Sharp LC-46LE824E - switches itself off when warm - Delta DPS-141CP1A - defect optocoupler in power supply

A friend brought me his Sharp TV with the symptom that it switches itself off when it gets warm. After cooling down it would restart and the problem would repeat itself.

I suspected a cold solder joint, but it was much more detective work. Thermal defects are the hardest.

A look inside:

Nice job, Sharp. Double speakers for each channel and one sub-woofer. Everything easily accessible.

We started testing and after pointing a hot air gun at the output part of the power supply, the device switched off. It recovered after a while and a more precise hot air attack located the area to the down right as most sensitive.

I removed the PS and checked for cold solder joints. None. Everything in good condition. The area around the three semiconductors under the heat sink was very sensitive to heating. The diode in the middle seemed to be the culprit, at least that's what I thought. So I removed the three, unscrewed them from the heat sink, soldered them back in without heat sink and then tested each in isolation. I replaced the diode with no success. The whole area was sensitive, with a tendency to the top towards the optocouplers. I spent a lot of time heating and shooting icing spray without any luck finding the precise spot.

In this Forum article I read about problems with optocouplers, so I narrowed my tests down. And lo and behold, the bottom coupler's sender diode freaked out when heated up. Part number was PC7703. It showed 1.9V break-through voltage at room temperature and quickly lost it when heated up slightly. A cross check with its neighbors settled the case.

Even a slight hot air wind made it fail, that's why I was fooled all the time when I pointed the hot air at other parts in its vicinity. 

I had plenty of spares from scrap boards and soldered in a new one. TV working fine again!

I never had a failed optocoupler. This is very rare. So what was wrong here? Look closely. The coupler sits in the main heat stream of the three semiconductors. And those get very hot! It literally gets grilled. I think this is the reason for its failure.

I glued a little piece of plastic onto the board to guide the heat from below around the couplers. This will relax the temperature situation.

Hallo Frau Johansson!

What I have learned from this case: A part with a thermal defect will reveal it measurably. If a suspected part behaves flawlessly, leave it and move on, even though it is hard to drop a promising fault hypothesis. I measured a double diode isolated from the circuit (forward voltage, reverse voltage) and it did not show any defect. Still I replaced it "just to be sure". It was a waste of time. The optocoupler however immediately revealed its problem very clearly.

Philips 42PFL9803 - Panel fault case study - vertical colored lines - black and white blocks in image

There is a lack of interesting TVs on eBay and eBay classifieds these days. It seems as if people only break their stupid thin-frame panels with kid toys or tip over the thing. Frustrating.

I found this Philips, a model I successfully repaired once, and gave it a shot. Maybe I could get lucky and it was the same fault. The seller said it would not produce an image at all. At pick-up time he told me that he got the TV from someone else who said it produced lines in the image. Damn it! I would neither buy stuff from a re-seller who never used the device, nor TVs with lines in the image - if I knew it.

Okay. Turned the device on. The activity indicator at the bottom flashed, which told me that booting was in progress. No image appeared and then the relays of the power supply clicked and the set went off. No blink code was emitted from the stand-by LED. That's not good. After checking the FETs on the main board voltage regulator section, which supply the main processor (those fail often in this model), I shorted the SDM testpoint on the main board to ground to activate the service default mode. In this mode, some safety checks are overridden. And I got an image:

Let's analyse this for a moment:
- The half-half characteristic would indicate a TCON failure, Same goes for the black block to the left. The right side shows no errors. It is just blank.
- The striped block however rather looks like source driver failure in the panel.

Look at this detail. The image had a kind of frame around it. The rounded corners looked funny:

As this was all around the image in perfect symmetry, I did not book that on the panel faults account.

So, the diagnose wasn't clear at all. The next experiment in such cases is to disconnect the connectors from the TCON to the panel one by one. Oh crap, the TCON was buried under the sub chassis on which all the other stuff was mounted. Thank you, Philips! :-(

Removing the chassis wasn't as difficult as it seemed:

The panel unit without the sub chassis. The TCON cover exposed:

Now I disconnected the left side:

And got this:

Hmmm, this really looked like a defect source driver! A panel cannot produce any patterns on the side where the connector isn't attached. Alas, upon close inspection of the TCON I noticed that some traces go across from the right control side of the TCON the left panel source drivers. And why was the right side still blank? Argh! These things are still a mystery to me.

Now left side back in, right side disconnected:

This is all so weird. The right side is disconnected and the image to the left changes. There were clearly some cross-effects, which I do not understand.

The lines looked interesting:

I did not get one single step further! It could still be the TCON, with a tendency to panel fault. On eBay the seller @loeterotto, who specializes in TCON boards, had the matching board and so I bought it to find certainty. Alas, the new board produced the same image. I returned it.

Right, so this TV is junk. At least the main board seemed to work fine and it is very rare. I have a chance for a good price.

While dismantling the panel, I finally found the reason for the fault. Two broken flat cables, where the source drivers are embedded into. So my first instinct was right.

LCD panels. I still don't understand how to reliably diagnose those suckers!

Here is an picture of the direct LED backlight. This was so much better (and more expensive) than the cheap crap that Samsung and LG produce these days. A vast array of less powerful LEDs is much more reliable then a small number of high power LEDs, which die after three years due to overheating.

Samsung LE40B541 - volume goes up and down erratically - defect volume-up switch

I heard about Samsung sets changing the audio volume uncontrollably. I recently got a chance to look at one myself. I switched on the device and after some minutes it started pulling the volume up and down unpredictably. During normal phases, the buttons on the remote and also on the TV would work just fine. Thus, my first instinct was not to check for broken switches.

Nevertheless, it was pretty likely that either the switch board or the control board, which also contains the IR sensor, was to blame.

So I went ahead and unplugged the switch board first. The TV was stable. That was interesting. How could a switch break without getting used at all? I still did not believe it. The IR board contained nothing but the sensor and the connector to the switches. Others have a controller on it. Not here, the controller was integrated on the main board. I swapped a 47µ cap on the board, just to be sure. No effect.

Ok, so the culprit must be the switch board. I removed it an put it on the table. While pressing down on the connector solder joints with my meter probes, the TV would not show any symptoms. Aha. I took the board in my hands and twisted it slightly. Clockwise twisting reliably produced the problem.

The board has two output lines. Each button is connected via ground and a pull down resistor to one of the lines. This produces an individual voltage for each button. The line with the volume buttons was erratic with an unstable voltage. I unsoldered a resistor, a cap and a diode. Still the same. So I finally measured the volume-up button and what can I say, it changed its resistance wildly. Not quite down to zero, just in-between so that the controller sometimes decided that the other button was pressed. 

What the...? The button worked totally fine, yet it produced a leakage current. I have never seen anything like this.

Button removed and everything was back to normal. I had no replacement at hand, but the owner did not mind to leave it out, as it had never been used anyway(!)

Some wire-redressing with fresh tape and the TV was good to go.

Let's take the switch apart and see what we find:

The body shows some shiny residue, which might be conductive. Not easy to tell, but that is the only irregularity I could find.

My theory is that some material had been creeping into the switch during the re-flow phase of production. On top of that just a slight amount of oxide during 7 years and that could be just enough to connect the outer contact with the central pin. Who knows?

So, I guess, these switches are actually fixable with a proper cleanup.

Philips 32PFL7695 - working but no standby light and non-functional remote and buttons - bad solder joint

The seller of this TV said that it would start and work properly when switched on via the main switch, but neither the remote nor the buttons at the TV had any effect. Also, the standby light wasn't illuminated and actions on the remote would not make it flicker.

The Philips support told her that a new main board was required. That was basically accurate, as we will see.

Now this did sound like a clear case of broken power supply of the control board, which links the remote sensor and the buttons.

A look inside. Lots of wiring. Good speakers!

This is the control board with the remote sensor. The left cable is coming from the buttons. The right cable goes to the main board.

Here we have the socket on the main board. Can you see how strangely imprecise it is mounted?

I shifted it a little bit and switched the TV on. It was working! The stand by light was red, the remote control did its job.

Aha! A look on the B-side of the main board revealed the problem. A production error with a terribly bad solder joint, which goes to ground. It took five years to finally break. And look at all that weird residue around it!

I reflowed everything, cleaned it up, tightened a number of totally loose screws of the power supply and the pretty little sucker was good to go :-)

Philips 42PFE0001D/H - distorted colors - defect AS15F chip on TCON board

Another AS15 repair :-) I was interested in this Philips TV, because it once was the absolute High-End. It has an external box with the power supply and main board. Two cables go the monitor, which contains the logic board, TCON and inverter.

The seller liked it so much that he offered to buy it back if I managed to fix it. Fair deal.

Switched it on and got this:

Really bad color distortion. The Menu titles even disappeared sometimes into the dark. A typical AS15 gamma chip fault.

Not much in the monitor:

Funny idea: a heat conducting block for a back cover made of plastic. Well, it worked for 7 years, so it wasn't that stupid, I suppose.

And here it is. The blue TCON board of doom made by AUO. The chip is on the upper right.

Time for gamma measurements. In brackets are the values of the new chip. Three test points were far off.

vgama1: 15V (15.5V)
vgama2: 14.7V (15.4V)
vgama3: 10.5V (10.8V)
vgama4: 10V (10V)
vgama5: 9.7V (9.8V)
vgama6: 7.7V (7.7V)
vgama7: 7.7V (7.7V)
vgama8: 7V (6.9V)
vgama9: 7V (6.9V)
vgama10: 9.2V (5V)
vgama11: 5.8V (4.7V)
vgama12: 10.7V(4.0V)
vgama13: 0.43V (0.36V)
vgama14: 0.38V (0.33V)

As always the heat conducting block stuck to the chip and broke off. I think we have a pattern here. It looks as if the material had disintegrated. Maybe that sealed the death by heat of those chips.

To solder the chip I used my new horseshoe solder tip and it worked fantastic. Much better than the spade type I had used before.

That's much better!

Perfect grades of black:

Being curious I peeked into the so-called hub, the control unit. I couldn't figure out how to get to the power supply below the main board. The chassis would not come out and I decided not to mess around with stuff that works. I just blew out the dust from the cooling fan on the left.

Hallo Frau Johansson!

Philips 42PFL8654 - no standby voltage - two shorted SB260 diodes - Power supply DPS-298 ticking noise

This Philips 8000 series of the generation 4 came in dead. It wasn't starting, no sign of life. This pointed to a stand-by power supply problem.

With the multi-meter in continuity mode I quickly located two shorted diodes on the secondary side of the stand-by supply. The diodes were Vishay SB260. I probed around the supply lines for shorts and found nothing. This is strange. Why did both of the diodes die? Maybe one blew first and the other would not hold on for much longer due to overload. I have never seen two diodes in parallel in any power supply so far.

I didn't have those diodes or similar on stock, but some BYV29 ultra fast diodes were available. One of these can cope with more current than two of the little ones, so I gave it a shot.

Plugged the TV in and it started just fine. The image came up all right. That was too easy, I thought. I left the TV running for about two minutes and suddenly, an intermittent ticking noise like from a Geiger counter came from the high voltage transformer, which feeds the backlight. A cold solder joint? I reflowed all the critical joints, even though none of them looked suspicious. As expected that didn't fix it.

So, those were the possibilities:

- The transformer had an intermittent short. I would expect it to go up in flames if that was the case.

- Some other weird problem with the transformer.

- At least one those blue ceramic 6kV capacitors was about to die:

I unsoldered all four caps and measured them with the capacitance meter. All had their specified values. Then I pulled out my funky Chinese DUOYI DY294 transistor and cap tester to check for break through voltage. This little device can produce high voltage and test whether a part withstands it or not, using 1mA leakage current.

All caps measured similar with about 1450V. I didn't know what to think about that. They are rated 6000V. It was outside the spec of the tester (max 1000V). 

As all caps were identical in that respect, I concluded that a fault was not measurable with my equipment. The output of the transformer produces much higher voltages, so it was still possible that at least one cap was faulty. Or I was on the wrong path altogether and the transformer was bad.

I checked eBay with the power supply board id DPS-298 and a number of repair kits came up! One contained the two diodes and the four caps. I ordered one, even though I was not convinced that the caps were the culprits. For 5€ there is not much harm done.

Compared to various images on the net and on eBay

like here:

or here:

my caps looked like new. I wonder why they burn up like this. I better not measured their temperature, for it is high voltage which either hurts me or destroys my meter.

This thread discusses the same problem my TV had:

http://www.badcaps.net/forum/showthread.php?t=35814

After a night's sleep I doubted that the four caps were responsible for the noise. I started measuring a little more. The inverter drives two windings in parallel. The primary side looked like it had a center tap, but the two coils are arranged symmetrically. The circuit is a mystery to me. The outer transformer pins are driven against ground and V+ by the FETs. The inner pins are coupled via 0.47µF to V+ and 0.47µF against ground. It must be some kind of resonance circuit.

I first located the noise in the transformer (because my mind said that it was logical) until I switched the light off to see if there were visible sparks. Nope. Nothing to see. Being blind, my ears overruled my mind and found the real noise source: The 0.47µF cap, which goes to ground from the test point.

The cap could not be the source of the problem, because it goes to ground and the measured pulse increased the voltage. However, the sudden voltage spike caused mechanical noise in the cap.
The two FETs are in a half-bridge configuration and they were my new suspects, because they control the voltage to the primary winding. Maybe one had intermittent failures?

After some searching, I found the circuit for another Delta power supply DPS-283 at Elektrotanya, which was close enough to the DPS-298. This design is missing the four caps on the secondary side, but the primary side seems identical.

For some reason, this resonance game had glitches. And this is what the clicking noise looks like on the scope. From flat 210V a spike of 230V.

This website offers a specific repair for the power supply and a number of spare parts for it:

http://www.electronic-doc.de/28-reparatur

The price for the FETs was too high for my taste and I found a cheaper source:

http://www.tv-ersatzteile.de/fqpf10n60ct.html

and ordered two pieces.

Screw that. It must be the transformer. Swapped all four caps and the high-side FET, which controls the positive voltage to the inverter transformer. Nope. Still noise after a some warm-up and bending the board changed the pattern. I shot freezer spray all over the inverter circuit and nothing reacted to it.

This power supply is an atrocious design. I know at least of four different regular failures by now (caps, diodes, clicking noise transformer, inverter control chip).

I gave up on this one.