Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN'S LOG Those summertime lightning blues Every summer, after the hot humid days, there follows the inevitable storm and whether it is the El Nino effect or not, these storms seem to be getting more violent. Or is it that electronic devices are becoming more vulnerable? One would have thought that, after all these years, lightn­ing would be better understood but it doesn’t appear to be. I watched a program recently that went on to show that lightning does indeed strike the same place twice – sometimes even more. And with older TV sets, the high humidity causes insulation breakdown in the EHT transformers and focus circuits. The day after a storm is always hec­ tic, with many confused people on the phone. Some are surprised that their electronic equipment isn’t working and fail to make any connection with mother nature’s major disturbance the night before. Most of them just want their gear fixed immediately but of course, that’s not always possible. For my part, I organise a stack of loan sets so that people don’t become too impatient. I never cease to be amazed by the path lightning takes through a set – it never seems to be logical or in a straight line to earth. Sometimes the damage can be particularly serious on a set that did not appear to be near the strike, while others right next to the hit remain totally unaffected. The fireball One such case occurred when Mrs Williams anxiously asked me to call. She was certain that the whole lot would be a complete write-off. When I arrived at her house, she told me that she was walking into the lounge room when the lightning struck, hitting the garage roof only a few metres outside the front door. 28  Silicon Chip And, according to the lady, a fireball had come through the front door and moved horizontally until it hit the TV set. I must admit that I was a bit sceptical about the fireball bit but, even so, I half-expected the AWA 2C6346 to be burnt to a frazzle. I removed the back at arm’s length to avoid getting soot on my clothes but, to my surprise, it was exceptionally clean for an 11-year old set. I poked about, removed the chassis and looked everywhere. There was no sign of any damage and the fuse was intact. In the end, there was nothing for it but to switch it on. Fearing the worst, I hit the power switch and stood well-back but it was a total anticlimax. The set came on perfectly except that it gave a snowy picture, which was fixed as soon as I recon­ nected the antenna. I then checked the VCR, a Panason­ ic NV-G16A. Neither the controls on the unit, nor on the remote control, were working. I switched it off while Mrs Williams brought in a cup of tea and after 15 minutes I switched it back on again. The controls were now working perfectly, as indeed was the remote control. The one explanation appears to be that a lightn­ing surge can tem­ porarily disable a microprocessor, by applying false voltages to a wide range of pins. And if one is lucky, it will restore itself after a few on/off cycles. Either that or there is a subtle in­ termittent lurking in it somewhere. In either case, there wasn’t much I could do about it then. If, or when, it reappears it will be time to try again. So Mrs Williams was exceptionally lucky; the only thing that had failed completely was the Foxtel STV-100 cable decoder, which was Telstra’s responsibility. A weak effort The next customer was lucky too, as mother nature’s effort here was also a pretty weak one. This time, the set was an Italian-made Fujitsu General, model FGS211 (BS950 chassis). This too was dead and I soon discovered that the mains fuse (F451) had blown. There was no visible damage and the cause turned out to be switching transistor T401 (BU508A), which was shorted. The bridge rectifier hadn’t been damaged, as I would have ex­ pected. There was, however, one weird side effect. The sound was fractionally low and this turned out to be transistor TR702 (BC547C) in the audio preamp circuit. This transistor had become leaky and so that problem was easily fixed. No lightning The next problem was not quite so easy. There are some customers who insist that their equipment was dam­ aged during a storm but subsequent investigation proves that this was not the real cause. In this case, the customer brought in a VCR that was now showing fine horizontal interference lines when playing tapes. Initially, it looked to me as though an electro had dried out in the power supply, thereby causing ripple in the power supply to the video head preamplifier. As it turned out, the true cause was rather sneaky. Underneath a small metal can, soldered as a component screen onto the printed board, was some of the no­ torious brown goo. It was completely concealed and it had corroded and shorted the tracks underneath a drop­ out compensation circuit. Removing it and cleaning the board restored the picture, so the problem was hardly due to a lightning strike! The 76cm Toshiba There was no doubt about the next job. Mr Johns owns a 1989 76cm Toshiba (model 329P8A) and lives on the first floor of a block of units. This is a huge set and weighs in at around 50kg. Faced with this, I elected to go to the mountain. The fault was described as a horizontal line across the screen. I removed the back and quickly dis­ covered that R327, a fusible 6.2Ω 1W resistor feeding 27V to pin 7 of the vertical output stage (IC303, AN5521), was open circuit. Access to most of this circuit is not easy but nevertheless, I decided to replace the resistor and IC together to save time. When I switched it back on, it only took a second or so before the resistor burnt out again. Naturally, it was the last resistor of that type in my toolbox. This time, I looked more carefully at the set and soon noticed that IC361 (TDA8145) on the U905B DPC-2 board was very hot. In addition, the PC board was dark all around it. I decided to leave the set where it was and lend Mr Johns a portable until I could order in the necessary parts. A few days later I returned and re­ placed IC361 and resistor R327. This time the resistor lasted a few seconds longer before failing and the IC was still getting extremely hot. It was then that I spied what had to be the real culprit: coil L464 on board U905C DPC-1. This coil (TLN3061) had be­ come so hot it had melted its plastic insulation and was burnt black. Well, I knew I didn’t have a replace­ ment coil and it might be a long wait for another one. Because the part was in the hori­zontal deflection correction circuit I decided to try it without the coil in circuit to see if I had some sort of vertical deflec­tion. This turned out to be a colossal mistake. Initially, at switch-on there were a few noises I hadn’t heard be­ fore, then a very loud one. After that, there was silence and all life ceased to exist inside the set. I had blown the line output transis­ tor (Q404) and vapourised the 145V rail which feeds Q404’s collector via resistor R444 (0.82Ω, 2W) and pin 2 of the horizontal output transformer. Again, the thought of negotiating a 76cm TV set down a flight of stairs, lifting it into a vehicle big enough to carry it and then getting it into the workshop was too horrible to con­ template. I left the set where it was and went home. The next day I ordered the parts which arrived almost immediately. Third time lucky? No, you guessed it; there was more trouble to come. Despite replacing all the parts, the line output transistor (Q404) got very hot and failed yet again. There had to be another short nearby and after a while I found that diode D440 (ERC0615) in Q404’s collector circuit was shorted. I didn’t have one of these to hand but I did have an FR307, another line output transistor and a 0.82Ω resistor for R444. Pincushion distortion This time, success at last! The pic­ ture was restored and I checked the HT rail, refitted the back and recon­ nected the exter­nal speaker and AV connections. With it all back together again I was enjoying a nice cup of tea, when Mr Johns and I both noticed that the tennis wasn’t quite right. No, I don’t mean that someone was chucking a wobbly; instead, I mean March 1998  29 Serviceman’s Log – continued lightning strikes. One way of quoting for such sets, where damage appears to be severe, is to price a replace­ ment chassis rather than fixing it to component level. Sometimes it is a safer proposition but in some cases, the cost of a re­placement chassis is simply too high to make it econom­ ically viable and the set either has to be repaired at component level or written off. The unluckiest victim that the picture was all wrong in the east/west direction, with noticeable pincushion distor­tion. “Oh bother!” I said (rather untruth­ fully) as I removed the back again. I quickly found the three east/west controls – R358 width, R356 keystone (trapezoid) and R359 DPC (pincush­ ion) – and marked their positions with a felt-tip pen before twiddling each one in turn. This had no effect; all three appeared to be totally inop­ erative. This seemed to suggest that IC361 had failed again; either that or I was indeed going to have to manhandle it back to the workshop so that I could use the CRO to trace the waveforms around this circuit. Please – not that! A stroke of luck Well, it was at this stage that my luck turned for the better. Earlier in the piece, I had sensibly had the foresight to fit a socket for IC361. This 30  Silicon Chip meant that I could easily replace it without having to first unsolder the module that it was on (U905B) in or­ der to gain access to the copper side. As it turned out, the IC had failed again and the new one finally fixed the last problem. But why had IC361 failed the sec­ ond time? The answer is that it was almost certainly a byproduct of L464’s failure and/or any of the other failures that occurred after replacing IC361 the first time. I realigned the three controls using a crosshatch generator and settled down to a replacement cup of tea before replacing the back. I then had another cuppa to make sure that it was still working – while watching the tennis – and I would probably have stayed for a third if Mr Johns hadn’t made some subtle comment about how busy I must be. The above story just goes to empha­ sise the good and bad luck aspects of The unluckiest of the lightning victims was undoubtedly Mr Evans, with Yours Truly suffering collateral damage on the side. He was watching the spectacle that night when there was a strike a few streets away. The lights and power flickered and his Fujit­su General TV set, an FG2012 with a Goldstar PC-04A chassis, went completely dead. When I called, I removed the back, expecting just the fuse to have gone. However, not only was fuse F581 in­ tact but the switchmode power supply was actually working. But that was as far as it went and I noticed that some of the printed wiring had evaporated. That was enough for me; it was back to the workshop. Later, I found that the missing print­ ed wiring involved the sub-switch on the mains power switch, SW851. The latter is a pushbutton toggle type which opens or closes the two con­ tacts directly in the incoming 240V power line to the power supply. It also performs several functions via the sub-switch (P801). These functions involve the various rails out of the power supply and also include the control of relay RL801 in the 112V rail from pin 17. The vaporised sub-switch wiring led to four NPN transistors – Q704, Q705, Q713 and Q714 – in the standby/on switching circuit and this had also been destroyed. All this damage was duly re­paired but this only allowed the standby LED LD701 to turn on. By this stage, there should have been 112V on the collector of the horizontal out­ put transistor (Q402), applied via relay RL801. However, the relay was not being activated. This set provides a test function here although this is not shown on the circuit. A 2-pin male connector, P702, is connected between pin 41 of the main IC (IC701) and chassis (pin 41 drives the damaged NPN transistor circuit mentioned above). By bridging these pins, the relay could be switched on via Q801S and Q713. Unfortunately, the horizontal output stage started but then died. My next stop was Q401, the hori­ zontal driver transistor. A waveform check at its collector indicated that it was delivering only low ampli­ tude pulses. The collector voltage is derived from pin 6 of the EHT trans­ former (T801) which should be at 28V. After following a number of false leads, I finished up at D401 which connects between the 18V rail from the switchmode power supply and the 28V rail. This diode was short circuit. At last, I seemed to be getting somewhere, as a snowy raster now appeared. Unfortunately, no control functions were available at the front panel or via the remote control and I still had no sound. On the other hand, I now had clear confirma­tion that IC701 was faulty; pin 41 still needed to be bypassed, whereas it should have been doing the job itself. So, regardless of any other faults in the set, this IC it had to be replaced. In fact, I finished up replacing two IC701s. No, I didn’t replace the same one twice; there are two separate ICs marked IC701 on the circuit – the original 42-pin PCA84C640P/030 device and an 8-pin PCD8572 device connected to it. But just to confuse matters, the 8-pin device in the set was marked X24C02P and designated IC702. Anyway, I now had a picture but there was still no sound and the au­ tomatic tuning wouldn’t lock in. It took the replace­ment of IC401 to fix most of this. This IC is a TDA1940 and, among other things, provides sync separation, a burst generator and vertical and horizontal sync outputs. Next on the replacement list was IC601 (TBA-120T) which provides the sound IF functions. Changing this restored the sound. I now had a working set – well, sort of. No remote control The one thing left not working was the remote control. This involves a 3-terminal package (PA1) which contains an infrared sensor and a preamp­ lifier. One terminal goes to chassis, one is fed from a 5.6V source and the other delivers pulses to pin 35 of IC701. At least, that is what should hap­ pen but no pulses could be seen on the CRO. I removed the PA1 and disassem­ bled it. It was fitted with a GL3274 IC (IC1) which is not obtainable. However, I was able to purchase the entire package for $32 plus tax. This fitted, the pulses could now be seen at pin 35 but there was still no remote control action. I checked the 5V rail for ripple and replaced ZD701, a 5.6V zener, as it was down to 4V but I was getting nowhere. I confirmed that the remote controller was working properly and transmitting the correct data patterns, using another TV set. And this is where the exercise came to an abrupt halt. I was forced to the conclusion that the faulty pre­ amplifier (IC1) had damaged the new microprocessor (IC701), this probably involv­ing the internal circuitry asso­ ciated with pin 35. And this meant that the microprocessor would have to be replaced a second time to restore the remote control function. But enough was enough. This IC is an expensive item and is time-con­ suming to fit. And even if I did fit it, I could not be 100% sure that that was the only fault still involved. So I had to call it quits. It was a difficult decision, con­ sidering the amount of time, money and frustra­ tion that had been expended but I had to make it. I told the owner that I couldn’t economically repair the set and advised him to settle with his insurance company. I would make no charge, of course; the loss was mine. By mutual agreement I acquired the set which was working but without the remote control function. I hoped that I might be able to salvage it in the long term and recoup some of my losses but right now, it was a write-off. In fact, this is a classic example of how a repair can get out of hand; the time taken to diagnose and repair each succes­sive stage eventually exceed­ ed the value of this 1992 48cm TV set. Sometimes, it is hard to decide where to stop; the next component replaced may be the last one needed to complete the job. And it is virtually impossible to estimate the total cost of the repair without actually making it. Talk about a catch 22 situation! It’s not the end End of story? Well, I thought so when I put the set aside a few weeks SILICON CHIP This advertisment is out of date and has been removed to prevent confusion. SMART ® FASTCHARGERS Brings you advanced technology at affordable prices As featured in ‘Silicon Chip’ Jan. ’96 This REFLEX® charger charges single cells or battery packs from 1.2V to 13.2V and 110mAh to 7Ah. VERY FAST CHARGING. Standard batteries in maximum 1 hour, fast charge batteries in max. 15 minutes AVOID THE WELL KNOWN MEMORY EFFECT. NO NEED TO DISCHARGE. Just top up. This saves time and also extends the life of the batteries. SAVE MONEY. Restore most Nicads with memory effect to remaining capacity and rejuvenate many 0V worn-out Nicads EXTEND THE LIFE OF YOUR BATTERIES Recharge them up to 3000 times. DESIGNED AND MADE IN AUSTRALIA 12V-24V Converters, P. supplies and dedicated, fully automatic chargers for industrial applications are also available. For a FREE detailed technical description please Ph: (03) 6492 1368 or Fax: (03) 6492 1329 2567 Wilmot Rd, Devenport, TAS 7310 March 1998  31 ago and then, more recently, when I had finished writ­ing the above notes which I have left exactly as I originally wrote them – ready to go to the Editor. In fact, it was only the intervention of a weekend which held them up. Boots and all And then, quite by chance, a similar model set came in for a minor service. I was sorely tempted – could I solve the mystery once for all? It wouldn’t cost anything, except time, so I jumped in, boots and all. I “borrowed” IC701 from the new arrival and fitted it to the written off set. The moment of truth had arrived; I switched it on and tried the re­ mote control. It still didn’t work – everything was the same as before with normal pulses into pin 35 of IC701 but with no response. Obvi­ 32  Silicon Chip ously, the fault wasn’t in the IC as I had thought. But where was it? The preamp had been replaced and was generating normal pulses; or was it? More to the point, did I have the correct preamp? I began to recall some discussion over the part number. Fujitsu’s spare parts division had been most helpful and had quoted me for a type 106-042A. I noted at the time that the unit in the chassis was a 106-042B and this was also shown in the manual. I did draw attention to this difference but I was assured that the “A” ver­sion was a direct replacement. Nothing unusu­ al about that; it happens all the time. But now I was more than a little suspicious. I went back to Fujitsu and began delving a little deeper. In fact, it didn’t take long to confirm my suspicions; the two units were not interchangeable. They both did the same kind of job and generated similar – but not identical – pulses. In greater detail, the pulses from the type “A” preamp are inverted compared to those from the type “B” unit. That’s fine for an IC designed to accept that type of pulse train but not for IC701. Fujitsu were most apologetic and promised to send a re­ p lacement immediately. But I couldn’t wait. I “pinched” the preamp from the set on the service bench and fitted it in my set. And bingo, it worked – at long last the set was complete and working normally. It’s hard to describe one’s feelings in situations like this. After all the hard work, there was finally a happy ending and I would be able to SC recoup some of my losses.