I've just come back from reinstalling a Sony SLV-EZ111AZ/HM video for the Buntings in an old folks' home. The set is barely out of its egg, as it was sold just two months ago, and I originally installed it in their unit which has a UHF community antenna system.

With so many satellite stations transmitting now, it is very difficult to find a slot on the UHF band for a video, and even more so for Foxtel as well. In this instance, the default Ch36 or 37, as well as Ch50, had co-channel interference with another station (the village was positioned on top of a plateau), so eventually I found I had to use Ch69. I then managed to tune in all the stations reasonably, though Ch7 was poor and the Teletext had a lot of errors (being hard of hearing, they needed the subtitle service on page 801).

But this was a piece of cake compared to what was about to happen. Recently, the Buntings were moved into another unit and I had to reinstall it all over again.

Piece of gateau, you may think. Wrong!

Their new place used VHF channels 2, 7, 9 & 10, with only Ch28 on UHF. Everything returned OK but the video was decidedly snowy – even on tape playback – whereas the TV reception was good.

I tried changing the VCR's output channel to various settings from one end of Band IV to the other end of Band V. I even made sure there was no receivable transmission on the channel selected by removing the antenna lead and connecting it to the video lead and checking for pure snow.

I then tried replacing the leads in case they were faulty – though if the reception on the TV off-air was good, I couldn't see how changing the leads could make any difference.

By now I was getting a little desperate and was suspecting a faulty RF modulator in the new video. I tried an AV lead and there was a big improvement all round on both playback and the off-air EE signal. And then, while playing a tape and swapping the RF lead to the video, I noticed that the picture was now excellent, with no snow at all.

Obviously, unplugging the RF lead did the trick. And that meant that there had to be something funny about the RF signal coming in from the community antenna system.

That's when it hit me – we now live in a digital era (something old analog types like me find hard to comprehend) and things are now very different. Somehow, the digital broadband energy of the VHF co-channels was affecting the RF modulator of the video.

Silly me for being so miserly – I had left my $10,000+ spectrum analyser back at the workshop along with my bank overdraft, mostly because I reckoned it would take me a thousand house calls to pay for it.

Anyway, the modern way, it turns out, to fix a very snowy picture is not to boost the signal but to attenuate it!

A 12dB attenuator in series with the video's RF input (antenna) socket improved the picture out of sight. Using the AV leads as well resulted in a picture that was as good as when the TV was connected directly to the antenna.

In case you're wondering, the Bunting's TV was also a Sony – in this case, an older model KV-G21S11 (BG-1S Chassis) – and the problem facing yours truly (and other TV service technicians) is the diverse ways various receivers behave when there are digital broadcasts.

After all, why should this particular Sony set behave differently to other sets, especially as the only other signal on UHF being transmitted on that community aerial system was Ch28 (hardly an adjacent channel). I fear we are in for a lot of this before it all settles down.

Anyway, the Buntings were happy with the reception but couldn't understand why it had taken me so long.

A Teac bounces in

A few months ago, I told you about a Teac Televideo MV1480MkII that I serviced. Well, it came bouncing back with a tag that said "No sound – customer not happy".

Last time, it had an intermittent open circuit loudspeaker, so I figured this had to be an easy repair. The only problem was, when I got the back off and checked the new speaker it was fine. Not only that but when playing a tape, the sound was fine too. It was only in the EE mode (tuner) that I was getting absolutely no sound at all and that included the record mode. Fortunately, I had the service manual and felt that although this was a little more complex, it wasn't insurmountable.

If you can recall the last article, the set turned out to have a faulty surface-mounted IF transistor, so I thought that the IF module would be a good place to start with this repair.

When you look at the back of this set, you could die of despair for lack of access. Luckily, when you get as intimate as I have (unfortunately) been with this one, you will find that it is not that bad – not brilliant but not too bad!

Undoing four screws lets you remove the tuner board (MCA), as well as the colour decoder, and turn it on its side to gain access to the underneath. That done, I used an audio amplifier with a test probe to establish that I did indeed have sound output from pin 7 of the IF module. This went via the AV input (which also didn't give any sound) to plug CL-4 pin 1 on the MCV-C board.

I traced the sound to R227 (27kΩ) and then C227 (0.47μF), before it disappeared altogether on pin 11 of IC201 (BA7767AS). This IC is some sort of custom audio record/playback control chip. I checked and replaced both R227 and C227 and also C217 but drew a blank. I then checked QR201, a digital audio muting transistor, but it too was OK.

Next, I shorted the base and emitter pins of QR201 and suddenly the sound was back. Ah ha! – the audio mute line is ON in the EE mode, my brain jumping quick as a flash to the bleeding obvious. I quickly checked the audio mute line and it measured 2.5V, so I now had to find out what was causing this.

Unfortunately, the next part took a lot of time and effort, especially as access was now appalling, but I had to trace where the audio muting came from. The circuit is also confusing, as it shows two separate paths in parallel from pin 18 of IC201. One is drawn via plug CL-10 pin 1 to pin 47 of IC501 (microprocessor 14D0899) on the MCV-B board. The other is from CL-4 pin 8 to CL-5 pin 3 and back to the same pin on the microprocessor (ie, pin 47).

On careful examination, I found that the former route was not connected, as links J519 and J503 were not fitted on the MCV-B. I can only assume that this was for different model options and function variations.

OK, I now knew for sure what was causing the problem but fixing it was another matter. After already spending a lot of time on this job, I was severely tempted to just leave it with CL-4 pin 8 unsoldered – after all, everything worked.

However, unsolved questions like this really bug me so I set about finding what it was that was causing the audio mute line to go high.

First, I checked every part of the rail, especially D402 which comes off the D-V line (whatever that is). Anyway, unsoldering this diode made no difference. It looked as though something was telling the microprocessor to mute the line but what?

I tried all sorts of things and even phoned Teac technical support but no-one could give me any ideas.

I finally had a bit of luck when I tried to tune in a signal generator. The preset search would tune in all the stations OK but wouldn't stop on any of them. This was a major clue, so I had to investigate how the self-seeking tuning system worked. This wasn't easy because, as you will have gathered by now, there is no glossary on the abbreviations used in this set for each of the lines.

Q704 controls the 0-33V feed to pin 4 (VC) of the tuner and the base of this transistor is controlled by T-DAC via CL-6 pin 3 from pin 60 of the microprocessor.

Once again, I hit a brick wall. I had no way of knowing what inputs to the microprocessor were controlling its outputs.

However, this time I was luckier in guessing that the SD line on pin 46 was the "Tuner Video Signal Sync Signal Output" which comes from IC703 LA7210 pin 10. The video signal goes into pin 6 and in the tuner search mode gives an output from its comparator to the microprocessor which locks in the stations and also controls the AFT up/down lines from IC502.

Though all the DC voltages were correct on IC703, there was no signal output from pin 10. I changed C713, C716 and C714 but it was actually the IC that was the cause of the problem. A new one fixed the self-seeking tuning as well as the audio muting and the sound was restored.

Panasonic MX-3 chassis

Just recently, I seem to have done the Panasonic MX-3 chassis to death, which is not what you would expect these days from a quality 5-year old product.

The first one was a TC-21510A that was "dead and whining" (who wouldn't be?). This luckily turned out to be just C805 and C825, two small electros that had dried up in the power supply.

The next one was worse. The same electros caused Q805 and R826 to destruct, as well as Q803 (the protection relay driver) and D835 (the 56V protection zener on the 42V rail). IC801 (an SE090 IC regulator) and optocoupler D820 (P82501) had also failed.

The third one of the same model was even worse. It came with the complaint "Dead and Pulsating" (surely an oxymoron?). This set had already been around a variety of different technicians before me so I felt like a bunny in taking it on.

It turned out that C805, C825, Q805, Q803 and D835 had already been replaced and we had 42V, 22V, 5V and 90V rails but that was all. Most people would say they could get a TV to work with that much but not on this model. The power supply is unusual (it's similar to the supply in some NEC sets) and is integrated with the line output stage. It's also all at live mains potential; ie, it's hot!

The secondaries are taken off via T802 and the flyback transformer T501. The only other transformer which separates the COLD from the HOT is the horizontal driver T566 – even the deflection coils are at HOT (horizontal) and COLD (vertical) levels! Because of this, precautions have to be taken if you don't want to get zapped.

Anyway, this particular set had no line drive, or rather, very little. There was enough at first to think you had some but a few checks with the CRO soon showed that Q565, the horizontal driver, just wasn't being driven hard enough. As a result, I spent some time trying to find out why there was insufficient line drive.

Eventually, I concluded that the problem involved either the jungle IC (IC601, AN5192K) or its associated circuits. First, I had to check if the voltages coming into the IC on pins 14 (9V), 23 (5V) and 47 (5V) were OK. They weren't because they are all derived from the line output stage via D511 (20V) and via IC805, IC806 and IC807. The horizontal oscillator is started by the 22V rail going through R553 pin 51 of IC501 (HOR-REG). The 6.6V was present and the oscillator crystal (X554) was on frequency.

Understanding the circuit and working out what was supposed to happen and what wasn't had taken a lot of time. By now I was fairly convinced that the problem was insufficient line drive from the jungle IC and so I fitted a new one.

This fixed the set – well, nearly. I now had sound and picture but no OSD (On Screen Display) and the set cut out after about five minutes.

I fixed the cutting out fault first when I accidentally burned myself on IC802, a 7805 5V regulator that was getting extremely hot. Freezing it stopped the set going off but finding the cause of the problem is a story in itself.

The IC still supplied 5V when it got hot and although the rail didn't measure short circuit, it was obviously low impedance. This 5V rail mainly supplies the microprocessor, the EEPROM and the power LED, all of which were working.

So there I was, suspecting and substituting components on the 5V rail but to no avail. I was about to change the microprocessor because of the OSD fault when I noticed a small whiff of smoke appearing from a very small black component next to IC1101. This black component turned out to be what was left of D1101, a MA4056L 5.6V zener diode.

Replacing this reduced the stress on IC802 and it no longer got very hot. This stopped the set from cutting out so I moved on to the OSD problem.

The red, green and blue signals normally come out of pins 33, 32 & 29 of the microprocessor but these were absent. The other signals I was looking for were the blanking pulses on pin 31 and the horizontal sync pulses on pin 30. The latter were also missing and are normally applied via Q1135 from pin 6 of the flyback transformer. When I traced them with the CRO, the pulses stopped at Q1135 and replacing this 2SC945 transistor restored the on-screen display.

Two televideos

I had two Panasonic TC-21SV10A Televideos (MX-3V chassis) in quick succession, with very similar faults. The first came from a school and, among other things, had suffered a cracked cabinet (apparently, it had fallen onto the floor). Both sets also had video tapes stuck inside them which couldn't be ejected.

The VCR chassis employs a K-mechanism and is tucked under the TV chassis and power supply. Invariably, by the time you get the chassis out, remove its covers, reconnect the four leads (E23, E24, E26 & E27) and switch on, the resultant jolting has freed the mechanism and the cassette ejects properly. And no matter how you try it, it is almost impossible to recreate the fault – unless of course, you send it back to the customer when it will do it immediately!

Fortunately, I am pretty familiar with the K-mecha video deck and the most likely culprits are the loading motor and the mode select switch. The deck is removed by undoing four screws at the bottom and three at on top and then unplugging all the connectors. The loading motor frame can then be removed by undoing one screw and two clips, after which the motor can be unclipped from its support frame.

The worm gear (DG0866) and the pulley bush VDP1434 (in particular) have to be replaced – the latter cracks and then slips on the motor shaft.

The mode select switch (VSS0365) is held on by one screw and several clips. It is important to replace it with the mark on the mode switch and the notch in the frame or the arrows pointing at SE (4.30), with the screw located at N (12.00).

In this particular case, after doing all this and reassembling it, I found that there was intermittently no picture on playback and there were no tape functions. When I removed it the second time, I eventually found a 7cm-long hairline fracture in the lower main board near IC2502. I repaired all the cracks in the board pattern and reassembled it again but I still wasn't out of the woods as it was stuck in Timer Record Mode, with the orange LED on.

Setting the clock fixed this problem but I also checked the TV and video by going into the service mode or market mode and self-check modes. To do this, you use the remote control to select MENU, FEATURES, OFF TIMER O and then set the time to either 30, 60, 90 or AUTO minutes.

You then simultaneously press RECALL on the remote and volume down on the set. There are four CHK settings to adjust all the controls. POWER OFF returns it to Normal Mode.

The self-check function is enabled by simultaneously holding down the right arrow on the remote (in the TIMER menu) with the volume down button on the set. This brings up a screen with DATA for the TV and VIDEO which is compared with a list in the service manual to indicate faults. You then switch the POWER OFF to resume Normal Mode.

Bizarre fault

As mentioned, the second TC-21SV10M also came in with a tape that wouldn't eject. Performing all the above fixed that except that it had a bizarre fault. Whenever there was a snowy picture, either in EE Tuner mode, or in video playback, the screen would go intensely white with a black horizontal line in the centre.

Normally, the set is designed to display a blue background screen if selected in the FEATURE menu but this wasn't happening. Strangely, after going through the Market Mode and self checking modes, the fault completely cleared itself. Just why it did this I don't know as no fault was ever reported in any of the menu pages.