Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN'S LOG That ‘80s gear and the art of printer repair Dave Thompson Regular readers of this column will know that I’m not one for throwing stuff away. They will also know that I’m a fan of stuff made a couple of decades ago. So when a client brought in his circa 30-year-old dot matrix printer, despite not normally repairing these things, I thought let’s give it a shot. Especially since it was described as ‘only’ having a broken wire. D on’t get me wrong; I’m not one of those hoarders who has to stand up to sleep because there is no more room in the house. However, my workshop is, shall we say, quite ‘busy’ with bits and bobs I’ve collected over the years. For example, a long-time client recently moved ‘up north’ [to Yorkshire? – Editor] and brought me a couple of presents before he left: a classic Avo meter and a Megger, both with original leather storage cases, user manuals and even leads. I couldn’t say no to those beauties, but they take up shelf space that I don’t really have. Still, I’m not the only one in this position. I fondly recall visits to my uncle’s electronics workshop in Melbourne when I could spend hours poking around, looking at all the gadgets and devices I’d not seen before. It was my idea of heaven. Recently, another long-time client brought in some familiar devices for me to repair – again. He has several Swedish-made electronic gadgets, built in the early 80s, using that nowclunky analog technology. I have had these in the workshop several times over the years for things like broken RS-232 socket wires or buttons that no longer work properly. Fortunately, the owner has a couple of extra similar devices that stopped working years before I was involved with them, and he now uses these for spares, especially as the buttons wear out frequently. The gadgets look surprisingly similar to my dad’s early HP calculator back in the 70s (Google tells me it was the HP-65 model). It boasted a swipe card with a magnetic strip that could ‘program’ the calculator’s functions, something I was quite taken with at the 84 Silicon Chip time. The red, bubble-lens LED display also looks similar and was very much of its time as well. When opened up, the Swedish gadgets contained three PCBs packed with EPROMS and other common chips of the time, plus room for a rechargeable 3V battery. I’m still not sure what my client does with these devices, but as he is retiring soon, he just wants to keep things going until that happens. I know he goes to various establishments, plugs these gadgets into machines (pokies maybe?) and downloads information onto them. He then takes the gadgets home and connects them via old serial-type cables to a green-screen computer I have also been keeping limping along. He can then print out the data he needs on a couple of old printers. What they do isn’t really important anyway; I just need to be able to keep them going. As with many handheld devices, he only uses a few of the keypad buttons to perform common tasks, so those wear out pretty quickly. If the plastic button’s top is popped off, a retained tension spring comes with it, and the contacts beneath are revealed. By today’s standards, I think they are a bit basic, but they do the job. Underneath the cap is a U-shaped copper spring contact about 15mm square overall but made from very thin metal, which looks quite flimsy. The top-left corner of this spring contact is connected through the plastic base of the switch to the button PCB underneath. When the button is pushed, the bottom right corner Australia’s electronics magazine of this piece deflects and touches the other contact, which is also moulded into the plastic base of the button. The obvious problem is that this main metal piece just wears out with use and eventually work hardens and breaks off, meaning the button stops working. Sometimes, I can just use a finetipped soldering iron to re-join the primary spring contact to the broken piece in the base, but this is a temporary fix only; the usual procedure is to replace the button itself. As I mentioned, he has several spare devices, and I have already used many of the buttons from these units. While the buttons are coloured and numbered, I just use the original plastic top from the broken one and put it onto the ‘new’ replacement base to restore functionality. Unlike a lot of stuff from that era, I got the distinct impression these weren’t designed to be worked on by anyone but the manufacturer. There is siliconchip.com.au Items Covered This Month • That ‘80s gear • Replacing the plugpacks in a • • dual-handset phone system A blown and charred mobile phone charger Repairing a 15-year-old Epson scanner *Dave Thompson runs PC Anytime in Christchurch, NZ. Website: www.pcanytime.co.nz Email: dave<at>pcanytime.co.nz no information about them anywhere; no circuits, schematics or other data except for the user manuals the client got with them when he bought them in the early ‘80s. To get an entire button out, for example, the whole device must be disassembled. Many through-wires must be disconnected/unsoldered with much gnashing of teeth and wringing of hands until the bottom of the button PCB is revealed. There is no other option, though, so we just soldier on. One good repair deserves another Recently, the same client called and asked me about having a look at his printer, which apparently had a “broken wire” and had stopped working. This sounded reasonably straightforward; I’m not bad at repairing broken wires, but printers in general are not my thing. I usually refer clients with printer issues to a specialist printer place. However, I said I’d have a look, and when he arrived with it, the job took on a different angle. This printer is 35 years old, one of two he uses every day to do whatever he does with them (my assumption is printing data!). Anyone who remembers these printers will know the type I’m talking about. They work very much like a typewriter, but instead of individual keys striking paper through an ink ribbon to make their mark, these printers have a matrix of pins within a print head which moves back and forth along the line and forms the characters instead. The major disadvantage is the sound they make; the name impact printer probably says it all. I recall going to offices and they’d have a special room set up just for these printers because they were so loud! One such office didn’t have a ‘printer room’, and the poor workers there had to suffer constant high-level noise; I could barely sit in the office chair for five minutes, let alone spend the day there with that racket going on! Still, these types of printers had their place and are still used by people who need duplicate (or even triplicate) copies, all done simultaneously. Of course, modern inkjets or laser printers can easily do multiple copies. However, some businesses still prefer that old tractor-feed printing that does it all at once (for POS systems where the customer gets one copy, for example). Anyway, the client brought the printer in and as a testament to the build quality of these older machines, it was the first time he’d ever had to take it in for a service. It showed; covered in dust and dirt, it really did need cleaning up. The biggest problem was this “broken wire” the client talked about. It wasn’t a broken electrical wire, but the main drive cable that connected the printer’s moving head assembly to the stepper motor, which of course, means nothing worked because the head no longer moved back and forth. This connecting cable is much like a bicycle’s brake cable; multi-stranded and designed to carry quite a substantial load. It is fixed to the movable print head of the machine, which is mounted on a couple of heavy-duty chromed rods, and the cable wraps around several pulleys and the stepper motor’s main drive shaft, allowing the print head to traverse the printer at some speed. You certainly wouldn’t want to get your fingers stuck in there when it’s going, which is why there is a removable plastic cover over the whole assembly. The cable was lying in the bottom of the printer, and it had obviously come unstuck somewhere. I’ve repaired a few cables like this before in different devices, so I said I’d take a look to see what I could do. ALSO AVAILABLE 10% OFF YOUR NEXT NOVEMBER ORDER WITH DISCOUNT CODE SCNOV10 FREE SHIPPING AUSTRALIA WIDE siliconchip.com.au THE TOOLS TO BUILD THE FUTURE w w w. p h i p p s e l e c t r o n i c s . co m Australia’s electronics magazine November 2021  85 Most repair people would just tell him it has had its day and to buy a new one, but there were a few problems with doing that. One, while still available, due to COVID19, there is not one to be purchased in New Zealand at the moment. Two, even if one was available, the cost is prohibitive now for a similar machine. Repair seemed like the only feasible option. Threading the cable back onto the various spools and pulleys wasn’t too onerous, except for the fact it has formed itself into very strong coils as it has been sitting in the same position for the last 35 years. At least this gave me a clue as to how it wound back on and where it could actually go, because it isn’t apparent just by looking at it from under the top cover. True to form, I couldn’t find a service manual for this printer anywhere on the web. I had only a vague idea of where everything should go back into place, so there was a bit of toing and froing as I tried different methods to get the cable back on track. Eventually, I got it into place – or where I thought it should be – and manually moved the print head back and forth. The cable was firmly connected to that, and I assumed that I didn’t want to move it because that might throw everything else out of whack. I theorised that if it had come off with the head in that fixed-cable position, it should go back on there as well. That was the theory anyway. With the cable in place, I manually moved the head along the shafts, and almost instantly, the cable pinged off the far-right hand pulley. I was back to square one. With a bit of fettling, I got the cable back into place and tried it again. Once again, everything just popped off; it was becoming evident that something else was wrong. The problem seemed to be with that far pulley, a plastic moulded disc running on what appeared to be a metal hub in the centre. It bore a lot of the cable load and, as soon as anything moved, the cable simply peeled off it. I took out the circlip holding it on and removed it for a better look. Straight away, I could see the problem; over the years, the cable had ground away at the channel it sits in, and it had simply worn through the plastic. Looking at it from the side, the bottom of the pulley channel was fine, but the top edge had disappeared completely, so there was nothing for the cable to run in anymore. It needed replacing, so I got on the phone and called a few printer places I know of to ask if perhaps they had an old dead unit sitting under the bench I could burgle parts from. No joy; as I suspected, most examples of this printer were skip fodder years ago. And new parts are not an option either, so it was back to looking at what I could do with this one. Jury-rigging it The pulley itself might be able to be repaired, and if I had a 3D printer, I might have been able to make a new one. But I didn’t, so I couldn’t. Still, I had an idea. The bottom edge was not worn through, so obviously, all the stress was on the top section, which had worn out. I thought I might be able to simply turn the pulley upsidedown and use it that way. Unfortunately, the cable wouldn’t stay on this way either, just falling off the bottom as soon as I moved the print head, so that wasn’t going to fly. But then I thought 86 Silicon Chip about using something to build up the missing bottom part of the pulley. I looked through my parts bins and found a large washer, which I know variously as a ‘penny’ washer or a ‘fender’ washer. Basically, it’s an oversized metal washer with a relatively small hole in the centre. I figured that if I bored out the hole to the size of the shaft the pulley ran on, I could glue this washer to the bottom of the pulley and essentially replace the worn-out edge. As it would be on the bottom now (because I would flip the pulley over), it wouldn’t be subject to the upward forces of the cable, and the now unworn plastic edge of the pulley would do the job its other half did for the last 35 years. Drilling the hole out on the washer was no problem, and I used a 24-hour epoxy resin and a press to glue the washer to the bottom of the pulley. I made sure to scuff the plastic to give the glue something to adhere to and centred everything up using an 8mm drill bit shank which I could remove easily later. It wasn’t brilliant, but I reckoned it should work. I reassembled everything once the glue had time to set, and other than a bit of cleaning of the glue around the cable run of the pulley, it looked pretty good. One problem was that now the pulley was around 50% thicker in cross-section, and I could no longer use the original circlip to hold this pulley to the shaft. I ended up drilling and tapping a small hole down the centre of the shaft, which turned out to be hardened steel, and mounted a small bolt and washer to hold the pulley in place. Threading the drive cable back into place was by now relatively straightforward, as I’d done it so many times, and the washer-bottomed pulley seemed to rotate and move just as well as it had before, with no binding or noise. The proof of any repair is whether it works or not, and so I plugged the printer in and hit the power button. The head shot back into its usual resting place, and all the right lights showed on the front panel. I didn’t have a parallel port with which to test it, so I asked the client if he could bring what he needed to try it out. He duly arrived with a handheld gadget and a bunch of cables and interfaces, which he plugged in, and soon we were rewarded with the dulcet tones of an impact printer doing its thing. The client has since reported that all is well and it is going better than it has in years. A job well done! Changing the plugpacks in a dual-handset phone system P. R., of Canterbury, New Zealand took pity on a Gigaset C300A dual-handset DECT phone system at a local church fair. It came with UK plugpacks, so it would not be usable for the average buyer, but it looked clean and tidy, and still in its original box, so he bought it... Once I got it home, I removed the two AAA cells from each handset, and they all showed signs of leakage, but the damage to the phones looked minimal. A new pair of cells brought life to one handset, but not the other. The next step was to sort out the plugpacks. I found two transformer-style plugpacks in my junk boxes, large enough to house the fairly small plugpack innards from the UK versions, held together by screws, not ultrasonic welding. Australia’s electronics magazine siliconchip.com.au It did not take long to cut open the plugpacks and reassemble them into the Australia/NZ style housings. I then cleaned up the battery housing of the working phone, set up the smaller charger and plugged the phone in. When I came back later, the battery was fully charged. It worked as a regular cordless phone, including an answerphone (I deleted four old messages). So that just left the non-working phone. There were no screws, so this would be one of those Humpty Dumpty jobs, crack the egg and try to end up with an intact, unblemished egg at the end. After watching a fuzzy video on the web, I could see that it was possible, and I knew which end was best to start from (the battery end). Lacking the usual tools for this, I finally managed to crack the case by apply gentle pressure using my bench vice on the battery end until I could fit a small steel ruler in the gap. I could then work my way up each side of the case with a small flat-bladed screwdriver until the whole thing gave up and opened. I could see some verdigris around the two battery terminals and other areas, but not so bad that it was not worth trying to save. A quick clean with warm water and a toothbrush, and it looked pretty good, but it was still not working. There is one LED between the two terminals that looked like the worst affected area. I removed the negative terminal, which was surfacemounted on two rectangular pads. Again, everything looked good, but an ohmmeter showed no continuity in the track joining the two pads together, nor in the zig-zag track through a low-ohm resistor to the ground plane of the phone. I could not see any breaks, but it turned out that both tracks were open at the point where they attached to the rectangular pads. This brought me right back to my early days servicing in the early 1970s. Back then, I discovered that some PCB designers had no concept of pad-to-track transition or sensible track widths. This was particularly apparent in an old pocket transistor radio that always suffered broken tracks when the user changed the batteries and accidentally applied pressure to vertical resistors. The designer of this phone would probably try to empty a football stadium through a standard house door. In the end, I had to repair seven tracks with fine wire, two on the negative terminal, three on the positive terminal and two on the LED. Only two tracks had an almost sensible width, and one of those was open-circuit. All these tracks could have been at least five times wider in the space available, and fillets are always good practice on pads that could be subject to stress (like battery terminals!). While this was a corrosion problem, not a stress fault, it could have saved all the re-work. Once reassembled, this phone joined the other and charged its old battery just fine – looking only a little battered along one edge from the case opening. Was it all worth it? Probably not, but for only $4 plus my time made this a reasonable upgrade from my old phone. A blown mobile phone charger B. P., of Dundathu, Qld got a rude surprise when one of his children handed him a blackened lump to diagnose and fix. It’s a good thing it didn’t burn the house down... My son handed me a phone charger and told me that siliconchip.com.au Australia’s electronics magazine November 2021  87 Even just a glance at the phone charger shows substantial burn marks which could be likely attributed to a blown diode and failed electrolytic capacitor. he was charging his phone after he went to bed last night, when the charger blew up with a loud bang, bright flash and smell of burning. I opened up the charger by undoing the two screws at the back. The inside of the top cover was blackened and I could see that the likely cause of the near-disaster was the failed 2.2μF 400V electrolytic capacitor that had a bulging top. Further investigation showed that one of the diodes had a piece blown out of its lead, and when I checked resistor R1, it was open-circuit. I have not previously seen one of these chargers blow up like this; they usually just fail silently and just don’t work anymore. It was only worth a few dollars, but I thought it would be interesting to see if it could be returned to service. I started by looking through my collection of reclaimed capacitors for a replacement 2.2μF 400V electrolytic capacitor, but the only 400V capacitors of the correct physical size I had were only 1μF. Then I thought to check another failed phone charger that I hadn’t been able to repair because of its more complex circuitry, and luckily, it had a good 2.2μF 400V electrolytic capacitor I could salvage. I did have a 1W 1W resistor in my collection, along with four 1N4004 diodes. I didn’t bother testing the remaining three diodes that hadn’t blown up, as I decided to just replace all four to be safe. With the replacement components fitted, I set the top back on the charger in case it decided to blow up again when plugged in. I plugged it into a power board, and nothing happened other than the indicator LED lighting up; a good sign. I plugged in a USB voltage tester and it showed that I had just over 5V, so the charger was now working again. This was somewhat surprising, as I had half expected it to blow up again when it was plugged in. I reassembled it, and it can now be returned to service. I wonder how many house fires are started by unattended chargers like this one failing spectacularly. Epson scanner repair A. L. S., of Turramurra, NSW went down a rabbit hole trying to fix his trusty old scanner. It probably wasn’t 88 Silicon Chip Australia’s electronics magazine siliconchip.com.au worth the time he spent given how cheap they are to buy these days, but it’s a pity to throw it away just because a couple of low-cost, easily replaced components have gone bad... My 15-year-old scanner started stopping halfway through a scan. Repeating the scan occasionally produced a complete image, but over time it got worse, eventually refusing to scan documents at all. I checked YouTube to see if anyone else had a similar problem. There was one person who had a sticking Epson scanner and described the repair in excellent detail. He described it as being “a reasonably easy repair” (https://youtu.be/RsmRNWoYkQI). His scanner was a different model to mine; the workings looked similar, but the repair involved disassembling all the mechanical components and cleaning the running rails which guide the scanning assembly. He also used replacement parts cannibalised from another scanner. I checked out available parts online for this scanner in the hope that I could find a dead one “for parts only” or some spare belts and pulleys, but all I could find was a used controller board for $90 including postage, which seemed way too steep. I thought it would be a waste of time and money anyway because I assumed my problem was mechanical, as it stopped scanning at the same location each time. If it was a mechanical failure then all I likely needed to do was free up the movement, so I split the case to get to the mechanical gear. It didn’t quite go according to plan because only two screws were holding it together, and as I prised the case open, I heard the horrible sound of plastic retainers snapping. There were several of these around the perimeter, and they all completely broke off. The plastic retainers required some sort of magic trick to separate, and I am still unsure what that was! I have repaired hundreds of similar devices and never came across a case like this. Fortunately, the absence of the plastic bits made no real difference, and the two halves of the case snapped back together again after I had finished cleaning the rail for the scanning assembly. The assembly moved very smoothly after I cleaned it. The belt and pulleys were all OK, and I was convinced that siliconchip.com.au Two faulty 10μF 6.3V SMD electolytics were removed from the controller board of the scanner and replaced with through-hole electrolytics rated at 25V. the scanner would now come back to life. After I plugged it in again and set it off for a scan – nothing had improved; it still scanned only half a document. I was really puzzled! If the mechanicals were operated smoothly, what else could be going wrong? I put it all down to a faulty motor or motor controller, but the curiosity was slowly killing me, so I eventually dragged it out, dusted it off and fired the thing up again. This time it scanned half a document immediately, but when I tried again, it stubbornly refused to obey any command. My thoughts turned to the plugpack; it might have a low output voltage or be overheating or somehow limiting the current delivered to the scanner. I checked it out by loading it with a 68W 10W resistor, but it maintained its rated voltage, so I had gone up another blind alley. It then struck me that I had not even inspected this controller board when cleaning it. That’s because it was entirely covered by a ‘full metal jacket’ that was screwed in place and also hidden by another PCB carrying the switches and the power indicator LED. Instead of opening it all up again and taking the time and effort to pull the circuit board out, I took a shortcut and zoomed in on the photo of the replacement control board on eBay. As I zoomed in on the tiny fuzzy photo, I had a light bulb moment – this thing had two 10μF 6.3V SMD electrolytics, the same type which caused me grief on a previous repair and that are famous for going bad! One of the symptoms was that the Australia’s electronics magazine scanner started to work when it was cold, but it would not work after being switched on for any length of time. I think that this might be because the electrolytic capacitors had degraded and become temperature-sensitive. So I extracted the circuit board and removed those two SMD electrolytic capacitors. This is easy if you know how to do it correctly. The best method is to gently rock the SMD electro from side to side with needle-nose pliers. This fatigues the pigtails, but you have to be patient and “gentle” is the operative word. After removing them, I checked them with a capacitance meter and they measured a measly 0.73μF and 2.3μF! I did not replace these with SMD capacitors but instead used conventional electros because modern through-hole types are small enough and are easy to solder, as you can see in the photo. These replacements also had a higher rating of 25V and were easily soldered to the SMD pads and insulated with some plastic tubing, mounted horizontally to allow space for the metal jacket to clear. I also replaced two other suspect conventional electrolytics with new, higher-rated ones and bingo! The scanner worked perfectly. I suspect it was only scanning half the page because it was getting to a point where greater demand was placed on the power supply, and with such poor bypassing, the voltage dropped too much and reset the controller electronics. This symptom was a live red herring designed by an ingenious gremlin. SC November 2021  89