820-3437 Power Issue

[iambehr]

Getting all g3hot rails. Not getting any s4 rails. ALL_SYS_PWRGOOD cycles between ~.5V and 0V.
Any ideas?

Thanks,

Stefan

 

[smiba]

How are your S5 rails doing?

Voltage of PP3V3_S5, PP5V_S5, S5_PWRGD, PM_SLP_S4_L, PM_DSW_PWRGD?

 

[iambehr]

So..

pp3v3_s5 and pp5v_s5 are present.
s5_pwrgd = 3.42
pm_dsw_pwg = 3.42
pm_slp_s4_l is fluctuating between 0 and 2 volts every couple of seconds.

s4 rails are cycling voltage similarly as well.

 

[dukefawks]

So it is power cycling. The usual suspects apply again: U1950 pin 8, U1900, U6100 and termination resistors/traces to it.

 

[iambehr]

u1950 getting 3.42 volts. All termination resistors test as ok (none failed high resistance anyways) and the traces connect thru on each end.
No issues with pads / traces at u1900, chip is getting power.
I will probably replace u6100 with a known working one and see what happens.

 

[iambehr]

Changed the bios chip w/ a known good one. No change in state at all. Maybe time to throw this one in the bin?

 

[dukefawks]

CPU Vcore pulsing too?

 

[iambehr]

Yes, 0-.5v in two second intervals.

 

[dukefawks]

I still say U6100 traces/resistors. HiRes picture of area please?

 

[iambehr]

No Title

Here ya go. "Hi-Res"... from the crappy amscope camera. I just need to buy the canon slr adapter.
I checked and double check all of the lines back to the resistors under the CPU. 3.3v getting to all of the pins on the bios chip that need it. U1900 getting all of the proper power as well.

 

[dukefawks]

There are a lot more termination resistors than in those pictures. Look at schematic, 16 of them or something. Also check for corrosion on J6100.

 

[iambehr]

Checked all resistors on page 61 "SPI Bus Series Termination". J6100 always gets removed and the pads wicked cleaned. I'll recheck monday again maybe I missed one.
I think originally the only visible signs of liquid damage were at j6100.

 

[iambehr]

I think it is probably PCH or SMC. I'm going to set this one aside and come back to it.

 

[dukefawks]

You checked the resistors but did you check the traces?

 

[iambehr]

So to revisit.. I have two of these with the same symptoms. Checked end to end continuity of all spi termination resistors and traces.
U6100 is getting power, U1900 is getting power.
Could the clock chip go bad?
Is this a common no fix?

 

[dukefawks]

It is a symptom that can have 100 causes. Is CPU Vcore pulsing?

 

[iambehr]

Yes.
The strangest bit - I accidentally disconnected SPI_MLBROM_CS_L from SPI_CS0_L and CPU Vcore and all rails stabilize, and the fan spins up full. No display out or signs of booting at this stage - which I do not expect would happen.
I will replace the flash chip next workday to see if it is bad.