Moving premises can be a real pain in the you-know-what but the move really has been worthwhile. Among other things, it gave me the chance to throw out a lot junk and other stuff that I didn't need, so that I can now work in an uncluttered environment.

It's amazing how much stuff you can accumulate in this business over the years and I really welcomed the excuse for a good "chuck out". Of course, I kept all the real treasure - you never know when it might come in handy!

Items Covered This Month JVC AV-G25AU 59cm (MZ2 chassis) Philips 25GR6765/75R (G110-S chassis) Panasonic TX-33V30X (M16MV3 chassis) Grundig M82-4986/9S/PIP (CUC 3840 chassis)

The whistling JVC

Mrs Blandford complained that her 59cm JVC AV-G25AU (MZ2 chassis) had no picture and "whistles". On the face of it, I thought that this might be quite interesting because I had never heard a TV whistle before. In reality, the set was actually pulsating and gave out a protesting noise that very loosely could be described as whistling.

Having repaired a troublesome AV-G21AU some months ago, I had a good idea what the problem was. The line output transistor (Q522, 2SD1878-YD) was short circuit and the usual cause was dry joints on blue ceramic resonator CF561.

I dutifully replaced the transistor, resoldered the resonator and, fully confident of the outcome, switched the set on. I then confirmed that all was Mickey Mouse and went to have a cup of coffee. When I returned 20 minutes later (I don't like to be rushed over sacred rituals), I was mortified to find that the set had rejoined the choir and was pulsating just as before.

So what had gone wrong? I installed another expensive transistor and then began resoldering the set in an effort to cure the fault. However, what I didn't realise was that the switchmode power supply reservoir capacitor (C910) remained charged at +295V for a very long time after the fault had occurred. As I worked on the set, my arm suddenly strayed across this vindictive component and yes, you guessed it - it bit me!

That 295V on my otherwise pristine arm was very painful but worse still, the solder I was carrying fell out of my hand and shorted out parts of the power supply. This resulted in a short fireworks display, even though the set had been turned off for quite a while.

When the smoke cleared, and after I had checked the obvious, I found that the power supply no longer whistled. It fact, it couldn't do anything at all - except give me another shock, if I was stupid enough to try. I wasn't - instead, I discharged the beast using a globe before carrying out a series of DC voltage checks on the circuit.

Eventually, I discovered that Q901 (2SD1853-T) and D908 were leaky. The former is a special Darlington pair and the latter a 7.5V zener diode. I had to order the transistor in but after I had fitted them both, the set was still dead.

Remembering to discharge the capacitor again, I then replaced IC901 (STR-S6707), after which the set agreed to fire up. This time, with my hand near the master on/off switch, I waited to see if there was anything untoward that was causing the destruction of the line output transistor. A few minutes later, the set started to make noises and the picture began to tear.

During this time, the +114V at test point TP91 remained constant on my DMM, so the power supply itself was obviously OK. And that in turn meant that something was affecting the line output stage.

My suspicion was again drawn to CR561 which I though might need replacing. However, before doing this, I resoldered all the pins to IC201 (the jungle IC), paying particular attention to pins 14 & 17 which are adjacent to the crystal.

This finally fixed the problem and Mrs Blandford was able to have her "telly" back after a good long soak testing.

Email in the fault

Amongst all the huge technological changes within this industry that happen daily, a slightly novel approach to reporting faults has arisen which I suspect will take off and be the way of the future.

A 1990 Philips 25GR6765/75R (G110-S chassis) appeared on my bench with a fault description "Intermittently, the top disappears when changing channels". So I switched it on, expecting a vertical linearity fault.

This set is well known and, although rather ancient these days, is not considered difficult to repair. In addition, you get used to the terrible and widely varying fault descriptions from clients, so you never quite know what to expect.

Well, I waited and waited but nothing happened. The picture was excellent considering the set's age, so I put it to one side but still within visual range while I got on with other jobs.

After three weeks, the set hadn't so much as blinked incorrectly and I put the "fault" down to either an antenna problem, or bad connections or interference. And so the set was returned to its owner with a note to that effect.

Days later, I got an email with a set of attachments. Apparently, it played up about 24 hours after going home and to prove it, our enterprising client had photographed the effects with his digital camera.

The photographs were quite clear, although the cause of the fault wasn't. The accompanying email said that it performed perfectly until there was an advertisement or switching from the studio to an outside broadcast, which seemed to invoke the symptom. And according to his email, the TV would sometimes come good when he approached it but would then revert to the fault when he walked away.

This was getting "curiouser and curiouser", as they say in the classics.

Anyway, the set was returned to the workshop but again refused to play up. However, the client's photographs showed horizontal lines of different contrast going up and down the screen, which definitely looked like a vertical timebase fault.

The vertical output stage of this set is pretty reliable but I did change C2813 and C2814 (1500μF 35V), as well as C2981 (1000μF), as these are known to give trouble. I also checked the PC board for electrolyte corrosion and dry joints, before putting the set aside to soak test.

It was two weeks later that I first got to see - if only momentarily - the fault, which was indeed like the photographs except that there was no colour when the fault occurred. I now suspected a vertical fault that was somehow impinging on the colour decoder, possibly involving the vertical blanking pulses.

To check this theory, I connected the CRO to the sandcastle line from IC7705 and monitored the combined vertical and horizontal pulses. However, the sandcastle pulse didn't vary and so I couldn't determine if it really was a vertical deflection fault at all.

Next, I decided to look at the supply rail voltages to see if problems were occurring there. Nothing showed but I did change a few electros that looked a bit suspect, just in case - especially C2175 (1000μF 50V) on the +32V rail. It made no difference.

Fortunately, I own several of these popular sets and so I decided to swap large chunks of the set in order to eliminate these areas.

First, I swapped and then removed the Teletext module. I also swapped the stereo decoder and even the CRT socket but the fault gradually became more frequent, which was good because I could measure and check more of the waveforms. However, because it was intermittent, there were many times when the fault would actually stop for days before reappearing, which meant tying up a lot of test equipment for a very long time, waiting for the fault to occur.

Next, I socketed and swapped IC7705 (TDA2579A jungle timebase), IC7278 (EEPROM), IC7550 (TDA3562A chroma decoder), IC7355 (TDA5850 video switching) and IC7325 (TDA8341 video IF). The fault continued, even with a generator connected to the AV input SCART socket.

And then, by sheer luck, I finally managed to get an idea of where to start when I checked the CVBS video input to pin 8 of IC7550 and saw the waveform vary wildly.

Having secured a toehold on the possible cause, I decided to monitor the waveform from pin 12 of IC7325 to pin 8 of IC7355, and from pin 5 (of IC7550) to pin 8 of IC7550. I then tried heating and freezing all the components but was continually thwarted by the fault intermittently coming and going. I even swapped over the SVHS module and some of the leads where I thought the problem might be lurking but it was just another blind alley.

By this stage, I had been working on this set on and off for about seven weeks. And then one day, an arrogant technician friend who was working with me at the time claimed to have fixed the fault in less than 10 minutes while I was out.

Incredulous, I asked him what it was and how he had managed to find it - especially as the set was giving a perfect picture again and had been doing so all morning. Seriously, he claimed that he just looked at the circuit and knew straight away where the problem was. He then showed where a lump of clear glue had held a wire link over the PC board. I then asked him where exactly on the circuit this was (you know, the one he looked at first) and I wanted to know what components were affected.

He proceeded to show me, in a large sweeping circle, an area of the circuit covering about 50 components. By now, he had lost all credibility and I told him he was talking rubbish. The fact is, he would have seen the glue first and the circuit second, and the glue was a later type used by Philips that no longer conducts with age like the old brown stuff.

In short, it was only by luck that the set was working.

Anyway, I decided to play along for a while and boxed the set up and left it on display in front of him. An hour later, his luck ran out - the fault was back and he fell extremely and uncharacteristically quiet.

Not being backward in coming forward, I chose my moment to rub it in further. His grim silence was eventually broken when a new original idea popped into his head. "Ah", he said, "I was only joking!"

Back to the grindstone - the fault, when I was able to measure it, seemed to start from near IC7550 but I had already changed that. The signal also went to the Teletext board, which I had removed, and it also went to pin 5 of IC7705 which I had replaced. From there, it went to a surface-mounted emitter follower transistor (Q7350) and then to the SVHS panel, before going through the luminance delay line to IC7550.

I had already changed the panel, so that left the transistor and the delay line. Because it was easier, I changed the delay line first but it was the surface-mount transistor that was causing the trouble and a new one fixed the fault.

Sony BG-1s chassis

The Sony BG-1S series of chassis are pretty reliable sets which employ a similar switchmode power supply to the JVC set mentioned earlier. However, when the power supply fails, it often takes out several parts.

Recently, I have had a series of these where IC601 (STRS6707) fails, disintegrating R629 (33Ω) and taking out D607 and/or D609 (DNL20). I suspect that C624 (1000μF) and C623 (220μF) may be the culprits as they are often also found to be faulty but guessing which failed first is beyond me.

When this lot fails, fuse F1610 3.15AT goes black too, though the other day I had a real beauty. The fuse had melted a gap of about 0.25mm in the middle, which is barely visible to the naked eye. Replacing this fuse was all that was necessary in that particular set.

Sometimes R601, R602 and R611 also fail. The BG-2S is more reliable but has a ceramic capacitor (C820, 1000pF 2kV) across the line output transistor which sometimes fails.

TV set sizes

When I refer to a large 80cm or 34-inch television, please bear with me, as these sorts of numbers can be confusing. For example, if a picture tube carried a label that read "M78KPH566X", you used to be able to rely on this as meaning that the viewable diagonal was 78cm. However, the manual for the Panasonic TX-33V30X set (M16MV3 chassis) that uses this tube states that it is a "type 33 (84cm) measured diagonally".

Other manufacturers use the imperial system and call it a 33-inch set. Because of these differences, I can only generalise and quote the advertised size where I can find it.

Anyway, there I was with another large heavy (60kg) "telly" on the bench (what do one-man-show TV repairers do about moving these? No wonder when we get old we all suffer from bad backs and poor eyesight!).

The picture and sound were great off-air but on AV there was little colour and no luminance. Instead, there was an effect I can only describe as like "shooting stars"! These appeared as lines with a bright spot at the end on the lefthand side - very bizarre.

I decided to start with the AV input "U Board" and trace the video signal with a CRO into the "C Board" (AV Control). The only trouble was access - basically, there was none! Fortunately, I had a set of extension cables and when I finally had the "C Board" out on the bench with the CRO probe in my hand, I switched the set on.

Unfortunately, the fault had now disappeared and the set was showing the colour bars perfectly. After examining it for dry joints, I refitted the module into the set to see if the fault would reappear. This was going to be tricky to get to the bottom of because it really needed to be in the set to give the fault.

I persevered again with the module on the bench (with the extensions) and tried to recreate the fault. It looked as though a component on the "C Board" was radiating an interfering cross-modulating signal. I ran my moist fingers all over the board and noticed that when I touched an area near the bottom corner (from IC3003 to the C1 plug corner), the fault began to recur.

I tried to nail it down but it wasn't possible with my big fingers, so I used a small screwdriver instead to pinpoint different components.

I started at "A/V 3.58/Other" analog switch IC3003 (TC4066) and found pins 6, 8, 11 & 12 to be sensitive, but it got more so when I touched the base of Q3039 and even more so when I got down to Q3046 (UN4213).

The latter is a surface-mount NPN transistor (with inbuilt 47Ω resistors in its base and emitter circuits), which controls the switching signal to the IC. Replacing this transistor and refitting the module fixed the problem.

A slow Grundig

A Grundig M82-4986/9S/PIP (CUC 3840 chassis) was brought into the workshop with the complaint that it was slow to come on.

Once I got the back off, I connected a true RMS heater meter to the heater filament pins (9 & 10) of the A76JTS90X03 picture tube and, as I suspected, the voltage was very low (about 3.5V) at switch on. However, there was also a noise coming from the motherboard, which suggested that something was under stress.

Next, I measured the main voltage rail (A) from D656k. This was spot on at +152V and as I was watching, the noise stopped and the heater voltage moved up to 6.3V. However, the "A" rail remained constant. This told me that the switchmode power supply was OK and so my attention moved to the deflection stages.

This set uses IC550 (TDA8140) as the horizontal driver and it goes direct to the line output transistor. It is fed by a + 12V rail to pin 2 and right next to it is a 100μF electrolytic filter capacitor (C507). Its cover had peeled back and it looked highly suspect. I froze the capacitor when the set was running normally and immediately the noise returned and the picture dimmed.

Replacing it fixed the problem completely.

Share this Article: 

Digg this story

Seed Newsvine

Post to Reddit

Save on del.icio.us

Email