Mr Greenspan won his TV set in a competition, so his perception of it probably differs from that of most customers. Because you haven't actually had to shell out your hard-earned shekels, you tend not to appreciate its true value. As a result, when it has the temerity to break down, your loyalty to it is limited. It didn't cost anything, it doesn't owe you anything, so if it is anything serious . . . well, easy come, easy go.

Items Covered This Month Philips Matchline TV set, Model 36ML8906/00B (FL 1.2AA chassis). Philips Matchline TV set, Model 33CE753/42R (3A chassis)

Mr Greenspan's good fortune was an expensive 1992 Philips Matchline 16 x 9 aspect ratio TV set, model 36ML8906/00B, employing an FL 1.2 AA chassis. The tube is a W67EW5001X42 but to me that is rather meaningless in conveying size because of its aspect ratio. The manual says there are four picture sizes - 21in, 25in, 28in & 33in - but I really don't know which one this is. Its diagonal measurement gives a 76cm (29in) viewable area.

Anyway, Mr Greenspan's set was dead and considering that he had won it and it was now nine years old, he wasn't sure whether he wanted to spend real money on it to have it fixed. I told him I would have a go and spend an hour's labour on it to come up with an estimate. He could then make a decision on that basis.

I removed the back to reveal a large flat split chassis interspersed with a few modules. Access to the copper side of the main board looked particularly awesome and I was so intimidated by the technology that I felt I just couldn't proceed without a service manual. Fortunately I was able to order one.

While waiting for it to arrive, I had a good look at the setup during coffee breaks and while dealing with other sets in the queue. The biggest problem appeared to be that I couldn't reach the copper side of the mother boards while the set was on. This is not the sort of set one can roll onto its side or even upside down.

The chassis could be removed by unplugging it and sliding it back after pressing four locks on either side of both boards. Later, when I acquired the manual, it suggested assembling the two panels with a red clip attached to the signal panel at 90 degrees, though exactly how this was supposed to help access escapes me.

In the meantime, I decided on a course of careful and detailed observation. In fact, no component could be found to have failed visually, and strategic measurements with an ohmmeter confirmed that the set should work. But I did discover a lot of faulty joints all over the power supply and horizontal stages.

Rather than wait for the circuit, I impatiently soldered everything in sight. I also checked for dried out electrolytic capacitors but could find none. Satisfied, I reassembled it all back in the cabinet. I wasn't really very optimistic that I had achieved anything much, so I was really surprised to find that I now had both picture and sound and that even the remote control worked.

However, I soon realised that the picture was far from perfect. It was dark, small and with a tendency to bloom. Nevertheless, I felt that I was probably half way home.

Voltage measurements

Later, when I received the poorly photocopied version of the service manual, the first thing I did was measure the main HT at TP57, the cathode of D6237. I can never understand why important test points do not provide easy access, with a real test point. D6237 is buried in among heatsinks and it is not easy to attach a crocodile clip to it, although I eventually managed. Instead of 141V there was only approximately 105V, depending on the picture tube beam current. Varying VR6370 on the SOPS (Self Oscillating Power Supply) Control Module varied the HT but not by much.

I then spent some time checking all the voltage sources in the power supply and established that all the secondary voltage rails were low. The only exception was that of the separate standby power supply (fSOPS). This gave a higher voltage output than was marked on the circuit. VSTART measured over 20V and is marked 17.5V at TP60 but 5V STBY at TP50 was spot on. I decided that these later observations were irrelevant. And the conclusion was that the control module was the most likely culprit.

The irritating thing was that this control module was shown only as a block diagram. There is no circuit, no part number, nor any reference to it in the chassis service manual. I contacted Philips for more information but no-one had any at all. The only thing I could get was the "SOPS REPAIR KIT 4822 310 31932" for a trade price of $134.09.

Well, it was crunch time for Mr Greenspan. I told him all I knew - which admittedly wasn't much - and the likely costs that might be involved. He sensibly decided to abandon the set and we struck a deal that I could keep the set in lieu of any labour costs I had incurred so far.

I then invested in the repair kit - taking a punt on whatever that was. Two weeks later I received a small box with some components, a list, and a new control module! The ES7028 multilingual list was extremely brief and insisted that I replace all 11 components and the module.

I hate being told what to do, so I carefully examined the circuit and noted which parts they were wanting me to replace.

The first five parts were the main fuse and bridge rectifiers. Then there was another fuse, a chopper transistor, two more diodes and capacitors and a regulator transistor - all of which were working perfectly in this set. Admittedly, a few of these had been modified but only superficially and I
really couldn't see this making much difference.

In the end, I just plugged in the new module and switched the set on. And would you believe it? - the picture was now perfect and the main HT was spot on at 141V.

At this point, I imagined that would be the end of the exercise. However, as a matter of course, I put the set on test and left it. Each day, I switched it on for the full working day and switched it off with the remote control before the master power switch was turned off.

The set baulks

After a few days I noticed that the set didn't want to switch on. By jiggling the power on/off switch several times I could sometimes make it come on and then it would stay on for the rest of the day with all functions working perfectly - including the remote control and main power switches.

Gradually, it became worse until finally I had to remove the back and look inside. Just jiggling the module was enough to get it started again and it stayed on. But the problem persisted; every time I switched the set on from cold in the morning, it baulked. Finally, I could only start it by putting in the old module, switching the set on, then switching it off and refitting the new module.

By now, I really wasn't sure whether it was the new module that was at fault or whether it was another problem altogether. Eventually, after much investigation, I concluded that the new module was faulty and should be returned.

Acting on this turned out to be a drama of its own. First, Philips Spares was in the middle of moving and second, they didn't really believe that the new part was faulty. Eventually, I managed to persuade the Spare Parts Manager to exchange it for another but in view of their move and also lack of stock, I felt that perhaps it wouldn't be a good idea to hold my breath.

He told me to keep breathing and eventually something would be done, so I posted the kit back and prepared to wait.

In the meantime, I regularly glared at the old module whenever I had spare time, trying to guess what the cause of the low HT might be. I particularly looked at the voltage comparator components around transistor 7370 and started to replace some of the surface mounted components such as the 6.2V zener in its emitter.

As the days turned into weeks, I gradually ended up by checking almost every component on this module. The module is actually split into two sections - one section is cold and the other hot and they are connected via an optocoupler very similar to many other Philips TV sets of that generation.

Well, during one coffee break I made an interesting observation and that was to do with the VSTART voltage. (I am talking about the old module now; the new one was still with Philips Spare Parts) This voltage comes into the module and goes to a 5.6V zener diode in series with a diode (1N4148) and then to transistor 7371 (all surface mounted). These two diodes were marked 6373 and 6372 respectively.

I then realised that these two diodes were also identically marked on the circuit as being on the motherboard. At first, I thought this was just a circuit mistake but a check confirmed that these two diodes were actually on the motherboard, so this configuration had been duplicated. However, when I looked at the circuit of the new module, which I had drawn earlier, these two components were missing!

So, to summarise, I had a set with a motherboard carrying these two components plus a module - apparently the original - in which these component were duplicated. But on the second (new) module (now with Philips), these components had been left out and had been replaced by links.

What a mix-up; and what did it all mean? In an attempt to clarify the issue, I replaced extra components on the old module with wire links and plugged it back into the set. And when I switched the set on, the picture was perfect and the HT could be adjusted so that it was spot on 141V.

A riddle

I felt elated that I had fixed the problem but recognised that I hadn't solved the riddle. There was a whole swag of questions still to be answered but it appears that Philips made two (at least) versions of this set: one with the two diodes mounted on the motherboard but missing from the module; and one with the diodes in the module but missing from the motherboard. It is hard to imagine a setup more likely to create a mix-up.

Supposedly, this set had worked properly for years with this module in it - or had it? Had Mr Greenspan always had a small picture but just wasn't fazed by it? After all, there is much confusion among owners as to what size and aspect ratio the picture should be: 16 x 9, 4 x 3, or letterbox. Besides, before January this year, it probably was always 4 x 3 and a 16 x 9 set was really superfluous in 1992. Also, Mr Greenspan may have thought it was normal that he was receiving a 4 x 3 picture with the set in 16 x 9 mode.

Or had someone else had a go at Mr Greenspan's set? Getting a set's history out of some owners is like pulling teeth. I'm not supposed to ask "silly questions" like that.

In any case I can't really raise the matter with Mr Greenspan now, so I can only conjecture as to why a set fitted with the wrong module turned up on my bench.

Finally, why did the new module fail intermittently? It will be interesting to see if the replacement module works when it returns from Philips. In the meantime, the set continues to work flawlessly.

Postscript

Subsequently, since writing the above, two events have occurred. First, a new SOPS Control Module arrived and, initially, appeared to be working properly. Unfortunately, it failed after a couple of days and I may be able to write more about this later.

But was this really a new module or had the previous one with the intermittent fault found its way back to me in the confusion caused by the move. Stranger things have happened.

Second, after spending months trying unsuccessfully to acquire a circuit of the SOPS Control Module from Philips, a colleague finally tracked one down. He is an experienced technician who works by himself from home and found it loose with the instruction book for the set. It is not for an FL1.2 chassis, but actually for an FL1.1S (Part No: 4822 727 6395.1/SV00.wk.9148).

So there we are - but where are we?

Another Philips

Mrs Adams' set was a 1989 Belgium-made Matchline 33CE753/42R (3A Chassis) which she had been trying to get fixed. I only found out later that she had had the set looked at by several technicians who didn't want to know about it. Had I known this I, too, would have been more cautious - much more cautious.

As it was, I was told that it was dead and I was cheaper than the authorised agent. This is always a bad omen - I hate the thought of being contracted just because I am cheap! Still, I take comfort from the fact that the NASA spacecraft components are also constructed (supposedly) by the companies that submit the lowest tenders.

So there I was, at her house, looking at this monster 80kg 84cm TV set, having been given the impression that it had only a faulty fuse - or something equally simple. The set had already been put aside and another placed in its place. My heart was in my boots - I just knew that this was going to be difficult but this is my profession and I need the money.

The back unclipped easily to expose a chassis which must have few rivals for poor access. The whole thing is enclosed with concealed plastic catches and supports. All I could confirm was that power was getting past the power switch and fuse and onto the main chassis. Apart from that, the set was completely dead and I could find no other easily accessible voltage to check.

The set obviously had to go back to the workshop and I also needed a service manual. Initially, I planned to order a new service manual, until I was advised of the cost for a photocopy - $300! I settled for renting it instead.

When the manual arrived, I could see why it was so expensive - it was huge. The main reason is that it covers all the models employing this chassis and all the possible options. So it covered the basic sets, the projection models, plus the medium and high-end versions. My first problem was identifying this particular set's chassis - and that wasn't very easy.

The serial number started with AGO6, which told me it was a later series after AGO3. It had PIP (Picture-in-Picture), SVHS and hifi sound, Teletext and Multisystem - in fact it seemed to have everything.

I never really did discover what VHP, IPP and SYMMETRIC WITHOUT LOUDSPEAKERS meant. More to the point, there was not one complete section or circuit which covered this model in detail. It was a case of a bit of circuit here and a bit there, making the whole thing a schemozzle. Also, while there were fault finding trees, they were just too difficult to comprehend.

In the end, I settled on using basic principles and muddling through. Removing the entire chassis from the cabinet involved raising two clips (B) at the rear and pulling it out. But this gives only limited access to the components on top and virtually nothing underneath.

I found a plug, M17, with a link between pins 5 and 6. This was the 140V HT feed to the horizontal output stage and there was no voltage on this - but neither were there any short circuits, which indicated that the line output transistor 7618 was probably OK. By removing this plug and connecting a 100W globe to pin 6, I could measure 140V.

By now, I had also inspected some of the components on the top of the main board and quite a lot didn't look healthy. I therefore felt that it was a good time to remove the board and give it a detailed examination.

Using an indelible felt tip marker to identify everything, I unplugged the sockets leading to the front controls, the tube and the peripheral modules, before getting it out. Once out, I had to remove two screws holding the AV connectors plus remove a fixing bracket E which holds the modules in. I then released two clips and six screws on either side before the main (mono-carrier) panel would slide out - grudgingly - to the rear.

Now that I could have a good look at what I had, I could see that several electrolytic capacitors needed replacing. I also noticed that R3698 (470W) was burnt out and there were a several faulty joints. I replaced capacitors C2638, C2555, C2644 and C2701 on spec, along with C2618 (.001μF 2kV) which had cracks in it.

C2609 (.0082μF, 1.6kV) had been replaced at some stage in the past with the wrong value and capacitor type, so this too was changed. The horizontal output transformer also looked a little the worse for wear but, for the time being, I just resoldered the pins and repaired a few cracks in the PC board near pin 14.

My main worry at this point was to determine what had destroyed R3698, which is from the gate of SCR6698 to chassis - it would need a fair bit of grunt to burn a 470Ω resistor, yet the SCR measured OK. I replaced the resistor and C2698 for good measure before putting it all back together and trying it.

Early on in this story, when I first checked the set in the customer's home, I had neglected to label one of the single plugs coming out of IK13 on the PIP interface panel 1710. Now I couldn't work out where it went and the service manual was completely useless. Eventually, I found it in the back of the 100Hz section where it is shown as going to plug IM27 on the main board (which is conveniently hidden under a plastic bracket).

On some circuits, this is also shown as 1M17 and is the 5V standby but quite where it goes to isn't shown on the circuit diagrams. I suspect it goes to the front panel but time prevented me from tracing this further. And I still had no power at all, even with M17 connected.

Next, I shorted the base and the heatsink of the horizontal output transistor, effectively cutting it off. The lamp still glowed when I did this, so it looked as though I was looking for some protection circuit fault.

I removed the chassis again and started an exhaustive DC resistance check. This revealed all sorts of bad news. First, D6702, a 47V zener, was short circuit and F1601, a 315mA button fuse, was open.

It was about now that I was beginning to realise that there was a significant similarity between the 3A chassis and the more popular 2B chassis. With F1601 blown, it was fairly obvious that the horizontal output transformer (5620) was shorted. And guess what - it is identical to the one used in 2BS chassis (EHT6140 substitute). A new one was fitted and the set reassembled but it still refused to start.

This time I followed the circuit from the gate of the SCR via plug M7 pin 2 and onto the Power Amplifier Sound (PAS) module 1003, where I found resistor R3229 burnt out on the collector of transistor 7226 (BC558), which was shorted. The difficulty here was that R3229 is not shown in any circuit at all and in the end I just fitted a link instead.

Now when I switched on, I had sound and picture but there were still problems. The purity was all over the place and I made a mental note to replace PTC 3653, the dual posistor to the degaussing circuit. The was a lot of intermittent chirping noises from the horizontal output stage and when I tapped the horizontal output transformer, the sound would change pitch.

I tried this several times, then suddenly I had no vertical deflection and the set died completely. The chassis came out again and this time the horizontal output transistor (7618) was shorted and fuse F1601 had blown. I replaced the 2SD1710 with a 2SD1577 as shown on the circuit and I found two long, barely visible cracks between adjacent pins of the horizontal output transformer which were breaking some of the copper tracks.

I repaired all these and replaced D6646 which supplies the 26V to the vertical output stage. I also found cracks in the degaussing sockets (M18 and M19), which I also replaced when replacing the PTC.

Getting There

That finally fixed the more urgent problems but unfortunately there were still some minor faults. I still hadn't fixed the intermittent chirping noise and traced that to the deflection yoke before deciding that the customer must have always had this and could live with it. In any case, it only happens during the first few minutes from switching on. This she subsequently admitted and she also told me that the set had a new tube fitted while it was still under warranty.

The next problem was that it took ages for the picture to come on and sometimes it didn't come on at all. This can be indicative of a "flat" (low emission) tube. However, in this case, the collectors of some of the colour output power amplifier transistors had broken their copper PC tracks.

I repaired these and made two modifications to improve the turn-on time, to reduce it to about 10 seconds. One modification involved changing R3492 to 3.3kΩ (was 1.8kΩ) for peak white and the other was to add 150kΩ from IP12 (12V) to 2P12 (Beam Current Information Limiter). This desensitises the auto cutoff control to IC7355 TDA4580.

And that was it. The repair wasn't cheap and even then didn't charge for all the time taken. However, Mrs Adams accepted the cost and, after a long soak test, I organised for it to be returned. She was extremely suspicious about every move I made while setting it up but t all worked properly and she paid promptly.

I do hope I don't get too many more of these.