noted that there were two other feeds off to houses behind this one. They had their own pole fuses, and I didn’t touch them. Fortunately, someone years ago had labelled them, and the owner knew which ones were hers. siliconchip.com.au So, I had power to there, but not to the apartment in question. There are three apartments, each on separate floors of the three-storey building. This was the top-floor apartment, the newest one built, about 10 years old. The others had power, and their pole fuses were also on the main switchboard downstairs. Flying blind The second-floor apartment was empty, so I pulled both pole fuses and swapped them into each other’s sockets. The third floor was still without power. The second floor still had power, so the pole fuses were both working. That didn’t bode well because it meant there must be a break in the mains feed to the third floor, somewhere between this switchboard and the one upstairs. The obvious thing now was to remove the screws from the upstairs switchboard, fold it down and have a look behind it. Maybe a connection there had shifted? At this point, I was really missing having any tools with me. The family has a few random tools collected over the years, but nothing aside from a flat-blade screwdriver and a shonky Phillips one that was already chewed out and looked to have a shaft made of lead. I managed to remove the overly-long screws so I could lift the panel away from the wall. Everything looked to be connected securely, so that was another dead end. I could have done the same thing at the street switchboard, but with live power coming into it, and not wanting to shut down three households, I wasn’t about to mess with it. I replaced the upstairs panel and went to break my news to the owner. Between the two connections, there was about 20 meters of cable – somewhere along it, something had happened and had killed the power. Where was anyone’s guess. My serviceman’s brain was already spinning with the potential fixes for something like this, and none looked very appealing. I was thinking by this time that we’d have to get a proper sparky in to resolve it because, without proper tools, I’d just be winging it and likely doing it badly. The professionals must be used to dealing with this sort of thing all the time. Australia's electronics magazine August 2023  83 The only way I could see to repair this would be to cut into the concrete somewhere along the track, if we knew where it was, and rummage around until we found the cause. Hired guns My mother-in-law made the call, and two likely lads duly turned up. They were your typical tradies – a van packed to the roof lining with tools and whatever else and the ‘nothing is a problem’ attitude. They initially did essentially what I’d done, checking the obvious and conferring a lot before coming to a similar conclusion. Like many of these places, there were no real plans available of where anybody put what and because the apartments had been built over a span of almost 50 years, what drawings or plans there might have once been around at the time were long gone. We discussed it and determined that the cables must run in the shortest line between the two structures – the garage downstairs and the apartment upstairs. We knew where the cables ended and guessed the people who put them in would follow the natural path between those points. The guys had to go away and get even more tools. They returned the next day with some hired concrete-cutting gear and longer extension leads. They drew out some lines on the path about a third of the way from the bottom junction and set about cutting in. There was a decent crack running through the concrete around that point, and they (like me) assumed that might be where the problem was. It was as good as any place to start. Who needs health and safety? One thing I noticed in Croatia with workmen is that it’s like the Wild West. Workplaces don’t appear to be as highly regulated as they are in New Zealand, with health and safety briefings, licensed operators on specialised machinery, that sort of thing. On a job site there, you won’t find anyone wearing safety boots, high-vis vests, ear defenders or safety glasses. 84 Silicon Chip One renovation project going on a few hundred metres from the house I was staying in has a huge crane perched half on the road and footpath. It was being used to lift pallets of white rock cladding tiles to install on an old municipal building which, like many there, had been built in the 1960s and hadn’t had a thing done to it since. This crane had sunken into the footpath, probably because that hadn’t been made properly either, and yet it was there when we got there and still there when we left, a bit cock-eyed and still being used. Not one guy on that site had a hard hat or any safety gear, with other workers using big, gas-powered cutters to shape these tiles right by the thin wire fence they’d built to seal off the site. The noise was deafening (literally!) all day, and dust and other rubbish were being blown all over the footpath and road. Pedestrians and motorists were going about their business as if nothing was out of the ordinary. I crossed the street because I didn’t want to walk anywhere near that noise, the flying debris or under that wonky crane! These guys were the same. No regard for safety; they just lit up this concrete saw and got stuck in. They probably sneered at the foreigner standing way back with his fingers in his ears, but I didn’t care; I’m still able to hear (albeit not perfectly) at 61! They made short work of the concrete and made a 150mm/6in-wide cut over some guidelines they’d drawn. A few good hits with a hammer and chisel had the chunks out, and there was a partial view of a conduit underneath. Good guess! Dodgy work revealed They made another cut, widening the slot and when the concrete was clear, they cut carefully into the conduit. The cables were all in there – phone and power. One of the guys hooked the power cable and gently pulled it. It kept coming, from the road end, until it cleared the conduit. There it was, a break. On closer inspection, we could see that the cable had been joined, and the joint had failed. It was poorly done, and by the disgust on these guys’ faces, they thought so too. Here it is against the code to join a mains cable; from what I understand, it must be a single piece between the house and the street junction. Whoever put this in originally had not done it correctly, so we now had to restore it. After much discussion, including joining the cable again, they decided – at our insistence – that the whole thing should be replaced. That meant disconnecting the cable at the apartment end, behind the switchboard, and after tying some strong twine to it, carefully pulling it back through the conduit to the hole they had made. This they did without too much fuss. They then retrieved what we call a ‘fish stick’ from their van. This extendable, flexible fibreglass rod can be fed down walls and conduits to catch wires and pull them through. Sections are added as necessary. The cut was only about six meters from the bottom switchboard and with just one bend, so it wasn’t too much work to get the rod through to the hole. The new cable was measured out and pulled back with the twine to the apartment end. Once terminated there, the remaining length was attached to the fish stick, and that end was pulled to the board at the road. It was terminated there carefully because the board was still live! Australia's electronics magazine siliconchip.com.au Brave men. The pole fuse was put in, and the apartment finally had power. All that remained was to tidy up the switchboards, re-­ attach the cut section of conduit with lots of silicone and fill in the hole with concrete. These guys did it all. It was a fun job to be part of and to see how others do it. Holidays! Repairing an obsolete Lenovo tablet B. D., of Mount Hunter, NSW encountered the pitfalls of repairing a device when you aren’t familiar with how it comes apart and goes back together again. Still, he overcame that, albeit at a higher cost than anticipated... I purchased a Lenovo MIIX 510-12ISK Windows tablet about six years ago to use as my main computer. It is quite a good unit, similar to a Microsoft Surface but cheaper. It has an i5 processor, 8GiB of memory and a 250GB SSD. I used it for a couple of years with no problem when I noticed the power switch no longer had a positive click when operated. I tended to use it often to shut the computer down rather than using the software controls. Not long after, the computer started to shut down and reboot spontaneously. This went on for a while until it got so bad that it was in a continuous reboot cycle, and the tablet was no longer usable. I tried a few things, but realised after a while that the power switch must be faulty. I thought I might try to fix it myself and managed to remove the screen and look inside. I found the power switch was on a small circuit board together with the volume control that would not be easy to replace. I rang Lenovo, who had a repair service (no longer operating). They quoted $200 and I would have to post it etc. I decided it was all too hard, and as the battery had also lost capacity, I purchased a HP Spectre as a replacement. It has virtually identical specifications and performance. I put the Lenovo tablet away in the cupboard and left it there. Later, when I had spare time, I decided to have another look at it. At the time, I didn’t realise that when removing the screen, it is better to lift the top and not the bottom to avoid disturbing the connecting cables attached to the screen. I managed to dislodge both cables (for the display and touchscreen) from the back of the screen. So I decided it was still too hard to fix. At this stage, I hadn’t worked out how to determine the correct way to reattach the cables. So naturally, I managed to reconnect the main display cable the wrong way around when putting the screen back and was greeted with a burning smell and a small column of smoke rising from the system board. The tablet was then dead. I returned it to the cupboard as it still had usable memory and the SSD, so I was reluctant to throw it away. About a year ago, I was browsing on eBay and noticed second-hand system boards for my exact model were available for a reasonable price. I also found a very good teardown manual on the Lenovo website for this model. It gave clear step-by-step instructions on removing each part, such as the battery, hard drive etc. However, it doesn’t give small details, such as the correct way to remove the screen or that the small gates on most connectors must be lifted before removing or connecting the cables. It also doesn’t mention that a white line on the end of the cable must be visible when reconnecting it to ensure it is the correct way around. siliconchip.com.au Australia's electronics magazine August 2023  85 The two cables that caused the problem are circled in red. Hard-to-see horizontal white lines on the cables under the connectors show the correct orientation. The power switch and volume control are in the top right-hand corner. The touchscreen control board with the broken connector (CN1) in the bottom right-hand corner. I was still unsure whether the screen had been damaged, and with no way to test it, it would be a gamble if I purchased one of these boards, as a screen replacement would make the whole exercise non-viable. I explained this to the vendor and told him that these tablets were now obsolete, and he accepted my offer of $60 plus postage for the system board. He also threw in a used power switch and other useful bits and pieces. When the parts arrived, I commenced the repair slowly and carefully. Everything came apart pretty well as described in the manual. Just about everything has to come out to replace the system board, including the SSD, wireless card, cameras, battery and switch. I was also helped by a YouTube video on a heatsink paste renewal for this tablet, which showed most of the steps. I installed the replacement system board and also replaced the troublesome power switch with the one supplied. It seemed OK, with a positive click. I put everything back in reverse order and then reattached the screen. The battery had to be charged after so long in the cupboard. While reattaching the screen, as I hadn’t worked out the best way to do that yet, the touchscreen cable was stretched beyond its limit, and the plastic connector on the back of the screen broke away beyond repair. It was attached to a small circuit board on the back of the screen, called the 86 Silicon Chip touchscreen controller. I carried on and managed to put the screen back. With some trepidation, I pressed the power switch, which needed to be held down for some time. To my great relief, the Lenovo logo appeared on the screen after about ten seconds. However, I had no keyboard and no touchscreen, just a mouse. This made logging in (which required a password) rather tricky. I removed the screen again and checked the keyboard connector to the system board. This was when I realised that the small locking gates must be pressed down. After doing that, I reassembled the screen and had a working keyboard. I was able to log in and connect to WiFi. Everything seemed to be working OK, except for the touchscreen. The part number was clearly written on the part, so I typed it into Google, as Lenovo no longer carried this control board. Google returned an AliExpress page with the control board listed, albeit quite expensive at $60, but having no alternative, I ordered the part. I also ordered a replacement battery to complete the package. The touchscreen controller eventually arrived, and by then, I was well-practised at removing the screen. The board is attached with double-sided tape and reconnecting the multi-pin connector to the panel was not as difficult as I had anticipated. I put it back together and switched on the tablet. To my dismay, there was still no touchscreen. I went into Device Manager, where it was listed as not working. Fortunately, when I clicked on one of the options there, it suddenly came to life, and everything then worked. The tablet was now pretty good; however, Windows Update insisted on installing the latest version of Windows. After that, I noticed that the tablet took about 15 seconds to wake from sleep, whereas with the old version of Windows, it woke up straight away. I went to the Lenovo Users’ Board and found that this tablet is incompatible with the latest version of Windows, so I reverted to the old version, and the problem went away. However, as the slow wake-up seemed to be the only problem, I decided to upgrade again and live with the slow wake-up. I now use the tablet quite a bit as a spare or when the HP is not available. It has a removable keyboard and is lighter than the HP. I am glad I persevered with the repair as it was satisfying to achieve a result, and electronics is my hobby, after all, so it was time well spent. All up, the parts cost around $200. Speaking about a sticky situation N. B., of Seven Hills, NSW discovered how much more difficult a repair becomes when the circuit boards are covered in glue. It also didn’t help that multiple assemblies were packed into a small space... I empathise with Dave Thompson’s wrestle with a perfectly functional carpet cleaner in his March 2023 column. I’ve also once or twice found myself disassembling and testing something under the assumption that something was broken, when it was just simple operator error. When I bother to read them, I often think that the ‘troubleshooting’ sections of most appliance manuals usually only cover the ‘bleeding obvious’ operator errors and generally give little if any advice on what to do when there is a real problem with the device, other than “contact the service agent”. But I suppose operator errors are much more common than real failures. Australia's electronics magazine siliconchip.com.au My write-up of the following repair saga has already helped a colleague fix another pair of similar speakers. The Tannoy Reveal 6D monitors are compact two-way active, bi-amped near-field speakers. The 6D is one of the second-generation Active Reveals, built from about 20052010, with a 6-inch (150mm) low/mid driver and 1-inch (25mm) dome tweeter, each driven by its own amp after the active crossover and filter circuitry. They have a digital input (hence the “D”) and cost about $1200 each when new. This pair started making odd noises and didn’t sound right. I asked for a fault description, and it was reported that they made a great pair; one had no HF driver output, and the other had no LF driver output! (The failed LF driver sometimes worked intermittently). I decided to make them a personal project and took them home. I had no idea what I was in for, but I like a challenge. The tweeter producing no sound had a resistance under 1W and connecting a battery resulted in an extremely quiet click. I then noticed a smoke stain on the front panel above its mounting hole. I also removed the low-frequency driver and powered it up with DMMs on the speaker leads. About half a volt of DC appeared on both speaker outputs, increasing to 5-6V. Not good. I would have to pull the amplifier module from the back. Everything is mounted on a 3mm aluminium backplate, with a large finned heatsink occupying about half its outside area. Inside, things were quite packed. The power input and transformer were at the bottom. Above them was the main board with the power supply and power amps, then the filter board, with the analog input, crossover and EQ circuits. It has a level control pot and a line-up of switches, including 20 DIP switches for setting the EQ characteristics. Finally, the digital input board is at the top, with a few surface-mounted devices on it. Almost everything was coated in glue to stop things from vibrating and rattling. Fasteners were covered in hard, clear glue, while connectors received tough black glue to ensure they didn’t separate. Components of any size had brown glue added to brace them. A bead of it was also used between each board and the backplate. There was evidence that an electrolytic capacitor had vented all over the main boards and filter boards in both speakers. There were 11 small electros on the main board, plus two large filter capacitors for the modular power amps. Half of those caps were related to the ±15V supply for the filter and digital boards, with others for the mute circuitry. All electros measured at least 10% low, including the power amp supply caps. But a few were very low in value, and not surprisingly, the vented cap was open circuit. Getting to the main board to replace the capacitors was almost impossible, with it sandwiched between the transformer and the other boards above. I found a forum thread on repairing the Reveal 6Ds, which helpfully included schematics of the main and filter boards. There were several comments about access difficulty and problems from all the glue, including that the brown glue can be corrosive and conductive, so you should try to remove it as much as possible. Great. Removing the digital and filter boards was relatively straightforward, other than the time to chisel, cut and scrape off glue as I went. After doing that, a couple of the caps were still in awkward places, under an aluminium heat conductor from the 7815 regulator to the backplate and heatsink. siliconchip.com.au Access to the bottom of the board was also quite limited, and two end-mounted power resistors below the board were just a few millimetres from the transformer. The whole assembly would have to be dismantled to obtain good access to work on the board, or I would at least have to remove the power transformer. The amplifier modules were clamped to the backplate by a steel bracket that also supported the board, with the heads of the screws fixing it under the heatsink. Most of the screws mounting the heatsink were accessible with their heads on the inside, although glue had to be removed from each. But one heatsink screw was under the power transformer. The transformer was mounted on screws coming through the backplate, with nuts on the inside, but the heads of two transformer screws were under the heatsink. Australia's electronics magazine The tightlypacked Reveal 6D amplifier assembly. The dreaded brown glue, with the exploded capacitor visible in the centre. The Reveal 6Ds with replacement tweeters installed. August 2023  87 The transformer sat on an adhesive pad, which did the job so well that the screws seemed almost redundant. I could loosen the nuts on the transformer screws, but the remaining glue in the threads presented so much resistance that the screws started turning once there was some slack. Clearly, the transformer was going nowhere, so I gave up on that idea; the board would have to be fixed in situ. I removed the power resistors to improve access to the bottom of the board, then proceeded to replace all the small electros, chiselling off all the brown glue I could. Clearing the plated through-holes is not easy when you can’t get a solder sucker onto them, but solder wick saved the day. After remounting the power resistors, I connected meters and prepared for a smoke test of the main board. When I powered it up, the +15V output was only about 8V, and one cap started to smoke and bulge. I quickly turned it off and checked it. Oops, I’d installed it back-to-front. The overlay showed the polarity for most of them, but a few had to be carefully checked, and I’d gotten that one wrong! After replacing it, the main board passed the smoke test, the supply voltages were all good, and with no input, the speaker output voltages were only a couple of mV AC and DC. So far, so good. I mounted and connected the filter board and performed another test. Now the speaker outputs were showing about 30V DC. Not so good! The ±15V supply was going to the filter board, but the output lines were obviously DC-biased. I hoped it wasn’t a failed op amp, as there were 22 of them on the board. I decided to put that one aside and work on the second speaker, figuring that if that one could be more easily brought to a functional condition, it would be a good reference to diagnose the problem with the first one. Now that I’d learned what not to attempt, work on the second module proceeded a bit faster. However, the power resistors didn’t want to move as easily as those on the first board, and there was more glue under their bases. When they finally moved, each took a solder pad and a section of track with it. The glue stuck the solder pad to the resistor better than the pad was fixed to the board! I also accidentally broke a zener diode mounted across the inputs of the 15V regulator ICs when I tried to move it to give better access to clean the capacitor mounting pad it was soldered to. So, extra repairs were required when reinstalling the components. I laid down the lead of the 88 Silicon Chip capacitor next to each power resistor to replace the missing tracks. It passed the initial smoke and voltage test. I cleaned the filter board a bit more and reinstalled it. Again no smoke, and the voltages were OK, so I fed a pink noise source to the analog input, and the speaker output voltages did the right thing. I connected it to the speakers in their box, and both drivers produced suitable sounds, so that was one win. Back to the first speaker. On the filter board, I fitted component lead off-cuts to the bottom of the connector pins for the cable from the main board. That enabled me to selectively plug the cable into the leads on the power pins and leave the signal leads bent out to monitor what was going on. The power amps would then not be stressed by a DC signal input while I diagnosed the problem. On power-up, the supply voltages were good, but the signal outputs of the filter board were around 13V DC. Further probing found that the +15V supply pins on each op amp were also around -13V. I powered it down and started doing resistance checks on the +15V rail. Connections between all the op amp VCC pins were all good, but there was about 60kW between any of them and the +15V input pin and filter cap on the board. The filter board’s supply caps were a pair of 100µF electros right next to the cable connector, which measured better than the others. Like on the main board, there was a liberal dose of brown glue on them. I removed some of the glue, but it was between and under the caps and cable socket. I removed the filter caps and chiselled off as much of the glue as I could. I could then see that the +15V track ran on the top of the board, back under the edge of the cable socket, came out the opposite end, and onto the first op amp. I’d found earlier that the cable sockets are not too well fixed to their pins, as one of them came off the pins because the connector bodies were stuck together by glue I could not access. So I got a small sharp screwdriver under it and levered it off the pins. Once that was out of the way, I chiselled off the remaining glue to clean up the area. With the area all clean, I saw a 2mm-wide charred spot on the +15V track where it ran under the socket. It seems that the glue had eaten away at the track until it heated and turned a small area of the board into a carbon resistor. I used the tip of a small drill to excavate the carbonised material, replaced the socket and fitted new caps. I added a bit of old telephone (solid-core) wire on the bottom of the board to bypass the broken track. On reinstalling the filter board, its supply and the speaker output voltages were all good. It now passed sound well, and connecting speakers showed it worked like the second module. I reinstalled the digital input boards but did not connect them, as we have not found them useful. The audio outputs from the sound cards in our studio computers have better sound quality than the digital-to-analog converters in the 6Ds. Tannoy seemed to agree, as the digital input was omitted in later Reveals. The original tweeters are no longer available, but similar drivers from Jaycar with a differently-shaped mounting plate could be made to work. Their differing sensitivity was not a problem as a trimpot sets the tweeter level. So with some front panel work, I replaced the tweeters, and they are working again as a pair of utility speakers. SC Australia's electronics magazine siliconchip.com.au