Items Covered This Month Philips KR638 (KL9A-1) Panasonic SA-PM01 stereo system Philips 29PT6361/79B A10 chassis Philips 21PT3822/79T (L9.2A chassis, BZ production) Sony KV2185 S (G3E chassis) NEC FS-6807S Panasonic TX-33V2X (M16 chassis)

A well known audiophile technocrat lives not far away from our workshop and owns an ancient Philips KR638 (KL9A-1) with a wooden cabinet (do you remember those?). He complained of an intermittent "no-sound" problem which he and his family had been putting up with for nearly two years. This was because it had taken this long to build an extension to his house, which is at the end of a long steep drive and we couldn’t arrange to pick the set up until the concrete had been laid and dried.

Now we had all been watching this development with great interest at the strange goings on. For example, there was the time he mowed the grass beautifully just before the builders arrived and heaped all their junk on it. As for us, we concluded that the main reason it took so long to pour the concrete for his drive was so that he could lay "jumbo cable" in conduits underneath it. But was this for sound or concrete reinforcement? As mere bystanders, this seemed to us to be taking "audiophilia" to extraordinary lengths.

Anyway, the situation with the telly finally became so bad that the only sound they could hear was through the headphone socket and then only on one channel. As a result, only one person in the household could listen at any one time. It was a case off "first in, best dressed", with the rest of the family left to lip-read! This finally forced them to do something about the problem and so the set eventually arrived at the workshop and a loan set despatched.

Personally, I am very fond of these particular sets, most probably because of their ease of access. Anyway, it didn’t take too long to remove the rear quarter panel and slide out the audio amplifier board, which revealed monstrous dry joints on all the audio output ICs. Resoldering them restored the sound but the picture wasn’t too good, with intermittent lines and poor focus. Examining the main board revealed many more dry joints, especially around the focus control.

Having attended to this, the picture was now quite good and I let it soak test before putting the back on. I then switched the set back on again and gave it a whack on the top, just to reassure myself that all was now robust and sturdy. Unfortunately, this hearty slap on the cabinet murdered the set, so I removed the back and found I had blown the 2A mains fuse.

Well, how do you do that with a "bit of a whack"? It turned out that the degaussing dual posistor was also dry-jointed and the vibration caused it to draw more current than the fuse could stand. Resoldering it and replacing the blown fuse finally fixed it properly.

In the meantime, our audiophile decided that the loan set, some 15 years its junior, really did give a better picture and so he bought it from us. The old Philips has been relegated to the rumpus room but at least it now gives stereo sound and you don’t need headphones.

The question on everyone’s lips is does the concrete driveway have 5.1 Dolby surround sound? And where is the sub-woofer buried? And will this help with the parking?

A gorgeous model

One day, this gorgeous-looking model glided into the shop – and I’m not talking about the Panasonic stereo she was carrying. She asked me whether I could fix it for her and I demurely replied that I would fix anything for her if she so wished.

At that point, Mrs Serviceman shoved me aside and rudely (I think) monopolised the conversation, pausing only occasionally to glare at me. At the time, I thought I was just being polite but she (Mrs Serviceman) later informed me that my tongue was actually dragging along the floor. There is no justice in this world!

Unlike her (the model), the Panasonic unit was a middle-of-the-road SA-PM01 stereo system and she was complaining that the CD door clunked loudly when opening and closing. She had taken it to two other service companies who had had it for a long time and now the CD didn’t work at all. Well if you saw her, you’d want to have her – er, I mean it – for a long time too.

Anyway, under strict supervision, Mrs Serviceman made sure I was going to do the right thing at all times.

When I had stripped it down, I found that someone had taken an
awful lot of time and trouble to install the flexi-cable the wrong way around into the connector (he too must have been distracted) and this was why the CD wouldn’t work.

The original problem was due to broken gears on the lid assembly, which is a recognised problem with this model. Replacing five components in the gear assembly usually fixes it for good (Part Numbers: REE0880, FKF0561-K2, RMA1291, XTV3T6JF2, RDG0473).

I was all for personally returning the unit to its owner after hours but unfortunately, the thought police scuppered that idea as well.

A real mystery

One of my main breadwinners is repairing Philips A10 chassis TVs and I have written a lot in the past about their "Small Signal Panel/Board". The other day I had an interesting side event involving a 29PT6361/79B and a Mrs Evans.

This set – in an unusual black cabinet set (early production?) – came in for the usual intermittent "no remote" and "dead" symptoms. The repair was quoted for and accepted, and so an exchange module was ordered and fitted. The set was then aligned and tested before being delivered back the customer.

However, the twist was that Mrs Evans phoned a few days later to report that all was going fine, except that (and there always is an exception) the "Rotation" control in the Features menu wasn’t working. Apparently, the control could be adjusted with the remote via the on-screen display but nothing was happening to the picture.

I have to admit that I didn’t even know about this feature on this particular model – it’s normally only fitted on the 34-inch version. As you’ve probably realised, it’s designed to correct for the effect of the Earth’s magnetic field on the picture tilt (the effect is more noticeable on large screen sizes).

I decided that the best place to tackle this problem was back at the workshop. And sure enough, Mrs Evans was absolutely correct and very knowledgeable – the picture would not tilt at all, although the on-screen display said it should be doing so as you pressed the remote.

So was this a microprocessor/EEPROM software problem, or was it hardware? I checked for error codes but there were none, so I examined the hardware to see if I could find anything visible like coil plug 0252/0235 and 0234 being left unplugged when the chassis was refitted.

To my amazement, I discovered that there was no rotation coil fitted and neither was the necessary accompanying circuit installed at all in the set. What’s more, no-one had taken it out as it was just never manufactured with it.

At this stage, I decided to go back into the SAM (Service Alignment Mode) menu and check the option codes. Unfortunately, though, Philips seem to no longer stick the option codes list inside their sets, so now you have to look up the Product Survey for the year the set was manufactured (2000). In fact, the PS2000-4 Yearbook is the best place to go for this latest information.

This laborious process, which is really only available to dealers, revealed that the options should have been 123 253 255 159 and 214 for this particular model. However, this set’s option byte value OB5 was 246 (for the 34-inch) instead of the published 214 (for the 29-inch). Going through the option code menu item by item, I eventually found what I was looking for. Option Abbreviation "ROTI" (Rotation) was ON (=1) and this option bit 6 gave a value of 32. Switching it off produced the correct option byte value for OB5 of 214 (214 + 32 = 246).

In my opinion, the way Philips has presented this in their service manual is very confusing, especially as their list on pages GB78 and 79 seems to be in the wrong order (in particular OB4 and OB5). Anyway, this was the easy part; the hard part was explaining to Mrs Evans that she had in fact never paid for nor received this particular feature, even if it is shown on page 27 of her instruction booklet (though to be fair, it does say "Note: applicable to certain models only"). If only Philips wouldn’t produce generic versions of this covering more than one model.

Obviously, when the Small Signal Panel was replaced, this feature had been left inadvertently enabled – most probably for the 34PT6361/79R with more features.

Murphy’s Law

Mrs Rodger’s brought her 1999 Philips 21PT3822/79T (L9.2A chassis, BZ production) in at a particularly busy time, when the usual people who book the work in were otherwise engaged. As a result, the fault stated on the job sheet was simply to the effect that the set was dead.

After removing the back and checking the fuse, it didn’t take long to find that there was no output from the power supply. A quick check then showed that diode D6550 (BY229X-600) which feeds the main supply (V_Batt – 95V) was short circuit. Unfortunately, this being a flat pack diode, I had to wait until a new one was ordered in.

In the meantime, I was also repairing another similar Philips TV with a faulty flyback transformer. To replace this, the chassis had to be removed to gain access to the PC board tracks.

There aren’t all that many plugs to be reinserted afterwards when reinstalling the chassis but they are all unmarked. Usually, however, it’s just common sense – most plugs marry up with the correct socket due to orientation, size or colour. However, when I switched this set on, it was now inexplicably dead.

It took a long time to figure out that the cause was the top control panel and by unplugging it, I could make the set work. The control panel is a very simple assembly with a collection of four microswitches and a 3-core cable connecting it to the main chassis. How could this cause the set to go dead?

Well, it took nearly an hour for this old idiot to realise that there was in fact nothing wrong with the assembly – I had inadvertently plugged it into the wrong socket. Once this mistake had been corrected, the set worked perfectly.

Getting back to the first set, the replacement BY229X-600 diode arrived in due course and I was annoyed to find that it made no discernible difference – the set was still dead. As a result, I spent a long time checking out this TV from stem to stern.

The flyback transformer and transistor were OK with no shorts and there were no shorts on any of the other voltage rails. I then discovered that power was getting to FET 7518 and IC 7520 (MC44603A) but the start-up voltage was low.

I checked almost all the components in this simple switchmode power supply – admittedly in-circuit – but everything appeared to be OK. I then began to get desperate and replaced the IC and FET but it still wouldn’t fire up. I even ordered and replaced the chopper transformer 5545 in case that was shorting inside but again it made no difference!

By now, I was running out of ideas, so I decided to start from scratch again and check all the components out of circuit. And this time I did find something – diode D6540 (BAV21) was leaky. I replaced it and at long last there was voltage coming out of the power supply – but otherwise the set was still dead.

With the power supply now working, I went back to the line output stage and found that transistor Q7400 (BF422), the line driver, was open circuit. I replaced it but still no joy. You would have thought by now that I would be at least a little closer to a result but this wretched set was going to fight me all the way.

Next, I discovered that there was a problem in the start-up circuitry which was preventing any horizontal drive to Q7400 (via Q7608). This latter transistor was switched off by IC7607 NE555D.

Mrs Rodgers was by now becoming annoyed at the long delay and I was thoroughly frustrated with the extent to which this set had been damaged. Finally, I got to talk with her and it turned out that this set may have suffered a power surge from a lightning strike during a storm.

Armed with this information, we mutually agreed it would be best to write this set off and approach her insurance company for a replacement and we would keep the old set for parts. And that’s what happened.

Many weeks later, when business was quiet, I decided to give it one more look over. While plugging the set back in, I noticed that in fact there were two places the lead to the side control panel 3-pin socket (0251) could plug into – either position 0267 on the main chassis or 0248 or the module on the lefthand side.

I was now beginning to hear a bell ringing somewhere in the back of my mind and thought it was worth investigating how this should really be hooked up. However, this wasn’t as simple as it sounds. Because of the way Philips like to draw their manuals for each chassis and include all the variants for all the models they make world-wide, it can get pretty confusing.

However the circuit shows the unmarked plug going from 0251, the audio output socket Ext1, to either 0260 on the main chassis or 0248 to the sound panel. There is no note in the diversity table but fortunately 0260 was a 4-pin socket and 0248 was a 3-pin socket on the ITT Audio decoding panel D1. 0267 on the other hand is the data communications socket for an external computer (COMPAIR). Connecting these two sockets causes the centre pin (SCL) to be connected directly to ground, which would be like killing one half of your brain.

Correctly connecting the sockets (0251 to 0248) fixed the problems, the set bursting into life with perfect picture and sound.

Of course, I could blame myself for not remembering having done this before and condemning this set to be written off because of it. Alternatively, I could blame the idiot who designed this ambiguous arrangement where one unmarked plug could easily be plugged into two different locations.

Even giving the same number identification to both sockets would have helped. And as for the service manual, it just confuses the situation even more. Obviously, they have never heard of Murphy’s Law.

I wonder how many other technicians have fallen into this trap?

The harder you try . . .

I find sometimes that the harder you try to do the right thing, the worse the outcome. On other occasions, the opposite applies – you just do a basic, simple operation and everything just falls into place.

To illustrate this, a 1993 Sony KV2185 S (G3E chassis) came in with the complaint that the picture "went black". The client also stated that he didn’t wish to spend much money (who does?) on an otherwise small and old TV.

I have repaired a lot of this series of TV and know that the G3E was not Sony’s best product when it comes to reliability. Nevertheless, the first thing I noticed about this set was that the fault was intermittent (the worst scenario).

My first step was to get stuck into the set and repair a large number of dry joints all over the board. That done, the set performed faultlessly but despite my best intentions, I was not in a position to say exactly where the fault lay.

Satisfied that the problem was solved, I replaced the back and put the set aside to soak test. Apart from that, the door flap that hides the front controls had broken hinges and so a new one was ordered. It took a couple of weeks for them to inform us that the door flap was no longer available, so the client was notified that the set was ready.

He called a week later to pick it up, by which time the set had been running perfectly every day for over three weeks. But three days later, an unhappy client returned the set with the same fault. There was one important difference, however – this time the fault was permanent. Despite our mutual disappointment, I was at least glad to be given a second opportunity to crack the case.

I started by checking all the voltage rails and found them all to be spot on. What’s more, the picture was fully scanned and the On Screen Display (OSD) was present. However, every few seconds, the OSD would appear to roll and retrace lines would appear as though there was no sync. In addition, there was no video output from the jungle IC, either in the TV or AV modes.

A quick check with the oscilloscope showed luminance, chrominance and clean sync pulses arriving at IC301 (CXA12138S), along with the blanking pulses. All crystal frequencies checked out, as did the all the voltages around the IC apart from the RGB outputs on pins 39, 38 and 37.

In the end, there was nothing else to do but change this $70 48-pin high-density jungle IC, which fortunately fixed the fault. After another week of soak testing, it went home to its owner – hopefully for longer than three days this time

But why did it fail after it had been tested for so long in the workshop and then so soon after it had gone home? It just doesn’t seem fair! Usually, such faults are discovered and rectified without any prolonged testing and you never hear of them again. There is no justice!

The elderly NEC

I was asked by Mr Beaumont to fix his mother’s television. He warned me that the set was getting old and she was a pensioner and really couldn’t afford to spend much money on it.

Not being a registered charity just yet, I enquired as to how he defined "much money" to which he replied "about $150". Of course, I am not a soothsayer and can’t really guess what it is likely to cost to fix an unknown TV located at his mother’s house. Instead, I told him that it would cost a service call for him to find out if I could fix it within a reasonable amount, after which she could make up her own mind.

He thought that this would be OK and so I phoned her to make an appointment. At the same time, I also managed to get the TV model number from her.

The set turned out to be a 1999 NEC FS-6807S that was pulsating but otherwise dead. Examining the copper side of the main board (PWC-4277) was difficult in the cramped surroundings but I could see a lot of dry joints, so perhaps I could fix this TV within half an hour.

Before I had even taken the back off, Mrs Beaumont, who never left the room, kept anxiously asking me what the problem was. However, I learnt a long time ago not to say what you think the problem is until you are certain of the facts. Unfortunately, Mrs Beaumont was persistent and after brushing off a few more questions, I was beginning to get frustrated with her for implying that I didn’t really know what I was doing. It’s an all-too-familiar theme – "surely you must know by now exactly what’s wrong?".

Well, there really were lots of bad dry joints, especially on IC501 and IC502, the 5V and 9V regulators, the transformers and power transistors, the switchmode power control IC (IC601, STRF6524) and transistor Q507 (2SB546). Unfortunately, resoldering these failed to fix the pulsating problem and nor did it stop the persistent questioning.

By now, I had been going for well over half an hour. Q507 looked as though it had been getting hot and measured leaky between collector and emitter so I replaced it but still no result.

By now, I was under some pressure, considering the restraints put on me regarding cost. And that – coupled with the never-ending barrage of "surely you know what’s wrong by now?" – made it difficult to think straight.

This set really needed to be in the workshop, where I could get some peace and quiet apart from anything else. By now, I suspected the line output transformer (T502), although I hadn’t eliminated the power supply as I wasn’t sure whether it was a load problem or a supply fault. I also had very little in the way of test equipment with me but I knew that she wouldn’t accept the cost of a workshop repair.

In the end, I decided to have one more go at the power supply and concentrate on the electrolytic capacitors. As luck would have it, apart from the main filter capacitor (CC603), there’s just C621 (100μF 63V) and I just happened to have one in my toolbox.

I soldered the new capacitor in and held my breath while I switched the set on. The set sprang into life and I could breathe again. But that wasn’t to be the end of my problems.

After doing a few sums in my head (I had been there an hour by then), I very calmly explained what the problem was and told her that it would only cost her $100!

I thought she would be delighted with this news but far from it. Instead, I got the full repertoire – "couldn’t I do it for less?" and "I’m and old age pensioner" and "what about less for cash", and "it’s not worth it" and "what about the guarantee?".

I was really quite dumbfounded. The repair was really worth a lot more than I was asking and I had come in at a lot less than their $150 ceiling

By now I had had enough. It was $100, take it or leave it. Her 68cm set was only five years old and worth a lot more than $100.

Begrudgingly, she accepted the cost and made me reinstall the TV and the video and berated me about the reception (which was all my fault) before opening her wallet and taking out a crisp new $100 note from a wad that would choke a horse.

I was only too happy to be out of there. I guess it takes all sorts!

Flickering Panasonic

An 80cm 1995 Panasonic TX-33V2X (M16 chassis) arrived on my bench with the complaint that the picture flickered up and down when cold and cut out when hot.

Of course, when I switched it on, it worked perfectly. Nevertheless, being pretty familiar with this model by now, I quoted to replace the 330μF capacitors (C885/C889) on the 5V rail and just for good measure, the (sometimes) dried-out 0.47μF 160V capacitor (C760) that smooths the horizontal drive supply. I also included the time necessary to resolder any dry joints on the vertical IC (IC451).

The quote was accepted and I did the work required, which wasn’t easy with such a heavy TV set with such poor access. I then left it to soak test but an hour after doing the work, the picture started to show line tear, with ragged edges down the verticals. Brilliant.

It took another hour to find that secondary smoothing capacitor C580 (220μF 200V) was open circuit. This capacitor is on the X Board, which supplies 140V to the line output stage. The large cylindrical capacitor looked brand new but it actually rattled inside its case!

So much for the breaks!