Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN'S LOG A story to support Murphy's Law Traditionally, by Murphy's Law, any servicing short cuts we attempt will always bypass the fault. And conversely, when we find a fault by conventional means, we can see an obvious short cut. But let's not take Murphy too seriously. The truth is, we live on short cuts of one form or another, from recognising the symptoms from previous jobs to the replacement of complete boards in order to save time. Generally, these short cuts work. It is only when they don't work that we blame Murphy (or bad luck). . This story would seem to support Murphy's Law, at least initially. It concerns a Thorn Model 9007 colour TV set, which uses the Mitsubishi G series chassis. This chassis was also marketed as the AWA model C608. These sets are now about 15 years old and this one belongs to a local motel, the owner being a relatively new customer. The owner brought it in with the complaint that it had lost colour and an initial check on the bench confirmed this. However, I found that it was possible to brute force it into colour - of a sort - by deliberately mistuning in the direction of the sound carrier with the colour saturation control wound full up. I used the term "colour - of a sort" because there was something funny about it; that being the only way I could describe it initially. To put it a little more scientifically, I knew it wasn't right but couldn't work out what was wrong with it visually. But there was more wrong with the set than that. The first thing I noticed, apart from the loss of colour, was that the picture was grossly over-scanning; mainly horizontally but also vertically to a somewhat lesser degree. There was also some obvious convergence error. FRACARRO SAVES ••• .. .more than $10. from every UHF television antenna you buy. Fracarro's 4PUN for better performance from weak UHF signals. Peter C. Lacey Services Pty. Ltd. P.O. Box 678 (74Fulton Rd.) Mount Eliza 3930 T el:03 787 2077 Fax: 03) 787 3460 ACN006893438 64 SILICON CHIP Colour bar pattern I fed in a colour bar signal from a TV pattern generator and, by careful mistuning, was able to bring this up in colour. But the degree of mistuning was very critical and it produced a strange effect. While each colour bar displayed the correct colour, it did so for only part of its width, starting with monochrome on the left- hand side and graduating to colour on the right. This probably accounted for the "funny" effect I had observed on program material. . My first reaction to the colour problem was to speculate on a possible front-end fault. Perhaps the tuner or the IF system was upsetting the response curve. However, I put these thoughts aside for the moment and decided to investigate the overscan problem. HT voltage The first thing I check in such cases is the HT rail voltage and this paid off. Normally, the HT for this set is 105V but it was actually sitting on 138V. This was far too high and suggested a fault in the power supply regulator system. This set is made up of a main board ("PCB-MAIN") and a number of auxiliary boards which plug into this via edge connectors. Fortunately, this being an old set, I have accumulated a fair stock of these auxiliary boards from scrapped sets and these prove very useful. First, they can help pinpoint a faulty board quite quickly by simple substitution. Second, leaving the substitute board in place is often the quickest and cheapest solution; an important factor in sets of this age. So I simply fished out a spare "PCBPOWER REG" board and plugged it into the main board. That didn't help much because the HT voltage was virtually identical. This made me suspect that one or both of the power ~~?sr, i"RA't>l1IONAU-'-f, 'B'-f MURPHY'S LAW, A~Y SE"R.V\CIN6 SHO'R.T"CUTS W~ A1TE:M?r W\U... At..WAVS "BVPASS ™e:. FAuL--r.. o regulator transistors , Q901 and Q902 (2SC935), might be leaky. These are mounted off the board, via a plug and socket connection. The plug and socket arrangement makes it easy to check them but I could find nothing wrong. I replaced the original power board and went to the "B-ADJ" pot, VR901, to see if (1) it was still working and (2) whether there was adequate adjustment available. Well, it was working and I had no trouble resetting the HT to the correct 105V with a perfectly reasonable pot setting. In short, all the indications were that there was nothing wrong with the power supply, except that the pot had been incorrectly set. Why this was so was a mystery at that point. And why hadn't the replacement board clarified the problem? It turned out that, for some reason, now lost in history, it was also wrongly set. In fact, I put it back in the set and was able to reset it to 105V as well. So that cured the gross overscan and a minor touch up with the appropriate controls produced a well proportioned picture. But we still had the colour problem. In fact, it was now worse than ever, in the sense that it was much harder to brute force than before (a lesson there, had I but realised it). The next logical board to check was the chroma board, "PCB-CHROMA". I have several on hand and it is a simple un-plug, plug-in operation. But no joy. Remembering how I had been caught with a faulty substitute power board, I tried a second one. But it was no use; this was not the source of the trouble. So what next? Still convinced that the best approach, at least initially, was board substitution, I picked what seemed the next logical one - the "PCB-VIDEO-CHROMA-OUT" board. Again, it was a simple operation, or at least it would have been ifl'd had one in stock. I didn't as it turned out and I finished up borrowing one-from a colleague. Unfortunately, the effort was wasted; there was no joy here either. And that just about exhausted the board swapping approach. All the major and likely sub-boards had been changed, without result. The two suspect areas remaining were the main board and the tuner. Swapping the main board was hardly a proposition, at least at this stage, but I had considered the possi- bility of a tuner problem causing a faulty response curve. Fortunately, I am well set-up to check tuner problems. As mentioned in my last set of notes, I have a self-contained tuner · and power supply which makes an ideal substitute in such situations. It's an old turret type VHF tuner from the monochrome era but it still works perfectly. The substitution was simple. The set's own VHF tuner connects to the IF system, the "PCB-HF" board, via pin 1 of a 4-pin plug (FA). I simply pulled the plug off and clipped the substitute tuner output lead to pin 1. And that disposed of that theory quick smart! There was absolutely no change in the set's behaviour. And it also put paid to the idea of a simple fix by substitution. I was going to have to start working now and track down the fault the hard way. So, harking back to my opening paragraphs, had my short cut attempts been a waste of time? Not at all; they had ruled out a very large part of the set, involving hundreds of components, and forced me to narrow my investigation to relatively few components. That said, I wasn't quite sure where to go from there. The only thing I could immediately think of was the colour burst and/or the associated gating pulse. It was a long shot, because failure here normally means total loss of colour. However, the CRO quickly ruled out that theory, as everything was normal here. What next? So what hadn't I checked? Perhaps I had missed something around the chroma board area and I so I began by checking the supply voltage. The supply to this board is a nominal 20V which is derived from the deflection out board ("PCB-DEFL-OUT") and fed in on pins 1 & 10. This voltage was spot on but I then began wondering about any bypass capacitors associated with this supply. ThE;lre is one on the chroma board (pin 10) but this was ruled out because the board had been changed. Further checks revealed a second bypass capacitor (C579) on the deflection out board, at point "B5". But again I drew a blank as changing it made no difference. I was feeling rather frustrated by now. In desperation, I went over the FEBRUARY 1993 65 a problem. It looked like I would need access to the control board but this, not to put too fine a point on it, is a swine of a thing to remove. What was needed was some other way to conFig.1: the faulty circuit firm my theory before I took such a section in the Thom 9007 colour TV set. Transistors drastic step. Q623 & Q624 are the first & So I tried another tack. By suitably second chroma amplifier juggling the chassis, I was able to reach stages respectively, while the underside of the main board and VR671 (top) is the colour bridge pins 15 & 16 of the edge consaturation control. This pot nector using a clip lead. This was connects to the chroma electrically equivalent to turning the board via a plug connector saturation control fully up which, to the main board (labelled mechanically, was where it had al"MN") & then via pins 15 & ready been set. 15 of an edge connector The result was that the picture imsocket. mediately jumped to full colour saturation. Not only that, but all symptoms of the fault had vanished. I could now bring in colour - albeit oversaturated - on all channels, without resorting to the critical tuning procedure previously required. So at last we had a breakthrough; it physical arrangement. The appropriate part of the cir- was somewhere in the saturation con•ce-•ow~ocus ( cuit is reproduced here as trol circuit. But where? It took only a , ~ I._ Fig.1. The saturation con- moment to shift the clip lead from trol is a soon pot, VR671, pins 15 & 16 on the edge connector to which connects to four pins pins 2 & 4 on the main board, whereon the control board. Pins 1 upon the fault was back exactly as and 3 connect to chassis, before. pin 2 to the moving arm, So what was there between the two and pin 4 to the active end set of pins? Answer: just two copper of the pot. Plug MN fits on tracks, about 70mm long, and the solthese pins and its cable ter- dered joints to them. I went over all minates on a second MN these joints but could find nothing plug which mates with four wrong. Just to make sure, I resoldered ~~i pins on the main board. them anyway but I wasn't really surWith me so far? Right. prised when this had no effect. The saturation control is But I was hot on the trail now. By eventually wired into the pulling the plug off the motherboard chroma amplifier chain pins, I was able to make a resistance aroµnd transistors Q624 measurement from pin 4 to pin 16. and Q625. This is achieved There was no problem there but pin 2 chroma circuit again but without re- by wiring pins 2 and 4 on the main to pin 15 was another story - this sult. Eventually, I finished up near board to pins 15 and 16 of an edge · measured no less than 100!2 And, the top lefthand corner of the chroma connector and this, in turn, carries seeing that I had already cleared the circuit, looking at a 4-pin plug on the the chroma board. Pins 1 and 3 con- soldered joints, this left a faulty copnect to chassis via the main board. per track as the only possibility. main board marked "MN". I had traced out all this circuitry Yet try as I might, I could see nothThis plug connects to the colour saturation control which, along with because I wondered whether the fault ing wrong with it, even with a magnithe contrast, brightness, vertical hold was somewhere in the saturation con- fying glass. Granted, the green lacand horizontal hold controls, is trol circuit. Ifit was, there were plenty quer covering the print didn't help mounted on a sub-board designated of places where it could occur: either and, in any case, I was well past car"PCB-CONTROL". It sits at the front of in the pot itself, in the plugs, between ing about the fine detail. All I wanted the set, beneath the picture tube, and the chroma board and the edge con- was to get the job finished. It was simple enough, of course. A the controls are all connected to the nector, or in the soldered connections appropriate parts of the circuit via to the boards. short length of tinned copper wire plugs, sockets and cables. between the two pins effectively In order to follow what happened Pure speculation wiped out the offending resistance next, it is important to get a clear The trouble was, this was all pure for all time. I then put everything picture of both the circuit and the speculation and proving it presented back together again, satisfied that the SERVICEMAN'S LOG - CTD r 66 SILICON CHIP _____ fault had been positively identif.ied and fixed. Why was the HT high? And that, as far as the job itself was concerned, was that. But there is a little more to the story than just the colour fault. What was the history of the fault and why had I found the HT rail wound up to 138V? As I mentioned at the beginning, this motel is a relatively new customer and this was the first time I had seen this particular set. Prior to that, the motel had employed a serviceman from an adjacent suburb. I know the party concerned. He's a nice enough bloke at a social level and, I have no doubt, is very kind to his mother. But his technical expertise is another matter. In short, I don't have a very high regard for it. So what happened? I can only speculate, of course, but I think it went something like this. The board was probably cracked right from the start but functioned OK while the two edges of the copper track were bright and touching. But eventually the copper corroded and this introduced It had him beaten until he discovered, TETIA TV TIP Panasonic NV-G22A VCR (G mechanism) Symptom: machine won't accept cassette. Capstan motor runs backwards and forwards for 1015 seconds, then machine switches off. If fault occurs while tape is loaded, machine will not eject. Cure: Check for "cut washer" clip on shaft under centre pulley. The pulley contains a spring loaded clutch which has a tendency to push the clip off the shaft. TETIA TV Tip is supplied by the Tasmanian branch of the Electronic Technician's Institute ofAustralia. Contact Jim Lawler, 16 Adina St, Gei/ston Bay 7015. More on computers enough resistance to upset the colour to some extent, at least on an intermittent basis. It was too much for our serviceman friend, although I would be the first to concede that it was a very nasty fault. Electronics Workbench® The electronics lab in a computer! A simple, intuitive and very powerful teaching tool, Electronics Workbench lets students and hol!byists design and test both analog and digital electronic circuits, without the delays and expense of a laboratory. FEATURES • Quick and simple circuit entry • Digital and Analog Modules included, complete with all components • Simulated instruments: dual trace scope, spectrum analyser, function generator, multimeter, digital word generator and logic analyser • Complete control over all component values and parameters • Print: circuit schematics, parts list, instrument readings, macros • Logic conversion - truth table to Boolean formula to logic gates • Customisable hypertext help system probably by accident, that he could cure it by jacking up the HT. So that was what he did, hoping that the resultant overscan would go unnoticed - which it did. Eventually, of course, the corrosion increased and the fault re-appeared, intermittently at first, then permanently. And that's where I came in. Am I being too hard? I don't think so. I would be the first to agree that anyone can be stumped by a stinker and I'm not criticising on that basis. But faking a cure is another matter. Apart from the fact that the customer is being cheated, it simply doesn't pay off in a purely business sense. Sooner or later it backfires and the customer forms his own conclusions. Anyway, I've scored a satisfied cus. tamer. THREE VERSIONS • Professional Version: EGA/VGA colour display; unlimited components • Personal Plus Version: Monochrome display; unlimited components • Personal Version: Monochrome display; limited to 20 components per ... EMONA INSTRUMENTS NSW ......................... (02) 519 3933 VIC ............................ (03) 889 0427 QLD .......................... (07) 397 7427 DISTRIBUTORS Perth ... ........ ..... ......... .. ... (09) 244 2777 Adelaide .. ...................... (08) 362 7548 Hobart .... ..... ........ ....... ... . (002) 34 2233 After my two previous stories about servicing computer monitors, I imagined that that would be the end of it; that I could go back to doing fair dinkum service work. But it was not to be; someone blabbed about my monitor repairs and before long I had EmonaLAB4 Four instruments in one package! The LAB4 has four full function test instruments: ■ DC POWER SUPPLY Triple output; 5V, 15V, 0-SOV ■ DIGITAL MULTIMETER ■ FUNCTION GENERATOR with neatly laid out and labelled front panel controls. Also, the LAB4 is compact and only takes up a small space, compared to four bench-top instruments. Full function, auto-ranging 0.02Hz-2MHz, with SWEEP ■ FREQUENCY COUNT,ER · 1Hz-100MHz input range INDEPENDENT INSTRUMENTS Each instrument in the LAB4 is completely independent and can be switched ON or OFF as required . The LAB4 has only one 240V mains input, saving on multiple power points. Each instrument is clearly defined EMONA INSTRUMENTS NSW ......................... (02) 519 3933 VIC ............................ (03) 889 0427 QLD .......................... (07) 397 7427 DISTRIBUTORS Perth ........................ ..... (09) 244 2777 Adelaide ....... (08) 362 7548 Hobart ......... (002) 34 2233 FEBR{!ARY 1993 67 SERVICEMAN'S LOG - CTD ~ c::=:::::::, ~ c:>C::::.~ another customer appealing for help with a suspect monitor. It turned out to be another Acer MM211 - a fact which made me feel almost cocky about tackling it. After all, I now knew where all the screws were. On the other hand, it seemed to be tempting fate to hope that it would be another power supply fault. ' Even so, the symptoms as related by the customer seemed to support that theory. The fault had first appeared as a collapsed vertical scan, resulting in a bright horizontal line across the middle of the screen. It had lasted only a second or so, the monitor then reverting to a normal display. In fact, the customer at that time wasn't even sure whether there was a fault or whether he had simply observed a brief mains power failure. When the fault subsequently reoccurred several more times, he knew he had a problem. On the other hand, he wasn't sure what to do about it. He knew enough about intermittent faults 68 SILICON CHIP to realise that tracking them down could take days, or even weeks. And as he was using the computer professionally, he couldn't afford to have it out of action for a long period. So he kept flogging it along while ever it would go until, inevitably, it failed completely. And that was the best thing that could have happened really. At least ·I had something to work on. With the cover removed, I went straight to the 12V rail where it emerges from the 12Vregulator. It was dead, so I moved back to the input and measured 20V which, from my previous experience, was spot on. Crook regulator My immediate conclusion was that the regulator had packed it in but I checked the 12V rail for shorts, just to be sure. There was nothing wrong there, which left only the regulator. This was mounted on a 75mm-long finned heatsink which is attached to the PC board by a couple of screws tapped into one end. The regulator IC was mounted near the bottom of this heatsink, with its three leads passing through holes in the board to the copper pattern on the underside. The first problem was to identify the IC. I can't remember the type number but it was a new one on me and so I sought the help of the friendly man behind the counter at the local Dick Smith Electronics store. It turned out that he was quite familiar with the unit and the circuit but not with the regulator IC. No problem, though. A quick phone call to the service department at headquarters and he had the answer. The original type number was not available but it could be replaced by a 7812. This is a regular stock item and he pushed one across the counter. I expected that that would virtually be the end of the story, the repair involving no more than a routine replacement. But not quite. Not being familiar with the faulty IC, I took the precaution of tracing the copper pattern around the pin connections before I removed it. And just as well. Having identified the pin connections, I compared them with the 7812 arrangement and found, as I feared, that they were not the same; two of the pin connections were transposed. This was an annoying development because it meant a messy repair. The IC would have to adopt a crossed leg attitude and I would probably have to attach an extension lead to one leg to facilitate this. I was trying to figure out how best to do this when I noticed an extra set of unused holes adjacent to those for the existing regulator but which had no apparent purpose. When I traced the copper pattern around these holes, it quickly confirmed what I had begun to hope. The makers had anticipated this situation and had provided a second set of connections to suit the alternative pin arrangement. From there on, it was plain sailing. I mounted the IC on the heatsink, soldered its leads to the alternative pads, and put everything back together again. And, of course, it worked and I had another happy customer. But I swore him to secrecy. I don't want any more computer monitors, thank you. And they're making them in colour now, aren't they? Ouch! SC