Author Topic: Help Finding Suitable Replacement for MOSFET  (Read 2226 times)

0 Members and 1 Guest are viewing this topic.

Offline canadaboy25Topic starter

  • Regular Contributor
  • *
  • Posts: 152
  • Country: ca
Help Finding Suitable Replacement for MOSFET
« on: June 18, 2021, 01:27:31 am »
 Hey everyone, I've been struggling with a repair for a while so I figured it was time to reach out to some more knowledgeable people.

I have an Akai 50" plasma TV which uses LG internals that I'm trying to repair.  I narrowed down the problem to a faulty IPM module on the Z-Sus board.  I don't have a picture handy but it is similar to the attached image that I found online.  These modules are quite atrociously priced online and most of them look like they have been salvaged from other TVs.  My IPM was full of potting compound so I dug it all out and it turns out there isn't much in there.  I found that the two main MOSFETs were shorted, along with some blown SMD resistors and diodes.  I managed to find the exact replacements for all of the components except the MOSFETs as they are bare die FETs covered in a glob of hard resin.

I scraped the resin covering the old FETs off and desoldered the rest.  I then replaced them with some Panasonic 5N2307 FETs that were salvaged from a dead Panasonic plasma.  There are no datasheets for these FETs but as far as I can tell they are 230V 25A.  I replaced the blown fuse on the supply rail and fired up the TV.  I got a good image and figured I had won.  However, ~1 min later, the TV suddenly shut off.  The fuse was blown again and both FETs were shorted.  No other components were damaged.  I figured I might have got unlucky and salvaged a partially damaged FET so I grabbed two more and tried again.  Same result.  TV worked for ~1 min then shorted both FETs.

So I figured these FETs were not up to the task so I tried some FDB12N50TM FETs that I had on hand.  These are 500V 11.5A FETs.  After Installing them, I powered up the TV and saw that the screen had a lot of ghosting and bad colors.  I let it run for a good while and the FETs never blew.  However, there must be something about them that is not suitable for the application.  I am thinking it has to do with the switching rise and fall times but I have no way to confirm this.

I found this video highlighting the same problem that I was having and the guy used STD12N60DM6 FETs.  These are 600V 10A FETs.  As a last resort, I ordered some of these and installed them.  The TV powered up and the image was perfect, no more ghosting.  However, after a few seconds one of the FETs forcefully detonated with the volume of a gunshot.

I have no idea what FET to use and trying random ones with seemingly appropriate ratings has not worked.  The supply voltage Vs is about 196V so I'm not sure why this would be shorting the 250V Panasonic FETs.  The FDB12N50TM FETs were not damaged, but something about them was causing a poor image.  The STD12N60DM6 has a higher voltage rating and only a slightly lower current rating than the FDB12N50TM, but they exploded forcefully when used.

Could someone please explain what specification I am neglecting to consider that would explain these behaviors?  And even better, could someone suggest a suitable replacement?

canadaboy25

- Sometimes the light at the end of a tunnel is an on-coming train
 

Offline bob91343

  • Super Contributor
  • ***
  • Posts: 2675
  • Country: us
Re: Help Finding Suitable Replacement for MOSFET
« Reply #1 on: June 18, 2021, 03:06:02 am »
Perhaps a blown flyback diode or snubber is allowing excessive spikes to damage the transistors.
 

Offline canadaboy25Topic starter

  • Regular Contributor
  • *
  • Posts: 152
  • Country: ca
Re: Help Finding Suitable Replacement for MOSFET
« Reply #2 on: June 18, 2021, 03:10:27 am »
Perhaps a blown flyback diode or snubber is allowing excessive spikes to damage the transistors.

I thought that too.  However, I have checked and double checked every component on the IPM and everything checks out.  There really aren't many parts on this board.  6 diodes, two dual transistor chips, and one driver chip.  There are also a few passives.  All the diodes and transistors test good and the driver chip must be working as I would not get an image otherwise.  The only inputs to the IPM are a voltage rail and two control signals.  I have checked all inputs with the scope and cannot find any evidence of spikes.  Also, spikes on the two control signals would kill the small driver chip, not the FETs.  So if there were spikes, they would have to be on the two supply rails and I see none when checking with the scope.
« Last Edit: July 29, 2021, 07:05:30 am by canadaboy25 »
canadaboy25

- Sometimes the light at the end of a tunnel is an on-coming train
 

Offline bob91343

  • Super Contributor
  • ***
  • Posts: 2675
  • Country: us
Re: Help Finding Suitable Replacement for MOSFET
« Reply #3 on: June 18, 2021, 04:23:07 am »
The only thing that kills the transistors is heat.  Either steady rise in temperature or localized hot spots.  That's why I thought a spike might be doing the damage.  Perhaps the transistors aren't properly connected thermally with some heat sink.  This latter possibility is suggested by the fact that the unit will work for a while and then pop, probably overheating.
 

Offline canadaboy25Topic starter

  • Regular Contributor
  • *
  • Posts: 152
  • Country: ca
Re: Help Finding Suitable Replacement for MOSFET
« Reply #4 on: June 18, 2021, 05:09:21 am »
The only thing that kills the transistors is heat.  Either steady rise in temperature or localized hot spots.  That's why I thought a spike might be doing the damage.  Perhaps the transistors aren't properly connected thermally with some heat sink.  This latter possibility is suggested by the fact that the unit will work for a while and then pop, probably overheating.

The transistors are soldered to the board which is bonded to a solid sheet of aluminum on the back side.  The back side is then attached to a large heatsink using thermal compound.  I am confident that I have a good thermal connection between the transistor and the board.  I used a hot air station on the back side of the aluminum sheet which melted the pool of solder on the pad for the transistor.  Lots of flux was used and it was clear that the solder made a good connection to the entire transistor.  The heatsink was then attached using new thermal paste.

Perhaps the Panasonic transistors are heating up more than the other ones?  Could this be due to the "on" resistance of the transistor being higher and therefore dissipating more energy?  It is frustrating that the FDB12N50TM transistors worked indefinitely and did not blow, but instead gave a poor image.  I am not sure what parameter of the transistor affects this.  I highly doubt that effect is a result of heat.

As for the STD12N60DM6, I don't think the explosion of the transistors could have been caused by heat as the TV hadn't even got past its startup splash screen yet.  Of course it is possible that it got extremely hot very quickly, but I don't see why this would happen when it is rated for 10A.  It couldn't have even heated up the whole heatsink in that time.

I will try and take another look at the datasheets for the two transistors and see what the differences between them are.
canadaboy25

- Sometimes the light at the end of a tunnel is an on-coming train
 

Offline canadaboy25Topic starter

  • Regular Contributor
  • *
  • Posts: 152
  • Country: ca
Re: Help Finding Suitable Replacement for MOSFET
« Reply #5 on: June 18, 2021, 05:16:18 am »
Here is an image of what the output waveform from these transistors should look like.  I checked with the scope and could not see anything obviously wrong with the waveform when the FDB12N50TM transistors were installed, but did not have a precise waveform to reference to.

Perhaps I could install another set of the Panasonic transistors long enough to capture a known good waveform and then reinstall the FDB12N50TM transistors and see where the discrepancies are.  That is a lot of soldering though and the Z-sus board is staring to look pretty rough where the IPM is soldered to.  All the pins are connected to huge power planes so it takes a lot of heat to get the solder melted.
canadaboy25

- Sometimes the light at the end of a tunnel is an on-coming train
 

Offline canadaboy25Topic starter

  • Regular Contributor
  • *
  • Posts: 152
  • Country: ca
Re: Help Finding Suitable Replacement for MOSFET
« Reply #6 on: June 19, 2021, 05:33:33 pm »
Anyone have any other ideas?
canadaboy25

- Sometimes the light at the end of a tunnel is an on-coming train
 

Offline RayRay

  • Frequent Contributor
  • **
  • Posts: 297
Re: Help Finding Suitable Replacement for MOSFET
« Reply #7 on: June 19, 2021, 09:00:32 pm »
Anyone have any other ideas?
When you've tested the diodes, did you do it both ways (reverse bias too)? Because if you've only done it on forward bias, the results could be misleading! Also, your board should have a part number on it somewhere (either directly printed on it, or on a sticker) and odds are you could find a matching replacement on ebay or aliexpress, so you have that option handy. Apart from that, if you could find a service manual for your particular TV model (or the LG model it's based on) that could help you troubleshoot it better (and images of the board, from both sides, could be of help too!)
« Last Edit: June 19, 2021, 09:02:17 pm by RayRay »
 

Offline Mario87

  • Regular Contributor
  • *
  • Posts: 247
  • Country: gb
Re: Help Finding Suitable Replacement for MOSFET
« Reply #8 on: June 23, 2021, 11:53:28 am »
The transistors are soldered to the board which is bonded to a solid sheet of aluminum on the back side.  The back side is then attached to a large heatsink using thermal compound.  I am confident that I have a good thermal connection between the transistor and the board.  I used a hot air station on the back side of the aluminum sheet which melted the pool of solder on the pad for the transistor.  Lots of flux was used and it was clear that the solder made a good connection to the entire transistor.  The heatsink was then attached using new thermal paste.

Do you have high res images of the repair? As if its anything like that image you took from online then I would certainly not say its got good connections / workmanship.

We need to see what exactly you are working with. Generic images online can only show so much.

Also, what are the original MOSFETs which were installed?

You can't just keep pulling random parts from random places. You need to determine the specs of the original and buy high quality replacements with similar specs. The voltages you specify will be for Vds, but what about the Vgs threshold? Is it a P-type or N-type MOSFET?

This is all information you need to have before replacing parts randomly.

If you have issues after replacing the MOSFET with an appropriately specced replacement (specced in all terms, not just Vds), then start looking at other parts of the circuit. Until then you're still at the first hurdle.
« Last Edit: June 23, 2021, 11:55:18 am by Mario87 »
 

Offline canadaboy25Topic starter

  • Regular Contributor
  • *
  • Posts: 152
  • Country: ca
Re: Help Finding Suitable Replacement for MOSFET
« Reply #9 on: June 23, 2021, 06:33:47 pm »
When you've tested the diodes, did you do it both ways (reverse bias too)? Because if you've only done it on forward bias, the results could be misleading! Also, your board should have a part number on it somewhere (either directly printed on it, or on a sticker) and odds are you could find a matching replacement on ebay or aliexpress, so you have that option handy. Apart from that, if you could find a service manual for your particular TV model (or the LG model it's based on) that could help you troubleshoot it better (and images of the board, from both sides, could be of help too!)

Yes, it is always my habit to test diodes in both directions.  They all test fine.  I am not interested in replacing the whole board or module since the cost will exceed what the TV is worth.  It is a big old heavy plasma after all.  I have the service manual but it is no help since the transistors are part of the integrated power module which is meant to be replaced as a whole unit.

Do you have high res images of the repair? As if its anything like that image you took from online then I would certainly not say its got good connections / workmanship.

We need to see what exactly you are working with. Generic images online can only show so much.

I do not have any at the moment.  I am away from home right now so I will not be able to get some for a while.  No, my repair does not look as bad as the generic photo I found.  I just used that photo as it gave an idea of what the board looks like.  When I am back home I will take some good images and sketch out a schematic for the module.


Also, what are the original MOSFETs which were installed?

You can't just keep pulling random parts from random places. You need to determine the specs of the original and buy high quality replacements with similar specs. The voltages you specify will be for Vds, but what about the Vgs threshold? Is it a P-type or N-type MOSFET?

If you look at the image or video I linked, you'll see that I can't answer that question.  The original MOSFETs are bare die chips with bond wires going from the pad to the die.  There are no numbers or anything and no way to identify them.  If I could tell what the original MOSFETs were, I would not be asking this question as I could just find a suitable replacement.  The reason I am having trouble is that I have no idea what the original specs are.

No idea what the Vgs spec should be for the reasons listed above.  I know the originals are N-channel as I was able to deduce that from the layout of the circuit.

I will post some high resolution images of the pcb and a schematic of the module when I get back home.
canadaboy25

- Sometimes the light at the end of a tunnel is an on-coming train
 

Offline canadaboy25Topic starter

  • Regular Contributor
  • *
  • Posts: 152
  • Country: ca
Re: Help Finding Suitable Replacement for MOSFET
« Reply #10 on: July 29, 2021, 07:01:32 am »
Okay, I finally had time to get the TV pulled back apart.  I made up a quick schematic of the IPM module.  Sorry, my Altium library for IC2 and IC3 is messed up.  They are both ZXTC2045E6 complementary transistor packages.  The IPM has 3 inputs: SUS_UP, SUS_DN, and +18V.  It switches the Vs rail to ground to produce the SUS_OUT waveform.  I didn't pull the caps off the board to measure them as it is very hard to get the components off without damaging them due to the huge thermal mass of the metal backed pcb.  They are also glued down with some serious adhesive which does not seem to soften with heat.

Hopefully the schematic will help give a better idea why the transistors I have tried are not up to the task.

I am currently replacing R1, R2, R7, R8, D1, D2, IC2, and IC3 since some of them were damaged when the STD12N60DM6 transistors exploded.  The high/low driver still seems fine with a diode mode check and the 39K resistors R5 and R9 still test fine.

I will reassemble the board and power it up with the FDB12N50TM transistors that didn't short but produced ghosting and capture some scope traces from the two input signals to the board.

If anyone has any comments or suggestions please let me know!

EDIT:  Whoops, made a mistake on the schematic.  D1 and D2 are UDZVTE-1718B 18V zener diodes.
« Last Edit: July 29, 2021, 08:05:49 am by canadaboy25 »
canadaboy25

- Sometimes the light at the end of a tunnel is an on-coming train
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf