[LA-J891P] Warm cpu on 19v rail injection, is it fried ?

[tamerelapute]

Hi,
I'm learning electronics(already have some basics) by fixing a FH52M LA-J891P rev1B. It come from an acer nitro 5. The charger was defaulting upon pluging. The 19V rail resistance is 0ohm (I believe that caps screw my measurement). When i inject 5A, I get a rail tension of 1.3V, and 0.8V vcccore.
Except the cpu, I cant feel anything getting warmer.
I believe that it is normal, because the cpu is "eating" half the power I'm imputing.
I dont know if the cpu is fried, because I did not remove the big capacitor on vcccore.
I believe that if the cpu was fried, it would be caused by a fried mosfet, but I believe the mosfets are fine, because the vcccore tension is lower than the rail tension (i.e. no mosfet is burned open or closed)
Can someone tell me if I'm thinking right, and what should be my next move ?
I think I will try to inject 1V into vcccore, so that I can get the cpu resistance without frying it, and without removing the capacitors.
Thank you in advance for your help and advices
Have a nice day !

[MOD EDIT] Link to schematic --> https://www.badcaps.net/forum/troubl...atic-boardview

[ktmmotocross]

if u inject more than 2V u burn cpu because of shorted highside mosfet

[tamerelapute]

Hi,
If the high side mosfets were shorted, I would get vcccore = v_rail , am I wrong ?
Thank you for your help

[ktmmotocross]

yes. but charger can detect it and not do harm. injecting is another story

[jasko_jacker]

When there is a short on the 19V drop resistor it is a good rule not to inject the 19V because if a mosfet of the CPU voltage line is damaged it could happen that all the 19V go to the CPU, if we burn it it also depends on the amperage as well as on the voltage. If you think something similar has happened you need to measure the ohm value on the CPU line. It may happen that although there are apparently no problems, the CPU may be damaged. You understand this because usually the chip that manages the 19V does not supply it, probably being blocked by a powergood signal.