Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN’S LOG Closed for Christmas! Dave’s on an early holiday, so this month we’re instead featuring some of our contributor’s items, starting with a weather station that was fixed up by Bryce Templeton. Despite being a solar-powered unit, it still needed non-rechargeable lithium cells. I bought a new weather station about three years ago to replace an old one that had fallen to bits. I purchased it online as, at the time, store-bought units did not offer what I thought was a handy option: solar cells to power the outdoor section. That would save me having to pull the unit down to change batteries. I should have done more homework, as when it arrived, I found that it still required batteries and I was warned in the instructions not to use rechargeable cells. In fact, they recommended using non-rechargeable lithium cells. I didn’t have any lithium cells on hand and, as I was keen to get it going, I used ordinary alkaline AA cells, which worked fine. The theory of the solar cells was that the unit will run on solar if it is available; otherwise, it is powered by the batteries. This results in the batteries needing to be changed about every six months. This was the situation for more than a year, when I noticed that a section of the indoor display was blank. This was the section that shows the intensity of the sunlight in W/m2, and the UV index. So the next time I had to change batteries, I decided to investigate this problem. An examination of the device showed that while it was well made mechanically, electronically, it was a different story. Getting it to connect to the home WiFi had been an arduous task, and it never managed to send anything to Weathercloud. Items Covered This Month • Fiddling with a finicky solar weather station • Bruce Pierson’s troubles: lights, fans and angle grinders • Repairing a foldback monitor speaker • The bargain bin 65-inch TV Dave Thompson runs PC Anytime in Christchurch, NZ. Website: www.pcanytime.co.nz Email: dave<at>pcanytime.co.nz Cartoonist – Louis Decrevel Website: loueee.com 86 Silicon Chip Anyway, I took the covers off and discovered that the sunlight sensor is a tiny disc-shaped PCB containing an unmarked IC encased in clear material. It lives in a tower with a small window on the top of the unit. Examination didn’t show anything unusual, so I did a quick re-solder of the ribbon cable that connects to the main board, which of course did nothing. Alongside the sunlight sensor is a bubble level, apparently so that the unit can be mounted truly level, which is important for tipping-bucket rain gauges. Unfortunately, in most installations it can only be seen from a helicopter or drone. I decided to try to get a replacement sunlight sensor and sent off an email to the firm I had purchased it from. After several very confusing emails, in which they never said if they actually sold parts or if they had this part, they came up with the best ‘catch-22’ I have heard: even if they did have the part, I would not be able to buy it from them unless I had previously bought one. At this point, I gave up on the sunlight sensor. However, on this battery change, I decided to try the lithium cells they recommended. It seems that alkaline cells have problems if the temperature drops below zero, whereas lithium cells will operate to -20°C. There is not much chance of even 0°C here, but I thought they might last longer, so at great expense, I purchased a pack of four. I installed three of these, and all was well. I was very surprised about a week later to find the outdoor unit not transmitting. Down it came again, and I found the 4.5V battery reduced to about 1.5V. Measuring each cell revealed that one cell was reading about 0.5V in reverse! I decided that I must have gotten a dud cell and replaced it with the remaining new cell, and we were away again. Australia's electronics magazine siliconchip.com.au The weather station, both mounted and lying around (shown left), and the output data provided on the internal display. But not for long; a week later, again no transmission. Again, down came the unit, and again, I found one cell with reversed voltage. Closer investigation this time revealed that with no batteries inserted, there was about 2V at the battery terminals. The penny dropped; I took the unit out into the sun, and the voltage at the empty battery terminals shot up to about 7V! After tracing the circuit as best I could, I came to the conclusion that the solar was an add-on, as there was no circuitry on the main PCB to do with it. It seems that the solar cells are just paralleled with the battery, using the battery as a voltage regulator to prevent the voltage from going too high in strong sunlight. Apparently, alkaline cells tolerate this treatment, but lithium cells strongly rebel. This would indicate that although the instruction book recommends lithium, it had never been tried in practice, or more likely, the book was written before the solar addition came about. The solution? Well, it’s currently running with no problems on alkaline cells. A proper solution would be to use some sort of charge regulator and fit rechargeable cells. Editor’s note: maybe this is a case of confusion between lithium and lithium-ion cells, as noted in my July 2025 editorial? Still, charging lithium-ion cells without current or voltage limiting is a bad idea! Bryce Templeton, Mudgeeraba, Qld. Bruce Pierson’s troubles around the globe My wife asked me to replace a light globe as it was not working. I checked our box of spare globes, but I did not have that wattage, only the next wattage up. I decided to use a globe from the lounge room to replace the failed globe and put the higher wattage globe in the lounge room. But when I put the new globe in the lounge room light, it did not work. It was brand new, so I was not impressed. On inspection, I could immediately see why it didn’t work. One of the contacts on the base was completely missing. I wondered if I could repair it by using a contact siliconchip.com.au from the failed globe. I managed to prise out the contact from the failed globe with the point of a knife, so it looked possible. I tried to push the contact into the hole in the base of the new globe, but it kept popping out. I could see that there was a springy wire in the hole that was causing this. It seemed that the wire should be on the side of the hole and not in the middle, so it was a manufacturing fault. This also explained why the original contact had come out. I managed to bend the wire into the correct position using a small flat-bladed screwdriver, and that enabled me to push the contact in and it remained in place very firmly. I put the globe into the light and it worked. Of course, I could have taken it back for a replacement, but I presume it would have ended up in landfill despite such a simple fault. That would have taken up more of my time, too. This was an unusual situation that I have not encountered previously. It must have been missed in the quality check, or maybe the contact fell out during transporting. It’s always pleasing to rescue these devices. A wall-mounted fan repair Our son had been using a Heller wall-mounted fan. My wife asked me if I could clean it and put it away, as our son no longer needed it. I asked her if it still worked, and she told me he’d said that it worked the last time he used it. I started by unclipping the front guard and removing it. Then I unscrewed the blade retaining nut, which is a left-hand thread. That enabled me to remove the blade for cleaning. Then I thought I had better test the fan before cleaning to ensure it did work. There was no use in cleaning it if it didn’t work. This is an electronically controlled fan, and sometimes such types decide to stop working for some reason. In this case, when I pressed the start button, nothing happened. I felt the blade spindle, and I could tell that the motor was trying to turn it, but it was not succeeding. This indicated that the bushes had run dry and seized, which is a Australia's electronics magazine February 2026  87 cosmetic condition, but they are a lot more reliable and longer-lasting than newly purchased fans, which often fail when they are just out of warranty. A bit of work, some lubrication and a good clean, restored this fan to good working order again. The photo opposite shows the fan after repair and cleaning. Following with a Bosch 9-inch angle grinder common thing to happen with fans, so I switched it off and unplugged it. I then refitted the blade so I could turn the fan, finding that it was very difficult to turn. I removed the blade again, then unscrewed the nut holding on the back guard and took the fan out to my workshop. With a #2 Phillips screwdriver, I removed the four screws holding on the front plate that the back guard is attached to, then I removed the rubber plug on the back of the motor guard and unscrewed the single screw and removed the motor guard. I refitted the blade so the fan could be turned over by hand. Next, I put a few drops of engine oil on the front bush and turned the fan multiple times. It was still difficult to turn, so I added a few drops of engine oil on the back bush and continued turning the fan by hand. Repeating this process several times eventually freed up the bushes, and the rotor spun freely, so I removed the fan blade again. After wiping up the excess oil, I plugged the fan in to test it, and it worked nicely, so I refitted the back motor guard and the front plate after cleaning them. It is better to use engine oil when servicing fans, rather than machine oil, which is too light for this purpose. I have used this process many times on various fans (including exhaust fans) with good results. With the fan working again, I finished the cleaning job. I used a brush to clean the front and back guards, then a damp cloth to wipe the blade clean. I dried it with a dry cloth. This blade cleaned up easily, as the fan was relatively new. I reassembled the fan and gave it a good test run on all three speeds. As it was now working correctly, I put a cover on it and put it away for future use. This particular fan has a bracket that is screwed to the wall, and the fan sits on that bracket, so it isn’t really portable. When cleaning fan blades, I’ve found in some cases that it is sometimes necessary to use a brush and soapy water if the dirt is really stuck to the blade; still, they mostly come clean with a damp cloth. The hardest fans to clean are used fans we pick up at the Tip Shop. These old fans are usually not in very good 88 Silicon Chip I can’t remember where I got this 9in Bosch angle grinder, but I’ve never used it. I’m used to 100mm and 125mm angle grinders; this one is much scarier at 230mm. Still, I needed to cut some concrete, and the smaller grinders just would not cut deeply enough. I managed to find a 230mm diamond blade on eBay for $27.20 (they are normally over $100!). The grinder did not come with a tool for replacing the blade. I found the correct tool on eBay, but it was $30, so I decided to make one. I got a section of power pole bracing that I’d picked up at the tip shop and cut it to length. I then drilled three holes in it, and I welded a pin in the two smaller holes. It was good enough to remove the grinding disc that came on the tool and fit the diamond disc. I also fitted the side handle to the grinder for added safety. The grinder would not start if the disc was in the vertical position, but it would start with it horizontal and kept running when turned vertical. After cutting the concrete, I checked the cable and brushes; they were all good. Later, when I needed to do some more concrete cutting, the grinder no longer worked. I removed the cover and checked the switch with my multimeter. The switch was open circuit with the trigger held in, so I would have to replace it. I suspect that the switch had just worn out. I found a switch listed for this model, but it was over $30. It was a little different from the original but looked like it should fit. I then changed my search criteria and found the same switch listed for a slightly different model grinder for $17.50. Having eventually received it, I compared it with the old switch. There were some differences, but the new switch looked like it should fit in the case the same way as the old Australia's electronics magazine siliconchip.com.au The repaired fan (left) and Bosch angle grinder (right). one. The main differences were the shape of the trigger and the X2 capacitor being in a different place. It fit nicely into the handle section of the grinder, which comes off the main body after removing four screws. However, when I tried to plug the internal plug on the main part of the grinder into the switch, it would not go over the pins. I then realised that the pins on the end of the switch were closer together on the new switch than on the old switch. I’m not sure if this was because the new switch was for a later version of the grinder, or because I had ordered a switch for a slightly different model. In any case, after making some modifications to the plug and the switch, they went together. Thankfully, the only real difference was the size of the blank section in the middle of the plug that spaced the terminals apart, so it was easy to modify. I cut out the middle spacing section with a utility knife, which left me with two separate insulated plugs. On my first attempt, I found that I could not get the plugs onto the pins, as the switch had a ridge in the middle of where the plugs plug in, which the original switch did not have. I used a utility knife to remove this ridge, then it all went together. It was quite a nightmare getting the grinder back together because of the way the two handle halves went over the switch and over the main body of the grinder. It took a lot of trial and error to get everything lined up, and the two handle sections correctly positioned on the main body of the grinder. Then it was just a matter of installing the four screws and the repair was complete. A quick search revealed that it would cost $450-500 to buy a new grinder like this. Because I got it for free and only spent $17.20 for a new switch and a bit of time, I ended up with a good quality Bosch angle grinder for a fraction of the cost of a new one. I will repair anything that I can get parts for! Bruce Pierson, Dundathu, Qld. Repairing a “VoiceSolo” foldback monitor speaker The TC Helicon VoiceSolo foldback monitor is a self-­ powered speaker designed to be used with a microphone stand, with the mic boom attached to the top of the monitor. I was recently presented with one of a set of four that was described as “dead” by the users. siliconchip.com.au Australia's electronics magazine February 2026  89 The VoiceSolo preamplifier (left) and power supply (right), with the failed electrolytic capacitors circled in red. The monitor case is of diecast aluminium, with the front assembly containing the speaker and input controls, secured to the main enclosure by four screws. Applying power to the monitor and connecting an input signal confirmed no power indicator LED and no sound. Opening up the monitor revealed a loudspeaker and four circuit boards, one attached to the front assembly and three within the rear enclosure. The power amplifier is a ‘BASH’ amplifier design where a Class-AB bridge amplifier module is supplied with a main DC supply that is modulated by a secondary switch-mode variable voltage supply tracking the amplifier audio input. This particular design has a 200W power supply and amplifier module built on two boards by Indigo Canada. A check with a multimeter at the main switch-mode supply PCB confirmed a steady +60V DC main supply, but none of the four low voltage rails, ±24V and ±15V, were present. Visual inspection revealed a ¼W resistor burned to a crisp and a 100μF 25V electrolytic ruptured. This power supply derives its low-voltage rails from an additional secondary winding on the main switching transformer. A group of diodes and electrolytic capacitors create unregulated positive and negative DC rails, which are then fed through a 7824 linear regulator and a discrete transistor regulator circuit to deliver ±24V rails for the BASH amplifier control circuits. A pair of 7815/7915 linear regulators supply the mixer/ preamplifier circuits. The ¼W resistor is connected between the bottom of the secondary winding and GND, acting as a fuse. A circuit diagram could not be found with extensive internet searches, so I needed a working monitor for reference. This duly arrived after a week, revealing the resistor to be 100W. I replaced the two failed parts and reassembled the monitor, unsure whether the fix would work. Servicing Stories Wanted Do you have any good servicing stories that you would like to share in The Serviceman column in SILICON CHIP? If so, why not send those stories in to us? It doesn’t matter what the story is about as long as it’s in some way related to the electronics or electrical industries, to computers or even to cars and similar. We pay for all contributions published but please note that your material must be original. Send your contribution by email to: editor<at>siliconchip.com.au Please be sure to include your full name and address details. 90 Silicon Chip I couldn’t spot any other components visually damaged; I needed to test the low voltage rails before doing anything more. Reconnecting the loom and powering on the speaker, I was greeted with a green power light on the front panel, so I grabbed the multimeter to check the rails. But before I could take any readings, the resistor began emitting magic smoke and failed again. So, we have a burned-out resistor and a ruptured capacitor on the output of the 7915 regulator. What was causing this? My best guess was a failed 7915 regulator, as it was feeding the failed electrolytic. Since it is secured to the heatsink with a shared mount to the other three regulators and a TIP30C power transistor that delivered the -24V rail, I decided to test the lot. I removed the four from the circuit board and tested the regulators on a breadboard. All were good, including the 7915. So I reassembled the power supply board and replaced the resistor again. This time, I was wiser and powered up the switch-mode supply board on its own on the bench. This time, the resistor didn’t fail, and all four low-voltage rails were within spec. So the root cause was elsewhere. The preamplifier, mixer and tone control board were mounted to the front panel in an assembly comprising the combined vent/carry handle. Dismantling the assembly revealed a board with eight NE5532 op amps, 13 electrolytic capacitors and a mix of SMD components. I was about to begin meter checking each of the NE5532 op amps when I noticed a very slightly bulging electrolytic hiding among a cluster of four identical ones. A closer look revealed that this was another 100μF 25V capacitor. In fact, all the electrolytics on the board were 100μF 25V, of the same make and type. I removed the bulging one and tested it on the component tester. Its value had risen to over 150μF, and it had a high ESR reading as well. The VoiceSolo speaker and I/O box. Australia's electronics magazine siliconchip.com.au Given that the four DC regulator circuits were working correctly, I began to suspect the electrolytic capacitors themselves. In for a penny, in for a pound, and out with the vacuum desoldering gun. I removed and tested all 13 of the capacitors. Only five tested good! Some of the capacitors were functioning as supply bypasses, while others were for audio coupling. I decided to replace all the 100μF capacitors in the monitor with good-quality low-­ leakage types. For the ‘acid test’, I reconnected the wiring harness and gingerly reached for the power switch. On powerup, it was a bit anti-climatic. No magic smoke, all supply rails within spec, green power LED on, good to go! I reassembled the monitor and gave it a thorough bench test, playing Sting at a modestly loud level. With so many failed capacitors, I began to wonder if this might be a manufacturing problem. Would it happen again soon with the other three monitors? I checked the spare that was used to identify the 100W resistor. In this one, the resistor hadn’t failed, but sure enough, there were another eight faulty electrolytics. In the end, I dismantled and replaced the capacitors in all four of these monitors. Testing revealed the majority in each to be on their way out, with strange capacitance and high ESR readings. In all, I replaced 60 capacitors. Ray Ellison, Dover Gardens, SA. The $19.00 65-inch television set Dave Thompson’s article in the August 2025 issue about repairing discarded devices struck a chord with me (siliconchip.au/Article/18644). Over the years, while walking around the streets of McCrae, I’ve rescued many perfectly good items discarded by their owners and left on the nature strip for council collection. The list includes a Jensen X-125 subwoofer, a 150mm reflector astronomical telescope, sundry computers and laptops, to name just a few. However, my most rewarding nature strip pickup was a large-screen Sony TV. It was buried under a pile of old plastic chairs, a mattress and other paraphernalia, and barely recognisable as a TV. I ventured onto the premises and asked the owner if it was a TV, and why he was throwing it out. He said it had simply stopped working, and as an expert IT consultant, he had concluded that “it was a transformer failure” and not worth repairing. I was welcome to take it since it would reduce the volume of his discards, and hence his fee to the council. siliconchip.com.au The set was a 65-inch Sony Bravia KD-65X7000E of about 2018 vintage. It was awkward to load into the car – the thing was simply enormous and quite heavy – but we made it home and successfully unloaded the device into the shack for further inspection. Fearful of flexing the set too much, I cautiously removed its rear cover to be confronted by three PCBs: an RF board, a motherboard and a power supply assembly. They were absolutely dwarfed by the screen itself, and I wondered how such a small set of electronics could drive such a monster screen. A quick check revealed that the power supply was not working, further confirmed by two blackened diodes, which had obviously ‘released their smoke’. Without really checking these diodes, I reckoned a couple of 400V 3A general-­ purpose silicon devices from my parts bin would make suitable replacements, so I quickly substituted a pair of new diodes. Switching on the TV produced screen images for about 20 seconds until my replacement diodes got very hot, also lost their smoke, then the set died again. Sony, in their quest to minimise power consumption, had specified high-speed schottky diodes for their power supplies in the KD-65X7000Es, which meant my substitutes were not suitable. So I bought a strip of 10 schottky diodes from Amazon for about $20 and replaced the two faulty ones, as well as the other two that made up the bridge rectifying circuit. This time, the set sprang into life and continued to operate satisfactorily. The new diodes were barely warm. The 4K picture was crystal clear, with no screen defects, and great audio – so I had acquired a marvellous 65-inch TV for about $4, plus another $15 for a new remote control! Unfortunately, the owner had discarded the mounting hardware for the set, so I had a bit more work to do to make a frame to support the set from my old audio cabinet. The attached photo shows the setup, with the Jensen X-125 at lower left. Happy days. SC Rob Fincher, McCrae, Vic. Australia's electronics magazine February 2026  91