Items Covered This Month JVC AV-S290AUT (BY-I chassis) stereo TV set Teac CT-M761ST 76cm TV set.

I don't normally see many JVC TV sets, a fact that I attribute more to their excellent reliability rather than them not being the market leaders. In fact, I feel that their circuitry is somewhat more complex than in other sets but they are certainly well built.

Anyway, I was suddenly privileged to have not just one JVC set in for repair but two of the same model. They were both 1989 68cm AV-S290AUT (BY-I chassis) stereo sets. I don't really want to dwell on how I became involved in fixing these but it went something on the lines of the old mates act and redeeming outstanding favours (I seem to get into far too many of these situations!).

The first set (Set No.1) arrived in my workshop six months ago with the complaint that it was dead. It wasn't really but the main 115V HT rail was extremely low and replacing capacitors C951 and C952 (220μF 160V) fixed the problem and I thought the customer had gone away happy.

Obviously I was wrong and had done something to offend because the set recently resurfaced at my mate's opposition workshop which isn't all that far away. And my mate, after having had a tinker, decided to call in a favour I owed him.

The fault was described as retrace lines and vertical top foldover. At my suggestion, the workshop's technician, who I might add is a very capable lad, had changed all the electrolytic capacitors in the vertical timebase but to no avail. I don't know what other "tinkering" might have taken place by the time the set finally came to me but I was now on my own.

I hate problems such as these and I am grateful that I do not do many audio amplifier repairs, as they are very similar to vertical output stages. Amplifiers have this chicken and egg problem involving feedback. One can never be quite sure where the problem starts in the feedback loop.

I started by checking the voltages against those marked on the circuit. They turned out to be slightly high all round, especially around Q404. However, the oscilloscope showed a perfect waveform arriving from pin 13 of the jungle IC (IC201) on the small signal panel to Q406 (the vertical driver), although my meter measured 0.7V at the base of this transistor instead of the 0.8V on the circuit.

The main question was whether this 0.1V was critical, because the waveform became distorted on the retrace part of the waveform, at the collector of Q406. Was the transistor being forward biased adequately? The distortion showed up as a small lump on the retrace pulse.

The circuit is conventional enough, with Q406 driving two transistors, Q401 and Q402, in push-pull. The unusual part of the circuit was a signal take-off between the two outputs via two zener diodes - D403 (MA4200, 20V) and D402 (MA4270, 27V). This goes to "vertical drive" transistor Q403 which in turn drives Q405. Q405 then provides vertical blanking pulses to Q404, which are fed back to Q401.

There are two errors in the circuit diagram. Q401 (2SD1271A) is an NPN transistor and not PNP as marked, with the collector going to the collector of Q404. There is also an additional diode (D408) between Q403 and Q405. And its anode is connected to R410, not R402.

Some in-circuit resistance checks here provided no further clues and I was unsure as to what to do next. And then, as luck would have it, the identical set (Set No.2) arrived - only this one was very sick. It was severely rusted and corroded but I decided to try to repair it in the hope that this would help to fix set No.1.

Unfortunately, after spending an hour on or so it, I realised the problems were too extensive to make a complete repair worthwhile. The set was dead, with no picture, no sound, no vertical timebase, no remote control and no teletext! And at least one component was overheating.

I fixed the power supply, patched up the overheated horizontal output transformer and made the vertical timebase work by replacing R552 (5.1Ω 3W). Fixing the picture was much harder but the fault turned out to be diode D321 in the base circuit of Q203 and Q306.

The remote control failure was due to IC004 TC4049BF, a surface-mounted inverter on the "S Select Module". I didn't fix that but I did replace four surface mounted electros - C010, C013, C016 & C017 - which affected the set's memory. I also let the teletext fault go but suspected Q104 on the module. The sound fault involved IC651 (TA7630P).

The same fault

Having done the basics to achieve a picture, it was disappointing to find that this set also had the same fault as the first set - retrace lines and foldover! Despite this, I still felt that I could use this second set to track down the problem.

I took some voltage readings and quickly found that although the second set displayed exactly the same symptoms, the cause was entirely different. First, R409 (1.5kΩ) was getting hot and the voltages around Q404 were low this time.

I replaced C403 (33μF), which decouples the other side of R409 (this resistor connects to the 115V HT rail), but it made no difference. It looked as though Q404 was being switched on too hard by Q403, which had voltage on its base. Shorting it to chassis reduced the strain immediately. The voltages across the output transistors were correct, so the obvious suspects were the zener diodes D402 and D403. Replacing these immediately fixed the fault.

This was important, because I now had a reference set that I could use to tackle set No.1. And naturally, I hoped that the fault would turn out to be the same. Unfortunately, when I replaced these zeners, it made no difference.

The resistance of the vertical deflection coils was the same (25.4Ω) on both Panasonic tubes. Nevertheless, I swapped the chassis over to confirm that the coils were OK. I was also able to swap the small signal panels to confirm that the vertical drive was correct for both. I had already checked that the voltage rail (supplied via D404, R522 and D552 from pin 7 of the horizontal output transformer) was correct at 39.8V (nominally 41V) and I had also checked that R552 (4.7Ω) was OK.

Naturally, I was still confident that I could quickly solve the problem with this simple 6-transistor circuit. It wasn't to be - over two hours later, after I had swapped every transistor and diode with set No.2 and checked almost all the resistors, I still hadn't found the fault. And there wasn't much left to change.

At this stage, I had a TV set with a complete new set of electros, transistors and diodes in the faulty circuit section. What's more, the supplies were correct and the waveform was fine going in. However, it was incorrect after the driver transistor and transistors Q403, Q404 and Q405 were not being switched on because of this distorted waveform. In fact, set No.2 showed that the retrace pulses should be very tall and thin whereas on set No.1, they were small and fat with a little lump on them.

I spent the next half hour checking all the possible ways this pulse could be attenuated so much, particularly concentrating on small capacitors like C404, C405, C409 and C410 but got nowhere. In the end, I was about to abandon the whole sorry mess and was contemplating whether to just swap the good parts and make one set out of the two.

And then it struck me - one thing I hadn't done was to check all those electros that had been replaced by the other technician. I knew he was conscientious and capable, so I didn't feel he was likely to have made a mistake. But now, having run out of all other ideas, I decided to go over his work.

He had in fact replaced over half a dozen capacitors but only five involved this part of the circuit - C401, C402, C403, C408 & C552. I decided to change them, as you can get faulty new parts occasionally. Replacing C401 made no difference but when I came to C402. I noticed that a 47μF capacitor had been substituted for the original 2.2μF (100V) unit. That's a huge difference - about 20 times bigger.

Fitting the correct capacitor value fixed the fault completely!

Well, the moral of this story really doesn't need reiterating as it hurts! I have no excuses. The sad part is that, in all the mess, the exact cause of the original fault was never found - it was fixed somewhere along the track when one of the other parts was changed.

So overall, it was an unsatisfactory end even though the set was fixed. And at least I got to keep set No.2 which can now be used as a source of spare parts for other similar sets - always assuming more come in.

An intermittent Teac

My next job was an equally difficult one. I was asked by the wife of one of my mates to fix their TV set. The only problem at first glance was that it was a 76cm job and weighed a few tonnes. But it was worse than that - it had an intermittent fault which meant that it really should be tackled on the workshop bench.

Unfortunately for me, my friend (an ex-technician who saw the light and made a successful career change some years ago) was overseas on business. So there was no way of moving this set to the workshop.

The set was a 1996 Chinese built Teac CT-M761ST and it had a weird intermittent video fault that varied the colour, brightness, contrast and definition. Being intermittent, it couldn't be made to perform to order and though I did see the fault in action, I really had no idea where to start.

However, after careful consideration, I decided on a strategy of replacing all the electros in the power supply and those on the 210V rail to the video output stages on the CRT board. Hopefully, this would cover enough likely suspects to catch the elusive fault. And so, one clear afternoon, I called around with a service manual and the five electros I intended to replace.

After removing about 50 screws, I took the back off and found the chassis moved out with the release of four more screws - but where could I put it? There is no service position - if I needed to do any work underneath, I would have to hold the chassis up with one hand and solder with the other. But how does one unsolder using solderwick or a solder sucker without a free hand?

Anyway, I eventually managed to replace the five capacitors - C7114, C717, C735, C713 & C616 - and switched the set on. Everything came on correctly and all was looking fine, so I switched the set to standby with the remote control and replaced the back.

However, when I switched it on again, there was no picture. I assumed that I must have done something silly putting the back on, like tearing a lead out of its socket or something. I removed the back and checked everything but the set just wouldn't cooperate. I could hear a varying "rustling" sound coming out of the power supply, the audio was fine, there was EHT and the CRT filaments were lit - but there was no picture!

Clearly, this was not a problem that I could solve on the spot. Apart from the humiliation of looking totally stupid in front of my mate's wife, I would be behind schedule with my remaining jobs for the afternoon if I persisted. The only thing I could do was take the chassis back to the workshop.

Back on the bench, I examined it carefully. There were a few doubtful joints but nothing significant. The new fault had to be due to something I had done, as it wasn't there before - unless some really unlucky coincidence had occurred.

I checked the polarity and values of the electrolytic capacitors I had replaced. They all had higher voltage ratings and lower leakage characteristics than the originals. Perhaps one was faulty?

Another five new capacitors was a small price to pay to solve the problem and it was easy to replace them again. I also examined a few other electros on the main HT rail - C719, C720 & C615 - but all were fine.

Back at the house, I reinstalled the chassis and switched on only to find that the fault was still there. I measured the main 125V HT rail, which I expected to be low (it should be 122V precisely). To my surprise, it was high and varying even higher! There was nothing for it but to take the chassis back to the workshop; this was too hard to tackle in the home.

No messing about

This time, I wasn't messing about (not that I was before). First, I replaced IC701 (TDA4601), the power supply controller. I then shorted the base and emitter of horizontal output transistor Q601 and hung a 100W globe between its collector and chassis.

Next, I reconnected the power and monitored the 122V rail. It started OK but began to rise fairly quickly. I measured the voltage out of the bridge rectifier to be a healthy 330V and changed C726, the main reservoir capacitor, just in case.

I then checked the voltages on all nine pins of IC701 and they were close to those marked on the circuit, the exception being pin 5 ("V" Protect) which I measured at 7.0V (it should be 8.2V). Another thing I noticed was that the rustling sound became worse with the meter on pin 3 (error feedback). The voltage here was 2.1V which was close enough to the 2.0V specified.

Ironically, I am very familiar with this circuit - it is an extremely popular design used in many brands of TV sets. The one I have dealt with the most is the Goldstar PC-04X chassis, so I decided to compare notes with that circuit.

The Goldstar has pin 5 at 7.2V and the most common fault here is C819, a 1μF 50V electro which gives precisely the same symptom as this Teac displayed now. I examined this carefully (C819 equates with C714 in the Teac) but made no progress.

Next, I tried heating and freezing the components around IC701 and found R706, D736, C714 and C709 to be most affected by temperature. Then I began to notice more and more that there were differences between the parts shown on Teac circuit and those that were actually fitted to the set; eg, C712 is shown as 0.1μF on the circuit but a 330pF 1kV unit is fitted. Also, R714 is shown as 10kΩbut a 15kΩ resistor is fitted instead. I found more differences in other parts of the chassis but they weren't particularly relevant to this fault.

So, was the circuit correct or were the parts on the chassis correct? In the end, I decided to go with what was fitted - after all, the set had worked for six years without any trouble until now. I replaced R706, D736 and C709 and even C714 for the third time but it still made no difference.

By now, I had reached the stage where I was drawing out the component layout in an effort to get to grips with it. And it was while I was drawing this layout that I noticed that C714 was shown on the circuit with its positive lead to chassis whereas in the set, it had been installed the other way round. I fitted another 1μF capacitor with the polarity reversed and tried the set again. This time, the 122V rail remained stable and there was no rustling sound - the PC board had been incorrectly marked!

Black cat

Triumphantly, I shot back with the chassis and refitted it but there was still no picture. I couldn't believe this - what black cat had I seen recently? I didn't recall walking under any ladder or being involved in any other such jinx.

But it did occur to me that because the fault was producing a high and rising HT, there could be components that might have been destroyed by the high voltage. I momentarily, shorted the red gun of the tube to chassis while the set was on and a bright red horizontal line appeared - the vertical IC (IC507, TDA3654) had been destroyed!

Once again, I took it back to the workshop, installed a replacement IC, then returned and refitted the chassis. This time - at long last - success! The picture was now fine and a few minor adjustments completed the job.

I told my friend that I still wasn't sure this had fixed the original intermittent fault but to keep an eye on it and let me know. Three weeks have gone by since and all is well but my fingers are still tightly crossed.