Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN’S LOG Tips on kits and bits Dave Thompson I’ve fixed so many faulty kits that I now have a pretty good idea of the pitfalls of kit and PCB assembly. Often, the fix is quite simple once I’ve spent a while poring over the board and located the fault, but it’s so much easier if you don’t make a mistake in the first place. So pull up a chair, dear reader, and let Uncle Dave tell you all about the ins, outs, dos and don’ts of PCB and kit assembly. I’ve been building electronic kits and projects since I was eight years old. How do I remember this age so precisely? Because dad, on one of his many travels, bought my brother and me what was then called a 10-in-1 electronics ‘Lab Kit’. These are still sold, with larger 50-in-1 and 100in-1 versions also available. This was the late 1960s, though, and that lab kit was my first real introduction to electronics as a hobby. It enabled me to clip in components and make a simple amplifier, oscillator, lamp flasher and similar projects. I was already an inquisitive child and soaked up as much knowledge as I could. Luckily, dad was doing a wide range of engineering, electrical and electronic jobs, and I often tagged along for the ride. I wasn’t always up with the play, though; for some time, I couldn’t figure out how noise came from a radio or a TV. Like many kids, I assumed there was someone in there somehow. Silly, I know! On my seventh birthday, I was given an eight-transistor radio. I wish I had it now, but it is long gone. I have similar models in my ‘collection’, but not the original one. At the time, I recall promptly pulling it apart to see how it worked. What I saw inside didn’t really clue me in much – but I could see that there were no tiny people in it! In that case, dad had to put it all back together because, like all good servicemen, I am better at taking stuff apart than I am at putting it back together. I worked for years to gain the skills required to put things back together again; it takes even longer if I wanted them to still work afterwards! Items Covered This Month • • • • • • The pitfalls of kit and PCB assembly Louvre rain sensor repair A dual-purpose intercom and ant colony unlocker Converting a torch to use Li-ion cells Repairing a Miele clothes dryer Three blind mice and an aircon 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 84 Silicon Chip An inauspicious start to a career, then. But that’s how I learned to do things; by actually doing, making mistakes, rectifying them, and then making some more. I make no claims of expertise, or even being the best electronics guy on my street; I just read and learned as much as possible from those around me. More projects than I’ve had meat pies In the intervening years, I have made literally thousands of projects – some from scratch and some from kits. Most worked straight out of the box, often because the project came with a PCB layout that could be replicated by the home constructor. In a commercial kit, the designer or kit manufacturer has already done the heavy lifting. We hope that most of the bugs and errors have been ironed out long before a kit is released. In theory, it should be as easy as ABC. While this is typically the case, as anybody who has ever purchased and built kits will tell you, that isn’t always how it works out. It stands to reason that the more complicated the circuit and the project, the more chances a constructor will do something wrong when assembling it, or when configuring it after the build. In many cases, this just means the project won’t work as intended. Still, in some cases (for example, in mains-powered projects), this can be a spectacular showstopper, and especially catastrophic if the PCB gets fried in the process. I’ve also built a lot of kits for many people over the years. It makes sense for someone who wants the device but doesn’t have the confidence to undertake the project. It’s easier to ‘farm’ the build out to someone who is proficient with a soldering iron and already has the mechanical and practical skills necessary to put it together. That said, guys being blokes, many of us take on projects that are obviously above our pay grade. It is no surprise that some of them just don’t work when they come out the other end. As a relatively experienced constructor, if I might be so bold, I’d like to offer some advice for people who might want to take on any of the projects featured within these hallowed pages. One thing to note is you usually don’t need to know much about electronics to build a well-produced kit and have it work. Plenty of people I know – with no previous experience – have built some of the many and varied Australia's electronics magazine siliconchip.com.au valve-based guitar amplifier kits marketed on the web, and they worked out very well. This is a popular way to get into a tube amp without the much-bigger price tag of a commercially produced amplifier. Most of these kits are time-tested and come with excellent documentation, support videos and other resources to ensure the build goes smoothly. Obviously, this isn’t the case for many projects, and it’s a matter of being ‘on our own’ if we decide to try building something a bit more obscure (or even from scratch). Community support can sometimes be available through the circuit designer, kit manufacturer, or even other enthusiasts, but it isn’t always guaranteed. Previous knowledge of electronics is not always a prerequisite. Studiously study soldering for success One skill you do need, however, is to be able to solder properly. Solder is typically how electronic parts are connected to printed circuit boards, terminals and to each other, so it stands to reason that this is a necessary skill constructors must at least be competent at before building anything electronic. Many of the problems I find when given a non-working project to troubleshoot are down to poor soldering, so this is something that shouldn’t be taken lightly, especially with major kit builds like large amplifiers and mains-­ powered devices. There are many tutorials available (including in this magazine) on how to solder correctly, so I won’t go into too many details here, except to say that if someone doesn’t know how to solder, they should learn to do so before starting any kit build. Obviously, there are a range of skill levels regarding soldering; if you are looking at making something that doesn’t require soldering a 100-pin surface-mount component, then there is no burning need (pun not intended) to learn that particular skill. However, people should learn enough to do their proposed job properly. I would say that 25% of the dead projects I get to troubleshoot have simple soldering mistakes. Get yourself a decent iron My number one top tip is to get a decent soldering iron. Using dad’s old plumber’s iron is inappropriate for this kind of work. A decent soldering iron fit for purpose is siliconchip.com.au Australia's electronics magazine April 2023  85 relatively cheap at the likes of Jaycar or Altronics. While it doesn’t have to be anything über fancy, like an expensive soldering station, it is a worthy investment to buy the best one you can afford before getting stuck in. Something in the 25-30W range, with a medium-to-fine tip, is ideal. Some have replaceable tips as well, which can make the iron a lot more versatile; it’s a good idea to have a larger tip, such as a chisel or screwdriver style, on hand in case you need it. Surprisingly, large flat tips can make soldering fine-pitched SMDs easier than the needle-like tips some people think you need for that job. Still, there are times when having a very fine tip is helpful, so you could probably justify having four or so tips to start with: fine, medium, large and flat-edged. Other stuff you’ll need While you’re at it, buy a proper cleaning sponge; never take a file or abrasives to a soldering tip – just a damp sponge will keep the tip in excellent condition. If it gets pitted or wears out (which it will over time), simply replace it (if you can). Keep the tips tinned with solder when you aren’t using them; it helps prevent oxidation. Once you have an iron sorted, the next requirement is some decent solder. The old lead bar granddad left in the shed for fixing a blown copper boiler is obviously not suitable for fine work like this, nor are some of the acidbased flux solders used in golden olden times. If you find a reel of this in the shed, I’d avoid using it on your electronics projects, as the acids in it can corrode PCB tracks and component legs. The best thing is to buy fresh solder while you are down at the store buying a soldering iron. The standard hobby solder available these days is lead-free, which is the best option. Made from copper and tin, with a rosin flux core, a small reel is not expensive, and having a reel in the workshop is always very handy anyway. I use two sizes: 0.5mm diameter for finer work and 0.71mm for larger jobs. While I’m on the subject, do yourself a favour and get a small (or even large) syringe of proper flux paste while you’re at it. Don’t use liquid flux, as it’s only suitable for specific jobs; thicker flux paste can be a real lifesaver, making seemingly impossible tasks possible, especially when working with tiny SMDs. Good soldering is critical because it doesn’t take much in many of today’s designs, kits and projects to cause a device to stop working because of a poorly soldered, high-­ resistance joint. My first port of call in any troubleshooting scenario is to go over all soldered joints one by one with a jeweller’s loupe, or in some cases, a USB microscope. It’s a painstaking job, but one that can nip a potential nightmare in the bud if the rogue joint is spotted earlier rather than later. Generally, if the soldering is good, I don’t have high hopes a dud joint will be the cause of the fault. Still, if the soldering overall is looking a bit dodgy, this is a likely place to find the problem. So it really does pay to learn to solder well before taking on any electronics project. The second most common problem I encounter is components inserted incorrectly. This is such a basic mistake, but even experienced constructors (me included) can put things in backwards. Diodes, electrolytic capacitors, transistors and ICs of all types are the most commonly misplaced components. After checking the soldering, my next step is to check component placement. If the soldering looks pretty good anyway, I might skip straight to this step. This part of the troubleshooting process is much easier if we have a circuit diagram, a PCB layout map and component designations screen-printed on the board itself. Sometimes, this information is not available, but the more information we have, the easier it will be to find the source of the problem. If all the information we have lines up and agrees with each other, the project’s eventual success should be just a matter of assembling it with good solder joints, then checking it and plugging it in to try it. Of course, Murphy and Sod are always testing us. It might be you get a dead component from the factory, or the PCB you are using has a fault in it (multi-layer boards can often have, or develop faults that are invisible to even the keenest eye). Also, many times, I’ve fired something up after building it, and despite checking and re-checking, I find that I have misinterpreted something in the instructions, or installed something backwards only to see the magic smoke coming out. With care, however, assembling a project and getting it working should be fun and rewarding. Attention to detail required When constructing any board assembly, I start with the components that lie flat first, like resistors, diodes and any SMD components. I like to arrange all the resistors with the colour bands facing the same way. This is not some obsessive-­compulsive disorder on my part; I’m just being tidy. It also helps if I (or someone else) need to troubleshoot the board later; constantly flipping it around to check colour bands or read part designations gets tiresome very quickly. 86 Silicon Chip Australia's electronics magazine siliconchip.com.au Editor’s note: it also pays to check resistor values with a DMM. It can be hard to tell black from brown, brown from red, red from orange and grey from white, especially if your lighting isn’t ideal. Once again, there are many tutorials out there on soldering SMD parts, but that is beyond the scope of this article. Needless to say, constructors should check very closely for solder bridges and connections that haven’t been made once they’ve completed soldering in all the SMDs. A decent, lighted magnifying glass or a good jeweller’s loupe will make this task a whole lot easier. Typically, in any reputable kit of parts, the PCB will have a screen print of the component layout depicted on it, which may also include the circuit diagram’s parts references or even the parts’ type numbers or values. This is usually foolproof, because everything should have been carefully worked out beforehand, thus avoiding potential errors. However, there are traps for younger players. Editor’s note: if building one of our PCBs, check the overlay diagram published in the magazine. Sometimes, changes to the PCB silkscreen can be missed after the prototyping stage, and values that were since changed might still be printed there. The overlay in the magazine is usually final and should have all the correct information. Transistors will usually be depicted asymmetrically, indicating they should only be fitted one way. This is all well and good, but it can get confusing if substitute components are used due to supply problems or expense. Kit manufacturers often swap out different types, but usually mention it in any documentation. Some even add a note in the bag with the parts. Pinouts are not always universal among different transistor types. It pays to check that the component you are soldering in has the same lead designation as any original part quoted. Many projects I’ve repaired over the years have had substituted components installed, and as these were inserted as per the instructions and PCB overlay, the project didn’t work. They’d had a different lead configuration. Data sheets for almost every component on the planet are available with a quick Google search, so it doesn’t take much effort and research to make sure you put things in the right way around. This is especially true for many of today’s multi-layer PCB projects; it’s a lot easier soldering something into these PCBs than getting them back out again! Putting components in backward has been a staple error of constructors since project building began. The mantra is to check, double-check, then triple-check before you solder anything in. This tip alone will save a lot of grief and hand-wringing out the other end. Another problem worthy of inclusion here is when working with wound inductors or transformers using enamelled copper wire. This wire is insulated with a very durable coating – it might not actually be enamel anymore, but the theory is the same. This wire is insulated to prevent shorts and flashovers in coils and transformers, so it is quite a thick coating by design. It can also withstand flexing and bending (to a certain extent) without cracking or failing. However, it is not designed to be soldered, and a standard soldering iron will not melt the material, no matter how long you hold the iron on it. This coating must be completely removed, exposing the siliconchip.com.au bare copper wire beneath, before a decent solder joint can be made. I have ‘fixed’ many a project using self-wound inductors where this enamel removal has not been done at all. This means there is no electrical connection between the inductor and the rest of the circuit. Kits and projects usually have specific instructions on the requirements to do this enamel removal, but some constructors don’t get the memo. I’ve found that taking this insulated coating off is best done very carefully with a sharp knife (like a ‘Stanley’ knife or box cutter). Yes, I know people will be eye-rolling and saying they have a better method, but for me, a sharp blade is my go-to tool. Some use sandpaper, or worse, try to ‘burn’ it off with a lighter or blowtorch; this is inefficient and messy, and often leaves soot all over the wire. Careful scraping is the only way to leave a decent, clean wire underneath, ready for soldering. Being too aggressive with the knife could also cut the soft copper wire, so like any task, care and finesse make the difference. With patience and care, even the most complex projects can be constructed and work the first time. By all means, ask questions where possible, and above all, have fun with electronics! Louvre rain sensor repair J. W., of Hillarys, WA is at it again. This time, the louvres on his house were playing up, and it turned out to be some of the usual suspects (but not faulty capacitors for once)... In 2003, I installed a Vergola Louvre Roof System across the rear of my house, which has a North orientation, to let the winter sun in and keep the summer sun out. The system has six separate banks of louvres with a Linak linear actuator for each bank, a rain sensor that shuts the louvres when it rains, an indoor control panel with six buttons and a 7-segment display. You can access each bank by cycling through the number on the 7-segment display and then pressing buttons to open or close it. With cooler weather upon us, it was time to let the sun in and warm the house, but when I activated it, the rain sensor always shut the louvres even though it was not raining. Australia's electronics magazine April 2023  87 I tagged all the wires and took a photo with my phone to ensure I got all the wires back in the correct positions. After disconnecting all the wires, I gave the top a good clean. After finishing the reassembly, I turned the power back on and waited out the required 15-minute delay. I was pleased to see the unserviceable condition go away, and the system worked normally. So now the sun can warm the house again, with the panels closing when it rains. Intercom and ant colony unlocker The louvre control box is shown above, with the rain sensor shown adjacent. The rain sensor consists of two stainless steel combs with teeth that mesh into each other, leaving about a 1mm gap so that a drop of rain will bridge the gap and cause a change in resistance from an open circuit to a few megohms. The control box senses this change and shuts the louvres until 15 minutes after the rain stops. I put the ladder up and examined the sensor and cable, which looked the worse for wear after sitting in the sun for all those years. I decided to remove the sensor and refurbish it with new silicone sealant and paint for the base. I managed to pull an extra 30cm of cable from under the tiles, so I cut off the sun-damaged section. After a final test to see that the sensor showed infinite resistance, I put it all back together and turned on the power. The control panel 7-segment display showed a flashing U for unserviceable. This is normal after a power loss so that if it’s raining, it won’t cause the louvres to open. If the sensor is still dry after the 15-minute delay, the louvres will cycle to open and then close. I waited the required 15 minutes and still had the unserviceable indication. After a further 15 minutes, I decided to disconnect the rain sensor and try again. The unserviceable condition persisted. The next step was to find the control box under the roof tiles in the eaves. After pulling several roof tiles back, I found a large Jiffy box with 16 4mm banana binding posts on the top and a mains transformer. The six linear actuators are connected to 12 high-current binding posts with the reset switch, and the rain sensor connected to four smaller posts. I could see what the problem was straight away. There was 17 years of dust and detritus build-up on the top of the Jiffy box, which looked damp. The separate power transformer must have been providing warmth to some rats by the number of droppings around it. So the dampness was probably rat urine, causing a low enough resistance across the metal base of the binding posts to simulate rain. 88 Silicon Chip P. B. E., of Heathcote, Vic thought he had an easy job as it was ‘probably just’ a dry joint. It turned out to be a few different things, including some unwanted guests... I volunteered to ‘have a look’ at a Fermax intercom and door unlocker. The intercom part worked most of the time, but the unlocker hadn’t worked for years. Intermittent faults are always a bigger problem than simply not working. However, it usually means there is a dry joint or broken wire. I was hoping for an easy fix along those lines. The unit was installed at a property in Melbourne, so I got the whole thing out: master, slave and door strike. That way, I could take it back to the workbench and look at it closer. I left the transformer behind as I knew it was working and it would be easy to supply 12V at home. There were five coloured wires from master to slave and two to the door strike. Strangely, the wire that was used was six-core, similar to alarm wire. I thought five-core trailer wire would have been better. This caused me some confusion as the yellow wire was connected to the slave but not to the master. I tried to get the schematic from Fermax, but it was a dead end. I then spent far too much time trying to find a PDF with the circuit diagram. After about an hour, I managed to find a manual for a similar unit from an intercom place in America. I downloaded the PDF manual and printed the page I needed. It was only then that I found that the yellow wire did nothing. On removing the master unit, I realised it was full of ants. I’m sure they didn’t help the situation. I didn’t have any insecticide, so I sprayed the unit with WD40 – that’s for water displacement, not insect displacement! – Editor. Alas, it turned out that the intercom runs on 12V AC, not DC as I’d assumed. I didn’t have a 12V AC supply, but I did have an old Triang model train transformer that put out 15V AC – close enough. I wired it up on the bench using the same colour codes. I got nothing, not even intermittent operation anymore. Oh dear. It was time to pull this thing apart as far as I dare and clean it. That turned out to be surprisingly easy. It was held together by just two screws and four clips, and once open, the PCB came right out. The speaker was connected with flying leads, so I desoldered them. I cleaned the speaker gently with metho. Knowing the PCB had been subjected to ant acid, I dipped it in a very weak caustic soda solution and washed away all the gunk with a long soft paintbrush. Then I gave it a quick metho bath and dried it using compressed air. I left it in the sun to dry properly. It was time for a coffee! On inspecting the printed side of the PCB, I found what I thought would be the problem, a dry joint. There were a few other joints that, in my opinion, were bad, so they got the resoldering treatment too. On testing the nameplate light, it was blown. A new 12V 5W festoon ‘trailer’ globe Australia's electronics magazine siliconchip.com.au fixed that. I’m not too fond of these festoon globes, but that’s how it’s designed. I then reassembled and wired it back up on the bench. The call button didn’t work very well, so I took it apart and cleaned it again, bending the two metal prongs to make better contact. It then all worked well. I only had to reinstall it in the client’s house back in Melbourne. Easy. After I reinstalled it, no go again. This time, the problem had to be the wiring in the house or underground. I guessed it would be in the hardest location to fix, underground! With a simple multimeter test, I discovered the wire from the master unit to the striker was open-circuit. After digging for only 10 minutes, I found a join in the conduit that I didn’t like. When I took it apart, I found it was full of ants and dirt. The wire was corroded at a three-way join in the conduit (never join wire underground). A new two-core wire had the unit working again. I’d spent about eight hours on this ‘simple’ fix. However, I got more satisfaction from it than many others I’ve done, probably because there were four separate faults. Another success! Converting a torch to use lithium-ion cells B. P., of Dundathu, Qld discovered that it’s pretty easy to convert some torches from using three disposable cells to a single rechargeable lithium-ion cell… Small pocket torches that take three AAA cells are very common. We have several at home, and I always carry one in my pocket. But I was getting a bit sick of replacing the AAA cells. Also, these torches can get a bit touchy with all the connections for the cells and the cell holder. There are eight different connection points; one on each end of each cell and one on each end of the cell holder. Sometimes you have to give the torch a bit of a jiggle before everything makes contact and works. I thought that there must be a better way! I was recently working with 18650 cells and realised that an 18650 cell should be able to power one of these small torches. The only problem is that they are too long to fit inside the torch. I needed a shorter 18650 cell, so I ordered some 18500 cells on eBay. They are 3.7V Li-ion cells like 18650s but are 50mm long instead of 65mm long. The cell holder for the three AAA cells is just over 50mm long, so the 18500 cell will fit inside the torch, but the 18500 cell is smaller in diameter than the threeAAA cell holder. siliconchip.com.au I thought of using 25mm electrical conduit, but it wouldn’t quite fit inside the torch, and the 18500 cell was loose inside it. After cutting a suitable length of conduit, I solved these problems by cutting a slot in the conduit and heating it with my heat gun, then wrapping it around the 18500 cell while it was soft and pliable. Then it was just a matter of assembling the torch with the conduit sleeve and the new 18500 cell. It all fits together nicely and now there are only two connection points instead of eight. With some torches, stretching the spring on the cap end may be necessary, but that was not required in my case. The sleeve can be made from thick cardboard if you do not have 25mm electrical conduit. Although 18500 cells are rated at 3.7V, a fully-charged cell has a similar voltage (4.2V) to three AAA cells in series (3-4.5V), so I didn’t find any need to change anything inside the torch. After converting three of our frequently used torches to 18500 cells, it’s now just a matter of grabbing a charged cell as needed and then re-charging the flat cell instead of having to buy AAA cells continually. Miele clothes dryer repair D. T., of Sylvania Southgate, NSW found out (if he didn’t already know) that buying electrical goods at an auction is a bit of a gamble. Still, that gamble paid off as the faulty device turned out to be relatively straightforward to fix... My wife bought a used Miele T7944C clothes dryer at a local auction. The dryer came with a matching washing machine, which we ran a few loads through, and it worked fine. However, the dryer only worked for about 10 minutes before it stopped with a “Clean out airways” LED illuminated on the front panel. The first thing I did was clean out the obvious filters in the chassis around the door opening. These weren’t too blocked, but it’s hard to know what the problem threshold is when you have a new piece of kit. That didn’t help. Then I found another pull-out filter in the door, which also wasn’t too bad, but cleaning that didn’t help either. Searching the internet revealed this dryer is a ‘condenser’ type, where the moisture from the clothes comes out as liquid in a pipe that you feed into a drain instead of being blown out the exhaust all over your laundry. To achieve this, it has a closed loop where heated air is blown through the clothes like a standard dryer, but instead of exhausting out to the atmosphere, it circulates through a condenser where it is cooled, causing the water in the air to turn into a liquid and drip into a tank/drain. The air is then reheated and passed back through the clothes. All this heating and cooling of air may seem inefficient, but consider that with a regular dryer, fresh air is continuously heated from room temperature and blown out as waste. The internet also revealed that the condenser can be pulled out and cleaned. It too had some accumulated Australia's electronics magazine April 2023  89 fluff, but it wasn’t downright awful. Cleaning it as per the instructions made no noticeable difference. It seemed likely to me that there was a sensor in the air loop somewhere that would show a high temperature if the filters were blocked, so I thought I’d see if I could find it. After passing through the condenser, the air travels up the back through a duct made from galvanised sheet steel that passes under a cover screwed to the back. I removed the cover to reveal a heater, an over-­temperature mains cutout, and something that looked like a sensor. I pulled the connector off and removed the sensor by bending a pair of chassis tabs. On the bench, it measured about 83kW at room temperature. This seemed reasonable for an NTC thermistor, but since I didn’t really know what it should be, I decided to have a go at opening it anyway. The sensor housing was made from two pieces of plastic with four tabs that had been melted over to keep them together. I sliced these off with a scalpel, and the halves came apart to reveal a two-wire sensor that had been spot welded to pair of brass bars – the bars formed the connector pins. Most significantly, there was evident corrosion on one of the joints. Both joints were still physically intact – the pins were still well attached to the sensor wires, but I decided to clean it and re-solder the connections anyway. It wasn’t hard to re-solder after I scraped all the corrosion away. The hard part was fitting it back into the housing with the extra solder. In the end, I cut away some of the plastic housing to make room for the solder, then cable tied it back together and reattached the duct. I didn’t have any washing that needed drying, so I tested it with an old towel I dunked in water. An hour or so later, I had a nice dry towel. I’m not sure if my soldering cured it or if it was the disconnection and reconnection of the plug onto it that ‘cleaned’ the connector (I suspect the latter). Still, I’m glad I removed the corrosion – it was a future failure waiting to happen. 90 Silicon Chip Cable management of an aircon P. B. E., of Heathcote, Vic was asked to ‘have a look’ at a Panasonic CU-624KR air conditioner by a friend. It had been ‘professionally’ repaired, but it turns out that being a professional doesn’t necessarily mean you know what you are doing... This unit was only about 15 years old. The owners said they don’t use it much, so it should be OK. Actually, the opposite is true. Air conditioners, both in homes and cars, should be fired up for about 15 minutes per month to allow the oil to circulate, keeping them in good condition. This Panasonic had been fixed before by ‘professionals’. The problem then was that a mouse (or mice) had decided that the fine control wiring was a good place to sharpen their teeth. The wires were poorly joined back together and insulated with thick tape. It’s amazing it worked at all, but it did for about 12 months. Then nothing again – absolutely nothing. No error codes, lights or relay(s) clicking. I checked the outside unit first, thinking that’s where mice could easily get into. After undoing silly little clips and many screws, it all seemed OK. Nothing obvious was wrong. I gave it a good clean, particularly the fan and evaporator. I then started on the inside unit; this was harder to take apart. The screws are cleverly hidden behind plastic clips. With the screws out, the plastic cover still needs to be un-clipped from the main housing. I couldn’t find the clips for some time due to them being on top and the unit close to the ceiling. After finally getting the plastic cover off, mouse poo and small bits of wire fell out. Oh dear, “there’s your problem”! I made a drawing of the mains wiring that I knew I would need to dismantle. There were two active red wires; I thought that was a bit strange, so I marked them separately. I doubt that it would work if I reversed them. With a lot of wriggling and gentle force up and down, I got the two PCBs out that should be connected with the chewed wire. To make things more of a challenge, Panasonic (bless them) decided to make all these wires the same colour, white. I took the boards home for scrutiny. There were ten wires, with only three still barely connected. What goes to what? All I could think of, and hope for, was that the wires were in the same order on each board. I know one shouldn’t assume, but I had no choice. I set about reconnecting all 10 wires, about 20mm longer than before. That would make it easier to slide the boards back into the plastic housing. I wrapped the new loom in three layers of thick tape, hoping this would discourage future mice attacks. Back on the job, the reassembly was easier than the dismantling. I also packed in some Scotch-Brite pads laced with a good amount of cayenne pepper around both PCBs, hoping that mice aren’t fans of spicy food. I then reassembled the rest of the indoor unit. I went outside to check that I hadn’t forgotten something silly. I turned the unit’s circuit breaker on in the meter box and its separate switch on the wall, then noticed a relay clicked in the outside unit. That sounded encouraging. Back inside, I programmed the remote for cooling at 20°C. I then hit the on button and was greeted with a pretty blue LED. After about one minute, the unit fired up, and it smacked me with cold air. After five minutes, we got too cold and had to turn the temperature up. Another success. SC Australia's electronics magazine siliconchip.com.au Rack Equipment Ideal for IT Networking, Small Offices, Recording Studios, Sound & PA Equipment. GREAT VALUE and IN STOCK at your conveniently located stores nationwide. ALL cabinets feature: • Key lockable front, back & sides • Tempered glass doors • 60kg static loading capacity • Wall mount without brackets Swing frame SWINGS OPEN FOR EASY MAINTENANCE OR INSTALLATION 19" Rack Mount Cabinets 6U Flat Packed HB5170 / Assembled HB5171 12U Flat Packed HB5174 FROM 239 $ Swing Frame 19" Rack Mount Cabinets 6U Assembled HB5180 12U Assembled HB5182 FROM 299 $ Here's just some of our Rack Mount Accessories See our range in-store or online Patch Cables • Cat5e, Cat6a & Cat7 available YN8200-YN8299 FROM $3.75 24 Port Cat6 Patch Panel • Numbered ports • Labelling area • 1U rack height YN8048 JUST 79 $ 95 Patch Lead Management Panel • Keeps leads tidy JUST • Removable cover $ 95 • 1U rack height HB5434 32 Fixed Rack Shelves • Powder coated steel • Supports up to 20kg • 1U and 2U available HB5452 - HB5454 FROM 49 $ 95 Professional 19" Rack Style Enclosures • Metal construction • Supplied flat pack • 1U, 2U and 3U available HB5120 - HB5130 Shop at Jaycar for your Computer Networking gear: • Computer Modems and Routers • Neworking Leads & Adaptors • Ethernet Switches • Keyboards & Mice • Networking Test Equipment • Storage, Docking & HDD Explore our full range of rack mount products, in stock at over 110 stores, or 130 resellers or on our website. jaycar.com.au/rack-mount 1800 022 888 Prices correct at time of publication but are subject to change. Jaycar reserves the right to change prices if and when required. FROM 139 $