Items Covered This Month Sony ST-E200 Panasonic TC-2670S M15D chassis Panasonic TC-63A61 M16M chassis Philips 14PT132A/75R L7.01 chassis Sony KV-3400MAS SCC-C91C-A GP-2A chassis

My next door neighbour Leo likes to reinforce the Old Pals Act (OPA) occasionally and get me to "look at" things that don't quite work right. Leo's no slouch for a bloke of his advanced age and hearing, fancies himself as a bit of an engineer and tends to "have a go".

Anyway, one Sunday morning while letting the cat in (at 7am!) - and still in my pyjamas - he asked me to have a look at his Sony ST-E200 stereo tuner. He had already removed the lid and pointing to some transistors (subsequently identified as Q15 and Q16), informed me that the left channel wasn't working because one of these emitter followers was crook.

Some hours later, when I was really awake, I humoured him by looking at it in proper focus. This also meant opening the shop to get the circuit - oh well; I hadn't anything else to do. Regretfully, this ruined my plans to go to church and Mrs Serviceman was unimpressed with Leo's tuner in pieces all over the kitchen table, especially with lamb roast on the menu.

Getting down to details, transistors Q15 and Q16 are conventional muting transistors in parallel but one could cause this symptom. I could confirm this by just disconnecting the collectors but I felt it would be easier to use an audio probe and see if the audio was coming out of IC41 LA1835 (detector and decoder).

I confirmed this; audio was OK on each channel - on pin 21 (R ch) and pin 22 (L ch). From here, each goes to one of two low pass-filters - LPF42 (R ch) and LPF41 (L ch). Audio goes into each LPF on pin 2 and comes out on pin 4. The audio was OK on the input and output of LPF42 (R ch) but was absent on the output of LPF41.

The DC voltages were the same on each and I could see no harm in patching pins 2 and 4 together on LPF41. I now had output from both channels which meant that LPF41 was almost certainly open circuit.

Well, I could have given it back to him like that and I doubt whether he would have heard the difference. But I knew he would have been unhappy with a "bodgie" fix and would have interrogated me as to how I had fixed it. So a replacement LPF41 went into the ordering process first thing Monday morning.

Postscript: believe it or not, this part was not available in Australia and had to be ordered from Singapore. It arrived the next day! The set was soak tested and returned to its owner the next weekend.

Well, that really was an easy one.

(Editor's comment: whatever happened to the policy of changing names to protect the guilty? I'm not too happy about this, especially the comments about age. Life's pretty crook if you can't have a little service job done without having it written up for the whole world to read about it ;-)

A shame to work

The following Monday was a great morning. The sun was up in a cloudless sky and even the kookaburras were telling everyone what a glorious day it was and how good it felt to be alive.

I contemplated over my coffee that it was a shame to have to go to work instead of going to the beach - especially as there wasn't really much on. (At work I mean, not at the beach).

I felt that today would be a good time to sort out some old stock and fix up some odd jobs left lying around. And so I turned to a couple of Panasonic TV sets that had been put aside for a very long time and needed investigating.

One was a 1990 TC-2670S M15D, which had been abandoned due to an horrendous intermittent fault of no video when hot (it was invariably OK when cold) and no sound.

The other was a 1992 TC-63A61 M16M, which I had been given a couple of years ago with the picture tube (M63JUA07X) having no red. Its cathode was completely stripped and I'd had no luck in obtaining a replacement because 63cm sets are no longer a popular size - it's now usually either 59cm or 68cm. The other problem was that the tube is an FST (Flat Square Tube).

Well, today was the day I was going to resolve these and hopefully other long standing problems.

The plan was to look at the older M15 and see if I could fix it with information I had acquired since abandoning it. If that turned out to be unsuccessful, my strategy was to try to fit the M15 picture tube into the later M16 set.

The audio problem was easily fixed. There was no 18V at IC2301 pin 1 AN7158N, due to R827 (0.22Ω) being open circuit.

It took longer to fix the picture failure when hot. I started with the usual suspects, namely R525 (100kΩ - check this out of circuit or waste a lot of time) and Q601 (UN1111, DTA114EA), but these were OK. The latter, by the way, can be replaced with a general purpose PNP transistor such as a BC558, with 10kΩ in the base and 10kΩ from base to emitter.

In the end, the fault turned out to be just faulty joints and some corrosion. The picture wasn't bad but the 12-year old set was corroded and dirty and very much the worse for wear.

Finally, I made a decision - I would scrap this set (with a good working chassis) for parts and transplant the tube to the later model set.

Removing the tube

Removing the tube was relatively easy, except for the time needed to remove the case. The tube is so large and heavy and the cabinet is very light and tends to move. Also, there is very little room for one's fingers to get a grip around the rimband. I put a plastic lunchbox underneath to take the weight of the tube as it came away from the four studs and the case supports. I could then rearrange my grip and get a better purchase.

I made the mistake of trying to put the replacement tube in the M16 cabinet face down instead of leaving it vertical (the empty cabinet was just too flimsy and kept moving). The result was that although I fitted the tube in, it was 12mm out but stuck fast. When I eventually moved it into the correct position, it fell heavily those last 12mm onto its face but a quick inspection later on indicated that everything was probably OK.

I was going to replace the deflection yoke but the one already on the tube looked identical to the original. Not only that but the plug fitted straight into the DY position. Also, the wire colours matched. I refitted the chassis and after a final check, switched the set on.

It tried to come on but then stuck in standby. What could I have done wrong? Well, lots as it turned out.

I removed D560 and switched on. Smoke rose out of the power supply board (D). It took a couple of hours to work out what I had done. As it turned out, the M15D deflection yoke (Part No: TLY15459F) differs from the M16M yoke (Part No: TLY15493F). Not only that, the DY plug is wired differently despite the fact that the wire colours are the same.

As a result, high-voltage horizontal pulses were injected into the output of the vertical IC (IC452, LA7838), destroying it and causing R570 and R451 to overheat and smoke. At the same time, additional current flowing through R578 in series with R570 caused Q557 to switch on and so the protection circuit was switched on.

But this was only the beginning. It took two ICs to find out that D460 (MA4360M), a 36V zener, was also short circuit.

Finally, after it was all fixed and the set was switched on, there was only a blurred smudge on the screen, a noise and the smell of burning. This turned out to be a fracture suddenly appearing in the tube neck under the yoke, with arcing everywhere. The tube was kaput!

I had waited a couple of years to replace this tube and I was now determined to fix the set. I tried all my colleagues (again) and was lucky to find one who had just scrapped an M15LW chassis, so I took the tube back and fitted it into the M16M, this time with a lot more care.

I was also careful while transplanting the deflection yoke and in fact everything associated with the tube. In the process, I removed all the old parts, including the chassis strap and degaussing coils, and put them on a shelf above the set, behind some aerosol cans.

The new tube worked well but when I reached up for a can of CRC 2-26, the old chassis strap got caught and fell straight onto the M16M chassis beneath. Unfortunately, it was switched on at the time and there was a small spark as I dashed for the main power switch.

After removing the offending chassis strap, I tried to work out what damage had been caused. The strap was nearly half a metre long and could have touched the righthand loudspeaker as well as the high voltage power supply.

On switch on, I confirmed there was now no audio and R819 (0.47Ω) in the 25V rail was open circuit. Replacing it caused the new one to smoke very quickly so I replaced the sound output IC (IC2301 MC13500T2) but there was still no sound and R819 was still getting hot.

Next, I measured the rail and found that it was only 14V when it should have been 25V. However, when I desoldered pin 9 of the audio IC, the voltage rose to its correct value.

I had fitted a TA8200AM audio IC instead of the MC13500T2 which should have been OK. So was there any modification required? I spent a lot of time on this red herring before I woke up that the lefthand loudspeaker voice coil was short circuit - not the righthand one, as I had suspected.

Unfortunately, the speaker is a custom 8Ω 10W oval type, so it will be difficult to get an exact match. In the meantime, I've substituted a similar unit which fitted after a session with the electric drill.

And what is the moral of this story? If it starts out to be a beautiful day and if the planned exercise looks easy, it can still end up like the pits.

Philips portable

I don't get many portable TV sets these days, as they are so cheap to buy now. Occasionally, they do arrive and when a 1997 Philips 14PT132A/75R (L7.01 A chassis) came in with no audio, I assumed it would be a simple fault. I mean, a TV set with no audio - how hard can that be? There was no audio from the loudspeakers or earphone socket on either TV or AV input.

This particular set also had Teletext and "SMART" controls which give a selection of preset Picture and Sound positions to enhance the type of program selected.

To get into the SDAM (Service Default Alignment Mode), I needed to short M24 and M25 and switch on. However, there were no error codes. Standby leaves this mode.

I measured loudspeaker continuity right back to pins 6 and 8 of IC 7120, a TDA7065B 3W audio amplifier. I measured 16V (Vcc) to pin 2 and an audio probe showed that audio was getting to the pin 3 input. That only left pin 5, the DC volume control - it had no voltage on it at all.

With the set switched on, I connected my old analog multimeter (on the x 1 resistance range) between pins 5 and 4 (chassis). Suddenly the room was filled with sound.

The DC volume is controlled by pin 2 of the microprocessor (IC7601, SAA5290ZP/072) and is connected by R3630, a 10kΩ SMD resistor. However, there are 11 other components hanging off this line, mostly diodes, capacitors and resistors.

Using a digital multimeter, I measured the impedance to chassis and it read over 3MΩ. So no short circuit, I thought. That left 11 components to check apart from the two ICs - a piece of cake.

First, I changed the audio output IC (it has only 9 pins), in case an internal diode protection circuit was dragging the voltage down. It wasn't.

Next, I measured 1V on pin 2 of IC7601 but nothing on the other side of R3630. The voltage on pin 2 varied from 0-3.3V, depending on the volume setting. I checked R3603 from the 5V rail to pin 2 as 8.2kΩ and R3630 was correct at 10kΩ.

I then disconnected the three diode clamps on the line and replaced the two electros. I also measured the remaining resistors - everything checked out OK. So where was the voltage being held down?

The remaining components were all surface mount devices (SMD). They are not only very small but are also glued to the PC board, making them difficult to remove. I tried heating and freezing them to see if any responded but nothing showed up.

Finally, I desoldered pin 5 of IC7120 from the board and soldered this pin (only) to a jumper lead connected directly to pin 2 of the microprocessor. Finally, I had voltage and sound which was now controllable.

That led me to my chief suspect - SMD capacitor C2121 (0.22μF). Even though I had checked the resistance of this capacitor several times in circuit, I now found after removing it that it measured only 586Ω. This was definitely the culprit but why didn't it measure low before?

One possibility is that this device might have been intermittent and only showed its true - and permanent - leakage after the stress of being removed.

Or was I deceived by the meter? Digital meters are fundamentally more accurate and have both advantages and disadvantages compared to analog types. One advantage is that they can often accurately measure component values in circuit, since the voltage they apply is too low to switch on active components such as diodes and transistors.

On the other hand, the measuring voltage from a digital meter is too low to place the component being checked under stress. Older analog meters use 9V or 22V batteries and can detect leakage better because components are being checked under more realistic working conditions.

A big Sony

An 80cm Sony TV monitor was dropped in for repair. I say "dropped" but I don't mean it literally - this set weighs in at 74kg and something would certainly have broken if it had really been dropped.

I had never previously seen this model - a KV-3400MAS SCC-C91C-A, using a GP-2A chassis. It was owned by a recording studio and someone had obviously had a look at it, as some boards were loose inside and some metal covers had been removed. All I was told was that it was dead and they asked if I could "please, please fix it ASAP". Fortunately, a service manual was supplied with the set.

The section where the covers had been removed was the power supply (F1) and I could see that all the electros had been replaced recently by someone who knew what he was doing; the work was clean and the soldering good. However, after a few resistance checks, I confirmed that PS653, a 2.7A IC link fuse that looks like a two-legged transistor, was open circuit. This is the source of the 15V B line going to CNO20 (pin 1) on the D Deflection Board and on to IC503 (STR90120), a large switchable 12V IC regulator.

With all the modules plugged back in, the symptoms at this stage were that the set would try to come on, with red LED D41 on the front indicator panel H coming on for a few seconds; otherwise the set was dead. There was no sign of distress anywhere else, which I thought might have caused the blown fuse.

My approach now was to follow the two rails, looking for something that was drawing too much current. I disconnected the modules one at a time as I progressed and when I disconnected the B Chroma Decoder board, the set fired.

Elated, I followed the 12V everywhere on the B board, desoldering devices everywhere in an attempt to find out what was causing the set to die. Finally, I reached the Comb Decoder IC (1310) and disconnected the 9V rail (via Q1360 from the 12V rail) to pin 1. Once again the set fired. However, after spending a lot of time up and down this path, desoldering and resoldering everything, I finally realised that the fault was intermittent and wasn't due to anything on this board.

It took a while to work out what was going on. Something was activating the protection circuit and switching the set to standby. Because of the PS653 fuse failure, I had assumed that it was a short circuit on the 12V or 15V rail that was responsible but that turned out to be a red herring.

I established that there was 135V on the horizontal output transistor (Q804) collector which was correct but the 135V on the horizontal drive transistor (Q805) was not right. Instead, it meant that this transistor was cut off and I found that it had no horizontal pulses coming in from the jungle IC (IC501, TEA2028B). However, the oscilloscope showed me it was trying to come on momentarily.

Among others, the circuit that interested me was the safety circuit involving Q806 and Q807 from pin 4 of the horizontal output transformer to pin 28 of IC501. Transistor Q806 detects the current flowing through R843 to the horizontal output stage and if it's excessive, applies voltage through inverter Q807 and switches off the oscillator in the jungle IC.

It was here that I found the circuit in the service manual was quite different from the one in the set. In particular, R842 turned out to be a zener diode but it checked out OK.

The safety circuit wasn't being activated, the collector of Q806 was correct at 0V and Q807 seemed correct too. Desoldering pin 28 of IC501 restored the picture. So what was happening between that and the collector of Q807? I traced the circuit and found it went near a mounting screw. Careful examination revealed a hairline fracture across seven PC tracks and it was this that was causing the intermittent fault.

The fracture was extremely fine and the tracks extremely narrow, so it took a lot of effort to solder tiny links across them. I was just completing this when I had an attack of stupidity. I had been using solderwick on and off on all the work I had been doing and when not using it, I had just placed it on top of the power supply which still didn't have its metal covers replaced.

When I had finished, I tidied up and removed the tools and reassembled the boards - but I hadn't noticed the solderwick.

I switched the set on and there was an almighty bang from the power supply. I immediately switched off. It was fairly obvious the copper wick had shorted out the power supply but what was the extent of the damage?

The violence of the short had melted two welts on the big aluminium heatsink. I removed the power supply and checked it all carefully - but amazingly could not find any fault. The copper braid had shorted the live side of the chopper transformer to the chassis side and the main circuit breaker had saved the day.

I reassembled everything with much greater care and switched on. This time the set came on properly. I connected video and audio and the picture and sound were good.

So that was it - but I was lucky with the solder wick incident.