Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN'S LOG Smoke, fire & confusion Yes, that is the only way I can describe the happenings which inspired this month’s notes. More to the point, some of these “happenings” were of my own creation; in hindsight, I should have done better. But, it’s easy to be wise after the event. In the continuous pursuit of brownie points, in the con­stant hope of a financial reward, or gratuity, I grasped the opportunity to repair one of SILICON CHIP’s monitors. Until now, my copy book wasn’t looking too good with them. One of their previous 21-inch monitors had the infuriating habit of dying at their premises but working constantly in my workshop. In the end, I concluded that my workshop was either drier or less polluted than their office. Or was it the other way round? Any­way, there didn’t seem to be any way of even starting the repair on that monitor. This time, it was a 1993 MAG MX17F 17-inch monitor with two distinct problems. First, there was an intermittent vertical jitter and height size problem and then, after about half an hour, the set would pulsate on and 28  Silicon Chip off in hiccup fashion (oh no! – more intermittents). Removing the case and metal covers gave fairly good access to the main printed circuit carrying the deflection circuits. The power supply, small signal and RGB circuitry boards turned out to be less accessible, in roughly that order. On examining the main board it didn’t take a mental giant to see that there were quite a few suspicious-looking solder joints. So out came the soldering iron and an hour later I had reworked almost the entire board. When I reconnected it, every­thing looked good – the vertical problem had vanished and the set was still running perfectly an hour later. I reboxed it and put it aside for a soak test, thinking that that was that. It was still working at closing time and half way through the next day but that was as far as it went. There was no longer any vertical jitter but by lunchtime, it had started to hiccup again. During the soldering procedure, I had noticed a dark spot around Q506, a 2SD799 low-voltage high-current TO-220 transistor. Thinking that maybe this transistor had become damaged due to high temperature from the dry joints, I decided to remove it for testing and possible replacement. The only problem was that I didn’t have a replacement or even a substitute. Fortunately, the set employed another one in a different circuit location and so I swapped them over (this wasn’t easy, as access was just under the yoke). But the swap made no difference – an hour or two later, it was hiccuping again. The scientific approach Next, I removed and reworked the power supply and small signal boards but the set still hiccuped. The brute force ap­proach was over; now for the scientific one! First, I summoned Mrs Serviceman to get on the blower and track down a circuit diagram – I always give her the easy tasks – while I dusted off the multimeter and warmed up the CRO. As it happened, she quickly traced the agency but that was as far as it went – the company wasn’t prepared to supply a circuit diagram which meant that I was on my own unless I could score one from somewhere else. But for now, I decided to let it ride while I delved deeper into the monitor. I was waiting for it to warm up when the phone rang and I had to deal with an elderly lady trying to book in a TV repair at home. It took some time to get all her details, the type of set, the fault, the likely costs involved and make an appointment. By the time I finished, there were two people queuing at the coun­ter. And then the phone rang again. All in all, by the time I dealt with everything it was over half an hour and when I went back into the workshop, the monitor was well and truly hiccuping. More precisely, this show­ed up on the supply rails which were pulsating in sympathy. In particular, I wanted to measure the 185V rail (the only voltage that wasn’t fluctuating was the 325V to the main electrolytic from the bridge rectifier). I also needed to figure out whether the hiccuping was due to an abnormal load, a safety circuit, or an excessively high voltage rail. I let everything cool and tried again and, as luck would have it, the fault was now permanent – it was hiccuping whether it was hot or cold and regardless as to how long the set had been running. Next, I shorted the horizontal output transistor’s base to its emitter and hung a 100W load on the 185V rail. It was still hiccuping. I then tried varying the input voltage with the Variac but it was a switchmode power supply that gave all or nothing. However, there was one thing – as the analog multimeter was swinging up and down, I noticed that the needle seemed to overshoot past 200V, although this might have been due to lack of damping in the meter. I marked the position of the 185V setup pot (VR303) and then ad­justed it. Interestingly, the set stopped hiccuping when I reached the lower end stop of VR303. Bingo! – it meant that this rail had to be too high. Circuit tracing My next step was to trace out the control feedback circuit – see Fig.1. This showed that VR303 was fed from the 185V rail via R331. The voltage on the wiper of VR303 then biased Fig.1: the voltage control feed­ back circuit in the MAG MX17F computer monitor. regulator IC303 which in turn fed optocoupler IC302 in the switchmode power supply primary. Resistor R331 was a 1W unit and, when I removed the brown glue that was covering it, I found that it was slightly disco­ loured. So, were its colour markings correct? It was now nearing closing time and I was about to shut the shop but I was eager to prove that I was on the right track. I read from left to right – brown, red, black = 12Ω and quickly fitted a new one. Big mis­take! The set tried to start, there was a crack – a puff of smoke – and WHOOPS! It was too late to investigate further. I had to leave it; there was some (supposedly) important engagement or other for which I had to be on time – a wedding anniversary or something. It wasn’t until next day that I realised my folly. Initial­ly, I had noted only the first three bands but there were actual­ ly five equally spaced bands – which I read as brown, red, black, brown and white. And the multimeter read 118kΩ. So was it really a 120kΩ resistor and had I misread the third colour, or had it been changed due to the brown goo or heat? But this didn’t make sense either. It was the white band that shook me out of my lethargy. As I had read it, the white band would be in the tolerance position. But there is no white tolerance band in the resistor scale. I had read the colours in the wrong order – the correct sequence was white, brown, black, red, brown. And that worked out to 91kΩ, 1%. So the multimeter reading of 118kΩ was correct; the resistor was originally a 91kΩ device but had gone high, to 118kΩ. This meant that the 12Ω resistor I had fitted was singular­ly inappropriate; as testified by the fate of IC303, which now consisted of just three legs sticking up out of the board. It took some time to find the rest of it in order to determine its type number. It turned out to be a TL431. I fitted a 91kΩ resistor for R331, replaced IC303, reset the pot to the mark I had made, connected a digital multimeter to the 185V rail, and gingerly switched on. The set functioned correctly and the rail was almost spot on 185V. Finally, I rehoused the chassis and put it aside for a long soak test, while trying to ignore the “was it ready?” October 1997  29 Serviceman’s Log – continued was marked. After such major surgery, I felt that it should have been checked. Anyway, at least it per­formed well after a prolonged soak test. The Magtron was made by AVAT, which stands for Advanced Video and Audio Technology Co. It complies with a newish Swedish safety standard (MPRII) for levels of magnetic and electrical radiation, which it proudly boasted on the front. Anyway, this was severely corroded by, of all things, coffee being poured into it. However, the damage was localised and it didn’t take much to clean up the mess. The reason it was dead was because D515 was short circuit and C527 had exploded. Apart from that, it worked well when the work had been completed. The ancient Sanyo pleas from SILICON CHIP – I wanted to thoroughly test it. It went back a few days later and hasn’t missed a beat since. Now, about that, er, possible emolument arising from pre­viously mentioned brownie points. No? I thought not. Oh, well. Two more monitors Surprisingly, the same week I had two other monitors in for repair. They were similar but dissimilar, if that makes sense. One was a 1993 MAG LX1564 and the other was a Magtron BMC-14SV4. They were similar in that the badges were the same font, size, and colour. The other similarity was that they were both dead due to corrosion. However, their chassis were completely different. The MAG came from a location near the beaches, was corroded by salt air and in serious trouble. The power supply was blown and had to be rebuilt component by component. The faulty items were: F301, 3.15AT fuse; Q301, 2SK955 FET; IC301, CS3842A IC controller; and ZD307, 18V zener. Once 30  Silicon Chip the power supply was working, it went into hiccup mode – like the previous MAG moni­tor. Interestingly, it used the same main rail regulation circuit and the same resistor but in this case the cause was a short circuit horizontal output transistor (2SC4747). By the time I found this, R153, a 1Ω resistor, had expired from the current flow. I also replaced a leaky electrolytic capacitor (C330, 1000µF 35V). It was a bit like me: old, haggard and worn out, but still functional – just! I was still not out of the woods. There was no picture and there appeared to be sparking inside the CRT socket. This monitor was fitted with a lot of internal metal mesh electrostatic screens which makes access difficult. I had to move several before I could get to the problem area. And yes, it was the CRT socket – it was corroded and sparking across its own internal spark gap. A new socket fixed that and it behaved like a new one. I haven’t been able to track down a circuit for this model either. I would have liked to have known how to set up the HT and EHT rails, as nothing Now for a change of scene – TV sets. Mr Woods is a long standing customer of mine who owns a number of sets in a large rambling 2-storey house. Unfortunately, the house is on the side of a hill; worse still, all his sets are 63cm models. The one he wanted me to look at was an ancient Sanyo CTP8631N employing a B7PJ chassis, though this is probably irrel­evant for this story. He said his grandchildren had come down for the weekend and jumbled up all the stations. What’s more, the video would not work. He implied that all that was needed was a retune. Quick and easy – now where have I heard that before? I arrived that afternoon and climbed, thankfully, only two flights of stairs to their rumpus room. The TV set sat in a rather dark, damp-looking corner. I switched it on and noticed that the touch sensors wouldn’t change channel nor was there any picture. In fact, I was just pointing this out when there was loud banging from the rear and bright flashes of green raster on the screen. I switched off immediately and asked Mr Woods if this the simple fault about which he was complaining? Well, actually it was – except it hadn’t made that banging noise before! I turned the set around and carefully removed the back. One has to be very careful with some Sanyo models; the heavy back can fall and take the neck off the tube in one quick guillotine action. This model was OK SILICON CHIP SOFTWARE Now available: the complete index to all SILICON CHIP articles since the first issue in November 1987. The Floppy Index comes with a handy file viewer that lets you look at the index line by line or page by page for quick browsing, or you can use the search function. All commands are listed on the screen, so you’ll always know what to do next. Notes & Errata also now available: this file lets you quickly check out the Notes & Errata (if any) for all articles published in SILICON CHIP. Not an index but a complete copy of all Notes & Errata text (diagrams not included). The file viewer is included in the price, so that you can quickly locate the item of interest. The Floppy Index and Notes & Errata files are supplied in ASCII format on a 3.5-inch or 5.25-inch floppy disc to suit PC-compatible computers. Note: the File Viewer requires MSDOS 3.3 or above. ORDER FORM PRICE ❏ Floppy Index (incl. file viewer): $A7 ❏ Notes & Errata (incl. file viewer): $A7 ❏ Alphanumeric LCD Demo Board Software (May 1993): $A7 ❏ Stepper Motor Controller Software (January 1994): $A7 ❏ Gamesbvm.bas /obj /exe (Nicad Battery Monitor, June 1994): $A7 ❏ Diskinfo.exe (Identifies IDE Hard Disc Parameters, August 1995): $A7 ❏ Computer Controlled Power Supply Software (Jan/Feb. 1997): $A7 ❏ Spacewri.exe & Spacewri.bas (for Spacewriter, May 1997): $A7 ❏ I/O Card (July 1997) + Stepper Motor Software (1997 series): $A7 POSTAGE & PACKING: Aust. & NZ add $A3 per order; elsewhere $A5 Disc size required:    ❏  3.5-inch disc   ❏ 5.25-inch disc TOTAL $A Enclosed is my cheque/money order for $­A__________ or please debit my ❏ Bankcard   ❏  Visa Card   ❏ MasterCard Card No. Signature­­­­­­­­­­­­_______________________________  Card expiry date______/______ Name ___________________________________________________________ PLEASE PRINT Street ___________________________________________________________ Suburb/town ________________________________ Postcode______________ Send your order to: SILICON CHIP, PO Box 139, Collaroy, NSW 2097; or fax your order to (02) 9979 6503; or ring (02) 9979 5644 and quote your credit card number (Bankcard, Visa Card or MasterCard). ✂ and once inside I could see that it was very damp and obviously tracking from the ultor cap. As my CRC 2-26 contact cleaner/ lubricant was in the car way down the street, I asked if he had anything similar handy. The best he could suggest was some WD40. Close enough, I thought – it would do. I sprayed the touch sensors, ultor cap, EHT transformer and tripler, then wiped up the excess where I could. When I switched on again, the channel selector worked and there was no more banging. But there was still no picture and I could hear hissing and spluttering noises from around the EHT transformer. I turned the set off and sprayed more WD40 in that area. Another big mistake: I switched on again and stuck my head in the back, trying to pinpoint the sparking which I could still hear. Then there was a loud clicking noise and a flashover near the EHT transformer connections. The WD40 ignited in huge ball of fire and I nearly lost my eyebrows and what hair I have left. The area had been fairly saturated with WD40 and it con­tinued to burn with 30cm long yellow flames which showed no signs of extinguishing. I realised I had only a few seconds before the wooden case might start to burn too and cause a major disaster. I switched the set off quickly. There was nothing to hand except a large rag which I grabbed and attempted to smother the flames. In what seemed like an eternity, I finally put the fire out and then turned around aghast to see that Mr Woods had watched the whole saga. He was far less concerned about it than I was. I think many people would have abandoned the situation there and then but I figured that if I did that I would be let­ting Mr Woods down as well as myself – like falling off a horse, you should get back on immediately to restore your self-con­fidence. Or so I’m told. In practice, I play it safe; I don’t get on horses. The trick was to pretend that it was all quite normal. The melee had lasted, perhaps, only 10 seconds but I had at least seen where the fire started. I pulled the chassis out and unscrewed the EHT transformer. This transformer is fitted with terminal pins (or “posts”), which protrude vertically in a semi­circular array from a plastic support above the main winding assembly. October 1997  31 Serviceman’s Log – continued These terminals are supply points for the various voltage leads leading to other parts of the circuit. They are fed, in turn, by short bare wires emanating from within the transformer. And one of these short wires – the one connected to the G2 (screen) terminal – had corroded and broken, creating a gap across which some 500V was producing a spark. And it was this spark which had ignited the WD40. With some difficulty, due to the short length of wire and the restricted space, I managed to solder a jumper across the gap. There was now no need to wipe away excess WD40 because the heat had evaporated it along with any moisture and I was confid­ent there were no more sparks to ignite it in any case. Neverthe­less, I was nervous when I switched the set on again. It was all an anticlimax – both picture and sound were fully restored, much to my relief. The various stations could now be tuned in correct­ly with their corresponding touch button numbers and even the video functioned. 32  Silicon Chip The fire area looked rather messy, with molten plastic leads and fittings, but I checked that their insulation was still OK and nothing was likely to short. After I had replaced the covers and put the set back, I explained the situation as best I could. I advised him that this damp corner was not the best place for it, that he shouldn’t leave the set on unattended, and that he should have some sort of fire appliances inside his house anyway (smoke detectors, fire extinguishers, etc). Finally, I emphasised that the set was old and corroded and really should be replaced. I didn’t dwell too much on the flamma­ble qualities of WD40 and left while the going was good and my reputation still intact. My final faux pas My final faux pas for this week – did I really need that many? – involved a Sony SLV-X57AS video recorder which came in with the fault “too fast”. More exactly, when playing a tape the horizontal hold was off speed and speech was fast, rather like slow Donald Duck. Both the capstan and drum motors were running fast and the auto-tracking was not working. I removed all the covers and the front panel and examined the set carefully all over. It looked fairly new and in good condition. I refitted the control knobs and switched on. According to the service manual, all the pulses in and out of the servo/system control (IC501) were at incorrect speed. These included the drum FG, PG and capstan FGs, as well as the control pulses from the ACE head and the outputs to the motors. I noticed that when I touched pins 7 and 6 of IC403, the drum motor could be slowed until the picture was locked. IC403 is the error integrator. I checked all the supply rails and “scoped” them for ripple, in case some of the electrolytics had gone in the power supply. All were present and correct. I then changed IC403, IC406, IC407 and was about to change IC405 before even trying to change IC501, a 100-pin surface-mounted high-density microprocessor. But by now I was beginning to smell a rat. The machine was in good condition and no-one had fiddled with it – or so I was assured. What drastic action could have caused this? Well, I had to order these two ICs and they were expensive. In the meantime, I would have to put the machine aside and so, rather than risk losing the screws, I decided to reassemble the whole thing. Fitting the front panel is a little fiddly, especially with the toggle control knobs and switches, but at least I had the whole thing back together. I was checking it to make sure I had done it all properly when I noticed a switch under the front panel, marked “Color System NTSC PB on PAL TV”. It was in the wrong position but I didn’t put any significance on this; it can easily happen when refitting the front panel. Instead, I simply moved the switch to its correct position and switched on for a check. However, when I pressed play, the picture from the tape came up perfect in all respects. Suddenly all was clear. This recorder is a dual-standard machine; it will play either an NTSC tape or a PAL tape into a PAL receiver and someone had inadvertently moved the switch into the wrong position. So much for my high-tech approach. I had found the fault and fixed it unwittingly – if only it hadn’t taken so long for me to wake up to it! SC