Panasonic TX-P42ST30E - TNPA5330 board defect - seven blinks and six blinks of death

The Panasonic Plasma 30 series (ST30, GT30, VT30) all suffer from a very silly bug: the ground screws of the SN (scan + buffer board) go loose and eventually produce sparks, which will kill a whole bunch of semi-conductors on the SN board and probably on the power supply board, too.

(In the meantime, I have analyzed this defect in more detail HERE)

The symptoms are 7 or 6 blinks of the standby LED. The service manual says:

A new SN board is prohibitively expensive. I did not find any under 230€ on eBay. In other models, Panasonic used to split the buffer board from the scan board. Here though, it is all in one and the buffers make it so expensive.

Luckily, there are repair kits on eBay tailored for this problem. I got one from GTV Watford in Great Britain. Including shipping it cost me 83€. The TV itself  was only 45€.

The following image shows the shot parts on the TNPA5330 SN board. The red marks are the parts covered by the repair kit, the yellow mark is a transistor, which is not included in the repair kit.

That's not all. Due to the diodes and transistors going short on the SN board, the power supply board blew a fuse and two FETs also died. See next image. For this, GTV offers another repair kit.

My brand new pre-heating plate and hot air gun finally got something to work on:

I covered the capacitors with capton tape and heated the board up to around 140°C. With the hot air gun set to 420°C the SMD-mounted large semi-conductors came off quickly and without a problem. For soldering, I used solder paste and the same temperature settings. It is fun to see the big thingies snap into place once the paste has reached its melting point.

Replacing the fuse and the FETs on the power supply:

After replacing the large diodes and transistors, I ran a test and still got 7 blinks. So I fitted the remaining two parts from the kit also. They are a switcher IC and a double transistor. This pair occurs in all four energy recovery or sustain circuits, respectively.

After I had used all the parts from the repair kit, the TV ran fine for a couple of minutes and then suddenly switched off again, this time with 6 blinks instead of 7 blinks! What was going on?

I checked all the scan and energy recovery circuits again and lo and behold, I found Q451 with a short between Gate and Collector. It is part of the RECOVERY-L section. Apparently, it had been already pre-damaged and finally broke during my first test run.

This particular transistor was not supplied by the repair kit. In foresight, I had ordered a working SS-Board on eBay, which is the little sister of the SN board and contains most of the same semi-conductors except for the expensive buffer chips. The SS boards are cheap. Mine was 30€. The SS board had two DG302 transistors on it and so I picked one from there.

In hindsight, it would have been cheaper to order two SS boards, which I think would cover even more than the necessary parts, instead of the repair kit.

This is what it looks like when one of the panel connectors is not seated properly. Thin horizontal lines all over the screen (the boxes are from the screen menu).

Those connectors are not so easy to seat. It takes some practice and I recommend using a magnifier glass to check that the little notches on the connectors are actually all the way in.

And now the Plasma does what it can do best: producing rich and beautiful colors:

It was quite some work this time:

To help with cooling, I glued small heatsinks on the large semiconductors and buffer chips. The heatsinks are not very powerful, but better than nothing at all. There is no space for large heatsinks, anyway.

A few words about the Panasonic service manual. The images for the boards and circuits are too small and not zoomable vector graphics and you can't see any of the small labels properly. Repairing a mainboard with this material is practically hopeless. The SN board is relatively simple and it did not matter. A mainboard is much more complex and you need all the fine details. Maybe there are copies around with better quality. I for sure did not find any.

The manuals from Sony and Philips which I have are much better. Personally, I would refrain from buying a defect Panasonic with a potential mainboard or logic board problem.

This TV is a keeper. It is missing DNLA features, but I will rather buy one of those Android quad-core devices with WiFi and HDMI, which have better performance and software on board than any of the Smart TVs out there.

Don't forget to improve the screws, folks! Replace the originals with new, longer screws with spring washers. Otherwise the same disaster might happen again.

Philips 37PFL6606K backlight failure

The Philips 37PFL6606K from ebay for 77€ started up fine, but then the screen went dark after roughly a minute. I checked the usual voltages and all looked ok, except for the LED backlight. After I had reseated the backlight connector at the power supply, the voltages had different values and the TV has worked without a glitch since then.

Sometimes it can be ridiculously simple. I still can't believe that this was everything, but the defect never showed again. So it has to be fixed :-)

Philips 42PFL7606H - won't switch on - no LED blinking - dead. TPS53126 converter chip dead - and then some...

The Philips 42PFL7606H I got from ebay for 113€ switched itself off, according to its seller, and did not boot up again.

This one was a lot more difficult than the others I had fixed before. The power supply did come up, 12V, 3.3V, Standby voltage, backlight voltage all measured ok, but after a few seconds, the machine powered down, tried one more time and then it gave up and went to protection state, resetting the "power good" line to the power supply. All voltages down, except for the stand by voltage.

So, the main board must have had a problem. This was no good news because there, diagnostics can become really hairy - and that is what happenend.

First, I spent way too much time checking the stand by voltage circuit on the power supply instead of reading the service manual properly. I paid the price for my impatience.

The first thing you want to do with a broken Philips TV is to switch it to Service Default Mode (SDM). It will produce a blink code with the stand by LED.

Switch the TV on, let it settle in protection state. Locate a large solder pad labeled SDM on the main board and connect it to ground for a couple of seconds. Then check the LED. My TV morsed me a 2-15 code. Even more bad news. The standby microprocessor could not boot.

So I started checking the voltages of the DC-DC converters on the main board and it didn't take long to find out that the only ok voltage was the 1.1V. The 1.8V was missing completely and the 3.3V and 5V where much too low.

The Philips service manuals are quite good and contain all the circuitry of the main board, which you might attempt to repair yourself. I realized that the TPS53126 dual DC/DC converter chip was not working, because it did not send the PWM signal to the output MOSFETs of the 1.8V line.

Next image: The two output chokes of the dual converter with the MOSFETs on top. The right one is the 1.8V output, which measured 0V.

Board flipped over exposes the converter chip:

I got a new chip from ebay, soldered it in and hooray, the TV booted and worked. For about 1 hour and then it was dead again. What the heck is going on?

The 1.8V was ok this time, 3.3 and 5V again as low as before.

To make a loooong story short, it took another 10 hours to identify a dead ceramic C (the one at the 6th pin from the upper right corner of the chip) and a dodgy SMD quad-resistor, which finally broke completely with a silent "click" after I douched the area with freezer spray. Until then this sucker fooled me all the time with unpredictable and intermittend reboots.

Here is the fix I applied. I soldered a normal resistor across the broken one and glued it to the board. Two convenient test points helped a lot:

The problem was as follows: the resistor is supposed to feed 12V to the enable line of the DC/DC converters. The double converter I replaced is happy with about 1.5V already, which I measured, but the 3.3V and 5V won't start. The resistor is supposed to have 22k and I measured 9k across it. So, I thought, it is probably ok. Difficult to tell when it is mounted. Then I reasoned that the double converter might pull the enable line low. But the chip was brand new and it happily produced both voltages! A bunch of transistors around it, which sense a proper 12V supply and manage some delayed power-up did not measure badly at all. Being desparate I bridged the resistor to beef up the enable line to a higher voltage and lo and behold, the tv worked!

Unsoldering the quad-resistor was no option. The possibility to break or lose it was close to 100% and I was not in the mood for finding a replacement in the vastness of the internet :-)

The TV is running fine now and I am enjoying the ambilight and 3D images.

Philips TVs seem to be quite repair friendly. The mainboard layouts in the PDF are vector graphics and zoomable. The orange circles show the corresponding areas:

The main board has plenty of test points, which are documented in the schematics, and they are exactly where you need them. However, Philips won't tell us what voltages to expect :-( You need to reason yourself with some help of datasheets.

The build quality of Philips is nowhere near a Sony. The tapes they use to dress the wiring comes loose after a few years already. It  all looks lumped together. The power supply in this model however contained only high-quality parts. It's not made by Philips.

I did an ESR check on all the electrolytics and found two dodgy 220µ 16V smoothing caps on the mainboard with .25 and .35 Ohms, respectively. That's just in the acceptable range, but same type, different values! Not good! Replaced both with decent 470µ, 0.05 Ohm ESR Panasonic FC.

Samsung LE37B650 - solarized picture - defect AS15F gamma chip

The 2010 Samsung LE37B650 I got from ebay for 87€ displayed what is called a solarized picture. Some colors are all wrong.

This is a common failure on the TCON board of this TV series. The voltage reference chip AS15F produces incorrect gamma reference for the panel. It needs to be replaced. This is difficult when you have no SMD hot air rework station. I didn't have one at the time, but I had Chip Quick, a low temperature solder, which stays liquid much longer than normal solder. Alas, the chip has an exposed pad on its belly and that is soldered to the board, too. It won't come off by just heating the pins.

As the chip was already dead, I carfully bent all the pins up to avoid ripping any leads off the board and applied some force on the chip while heating it with the soldering iron. Then, the pad broke off from the chip. No worries. Desoldering it from the board then was an easy task.

Today, I would use my preheating plate and 400°C hot air and it would come off no problem. I am learning :-)

Anyway, got a new chip from ebay for 8€. As I couldn't solder the bottom pad properly (solder paste and hot air required) I put a little heat-conducting block (salvaged from the Sony TCON) on top of the chip. It made contact to the metal cover and should dissipate the heat nicely.

And it is working again:

After this adventure, I decided to get a hot air gun, a preheater plate, and some SMD solder paste :-)

Sony KDL-46W5800 not starting - standby LED blinks 13 times - dead TCON board

I got a nice 2009 model Sony KDL-46W5800 from ebay for 101€.

The TV switched on, but after a few seconds it went down again with the LED blinking 13 times. According to the service manual, this indicates a dead TCON board, which does not communicate with the main processor anymore.

I measured the 12V supply line to the TCON and that read ok. Inspecting the board, I found an open SMD fuse. The supply line did not measure completely short, so I tried my luck and soldered a normal fuse on top of it for a quick check. (Don't mind the funky soldering of some parts, I've used this board for SMD soldering exercises...)

After switching on the TV, I heard a popping noise. On the bottom side of the board, I found a blown ceramic decoupling capacitor. That must have been the culprit all along. To find its original value, I looked up the data sheet of the DC-DC converter chip it was connected to and tried again with a replacement, but the board was still shorted and pulled the 12V supply line down. This cap may have killed some other voltage regulator upstream in the supply chain. On the internet I found one picture in a forum with the exact same problem. This might be a common failure.

I found a new TCON on ebay and that fixed the TV. As a precaution, I backed up this apparently undersized capacitor with a bigger one, taken from the old board. The "big" guy (still microscopic) has a lower ESR and should be able to handle the current pulses better.

My theory is that the DC-DC converter is pulsing heavily and the little ceramic (1µF) is the closest to the chip. So it is taking most of the pulses and that is getting too much after a certain runtime. My set had a whopping 12000 hours since 2009 (in Jan 2015).

The Sony is now in my living room and I love its image quality. Generally, Sony TVs seem to be very solidly built and I found only highest quality capacitors on the boards. No junk anywhere.

The TCON's number is 46NN_MB4C4LV0.7. The metal cover has the number LJ63-06901A