I was called to Andrew Twist’s 2002 Teac CTW3250S (using the Turkish Vestel 11AK19PRO chassis), which was dead.

This is a 76cm widescreen TV and I knew that if the fault wasn’t simple, it would have to go to the workshop. And that would be difficult considering the size of the small hatchback I was driving.

Apparently, Andrew had decided to switch the set off and disconnect it from the mains just as a storm was beginning. However, before he could get to the set, a nearby lightning strike did the job for him!

I removed the back and checked the power supply. Ironically (and typically these days), the 3.15A mains fuse (F801) hadn’t failed but R817 (2.2W, 5W) had. The bridge rectifier was OK but FET Q802 was short circuit.

Because this FET is DC-coupled to IC802 (MC44604P), the question was had that also blown, along with anything else?

Anyway, that was as far as I was prepared to go in Andrew’s home. And having informed him of the facts, I suggested that he contact his insurance assessors.

Unfortunately, he had been going through a rough trot and was only just beginning to get back on his feet. He wasn’t insured and really couldn’t afford to spend any money. Instead, he simply paid for the service call and told me he would wait until he could afford to have it fixed.

Three months later, he finally dropped the set around to the workshop. Unfortunately, luck hadn’t been on his side during the storm. Not only had the control IC been destroyed but also D827 (BA159) which connects to the source of the FET.

This was rather puzzling. D827 is in series with D826 and both are identical BA159 diodes but only D827 had short circuit. I cannot really imagine why – I guess it’s just part of the rich tapestry of life and chaos theory.

After replacing D827, the power supply came on but was now pulsating slowly, with a low output. In addition, the red LED wasn’t lighting. I checked for shorts in the six secondaries but could find none.

Next, I turned my attention to the +150V rail to the line output stage, where I checked D816 BYM26D (now BYM36D), the line output transistor Q605 and transformer TR602 (which was unlikely to have failed in this instance). I then shorted Q605’s base and emitter and hung a 60W globe from its collector to ground. However, this globe merely blinked gently when the power was applied.

Further checking showed that there were low pulsating outputs on all the rails, even when I disconnected them one at a time. This meant that it was time to take a close look at the control circuitry in the switchmode power supply and at microprocessor IC501 – in particular, the circuitry to pin 22 (ST-BY) and pin 33 (RESET).

The supplies to the microprocessor are 3.3V and 2.5V and both these were extremely low. Disconnecting the links to the microprocessor brought these voltages up but still there was no life. Furthermore, shorting one side of R831 to ground should turn the power supply on from standby but it made no difference.

Next, I checked the optocoupler and its control IC zener (TL431), as well as IC804, C818, C704 and C829. However, I wasn’t really getting anywhere.

Biting the bullet

It was time to bite the bullet and replace microprocessor IC501 and the EEPROM (IC502). This was all rather confusing, as the service manual lists not one but two different part numbers – neither of which bore much resemblance to the one fitted in the set, the one on the circuit diagram or the ones supplied. I tried them one at a time until I found, naturally, that the most expensive part worked!

I now had some control of the set. The red LED would come on in Standby and flash four times when the set was on and I could do this with the remote. The voltage rails were up a little as well, especially the low 2.5V, 3.3V and 5V rails. However, the 150V rail was still pulsating around the 80V level.

Well, where to next? There was only the line driver and the jungle IC (IC401) left. The latter device is a TDA88441NZ, a big 56-pin device with high-density pins. I had already checked the voltages into this IC and the clock crystals, so it now had to be changed.

Replacing this IC finally brought the set to life, with picture and sound, but the geometry and picture were all over the place. Just in case I had misdiagnosed, I refitted the old micro and EEPROM which were socketed but it was obvious that these had been destroyed.

Now all I had to do was reprogram the EEPROM. First, you have to get into the Service Mode by pressing the blue INSTALL button and then 4, 7, 2, 5 and go to the Options menu. Once again, the inadequate service manual (which, I might add, lists the CTW3250S as an 11AK19P3 – not an 11AK19PRO as written on the back of the set – and the CTW32505-1 as an 11AK33) does not have a list of option codes for each model. Instead, it has a list of 60 options with eight bits each and you have to work out what is appropriate for your TV.

Fortunately, or unfortunately, options 30-37 and 39-49 are missing, so what do you do about these? And the manual refers to each bit as having a value of 1, 0 or X – X presumably being "I’m not sure?" (many of the choices you have to make from this total lack of information would surely come under this category). The only thing is, the TV won’t let you put in X as an option.

For example, Option Ø, Bit B4 Akb = 0 ) )0x02) Hue B6 (Black Current Stabilisation)? and Option 01 Bits B7 & B6 Forf = Fors = 1 (Forced Field frequency auto) (50Hz when line not synchronised)?

As you can see, it’s not easy to make decisions based on these criteria! However, I did the best I could, making intelligent guesses, before moving onto the adjustment menus. Here we now have 65 controls to align for 4:3, 16:9, Cinema, Sub-title and Super Zoom settings, again with no real specification to compare with.

Ironically, you cannot use a colour bar generator connection to an AV input as AV is not selectable in the Service Mode. Instead, you have to tune it in first on an RF channel and store it as a program number. For each of the aspect ratios, it doesn’t make it clear whether you need a generator that can give signals other than 4:3, nor does it make it clear whether and by how much you must overscan the picture.

By trial and error, I managed to get what I thought was the correct geometry for each size but found the S-correction (adjust 14, 26, 37, 48 and 59) made no difference and I assume that Vertical Slope (adjust 14, 25, 36, 47 and 58) means Vertical Linearity and that EW Parabola Width (adjust 18, 29, 40, 51 and 62) means pincushion.

When it came to setting up "Y" delay for SECAM, NTSC and others, as well as IF-PLL negative adjustment with PLL tuner, I chickened out and left the default values that were already there. Adjust 03 is AGC and was set to 015. After that, I tuned in all the local stations and was quite happy with the results while it was on test.

Andrew managed to con me into redelivering the set and reinstalling it. No sweat – I felt that this shouldn’t be too difficult. All I had to do was retune the NEC VCR and connect his Foxtel box.

It all initially went well, despite the millions of stairs and the hassle of moving his other 68cm set out of the way. However, when it came to tuning the set, I ran into all sorts of grief. I selected the Foxtel MENU setting so that I could easily identify it and that went into the pre-tuned VCR (on which I also selected its own menu) and into the Teac.

I was expecting it to come up in two places on the TV and sure enough, it came up fine on Ch65 for the VCR (too close for comfort for Channel 10 on UHF Ch66) and Ch55 for the Foxtel. However, when I turned off the menu, the picture was shocking, with severe herringbone patterning on both locations.

Andrew kept assuring me that all had been perfect before and that the reception was also perfect on the temporary set he had been using. The only thing I could think of was that perhaps the AGC hadn’t been set correctly and that it might be critical in this location.

I went back into the Service Menu and spent what seemed like hours messing with the tuner options and adjustments – all to no avail. In the end I did the "De Bono thing" and thought laterally – if the tuner/tuning is crook, connect everything via the AV sockets. This meant supplying extra AV leads and a SCART adaptor – but it was worth it if would save time to fix the fault.

Well, I did that but Foxtel was still showing a terrible picture. But why? Well, the first thing I noticed was that the simple blue and white menu was fine. It was only while actually watching Foxtel Ch13 that the problem occurred, so I changed the channel and found that almost 90% the other channels were perfect. It had to be a cable fault!

Luckily, Andrew rarely watched Ch13 and someone must have selected it by chance. Anyway, he was delirious and so was I that at last I was out of this particular nightmare.

Another Teac

Later that week, I was forced into attending (for political reasons, to do with Mrs Serviceman) a Teac EU-80ST (Vestel 11AK19) for another client. This was dead and stuck in standby, with the red standby LED on and no other functions working.

As luck would have it, the 80cm set was wedged into an entertainment centre that had to be unscrewed before it could be moved! I can’t tell you how much I didn’t want to be there but there are times when you just have to do jobs you don’t want to do, whether you like it or not (it’s called appeasement)!

Anyway, after a struggle, the client and I eventually managed to dismantle his elaborate entertainment centre console and place the TV on a coffee table. I then removed the back and took a "peek-a-boo" inside. The first thing I noticed was that, despite the furniture looking pristine, the cabinet of this 4-year old TV had been badly scratched.

When I looked at the chassis, I also noticed a large number of cracks in the PC board that someone had bridged with jumpers. Faced with this evidence, the client finally cracked (excuse the pun!) and admitted that the set had been dropped some time ago – but it had been working perfectly for a very long time since (where does Mrs Serviceman pick up these people?).

He also admitted that the front controls had not been working for a very long time. However, the remote was fine so he hadn’t worried about this.

Well, careful examination of the main chassis and the front control board, which also carried the infrared remote control receiver, showed that plug PL502 had a series of very bad dry joints. I resoldered these and – hey presto – the remote control was back in business and the set could be switched on again. The front control buttons still couldn’t be used but I didn’t care (it was probably due to a hairline crack). After all, the client was happy and if he was happy, Mrs Serviceman would be happy and perhaps I would get something special for dinner that night!

Microwave ovens too!

Our amazing technician Michael was telling me about an interesting problem he had with a Panasonic Series II Inverter Microwave Oven NN-T790SF.

This problem involved the display. When the START button was pressed, the word prompting display would sometimes show "Welcome to Panasonic", as though the power had been switched off and the microprocessor reset.

He traced this to the mounting screw for PC board 4W7QPT, which had failed to make a good ground connection. The circuit diagram shows pin 2 (centre) of plug CN4 as an alternate ground lead (yellow) but, in fact, the matching pin hadn’t been fitted to the PC board in the factory. The fix was to cut and solder the yellow lead directly to D27 instead.

Stubborn Sanyo

Sometimes, the problems I encounter are straightforward and sometimes they are extremely complex. Unfortunately, when a set first comes in, you can never tell which category the problems are going to be in.

Recently, I had a "newish" Sanyo CP14SR1(G)-50 (AC5-G14 chassis) come in with the complaint that there was no colour when a DVD or VCR was plugged into it. Well, straight away, I thought that this was going to be easy – the wrong colour format must have been selected for the AV input.

I pressed the Menu button, picture and tuning menus but nowhere could I find a system menu. OK, this isn’t unreasonable – after all, this is only a cheap 34cm portable so perhaps it doesn’t have this feature.

Next, I had a good look for dry joints, especially around the crystal, and looked for a reference oscillator control that might be slightly out. I couldn’t find anything.

Eventually, I noticed that if I ran my fingers down the pins of the jungle IC (IC201, LA76818A), the colour appeared but I couldn’t pinpoint the problem. In the end, I assumed that it was probably the IC, so I ordered a new one. I fitted it as soon as it arrived but "quelle horreur"! – there was no change!

A service manual for this chassis had been ordered at the same time but I could find nothing in it that gave me a clue. The oscilloscope showed the full colour waveform arriving on pin 42 of the chip and the reference oscillator was spot on.

Fortunately, by mistake, an instruction booklet had also been supplied with the Service Manual. And page 17 (near the very back) had the solution!

What you have to do is select the AV input and then the RECALL button on the remote control. This button is marked with a symbol of a TV screen and a bar in the bottom righthand corner and it reveals an AV system menu offering AUTO, PAL and NTSC. Of course, NTSC had been selected and changing this to AUTO fixed the colour problem.

Husband bashing

A husband and wife were having a domestic dispute that began to get a bit out of hand. In frustration, she picked up a glass and threw it at her husband (as you would expect) and naturally he ducked. The glass continued its trajectory unhindered until it came into contact with their brand new Panasonic TX51P800H Rear Projection Television (GP1VP chassis). It hit the middle of the set’s 130cm diffuser, severely cracking it.

The diffuser is made of two layers: a front lenticular lens that’s worth $300 and a rear Fresnel lens that’s worth $400. Add in an extra $220 for the labour charge and . . .
OUCH!

Anyway, they lived with this for a while until the set blacked out into Service Mode, with the red LED flashing. This problem was even more severe. The red picture tube had been broken in two when the EHT arced through the neck and onto the deflection yoke winding. Fortunately, these latter items could be fixed under warranty but they had to cough up just under a grand for the diffuser lenses!

Dead Orion

A dead Orion 32 DUA TV set (T7800 PRO series chassis) was delivered to my workshop without its instruction booklet or remote control, which I always request for obvious reasons. The fault was fairly easy to diagnose – the line output transistor had gone short circuit due to an arcing flyback transformer. Both were replaced but on switch-on, there was just a black screen with the program number "PO2" displayed and "Child Security Active".

I couldn’t get into the menu to unlock this as only the remote control works in this mode. That meant a delay of a couple of days while I contacted the client, to ask him to drop the remote in.

When I finally got the remote, I found I could activate the menu and go to child safety but I still couldn’t remove the message. It was only when I went I went into the service mode (press menu and –vol and switch on with the remote control) and entered the VG2 set-up that I saw "error" next to it. Adjusting the screen control of the flyback transformer changed this message from "increase" to "OK" to "decrease". Leaving it on "OK" finally removed the "Child Security Active" message.

It’s easy once you know why and how!

An expensive repair

A dead Sony KV-PF21L70 (BG-3S chassis) was brought in and I immediately thought "piece of cake".

Its line output transistor (Q511, 2SD2578-CA) was short circuit. I replaced it, resoldered the line drive transformer (T501) to eliminate any potential dry joints and checked C528 which is across transistor Q511. I then switched the set on, thinking that that would be the end of the matter.

So much for my confidence – the new (and rather expensive) transistor failed again and I was back to square one!

This time, I removed the flyback transformer (T503) and checked it with a shorted turns tester. It checked OK but just to be sure, I substituted a flyback transformer from a similar set with a broken tube. I then shorted out the base and emitter terminals of the line output transistor and hung a 60W globe on its collector, along with a DVM.

When I applied power, the globe lit and the DVM measured +135V. I then touched the CRO probe on the collector of line driver transistor Q506 and momentarily saw a good waveform appear.

Suitably encouraged, I removed the globe and the base-emitter short from the line output transistor, attached the CRO probe to its collector and set the DVM for maximum hold on the 135V rail. Once again, the set tried to fire up but after an initial rush of static, it immediately died. The second replacement transistor had also failed.

Both the waveform and voltage were correct, so I checked for shorts on the secondary of the flyback transformer, replaced the transistor again and tried another deflection yoke. This third transistor also failed when power was applied.

By now I was beginning to feel that excessive EHT was killing these expensive transistors. I had already used up my entire stock, so I switched temporarily to cheaper BU2508DF transistor equivalents.

Unfortunately, the failed start-ups were also causing further damage. D504 and Q505 (an IRF614 FET in the east-west circuit) had gone short circuit and IC301 CXA21395 jungle had also been damaged. This latter IC is extremely expensive, costing just over $110.

Even worse, I still I hadn’t got within cooee of solving this but my pride wouldn’t let me stop. As a last-ditch effort, I started removing and checking all the capacitors in Q511’s collector circuit. Fortunately, the third one I checked, C538 (8.5nF), was faulty, measuring just 1nF. This is a critical capacitor and as the EHT is inversely proportional to its value, it’s a wonder the tube hadn’t been damaged by flashover.

I fitted a new capacitor and switched on with some trepidation. This time the picture and sound came on perfectly. However, I wasn’t completely out of the woods because there was still one more fault.

On high-contrast pictures (ie, high beam current), the picture would pull to the left and the set would switch off with the red LED flashing twice, denoting overcurrent protection. The self-diagnosis showed errors 002, 003 and 004 as all being 001.

I spent a lot of time analysing and measuring the ABL (Automatic Brightness Limiter) circuit, especially D315, D316 and Q312, before finally concluding there was nothing wrong with it. Eventually, I realised that it was a protection fault and that Q604 and Q503 were being switched on.

It was then that the penny dropped – R615 (1.2W) had gone high due to the repeated excessive current being drawn by the short-circuit horizontal output transistors. I replaced it and this time the set performed correctly – but was this all really worth it?

Oh, well – you win some and you lose some.

Items Covered This Month Teac CTW3250S TV set (Turkish Vestel 11AK19PRO chassis) Teac EU-80ST TV set (Vestel 11AK19) Sanyo CP14SR1-50 TV set (AC5-G14 chassis Panasonic Series II Inverter Microwave Oven NN-T790SF Panasonic TX51P800H Rear Projection Television (chassis GP1VP) Orion 32 DVA TV set (T7800 PRO Series chassis)