Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN'S LOG Chasing wild geese isn’t as fun as it sounds I don’t know about other servicemen, but there always seems to be something in my household that needs fixing. I’m not just talking about stuff I encounter in my day job, or even the computers or phones in the office that get messed up with updates or apps that don’t work. I mean those domestic jobs that always crop up that often need a serviceman’s touch. For example, we were experiencing an intermittent problem with some of the devices in our home theatre system. Now and then, we’d lose power and while the TV still worked, the amplifier and disc player would go dark. All the plugs were fully pushed into a four-socket power board, except the TV, which plugged in further along the wall. It didn’t take long to discover that this power board suffered from the same problems that I’ve seen readers mention in several letters published in Silicon Chip; in other words, it was cheap, nasty rubbish. A tap with my foot on one edge of the board resulted in power dropping out. Another tap in a different place restored it. I couldn’t be bothered tearing it apart to find the root cause; I’ve been down that road before and there is typically nothing fixable inside anyway. What really ailed it was poor design and shoddy manufacturing. siliconchip.com.au I solved the problems by replacing the power board with a new, betterquality model. While this issue was easy enough to deal with, it still took time and effort to track the fault down. There was a more trying example recently when we awoke to lukewarm hot water and struggled to get in a couple of showers before the water was too cold. This is unusual as we are on a night-rate power plan; heating our water overnight takes advantage of the much cheaper off-peak electricity rates. While this usually works out well, something appeared to have gone awry. As usual, my serviceman brain immediately kicked into gear, mentally troubleshooting the possible causes. But there were some “wild card” factors muddying the waters. Around ten months ago, it was announced the drinking water in Christchurch, long prided on being the clearest and cleanest in the world (if local lore is to be believed), was to be chlorinated. This caused quite the backlash from the masses, including me, who strug- Australia’s electronics magazine Dave Thompson Items Covered This Month • • • • • Cold showers in Christchurch BWD 275 dual-range 36/72V power supply repair Static from a Codan X2 highfrequency transceiver Cordless vacuum repair A not so steamy kettle *Dave Thompson runs PC Anytime in Christchurch, NZ. Website: www.pcanytime.co.nz Email: dave<at>pcanytime.co.nz gled with the reasoning behind it. Due to a gastro outbreak in a city in the North Island – the result of contaminated tap water – our local council got spooked and decided that for the public good, chlorine must be introduced immediately into our water supply. What that outbreak had to do with us, a whole island and a half away, baffled me. I suppose that any of the hundreds of bores that tap vast aquifers deep under Christchurch might suffer the same fate as those ‘up north’, requiring the bore’s hardware (some of which dates back almost a hundred years) to be replaced (no doubt at a huge cost to the taxpayer). But adding chlorine to our water just seems like a solution looking for a problem. To placate the nay-sayers, the council claimed chlorination would only be required for a short period, and only in a few problem areas while all the bore heads were tested and/or upgraded. It all sounds plausible, especially as the quakes might well have had some impact on the state of these aquifers and bores. But the latest news is that the water could be chlorinated for years, Christchurch-wide, which is just adding to everyone’s anxiety over the issue. Anyway, the point to this backstory is that since this chlorination program started, more than 2000 hot January 2019  61 water tanks in homes around the city have corroded through and been ruined, apparently by the chlorine in the water. This number doesn’t take into account the hundreds of plumbing leaks and pipe failures that are also attributed to the chlorination of the water supply. Now I’m no plumber or water-tank guru, so I can’t say for sure if this was a just glitch in the matrix (ie, a coincidence) or merely a problem with some older pipes and cylinders, of which there are likely many still around Christchurch. Either way, it’s an unexpected boon for water heater installers and manufacturers. The rest of us are mostly just concerned about how our own cylinders and pipes will fare in this situation. The game is afoot Of course, chlorine-driven corrosion sprang to mind as a possible cause for our lack of water. The first thing I did was to make sure that we didn’t have a new indoor swimming pool beneath the hot water tank. The second thing was to check the breaker at the switch-box to make sure it hadn’t tripped. Everything looked OK, so with those two causes ruled out, I’d have to look further afield. Now before I get flamed by those far more knowledgeable than I am about low-pressure home hot water systems, let me qualify my troubleshooting process with the fact I know next to nothing about how it all works. I realised there could only be a few possible causes of no hot water, the most obvious of which would be the heating element itself failing. The element in our heater is a resistive immersion type, so its continuity should be easily measurable, and the wiring to, and within, the heater panel should ring out as well. I’d need a multimeter for these tasks; I chose my analog model as it is easier to read in tight corners. I then did what all servicemen love to do; break out the tools and remove whatever covers I could get off in the hope I’d see something really obvious, such as soot deposits, a broken wire or a blown fuse. Maybe I’d get lucky! The water heater looked to be relatively new and sits in a cupboard upstairs. I’m guessing it replaced the uninsulated, low-tech, late-fifties original when the house’s second storey was 62 Silicon Chip added in the mid-nineties and everything was relocated from downstairs to the upper level. I removed the large, circular “biscuit-tin” lid that shields the wiring and electrics. This is held on with two diametrically-opposite PK style screws, attaching it to a round housing that is spot-welded to the side of the cylinder. Once the screws were removed, it was relatively easy to pry the cover away with the edge of a screwdriver (you could use your fingernails if they are tough enough!). As the shaft for the thermostat control protrudes through the cover, the temperature-setting knob must also be removed for the cover to come away completely. In this case, the knob just pulled off the pot shaft with a little outward pressure. Once the cover is off, the wiring to the element and thermostat is very easy to access. Everything looked perfect, as if it had been installed yesterday, so no obvious fault presented itself here. As I was checking this during the day, theoretically there shouldn’t be any power present on any of the terminals, but while I might sometimes be an imbecile, I’m not insane, so I tripped the breaker at the switchboard. And though there is an isolating switch on the wall of the hot water cupboard, I wasn’t about to take it on faith that it was wired correctly either. Instead, I used my mains-detector tool to check for mains-level voltage in the wiring at all the points from the wallswitch to the element. When the tool started screaming at me, indicating voltage was present, I considered my caution justified. However, I know from experience that this tool can sometimes be too sensitive. I suspected it was picking up stray emissions from a mains-wiring loom that ran through the floor cavity just beneath the heater cupboard on its way to the main switchboard downstairs. In reality, I couldn’t get anywhere in the cupboard without the detector going off. To be 100% sure whether mains voltage was present, I’d have to measure it, so after removing the two retaining screws for the isolating switch plate and dropping it clear of the wall, I used my multimeter to ring out the system. I measured zero volts on all points, regardless of the wall switch or therAustralia’s electronics magazine mostat control’s position, so I was confident there was no power flowing to the heater and that my detector was indeed picking up stray emissions. With everything now electrically dead, I removed one of the Active leads from one side of the element and with my meter set to the ohms range, measured the element itself. This type of element actually has four terminals; I assume that there are two separate-but-identical elements as they were wired in parallel, with each pair of terminals bound to the adjacent pair with heavy brass links. All I needed to do was put a lead on each bus bar to measure the resistance. I got a reading of around 12W. I didn’t know what it should be, but that sounded about right. At least it wasn’t open circuit, and as the main breaker hadn’t blown, I knew it was unlikely to be shorted out. I also tested the thermostat and it clicked in and out fine, with continuity from the power leads to the elements when it was on. Since I had reasonable element resistance and all the wiring looked good to and from the switch plate, and with the switch properly isolating the mains feed to the system, nothing appeared untoward here. If there was no power getting this far, there must be something else somewhere upstream preventing it from getting to this point. Doing a sparky’s leg-work I digress now to another back-story that may have a bearing on this problem; when we renovated this house, we replaced a lot of the ropey old wiring we found in the walls and ceilings with new cables. I did most of this work under the careful scrutiny of a licensed electrician. He’d just had surgery and couldn’t do the monkey work, so I did it all while he sat watching, drinking lots of coffee, all the while telling yarns and talking the usual tradie rubbish. Once I finished each job, he’d hobble over, check and sign off on whatever I’d done. When the water went cold, I had a sudden thought that I might have messed something up wiring-wise and it had failed. While it wasn’t all that likely, given we are now two years down the road from doing all that work, the possibility did cross my mind. With the fuse and wiring apparently siliconchip.com.au OK, there wasn’t much left to check on. What I did next was what I usually do when I need to know something; I phoned my friend, Google. I searched for and found a link to the New Zealand Standards for Storage Water Heaters, but after downloading the file and discovering it was only a preview and that they wanted $61.20 for the actual PDF, I widened my search instead. Fortunately, I found many installation brochures for heaters similar to ours and most included schematics and wiring diagrams for installers. Just what I needed! With my newly-acquired knowledge, I looked at the system again. It seemed to me that if no power was getting to the switchboard, and then on to the heater, there could be something wrong with the ripple-control system, part of which is located in the meter box on the side of the house. In the two years we’ve lived at this address, this was only the second 64 Silicon Chip time I’d opened that box. The meter was changed for a smart meter just as we moved in (and had a wire fall off the pole fuse, as described previously in Serviceman’s Log). Since then, nobody has had cause to disturb the contents. The tone control blues Sitting next to the meter in the box is the ripple control relay. This heavyduty contactor controls when the hot water is switched on, based on a signal superimposed on the mains power waveform. When the signal is received, the relay turns on or off as instructed. Clearly if this was playing up, and failing to switch on at the start of the off-peak period, we’d get no water heating. The trouble was, I couldn’t think of any way to test it other than to check whether there was power to the heater element when there should be power to it – ie, during the off-peak period. But the point of the tone control sysAustralia’s electronics magazine tem is that it can vary from day to day (or more realistically, night to night). Precisely when that should occur on a given evening is anybody’s guess. Ripple control has been used for years in most big cities to prevent the electrical system from being overloaded. If everyone in the neighbourhood turned their water heaters on at the same time, something would blow out at the sub-station. To avoid this, the ability to heat water tanks on-demand was removed from the consumer and instead, households would be assigned certain times that their cylinder would be switched on and off, all controlled by ‘the man’ somewhere out on the grid. In practice, this works very well, but it makes things tricky when an amateur like me wants to test the system. Given that we’re on a night-rate plan, I could stay up all night next with a multimeter across the element, waiting for signs of voltage as the ripple-control instructs it to switch on. Or I could poke about the various terminals on the ripple control hardware in the meter-box during the day to see if (maybe) there was power going in, ready to be switched to the cylinder, or not. However, neither of these options seemed particularly attractive, especially given my reluctance to be electrocuted while playing around with high-voltage hardware, about which I know next to nothing. I closed the meter box and resigned myself to getting my electrician friend to troubleshoot the system for me. This was bad news in many respects, not the least of which was the fact we’d likely not be showering for at least the next day while we waited for the electrician to get his rear end into gear, so it was with some reluctance that I fired up my computer again to find his number. Getting to the bottom of it However, shortly after my machine woke up, an email alert popped up advising me I had a communication from my electricity provider. When I downloaded and read the email, all became clear. The message was an apology, detailing how they’d been having problems with their network and that some households would not have water heating at the usual times, or at all. This was good news, as it meant that siliconchip.com.au all we had to do was wait, and theoretically, it would all just work again – as long as I hadn’t messed anything up while stumbling around in the figurative darkness! The next morning, it was gratifying to feel the hot water back up to temperature. Typical serviceman that I am, I just assumed something had to have failed, or gone catastrophically wrong for the system to have gone down and that I had to get into it to find out why. In the end, all I had to do was nothing. The serviceman’s curse strikes again! BWD Power supply repair J. C., of Murrumbeena, Vic, loves a challenge. He recently acquired two BWD power supplies for free. That sounds cheap but it was because neither of them worked! Here is how he fixed the BWD model 275 power supply... The BWD 275 would switch on but did nothing else; the current and voltage controls didn’t do anything and the output voltage was zero. I took the covers off and had a look for clues of which there were three: one transistor was missing (and all of the others on the circuit board had been replaced), two of the four screws holding the transformer down were missing and the ammeter pointer had been repaired. I found two suitable screws and fixed the transformer securely. The missing transistor was a JFET that the circuit diagram said was a “Selected Component”. It is used as a constant current diode in the +16V section of the auxiliary supply. I made a replacement +16V supply on a piece of Veroboard with an LM317 voltage regulator, two resistors and a trimpot. To fit this, I had to remove two other transistors and two resistors from the main board. Turning on the power supply revealed that the +16V rail measured 0V. Next, I checked the 30VAC output from the transformer to the auxiliary supply. That measured zero too. When I touched one of the wires, it was loose, every strand broken. I stripped it, tinned it and re-attached it and then the power supply started working again. I don’t know why two transformer mounting screws were missing. Maybe it was a botched repair attempt, prompted by the failure of the JFET. There were marks to show that they siliconchip.com.au were originally fitted and without them, the transformer moved a little bit every time the power supply was picked up and put down. This caused the short wires to flex at the circuit board end and eventually break. Because all of the other transistors had been replaced, including two other “Selected component” JFETs, I had to change one resistor to get the Vmax for the 36 and 72V ranges set correctly. At the same time, I replaced the respective trimpots with 10-turn types. The ammeter also required a resistor change to allow accurate adjustment of the two ranges. I had forgotten about the repaired ammeter pointer. It broke again when the pointer went full scale very quickly. I re-glued the broken pointer and changed another resistor and was able to adjust the maximum current ranges such that the pointer won’t break itself again (hopefully). I later noticed that the Vmax settings became intermittent and I suspected that IC1 was the culprit; it contains six transistors arranged as two differential amplifiers. I pulled it out of its socket and cleaned its legs with a glass fibre brush, applied some contact cleaner and re-installed it. The problem went away. This is why I’m not keen on IC sockets; they tend to become intermittent after a few decades. I learned a few things during this repair. The three JFETs in the circuit marked as “Selected components” were chosen at manufacturing time based on the Vgs (gate-source voltage switch-on threshold). Without knowing what this was and the Vgs of the replacement FET, you will have to change some resistor values to get the circuit to work correctly. Also, before setting the Terminal Switch to “SET I”, it’s a good idea to connect a 5-10W power resistor across the output terminals and set the Terminal Switch to “USE”, to verify that the current control does what it’s supposed to do. In the “SET I” position, it connects a 0.1W resistor across the output and if something is wrong, this makes it easy to blow the fuse or even the output transistors. Cold weather Codan X2 HF transceiver fault R. M., of Sydney, NSW, got a bit of a shock as he was driving along when his HF transceiver decided to give him Australia’s electronics magazine January 2019  65 a blast of static for no particular reason. This very annoying fault would have to be addressed so despite being a relative amateur, he decided to have a go at fixing it... I’m not what you would call a technician but from a very young age, I’ve had an interest in electronics. As a licensed amateur radio operator, I have just enough electronics knowledge to do simple repairs to my own equipment. For more complex repairs, I turn to my friend for help, a very experienced technician. So, when a strange intermittent fault developed in my Codan X2 transceiver, I thought I might have a go at tracking it down. The Codan X2 is a 25-year-old 10-channel commercially made 100W HF transceiver which I use with my amateur radio Automatic Packet Reporting System (APRS). This allows my family and friends to keep track of my movements via the internet when I’m travelling. I had to make a few minor modifications to the Codan X2 to interface it with a Byonics Tiny Track 3 APRS kit. When on the move, it turns the X2 on, transmits a position information packet, then switches it off again. This is an excellent arrangement as it conserves my vehicle’s auxiliary battery, which powers other equipment. I have been using this arrangement for about two years. Most of the time it works perfectly, however, the X2 always has had one strange fault that caused it to intermittently break the mute and open the speaker at full volume. As is typical of intermittent faults, it would never appear when the X2 was on the bench. I tried varying the supply voltage, changing the antenna, transmitting, turning it on and off, but they all had no effect. The system would run perfectly for days just sitting on the bench. Months would go by without the fault appearing then suddenly there it was again, the audio at full volume blaring away. Every week I travel from Sydney to the Southern Highlands of NSW to work on my parents’ property. My truck sits outside for a day or two, not being used until my return journey to Sydney. Over time, I noticed that the fault would mainly appear on cold winter mornings when I was leaving the Highlands. At full volume, the noise was so bad that I had to switch the X2 off. 66 Silicon Chip By the time I had travelled 30km and stopped for breakfast, on turning the X2 back on, the fault had cleared itself. This led me to believe that the fault was temperature related. On the suggestion of my technical friend, I tried placing it in the refrigerator for an hour. But even that didn’t activate the fault as, by the time I had it set up on the bench, the unit had warmed up again. Reading the X2 service manual, I was able to locate the TDA1020 audio IC on the circuit diagram. My friend had suggested that whatever was causing the fault would probably be associated with this IC. The IC wasn’t difficult to find as the audio board is mounted inside the face panel of the unit. It was relatively easy to remove the front panel to expose the board and the unit could be run with access to the IC. So I sprayed the IC with a can of freezer spray to see what would happen. This produced a low-level hiss from the speaker but it did not cause the aforementioned fault to occur. Methodically spraying components associated with the IC finally reproduced the fault. When I sprayed a 10µF electrolytic bypass capacitor, the mute suddenly opened with the speaker at full volume, which tapered off as it warmed back up again. The audio IC is riveted to a small block of aluminium that serves as a heatsink, which in turn is riveted to the chassis. Drilling the rivet out and removing the knobs allowed the audio board to be lifted out of the front panel. Carefully examining the underside of the board revealed nothing unusual except for a bit of corrosion around the pins of the audio IC. I removed the corrosion with a toothbrush and some methylated spirits and lightly re-soldered all the pins. Using solder wick, I removed and replaced all four of the ageing electrolytic capacitors on the board. After reassembling, I again sprayed the section of the board which had caused the problem to appear last time. All was silent, as it should be. Cordless vacuum repair B. P., of Dundathu, Qld, had the unfortunate experience of buying a second-hand vacuum cleaner that worked fine when initially tested but then was found to be faulty when he got it home. Luckily, he isn’t afraid of Australia’s electronics magazine disassembling something right back to its constituent parts and he managed to get it going again... After visiting our daughter and using her cordless vacuum cleaner, my wife decided that she would like one of her own. She looked on Gumtree and found one locally, which we went and examined. It seemed to be working OK and we got it for a reasonable price and headed off. When we tested it at home, I noticed several problems with it. For some reason, the three-position switch was now jamming and the brush wasn’t turning. This had not happened when we first looked at the unit, so they must have happened on the way home. I decided to dismantle the unit and try to fix it. I started with the handle, which is attached to the main body by a single screw. The next thing was to dismantle the handle. This was achieved by first pulling off the decorative covering, which revealed several screws, which I then removed. With the handle now apart, I could see that the rotary switch operated a lever which in turn operated a PCBmounted three-position slide switch, which was supposed to be held on by two screws. Somehow, one of the plastic screw holes had disintegrated and the screw had fallen out and jammed the switch. I removed the four broken plastic pieces and the loose screw. Then I screwed the PCB back on with the remaining screw and fortunately, this was enough to hold it securely, so I reassembled the handle, with the switch now working. However, the brush still did not turn when the switch was set appropriately, so I would have to look further to find why. I started by removing the bottom cover of the cleaner head, where the brush was. This involved removing seven screws and this would be routine for cleaning the brush. Everything seemed to be in order here, so with the brush still out, I turned the switch to operate the brush motor. I found that by swivelling the head, I could get the motor to run intermittently, so this indicated a broken wire or loose connection. I then dismantled the cleaning head by removing eight more screws and checked the wires and the motor but they were all OK. I then unplugged the head from the siliconchip.com.au unit and I noticed that there had been some arcing on the two connecting pins, indicating a loose connection. I needed to dismantle the main vacuum cleaner body to find out more. First I un-clipped the decorative front panel and then undid six screws to separate the body into the two halves. I located the clip connectors at the base of the body and tested the fit of the pins, which were very loose. After bending the clip connectors to ensure that they had a firm grip on the pins, I reassembled the main body and refitted the handle assembly. Then I reassembled the cleaning head and the connector, making sure to turn the connecting pins 90° so that the damaged area was no longer the section that would make contact with the connecting clips. With the unit now back together again, it worked correctly, just like new. Smart kettle repair R. S., of Hoppers Crossing, Vic, makes problem-solving and repairing various electronic items his hobby. His neighbour is aware of this hobby and decided to take advantage of his generosity with his time when the household kettle decided to go on the blink, as follows... My neighbour Phil and his wife were having trouble with their kettle, a “bells and whistles” type. The base is used to control the final water temperature in five steps, from 70°C to 100°C. As is typical for this type of kettle, the base has a central metallic post, about 2mm in diameter, with several concentric rings around it. These fit into slots in the base of the jug and supply power to the heating element along with feedback from the built-in temperature sensor. I noted that none of the pretty blue LEDs in the base were illuminated when power was applied but the GPO seemed to be working correctly. The base was held together using some annoying tri-wing screws but luckily I had a suitable bit in a Dick Smith toolkit, so I managed to get it apart without further drama. Inside, I found two separate PCBs. One is best described as the power supply and the other, the control board. They are linked together by a five-way flexible flat cable. The power supply has a two-way flexible flat cable going the main jug connector. siliconchip.com.au The power board was the main suspect, so I removed it and carefully examined it. The circuit is simple; it is the now-familiar capacitor/rectifier/filter/zener type of supply which requires no transformer. The power board also incorporates a 12V DC coil relay to switch 230VAC to the kettle element, plus a 78L05 5V regulator. Both the 5V and 12V power rails are fed to the control board as well, along with control signals for the buzzer and relay and a feed from the temperature sensor. I checked the diodes in the bridge rectifier but they seemed OK. However, the zener diode which regulates the 12V supply was surrounded by some charring on the PCB. I carefully applied 230VAC to the board and measured the voltage across the zener diode. It was only about 2V DC; way too low. So, I disconnected power and desoldered the zener diode from the board. I measured the resistance across it with my DMM and regardless of the way I connected the probes, I got a reading of just a few ohms. So it seems like this component had shorted out, possibly due to overheating, given the charring I noted earlier. I guess it was better that it went short-circuit rather than open-circuit as otherwise, the 78L05 could have had a much higher-than-expected voltage applied to its input and that could have fried it, and possibly other components too. The only 1W zener I had handy was rated at 11V. I figured that was close enough that it should work, so I soldered it to the board and re-applied 230VAC. The 12V rail then came up (to 11V) and I could now measure 5V DC from the output of the 78L05 regulator. I powered it down, re-connected the control board and once again applied mains power. The pretty blue LEDs began to flash and the buzzer beeped. That was a good sign, so I decided to pop down to my local Jaycar to get a 12V zener. I figured I would buy a 5W type, seeing as the 1W zener originally installed seemed to have burned out. The pigtails of the 5W zener are larger in diameter than those of a 1W type; I had to drill out the through-holes to 1.2mm. I mounted it about 7mm proud of the PCB, to allow for better cooling air circulation. I could then finally re-assemble the entire unit and return it to Phil’s wife, Helen. She filled it with water and confirmed that it was back to normal. Now I just need to get Phil to reimburse me the $1.75 that I spent on replacement components! SC At right is the base of the smart kettle with the yellow power PCB before any changes. Below is the power board after the blown zener diode was replaced (marked ZD1 on the PCB). It was replaced with a 12V 5W zener (circled in red) that had slightly larger leads, so the holes had to be enlarged. Australia’s electronics magazine January 2019  67