Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN’S LOG Carpet vacuums suck, too Dave Thompson I don’t know what it is with vacuum cleaners and this household lately; if it isn’t one thing, it’s another. The symptom this time was a lot like the last – pull the trigger and nothing happens – but as it turned out, the cause was altogether different. After my previous repair of the Bissell AirRam, detailed last month, everything seemed fine. However, we ran into problems with another of our cleaning appliances. While the ‘repair’ was simple enough, it wasn’t an overly simple process, especially once we discovered what the problem was. Me! Let me explain in my usual roundabout fashion... Once a year, during the height of the summer months, we usually do a complete wet carpet clean throughout the house, just to spruce things up a bit. This year was no different. The problem is, last summer, we had some very unpredictable and inclement weather. We prefer to have a few nice consecutive sunny days to open all the windows and thoroughly dry the carpets out after cleaning them. As we felt we wouldn’t have that opportunity, we ended up not doing it at all that year. However, this year, we had to do a proper shampoo as the carpets were starting to show their true colours! Even this year, summer has been grey and unseasonably damp, with very few decent spells of warm weather. We’ve seen some extreme weather in other cities around the country (and indeed, other countries, as many of you know all too well). As it turned out, we were quite lucky not to have the severe storms, rain and floods that other towns and cities here were subjected to (and still are). We have had bursts of finer days, though, with the ‘mercury’ in the high twenties and low thirties, so we decided to take the opportunity to break out our wet vacuum/carpet cleaner and finally get our carpet clean. Once again, the machine is a Bissell appliance – we’ve had several Bissell vacuums of different types over the years, and all have been pretty good machines. Even though this one is getting on a bit now, it hasn’t done a tremendous amount of work because we only shampoo the carpets once a year (or thereabouts). Items Covered This Month • • • A carpet vacuum magic trick Remembering core memory Daikin three-phase aircon repair 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 siliconchip.com.au We try to keep the carpets clean regardless, with a ‘no shoes in the house’ policy and only the odd pet accident to clean up, meaning it is probably not essential they are done every year. Still, we try to keep to that schedule. Which brings me to my point; this machine does a lot of sitting around doing nothing. That also means that when we go to use it, we have to relearn how to operate it all over again. What’s the old saying? Use it or lose it? Well, that applies here, because the machine is relatively tricky to use and, to get the best results, it needs to be operated while taking all those quirks into account. Fortunately, we keep all the user manuals (they are also available for download on the web anyway). Since this was the first time we’d had problems when we fired up the machine, we had to find and dig the user manual out. To begin with, this is a very typical carpet shampoo vacuum cleaner. They are usually quite bulky devices and relatively hard to manoeuvre, especially around corners or in tight spaces. Ours is no different, although, to offset this and make it more appealing to the home user, our model boasts a removable motor assembly. That assembly has a smaller nozzle attached so it can be more easily used to shampoo and clean the likes of stairs and upholstery, where the original machine would not have a hope of reaching. Australia's electronics magazine March 2023  83 When the removable unit is ‘unplugged’ from the main body of the cleaner, a mechanically-operated valve diverts the water and shampoo mixture and vacuum to the small hand-held nozzle instead of the main head unit. This machine overall does a very good job for a domestic cleaner and has given us good service over the years. Rinse, lather, repeat In these devices, a specifically measured water/shampoo mixture is loaded into one of the two onboard tanks and pumped down through the head assembly onto the carpet (or upholstery) when a trigger is pressed on the operating handle. This mixture is driven deep into the pile of the carpet in a swirling motion, due to the head’s water-jet placement and the water pump pressure. When the carpet is well-shampooed, the trigger is released. The mixture stops pumping out, and the machine then essentially becomes a regular wet vacuum that pulls all the dirty water back out of the carpets, leaving them as dry and as clean as possible. Multiple passes are usually required for both the shampooing and vacuuming phases of the process. The dirty water (usually astonishingly dark and filthy looking) is collected and dumped into a clear plastic reservoir, which sits adjacent to the water and shampoo tank on our particular model. Both these tanks are removable for filling and emptying by means of clever locking levers. Therein lies the rub My wife got our shampooer out of the cupboard, installed all the hoses and bits and bobs and set it all up. She filled the tank with the correct shampoo/water mixture, plugged it in and switched it on. The problem she encountered was that nothing happened when she pulled the trigger. There was none of the usual water pump start-up noise from this machine when the trigger was pulled, but there was still plenty of suction at the cleaning head. Evidently, the vacuum part was working fine, but something was not quite right. This didn’t bode well, and was all I needed after the last vacuum-cleaner-related fiasco. 84 Silicon Chip The first thing I did was pull both the tanks off and check all the filters underneath. All are removable, but some only with screwdrivers, so I went to the workshop and tooled up for the coming disassembly. Like all our other Bissell vacuums, this one is also as over-engineered as a Bugatti Veyron. Everything removable is held on with many screws or bolts, making it quite a substantial and hard-to-disassemble unit. I hit the internet for a service manual but, as usual, found nothing (except those for sale on some manual sites). Still, I did find a couple of service videos on YouTube that vaguely included this model, though not in any great detail. It did get me up to speed on the filter checks and removals, however, so naturally, that’s where I started. Once I got them out, I could see the filters in question were relatively clean, with a little fine dust in some and a few pet hairs in others, but not enough to stop it from working. Regardless, I fired up the air compressor and gave them all a good clean-out before refitting them. I didn’t expect that solution to work, but I tried it anyway; still nothing. I hoped the pump hadn’t failed, because I wasn’t sure I’d be able to get another one, at least at a reasonable price. A pump failure could make the whole thing redundant, and it would then be just landfill fodder, a very unpalatable option. Time to take it apart The only thing for it was to strip the machine down to the pump and see what was happening. As I had no service manual, I was going in blind. That is not unusual in my line of work, but I would rather have at least some diagrams to follow, especially if it all springs apart somewhere and bits go flying. I’ve been there before, trying to reassemble something without any direction or idea of how it goes back together. It is incredibly time-consuming and frustrating; potentially, it might never go back together the same way. If that were the case, I’d also have to discard it. No pressure, then. I began by removing all the outer panels that could be removed; by past experiences, that might not have any real benefit, but I thought I might be able to see what was going on underneath them and find a way to burrow down to the pump. As is typical, the screws were very tight and some were hard to access, but I managed to get the panels off using several of my dozens of screwdrivers. Fortunately, there were no security-type fasteners, as seems to be the Bissell way, because that makes things so much harder. Underneath, I found two other sections I needed to remove to get to the pump, or at least where I thought the pump was. One was particularly difficult as it seemed to be interlocked with clips to the part next to it, but I finally finagled it off with much blue language and gnashing of teeth. There were also a couple of smaller filters in this area that I could remove and check. Both were a bit grubby but still relatively clear. I cleaned them with an old toothbrush in a bit of water and used compressed air to blow them out anyway. I finally found the pump, a compact self-enclosed unit, and removed the three screws holding the assembly in place. It spun freely, but without a service manual, I had no idea what voltage it ran on, and as it had no visible Australia's electronics magazine siliconchip.com.au markings on it, I had no way of finding out exactly what it was. Another Google search showed up similar items that looked like it, but with no specs available; I wasn’t about to just throw power at it to test it. Frustrating! At least it hadn’t jammed up, and while I had access, I carefully blew the water lines going to and from the pump through with low-pressure air to ensure everything was clear. It all seemed as expected, so I left it at that. Testing the trigger At this point, I considered that the trigger mechanism itself might be where the fault lay, and that it simply wasn’t switching the pump in and out, so I set about disassembling the handle assembly. This was an act in itself, with several screws holding it together that were quite challenging to get to, even with all my screwdrivers. I got it apart, though, and could see all the wires and the switch inside the handle. I used my trusty multimeter to ring the leads out, all the way down as far as I could see, and all seemed to be connected correctly. The trigger was working and switching, according to the meter. I tried ringing out the circuit all the way down to the pump motor leads, and while I got no reading, I wasn’t too surprised, as I knew there must be other sensors that controlled pump operation, such as the water tank being empty or the waste tank being full. I know this because the user manual mentions these as safety features. I couldn’t see these sensors in any part I had taken apart, and I guessed they’d be situated in the pipe assemblies somewhere beneath the tanks. Since the water tank was full and the waste tank empty, I didn’t think it would be one of those sensors preventing the pump from working. In my mind, it was still looking like the pump. I found what appeared to be a replacement unit on the web; the model of the cleaner matched, and the picture of it looked very similar, but I’d have to import it as there was nothing I could find locally. What a pain! And there was still the risk that replacing the pump assembly wouldn’t fix the problem anyway, so I held off going down that particular route for the time being. I still had niggling doubts about the whole deal – something seemed off about it. The cleaner had worked perfectly the last time we’d used it, and the pump wasn’t jammed by build-up or any foreign objects. The lines were all clear and the wiring was intact; it just didn’t make sense that it would stop working while sitting in the cupboard doing nothing. In the end, I reassembled the machine. The last thing I wanted was to have to wait for a new pump and then forget how it all went back together! At this point, there wasn’t much else I could do except try to find a service manual, or perhaps post in some online forums I had found, to find out what the experts had to say. 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. everything back together for what I knew would be a futile test of whether what I had done had made any difference, she suddenly said something that made perfect sense. She said, and I quote: “I think I put the shampoo and water in the wrong tank.” That did make good sense – as I mentioned, when that waste tank is full, the pump won’t switch on, and when the water tank is empty, the pump will not operate either. If I’d had even an ounce of sense, I would have figured this out before wasting all that time and expending so much bad language chasing a ghost. It transpired that she had forgotten how to use this machine as much as I had, and just filled the waste tank with the shampoo and water mixture, thinking that was the water tank. To be fair, the water tank system is a little unintuitive; just looking at it, anyone would think the waste tank is for water and shampoo. Still, at the end of the day, I was the one who went all repairman on it without looking for obvious solutions first. We poured the mixture from the waste tank into a jug and filled the water tank with it (pouring directly between them was not going to work due to the design), then put them both back in place. This time, well, you know what happened; everything worked perfectly. ...continued on page 88 Eureka! While I’d been doing all this, my wife was also hitting the web, trying to find anything relevant that might help. This is not uncommon; she often looks at things I wouldn’t think of, and vice versa, so between us, we can usually get to the bottom of something if we look long and hard enough. As I was turning the last few screws in and putting siliconchip.com.au Australia's electronics magazine March 2023  85 I was kicking myself. I’m always banging on about hearing hoof beats and looking for horses, not zebras, but in this case, I just assumed that everything was right and there must be a fault if it wasn’t working. Thus, it is a very cautionary tale, then, and one I’ll (hopefully) learn a lot from. We’ll see! Remembering core memory R. E. of Freshwater, NSW enjoyed the article on “The History of Computer Memory” (January 2023; siliconchip. au/Series/393), particularly the section on magnetic core memory. It made me laugh, as the Australian Navy was still using core memory as ROM in 1990 and even later. In fact, I was informed that the bootstrap loader for the ROM was a punched paper tape. How embarrassing! 26 years earlier, I was introduced to the obsolete technology of saturable core reactors and magnetic amplifiers in my “Industrial and Automation Electronics” course at a college in Toronto. We learned that if you saturate a transformer or other magnetic material, it will no longer pass a signal from primary to secondary, but rather leave the secondary at 0V AC. In 1990, I was working for Bellinger Instrument Pty Ltd, a small defence contractor in Rydalmere, Sydney. The policy of the company was that we would repair and refurbish anything that the three defence arms could throw at us. “If it is small enough to fit in the building, we will fix it” was the motto. In came six boards of unknown function or use, with absolutely no documentation and most likely security restrictions on the system it came from. I must have been standing in the line of fire, as I got the job of determining what they were, their use and reverse engineering, writing a test, and repairing any faulty boards. I deduced that these were 32-bit core memories of unknown capacity. With my introductions to the saturable core reactors, I deduced that a core memory can be used to store data. A core that has been forced into saturation with 88 Silicon Chip an excessive current will not pass any voltage transients to the secondary, but it will retain the received energy once the energising current is removed. This energy is transferred to the secondary as back-EMF and can be measured as a several-millisecond-long pulse in the secondary, and a shoulder in the primary voltage directly after removing of the saturating current. The ROM boards I had were a three-wire system, where each toroid had three wires affecting its function. One was a biasing wire that energised each toroid with a direct current to just below the saturation point. The second was an address line for the 32-bit memory (or primary winding of the toroids). The third line was the read wire or secondary winding. By applying a pulse to the address line with enough current to put the toroids into saturation, the read line received a several-millisecond-long back-EMF pulse after the address pulse was removed. The bias and read wires are fed through the centre of the toroids, but the address wires were fed through the centre only for a ‘one’ bit and fed around the outside of the toroid where they did not send it into saturation, with no backEMF pulse, for a ‘zero’ bit. With this in mind, I developed a system to send pulses through the all addresses, then read the ROM contents with a sample-and-hold circuit timed to the back-EMF pulse, and present the result to a 32-bit logic analyser as a waveform pattern. The analyser had the capacity to recall previously stored memory maps and compare them to new data coming in from my ROM reader, thereby highlighting any errors. Errors could have been caused by a broken toroid, or one no longer able to be saturated. We found several faulty boards and were able to get replacement toroids from the Navy. Then came the task of removing three wires from a 32-toroid strip and any faulty toroids, noting the in/out sequence of the address line and rethreading three new wires through all 32 good toroids in the noted sequence with a needle and fine wire. Not easy task for a technician with fat fingers, as the toroids were less than 5mm in diameter! Without a memory map from the manufacturer, we relied on the principle that if five boards were exactly the same, and one was different, we had found one faulty board. This principle had served us well in the past without any consequences. As a technician with limited design experience, I found the process quite challenging, but I was able to continue the company policy of “we can fix anything”. Unfortunately, the company was unable to continue providing these services due to changes in government policy, resulting in insufficient work from the defence force for small contractors. Daikin three-phase aircon repair K. W., of Craigburn Farm, SA found that sometimes faults in seemingly complicated devices can be simple enough to find and fix with just a bit of investigation. It sure beats having to buy a whole new control board... On a hot Sunday, my Daikin FDY “F series” three-phase air conditioner started playing up. It gave me Error E3, which an internet search revealed was likely due to a compressor over-pressure condition. To survive the day (and Australia's electronics magazine siliconchip.com.au keep SWMBO [she who must be obeyed] happy), I periodically sprayed the compressor cooling fins with water and powered the unit off and then on again. The problem was that one fan of two wasn’t running, so the compressor was overheating. Direct sunlight on the unit and the wall behind it didn’t help either. My wife wanted me to call in a pro (oh, she of little faith!). After removing a bunch of screws from the outside unit, the top and a small front panel came off easily, exposing the electronics for inspection and probing. I made some measurements, then phoned an electrician mate who has a refrigeration ticket. He came over and had a look, but went away with the model number, expecting to get the price of a new controller board for me. I thought I’d investigate further. First, I went next door, where there is an identical unit and got my neighbour to start his air conditioner to see if both fans ran at startup. They did. I’d already unplugged the fans and swapped them over. The fault didn’t follow the leads, so the fan motor was OK, even though the winding resistances on one were a bit higher than the other. I then swapped the leads to the two start capacitors. No change, so they were both good. The circuit diagram inside the cover showed that the wire into each motor (excluding the capacitor connections) was fed from a relay. I prised the control board out and measured the coil resistances on all those PCB-mounted relays and the (I think) 1kW resistors going to them. All were roughly the same. I thought maybe a relay had crook contacts. I’d had to pull leads off a couple of connectors to move the PCB, and I noticed one was very wobbly. It definitely needed resoldering, thanks to my rough handling. That made me wonder if any other connector solder joints were dry/broken. One lead to the non-working fan was near zero volts, while on the working fan, that same lead was at 240V AC. Between those connector pins was a copper trace. So out with the magnifier and torch; sure enough, I found a broken-­off solder joint on one of those pins. After scraping the nearby tracks with a small cutting tool to expose the copper, I repaired the bad joints I could see and then touched up a few others for good measure. Both fans were running with the power back on and the air conditioner set to cool. Success! I put the covers back on, and SMS’d my mate so he didn’t have to chase up a new controller board. The moral of the story is: if you have experience with mains-powered equipment, have a go. The fault might be trivial, even where a microprocessor is involved. I felt a bit silly not suspecting a dry or broken joint in the first place. A PCB, connector pins, and vibration are a recipe for eventual failure. A few years back, I fixed our front-loading washing machine with the same type of problem. The controller board was mounted on top of the drum! The symptoms were intermittent wash operation. I couldn’t see the dry joint except with a powerful magnifier; resoldering the high-current joints fixed it. The ordinary repairman (not Dave, of course) simply replaces controller boards at great expense and waste. When your time is free, a deeper investigation is warranted. SC siliconchip.com.au Australia's electronics magazine March 2023  89