820-3115-B board only works on SMC bypass mode

[sangolossyfon]

	regular mode, plugged in	SMC bypass mode
CPU_ISENSE	0v	0v
DCIN_ISENSE	0.04v	0.04v
HDD_ISENSE	0v	0.01v
BMON_ISENSE	0v	0v
CPU_HI_ISENSE	0.07v	0.01v
OTHER_HI_ISENSE	0.08v	0.02v
MEM_ISENSE	0v	0v
CPUVCCIO_ISENSE	0v	0v
AXG_ISENSE	0v	0v
CPU_VSENSE	0v	0v
DCIN_VSENSE	0v	2.08v
PBUS_VSENSE	0v	1.4v
AXG_VSENSE	0v	0v

I could compare these to working board if you want but they seem to be mostly 0v or near it except for the two that work in SMC bypass mode.

 

[2informaticos]

Why would you want to check them in normal mode, if board stays off???

Compare the readings in SMC bypass with any other board which turns on.

 

[sangolossyfon]

	regular mode, plugged in	SMC bypass mode	Working Board
CPU_ISENSE	0v	0v	0.10v
DCIN_ISENSE	0.04v	0.04v	0.33v
HDD_ISENSE	0v	0.01v	0v
BMON_ISENSE	0v	0v	0.06v
CPU_HI_ISENSE	0.07v	0.01v	0.12v
OTHER_HI_ISENSE	0.08v	0.02v	0.09v
MEM_ISENSE	0v	0v	2.64v
CPUVCCIO_ISENSE	0v	0v	0v
AXG_ISENSE	0v	0v	0v
CPU_VSENSE	0v	0v	1.11v
DCIN_VSENSE	0v	2.08v	2.72v
PBUS_VSENSE	0v	1.4v	2.09v
AXG_VSENSE	0v	0v	0.01v

Updated the chart. CHGR_BMON on good board was 0.07v.

 

[2informaticos]

DCIN/PBUS_VSENSE are both too low.

Post Q5480/90 voltages.
Be sure R/C5489/99 are not corroded.

 

[sangolossyfon]

resistors had no sign of damage
below I am posting first broken board voltage in SMC bypass mode, versus working board voltage for comparison

Q5480
PBUSVSENS_EN_L- 0v(broken) vs 0v(working)
PBUSVSENS_EN_L_DIV- 6.09v vs 6.25v
PPBUS_G3H- 12.26v vs 12.55v
PM_SLP_S3_R_L- 3.3v vs 3.3v
PBUS_S0_VSENSE- 12.27v vs 12.57v
Q5490
DCINVSENS_EN_L- 0v vs 0v
DCINVSENS_EN_L_DIV- 8.02v vs 8.28v
PPDCIN_G3H- 16.81v vs 16.62v
DCIN_S5_VSENSE- 16.81v vs 16.62v
DCIN_VSENSE_EN- 3.3v vs 3.3v

only difference I see is overall G3H voltage and PBUS_S0_VSENSE- so maybe bad isl6259 as SMC cannot communicate with it? Or is 12.26v always the G3H voltage in SMC bypass mode?

 

[2informaticos]

"is 12.26v always the G3H voltage in SMC bypass mode?"
Yes!

Recheck C5489/99 voltage.
If still low, remove these capacitors and test again.

 

[sangolossyfon]

PBUS_VSENSE and DCIN lines had same resistance to ground as good board

still 1.46v and 2.11v on the two lines even with capacitors removed

r5489, 5488, and r5499/98 all are correct values on broken board

 

[2informaticos]

Check again the voltages from these areas.

Posted values do not comply with bottom equation.
SMC_PBUS_VSENSE=PBUS_S0_VSENSExR5489/(R5489+R5488).

Something similar must be accomplished for SMC_DCIN_VSENSE.

 

[sangolossyfon]

I verified every voltage and resistance, and also checked that resistance to ground on the PBUS and DCIN lines were the same as working board.

Is it possible that this is just bad SMC or bad connections below SMC? All the other components are fine.

EDIT: reflowed SMC- board now boots and draws 15w! so likely working!

 

[2informaticos]

Bad solders or/and accumulated crap underneath SMC.
That's the reason of low SMC_DCIN_VSENSE, preventing the board to turn on.

 

[sangolossyfon]

unfortunately the board did stop working when I tested it again. It kind of tries to boot and spins the fan then stops then tries again. In SMC bypass mode it sometimes gets to a full boot and I hear the chime, but then it dies again. I have never gotten it to show anything on the display either even when it chimes. Sure enough, if I merely heat up the SMC, it boots more consistently- so definitely just a bad SMC. I even took off the SMC, it looked perfect underneath, but same behavior when I put it back on. Voltages are the same as before. So it looks like the voltage/current monitoring portion of the SMC is bad. Replacing it and reballing it is beyond my skill level at the moment.

I am also sunsetting my macbook repair business as I go off to college but I do appreciate the help on all these boards, I have learned a lot from this forum!

 

[2informaticos]

You must be sure AVREF_SMC (C4921) is steady high (3.3V).