Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN'S LOG Lightning can cause strange faults Further to last month’s story on lightning damage, another one comes to mind, along with a story about of a couple of VCRs and a service manual that went walkabout. First, another lightning story. Mr Knight’s wife was in the living room looking out the window at the never-ending sheets of rain when there was an unbelievably deafening crack as lightning struck about 100 metres down the street. Everything electrical went off and then the lights slowly came back on. The initial shock of being so close to so much power cannot be understated and one is always amazed when people are actually struck by lightning and survive! I suppose the immediate physical effect is the deafness from such a loud noise plus the intensity of the flash, which may cause temporary blind­ness. However, these effects wear off and you begin to weigh up the cost of the damage inevitably inflicted by nature’s fury. In Mr and Mrs Knight’s case, the telephone was dead, as were some of 28  Silicon Chip the house lights and appliances. The outside light proximity sensor switch had also gone. And all the neighbours suffered multiple failures. The TV set, a Panasonic TC-29V50A (MX-2A chassis), had been on at the time but, as soon as the lightning hit, reverted to standby mode. As I later learned, the remote control restored the picture perfectly but there was no sound. Because the phone had been knocked out, Mr Knight had to drive to my shop to arrange for me to fix the set. I showed up that afternoon, intrigued as to why only the sound had failed. I suspected that, in the confusion after the strike, someone may have pressed the wrong buttons on the set or the remote control. However, after spending 10 minutes checking all the controls, only faint clicking noises could be head in the speakers. I didn’t have a circuit for the set and, as it was another large model, I was reluctant to move it to the workshop – as was Mr Knight. I hoped that a temporary fix could be organised until I could better prepare myself. First, I tried feeding in signals from their VCR but as expected, there was still no sound. However, the TV set on-screen displays showed that the set’s stereo decoder was working and could distinguish between mono signals from the VCR and stereo off-air transmissions. Next, I decided to try feeding the AV (audio/video) outputs from the VCR directly to the AV inputs of the TV set. I fetched some RCA leads from the truck, connected the two machines togeth­er and selected the AV mode. As before, the picture was fine but there was still no sound. This could only mean that the problem lay somewhere in the audio amplifier stage. I took the back off and, by tracing the speaker connec­tions, established that IC2303 (AN7169) was the stereo output amplifier. Rubbing my fingers over the solder produced hissing noises from each channel. Although not a definitive test, it did suggest, even without the benefit of a circuit diagram, that the fault lay between the input AV sockets and this chip – possibly in the volume control and mute circuits. Fortunately, the family had a portable radio/cassette player with line inputs for recording. By connecting it to the audio output sockets on the TV set and pressing the cassette record buttons, I was able to hear sound from the TV set at last. I decided to leave things set up in this manner while I ordered a circuit diagram. The only inconvenience the family had with this arrangement was that they had to physically adjust the volume control on the cassette player to the level they pre­ferred, as the remote control had no effect. And of course, the cassette player had to be switched on and left permanently in the record mode. I received the circuit about one week later, only to find that there were no less than seven ICs involved with the sound circuits (not to mention the muting and control processors). These were: sound IF IC2206, stereo decoder IC2201, AV control IC3001, surround sound IC2301, audio control IC2302, preamp IC2306 and output amplifier IC2303. The audio muting, simply put, was controlled by IC1102 to Q2301 and Q2302, as well as Q3015, Q3016 and other circuits, such as Q1113 audio defeat and Q1111 volume. As my provisional sortie had already eliminated half of these, I decided to take a signal tracer (a little battery-pow­ered amplifier) and a signal generator on my next trip. Mr Knight was delighted to see me back but exasperated to learn that I still didn’t know where the problem was and that I was only there to attempt to identify the faulty part(s). After all, this was just a simple sound failure – at least, as far as he was con­cerned. Selecting the left channel and using the tracer, I managed to monitor sound from pin 5 of AV control IC3001 (pin 1 was for the AV OUT) to pin 8 of surround sound IC2301. There was also sound from pin 3 of IC2301 to pin 6 of preamp IC2306 and from pin 7 to pin 3 of audio control IC2302. But there was nothing from pin 9 of IC2302 to pin 2 of output amplifier IC2303. I unsoldered the collector of Q2301 to ensure that the muting circuit wasn’t doing its thing but there was still no sound. So, by a process of elimination, the fault had to be in IC2302, a CXA1279AS, and/or its control circuits. A meter check established that the control voltage to pin 16 varied with the volume control, which was correct. I felt I had to be pretty certain as to which part to order, as Mr Knight was becoming rather “tetchy” about the speed of this “simple” repair. To be safe, I decided the best course was to order the IC chipset in case of a misdiagnosis. If nothing else, I would have them in stock for what is a fairly popular model. Anyway, as luck would have it and to my great relief, my diagnosis was spot on – replacing IC2302 fixed the problem and restored the sound completely. But why, in the multitude of components in this TV set (there are 25 ICs in all), did the lightning destroy only this IC and nothing else? Unfortunately, this is one aspect of the job I am not qualified in so I don’t have the means to explain it. Perhaps no mortal can! A tale of two VCRs My next story is about two VCRs, both Akai VS-F10EA models. This model VCR is old by present day standards but is a reliable performer and a popular choice as a rental unit. And this is the story of two such rental units which landed in the workshop to­gether. To minimise confusion, I have designated them as VCR 1 and VCR 2. Mr Carton’s set was VCR 1 and the symptoms were no video on playback or even AV in or out. Mr Darnay’s set was VCR 2 and the symptoms were described as intermittent stopping when playing back. But by the time I tried it, it was completely dead. These units presented a major problem; I had no circuit. Originally, I did have a complete manual but this had gone walk­about. I had a good idea as April 1998  29 Serviceman’s Log – continued to where it had gone but recovering it called for some diplomacy. More of that later but, for now, I was trying to manage with a slightly different circuit, namely for a VS-F16. Though close, this was still significantly differ­ent in parts and led me to doubt conclusions I had made on the basis of this schematic. Mr Darnay’s set, VCR 2, was the more urgent so I tackled it first. This model VCR features two power supplies: (1) a main switchmode power supply which provides seven rails (23V, 16V, 3 x 12V and 2 x 5V); and (2) an auxiliary miniature switchmode power supply on the motherboard which generates a -35V rail and a 5V rail, the latter called a “filament” supply for the display system. The auxiliary supply operates from the 23V rail. I checked all seven voltage rails at the output plug (WP201) of the main power supply. There were voltages on all seven, though not exactly correct. However, I often find that Akai’s marked voltages are not necessarily exact, often contradicting themselves on various parts of the circuit. 30  Silicon Chip I checked both voltages generated by the auxiliary supply. Both were present but somewhat low. The question was the degree of error and what was critical. The 5V rail was less than 4.5V, while the -35V was down to -27V. Akai service bulletins warn that low or dark displays may be due to two electros drying out in the auxiliary power supply. I decided to replace C446 and C447 with two 100µF electros (they are marked 47µF in the circuit diagram but 120µF had already been fitted by the factory). It was a futile gesture which made no real difference. I next checked all the crystal clocks with an oscilloscope, especially X701 (4.43MHz) on the video board. This also supplies a clock signal (fsc) to the digital servo (IV401, pin 22) on the motherboard. Everything seemed OK. My next step was to see if the loading motor mechanism was aligned correctly but as expected, I could find nothing wrong here (after all, when it worked, all functions worked properly). By now, I was coming around to the idea that either a micropro­cessor or the servo itself was intermittent. At this point, I decided to switch my attention to Mr Car­ton’s set, VCR 1 (the one with no video input). The plan was to tackle what now appeared to be the simpler fault, then use this set as a donor to fix the problem in VCR 2. The latter could then be returned, while VCR 1 could wait for parts to be ordered and installed (I hope all this makes sense). Because the fault was lack of video, I decided to work with a colour bar generator rather than risk a faulty tuner. The only problem was that I stupidly plugged the generator RCA plug into the wrong socket on the rear panel, namely the audio out (the back was facing away from me, it was dark and the sockets all looked the same – well, that’s my excuse anyway and I’m sticking to it). So, following the colour bar signal with the CRO, I tried tracing the colour bars to pin 1 of IC101, a TC4066 analog switching IC. When I found that it never reached it, as the VS-F16 schematic showed, I abandoned this approach and assumed the VS-F16 circuit differed from the VS-F10. However, I did find a video signal on pins 2 and 4 of IC101. I followed this video signal all over the motherboard to pin 5 of IC602 (AN3247K) and out again on pin 9 to pin 13 of character control IC102. And that’s where the trail went cold, with no signal out from pin 12 to the video output. It looked as though IC102 was the culprit and so, to confirm this, I momentarily shorted pins 12 and 13 together and the picture was restored. Of course, it was possible that the fault could still be external to the IC. But I was happy to accept that it was the IC and so I deso­ldered the corresponding IC from VCR 2 and donated it to VCR 1. Success – well, sort of; the picture was fine but there was a buzz in the sound and I still had to solve the mystery of no video in. The missing manual At this point, it is appropriate to reintroduce the subplot of the VSF10EA service manual which had gone walkabout and detail the history and order of events. Being a small service organisation, it is impossible to stock all circuits for all models, especially as new ranges appear about every five months from every manufacturer. The only way to survive in this environment is to co-operate with the opposition – you lend me your manuals and I’ll lend you mine. Normally, this arrangement works well but occasionally, when you deal with a large service centre where there are many people involved, manuals can get lost. And so it was with my VS-F10EA manual – I lent it to this centre about six months ago but they didn’t return it. When I reminded them, they didn’t think that they had ever bor­ rowed it. Anyway, I didn’t want to alienate them by pressing the point too strongly and simply assumed that it would eventually turn up and be returned. As it turned out, my luck was with me. I called into the centre recently to borrow some other circuits and on the spur of the moment I asked if I could borrow their VS-F10 service manual. Obligingly, the technician went to the filing cabinets and pulled it out, only to discover that it was my copy with my writing all over it! The technician was most apologetic and so I departed, much relieved at recovering my lost manu­al. And none too soon, because I was still puzzling over the buzz in the sound and the confusion over the RCA sockets and the colour bar generator. Fortunately, it didn’t take long to realise my error and sort out the confusion. So, with VCR 1 working properly at last, it was now reassigned as a donor and I could swap parts out of it and into VCR 2. The first step, of course, was to refit IC101 into VCR 2 (I know it sounds silly but that was the way it had to be). This done, I swapped the entire front panel with the timer micropro­cessor and display on it. I thought I had correctly diagnosed this too but it wasn’t long before it started to fail intermit­tently and ultimately failed completely. Next, I swapped IC403 (syscon), which means desoldering and resold­ ering 64 pins twice over. Once again, it started to work and then died. I was becoming rather dispirited but decided to swap the digital servo IC (IC401). This had exactly the same effect as before and so I put the machine aside and waited for inspiration. While I was catching up with routine work, I kept thinking about the symptoms of this rogue set. More often than not, it worked when cold rather than hot or failed after it had been on for a while. So why not try the freezer treatment? To cut a long story short, I expended an expensive can of freezer and achieved nothing. Well, not quite; there was some momentary activity in the power supply, which made me put my thinking cap on again. Perhaps some of these voltage rails were more critical than others but the question was, which ones? It was at this time that I had the chance to talk again to my mate from the opposition. He was very familiar with Akai VCRs and told me the 23V rail from the main power supply was the one to watch and if it was down to 19V to change the bridge rectifier (D1, D2, D3 & D4). This was the lead I needed and when I measured it, it read only 20.5V. Unfortunately, replacing the diodes made no difference and so I decided to check the main filter electro (C3). Why not connect another electro across C3 and see if that made any dif­ference? To my delight and surprise, it fixed the problem com­pletely. I replaced C3 (2200µF 35V) with a new one and reassem­ bled the VCR. It was now working perfectly. I left all the good parts in VCR2 and put VCR1 aside to wait for the new IC101 to be delivered. And so it all ended happily – for the customers. But in retrospect, I didn’t come out of it particularly well, either financially or technically. I had missed the obvious; ie, the need to follow up any suggestion of a power supply fault. My only excuse is that I got sidetracked by the need to work (initially) with a substitute circuit, an apparently inter­mittent fault and by my confusion over the voltage values. Still, I should have known better and I do know better. I simply didn’t follow the rules and paid the price. The flea-marker computer To finish up, here is a story on a brighter note. It comes from a reader and was inspired in part by these notes in the December 1997 issue, describing a service job on an AST Ascentia Colour Notebook computer. It comes from a VK5 amateur, S. M. of Elizabeth Downs, South Australia, describing a tentative approach to laptop computer servicing. This is how he tells it. It all started when my 14-year old stepson, Peter (not his real name), at high school and up to his ears in computers, wanted to visit the local computer flea market. My wife said OK; she hoped she might get some clip art for her craft hobby, while Peter might get a CD ROM or some more SIMMs. The place was chock-a-block when we arrived. There were trestles sag­ ging with games, programs, old computers, VDUs; you name it, it was there. After it had thinned out a little EVERYONE KNOWS... 2694.VET.SIL.1/4.1 If you‘re concerned about viruses while surfing the Net, you can rely on Vet –the all-Australian software that offers you superior protection and full local support. Vet protects against thousands of conventional and macro viruses and is suited to any PC platform as well as Novell NetWare & NT Server. Unlike other software developers, we only make anti-virus software so it‘s natural that we‘re the experts at it. Confidence in Vet‘s abilities extends from fellow surfers to governments, banks and companies in over 30 countries world wide. So if you‘re on email or surfing the Net, Vet has all the protection you‘ll need, including products with free mailed upgrades to make sure you stay protected. And, after all, when you go surfing who wants to worry about the quality of the water. Evaluate Vet for Windows 95, Windows NT Workstation and Windows 3.x at www.vet.com.au For your nearest reseller or an information pack telephone 1300 364 750 Email: info<at>vet.com.au All the Anti-Virus you need April 1998  31 I wandered around and saw a chap offering a laptop. A notice on it said, “A Mr Fixit Special”. It was a Tandy 1100HD with an LCD green screen, a 20Mb hard disc drive, a 3.5in floppy disc drive and MS DOS version 5. It included two batteries (one of them new) and all manuals and discs. The notice said, “power supply will not operate the PC, or charge the battery”. He wanted $45.00 for it. I pointed it out to Peter. With a gleam in his eye he asked, “Do you think you can fix it?” I asked the vendor if I could have look at the instruction book. “Yes, go ahead”, he said. “It just won’t run, that’s why I bought a new battery”. The book indicated a 6V battery, and the external power supply unit (PSU) was 9.5V at 1.2A. This went into the computer via a standard DC connector, similar to most plugpack PSUs. I said to Peter, “If it’s the PSU, I could easily make a new one”. By this time my wife had taken some interest in it. And, in answer to Peter’s unspoken question, replied, “Yes”, and handed him the money. We 32  Silicon Chip wandered around a bit, then headed for home. On the way, my wife and I called in at a delicatessen, leaving Peter in the car. When we came out, Peter could not contain his excitement. “It works, it works! I connected the second battery, turned it on, the ‘Charge Light’ blinked, the display came up with a start-up routine, and the beeper beep­ ed. Then it died”. When we arrived home Peter wanted to fix everything straight away. I persuaded him put the battery on charge while we had lunch. With the battery removed and a couple of makeshift pins in the battery plug, we applied 7.5V across the battery at 500mA from a constant current regulated supply and left it for about 45 minutes. When the charged battery was installed, the computer fired up straight away. Peter’s fingers flew over the keyboard. Every­thing appeared OK. “Right”, I said, “let’s check the PSU”. I plugged it into the mains and switched on. There was no smoke and the DVM indi­cated 10V DC at the plug. A 12V 3W festoon lamp lit up when connected across the plug, with 9.2V still indicated on the meter. So it wasn’t the PSU. “Ah”, I said to Peter, “Are you game. Shall we take it apart?” “Yes; what have we to lose?” It took us 10 minutes to undo all the screws and the little plugs and sockets, after which we were able to remove the covers. Continuity tests with a DVM and a DC plug with test wires showed that the switching action of the DC socket was OK. Further checks showed that the DC was applied to the PC board and that the battery plug was connected to the board. So what was wrong? I was inspecting the board for burnt or damaged components when I spied four miniature fuses marked F1, F2, F3 & F4, each about the size of a 0.25W resistor. A quick check with the DVM revealed that F2 (2.5A) was OC. I decided to use one strand of multi-strand hookup wire which I guessed would fuse at about 2A. This was soldered to one end of the dodgy fuse and a short piece of plastic insulation slipped over it (in case it blew and splattered everywhere). The other end was then soldered to the other end of the fuse and a DVM used to confirm that it was intact. It took us a careful 15 minutes to get it all back together again, with all the right screws and bits in the right places. No wonder computer techs charge $50-$60 just to look at a repair. With it all back together we tried it on the battery first. It worked OK. We took the battery out and tried it on the PSU – OK again. We then reinstalled the battery and the battery charge light came on, so all functions were OK. Peter couldn’t get inside quickly enough to show his Mum. I obtained two spare fuses, one for the computer and a 4A one for the other battery pack (a miniature one under the heatshrink cover). They cost $1.50 each. We subsequently checked various Tandy stores and established that the 1100HD was on the market in 1991 for around $2000 and that it uses a 386 pro­cessor. It prints OK on two Canon printers and an old Panasonic dot printer – all for $45.00 for the unit and $3.00 for fuses. Thanks S. M. for an interesting story. I wonder how many other old (and not-so-old) machines have been consigned to the scrap heap for relaSC tively minor faults.