Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN’S LOG Doing the dirty work When a kitchen appliance fails, among the most dreaded must be the dishwasher. When it stops midcycle and refuses to proceed, you can’t troubleshoot without first removing the dirty dishes and bailing out the greasy, soupy water in the sump. It is tedious and unpleasant, to say the least. The second, more physically demanding chore, is to extract the beast from its cavity under the kitchen bench. At some 50kg, our failed dishwasher is no lightweight, and its German designers never thought to equip it with roller wheels (for some reason). Despite putting up stiff resistance, with a combination of tugging while simultaneously rocking it side to side, it gradually emerged, exposing 10 years’ worth of grime and dust. Now at last I could remove the metal side panels and get to the inner workings. This unit had performed faultlessly for 10 years and had always delivered great results. Its sudden and unexpected failure suggested a problem that might be simple or obvious. At least that’s what I hoped. I decided to run a short cycle to better observe its behaviour leading up to the point where it would stop. It began normally. The drain pump cleared the residual water, then it refilled and began the pre-wash cycle. However, after a few minutes, it just stopped with the time remaining indicator showing zero. It had failed to progress to the main wash, which was baffling. The water was getting to where it should be, the circulation motor was running and the water was pumping out during the initial drain. All conditions necessary to proceed seemed to have been satisfied, so why did it stop prematurely? 84 Silicon Chip I hoped that a Google search might help me locate some service information or a schematic, but the corporate world protects its secrets. However, I found an abundance of YouTube videos pointing toward the usual suspects being blockages, pump failures, or a failed heating element. These all checked out OK on our machine. The heating element is part of the main pump unit. Being connected to the control unit by heavy-duty wiring makes it easy to identify for the purpose of checking the resistance. This was as it should be, at around 20W. Having eliminated the prime suspects, the remaining possibilities seemed to be that the control module itself may have failed, or perhaps a malfunctioning sensor could have confused the control module, leading to a shutdown. I removed the control module for close inspection, but it looked pristine with no components damaged or burnt. From past experience, I knew that a sensor that failed to return the expected signal could cause a dishwasher to stop mid-cycle. Years ago, many dishwashers (and washing machines for that matter) had a motor-driven switching mechanism. The motor advanced the mechanism, and a large knob on the front of the dishwasher rotated accordingly to indicate the progress through the program. I had a dishwasher like that; it had both cold and a hot tap connections. The main wash cycle used only hot water. When the hot water solenoid went open-circuit, the motorised switch would be paused, waiting for a signal from the water level sensor. With a failed solenoid, the required water level would not be reached and, in the absence of a signal, the motor would not be powered on. At that point, the program would be abruptly halted. The result was not unlike what I was experiencing with the current unit. Motorised switches have long been replaced with microprocessor electronics, enabling more advanced functions, including the reporting of error codes. Unfortunately, I had no error codes for guidance and no service information that might have given a clue about how many sensors there were or what functions they performed. Australia's electronics magazine siliconchip.com.au I was on the verge of giving up. Judging by the wiring that snaked throughout the machine, most of the sensors appeared to be well buried in the bowels of the device. However, there was one sensor that stood out. Most of the space on one side of the dishwasher is taken up with a large, translucent plastic box containing an intricate labyrinth of water galleries. I’m guessing that its purpose is to store and regulate the inflow of water, and it may also save energy by warming the stored water using heat released during the wash cycle. Mounted in a recess on the plastic box is one very obvious sensor that is easily accessible, shown in the accompanying photo. I decided it was worthy of closer examination as a last resort. It appears to be a flow meter, which lives in the water inlet path. A small, bladed impeller with an embedded magnet rotates with the flow of incoming water. With a torch, I could faintly see it spin inside its translucent housing. Sitting outside in a recess was a reed switch, a glass capsule with metallic contacts that should close each time the impeller magnet passes. I can only speculate about the purpose of this sensor. It is obviously able to inform the controller about the flow rate and volume of water entering. Maybe it’s a safety device. Perhaps a runaway count might signal an overflow of water, prompting a shutdown. Alternatively, if the impeller seized or the reed switch failed to register, the controller might be programmed to halt the process due to failure of the device for safety reasons or because insufficient water had been received. The reed switch was mounted on a small circuit board and was easily removed for testing using a simple magnet and multimeter. The contacts inside the glass envelope would close as they should when the magnet came near. I could not fault it, but decided to reinstall it anyway. I ran the short cycle again. This time, the cycle progressed properly, and the machine ran for the full duration of the main wash but then stopped, showing an E14 error code for the first time. It had failed to perform the final, critical drain. It wasn’t a complete cure, but it was pleasing progress nevertheless. siliconchip.com.au A quick consultation with the internet confirmed that E14 was indeed associated with a flow meter fault or a problem with the water intake. Could it be that the reed switch had aged and become unreliable? I have had some previous experience with reed switch faults. In the 1980s, new telephone exchange equipment was installed that employed reed switches. Error reports showed that some reeds mounted on circuit cards were prone to sticking. To prove the fault, the cards needed to be very gently removed and the suspect reeds checked with a multimeter. Sure enough, certain reeds on the board were sticking with contacts closed. The gentlest tap on the card was enough to cause the contacts to open with an audible click. A quick trip to Jaycar and I obtained a visually identical reed switch for less than the cost of a cup of coffee. I soldered the new one in place of the old one on its circuit board and reinstalled it. I pressed Start. Success! The dishwasher ran perfectly, advancing through every stage, including the final drain. Incredibly, an expensive dishwasher had been brought to its knees by quite possibly the cheapest component in the entire device. Alan Preacher, Briar Hill, Vic. The red arrow points to the recessed sensor located in a plastic housing. This sensor is likely a flow meter. Australia's electronics magazine March 2026  85 Dredge boat radio repair In the distant past, I was married to a Sydney girl, but we lived in Newcastle. One weekend, we were visiting the relos in Sydney when the brother-in-law, an electrician, stated he was on call for the port of Botany Bay for the weekend, and had a call out for the dredge that had twoway radio problems. He said, you know more about electronics than I do; can you come with me? We went to the dock, and a tug was waiting to take us to the dredge. When onboard and taken to the radio room, we found that the radio was relaying all calls received and making an echo on the network. I picked up the mic and asked the harbour master for a radio check. He told me that the problem was still there. I then realised that I had not heard the clunk of the relay switching to transmit mode. We opened the box (about half a meter square) and found the relay. It was plugged into the circuit board. The points were fused together. We managed to separate them with a screwdriver. We found the onboard chippy and got some sandpaper from him to smooth them out. That got it working again, for the time being. We also asked the chippy to tell the electrician, on return from his break, to order a new relay. They fed us dinner and gave us some beer. As my brother-­ in-law was on call, he was not allowed to drink, but I was, so that was my pay for doing the job! We had to wait about two hours for the tug to come back. Mick Toomey, Newcastle, NSW. Breville microwave repair My wife told me that there was something wrong with our microwave, as it sounded different from usual. I could tell from how it sounded that the cooling fan had stopped working. I was hoping that it was just the fan motor, as that would be an easy fix as long as I could find a replacement part. We found this microwave at one of the local tip shops about four years ago. It was in almost-new condition but needed a good clean as it had obviously been stored for an extended period. Until now, it had been very reliable. I took the microwave to my workshop and removed the six screws that hold the cover on. The fan is located in the back right-hand corner. I disconnected the two wires going to it and got out my multimeter to check the resistance of the winding. The winding was open-circuit, so that explained why the fan no longer worked. The next question was whether I could find a replacement fan motor. I started by unscrewing the circuit board on top of the fan housing, then I removed the screw from the back of the microwave that was holding the fan motor housing in place. To remove the fan motor housing from the microwave, I had to first remove the magnetron, as it was stopping the fan motor housing from tilting forward enough to remove it. An eBay search for a fan motor to suit this model of the microwave was not successful, although I did see one or two listings. I changed my search to the part number of the motor, and I found a couple more listings from China, but when I switched from default to Australia only, there were none. I changed my search to worldwide and set the search to price plus postage, lowest first. This showed many listings for this part, and it also showed that this exact fan motor is used in a multitude of different brand microwaves. I was able to purchase a replacement fan motor for $15.27 with free postage, but I would now have to wait for it to arrive from China, which could take up to four weeks. In the meantime, my wife found another microwave at the local tip shop for $10 and it was still in very good condition. It had been tested before she purchased it to make sure that it worked. Amazingly, the new fan motor arrived in just 11 days from China. It was obviously an after-market replacement, as it did not have the part number on the side like the original motor. The next problem was that I was unable to remove the fan from the old motor. However, I was able to change the rotor, with the fan blade attached to it, into the new stator. The two motors were almost exactly identical, enabling me to swap the parts, and the rebuilt motor worked as expected. I reassembled the microwave and tested it by putting a cup of water in it and running it. I could hear that the fan was running, and the microwave sounded the same as it did before the fan failed. Good-sized new microwaves cost around $150-300, so being able to repair this one for less than $20 was well worthwhile. The spare microwave for just $10 was a bonus. The interior of the Breville microwave (left), fan motor (above) and the repaired device (right). 86 Silicon Chip Australia's electronics magazine siliconchip.com.au WARNING: Microwave ovens contain very high voltages that are extremely dangerous. A microwave can retain these high voltages even after it has been turned off and unplugged, and even a dead microwave can kill you, as this high voltage may not dissipate for a long time in some circumstances. So if you are not experienced in repairing microwave ovens, do not remove the cover. Bruce Pierson, Dundathu, Qld. Turntable inverter repair This story has a lesson about buying semiconductors from online vendors. I recently built the turntable inverter from the May 2016 issue (siliconchip.au/Article/9930). Everything went well with the assembly, and I decided to benchtest it before fitting it into its diecast box. This meant that the Mosfets were not yet heatsinked. I connected a 12V car test lamp, drawing approximately 180mA at 12V, across the transformer terminals on the circuit board, with an oscilloscope probe connected as well. I connected a linear power supply set at 14V DC. To my disappointment, the test lamp flickered randomly, and within about 10 seconds, there was a burning smell. I quickly switched it off and found that the IRF9540 Mosfets were stinking hot. Surprisingly, the IRF540 Mosfets were at about room temperature. Troubleshooting was going to be difficult as I had to connect scope probes to the circuit, switch it on, quickly make measurements, then switch off. I tried tracing waveforms using this tedious procedure. It was getting ridiculous, so I removed IC3 from its socket, then bent pins 1 and 7 horizontally so they would be disconnected when I plugged IC3 back in. I was now able to trace all the waveforms up to the inputs of IC3. They were all correct 50Hz sinewaves (I had set the inverter to 50Hz mode). The output pins of IC3 (1 and 7) were putting out square waves, but they were not driving the transistor section, so I desoldered all four Mosfets, bent the pins of IC3 back to their original positions and powered it up again. I could detect sinewaves at the bases of Q5, Q6, Q7 and Q8. I powered it down and performed continuity checks to verify that each component in the circuit was connected correctly. I couldn’t find any faults; all components tested OK and were in the correct places. To make troubleshooting easier, I soldered header sockets siliconchip.com.au Items Covered This Month • A prematurely stopping dishwasher • Dredging up a boat radio • Breville microwave repair • Finding the culprit in a turntable inverter • Following the breadcrumb trail • Repairing a NAD 701 stereo receiver 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 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. in the Mosfet positions so I could easily swap them over. I put in all new Mosfets and powered it up. The same fault appeared; the Mosfets got hot again quickly. I then tried bypassing IC3 by removing it from its socket and using tinned copper wire. I bridged IC3b pins 5 to 7 and IC3a pins 3 to 1. This produced a square wave at the transformer terminals, and the Mosfets ran cool. I suspected that IC3 could be faulty and replaced it with the one spare I had on hand, with no improvement. As the pinout of IC3 is the same as an LM358, I tried that again with no success. I then noticed that the output waveform to the transformer terminals was a half sinewave, but it had oscillations superimposed on it at about 15kHz. If I powered the circuit off and then on again, the circuit would start up without oscillations, and the Mosfets would run cool with a clean sinewave at the transformer terminals. As soon as I put a scope probe on the output or anywhere in the circuit, the oscillations would start and the Mosfets would again overheat. Everything checked out up to IC3, the Mosfets had been replaced and I could find no faults in the driver circuit, so I purchased a couple of LMC6482AIN op amps from Jaycar (to replace IC3) and that was it. The inverter sprang to life, and I could not get it to oscillate anymore. I have purchased lots of components online over the years and had good luck with only one dodgy purchase (previous to this one) in that time. I have noticed that there are a lot of YouTube videos online regarding testing for dud op amps, Mosfets etc. Online bargains could be duds, so you should put the device through its paces and ensure that the test results agree with the device datasheet specifications before installing it. This is time-consuming but can save you a lot of work and time later. Australia's electronics magazine March 2026  87 In spite of all the frustration, I did learn a lot about this circuit, and it was very satisfying to be able to finally nail the culprit. Geoff Coppa, Toormina, NSW. Russell Hobbs toaster repair We have had a Russell Hobbs four-slice toaster for some years. While a little bulky for the benchtop, its saving grace is its capacity to toast all four slices evenly on both sides at the same time. Previous toasters had achieved only various patterns of brown and tan, sometimes black, so when the Russell Hobbs ceased to function, it was a sad day indeed. While one side of the toaster still functioned, the other, much more frequently used side (the one closest to reach!) refused to light up. It had started to work erratically a few days leading up to the final failure. There is a history in the house of repairing white goods and appliances instead of replacing them, so the toaster duly made its way to the workbench. Taking the top cover off, the workbench and the technician was quickly covered in crumbs – quite an amazing amount, really. The internals of the plastic chassis included the toasting chamber with the heater elements and components of the two bread carriages, including springs and bread racks. In front of the chamber, a set of contacts for each of the carriages is connected to the mains supply. These close when the bread carriage lever is depressed. Beside the contacts, a release solenoid, visible at the top of the photo (shown below), holds the carriage down when energised until the desired level of toastiness has been achieved, then releases the carriage. On each side of the chamber, a small circuit board receives mains from the carriage contacts. It has various components, including an DPST relay that energises the carriage solenoid. This board also supplies 12V DC to, and is controlled by, a timing circuit board mounted in the top cover. The timing board has various functions, such as defrost and warming, and energises the relay. At the top centre of this photo, you can see just a bit of the release solenoid, while below it is the 12V DC DPST relay. 88 Silicon Chip Once all the breadcrumbs were cleaned up, the diagnosis of the fault was quick. On the side that wasn’t working, the DPST relay had lost a large portion of its plastic cover due to contact arcing, judging by the look of the relay contacts. As with most appliances these days, they are not made to be repaired, and extracting the circuit board without breaking all the plastic fastening tabs was a mission in itself. Once removed, further examination revealed that the board had got rather hot underneath the bridge rectifier, which was mounted flush on the board. The rectifier failed testing and was replaced but elevated off the board. The electrolytic cap also tested bad and was replaced. All other components tested OK. The DPST relay was not so easy to replace due to its size and pin layout. A replacement was found from one of the major component suppliers, but with a rather eye-watering cost once postage was included. The parts cost less than a new toaster, but not by much! After testing the other circuit board, replacing the bridge rectifier and capacitor, two replacement relays were ordered – might as well replace both. Delivery was prompt, and the new relays were installed. The opportunity to check the mechanical operation of the toaster was also taken, cleaning the carriage contacts, straightening some bent components of the bread carriage, and general de-crumbing. After reassembly and testing, both sides were found to be working correctly. All up, the time to repair the toaster was around two hours. Still, there is the satisfaction of keeping an appliance in service and not going into the hard waste collection. Richard Dilena, Ocean Grove, Vic. NAD 701 stereo receiver repair The ad stated, “NAD 701 parts only untested with the display not working”. For $50, it had to be worth a try. In the worst case, replacement displays are available, and swapping them can’t be that hard. So I bought it, opened up the case and looked at the display PCB. The display is backlit, and one of the incandescent lights had failed. The power to the lamp was 12V DC, so I replaced the 4.7W current-limiting resistor with a 560W value and inserted a high-intensity green LED in place of the incandescents (see the photo opposite). That got the display working. The power amplifier voltages are regulated (along with most supply rails in this amplifier) and they were all within tolerance. I checked out the amplifier by playing a CD, and the amplifier was working with no apparent distortion. I also tested the phono preamp, which was also functional. Next, I tested the tuner. It is built around three ICs and an FM tuner front-end from Mitsui. I could change the AM & FM tuning frequency on the display, but there was no change in the sound. It was like the oscillator was not working. In this receiver, AM and FM are tuned with varactor diodes powered from an LM7000 IC through a Darlington transistor pair amplifier. The LM7000 has an onboard oscillator with an external 7.2MHz crystal followed by a frequency divider of 145 times for AM and 1007 times for FM. There is a second divider programmed from the display output that is fed by the AM or FM voltage-controlled oscillator (VCO). Australia's electronics magazine siliconchip.com.au Shown above is the circuit diagram for the tuner and the replacement I made for the backlight, using a high-intensity green LED and 560W series resistor. Below is the front panel of the NAD 701 stereo receiver and a close-up of the tuner section of the board showing some of the adjustment points. The frequency and phase difference between the two divider outputs is compared and converted to a voltage proportional to the phase and frequency difference. The amplified voltage determines the VCO frequency; as the comparator ramps the voltage up (or down), the AM (or FM) oscillator frequency changes until they match. Test point 1 (TP1) is the voltage applied to the VCO (AM or FM) and it was stuck at 40V. Pin 17 of the IC (charge pump output) was also fixed, so the transistors were likely alright. An oscilloscope on pin 20 of the IC showed no oscillator output. Adjusting the small variable capacitor in the crystal circuitry made no difference, so I replaced the LM7000. Once the decision to remove an IC is made, avoiding siliconchip.com.au damage to the PCB is the highest priority. So I cut every leg of the IC near the chip and desoldered them individually. I then soldered a socket in place and inserted a replacement IC. Adjusting the trimmer capacitor brought the circuit into oscillation. The AM tuner then worked well, but there was no change in the FM behaviour. The supply to the FM frontend module is 12V through an inductor/capacitor RF filter. There was 12V on the supply side of the 2.2μH inductor but not the tuner side. Replacing it brought the FM tuner to life as well. The story could have ended very differently, but in this case, it was $50 well spent. SC Jim Greig, Mount Helen, Vic. Australia's electronics magazine March 2026  89