00138 won't turn back on after initial boot

[grmacs]

This board will usually boot fine the first time but will often not turn back on with or without battery plugged in. On bench nothing plugged in if I shutoff by grounding SMC_ONOFF_L it wont restart. IN this state, I get 3.4V at SMC_ONOFF_L with and without keyboard plugged in. When triggering on/off I get voltage drop at PM_PWRBTN_L. When it won't restart I tested voltages at U5110 and everything looks good there. Board always starts when I unplug and replug charger back in. Somerhing appears ro not be resetting when board stays connected to power but shut down manually or via OS. Only if removing all power.

PM_SLP_S4_L is not present. From what I can tell the below are all present:

PPVRTC_G3H
RTC_RESET_L
PCH_SRTCRST_L
PCH_INTVRMEN_L
PCH_DSWVRMEN_L
PP3V3_S5
PM_DSW_PWRGD
PP3V3_SUS
PM_BATLOW_L
PM_RSMRST_L
PM_PWRBTN_L

 

[2informaticos]

PM_PWRBTN_L reacts even when doesn't want to turn on?
Any pulse at PM_SLP_S4_L then (at same time)?

You've forgot to check PCH_INTRUDER_L?

 

[grmacs]

I did forget to add to list: yes PCH_INTRUDER_L is good also. PM_SLP_S4_L does go to 3.3V momentarily when toggling on/off while board is in state of being unable to start after shutdown.

 

[2informaticos]

Then check for pulse at S3 voltages.
Do the same with PM_SLP_S3_L abd S0 voltages.
Check if ALL_SYS_PWRGD pulses too.

Inspect U1950 for corrosion.
Remove Q8150 and test again.

 

[grmacs]

Well you helped me discover some things I realized the board is cycling the rails even without triggering SMC_ONOFF_L.

PLT_RESET_L is jumping to 3.3V every couple of seconds also. So if I shut the board off it seems to go to normal s5 state (rails are not pulsing), but once I trigger the board to start again, the board is continously trying to restart and I get pulse at PLT_RESET_L.

I couldn't get the board to PRAM reset when it was in the Mac so I PRAM reset via Medusa and also did cleanME. The board still won't start again without disconnecting power.

 

[2informaticos]

Did all mentioned tests?

Just in case, test starting in SMC bypass mode.

 

[grmacs]

Yes U1950 good, Tried without Q8150.

If I start in SMC bypass, then force off by grounding SMC_ONOFF, then triggering did start the board again but only briefly. Then you can see fans periodically start to spin but stop again and this continues.

You think SMC is shit maybe?

 

[2informaticos]

Any history available?

You can try heating PCH area with quality flux; in case of accumulated crap underneath
Reballing SMC can also be done.

Not sure if is related with FDMF6808N chips, but you can try replacing them too.
Use FDMD6708/3030 from reliable source (not China).

 

[grmacs]

No history. I tried heating PCH and also replacing SMC and new BIOS but no change. I think I will just replace the board unfortunately.

 

[2informaticos]

Did you test removing Q8150
Also try pressing power button 2-3s.

 

[grmacs]

I ended up as last hope, reflowing PCH w/ flux and high temp enough to melt solder and it now is restarting normally.