LG TV PSU MOSFETs keep exploding

[cmlewis89]

Hi folks,

So I need some help repairing an LG LCD 47LB5DF PSU. The main caps finally gave out after 17 years and I thought it would be a fun challenge to try to fix. I replaced the main 2 caps, both main MOSFETs (one (Q506) was shorted both gate ~ drain and drain ~ source, other (Q507) was fine), main 8A fuse, standby IC (U502), resistor to standby IC (R508), the 4 thermistors (one exploded - TH502), and a small cap (C510).

When the board is isolated, I'm getting stable voltages throughout. The main caps are charging at a stable 167V. I'm also getting a very stable 5.01V DC out of the standby voltage output pins (nothing out of anything else). Everything looks fine until I connect the PSU to the main board. Doing so causes one of the main MOSFETS to immediately explode followed by the main 8A fuse. It's always just the one MOSFET (Q506) that fails (shorts open gate ~ drain). The drain ~ short diode remains good.

I've tested basically all of the MOSFET drive circuit components and even replaced the drive IC (U600). I'm not sure where to look next for potential faults. I couldn't find a schematic, unfortunately, so I just uploaded some pics (I tidied up the solder joints btw). I'd appreciate any pointers in the right direction.

Thanks!

[Harry_22]

Hi!

It seems U502 is PFC controller and Q506 and Q507 are running in parallel.
Could you clarify U502 name.

[cmlewis89]

Hi,

U502 is an Infineon ICE3B1565J. The original had a shorted internal MOSFET from the power surge, so it was shorted closed CS ~ drain (pins 3 ~ 4). This is the datasheet I found on it: https://www.alldatasheet.com/html-pdf/152673/INFINEON/ICE3B1565J/1959/5/ICE3B1565J.html

That's correct, the 2 MOSFETs in question are running in parallel. The model # is: Infineon SD20N60.

[Harry_22]

Sorry PFC is another one U600, please see the picture.
Please send me the name as well.

[cmlewis89]

Certainly, U600 is a Fuji FA5501AN. This should be the datasheet: https://pl-1.org/getproductfile.axd?id=11723&filename=FA5501AN.pdf

[Circlotron]

Maybe whatever circuitry that drives Q506 gate is also damaged and is providing wonky gate drive, e.g. slow fall time. Just good enough to make things work with no load, but full load pushes Q506 over the edge.

[cmlewis89]

Agreed. I'm just hoping there isn't an issue with the load (aka the main board) - not sure how likely it would be for a power surge to travel that far downstream.

I've uploaded another pic of the circuit with better lighting in case that helps. Please excuse the messy desoldering job/flux around the filter caps/MOSFETs (both removed in pic). Obviously, some of the traces got damaged in frustration during initial desoldering. I've since invested in a proper desoldering station (sooo much easier) - bridges should be good.

Also, sorry, looking again closer at the pics, that should be Q606/607, not Q506/507.

[Harry_22]

Hi!

Check the following parts in red circles. Please use Paint.NET to see the drawing.
https://drive.google.com/file/d/1dd0aFyBGRpKkHC1cE7l4R7dtfifvHPu5/view?usp=sharing

Use laboratory DC power supply as a mains to avoid exploding. 90 VDC should be Ok.

[cmlewis89]

Wow, this is awesome! I really appreciate the time you put in to patch together all the images and label them. This is my first time trying to repair a TV, so this is all very helpful.

Ok, sounds like a good idea on testing with a DC power supply. I've honestly never done that before and will have to buy a unit. I'm assuming it's fine to just connect straight to the AC input pins (maintaining ground)? I'll try the lower/90V DC - hopefully the board still works without exploding components, obviously. Basic question: what about amperage? The board states it draws 4A in AC input.

I will begin testing all the circled components and will let you know what I find. Thanks again!