820-3437-B fan spins, ramps up to high speed, no display/backlight

Fanfwe

Member
Hi,
I was given a not working MBA in which coffee has been spilled.
First inspection revealed a bit of corrosion next to U7200. I cleaned that, and replaced the caps and resistors involved.
The rest of the board has traces of coffee, but nothing too bad and no other trace of corrosion. I cleaned the coffee with alcohol.

When I power the board, I get a green light and then orange light on the charger, so SMC talks to the charger properly.
The fan then starts spinning, which is usually a good sign. Board draws 156mA at that point.

Then about 8 seconds later, the fan slowly starts ramping up to full speed. Once it is at full speed, I have a 192mA current draw on the board.
If I connect the board to the display, I get nothing there. No backlight, and also no image. I have no way to know if the display is working fine, and I don't have another one of that kind.

I checked most power rails and they look good. I have working Vcore.
CPU doesn't get hot.

So at that point I'm not sure if I have a working board with a defective screen, or a defective board. Any suggestion what I could try next ?
 

Fanfwe

Member
I put the board back into the topcase, with all ribbon cables attached, and now the fan spins at normal speed. Board draws 156mA.
I assume the SMC wasn't happy while outside of the topcase due to missing sensors of some kind and commanded the fan to spin at full speed.

So now it looks like I have a board with a brain but no display.
 

2informaticos

Administrator
Staff member
Air boards fan spins at max if trackpad is not connected.
Tell us if you get USB activity.
You can test with USB mouse, see if gets steady light.
Always better to use an USB OSX installed in a pendrive with Link LED; if flashes, there is USB activity 100%.
 

Fanfwe

Member
OK, I used a pendrive with OSX installed. Boots fine on other apple laptops, just to be sure.
On this board, I get no activity when I turn power on. Sounds bad I guess.

Anything else to test before trying to reflow CPU ?
 

2informaticos

Administrator
Staff member
Post all S3 and S0 voltages which appear on the board.
Don't forget to check pages 57/58, apart of big coils.

This board should make three start attempts, before stays on.
You can also try a known good BIOS file.
 

Fanfwe

Member
All PP5V_S3s read 5.12V
PP5V_S0 reads 5.12V
PP3V3_S3 reads 3.33V
PP3V3_S0 reads 3.3V
PP1V8_S3 reads 1.79V
PP1V5_S0 reads 1.50V
PP1V2_S3 reads 1.20V
PP1V2_S0_SD_AVDD12 reads 0.95V
PP1V2_S0_SD_VUHS1 reads 3.33V

PP1V05_S0 reads 1.05V

Checked all big coils. I have power everywhere except for LCD backlight and the Thunderbolt 15V coils.
What do you mean with the three start attempts before stays on ? How can I see those three attempts ?
 

2informaticos

Administrator
Staff member
You should see fan spining on/off two times, before the board stays on.
Try full known good BIOS file; some link should be already on the forum.
 

Fanfwe

Member
I haven't touched the BIOS yet (need to find a SPI ROM programmer, or build one out of a Raspberry Pi and flashrom), but I inspected the board under the microscope thoroughly once more.
I found an area near U1950 where there were still traces of coffee spill.
Although there was no sign of corrosion, I suspected that the underside of U1950 would be very dirty, so I decided to remove it, in order to check.
This revealed that pin 8 (PP3V42_G3H) was totally gone, both on the chip itself, and the pad on the board. Also there was a capacitor next to that pin which was totally destroyed (came into pieces with just a little push of tweezers)
Interestingly, with the chip off the board, I now can see what you described, with the 3 start attempts.
I have repaired the missing pad with a tiny bit of wire, and I now will need to get my hands on a replacement part for U1950 (could take a little while).
Could a bad PCH_PWROK/PCH_SYS_PWROK have explained the previous symptoms ?
 
Top