[SOLVED] 820-2936 Quarter fan spin - yay!

Status
Not open for further replies.

MUUK

Member
820-2936

liquid damage to SMC area, SMC reset chip and clock chip area.
Had no green light.
Replaced SMC reset chip, reflowed SMC. Now get green light and PPBUS is 12.56v

Now get quarter fan spin


Replaced U2800
replaced Y2805
replaced caps etc in that area

no change.

Checked CPU Vcore. It is not pulsing on with fan spin.
PPBUS_S5_HS_COMPUTING_ISNS is present

U7400

pin1 3.3v jumps
pin2 12.46v jumps
pin15/24 5v jumps
pin 20 jumps all over, difficult to tell but above 2v
pins 21/22/23 0v
pin40 0.25v jumps

so i am lost.

Not sure if this is clock related or should be looking more at U7400. I think that Vcore should pulse with the fan spin.

Grateful as always for any help.
 

larossmann

Administrator
Staff member
1) Resistance between pins 17/18 of U7000, and 27/28 of U7000?

2) Is there a short on any S0 rail(do not call vcore a short)? PP5V_S0, PP3V3_S0, etc.

vcore should pulse with fan spin so its lack of pulsing is a bad sign.
 

MUUK

Member
Thank you Louis.


1. U7000

pin 17-18 2.8ohm
pin 27-28 20.3ohm


2. S0 rails not jumping to correct voltage (all others present)

pp5v_S0 jumping but only to around 0.2v
PP0V75_S0_DDRVTT only jumps to 0.02v
PPVCORE_S0_CPU 0v
PPVCORE_S0_AXG 0v
diode mode on pp5v_S0 0.398v drop to ground.

checked Q7860
pins1,2,3 (pp5v_s3) first jump high - like seems to jump to 10v (can that be right) then went to 0v
pins ,5,6,7,8 (pp5v_s0) jumps only to 0.2v
R7862 + R7860 measure 150k over all but lower resistance than expected so ignore.

So it seems pp5v_s3 is missing. Checking U7200

pin3 2v (??)
pin13 (en) 3.4v
pin16 12.56v
pin17 5v
pin20 - seems like it jumps to 10v - ??
pin23 - jumps to 3.4v

Tried replacing U7200 but no change.


Tried lifting L7260,
Diode mode pin2 is 0.120v, pin1 is 0.48v (v similar to other board).
So its a problem with generation.

Replaced Q7260 / Q7261
Diode mode now matches other board, but still same quarter fan spin.

Thanks for any help you might give!
 
Last edited:

larossmann

Administrator
Staff member
WHAT! I asked for resistance to ground on S0 rails and you replace U7200? WHY??? I do not need all this information, just information requested. :( Replacing Q7260 makes no sense since we have no reason to believe it is bad yet.

Answer my question and as punishment for providing all this unnecessary info and replacing U7200, check for shorts on every s3 rail too. PP5V_S3, PP3V3_S3, etc.
 

MUUK

Member
Sorry Louis.

Resistance to ground S3 rails.

PP5V_S3 20.3k
PP3V3_S3 19.8k
PP1V5_S3 115ohm
PPVTTDDR_S3 25.1k
PP1V5_S3_CPU_VCCDQ 113ohm

Resistance to ground S0 rails

PP5V_S0 29.5k
PP3V3_S0 8.8k
PP1V8_S0 295ohm
PP1V5_S0 14.68k
PP0V75_S0_DDRVTT 157k
PPVCCSA_S0_CPU 33ohm
PP1V05_S0 22.1ohm
PP1V05_S0_PCH_VCCADPLL 1.25k
PPVCORE_S0_CPU 9.3ohm
PPVCORE_S0_AXG 10.8ohm
PP1V05_S0_CPU_VCCPQE 21.5ohm

Thank you.
 

MUUK

Member
Hi Louis,

Yes. ALL_SYS_PWRGD does pulse.

I checked a lot of things. The only things that looked odd were:

1. pp5v_S0 jumping but only to around 0.2v (which sent me on my U7200 journey as it appeared to be the only rail not jumping to the voltage it should be)

2. MEMVTT_EN is missing and not jumping. Although PM_SLP_S3_L is jumping (so for R3210 i get jumping 3.4v at pin 1 but 0v at pin 2).

3. May be related to 2 but PLT__RESET_L is a steady 0v and does not jump. I get a flash of voltage on PP5v_S3 at R3202 (too fast to read only get a lightning bolt on my fluke)

appreciate the help.
 

MUUK

Member
Hi Duke thanks for the reply.

R7402 measures fine, voltages correct either side. (12.56v/12.46v)
R7401 also measures fine.

is 0v (not jumping) on PLT__RESET_L and MEMVTT_EN worth investigation or should they come up after Vcore ?
 

dukefawks

Administrator
plt_reset will only be released after Vcore has come on. At this point it could be anything, most likely is crap under SMC or PCH, or corroded trace/resistor.
 

MUUK

Member
Was the bios chip (U6100). Donor chip boots all the way through - was literally last thing i was going to do before donoring it no signs of damage either. Some ME fuckery to do but can mark this as solved! thank you both for your help.
 
Status
Not open for further replies.
Top