Email Address:

Lost your password?

This is the legacy website; please use the new website.

Serviceman's Log

DOA - that's "Dead On Arrival"

It’s not uncommon for sets to turn up in the workshop DOA (dead on arrival). But there are varying degrees of DOA, ranging all the way from completely dead to not quite dead!

A large 2002 Philips 32PW4523 TV set was brought into the workshop and pronounced DOA. This popular set uses an L01.1A chassis which isn’t quite so popular with technicians as it can sometimes be a bit hairy to fix.

In this instance, I seemed to have got the full treatment. Fuse 1500 had blown and this told me that the switchmode power supply was also blown. I replaced it, along with chopper FET 7521, its driver Q7522, the diode in its base (D6523 and D6525) and R3530.

To add to my woes, IC7520 (TEA-1507), R3521 and R3523 were also faulty but it still wasn’t done with me. The set was still dead, with no red LED. There was also no drive pulse on pin 6 of IC7520, the power supply controller.

Click for larger image

The voltage on pin 1 (Vcc) of this IC should have been greater than 11V (13.8-16.8V is marked on the circuit). In fact, it was less than 5V and this turned out to be due to C2526 (470nF) being leaky. A new one brought on a little life, with the red LED now flashing.

I then replaced the infamous C2455 (47mF) in the line driver stage as a matter of course, as it dries out. Then I noticed a very obvious pregnant flyback transformer TR5445 (see photo). It must have got very hot to melt the plastic case in this fashion. I replaced this and the line output transistor, before attending to the east-west FET (7400) and R3411 and D6460. But there was more to come.

R3344 and R3346 (22W) in parallel were burnt out on the EHT-info line on the CRT Panel B1 from the CRT aquadag, while VR3345 which connects the aquadag to ground was also faulty.

When this set spilt its guts, it must have been spectacular, with so much "blown up". It is really hard to know the sequence of events but fortunately, that was the end of its troubles.

Another Philips

The 2000 Philips 29PT9418 TV is quite a popular set, using the MG3.1A chassis. It’s a top-of-the-line TV with some additional options such as cordless/wireless Dolby 5.1 surround sound (it has about eight microprocessors), not to mention full digital features. Consequently, it is a very complex TV and repairing it is not for the faint-hearted.

Click for larger image

Until now, I had only seen a smattering of sets with fortunately fairly simple problems like dry joints on pin 2 of TR5204 and the vertical output IC (IC7600). Recently, however, two sets turned up with more complicated problems.

The first set I had arrived from a colleague who had given up on it. Its symptom was that the set was dead with a flashing red LED. The sequence for switch-on is: solid red = standby, amber = start-up sequence, solid green = on and flashing red = protect mode.

This was one of a series of models that Philips, with great fanfare, introduced with a Dealer Service Tool (DST), which is a remote control (RC7150) that can interrogate the set and communicate with the microprocessor and EEPROM, even when the set is in protect mode. It also has the ability to automatically tune the set to a predetermined arrangement.

Unfortunately, neither this, nor the ComPair (Computer Aided Repair) I2C system designed for these models is available in Australia. The reason, I gather, is that they are too expensive and now possibly not available. Besides that, the management here believe that if you are a trained technician, you just don’t need them.

I disagree. Anything that makes our lives easier and the jobs quicker must be economically viable. Indeed, in the early days of colour TV, most German TVs had plug-in diagnostic systems which worked very well.

Because these useful tools are not available, the SAM (Service Alignment Mode) and SDM (Service Default Mode) are only available by delving deeply in and finding (often unmarked) service pins near the microprocessor. When these modes are activated, some of the protection circuits are made less critical, occasionally even allowing the set to actually come on.

As I also had a working TV in stock, I was able to quickly establish that there was a problem with the power supply – or more specifically, "Top Supply Panel B". I immediately checked for the dry joint on pin 2 of transformer 5244 but it was OK, so I then ran this SOPS (Self Oscillating Power Supply) module on the bench alone, with a 100W globe across D6224 to monitor Vbat (+141V). This showed that the power supply was pulsating on this rail. I then found that all the other voltage rails (16V, 11V, 8.6V, 5.2V and ±VS) were also pulsating and there were no obvious shorts.

Next, I tried swapping components with the good supply, starting with the electrolytic capacitors. I then made sure that the DC-protection was disabled by shorting the line to ground, after which I tried removing or substituting the crowbar circuits (SCR D7232 and various zener diodes) on all the rails. I also checked out the feedback control circuit involving D7212 (TL431CLP) and the optocoupler, before swapping the control IC, C2204 (680pF) and also L5240, L5211 & L5212 and their series capacitors.

Going nowhere

I was really going nowhere until I noticed that "Va", which should be at 18V and is the "take over" supply to the control IC, was very low and pulsating. C2203 and D6210 were both OK which made me suspect the chopper transformer (5202). This was swapped over and it turned out to be the culprit, the power supply now bursting into life.

So what was wrong with the faulty transformer? I fully expected to find shorted turns where the wires come out to the PC board lugs but these were all in order.

Click for larger image

However, when I looked on the top side of the transformer, there were two wires that were part of the "Va" voltage winding and where they crossed over a plastic ridge, there was a very faint imperfection in the wire. On very close examination, the single core leads were both cut but were still making a "touch connection". I can’t be sure whether these had been cut on purpose or by accident but resoldering them fixed the problem.

Reinstalling the supply in the set was straightforward but there were still other problems to solve. The set would try and come on, going through the red, yellow and green phases of the standby LED but ending with it flashing (the red LED is supplied by switching the +5V Standby rail from the micro and the green LED by switching the +8.6V rail).

I then found that initiating the SAM or SDM modes by shorting pins 1 & 2 or 2 & 3 of connector O356 on the Small Signal Panel (SSP K board – K7 circuit) allowed the set to come on. The error buffer then showed a 068 error, denoting a fault on the 8.6V rail. Clearing the buffer and the doing a self-diagnostic test cleared this error and no other errors were indicated.

Switching the set off to standby with the remote and then switching the set on again produced the same results as above, with the 068 error reappearing. Occasionally when switching the set on from cold, it would fire up without putting it into the Service Mode.

The east-west pincushion correction wasn’t working and this was fixed by replacing FET Q7480 (STP16NE06). I also found that there was nothing coming out of the external surround sound speaker connections and so I replaced IC7760 (TDA2616Q). However, neither of these faults affected the 068 error and the failure to start up normally.

Feeling somewhat frustrated by now, I then swapped all the remaining boards with those in the good set until the fault was transposed with the SSP (small signal panel). I then measured all the voltages on the SSP – 3.3V standby, 3.3s, 3.3VA, 3.3VB, 5DA, 5DB, 5DC, 5.2, 7.7, 8VA, 8VB, 8VR and 8.6V.

As far as I can make out, the OTCuP (Onscreen display and Teletext Controller Microprocessor – IC7003, SAA5800/1) detects the 8.6V on pin 105 via resistors R3006 and R3009. At the junction of these two resistors is a double zener diode arrangement. Anyway, I checked all these components and could find nothing wrong – even when heating and freezing them.

In the end, I could only assume that the SSP (small signal panel) was faulty. However, I’m not prepared to try replacing the 120-pin surface-mounted processor, although I did change the EEPROM IC7008.

Anyway, I passed the diagnosis on to my colleague and unless he can find an alternative strategy, he will have to purchase an exchange module.

Same symptoms

The second set had the same symptoms – dead with the red LED flashing. Putting it into the Service Modes made the LED go full green and the sound come on but there was no picture and the CRT heaters were off.

As before, I started to swap boards and found this time that it was A1, the deflection module. And when I got a picture, I could see that I had an error 073 line deflection protection fault.

The line deflection board is large, with a lot that could go wrong. I noticed that in the green LED mode, you could hear the rush of the +32kV EHT charging the picture tube, so I was fooled into thinking there was EHT. However, the +13D rail was very low (8V) and then dropped right off.

I was further convinced there was EHT when I got a shock removing the final anode cap. It wasn’t until I put an EHT meter on it that I realised that the EHT came on initially and then very slowly decayed off.

Next, I removed plug O320 to the B board (O328), thus disabling the audio output ICs and the DC-protection circuit. I could not find any shorts on this board and tried swapping IC7484 (LM358N). The east-west circuitry also appeared to be OK.

Click for larger image

The circuit I had showed a convoluted diagram (A2) for the CRT filaments called FBCSO (Fixed Beam Current Switch Off), involving a regulated FET (Q7340) for power saving and a crowbar protection circuit. In reality, none of this was fitted to the board I was working on.

I was beginning to suspect the TDA8177 vertical output IC (IC7600), even though I could find no shorts on the 13V supply rail to it. Finally, I decided to unplug IC7484 which also operates the x-ray protection circuit. This then allowed the line deflection stage to try to operate uninhibited.

It immediately went into a pulsating mode, with all the line-derived voltages rising and decaying. The flyback transformer was groaning and the line output transistor (Q7421, BUX2520DX) was getting hot. However, before I could connect a CRO to Q7421’s collector, it decided to expire by going short circuit.

You can purchase repair kits for this set for each of the major boards but in this case a new BUX2520DX output transistor and a new flyback transformer fixed the problem. Make sure you get the correct part number for the flyback transformer, as some sets have the focus and screen controls built in and some have them on a separate part.

A couple of things make life difficult with this series of sets. Access to the solder sides of the motherboards is difficult and the non-Murphy-proof plugs are also a problem. The latter are not only unmarked but are very easy to plug into the wrong sockets. Some are not even colour-coded!

On one occasion, I managed to plug the vertical deflection coils (0325) into the Frame Rotation 0390 socket. Consequently, this blew R3447 and Q7442.

The DOA Panasonic

The 1995 Panasonic TC25V50A TV set that came in was also DOA. The set was an MX2A and is a well-known model which gives very little trouble.

I guess the most common fault is the failure of R833, a 22W resistor in the emitter of Q802 in the power supply.

Interestingly, if Q802 is short circuit, 90V will be applied to D820 (a 36V zener diode in the collector circuit of Q802), eventually destroying it and resistor R1111. Also, if an earth jumper is routed near IC1106 (Power-on-reset 5V IC to pin 7 of the microprocessor IC1102), Q802 will be turned on every time a reset pulse is sent. This causes stress in D820 and eventually a consequent chain reaction.

This particular set gave the impression of switching on fully when the power switch was depressed. You could hear the rush of static to the CRT and the sound coming on. However, no picture ever appeared as the line output stage was immediately being switched back off by the lack of line drive pulses from pin 19 of IC607 (AN5607NK).

The horizontal oscillator is switched off via pin 20 (x-ray protect), which monitors the 26V at TP-E4 and the current flow through R411 on its way to pin 6 of IC451 (LA7833S; TA8403K in other models). It also monitors the vertical output and you can override this protection by shorting pin 20 to ground.

The vertical ICs in some Panasonic sets can be dry-jointed due to their small solder pads, which if not fixed will cause the IC to fail. And when it does, a voltage appears on pin 20, causing the set to close down.

Replacing the vertical output IC and the electrolytic capacitors around it (as they dry out due to their proximity to the heatsinks) fixed the fault.

Dry joints

God bless dry joints! Without them, I would have been on the streets years ago, as they are still the most common cause of faults in TV sets regardless of manufacture.

Recently, I had a Sony KV-S34SN1 Kirara Basso (G1 chassis) come in. This is a very expensive and complicated 80cm TV set which weighs in at 81kg.

The owner’s complaint was that it wouldn’t turn on. Actually it would to some extent but unfortunately it was only the flashing red LED that was coming on, the set immediately going into the "protect mode".

Fortunately, most of the faults in this model are well known and mostly involve dry joints – in particular to the IC regulators spread out on almost every board. Eventually, I got the power supply board F out and found hairline fractures around D610, a common cathode double diode which feeds the +15V rail. I also reworked the solder on the board and all the other known problems, such as around IC208 and IC209 on the A1 board, IC2603 on the A board and IC503 (STV9379) on the D board, as well as all the VC board soldering.

Finally, I did the C570 (220mF) modification on the D board.

That fixed the set and I suspect this particular monolith will now do another 100,000km without requiring another service.


Finally, I have a postscript to last month’s story on the Sony KV-E29SN11 (BG1L). Afterwards, I had another go at the problem of the burning components in the 5V backup and standby circuit.

Basically, R615 (0.47W) and D608 supply 70V to the collector of Q601 (2SA1315) via R606 (18W). Q601’s emitter then supplies 10V to 5V IC regulator IC002 via the 22W resistor in the JW158 link position. These components were badly overheating and failing.

So what was the common part in all this? The answer is transistor Q602 (2SC3209), which drives Q601. Well, actually it doesn’t normally drive it as it is switched off and Q601 is biased to give 10V out.

However, when Q602 is switched on, Q601 gives a much higher output voltage at its emitter. And if Q602 was faulty, perhaps it would cause the full 70V to go through to the 5V IC regulator.

The only thing was that Q602 measured perfectly in circuit and even out of circuit was only slightly leaky. But when measured on a PEAK Atlas Component Analyser (DCA55), this NPN transistor was found to be behaving like a common cathode double diode!

Replacing it stopped the pyrotechnics immediately.

Items Covered This Month

  • Philips 32PW4523 TV set, L01.1A chassis

  • Philips 29PT9418 TV set, MG3.1A chassis

  • Panasonic TC25V50A TV set; MX2A chassis

  • Sony KV-S34SN1 Kirara Basso TV set (G1 chassis)

  • Sony KV-E29SN11 TV set (BG1L chassis) – postscript

  • Share this Article: 

    Privacy Policy  |  Advertise  |  Contact Us

    Copyright © 1996-2019 Silicon Chip Publications Pty Ltd All Rights Reserved