One evening, just as I was closing up to go home, a young man in a fairly excited state arrived and began knocking on the door. I let him in and asked him what he wanted and he replied "What do you charge to provide and replace a part for a TV?"

It was the way he phrased this that threw me. So I asked him what part, which set was it for and where did he want it putting?

"Look", he explained, "I have a TV - in the car - it has a part missing and I want you to put it in".

I thought I had better check this out and so we went over to a small 3-door hatchback. He opened the tailgate and inside was a monster TV. Truly, I don't know how he managed to get it into such a tiny car.

The set must have been an 80cm model and was face down. The brand was Grundig but something wasn't quite right about the whole thing. I persuaded the animated young man to start at the beginning and tell me what happened.

Items Covered This Month Sanyo C25PG51 TV set (AA1-A25 chassis) Sanyo CPP-2140 TV set (A3-C21 chassis series) Sony KV-2113GE TV set Panasonic TC-36PM10A (MX-7 chassis) TV set STA TVP-50505K 46cm TV set Philips 29PT9418/79R TV set (MG3 1A chassis)

Eventually, the story came out. The set had belonged to a company and had been sent out for repair by a service organisation. However, the quote to fix the set had been too high and the company decided to buy a new TV instead. Our young man had offered them $200 to purchase the old set and brought it to me to be fixed.

It was then I noticed what was really wrong with the set - there were large holes in the back. These are normally filled with an A-V connection panel and a closer inspection revealed that the entire chassis was missing! I didn't think my potential client was prepared to pay for me to replace that!

Somewhat humbled, the young man went back to find out where the rest of his set was - I never saw him or the set again!

Faulty Sanyo

It seems that most problems associated with Sanyo CTVs, which are generally very reliable, are caused by faulty high-value resistors, in particular 120kΩ 0.5W units. For years, these resistors have failed in the start-up circuit for the switchmode power supplies from the A1 series to the A8 - well over 10 years, in fact. Perhaps it's Australia's harsh weather conditions that make this one value fail so often?

Recently, I had a Sanyo C25PG51 (AA1-A25 chassis series) with no picture or on-screen display. It didn't take long with an oscilloscope to realise that the signal was getting as far as the CRT panel where it disappeared into a morass of transistors. All the voltages seemed correct everywhere I measured and I was beside myself with what could be wrong.

I was only saved when I spied what I thought was one of these resistors. Actually, it turned out to be a 220kΩ unit and it was marked R792 on the PC board. It measured open circuit and replacing it fixed the problem.

Later, I tried to identify it on the circuit diagram but was unable to do so with any degree of certainty. That's because all of the components are marked on the circuit with a 4-digit code after the "R' - ie, there is no R792. I can only assume that it is either R2792 or R1792, the latter a 120kΩ unit from the +210V rail to the base of each of the three push pull video output stages.

Another Sanyo

Mrs Talbot's Sanyo CPP-2140 (A3-C21 chassis series) had no colour on any channels, including the AV inputs (I always interrogate the customer on these little details as it may only be an aerial or tuning problem). Sometimes, you can be lucky and troubleshoot these in the home but, more often than not, it needs to be in the workshop with an oscilloscope and a service manual. Well, at least I had the former and a circuit for a similar chassis model.

The main difficulty, as far as I am concerned, is that there is no easy way to troubleshoot the jungle IC (IC101, LA7680) for colour faults. The colour killer circuit at pins 41 (6.2V) and 39 (5.6V) could not be overridden, either by tying them to the 9V rail or to ground. This would have been really useful as I could then work out from the picture which area to attack.

For example, if the colour was running, I could examine the reference oscillator, burst and ident and line pulses for synchronisation.

I started by checking the DC voltages around the jungle IC to find them all pretty well spot on. I then checked the reference oscillator crystal on pin 16 with the CRO and a frequency counter - it was spot on at 4.43MHz.

By connecting a colour bar generator to the AV socket, I could see the chroma signal on pin 40 but nothing was coming out at pin 14. Obviously, the bandpass amplifiers were not being switched on, but why?

Because of the nature of large scale integration, it is not possible to have access to a lot of the circuitry inside such an IC or even guess how it works. Because of the work involved, I was extremely reluctant to change the IC but was finally about to do this when I noticed that by freezing and touching the components feeding pin 17, I could occasionally could get a flash of colour.

Pin 17 APC F (6.2V) feeds the VCO and ACC parts of the decoder. This in turn is fed by a twin resistor divider from the Vcc rail - R263 (220kΩ) and R264 (3909kΩ). The latter had gone very high in value and replacing it dropped the voltage to 5.9V and brought the colour back. However, just why lowering the voltage to an unpublished figure should fix the problem is beyond me!

The old Sony

An old Sony KV-2113GE came into the workshop with the fault description "sound low and unclear".

When I switched it on, the picture was fine but the sound was intermittently poor. Suspecting electrolytic capacitors and dry joints, I gave the power supplies and audio output stages a good going over but the fault remained, even when I applied freezer and heat.

I was about to replace the audio output IC when I decided to remove the loudspeaker, as the leads weren't long enough to enable me to work on the chassis and still have it connected. It was then that I noticed that touching the cone of the loudspeaker made a difference.

A close examination revealed that there was a bad connection between the copper braid and the voice coil that's attached to the cone. After messing around with it for a while, I ended up replacing the loudspeaker, which fixed the problem.

Bread and butter

Panasonic IC451 vertical output ICs have helped put bread on the table for me and others for many years; in particular, the dry joints on the inadequate solder pads provided on the PC boards on all models. That is, until recently - from the MX-7 chassis onwards, the soldering has been perfect, only now on the MX-8 chassis, the 1W 1W fusible resistors randomly blow and on the M-17 chassis, the LA7833 destroys itself, taking the feed resistor as well.

On the EURO3 chassis, the resistor is replaced by a "button" fuse which blows when the IC fails. The C150A chassis has the worst combination of faults. As the dry joints deteriorate, IC451 (LA7838) and C455 (100μF, 35V) begin to fail and the set's protection circuit starts to operate as the 12V rail becomes loaded down under the strain. This gives the effect of the set intermittently not wanting to start.

Ironically, when the set does start, the picture and sound are excellent. If, at this stage, the dry joints aren't fixed and C455 immediately replaced, the IC will fail completely.

The problem is that you spend hours trying to troubleshoot an elaborate power supply when it is a vertical timebase fault. This is compounded with an intermittent shutdown problem on earlier models due to the overcurrent protection circuit (D836) being too sensitive. R835, R848, C831 and C838 were changed in values to overcome this.

Mrs Lawson's Panasonic

Mrs Lawson brought in her Panasonic TC-36PM10A (MX-7 chassis) portable, complaining that the set wouldn't go. Fortunately for me, she mentioned that just before it finally wouldn't work anymore, the width was intermittently too wide.

When I switched it on, all I got was the LED flashing slowly - obviously the protection circuit was working. The problem with protection circuits is that they are difficult to find, understand and disable without causing major damage, and all too often are faulty themselves.

In this instance, I was extremely grateful for Mrs Lawson's final comment about the width. After measuring the main B+ rail to see it was correct and steady, I concentrated my search around the horizontal deflection circuit. Unplugging the horizontal defection yoke plug (DYH) restored the sound and I had a vertical white line down the screen.

Well, it could have been that the yoke was short circuit but it was more likely to be a fault in the east-west correction circuit. I checked D556 and D557 and was working my way towards IC701 (TEA2031A) when I examined L701. There was a slight discolouration on the outside and it looked as though it had been under stress. My circuit showed it as an 8.2mH coil but my meter read only 0.1mH. A new one fixed the problem completely.

Ooh, la la

An interesting set came in last week. It was a French Samsung televideo all the way from La Belle France. Mrs Serviceman was completely taken in by Jean-Paul's accent and there were hormones and pheromones flying everywhere. In fact, she was so besotted by him that she barely noticed what it was she was booking in.

It, in fact, was a 1997 TVP-50505K STA 48cm set using a SCV11A chassis which was made in Spain (a multistandard model). The fault was that it tried to start and cut out after five seconds.

I managed to hire a service manual for a similar model using the SCV11 chassis but its power supply bore no resemblance to this set. I knew that the main B+ should be +125V and I soon discovered that it came from D812k and the switchmode power supply. My meter showed this momentarily reading +170V before it closed down.

Without a close circuit diagram to refer to, I knew it was going to be difficult. I started by replacing four electros in the "hot" side of the power supply and two on the secondary to no avail. I then replaced the STRS6707 chopper IC (IC801) but this also proved futile.

I was trying to figure out how the switchmode power supply worked when I realised there was an optocoupler and the control circuit was on the secondary side. It was at this point that I was exceedingly lucky to see and recognise Q803 as an SE125 - a 125V control IC driver. These are notoriously unreliable in all brands of sets and replacing it cured the fault.

The next problem was trying to tune the set. It is fitted with two parallel digitally operated tuners on the main chassis.

I can only speculate that one was for the TV and the other for the VCR. Of course, the menus on the OSD were all in French and so was the instruction booklet. Eventually I found the menu to change the language to English and was then able to set up the tuning and controls, though I didn't discover how to tune the second tuner.

Jean-Paul was ever so delighted and Mrs Serviceman was on the verge of eloping with him when his gorgeous better half arrived and put them both in order!

Dead & urgent

Today, I had a very unsatisfactory outcome with Mrs Taylor's Philips 29PT9418/79R (MG3.1A chassis) which was brought in as dead and marked urgent.

When I checked it, the set would switch on to the Standby Red LED mode. When the remote power button was pressed, the Red LED became green, then yellow and then red before flashing.

According to the service manual, the way to fix this is to just plug in your DST (Dealer Service Tool) and it will tell you what is wrong. The trouble is that this is rather expensive to buy at around $500 - assuming that it's available. It is rather like an ordinary remote control, except that it has 2-way communication with the TV and can read and write data to and from the set.

Anyway, I had to troubleshoot this set the old-fashioned way with a multimeter.

Access to the underside of the PC boards is extremely difficult, the exception being the power supply B board on the righthand side. What I needed was good access to the line output transistor but this required removing an awful lot of plastic architecture and support frames for each of the boards. What's more, these boards are interconnected via concealed locking plastic clips as well as by screws, not to mention cables in cable ties. It is very easy to break a lot of this if you cannot fathom out how they all interlock.

The first thing was to establish to what extent the power supply was working. It had to be working a little for the LED to come on at all (ie, the +5V standby). I tried to measure the nine voltage output rails but the power supply closed down too soon to allow this. I then tried overriding the protection system by unsoldering D6271 to transistor 7250 but this made no difference. However, disconnecting D6270 from the Standby rail made the power supply stay on and I was able to confirm that all rails were working OK.

The next step was to examine the deflection board (A1). After a very long time, I managed to detangle the wiring and plastic support brackets to check out the soldering. All was OK and so I checked the line output transistor (7421, BU2520DX) with an ohmmeter - it too was OK.

Next, I checked to see if there was +141V to the collector of this transistor. This was correct so I decided to check the line drive. First, I had to reconnect protection diode D6270 as there would be no output from the oscillator. I then traced the line drive signal route from the small-signal panel surface-mounted IC via transistors 7407 & 7411 and transformer L5410 to the line output transistor.

When I switched the set on, the waveform was only there momentarily but it was good and strong until finally I connected the probe to the collector of the line output transistor. At first it was like the rest of the set, with it trying to come on but not staying on. However, the second time I switched on, the whole set suddenly came on and stayed on.

No matter what I tried to do now, the set just wouldn't go off other than by using the appropriate remote control or main switch. I bashed it, froze it, heated it - I tried everything. The !@#$% set had fixed itself!

Finally, I reassembled it and left it on test but it just refused to show the fault again.

Getting this far had taken me three hours and now the set was fixed for no apparent reason. I was left to contemplate what to charge, what to say and what to guarantee. Somehow, measuring the waveform on the collector of the line output transistor had shocked a faulty component into working - but which one? I decided to replace the line output transistor and I told Mrs Taylor what had happened and why I could only apply a limited warranty. She was quite happy with that and she paid for the transistor and labour.

I was of course extremely unhappy and am now just waiting for the recall. Subsequently, we have had a lot of the later Philips 29PT6361/79R sets which employ an A10A chassis made in China. This set has a lot of intermittent faults which can be identified if you can get into the SDM (Service Dealer Mode) and then the SAM (Service Adjustment Menu) and read the error codes.

Inevitably, it turns out to be the SSP (Small Signal Panel) that needs to be replaced, as this cannot be repaired (the surface mounted chips are just too small). However, the MG3.1 hasn't yet needed this.

Share this Article: 

Digg this story

Seed Newsvine

Post to Reddit

Save on del.icio.us

Email