Items Covered This Month Sony KV-T25SZ8 TV set (BG-1S chassis). Sanyo VHR-190 VCR. Sony Playstation. Mitsubishi CT-2804AST (ST) TV set (ASV664 chassis).

That preamble was inspired by my first story this month. It concerns a Sony KV-T25SZ8 with a BG-1S chassis, which came to me from an agency. This model is only three years old and looks brand new except perhaps for it’s styling – it is black plastic instead of the current silver look.

The complaint written on the service card said "dead". Well, it wasn’t; at least not completely. Switching it on, the set fired up and with an antenna connected, even gave forth sound – until it cut out a few seconds later. Using the RM870 remote control, it could be switched on again immediately and the stations se­lected but it closed down again shortly afterwards. By doing this a few times, coaxing the tube to warm up, I managed to get a white line at the top of the TV screen but it still closed down.

I was discussing this with a colleague as we shared a brew of morning coffee. He reckoned the fault was a flashover in the picture tube – one of a batch of tubes used in this model, and which was supposedly faulty. I thought otherwise, having had a similar set give the same symptoms. That set was a KV-G21S1 with the same chassis, and the cause was the vertical output IC, IC551, and pincushion control amplifier, IC801.

Well, we argued back and forth but the set wasn’t being fixed and, as my coffee was finished, I whipped the back off and had a poke about. My friend kept reminding me that this was a well known fault – but that didn’t explain the line at the top of the screen.

I pulled out the file on the BG-1S chassis to find the circuit for this model. Unfortunately, I had only the one for the KV-T25SF8, as well as for the set I had fixed before. The first thing I noticed was that the jungle chip IC300 was now a TDA8375A instead of TDA8366N3D but pin 50 was still the EHT X-ray protec­tor, which I unsoldered.

This allowed me to switch the set on without it closing down – but I had to keep an eye on it. If something serious was causing it to close down, this was now free to vent its wrath on associated components! On the other hand, this gave the tube a chance to warm up properly and I could see that there was about 100mm of somewhat distorted scanned raster at the top of the screen. But there was no sign of distress – heat, smoke, or whatever – from the rest of the set. Whatever was crook wasn’t effecting anything else at present.

I followed the X-ray protection line and found that, by desoldering zener diode D1505 and resoldering pin 50, the set would remain on but with the same fault. This suggested that the fault was in the vertical deflection and correction circuits.

I measured the main supply rails to IC551 (LA7830) and IC801 (5PC4558G2-E – surface mount) to find them spot on at -13V and +15V, so I confidently ordered these two devices on spec.

I fitted the parts when they arrived, one at a time, only to find that they made no difference.

"I told you so", my friend said, when I acquainted him with this development.

"OK, you fix it, big shot" I retorted.

"Well, all right – let me have a squiz at it".

There were a lot of "ers" and "ums" coming out of him for the next half hour. That will keep him out of trouble for a while I thought, while I went on to the next job.

But give him his due, he solved the problem in one go – and within that half-hour – but he did have to bring in the heavy artillery; ie, the oscilloscope.

And he now conceded that a flashover seemed unlikely, since the set continued to function without any sign of sparking. His first step was to measure the voltage on all the pins of the vertical output and pincushion ICs and he found them to be correct.

So the CRO was brought in. This revealed that there was a nice clean vertical sawtooth on pin 46 of the jungle IC, the vertical positive output, but nothing so clean was arriving at IC801 or IC551. Following this path it appeared to go crook at D315, a 9.1V zener. Removing this established that it was leaky. Replacing it fixed the problem.

I have to admit that the man appears to be a genius – give him a beer! (Of course, I had done all the spade work first. He was just dotting the i’s and crossing the t’s. Ahem!).

Now, what was that I was saying about a sheep’s head?

The Sanyo VCR

My next story involves quite a change of scene. Tony Black came in looking as white as a sheet. He was clutching a VCR but somehow I didn’t think that was the cause. Tony is a young farmer who lives at the back of Woop-Woop and drives an extremely dodgy Kingswood wagon.

"What’s the matter?" I asked. "It’s those truck drivers," he finally gasped. "What about them?" I enquired. "I’ve just had a whole lot of them monster me up the highway. I was driving at 110km/h and they were still right on my rear end all the away".

I soothed him down and carefully removed the VCR from his white knuckles. Obviously, the truckies had scared him, as they do everybody, and having come this far, it would have been a pity to drop the VCR, a Sanyo VHR-190.

After a suitable interval, I asked what was wrong with it. As he described it, it died every time a tape was put into it. I told him to leave it with me and I would check it out. I didn’t know what to suggest about the speeding truckies; his story reminded me very much of Spielberg’s "Duel" sequence. A newer, faster car might be one way but if one is at the speed limit and the truckies drive well over the limit, what does one do? Pull over and let them pass, I guess.

Back at the bench, I removed the covers, and could see his description of the fault to be quite accurate. This model has been quite successful and employs the Sanyo P88 deck but this unit is getting on a bit now.

I didn’t have the service manual for this set and was disinclined to buy one because of its age but the problem looked to be a power supply fault, probably due to leaky electros and a low power rail.

I could see that the loading motor was working but I made a note that I should change the belts anyway. I spoke to Tony about it, emphasising how old it was and, with new ones starting at less than $250, it was hardly worth it. Anyway he started bleating about needing a new car that he couldn’t afford and could I please see what I could do?

How do I get myself into such situations? I had listened to him whinging about the truckies and had become distracted.

To tackle the power supply, I would really need a circuit. I could buy one of course but hopefully I would be able to beg, borrow, or steal one from the opposition.

In the meantime, I needed replace those two belts anyway. The one underneath the deck was easy but the top one was diffi­cult, not only because access is poor but because the loading gears have to be dismantled and can easily be reassembled incor­rectly, giving horrible timing problems and jamming the eject mechanism completely.

Firstly, the PC boards, which partly cover the loading motor assembly, have to be removed. Then, making sure the deck is in the full unloaded and ejected mode, you remove three small screws, release a white plastic cam and finally remove the squirrel gears. (It is a good idea to mark these with a felt tip pen to ensure they are reassembled in exactly the same position). The next problem is to remove the belt – to save time, I just cut it. To fit the new one over the lower pulley is fiddly and a tight fit. Wetting it with metho can sometimes make it easier.

When the belt is on, one has to reassemble the rest in the reverse order, making sure the ejector hasn’t moved and is still in its maximum eject position.

Fortunately, and probably due to experience, everything went well until I tested it. Initially it wouldn’t work at all but this was because the belt was slipping over the metho. When it evaporated and the belt gripped, not only did it load and play but also it performed all its functions without stopping or turning off.

What a bonus – the power supply didn’t need attention after all. Presumably, the stretched belt was slipping and, as a result, the loading sequence was not being completed within the required time and the system would shut down.

When I told Tony, he was delighted and now has a few more pennies towards a new car.

Sony Playstations

And now a few thoughts on Sony Playstations. I have had several through the workshop over the last couple of years, initially for modifications to play overseas games.

As the supply of chips has dried up and newer models are now available, the emphasis has been more in repairing the power supplies, in particular the SCPH-700Z. More often than not, the repairs have been straightforward in this simple switchmode power supply. If fuse F001 is blown, for example, the problem is normally confined to a diode in the bridge rectifier; eg, D002 being short circuit. Replacing it with an 1N4007 usually solves this.

More elaborate failures will sometimes take out the FET. If the power supply looks otherwise immaculate and it won’t go, switch it off and check for 325V on the main electrolytic. If the voltage is still there and not decaying rapidly it can be assumed the start-up resistor R003 (820k) has gone high. Make sure that you discharge the capacitor before replacing R003.

Mitsubishi TV set

Mrs Thompson’s Mitsubishi 1988 CT-2804AST (ST) ASV664 chassis had ceased to function after a power surge from a recent storm. Unfortunately, she didn’t have household contents insurance and was also an old-age pensioner. It was my task, therefore, to resurrect her dead set for as little as possible. It also had to be done in the workshop because, when I called, I had noticed there were a few burnt patches of components on the separate power supply board, implying a fairly extensive failure.

The power supply is a dual one – there is a main switchmode power supply providing 115V and 18V, and a secondary one employ­ing a small 50Hz transformer (T9K1), providing 5V, 30V and 14V rails. The secondary supply has to operate first in order to activate a relay which operates from the 14V rail.

When the relay closes and the set fires, the relay voltage is supplemented by one from the horizontal output transformer (pin 4 of connec­tor PE, which mates with connector PF on the selector board, and the collector of the relay-drive transistor, Q702).

But the relay wasn’t working, simply because there was no 14V to operate it. In fact, none of these rails was working. And, in turn, this was because the primary of mains transformer T9K1 was open circuit. This is a small PC-mounted trans­former which, unfortunately, is no longer available as a spare part. The secondary supplies a voltage doubler via R9H1 and C9H1 to the collector of Q9H1, which is a voltage regulator delivering 30V. This secondary winding also applies 12V to the collector of Q9H2 via D9H4.

So what was the secondary voltage supposed to be? It had to be at least 12V but it could be higher. Eventually, I settled for 12V and bought a conventional 300mA trans­former, which mounted well where the old one had been. I also replaced a few dodgy looking electros on the secondary and I now had the voltage for the relay and could switch on the other power supply.

This now produced the 18V supply but there was no 115V rail from the emitter of Q9A6. That left only the area around Q9A6 and Q9A7. And this was one of the areas that was burnt. Resistor R9E1, the two transistors, three electros and diode D9D1, all needed replacing

Once this was done, the whole set fired up and there was picture and sound.

Everything appeared to be fine, except for the -30V rail, which was down to -22V. I then found that the voltage on the collector of Q9H1 was low at the same voltage; ie, -22V. I spent some time checking all the parts in this circuit before concluding that perhaps my choice of power transformer hadn’t been the best.

I phoned technical support at Mit­subishi to find that the secondary of the transform­er should really be 24V and not 12V. However, if I substituted a 24V transformer, it would no longer fit on the PC board and it would be a messy repair. The other two rails and associated circuits were functioning correctly and I pondered what to do.

In the end, I decided to leave it as it was – the -30V rail is used only for memory on IV702 M58659P (pin 2) and the set was remembering everything perfectly so I figured it was best left alone.

Mrs Thompson was relieved to have her set back, and the bill wasn’t too heavy.