A2141 820-01700 Is this dead CPU - 0.5V 0.01A Short on PPBUS_G3H

it-solve

Member
Short on ppbus_g3h line
Removed the fuses and injected 1V, CPU glares up. Customer needs data desperatly.

1669495435509.png

Dead CPU?
 

2informaticos

Administrator
Staff member
Very likely...

How many times need someone to tell you STOP to inject voltage into PPBUS_G3H in this manner.
Don't say 1V is safe voltage; just a childish excuse.
You only need ohm scale measurment to detect CPU power supply problem; why to inject voltage first???

Now find the burnt CPU power supply chip and check VCORE resistance to ground, after removed it.
 

it-solve

Member
I just got A1707 with the same type of short circuit on the PPBUS_G3H line, and 1V injection detected a shorted cap on the audio line.

I agree with what you said, but can we clarify what internal ohm shortage or case scenarios we should look for before any injection? Even on low voltages. Removing all CPU GPU MOSFETs/Coils before any injection surely is not cost-effective.

We surely cannot open a new thread each time we want to diagnose a short.
 
Last edited:

2informaticos

Administrator
Staff member
You don't need to completely remove anything.
Once you detect the short between PPBUS_G3H and any big coil, just lift (one side) the corresponding coil.
PPBUS_G3H resistance to ground should go up then.
Be aware, CPU and GPU have 2-3 coils for the same rail, so need to lift all common coils.

In some cases (rarely), the burnt chip/MOSFET cause short to ground too.
Lifting coils don't help like above then.
 
Top