820-00165 pulsing and stuck at S5

[speggio91]

Hello everyone!

I have this board that got slightly water damanged and it is stuck at S5 state. It pulses once every 5 seconds.
When connected the magsafe 2 to the bench psu, the current goes from 0.02A to 0.10A when is pulsing.

When using the genuine magsafe 2 charger, the light turn on and it became green and then amber. The fan is not spinning.
When I tried the SMC bypass mode, the fan spins a quarter every 5 seconds and the pulsing is still present every time the fan spin.

The pulses are present in the CPU Vcore rails on L7320 and L7310.

At a visual inspection I noted that:
U1950 had the c1950 burnt on pin #1 and it didn't have contact with the PP3V42_G3H line. I removed that cap and jumped a wire from the pin #8 of U1950 to the nearest spot where PP3V42_G3H line is present.
R5620 has a burn mark on pin #2 and C5620 is corroded. Some traces of corrosion are also present around U5620.
C7319 has a burn mark on pin #1, resistance to ground of PPBUS_S5_HS_COMPUTING_ISNS is over than 1M ohm and it rapidly decreases to Kohm range after few seconds

I took a look at the wiki page about the pulsing condition and I've checked U5110. Its looking good and all the resistors and capacitors read good values with no shorts, even when compared to a functioning board.
Also R1275 is measuring a good 0ohm value.

I took some measurements:

pp3v3_s5_avref_smc 3.3V
ppbus_g3h 8.6V
ppvrtc_g3h 3.3V

PCH_SRTCRST_L 3.33V
PCH_INTVRMEN 3.23V
PCH_DSWVRMEN 3.23V
SYSCLK_CLK32K_RTCX1 0.44V
PM_DSW_PWRGD 3.39V
PM_BATLOW_L 3.38V
PM_RSMRST_L 3.30V
SPIROM_USE_MLB 3.30V
pp5v_s5 5V
S5_PWRGD 3.3V
pp3v3_s5 3.3V
pp3v3_sus 3.3V
PP1V05_SUS 1.050V
PM_RSMRST_L 3.33V
PM_PWRBTN_L 3.39V
PM_SLP_S5_L pulsing
PM_SLP_S4_L pulsing
PP3v3_S4 240mV pulsing to 1.5V
RTC_RESET_L 3.33V
PLT_RESET_L pulsing
PM_S0_PGOOD pulsing
ALL_SYS_PWRGD pulsing
PM_PCH_SYS_PWROK pulsing
PM_SLP_SUS_L 3.33V

I took a video with the infrared camera and I noticed that U3900 get warm with the peak temperature of 27 celsius in a partcular spot when the lines are pulsing. Maybe it's nothing relevant but I was curious to know if it had something to do.

Any ideas of what else to check? Could be the SPI chip corrupted?

[mon2]

https://repair.wiki/w/A1466_2015-2017_13%E2%80%9D_MacBook_Air

https://logi.wiki/index.php/820-00165_Pulsing

[reformatt]

Since you have Vcore, pulsing is either shorted caps above the CPU (C1270 etc) or not loading the BIOS from U6100. Check your termination resistors and also meter them from U6100 side to CPU side. For example R6120/6125 to R6110 (SPI_CS0_L). This one is a common fault due to a corroded test pad to the right of R6116.

I have had one board where this still occurred despite the resistors etc being ok. In this case, the CPU needs reballing to fix.

[speggio91]

All the caps above the CPU are measuring good values, no short have been found.

All these SPI lines are connected to the respective resistors.
SPI_ALT_CS0_L
SPI_CLK
SPI_MOSI
SPI__MISO_R
SPI_IO2_R
SPI_IO3_R
All the resistance values are correct, no traces of corrosion or liquid damage has been found on the SPI area or R8020 one.

Measuring the resistance of all the main S4, S3 and S0 power lines didn't show any shorts to ground.

I've tried to replace from a working board U1950 and u7400, nothing has changed, the board is still pulsing.
Following the wiki troubleshooting (No power, PM_SLP_S4_L is missing, 23-24 mA draw), I've removed R8111 and all the S4, S3 and most of the S0 powerlines (exept for the Vcore related ones) started to show up and the pulses stopped from happening.
The current consumption went stable to 130mA.

Is it safe to assume that the CPU is the problem at this point?
Is it a waste of time trying to reflow the CPU with just a heatgun, some good quality flux and a thermocouple?

[reformatt]

Also check J6100 is clean and free of corrosion etc. Forgot to mention this one as it is also a common reason for pulsing as the SPI lines are exposed there. BIOS is also Quad Enabled, so check U6101 is ok.

You need to check continuity for each track from BIOS to the CPU as I said in my earlier post.

R6110 --> R6120
R6111 --> R6121
R6112 --> R6122
R6113 --> R6123
R6118 --> R6130
R6119 --> R6131

Common on most Air boards to have some type of corrosion at the bottom of the board, so eyeball everything from side on. Things like U1930, U8005 etc. If everything checks out, I'll try a donor BIOS, reflow/reball the SMC and then I will blame the CPU. Use of a heatgun is a good way to consign the board to the donor pile. These CPU's have a tendency to warp when heated I'm told.

[speggio91]

Hello again,

J6100 is completely clean, no water went on it.
In the SPI area there wasn't any water infliltration either. I've checked all the connection between all the resistors and they are all interconnected.
I've replaced U6100, U5620 from a donor board, nothing has changed.

On the bottom side of the board there is nothing that got touched by the liquid, no corrosion was found.

I've first reflowed the SMC, nothing has changed.
I've then desoldered the SMC and I've found that the pad 117 (PP3V42_G3H) is gone. I've scraped underneath it and the copper is still there and it is still connected to that line.

Should I try to replace the SMC with a donor one or should I try with the reball of the original one? Could that missing pad suggest that the SMC is dead?
I think that pad must be restored in order to make the SMC function again. Its the J12 from the schematics if I traced it right. It should bring PP3v3_S4 to the SMC, right?

[reformatt]

Pad 117 = K12. Letter I for SMC is skipped in the schematics as it's likely too similar to 1.

So when you say the pad is missing, do you mean just corroded or the only the VIA remains? Uncertain for me as whether this would have caused your issue, it may be a result of the removal of the SMC possibly. If you want to be sure, repair the pad and use a known good donor. But it's sounding like its a bad connection under the CPU for the SPI lines.