820-00165, green & orange light, no s4 rails

[jpadie]

Any help with this board would be appreciated.

no history to help. the board was clean; no evidence of corrosion nor any tarnished test points.

all g3h rails are present and correct. Good 32.7kHz pulse coming from pin 12 of the RTC (replaced RTC in any case as I was seeing a pulsed voltage - gunk on the leads in the end).

All s5 rails are there.

PM_SLP_S4_L is not present. Ditto PM_SLP_S5_L. not seeing anything change when the SMC_ONOFF_L is shorted.

I've reflowed the SMC and checked all the SPI resistors and traces.

is this probably a dead PCH/CPU?

One thing to note is that the resistance to ground on L7520 (PP5S4RS3) is very low when the IO board is plugged in. With one IO board it was less than 1Ohm, on another I tried it was 20Ohm. When unplugged the resistance is OL. I don't have any other IO boards for this model to test against but this does seem odd. Nothing getting hot though. Current draw is only 0.02A at 14.5v.

thanks
Justin

 

[2informaticos]

Disconnect LIO flex, leave only 5 pins DC cable connected.
The MLB should start when you plug the magsafe2 charger.

Do you get high level at PM_SLP_SUS_L and PM_RSMRST_L?
PM_PWRBTN_L reacts when you trigger SMC_ONOFF_L?
3V3_SUS present?

 

[jpadie]

thanks for the reply,
without the flex cable:

PM_SLP_SUS_L 0mV
PM_RSM_RST_L 0mV
PM_PWR_BTN_L does not react when R5115 is shorted. stays at a stable 3.41v .

 

[2informaticos]

Check the 4 PCH's RTC signals.
Also check SMC_ADAPTER_EN, PM_BATLOW_L, S5_PWRGD, PM_DSW_PRGD and PCH_DSWVRMEN.
Do you get 3V3_S5 and correct PPBUS_G3H?

 

[jpadie]

PP_3V3_S5 is present at 3.27v
SMC_ADAPTER_EN is present at 3.4v
PM_BATLOW_L is not present (3mV)
S5_PWRGD is not present
PM_DSW_PRGD is not present
PCH_DSWVRMEN is not present
PPBUS_G3H is 8.16v

re the PCH RTC signals, which ones did you mean?
PCH_SRTCRST_L is low
RTC_RESET_L is low


when I measured PCH_INTRUDER_L the current jumped to 260mA however (from 10mA). so something has changed! and the orange light turns to green after 5 ish seconds. then back to orange again. as if the board is power-cycling. the CPU is also getting very hot. but no fan spin.

 

[jpadie]

NB it is the left hand of the chips that is getting super-hot (as you look at the board with the fan on the right). I think this is the PCH.

 

[jpadie]

i plugged it back in over lunch to test again - flames leapt from the PCH so I think this is now officially a dead board.

thanks!

 

[2informaticos]

Probably overvoltage on VCCRTC killed PCH...