820-3023 - No Vcore, has fanspin and ALL_SYS_GOOD

smiba

New member
ALL_SYS_PWRGD high, fan spinning

U7400 power on pin 15, 24, 40. CPUIMVP_VR_ON high.
R7402/03 within specification

CPU not shorted to ground.

First sighting of a dead U7400? Am I missing something?
The SMC has seen better days but since we get into S0 I don't think there is reason for concern.

Cheers
 

smiba

New member
No idea about the history

Its a board I've had here for almost two years which I had totally forgotten about. It does not look water damaged because all indicators are still white and there is no corrosion to be found.
It looks like the SMC was replaced(, probably by me). Can this issue be caused by a bad SMC?

If you have no idea and it could be anything I'll just replace U7400 from a donor just to confirm its not that so I can get it out of my mind.
Thanks
 

smiba

New member
The system does not seem to fully get into S0

PP0V75_S0_DDRVTT is not there. MEM_VTT_EN is low because PLT_RESET_L and ISOLATE_CPU_MEM_L is low so none of them pull MEMVTT_EN_L to ground.

As I understand the PCH is responsible for these signals, I'll keep investigating

EDIT:

Missing PM_PCH_SYSPWROK due to CPUIMVP_PGOOD being low.
Not surprising since we have no CPUIMVP out of our U7400
 
Last edited:

smiba

New member
Replaced U7400 (With a MAX17511 from a 2936 board which is a suitable replacement) and CPU Vcore is there :)

Solved! U7400 can die apparently!

EDIT: It seems to be brain dead though, no POST :(
Anything I can do to rule out a dead CPU? All the other boards I have at the moment are on hold while waiting for parts so I have the time
 
Last edited:

dukefawks

Administrator
Could be anything. This has been in a pile for 2 years, how much chance is there that something got knocked off?
 

smiba

New member
Looked for quite a while under the microscope but all parts seem to be there.

I flashed a new bios with clean me region but no luck with that either

I'll see if there are any signals going into the CPU that might be preventing it from posting, but most likely not.
There is small discolouration around the CPU so maybe some liquid was into play but I can't be sure. I've seen this as well on boards without liquid
Is there a way to reflow the CPU with a hot air station? Or will I just crack the die without pre-heating from underneath?
 

smiba

New member
All the signals to U7400 were alright as far as I could see, at best the heat woke up the CPU/PCH and made it send the right signals. Confident that U7400 was messed up though
Board is still working even after being fully cooled down again.

The last try would be reflowing the CPU at this point probably? I checked again and all S0 signals are there and solidly within specification. Would reflowing the CPU go just as I would reflow a SMC or does it require additional precaution?
 

smiba

New member
No luck, used a nice amount of flux and heated it up well enough to activate it

Still no POST :(

Any chance this might be clock related? It gets all the way to S0 though so I doubt it. Just a dead cpu?

EDIT:

Just gave the ugly looking SMC a reflow as well because it is somewhat involved in controlling aspects the PCH/CPU and it was my last hope. Sadly even through I was extra nice by carefully preheating the board the SMC is now fully dead. Missing SMC_PM_G2_EN but PP3V42_G2H and everything needed to get it going within specification

I'll see if I have a SMC for this one around but I'll probably just throw it back on the pile as a donor. Just wondering could there be a chance that it actually would have been the SMC messing up the CPU? I'm not sure if it has low enough access to the CPU to get it stuck before POSTing
 
Last edited:
Top