Fullscreen HTML5Med Res (??MB)HTML4

SERVICEMAN'S LOG The path of TV service never did run smooth Ever had one of those jobs where nothing seemed to go right; where each lead turned out to be false? Of course; we all have. Well, I had one recently which produced just such a sequence of false leads before I finally cracked it. The story concerns a 46cm National TC-1809 colour TV set fitted with an M9 series chassis. It is one of about 12 used by a local motel and I have been servicing them since they were installed about 11 years ago. They have proven to be a very reliable set with a minimum of problems over this period. ' But, like most sets, they do have some particular weaknesses. One which I cottoned onto quite early in the piece involves C519, a lOµF 250V electrolytic smoothing capacitor on the supply rail for the RGB driver transistors. This can dry out and in some cases, one of the terminal lugs can actually come adrift. In any case, inadequate smoothing on this rail can produce some funny effects. In fact, this particular set had been FRACARRO DELIVERS•.. .. .best television antenna performance for less cost. Fracarro 10BL4, 10BL5, 10BL45, 10 element yagrs that ~ work. I Peter C. Lacey Services Pty. Ltd. P.O. Box 678 (74Fulton Rd.) Mount Eliza 3930 Tel:03 787 2077 Fax:(03) 787 3460 ACN006893438 32 SILICON CHIP in the workshop about a year ago for just this fault, at which time I had given it the usual once over for any other obvious problems. But now the motel proprietor was on the phone with a new fault which, with a bit of guesswork, I interpreted as severe vertical overscan. So I said "bring it in and we'll have a look at it". When he turned up, I plugged the set in immediately and, yes, that was the problem. And it was quite severe; on an SBS test pattern, the SBS logo at the top of the screen had all but vanished, with much the same degree of error at the bottom. This drew my attention to another point: the scan, though excessive, was still quite linear. It was almost as if the height control had simply been wound up. In fact, that was the first thing I tried - not that I believed that the problem was goipg to be solved that easily. But I wanted to check a couple of points. First, was the control working? And if it was, was it set correctly, somewhere near its mid-position? The answer to both questions was yes; it appeared to be working correctly and it was correctly set. In fact, by winding it right back, I was able to reduce the scan to almost normal but not quite. But that was not the answer. The need for such an extreme setting clearly indicated a basic fault - one which might well get worse with time. So where to from here? The horizontal scan appeared to be normal; perhaps overscanning by a whisker but nothing serious. So I went first the main HT rail, which is always a good starting point. This should have been at 111 V but was a trifle high at 115V. I reset it as a matter of routine but this had only a marginal effect. Circuit details At this stage, it was time to get out the circuit and try to come up with some ideas about the fault. The relevant section is reproduced here and I will try to give the reader a broad picture of what is involved. On the left of the diagram is the horizontal output transformer and near the bottom is a winding terminating in pin 2. This point feeds diode D503 and capacitor C517 (lO00µF 35VW} to provide a +20V rail (test point E14}. This rail supplies several sections of the set and is also used to derive a +12V rail. This is achieved using resistor R516 (immediately to the right of C517} and 6V. zener diodes D506 and D507. And that brings us to the next section of the circuit: jungle chip IC501 (AN5431}. Among other things, this chip provides the sync separator, horizontal AFC, and the horizontal oscillator, vertical oscillator and ramp generator circuits. These latter functions are accessed via pins 10, 11 & 12, while the +12V supply is fed directly to pin 13. Pin 10 is also fed from the 12V rail, in this case via R428, diode D406 and the height control (R406, a 30kQ pot}. Having digested the circuit arrangement, the first thing I did was check the +20V rail at test point E14. This was at +19.5V which was quite acceptable. My next stop was the +12V rail; this was checked at pin 13 of the IC and came out a fraction high at 12.5V, which again is an acceptable figure. Neither observation helped very much and I was pondering on my next logical move when I realised that the fault condition had changed. The degree of overscan was not constant; while I was watching the screen, the height suddenly decreased slightly. ~EIGHT 05 q408 14J l.3K C40i 16VIOO .. CSII 25V33 I ii R506 !OK l.T i zw ' Fig.1: relevant section of the National TC-1809 colour TV receiver. The +20V rail components are associated with pin 2 of the EHT transformer (lower left), while the +12V rail is derived via R516 & zener diodes D506 & D507, The height & linearity components are associated with pins 9-12 of jungle chip IC501. was further encouraged when, while probing at the diodes in an effort to measure their individual voltages, the scan began varying erratically. So did I have a faulty diode or a dry But before I could follow up on before I came back and, when I turned joint? To save mucking about, I pulled this, I was called away on another the set on again, it was no longer them both out and fitted two new matter. It was half an hour or more overscanning. In fact, it was now ones. These gave a slightly lower voltunderscanning slightly but . age than before (about 11.8V) but again Y- the fact that I had previously I didn't regard this as important. It left the height control at its did, however, reduce the overscan minimum setting partly ex- slightly and I found that I could actuplained this effect. ally set the height control to give a More importantly, as I normal scan by backing it off to its watched the screen over the extremity. next few minutes, the height That was no solution of course and slowly increased, eventu- I still suspected a thermal fault someally settling down to its pre- where in the set. In the meantime, it vious gross overscan posi- was time for lunch and so I turned the tion. So it looked like we set off to let it cool down. The lunch had a thermal problem · of break would also give me time to think some kind. about the problem. At this point, I decided to re-check the+ 12V rail, only Doubts dispelled this time I went to the two When I switched the set on again zener diodes (D506 & D507). about an hour later, any doubts I may I had tried to find these ear- have had about a thermal fault were lier without success but this immediately dispelled. The picture time I eventually tracked was now underscanning substantially, them down under the verti- by about 35mm at both the top and cal hold control. bottom of the screen. It then started to And now I seemed to be creep up and, in a few minutes, was on the track of something. back to where it had been before I For one thing, the voltage switched it off. across them had increased Fortunately, I had taken the preslightly to about 13V. While caution of leaving the meter across this increase was not im- the 12V rail, so I was able to monitor 'SO 1-r t - ~ t - 1 ~ portant in itself, I wondered it while the scan increased. For all whether it was a sign of practical purposes, it didn't vary. H~0 A. ~~~L. PROSL.E:M something that was. And I So, summarising the situation, I was OF SOW\E. \<.\tJ'O••• APRIL 1993 33 SERVICEMAN'S LOG - CTD • e~ • • • HA.)l\t-lG DlG5-STE.t> '""•r • • "' ~• '.., , • . ,.. • • ~ 1\-\E: CIR(.Ul""f A~~~NGe:N\E.N,, iHE. FlR..S, 'TH\ NG 'I 'DIP WA.'S CHE:.CK -rHE: +~OV "RA\\.A-r ,-i;.s-, 'PO\N,14- ... e: able to rule out both the +20V rail and the + 12V rail. And if there had been a fault in the zener diodes (and I was no longer sure about this), it was incidental anyway. False lead number one. Height control circuit So where to now? Pin 10 of IC501 and associated components in the height control circuit seemed to be the next best bet. As already mentioned, the main components here are R428, diode D406 and the height control pot (R406). Also connected to pin 10 is diode D403, the other side of which goes to the +12V rail at pin 13. My first step here was to measure the voltages on pins 10, 11 & 12. According to the manual, pin 10 should be at 1V, pin 11 at 6.3V and pin 12 at 7V. All these came out pretty well spot on, so there were no clues here. While making these measurements and studying the circuit, I noticed an interesting diversion. If we go back to the junction of R428 and D406, we 34 SILICON CHIP encounter another resistor - R405 . And if we trace the circuit from this point, to the right, down and left, we come back to the +20V rail, near E14. As I said, this is a diversion and, fortunately, I dicl not allow myself to be side-tracked by it. Nor should the reader, because it has nothing to do with the final outcome. But I am still wondering about its function. My best guess is that it is a belt-and-braces arrangement to ensure that the jungle chip keeps working - at least to some degree - in the event that the main +12V rail goes out. But back to the height control cir- . cuit around pin 10. Since it was obviously a thermal fault, I went over each of the components in this section with freezer spray, fully expecting that cooling one of them would cause the height to drop. Alas, no joy. Even when I went further afield, spraying C405, C406 & D404 in the linearity circuit plus several components around pins 11 & 12, there was still no response. My n ext move may seem a little unusual but was quite logical - I removed the height pot from the board. One reason was simply to check it for any funny behaviour. A Samsung set I had serviced a few weeks earlier had produced all kinds of weird intermittent symptoms. After much heartbreak, the problem was traced to a crook pot; one which measured OK when tested initially but which varied its value drastically in operation. It took quite an effort to nail it. So I gave this pot a thorough going over. And as far as I could tell, there was nothing wrong with it. My second reason for removing the pot was that I could then check a number of surrounding components without removing them. R428 checked out OK but D406 presented a query. It had a very high forward resistance reading and no reverse reading. Not only was this unusual but the diode itself was something of a mystery. Designated MA26TA, I suspect that it is a special purpose device of some kind but I could find no reference to it in the regular manuals. Nor is there anything in the parts list to provide a clue. It is quite tiny physically and is colour coded in green and cream. But since it didn't seem to be working as a diode should, I tried replacing it with a regular 1N914 small signal diode. I didn't know it then but that was another wasted effort, to say nothing of the frustration of not knowing what the device was. The other components - diodes D403, D404 & D405, capacitors C405 & C406 in the linearity circuit, and the resistors in the vertical hold circuit - all checked out OK. And this was rather significant, in view of the final outcome. No frame collapse At this point, I switched the set on again. Because the height control pot was still out of circuit, I had backed off the brightness in anticipation of total .frame collapse and the inevitable bright line across the screen. But this didn't happen; at least not totally. There was now a scan about 50mm high across the centre of the screen. And this was the first real step towards a breakthrough because, on the face ofit, it was impossible; how could we have a scan with no voltage on pin 10? But this wasn't strictly true; the meter showed that there was a voltage on pin 10. It wasn't much - a mere 0.086V - but there should not have been any voltage at all. And was it enough to give the 50mm or so of scan? Probably. Anyway, I knew I was onto something. At this point, I had to leave the set to complete another job - this time for a couple of hours. When I came back and turned it on again, there was considerably more scan than before. It now measured about 150mm and varied quite randomly. That didn't make much sense in view of the previous behaviour but I'd more or less given up trying rationalise what was going on. But the real query was the source of the spurious voltage. When I looked at the circuit, one component stood out above all the others: diode D403, between pin 10 and the +12V rail. Ifit was suffering reverse leakage, it could apply voltage to pin 10. Granted, I had already tested this component on the board and it had checked OK, but stranger things had been happening with this job. There was one quick way to find out; It took only a moment to unsolder one leg, whereupon we had total frame collapse. I fitted a replacement diode and we still had total frame collapse. I re-fitted the height pot and we had normal scan with the pot set near its mid-position. More importantly, it remained rock steady over the next couple of hours and through a couple of on-off cycles. Problem solved. But there was a bit more to it than that. When I checked D403 again on the meter, it now showed substantial reverse leakage. So why hadn't it done that when I first tested it? It would have saved me several hours and much frustration. Then there were the two suspect zener diodes. Now that I had found the real culprit, it seemed unlikely that there was anything wrong with them. It was a simple job to replace the original pair, so they went back in and behaved perfectly. And finally, the mystery diode D406, MA26TA. This was re-fitted and, again, the set continued to perform without a hitch. And it continued to behave over the next couple of days and through several on/off cycles. It has now been back with the cus- TETIA TV TIP Akai VS200 VCR (and Samsung equivalents) Symptom: Intermittent orno clock when the set is on standby. Also EE picture shows a small amount of hum. Cure: C15 in the power supply (an electrolytic capacitor on the 32V rail) is defective. The manual shows th is capacitor to be a 1OµF 50V unit but this set had a 22µF 50V fitted. It didn't matter which value replaced the faulty unit. TETIA TV Tip is supplied by the Tasmanian branch of the Electronic Tec;hnician's Institute ofAustralia. Contact Jim Lawler, 16 Adina St, Geilston Bay 7015. tomer for several weeks and, at last report, has not missed a beat. So after a couple of false leads, all ended happily. But it was not one of my most satisfying jobs. Filling in the gap Now, for a change of scene, here is something from our regular contributor, J. L. of Tasmania. He starts off with the cryptic comment, "this story started a week before it began". Well, if you say so J. L. , who am I to argue, but I must confess that these ancient brain cells had some trouble working that out. No matter; here's the rest of his story. I was asked to repair a TV set for an elderly customer and , so that she would not be without entertainment while I worked on it, I left one of my older loan sets with her. I went ahead with her job and had it completed within a couple of days. However, it wasn't convenient to take it back immediately and, before I was able to do so, she rang to say that my set had broken down. This was rather a surprise; it is an old HMV 48cm model - a B4803 fitted with a Rank Arena Dl chassis. It had given no trouble in the several years since I inherited it. Naturally I wasted no time in delivering her set and, while I was packing up the old HMV, I asked her what had happened to it. She told me that the screen had gone bright red, with even brighter lines across it. There was no sign of a picture, although the sound had continued uninterrupted. I groaned inwardly when I heard this because these symptoms in older sets often mean that the picture tube has developed a heater-cathode short. If this was the trouble , then the set would be a write-off because it wouldn't be worth fitting a new tube to the old chassis. Back at the shop, I wasted no time getting the wreck onto the bench to see ifI could rescue the situation. The first thing I did was measure the voltage on the collector of the red output transistor. This was down to 30V; over 100V less than normal. A check on the other two collectors showed a much more likely 190V. So it certainly did look like a short in the tube. There is one sure way to prove a heater-cathode short. If the collector voltage reverts to normal when the socket is pulled off the tube , then the problem has to be in the gun assembly. But when I pulled the socket off in this case, the voltage remained low. I gave a little cheer and began looking for the true cause of the problem. If the output transistor had become leaky, it could account for the symptoms. But this would be a fairly unusual fault , since output transistors usually fail completely- a dead short or an open circuit. In my experience, most leaky transistors are small signal, low voltage types. Anyway, I removed the transistor and checked it on the tester. It was perfect! No leakage and normal gain for the type. I re-fitted it to the board and hunted for any other fault that could cause a red screen. I checked the base and emitter voltages on all three transistors and found nothing amiss. Well, nothing amiss in that they were all the same but nothing like the figures given in the circuit diagram. The bases read 1V and the emitters 0.4V, instead of 2.3V and 1.8V respectively. This suggested that all three tube cathodes were drawing less than normal current. What's more, the voltages on all three transistors remained more or less the same when the socket was pulled off the tube. Where's the current going? This left me with quite a puzzle. The fact that the red collector voltage was very low suggested that excessive current was being drawn. But it APRIL 1993 35 It looked as though it was only a cou- SERVICEMAN'S LOG - CTD wasn't being drawn through the transistor, nor through the tube. So where was the current going? Nothing appeared to be getting unduly hot. There were no sparks or flames to indicate where the ergs were going. Yet the power was being dissipated somewhere. I checked all the resistors associated with the output side of the transistors and found a slight anomaly. The circuit diagram showed 10kQ load resistors and 2. 7kQ feed resistors to the tube cathodes. The values I measured were 15kQ and 10kQ respectively. But they hadn't gone high; they were the actual values fitted, so I had to accept them as correct. Leakage checks The only explanation I could suggest for the low collector voltage was that there was a leakage path somewhere on the board. I checked leakage on the red channel, using a high impedance digital meter, and compared it with the blue and green channels. There was no leakage that I could detect on any of them. Yet there had to be something that was leaky. And if it couldn't be detected with a multimeter, then perhaps it was only present when working voltages were applied. Leakage of this kind is usually found only in capacitors. Resistors go high, not low, and other devices usually short out completely. In this circuit there was only a lO00pF ceramic capacitor in the emitter circuit and a spark gap off the collector. It took no time to replace the capacitor, without any result. So next came the spark gap. I put Z0V from a bench power supply across it and it showed no sign ofleakage. Yet when I removed it from the board, the collector voltage came up to 170V and the picture returned in all its glory. The spark gap was totally enclosed in heatshrink plastic and, after I cut this away, I could see through the gap. ANTRIM TOROIDAL TRANS 0 ple of thousandths of an inch across but, by holding it up against the light, I could see that there was something inside the gap. (I trust I will be forgiven for reverting to the imperial system of measurement but it was more appropriate in this case). At first I couldn't find anything thin enough to go through the gap. Then I thought of the feeler gauge in the car toolbox. The .002-inch gauge fitted perfectly and when I looked again the gap was perfectly clear. I re-fitted the gap to the board and the picture remained bright and clear, so I declared the fault found and fixed. I finished the job by putting a short length of heatshrink tubing over the gap and shrinking it into place. I have no doubt that the gap will continue to work perfectly but I can't help wondering what it was that caused the voltage dependent short inside a spark gap that was very effectively sealed inside a thick layer of heatshrink plastic. Well thanks J. L., and another happy result. And I trust the story finished at the same time as it ended! SC AUSTRALIAN MADE TV TEST EQUIPMENT 10% discount on all test equipment. 15% discount on two or more items. Built-in meter to check EHT transformers including split diode type, yokes and drive transformers. $78.00 + $4.00 p&p ERS ~ V Built-in meter reads positive or negative voltages from 0-50kV. For checking EHT and other HT voltages. $98.00 + $5.00 p&p Strong magnetic field. Double insulated for your safety with momentary switch operation . 240V AC 2.2 amps . As important as a soldering iron! $75.00 + $10.00 p&p F 10 • Now made in Australia. • Standard range 15VA to 2kVA. • Large ex-stock quantities available. • Competitive pricing available for OEMs & resellers. • Simply the best transformers all round. Designed to test infrared or ultrasonic control units . Supplied with extension infrared detector lead. Output is via a LED and piezo speaker. $84.00 + $4.00 p&p - ::sc-- ~ ~~ ~ !UNER REPAlf!~ From $19 exchange "':'z: ~ pluspost&pack Cheque, Money Order, Visa, Bankcard or Mastercard 216 Canterbury Rd, Revesby, NSW 2212, Australia. PTYLm HARBUCH 90 George Street, Hornsby NSW 2077. Telephone (02) 476 5854. Fax (02) 476 3231. 36 SILICON CHIP Phone (02) 774 1154 Fax (02) 7741154