LG 60PZ950T Z-sustain repair

[tom66]

So... This was an interesting one.

Bought this TV for a few quid from a member on here who gave up on it. Essentially, poor picture quality, looked like a Z-sustain fault. Originally had problem with colour distortion due to damaged Z-sus ERC inductors. Previous owner replaced Z-sustain but failed to install all screws, this appears to have caused the new board to fail taking the SUS IGBTs out with it.

So, originally, removed all IGBTs and tested using o'scope for gate drive waveforms: definitely good on low side, hard to tell on high side...decided to give it a shot anyway.

Replaced six IGBTs in SUS path and powered up with bulb in place of fuse on board. Bulb glowed dim then full bright, and same fault presented itself: dead short Z-sus. Found one lower IGBT in sus path to be shorted and both gate drive resistors measure 50kohm or so, should be around 10 ohm each. Luckily, the bulb seemed to have saved the other five IGBTs which all tested good. Replaced that IGBT and the resistors. Powered it up again...same IGBT went bad! Realised I had left some screw out on the board...d'oh!! Replaced that screw and the IGBT a third time...success! Set gave a decent image with bulb in series (some maldischarge); taking bulb out gave a perfect picture.

Lessons learned: check all screws are installed, and use a bulb tester, saved me twice from replacing a whole load of IGBTs!

Biggest TV so far. Decently made! Ultimately this only occurred because all the screws weren't installed. On these boards it is absolutely vital to install all screws or failure WILL occur!

[Nick's Tvs]

Re: LG 60PZ950T Z-sustain repair

Just fixed 2 of these this week alone, its a fun repair job and the failure is always the exact same. At least so far that I have experienced.

[tom66]

Re: LG 60PZ950T Z-sustain repair

What was the fault on your set, cold solder on the ER inductors or something else?

[Nick's Tvs]

Re: LG 60PZ950T Z-sustain repair

Every time so far the inductors had bad solder joints, had shorted out fuse, shorted out IGBT's so I ended up just going with the Shopjimmy repair kit on both

http://www.shopjimmy.com/lg-ebr73561...repair-kit.htm

I ended up replacing all parts from the kit including the 4 IC's and all the surface mount resistors between the IGBT's. I wasn't sure how to test the IC's , so I figured might as well. Did you end up replacing those as well or just the IGBT'S fuse and resistors? Was your fuse even blown on your board? I don't think you mentioned it.

[tom66]

Re: LG 60PZ950T Z-sustain repair

Fuse was blown on mine but the inductors were ok--the fault was caused by loose screws. I'm fairly sure that if the inductor solder joints fail the IGBTs won't be too far behind.

I used my oscilloscope to test the low side gate driver, I wasn't able to test the high side apart from ohming out the output pin to the supply and ground. I'm not able to test that because without an IGBT in place the drive waveform doesn't pump the gate driver charge pump. I decided since I had a bulb in series just to go ahead and it seemed fine, the gate driver on the low side actually survived three shorted IGBTs in a row and was still good, which was quite a surprise.

[Nick's Tvs]

Re: LG 60PZ950T Z-sustain repair

Just had a customer send in his 2 boards for repair, this time first original board only 2 IGBT's out, and the joints on the inductors were god awful burnt.
Board number 2 was the replacement he bought, and probably forgot to screw in a screw since all inductors are fine. Same thing fuse is fine, but I don't know how many IGBT's failed, all I know is that some did since they measure shorted.
I guess this leads to a few questions,
Have you been able to figure out which failed without taking them out of circuit?
Second questions is so far I have just been replacing all 6 every time as a precaution, but do they actually get any weaker after one fails? Would it be ethical to only replace the ones that failed?

[ljintexas]

Re: LG 60PZ950T Z-sustain repair

I will inspect and test board before I send it to be repaired..I do not have the gear to do any extensive SMD component replacement..if this trend continues, I may consider a hot air station..I just don't do enough of it to justify the cost right now.
I will report my findings..
Thanks

LJ

[ljintexas]

Re: LG 60PZ950T Z-sustain repair

I am taking for granted that the sound no video condition with all boards iinstalled and connected indicates a failed Z-sus.

[Nick's Tvs]

Re: LG 60PZ950T Z-sustain repair

Got it figured out. Both boards ended up only having 2 bad IGBT's both boards had different ones affected however and it seems random because a third board i worked on today had a different pattern of which were bad and also blown fuse with blown resistors. It seems like if the fuse blows then bad resistors, if no blow then good resistors only bad IGBT's.
Lastly, it makes so much sense and is so easy to figure out which are good and which are bad while still in circuit. Since both rows of three are in parallel, all will test shorted in the dmm beep mode. All emitter pins are shorted to one another, but the collector pins are not. they have the resistors. So pretty much you measure all three and which ever one measures 26 ohms less of resistance between emitter and collector pin will be your bad IGBT. However the boards with blown fuses and thus bad resistors will also show shorted in the same way, so you have to replace the shorted resistors first before you can know which IGBT went out.
I know this is not a good explanation, but I never have to explain this stuff to anyone so I don't really get practice doing so ha.

[lookimback]

Re: LG 60PZ950T Z-sustain repair

Thanks Tom66! I would have never fixed this one if you hadn't made this thread. Mines a 60PV250. I've been working on this one off and on for a few months. I replaced the a few bad RF2001 diodes and FGPF4633 IGBTs on the z and on 2 different Ys, as well as the FAN73711a and those SOT-25 transistors marked 5k(replaced with CPH5506). Tested and saw the OSD, and pop. Did it again, this time I scoped out the 73711s and replaced as needed, and pop. And again, pop. If the definition of insanity is doing the same thing over and over expecting different results, then I am apparently insane. So, I have been waiting for 3 days now for more of the CHP5506s from Digikey and they came in today. Replaced the 2 on the Z, this time I put ALL the screws in, and nothing. But nothing is better than pop. Actually, the screen went a fuzzy dark red for a second. Was still pretty sure it was the Z. Soldered wires onto the outputs of all four 73711s and looked at the waveforms on the scope. The two high side ones were all crazy, but maybe my scope just isn't fast enough. One low side was clearly outputting a nice waveform, and the other low side was outputting nothing. Replaced all 3 because I had them. And it's alive! In fact, not only is it alive, but the second Y SUS is also working!! I have a second set of Y buffers I'm about to test in a minute. I never would have guessed that only putting in one screw when testing would have been the reason the IGBTs and diodes kept failing. I can also say this is probably the first TV I've successfully troubleshot with the scope.

[freakaftr8]

Re: LG 60PZ950T Z-sustain repair

Sticky on LG YSUS/ZSUS floating grounds???

Throw Panasonic up there too!!

[lookimback]

Re: LG 60PZ950T Z-sustain repair

Sticky on LG YSUS/ZSUS floating grounds???

Throw Panasonic up there too!!

Good idea. I did the same thing on a Samsung PN43D440A5D. Got another one open and dug out the instructions from the repair kit I'd tried on the other one a couple years ago(kept it in case I wanted to order the components) and it says tighten all screws. Guess I should have read the instructions.

[lookimback]

Re: LG 60PZ950T Z-sustain repair

Anyone got a cross-reference for panel driver ICs? I can never even find datasheets for them. I need 2 R2A20296FTs to fix the spare buffers. TV will be gone tomorrow so it's either fix them tonight with something from a donor board, or toss them.

[tom66]

Re: LG 60PZ950T Z-sustain repair

There's no floating ground on a Z-sustain board.

There's also no datasheets for the scan driver ICs but I can describe some things about them based on reverse engineering.

[lookimback]

Re: LG 60PZ950T Z-sustain repair

There's no floating ground on a Z-sustain board.

There's also no datasheets for the scan driver ICs but I can describe some things about them based on reverse engineering.

What causes the IGBTs to pop when all the screws aren't in then?

[lookimback]

Re: LG 60PZ950T Z-sustain repair

Well, that was short lived. Now it shuts off after a few seconds. It will turn back on with picture, but same thing again. I'm about to take the back off again and check the VA and VS voltages. I know I should have already. I haven't fixed a plasma in a long while. I'm rusty. With no input it was staying on.

[tom66]

Re: LG 60PZ950T Z-sustain repair

What causes the IGBTs to pop when all the screws aren't in then?

No ground, electrons get lost trying to find ground, causing high voltages all around, alternative paths are found through other sets of IGBTs ... bang.

Or, the gate drive signals go skew-iffy and the IGBTs get half turned on as a gate driver IC doesn't have the right supply voltage ... bang.

Or, the inductance of the remaining ground is too high, causing very high source transients, due to the leakage inductance, which damages the gate of the IGBT ... fizzle.

[lookimback]

Re: LG 60PZ950T Z-sustain repair

I see. So what is a floating ground? TVs working again btw. The one Y sus had a crack in it. I'd bridged it, but it seems not to like it when it warms up. I guess that one will end up back in my scrap bin where I found it.

[tom66]

Re: LG 60PZ950T Z-sustain repair

I see. So what is a floating ground? TVs working again btw. The one Y sus had a crack in it. I'd bridged it, but it seems not to like it when it warms up. I guess that one will end up back in my scrap bin where I found it.

I've explained this before but the floating ground is only used on the Y-sustain. It is the output voltage generated by the Y-sustain board. It's what all the IGBTs on the board are working to make.

So if you ever see an oscilloscope waveform like this:
http://4.bp.blogspot.com/-8RDK8j1TqP...600/fig-11.jpg

then that is the floating ground output.

Hmm, you say, that doesn't look very "ground" to me...!

That is because it is not a true ground but rather the ground reference for the scan driver ICs (buffer board ICs). These scan drivers select between the VFG (floating ground) and the VFG + VH voltage (which is floating ground + 150V). But as far as these ICs know, it's ground. They are fully isolated digitally and electrically from the normal ground. Though a path to ground from the sustain board output will cause a failure of the sustain board or buffer boards. *THE ONLY PURPOSE* of the scan driver ICs is to select a line to scan on the panel. They apply the 150V VH voltage to select a line and the address / data drivers which run along the bottom of the TV screen program the line with the appropriate pixel data in the given subfield. In this sense, the scan driver ICs are similar to the LCD gate driver ICs, though failure symptoms do vary.

The VH is often called Vsc, or Vscan... And the floating ground is the same ground reference as used for -Vy or -Vad voltages... Though Samsung calls -Vy the -Vsc voltage just to be confusing... But in general the negative ramp voltage/floor voltage -Vy is around -180V to -220V. And the positive line scan voltage Vsc is +140V to +160V.

Plasma TVs before 2005 used some different designs that were more expensive to implement but from roughly 2006 onwards all manufacturers standardised on floating ground designs. Hitachi panels are the only exception because they use TERES and ALiS technology which is...strange... to say the least.

[lookimback]

Re: LG 60PZ950T Z-sustain repair

Thank you for the very detailed explanation.