Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN’S LOG Designing for unrepairability Dave Thompson A lot of equipment these days is designed to be compact and easy to assemble, but little attention is given to repairability. That’s why so many devices are essentially disposable. There’s no information like circuit diagrams or repair manuals available, so once something goes wrong, in the bin it goes. It’s sad. I ’ve mentioned before that I get a lot of random devices into the workshop for repair. Once people hear I can do this sort of thing, they all seem to pour out of the woodwork with all manner of strange and wonderful things for me to fix. One of the problems I have with this is that many of these devices are no longer supported, built to fail and designed not to be repaired, or simply not worth the considerable time it would take to fix them. The other day, a guy brought in a hand-held Garmin Rino 650 GPS siliconchip.com.au transceiver. This is a very handy unit because not only is it a fully-featured satnav system, it is also a two-way radio, so hunters who have other compatible transceivers can talk to each other over quite long distances in the bush. These devices have a touchscreen menu system and a PTT (push-to-talk) button on the side of the rubberised case, so they work much like the typical ‘walkie-talkie’ type devices we all know and love. The problem with this one was that there was no sound output, so while it can still be used as a standard navigation system, there were no audio Australia's electronics magazine prompts and obviously no radio transceiver function. These devices are designed to be rugged. They use heavy rubberised plastics and lots of rubber bungs and stoppers to provide basic weatherproofing. Pulling them apart isn’t too onerous, with the usual long Torx-type screws (which require the exact driver bit or they aren’t going anywhere) and the increasingly typical hidden screws buried beneath barely-removable case flashings. If one isn’t thorough, trying to pry apart a case that still has screws holding it together can end up a real mess, especially if one isn’t having a good January 2022  89 day and the red mist comes down in frustration! With all the screws removed, the two halves of the case separate relatively easily, but then the space is so tight that you have to hold your tongue one way while the planets align before the board assembly can be eased out. There is literally only one way it can come out, and finding just where that sweet spot is can be maddening. Once I got the board out, it was pretty obvious I was not going to be able to do much with that part of it unless the problem was caused by something simple like a blown or faulty speaker. The speaker was one of those super-thin permanent-­magnet types, about 25mm in diameter and plastic-welded into the front of the case. There was no physical wiring connecting the speaker; the board simply pressed against it, and a couple of tiny gold spring contacts on the PCB rested against the speaker’s terminals. I guess with everything crammed in so tightly, this system does work well, but I’d imagine any moisture that gets inside might interrupt this type of connection, so I resolved to take that into account as well. I decided to check the speaker first. My trusty multimeter could handle this with a quick touch to the speaker terminals with the meter set to x1 on the ohms scale. A click from the speaker told me it was working, which, to be honest, made my heart sink a little because I was hoping it would be one of those silly fixes that resolves everything with a minimum of fuss and expense. But no. There was obviously something further up the audio chain that was stopping this from making noise. As is usual in these situations, I hit Google to see if I could locate any circuit diagrams or schematics that may help. I found nothing but a wealth of misinformation. However, one theme kept cropping up: that on this model, it is easy to mute the output by accident whilst perusing the menu system, so many people had resolved the no-sound issue by simply un-muting the audio. I reassembled the GPS to the point that I could fire it up and go through the menu options on the touchscreen. The sound was not muted, which was disappointing, but there was also an option to set the handset to vibrate. I had seen the tiny vibrator motor mounted on the circuit board, so I knew it had that feature available. 90 Silicon Chip I set that to on, so at least there was some haptic feedback when something happened, though that wasn’t going to help get the two-way radio working... Breaking it down the second time wasn’t as finicky a job as the first time, so that was something. But trying to track anything back from the speaker side was a nonstarter. The multi-layer PCB was stacked with the smallest SMD components I’ve seen for a while, and though some did have numbers printed on them, I could find nothing about what they were. Even if I could find a replacement, removing them would require specialised tools that I don’t have, and if I tried to do it, I’d likely have just damaged the board further. So, this was one that I couldn’t help with. Garmin no longer runs a swap-out refurbishment program for this model, so they were no help. While I did check the usual auction sites for spares and replacement units, I could only find models being sold for spare parts, and they could very well have the same problem as this one – the vendors couldn’t tell me what had gone wrong with them. At almost half the price of a new unit, buying dead handsets in the hope that something might work was just not feasible (or sensible). The owner was philosophical about it; at least we’d had a look and determined it wasn’t worth pursuing. He’d still use it as a GPS but would have to do without the radio/audio side of it. Next! Another client brought in an old ’70s clock radio. It had been in the family for years, and though it worked, one of the red seven-segment LED displays had faded enough that it was difficult to read the time. This is a classic example of whether to repair or not, and why each case should be taken on its own merits. The clock radio had been bought for the current client back in the day by his dad, and so it had a lot of sentimental value. He had used it in his workshop for many years and wanted to see it going properly. I advised him that I could likely fix it, but he might be looking at way more than a replacement clock radio from some big-box store might cost. He was OK with that because the sentimental value was greater than that for him. I told him I’d see what I could do. Working on these older devices is so much nicer than a lot of today’s stuff: plain screws, simple engineering holding it all together and basic analog components with designations printed on them. Almost anything inside it could be repaired, or even fabricated to fit if that’s what it takes. This clock just had a single dim display from the four onboard, and I guessed that it had simply faded with age, as many LEDs of all types of this era do. The big question was what type of display it is and the pin configuration, because I literally have a parts drawer full of reclaimed and NOS (new old stock) red seven-segment LED displays, and I felt confident one would fit in this old-timer. I’ve said before; I’m not exactly a hoarder – at least not to the extent I have to sleep standing up in the laundry because all the rooms are stacked floor to ceiling – but my workshop is quite ‘busy’, with parts drawers and shelves groaning under the weight of stuff I’ve accumulated over the last 40 years or so. That’s not including all the other things I inherited from dad’s very similarly-­appointed workshop. Australia's electronics magazine siliconchip.com.au The new display was a little brighter, but not too bad. I guess I could have installed a resistor in the ground line in an effort to dull it, but I thought that once the plastic front was in place, it would look OK. I was right; once reassembled, the red plastic cover the LEDs shone through, which was a bit faded and scratched itself, tempered any bright spots on the new display, and it looked just as good as the others. The customer was pleased, and I’m trimming down my component stocks the right way, one at a time. On the slow boat from China The rationale is that if I come across a new-old-stock seven-segment display, I’m not about to throw it out just because I’m running out of room, so it gets squeezed into the drawer with the others. While there isn’t a huge call for components like this anymore, I could almost guarantee that if I did have a colossal clean-out and biffed a lot of this stuff away, the very next day, a job would come in that requires something I have just binned. This clock-radio job has proven that it would have been folly for me to throw these displays away because, as luck would have it, I had several that could do the job. My main concern was that if I fitted it successfully, this ‘new’ one would be much brighter than the remaining original displays. As I didn’t have four the same, I couldn’t replace them all. Still, I’d cross that bridge when I got there. It was a bit tricky to manipulate the various PCBs into a position that I could de-solder the dud display. The separate boards were all linked together using that multi-stranded, hard plastic insulated joining cable. It’s great stuff for a strong interconnecting joint, but over time it gets brittle and breaks easily. If I did break a link, I could always replace it, but it’s better not to bend these old parts around too much. The smell of the old solder brought back memories of watching dad in his workshop when I was a little kid; it’s strange how some odours stay with you. It’s likely seriously unhealthy, with all the fumes that come off when heated, but it smells of home to me. Getting the old dim display out was easy; I just wiggled it free after removing all the solder. The PCBs are pretty hardy from those days, but like all electronics, excess heat can do a lot of damage. So I just took care not to overcook it. I lined the new one up with the others and soldered it in – it was really that simple. I flexed the boards back into their original positions and sat it all carefully on the bench before plugging it in and lighting it up. siliconchip.com.au You might recall a story a while back about an electric bike that I couldn’t finish repairing as I was waiting on parts from China (June 2021 issue; siliconchip.com. au/Article/14895). More specifically, I was waiting for a new speed controller, because the old one had gone up in a puff of smoke. The problem was that I didn’t know if it was just the controller that had failed, or whether the motor assembly built into the back wheel had shorted and burned things out, or both. The controller was far cheaper to replace than the wheel/motor assembly, so that’s the bit I bought first. I’d sourced one easily enough, as they appear to be at least partially standard devices, but it took forever to get here. When it did arrive, I installed it – thankfully, most of the connections are also relatively standardised – though I’d taken lots of photos before I pulled the old one out as a precaution. I charged the battery, which had been sitting for a while and was discharged, and when all was ready, turned the key and wound in a bit of throttle. In a flash, the new controller was toast. As I’d already previously checked the external wiring to the motor for obvious shorts between themselves and to ground, I thought nothing was apparently wrong with it. Still, without a compatible controller, I couldn’t check it properly. Therefore, the controller was intended to be a sacrificial lamb and did its job by telling us that there was no point in carrying on and sinking even more money into it. It was a shame, really, as it was a cute little thing and not cheap to buy in the first place. Last call And finally, I had a guy bring in a PCB from a heat pump compressor. These are pretty large, and he said no one in town is repairing them anymore after a well-known repairer shut up shop due to the pandemic. Items Covered This Month • • • • The art of unrepair Macbook Air repairs Repairing a double-clicking computer mouse Replacing damaged varistors in two Panasonic microwaves Dave Thompson runs PC Anytime in Christchurch, NZ. Website: www.pcanytime.co.nz Email: dave<at>pcanytime.co.nz Australia's electronics magazine January 2022  91 I know next to nothing about these things, but how hard can it be? Circuit diagrams for common models are widely available for service guys, and the bloke (an installer/engineer) said he could get a circuit if I needed one. The fault was that the compressor motor no longer ran, and he suspected the motor driver section of the PCB had failed. To start with, I asked him if he’d tested the motor itself. He hadn’t but claimed not many fail, so it was most likely the board. He left it with me, and I got out the magnifying glass to see if I could find something obvious. I could see no burned areas and there was no acrid ‘dead component’ smell, so I doubted this was the problem. There were two onboard fuses, though, so I checked them. One was blown. I replaced it with one of the same specifications and called the installer chap to say there was nothing obvious here. He went back to the house and pulled the motor, bringing it straight around. It was shorted every which way I could measure it according to my meter, so that was likely the problem. Whether it had smoked the PCB itself was anyone’s guess. But again, it is way cheaper to replace a motor than a PCB in these units, so he took both away and installed a new motor, wired in the old PCB and voila! The thing fired up and is still running well today. It just goes to show that sometimes the simplest things can go wrong, but our tendency to over-complicate things can point us off in the wrong direction. He cursed himself for not checking those onboard fuses instead of wasting time taking the board out, but he made his assumptions on his own experiences. If he’d tested the motor, he’d have found the fault straight away, so I can be sure that’s what he’ll do next time. Either way, it was a super-easy fix for me, and everyone was happy. Macbook Air repairs B. P., of Dundathu, Qld previously wrote in to describe several MacBooks that he brought back into service. Well, he’s up to it again, this time rescuing some MacBook Airs from the rubbish tip... Some time back, a friend gave me several MacBook notebooks. I was able to repair three using parts from those 92 Silicon Chip and others that I already had. There was also a MacBook Air, but I’d been unable to test it because I only had Magsafe 1 chargers and the MacBook Air uses a Magsafe 2 charger. More recently, I was given another MacBook Air notebook, but I still didn’t have a Magsafe 2 charger. Now I had two of these MacBook Air notebooks and no way to test them. I’d previously looked into the price of a charger, but as they were around $40 or more, I didn’t want to spend that sort of money without knowing if these notebooks even worked. I decided to look on eBay for a replacement Magsafe 2 cable, and I found one for $11.95, so I ordered it. When it arrived, I dug out a 14.5V charger with a Magsafe 1 cable, and I cracked the charger apart using circlip pliers. One side of the shell came off fairly easily, but the inside of the charger proved challenging to get out of the other half of the shell. After I managed to remove it, I desoldered a wire and removed the copper wrap. I was then able to desolder the old cable and solder in the new one. Then I refitted the copper wrap and re-soldered the wire, and put the charger back together. The case clipped back together nicely, without needing to glue it. I grabbed one of the notebooks and connected the charger to it. I waited a minute and then pressed the power button. I heard the familiar Mac boom, so that was a good sign. The MacBook loaded up with the previous user’s account without needing a password, but something wasn’t right. There was no dock. Then I discovered that the keyboard and trackpad didn’t work, and I wondered if they had been disabled, so I decided to boot from a USB installer and check. I had an earlier version of Catalina on a USB, so I’d use that for now and update it later. I pressed the Option key and got the option of booting from the HDD or the USB drive. The trackpad now worked, so I chose the USB installer. I returned later, but now the keyboard and trackpad no longer worked; they were obviously faulty and only worked intermittently. I put this MacBook aside and grabbed the other one to check it. Once again, it started up, but I got a folder with a question mark in it, indicating that the eSATA SSD had either been wiped or removed, so I booted from the USB installer. There was still no HDD present, so I guessed it had been removed. I took the back off, but I found that the SSD was actually present. I removed it and replaced it and tried again, but it still didn’t show up. I suspected it might be faulty, so I took the back off the first MacBook Air, removed the SSD and installed it in the second MacBook. But it still didn’t show up; I knew it was good, which meant there was a fault with the motherboard. I decided that the best option was to swap the good motherboard from the first MacBook Air into the second MacBook Air. But this second MacBook had been dropped and there was a significant dent in the front righthand side of the base and the lid, as well as the base being bent where the left USB port is, and the lid would not close properly. The other shell was in much better shape, but I didn’t want to swap over the keyboard, which is a massive job. From what I’ve seen on YouTube videos with MacBook repairs, the keyboards are held in with a million tiny screws, so I would have to repair the damaged case. I managed to tap the front corner of the top case back into shape and straighten the bent area at the USB port, but I didn’t want to try to fix the lid, as I could risk breaking the screen. Instead, I would swap over the good lid. I started by dismantling the first MacBook with the working motherboard. I removed the battery, then the motherboard. I put the battery, motherboard and back of the shell aside to Servicing Stories Wanted Do you have any good servicing stories that you would like to share in The Serviceman column? If so, why not send those stories in to us? 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. Australia's electronics magazine siliconchip.com.au Left: the left-most charger is a Magsafe 1, while the one on the right is Magsafe 2. The Magsafe 2 charger uses the same pins, but has a slightly longer connector, fit to the repaired case. The reason for using this back is that it has the correct serial number of the motherboard on it. I’d use this battery, as it was already fully charged, and I knew it was good. I disassembled the second MacBook similarly, then swapped in the good lid and fitted the good parts. Now the lid closed properly after the shell repair. I also put fresh thermal compound on the heatsink and cleaned the fan at the same time. For the moment, I just sat the back on, in case I had to do anything else inside, and I turned it over and connected the charger. The MacBook started up, and this time, all was well with the keyboard and trackpad and I finished the installation. I decided to check for updates and was offered the latest version of macOS as an upgrade. The download finally finished, so I copied the installer to an external hard drive, and I also copied it to a folder on the desktop before running it. After some time, it was installed and I was greeted with a terrible wallpaper, which I quickly changed to the Big Sur photo after I logged on. Further testing indicated that this MacBook Air notebook was working correctly, so the $11.95 investment for a replacement Magsafe 2 cable was a good move. I screwed the back on permanently, gave it a quick clean and it was complete. This was a successful exercise in taking two unusable notebooks and making one good one. I don’t yet know if the leftover parts will be of any use, but I have kept them in case I get another broken MacBook Air in the future. When I’d had the other MacBook Air apart, I could see some corrosion around the connection for the trackpad, so that explained why it didn’t work reliably. It looks like someone had spilled liquid on the trackpad, which had seeped through and got into the connector. I haven’t looked into it any further to see whether that damage is fixable. It might be. Apparently, the small chip in the trackpad cable also controls the keyboard, explaining why both had stopped working. Working on Mac computers and MacBook notebooks is vastly different to working on PCs and Windows laptops. Mac computers and Macbooks need a lot more different tools, and they are a lot more compact. siliconchip.com.au Below: the Macbook Air opened up, so that the motherboard could be swapped. Australia's electronics magazine January 2022  93 Repairing a computer mouse that double-clicks D. S., of Maryborough, Qld made a similar repair to our own serviceman, who described fixing a computer mouse in the May 2021 issue... Dave’s mouse repair story made me chuckle. A couple of weeks ago, I was in a similar situation. A young fella had a problem with his mouse; apparently, it was double-clicking on the right mouse button whenever it was pressed. This, I was told, was “not a good thing when facing off with opponents in various online games”. I asked if he had played with any of the settings in the PC or the mouse’s installed software. He said he hadn’t, and without the PC, I had to take his word for it. I did ask if it was worth repairing, given that the repair might cost more than a new mouse. I was told in no uncertain terms that this was his “gaming mouse”, and he simply could not replace it, as it had cost over $100 new. I plugged it into my PC and ran it through its paces. It certainly had many buttons, and they all worked as expected, except the right button. It did indeed double-click with every single press. The screws holding it all together were hidden under more of those Teflon coated cushions, which I managed to save. Internally, the mouse looked like it was a nesting home for a cat or other small pet. The internal optics for the scroll wheel were buried under a soft blanket of pet hair, which extended across most of the mainboard. After I removed all that, I could test the microswitch on the board, and sure enough, it was faulty. After removing the small daughterboard, the mainboard came out, and I removed the offending microswitch. It was a fairly common part, even though it did have the Logitech logo on the side, so I quickly found a replacement. It all worked fine once reassembled, and the young gentleman was very happy. I did mention the pet hair inside the mouse (shouldn’t that be the other way around, mouse hair inside the cat!), and he smiled and said the family cat often slept on his desk, so that answered that question. I have also repaired his monitor twice, once for a bad tactile on/off switch and again when the mainboard stopped working. Both were easy fixes, although I had to buy 250 tactile switches in a nice neat little case; I will soon be ordering more, as I replace more and more of these switches. Like Dave Thompson, my eyes are not so good nowadays, so soldering the 2.1 x 2.8mm switches required the use of a desktop magnifier. I also have a repair story involving Coca Cola and a JVC 65cm LED TV. I won’t bore you with the details; suffice to say, it no longer worked. The photo below shows what I found after checking the power supply board and the T Con board. Coke always makes a mess of electronics. When these accidents happen, turn it off and get it to a service person ASAP. Don’t ignore it just because the TV (or whatever it is) still works. That’s what this teenage customer did, and check out the resulting damage. After I explained the fault and the time it may take to repair, I got the feeling that this young fellow would be mowing the lawn to pay Dad back for the cost of the repair. It took a couple of hours to clean it up and get it working again. Rest assured that the offending teenager learned a lesson and got Dad out of mowing for a while... Damaged varistors in two microwave ovens R. S., of Fig Tree Pocket, Qld has been busy fixing many appliances, including two microwaves. Despite being different models (both by Panasonic), they failed for the same reason... Both microwave ovens had intact mains fuses, but the protective varistor across the mains on both microwave ovens’ control/display boards were damaged. I think this could be due to mains surges. This varistor is a VDR10D511 10mm diameter, 511V varistor. There are also protective capacitors on the mains supply input board. The NN-SF574 oven (quite new) uses an inverter for the control board supply with a Panasonic flyback control IC and a high-frequency transformer. The varistor across the mains is fed by a fine track marked PF1, which acts as a fuse. This saved the board; replacing the track with some fine wire and fitting a new varistor got the oven going again. A new control board is about $110, and a new oven about $225. The NN-ST671 oven is an older design, with The doubleclicking mouse’s PCB, shown at left, had faulty microswitches. At right is a section of the power supply board of a 65in monitor. Some corrosion can be seen on the connectors from a soft drink spillage. 94 Silicon Chip Australia's electronics magazine siliconchip.com.au a small 50Hz transformer on the control board. Again, the mains varistor is fed by a fine track, which I had to replace along with the varistor. In this case, the varistor was sleeved with a high-temperature fabric. Perhaps varistors have caught fire or exploded in the past. I also repaired a Dome 24 wine cooler, Item No 900096. Taking the back cover off showed it was a Peltier Effect cooler with a large heatsink. There were two fans on the outside to remove the outside heat, plus one on the inside to distribute the cold air. The constant current switching power supply for the Peltier device had a blown fuse on the mains side. The supply uses a TL494 switching regulator driving two 13005 400V transistors connected in a totem-pole arrangement, probably to increase the voltage rating. Both 13005 transistors measured short circuit, as did the STPS2045 dual schottky output rectifier (2 x 10A 45V). After replacing these, I put a light bulb across the blown fuse and connected the mains. The bulb stayed on, so there was still a problem. I was surprised to find that two of the 1N4007 rectifiers in the full-wave bridge were also shorted on the mains input side. I am used to the large, sturdy rectifier bridges used in microwave oven inverters which never fail; the switching IGBT goes first. Replacing the 1N4007s and the fuse got the cooler working again, with about 1.5V across the Peltier Effect device. I did not measure the current. The big test will be when summer comes, to see if the wine stays cool. I also fixed a Kambrook K1780 steam iron. The series capacitor in its power supply was a 560nF 250V DC rated type that was down to about 430nF, so there was not enough voltage to operate the 24V relay that connects the mains to the iron heating element. So the iron would not heat. I notice that the capacitor manufacturers derate the voltage rating for AC, to about 60% of the DC voltage rating. For example, a 400V DC capacitor has an AC rating of about 240V. On that basis, the original 250V DC rated capacitor was not adequate for the task. It seemed to have been chosen for its small size, to fit in the space, rather than for a suitSC able voltage rating. The damaged VDR10D511 511V varistors taken from the Panasonic microwave ovens. siliconchip.com.au Australia's electronics magazine January 2022  95