820-4924 freezes with battery plugged in

[fijolin]

I have 2 non-liquid damage 820-4924 boards with the same exact issue it seems.

When the battery is plugged it, they freeze whenever there's enough load on the CPU.
Sometimes loading OS causes that, sometimes it doesn't. However, if I manage to boot,
anything that causes the clockspeed to go up will freeze the computer. All power-rails incl. CPU Vcore stays on.

Without battery the CPU is limited to some 0,8GHz. That way there is no issues whatsoever apart from the CPU being clocked down.

I couldn't find similar problems on the forum.

Any ideas?

 

[2informaticos]

Try known good battery first.
Also discard a software issue; using another SSD, or an USB pendrive with test OSX.

 

[fijolin]

Try known good battery first.
Also discard a software issue; using another SSD, or an USB pendrive with test OSX.

All of the aforementioned is done. Swapping the logic board to a known working one= no issues.

Have tried to boot with/wo SSD, also from USB on logic board port & on IO port. No change.

 

[2informaticos]

Then try known good BIOS, with clean ME of course.

Do NOT quote an entire post, visible few centimeters above.
Maintain forum aesthetic, please.

 

[fijolin]

Tried with known good BIOS.

Same story, without battery=no issues.
KNOWN GOOD battery plugged in= doesn't boot or if does, freezes almost instantly.

Usually when there's 2 or more instances of the same problem, there's hundreds of others having it too.

Anybody?

 

[2informaticos]

Try starting on SMC bypass mode.

 

[fijolin]

I've tried SMC bypass as well.
Same story if battery is plugged in.

The shitty build in "diagnostics" says there's no problems.
No error report on reboot either.

 

[2informaticos]

Any history available?

Try changing C1031.

 

[fijolin]

Any history available?

Try changing C1031.

Both machines stopped working gradually with the customers.
According to them, no drops etc. There's def. no liquid in them.

Changed C1031. Worked for a few minutes, then froze.
Heating up the board close to the CPU probably made it work for a while.

Would that indicate that there's a problem with the cpu itself?


EDIT: Continued trying & testing, there's def. a clear improvement in the system not constantly freezing.
Have been running cpu heavy tasks now for some 15minutes without a single freeze.

Will continue testing...

 

[2informaticos]

You can consider to heat CPU at 180-200 degrees (no more) using quality flux.
Just in case, to burn eventually crap accumulated underneath.
I say this because of the board reaction, if changed C1031 using hot air.
Test the board more time first, maybe was just C1031 issue (incresead ESR factor).

 

[fijolin]

Had the machine running overnight, no issues. Running more benchmarks now, no additional freezes after the first one after replacing C1031.

Guess I'll hand it back to the customer and pray that it'll continue working.

Will try this on the other board as well.

2informaticos Thanks for your input