Items Covered This Month Philips KS786 TV set. Grundig ST 95-775/9 TV set. Panasonic NN-C855B microwave oven. Akai CT2869AUS TV set.

My first story was a typical example of the last point, although I didn’t realise this immediately. Instead, I took the customer’s observations at face value.

Mrs Edwards needed a house call for her TV set, since it was a 95cm (37-inch) model that was too big to bring in. And to help make the house call worthwhile, she also had another TV set which needed fixing. Both sets had sound faults.

The larger set turned out to be a 1992 Grundig ST 95-775/9 PIP/TOP with a CUC7890 chassis. The other was a 1987 Philips KS786 with a KL9A-3 chassis. But how hard can a sound fault be? If nothing else, I could fix the Philips and bail out of the ageing Grundig, so I agreed to go.

The lady had switched both sets on before I arrived and neither showed any problem at all. I was somewhat miffed at this but she insisted I stay until the faults occurred. Fortunately, the Philips didn’t take long – there was a crackling coming from the loudspeaker and when I removed the back, I could see the EHT had been sparking from a very dirty ultor cap.

I removed and cleaned the EHT lead and made sure it was making good contact before sealing it with silicone. A piece of cake, I thought – now for the Grundig which was running in another room. This, too, was beginning to play up with slightly distorted crackling but only in the left channel.

Using the remote control and selecting the audio menu, I moved the stereo balance to the right and the problem stopped. I told Mrs Edwards that if she wanted this fixed, she would have to wait while I ordered the circuit manual. I explained about the balance control and she seemed quite happy with this.

The service manual duly arrived a fortnight later and I phoned to make an appointment. However, Mrs Edwards said that the set was now working OK and that my services were no longer needed. Such is life – I filed the manual and moved on to other jobs.

It didn’t last, of course. Mrs Edwards was back on the phone three months later, saying the Grundig was "still playing up". Patiently I asked her to be more specific. Well, she said, the sound was distorted in both speakers, as before!

Because I now had the manual and the audio output ICs in stock, I decided to give it another go. But I was more than a little concerned about how vague Mrs Edwards was with some of her comments.

When I arrived and checked the set out, the distortion was still only in the left channel – it was just that the balance control had been returned to the centre. Initially, I decided to replace the left channel audio output IC (IC4200, TDA2502H) and solder any suspect joints I came across to see if that fixed the problem.

The first obstacle I had to overcome was the need to swing the TV set away from the wall. Unfortunately, it was sitting on a stand that didn’t have castors, weighed about 100kg and had to be slid across thick shag-pile carpet. However, I eventually managed to achieve this and remove the back.

The next problem was to figure out how to remove the sound output module, which was held in with a series of ingenious plastic clips (the service manual doesn’t mentioned these). After a struggle, I finally managed to remove the module and replace the left-channel IC. I then put it all back together again and switched on but it was still distorted in the left channel.

I then did what I should have done when I first arrived and that was to play around with the accessories on this rather complicated TV set. The problem was trying to familiarise myself with such a complex set without altering any settings that I might not be able to restore.

I checked whether the TV system had been correctly set as B/G for Australia, which it was, but when I selected the video via the AV (Audio/Video output), I found that the sound was fine in both channels. This proved the problem wasn’t in the audio amplifier stages at all.

Next, I tried to determine whether the fault was the same in mono transmission as well as stereo but I couldn’t get any stereo reception! The stereo indicators on the display and in the menus wouldn’t alter at all. The problem seemed to be in the stereo decoder which was, as far as I could see, part of the multi IF module.

By now, I had spent a lot of time on this set without achieving anything, so I decided I would go back to the workshop and have a good think. If only Mrs Edwards had been a little more observant and more detailed in her description of the fault – after all they watch the set all the time. And if only the set wasn’t so big and heavy, so that I could get it into the workshop.

Finally I contacted the agents to find out the price and availability of the ICs and modules. The parts were all available but expensive and I didn’t really have a clue as to what I might need. I asked if I could speak to technical support and was delightfully surprised to find out that not only could I speak to someone but that he actually knew – or was pretty sure he knew – what the fault was.

In this set, there are two surface mounted BC858 PNP transistors (CT2524 and CT2527) on the copper side of the sound output module PC board – one for the left channel and one for the right – and these were the nominated culprits. Armed with this sort of 24-carat knowledge, I decided to give it one more try in the house and made an appointment after the ordered transistors had arrived.

Back at the Edwards’ residence, I removed the sound output module and sat down with it at a table. The first problem was finding the two transistors, as they aren’t marked and there is no component layout. Eventually, I found them under a metal screen soldered on the top lefthand side of the PC board.

The next drama involved removing these surface-mounted devices (which are soldered and glued) without damaging the tracks on the board. After that, I had to hold each of the tiny new transistors in position with a pair of tweezers while I soldered their legs into place without applying too much heat.

Finally, I reassembled everything and was ever so relieved to hear clear stereo sound from both channels.

Because the circuits don’t show the interior block diagrams of the associated ICs, I can only speculate on their roles. IC2530 appears to be part of the digital surround sound preamplifier processor, while IC2350, IC2520 and IC2510 make up the stereo decoder (both Nicam and Zweiton) – this being a multi-system TV set. Why the two transistors failed I don’t know – I was just happy to get out of there with the job done.

A spooky microwave

Many electronic items are now so cheap that they are no longer economical to repair at all. Unfortunately, I am old enough to remember fixing 6-transistor AM radios, whereas a modern "boom box" – usually consisting of a CD player, dual cassette and stereo AM/FM tuner – is no longer worth fixing.

Video cassette players are borderline too, with new VCRs retailing at just over $200 and the hifi models starting at about $300. It is hard advising someone who paid in excess of $1000 many years ago that a better performing machine is now available at a third of the cost.

The same can also apply to computers and to microwave ovens. I recently saw a microwave oven advertised for just $120! One wonders how they can purchase the raw materials for that! At the same time, the power output has increased in these ovens to almost double what it once was, although many are now so small that I wonder whether my coffee cup would fit inside.

So what is all this leading up to? Betty Carver brought in her rather up-market Panasonic microwave oven, somewhat grandly called "The Genius Dimension 4". This particular model (NN-C855B) also includes a convection oven, is about five years old and would cost in excess of $600 to replace.

The problem, according Betty, was that it intermittently "cooks by itself" (ie, turns itself on) and sometimes she couldn’t set the clock.

With all the safety circuits built into ovens, I found it hard to believe that it could turn itself on unaided and suspected there were some two-legged gremlins in her house. But she swore that she had been woken in the middle of the night by the audible alarm noises that the oven makes when it has finished cooking. Apparently, she even went into the kitchen once to find it on and cooking! More specifically, the display read "7/8 Beef/Pork".

Intrigued by this somewhat spooky story, I agreed to have a look at it. This oven has a "Word Prompting Display" with an operation guide built into the microprocessor. When I switched it on it said, "Welcome to Word Prompting", and I found that everything from setting the time to convection cooking all worked perfectly for me.

I left it on the soakbench and days later, when I was about to phone Betty and tell her that I could find nothing wrong, it finally started its tricks. I was in another part of the shop and suddenly I could hear the oven beeping, followed by the noise of the relays cutting in, the magnetron powering up and the tray rotating. It was all very X-files stuff, as no one else was around. And why had it selected the food category "8 Pork".

I tried to stop it but it wouldn’t immediately. After four tries it did stop and then the oven would work normally. It didn’t play up for the rest of the day, and I was puzzled as to how this had happened. Normally, selecting "Pork" would require, as a minimum, someone to press the 7/8 button twice and then press Start.

The next day, when powered up, it wouldn’t let me change the clock or do anything much. In the days that followed, I found that this apparition was also likely to occur when the oven was switched on in the mornings, intermittently giving exactly the symptoms Betty had described.

I opened the oven and found it to be exceptionally clean. I then decided to stop our "ghost" in his tracks by disconnecting the membrane switch panel from CN4 on the Digital Programmer circuit board.

And this, at least, provided a clue. The oven no longer operated by itself but neither could I make it operate. I then left it connected and the next time it tried to cook by itself, I unplugged it before it could complete its sequence. Gradually, by plugging and unplugging it, I convinced myself that the source of the phantom cooker was indeed the switch panel itself.

I ordered a new one and when it arrived, I dismantled the front panel and carefully pulled off the old switch panel. I tried to examine the internal parts of the switch membrane assembly (visible through the transparent underside) but everything looked OK.

Finally, I pulled the paper backing off from the new panel and glued it to the front escutcheon before reassembling the unit. The unit was then soak tested for three weeks and it now worked perfectly.

Intrigued as to what caused the problem, I checked for high resistance between the contacts on the flex cable but couldn’t find any problems. So what happened? I can only speculate that Betty used detergent and water to wipe the front panel and some had penetrated around the edge and crept inside by capillary action. After that, depending on the temperature and humidity, this would intermittently activate some of the switches, causing the oven to start cooking.

A crook Akai

Long, long ago (even before Star Wars), Akai Japan used to make up-market reel-to-reel tape recorders, then moved into VCRs and TV receivers. They became very big and other factories in Korea (Samsung) started making TV sets for them, after which they bought sets made in China. Most of these were for the small-screen budget market and Akai felt that there was a gap they could fill by also supplying upmarket large-screen TV receivers.

Akai turned to Europe to produce these and they were made by ITT (also Nokia/Salora/Schaub Lorens/NAD/Dual/Luxor, etc). Only a dozen of these sets were sold in Australia and these were early digital types (the term "digital" is used very loosely in TV technology and is mostly confined to the tuning and control end of the set – the scanning and audio is essentially analog. Most use the Philips I2C bus system).

The problem is that when I encounter these sets, I am intimidated by the often unrecognisable – so called – international symbols, the sheer number of features and the need to figure out how to use them before the function times out. Often, I wish that they would print an instruction booklet for the instruction book – in English!

Anyway, there I was behind Mr Smithfield’s Akai CT2869AUS TV set, grateful that this was only a basic "poverty-pack" model. The fault ticket said succinctly that the set was dead which was partially true – there was no sound or picture, just a ghastly expectorant sort of noise from the rear, as though the horizontal output transformer was dead or in the process of dying. Connecting the CRO to the collector of the horizontal output transistor (T501) revealed significant ringing in the horizontal output transformer – that is before T501 decided to cark it, right in front of me.

This was a not a good start. Convinced that the horizontal output transformer was the main fault I ordered a new one, only to find it was no longer available. After a lot of searching, I found that an HR6160 or EHT3790 was probably an equivalent and ordered one. When it arrived, I duly slipped it in with supreme confidence, along with a new horizontal output transistor, fully expecting that that would be the end of the matter.

Well, I was wrong! – it made no difference at all and the set was still dead and buzzing.

I subsequently spent an lot of time checking the tuning capacitors around the collector of T501 and the deflection yoke, to see if any had changed in value. After all, it takes a lot to destroy a horizontal output transistor which is rated at 1500V and 5A. The horizontal pulses are normally 1200V but if a capacitor changes in value (either going lower or open circuit), these pulses can go a lot higher and exceed the collector-emitter rating of the transistor, destroying it instantly.

Alternatively, if there is a short circuit and the current exceeds 5A, the transistor gets very hot and fails. However, the switchmode power supply normally senses the excess current drain and will reduce its output – which was what was happening here. In fact, the 155V rail was down by about one third.

After some hunting about, I found that D573, a 33V zener diode, was short circuit. This zener diode is in the collector circuit of the east/west output transistor (T573) but not shown on the diagram.

Another component not shown on the circuit diagram was C516, a 470pF 1.6kV blue capacitor, which appeared to be somewhat worse for wear. Unfortunately, replacing these two components had no significant effect on the symptoms. However, the situation improved when I removed the deflection yoke and I subsequently wasted a lot of time substituting the yoke from another set only to find it made no difference.

In the end, I turned my attention to the power supply. This really hasn’t changed all that much in 25 years and is not all that dissimilar to that used by Sunbeam Graetz TV sets in 1975!

The switchmode power supply is based on a TEA2164/G control IC (IC701) and chopper transistor T701. Among other things, the output from the transformer supplies the horizontal output transistor (T501). This means that if anything happens to the power supply, it is likely to have some affect on the horizontal output stage.

With this in mind, I decided to replace all the electrolytics in the switchmode supply, from C700 to C734 (about 10 capacitors in all; the others in the sequence are elsewhere in the circuit). None of the electrolytics on the primary side made any difference but when I replaced C732 (1000μF) on the 31V rail, the set was fixed – well nearly!

Now you could have knocked me down with a feather at this point! Why should ripple on the 31V rail be so significant, especially as all this rail seems to feed are the audio output ICs?

I wasn’t quite out of the woods yet, as the set still showed a picture that was distorted with what looked like hum. Hot on the trail now, I continued replacing the electros and eventually found that C722 (470μF) on the input to the 5V regulator (IC721) and C522 on the output of the 12V regulator (IC521) were the main culprits.

Finally, I set the 155V rail exactly with R612, making sure there was no picture (ie, minimum contrast and brightness) during this procedure. I also touched up a few potential faulty joints before composing a suitable bill for Mr Smithfield.