Lenovo Legion Y540 burned PCB repair

[raill]

Hello,

I'd like to ask someone more experienced than me for advice - I'm trying to repair this Legion Y540 with a burnt mosfet (PU3001) through the PCB (damaged/shorted capacitor was also on the other side).

I am able to drill out the damaged part of the PCB, repair the surface paths and solder the components according to the available Board view data, but it is very difficult to find out if some internal paths are broken.

1) Any chance there is/will be CAD documentation available for the PCB layers for this Lenovo, such as for example for the Dell E7470 (.brd file openable in Allegro Free Physical Viewer)? (https://www.badcaps.net/forum/troubl...7-need-brd-for -dell-la-c461p)

2) Is there any theoretical chance that there are no internal paths in this particular damaged spot on the PCB?

3) I will be grateful for any suggestions.

Thank you!

[piernov]

Not much point wasting time here. CPU will be dead, and in my opinion PCB damage is too extensive to be rebuilt reliably.

[raill]

Thank You, although this is not very optimistic news

[raill]

Hello again,

trying to repair it in my free time even if it will require CPU replacement. Doing it for my own education...

Already "cleaned" burned hole and replaced affected components around + replaced MP2979A chip.

All shorts resolved, battery charging is working, all voltages present (EC and GPU is warming up when charger is connected), except one serious problem - on CPU VCC rail (coils) i measure (almost) main power voltage - see pictures.

May i please ask some experienced with this type of power supply knowledge, wht can cause this?

Thank you for any help!

[Sephir0th]

Most likely means, at least one of these three DrMOS is internal shorted.

If only the phase, you are measuring here, then it must be the DrMOS right next to the probe.

[raill]

ok, ill try to remove them one by one and see.

Thank You very much!

[ktmmotocross]

burned welded layers together. must grind it out

[raill]

Hmm, i measure same 19V on all phases, so it seems i really missed some layers shorted when grinding burned hole...
Really happy
Thank You!

[raill]

Most likely means, at least one of these three DrMOS is internal shorted.

If only the phase, you are measuring here, then it must be the DrMOS right next to the probe.

It seems, You were right!
Removed all of them one by one, then replaced one and it produces correct voltage! Maybe all of them were partitially shorted...
During the weekend ill try to solder rest of them and check if CPU is alive

Thank You one more time!

[raill]

Some sad progress here - after replacing all DrMOSes still measuring 19V on all of them. Cleaning/increasing hole did not help...

[raill]

Another possibility came to my mind - the MP2979A controller i replaced for new one (from Aliexpress) needs to be programmed to set proper voltages?

[FriedFred]

Some use the factory presets or gets the configuration bits from EC chip while booting.
With a thermal camera, you can find the hot-spot more easy. I have some experience with burned areas and the high current makes the FR-4 material carbonising. It's necessary to remove all residue of that carbonised material until the short is gone.
For my experience, about 50% of those main power rail shorts, the GPU or CPU survives it.