Mr Cleary was having his wooden floor-boards sanded, sealed and polished in his small but expensive ultra-modern kitchen and lounge room. Naturally, everything had to be lifted off the ground in order for the sander/polisher to do its work.

Being such a small apartment, space was at a premium. Most things were put in spare rooms or on shelves, but also on the kitchen benchtops and sink.

He left his Panasonic TX68PS72A (MX12) flat-screen TV until last but by now, there was no room left to stow it. Ever resourceful, he eventually decided to put it face down on a soft thick blanket on top of his high-tech induction cooktop – but only after making sure the cooktop was turned off. Not wanting to take any chances, he also turned off the circuit breaker to the cooktop – just in case.

Now, this TV is pretty heavy and requires a lot of effort to lift, manoeuvre and roll over onto its front. This meant that is was impossible determine whether any of the cooktop controls had moved under the blanket but if they had, the circuit breaker was still off.

When he had finished, he checked and covered everything up before leaving just as the floor polishers arrived. The men wasted no time and after a couple of hours, the job was nearly done when the old sanding machine decided to stop – it was pretty hot after a good workout.

The operator cleaned out the dust that was clogging the machine but could find nothing otherwise wrong with it. However, it still wasn’t working so he went over to the fusebox to check things out. Here he found that not only one circuit breaker had tripped but two. Unaware of Mr Cleary’s failsafe approach to safety and not knowing which circuit breaker did what, he reset both.

Well, his sander was now back in business so he completed his task unaware of any other problem.

Predictably, Murphy’s Law kicked in – the cooktop controls had moved when the TV was placed on top of it and unbeknownst to anyone, was now powered on and cooking.

In operation, an induction cook plate runs cool and only warms up metallic objects placed on top of it to cook. The problem was, the metal shadowmask of the M68LVQ087X real-flat picture tube was only centimetres away and it got hot.

After a while, there was suddenly a very large bang, followed by a loud hissing noise. When the TV was lifted up and the screen was visible, there was a large crack roughly in the shape of the cook top. The plastic cabinet was in perfect condition but the picture tube had been completely destroyed!

I’m not sure of the insurance implications but it does look as though the high cost of installing a new picture tube will render this set a complete write-off. So despite Mr Cleary’s precautions, this has been what you might call a first-rate high-tech accident!

Items Covered This Month  Panasonic TX68PS72A TV set (MX12 chassis)  Technics SE-A5K-XL amplifier and S4-A6K-XL preamplifier  Boston Acoustic P0054 SA1 VR2000 Subwoofer  Sanyo CP28WF2 widescreen TV set (FAW-128 chassis)  Loewe Concept 770 TV set (Art. No. 53470 Chassis C9003)  Toshiba 50D9UXE rear projection TV set (RB9731 chassis)  Panasonic TC48M90A TV set (MX5A chassis) Philips VR6740/75 VCR

Fussy audiophile

David, our main man when it comes to audio, had a very fussy audiophile who brought in his Technics SE-A5K-XL amplifier, complaining that the righthand channel was distorting and intermittently cutting out.

David insisted that he also bring in his S4-A6K-XL preamplifier as well, despite Tony’s assurance that it was working perfectly and couldn’t possibly be the problem. Eventually, David prevailed and the preamplifier was brought in.

Despite its 23 years, the SE-A5K-XL amplifier is still an impressive piece of equipment to look at. It also performs pretty well, with 100 watts true RMS per channel into 8Ω at just 0.001% THD.

It didn’t take long for “super-tech” to trace the reason why the amplifier was intermittently cutting out – the tone defeat switches were dirty and creating noise, causing the main protection circuit in the amplifier to activate. David 1, Tony 0.

The distorted amplifier was a bit more tricky and very intermittent after 5-10 minutes operation. Dave went for the heat/freeze approach but you have to be very careful with this because sometimes the ensuing condensation can switch the amplifier into an avalanche that the protection circuits cannot save.

What he did was squirt freezer onto a cotton bud and then touch individual active components until he found the most sensitive one. That done, he then reheated the device with the hot air from our Hakko 852 desoldering station.

This approach soon narrowed the problem down to driver transistor Q316, a 2SC2632. Replacing it completely fixed the problem.

After soak testing both units with various sources, it was ready to be collected. Tony was still arguing the toss as he departed with his precious gear but at least he was happier now.

Woof, woof

I had an expensive 1996 Boston Acoustic P0054 SA1 VR2000 Subwoofer come in, where a transistor insulating pad had punctured under one of the transistors (Q12, 2SD1138) and had in turn blown all nine output transistors, along with nine resistors. The repair work was really straightforward and I made extensive DC checks of all the components in the circuit.

I fitted the output transistors with large mica wafers (cut to size) and smeared all mating surfaces with silicone heat transfer compound for good measure. The only difficult part was soldering the output transistors onto the copper side of the PC board, because the copper tracks were lifting. Eventually though, I was pretty confident that all was OK and switched it on.

The subsequent explosion was probably heard in Perth – it was about 10 on the Richter scale!

Back at square one, I measured the hfe of all the new transistors and checked the resistors. Exactly the same components had failed as before. So what was going on?

This time I wasn’t going to pussy-foot about. Before fitting the output transistors, I checked the ±80V, ±12V and ±23V rails but these seemed to be OK. I also fitted additional fuseholders and 3A fuses and when it was all reassembled, I connected a 200W globe in series with the 240V into the power transformer. Finally, with no preamp or speaker connected, and with meters monitoring the offset voltage and the power rails, I gingerly switched it on.

What an anti-climax! The globe didn’t light – instead it powered up perfectly and all the voltages were correct. I adjusted R29 for 7mV across the emitter resistors but anything from 8-13mV gives a quiescent current of 35-49mA which is acceptable.

I then checked the subwoofer loudspeaker with another amplifier and it worked perfectly. That done, I connected it to the subwoofer amplifier and again switched it on. Once again it fired up flawlessly. The power rails measured ±74V but the quiescent current remained the same.

Finally, I connected the preamp and rechecked everything. The 200W lamp was only just beginning to glow very slightly, all the voltages were correct and there were no bad noises coming from the speaker. The amplifier was warm but not hot after half an hour. And so I decided to turn it off and reassemble it, happy that everything was now OK.

My happiness didn’t last long. As I turned it off, there was an unexpected loud “pop” from the speaker. I didn’t like that – it spelt trouble!

I examined the components for any visible damage and after a few minutes, I decided to try it once again. The globe lit up and smoke poured out of resistors R25, R27 and R36. Aggghhhhh!

Well, that was that. I couldn’t afford any more time and it was obvious that the solder tracks couldn’t handle any more component replacements. It was a write-off.

But what had caused all this? Of all the genuine transistors I fitted, the hfe values for Q11 and Q13 (2SC3281) varied widely and were very low compared to the other transistors. I selected two at hfe = 40 and 55 but some were a lot lower. The rest of the transistors measured over 100.

I suspect that these transistors from my local wholesaler were substandard. In fact, Q8 was hard to get but eventually I managed to get originals and they measured OK. The mica washers were all OK, as were the speakers and preamp.

I guess I should have got Dave to fix it!

Intermittent Sanyo

Mr Vincent’s new Sanyo CP28WF2 (FAW-128 chassis) was brought in suffering from intermittent vertical deflection. The slightest knock to this flat widescreen 66cm TV would cause all sorts of crazy lines as the vertical scan broke up.

This looked like a straightforward case of dry joints or a fracture in the PC board. However, a close inspection on the workbench under a good bright light failed to reveal any problems.

I wasn’t unduly fazed by this – after all, many dry joints and hairline cracks are invisible to the eye. As a result, I started soldering everything in sight around the vertical timebase circuit and then carefully examined both sides of the boards for cracks with a microscope but found absolutely nothing.

By now, I was convinced that the problem was around the vertical output IC (IC501, LA7846N). I replaced it in case of an internal intermittent problem but even this didn’t fix the problem. I then spent an awful lot of time checking, testing and replacing all the components around this IC. Unfortunately, measuring the voltages and waveforms didn’t reveal much because the fault was so intermittent. In addition, with a feedback amplifier, a faulty part will cause a ripple-on effect around the whole circuit, making the culprit difficult to pinpoint.

Finally, I tried the heating and freezing technique. This was done by propping the boards on edge while the set was on and then freezing and heating individual parts while watching the results in a mirror. And that finally pinpointed the trouble.

The problem was C503, a surface-mounted 0.01µF capacitor between pin 6 of IC501 and ground. It had fractured and I replaced it with an ordinary tagged ceramic capacitor.

Crook Loewe

I had a Loewe Concept 770 (Art. No. 53470 Chassis C9003) brought in with a partial vertical collapse.

I began by checking diodes D561, D562 and D563 in the vertical circuit, then replaced all the electrolytic capacitors (C561, C563, C574 and C558). However, the problem persisted so I figured that the vertical output IC (IC561, TDA8175) must be faulty. I ordered a new one from International Dynamics but as the TDA8175 is no longer available, I received a TDA8172 and a modification kit of 15 components plus an instruction sheet for each chassis that took this modification.

If the modification isn’t done, it will blow the power supply up. As a result, I followed the information religiously, which involved removing some components, changing others, and adding R002, R003 and C002 on the copper side of the board. These latter parts bias the reference voltage on pin 7 of the vertical IC to approximately 1.7V from the 5V rail. As part of this job, the track to pin 7 is cut and a resistor R001 added to pin 1 instead.

Afterwards, the vertical geometry needs realigning and storing for both PAL and NTSC. However, the notes say that if the height is still too great, then R001 (22kΩ) should be changed to 39kΩ.

Unfortunately, even after trying all this, I still ended up with a screen that was only three-quarters full, with foldover occupying the bottom quarter of the picture. And even the top had poor vertical linearity.

Clearly, I still had a serious problem but after checking all my work twice over, I couldn’t find any measurable problem. In the end, I felt that there had to be something wrong with the modifications and so I went back to International Dynamics.

They soon got back to me and informed me that the factory had made a few misprints in the instructions! R578, which was originally 68Ω, should not be replaced with a 120kΩ resistor, as noted. Instead, it was R577 that should be changed – from 13kΩ to 120kΩ. Correcting this fixed the problem, much to my relief.

Afterwards, I examined the multilingual 16-page instructions and noticed an additional stamp on the front of the German section, which is also multilingual on alternate lines. In the middle it says “At each actual table the appellation of R578 (120kΩ) must be R577 (120kΩ)”. That’s pretty torturous “German-English” for “I meant R577 not R578”, or “please change R578 in the instructions to R577”, or “error in parts list: R578 should read R577”!

Isn’t English hard?

Toshiba rear-projection

One morning, I found this massive 50-inch Toshiba 50D9UXE rear projection set (RB9731 chassis) plonked right in front of my bench. The note stuck on it said it was dead but ever observant me noticed that you could actually turn it on. However, it would then click off after a few seconds, which indicated one or more protection circuits cutting in like last night’s hangover.

I was told that most faults on these models are on the U909 convergence output board (P139488). This board is located on the lefthand side of the chassis and is hard to remove (the three convergence coil plugs are a very tight fit). Eventually, however, I eventually got it out and checked it for dry joints.

I then connected it back in without the convergence coil plugs (Red P713, Green P714 and Blue P715). This time the set came on and stayed on, giving three separate pictures. It was then a matter of plugging each pair of coils in separately to track the problem down. This indicated that there was a problem only in the green pair of output amplifiers, which are embedded in two ICs (along with the other colours) – Q750 and Q751 (both STK392-110).

This meant that it was necessary to replace them both. Fortunately, that fixed all this baby’s problems – at least, after spending a couple of hours realigning the convergence.

Sick Panasonic

I don’t see many small-size TVs any more because the price of new sets is now so cheap. However, I did get a 1999 48cm Panasonic TC48M90A using an MX5A chassis in for repair.

This set was dead except for the red Standby LED but it was otherwise pristine inside and out. It uses the same type of weird power supply as the MX3, with the main 90V rail being derived and delivered to the line output stage all on the hot side (primary) of the switchmode power supply.

A quick check soon showed that there was no voltage on this line. And on the secondary, the 41V rail was down to 14V and the 24V rail measured just 7V.

There are two optocouplers in the power supply, one for voltage regulation using an SE090 IC and the other controlled by the power on/off function. I briefly dallied with the 90V rail, checking it for shorts and replacing the IC regulator (IC801) before moving on to the power-on circuit. In this circuit, a high from pin 26 of microprocessor IC1101 (MN1871681TFH) drives Q851, Q850 and the optocoupler D840.

I shorted out Q851’s collector and emitter (ground) and the set powered up but no 5V was coming out of the microprocessor no matter what I did. I checked to see that 5V was available on pin 39 but there was no 5V from the reset pin of IC1103.

Next, I tried shorting out pins 2 and 1 of IC1103 which was supplying 4.5V instead of 5V to the microprocessor on reset pin 34. This told me (and the ohmmeter confirmed it) that the microprocessor was faulty and drawing too much current on this pin.

Fitting a replacement fixed the problem.

Spitting chips

This final story is really weird. It’s not from my own workbench though – instead, it comes from J. A. of Canberra. I’ll let him tell it in his own words . . .

I swear that the following story is true but you’re going to find it hard to believe.

I have been in the electronics industry for about 35 years, specifically in computing. And because I “know” about electronic things, it automatically makes me qualified to fix anything that has a mains cord or batteries!

As anyone in this field knows, when a mate says (over a few beers) “you know about computers and electronic stuff, my [insert name of electronic equipment here] has blown up. Can you fix it?”. Being an accommodating kind of chap, and not wishing to let them down, much less degrade my reputation, I usually accept. Even more so if it’s a “damsel” that’s in distress.

As you can imagine, this leads to my fixing some interesting stuff – and some not so interesting stuff..

In this case, it was a Philips VR6740/75 VCR. I have had a look at this machine a few times before. It belonged to a mate and it originally suffered interference from a nearby wireless station, then some time later it developed what looked like worn heads. (That was diagnosed by a repair shop. It turned out to be simply dirty heads).

Anyway, my mate and his wife split up some time ago and she scored the VCR. She dropped it off recently, saying it had the same fault as before – ie, “doesn’t work”.

It stayed in the corner of my workshop for a few weeks, waiting for me to get around to it. When I did have a look at it, the fault was interesting: it would load a tape OK and fast forward and rewind worked, as did picture search in forward and rewind. But if I pressed Play, it would eject the tape!

Removing the top cover confirmed that the capstan wasn’t turning. I removed the bottom cover and checked the belt – it was OK and the motor turned the capstan when I turned its shaft. I then checked for voltage on the capstan motor but it was at 0V at all times.

Not having a circuit diagram, I traced the motor leads back to an interconnecting board, then through a loom to the motor control board. This board is mounted upside down – ie, with its solder side uppermost. I could plainly see the PC tracks from the wiring loom connector to the motor control chip (IC704) solder pad but there was no chip – just a row of holes!

No chip! What the hell was going on? It looked liked someone had had a go at this. I quizzed my wife who is good friends with the VCR’s owner: “Are you sure she hasn’t had this looked at by anyone?” A definite “no” was the answer, “she can’t afford it”.

So where was the chip? Did it fall out?

I reasoned that if it had indeed fallen out, then it should be still in the VCR somewhere. I picked it up and turned it over and around. At first, I could just hear a rattle and then a chip fell out!

A close inspection showed some solder on the pins and some on the PC pads, as you would expect, but there were no signs of it being removed from the board intentionally. Besides, it was still covered in the same layer of dust that was on everything else in the VCR.

Anyway, I cleaned up the solder pads, replaced the chip and the VCR worked perfectly. I was very impressed with the quality of the picture. This must have been an expensive unit in its day, with stereo sound, Dolby noise reduction and so on.

As to how the chip came to fall out, I can only surmise that it’s proximity to the power supply makes it run fairly hot and over the years the solder joints became dry to the point where it simply fell out.