820-3209-A - Fan spins high.

[gcain]

Customer said laptop was in cupboard for a few months. Laptop is suspiciously clean...
MagSafe gets Green & Orange light.

Fan doesn't do the start-stop-start-stop thing; it just turns on then slowly spins faster and faster till it reaches warp-speed.

No backlight or any other signs of life.

I don't even know where to start looking; 99% of my repairs are dead or liquid damage.

 

[peterfixit]

check power rails, find whats missing, you can skip s5 s4 and s3

 

[larossmann]

measure cpu vcore, check r7402 if no vcore.

 

[gcain]

PP3v3_S0 = 3v3
PP3V3_S0_HS_COMPUTING_ISNS = 3v274 - 3v28. always moving.
PP3v3_SUS = 3v330
PP1v8_S0 = 1v78


R7402::1 = 360mV
R7402::2 = 8v39

PPVCORE_S0_CPU: 0.816
PPVCORE_S0_CPU_PH1 = 0.822
PPVCORE_S0_AXG = 0;

 

[peterfixit]

measure cpu vcore, check r7402 if no vcore.

do you sleep?

 

[gcain]

do you sleep?

Not at the moment, no. haha.

Anyway, back to the point;

R7402 == 91k. So it's within spec.

 

[peterfixit]

i ment louis, its 2am there,, if it was me id change r7402 or check voltage on working board

 

[gcain]

i ment louis, its 2am there,, if it was me id change r7402 or check voltage on working board

I thought it was an odd question. lol

I don't have a board to test with. It measures right, I think U7400 might be pulling it low?

 

[peterfixit]

its unusual for a resistor to pull down the voltage that much, does it visually look fucked up?

 

[gcain]

It wouldn't be the resistor pulling it down, wouldn't U7400 be pulling it down?

Just to check my thought process;
That's why R7402::2 is higher than R7402::1. The resister is restricting the flow, allowing the difference.
If R7402 had 0 resistance (just for example), then voltage would be the same both sides and some component further up the ::2 side of the line would be getting hot.

Board looks spotless.

 

[peterfixit]

resistors usually fail by increasing resistance, caps turn into a wire. so, resistance to ground?

 

[dukefawks]

Fan ramping up is just that the CPU is not running. This one is going to suck probably. I would first try a known good BIOS chip, besides that there is not much you can do if it is not a liquid damage. Save the original BIOS file!

 

[gcain]

Is it possible to buy the header clip for the diagnostic port?

Would make flashing the bios a lot quicker if I could build a little adapter for it.

 

[dukefawks]

Pretty sure you can find it, but it would require research. I have no time to mess with that. I doubt it will be the BIOS though, just a last ditch effort.

 

[gcain]

I managed to find the connector. I flash these things just often enough that it's probably worth having a jig.

I have two boards now that are doing the same thing.
Should I throw then in the bin, or keep trying?

 

[SMMRepair]

FYI, those connectors are garbage. You'll go crazy getting corrupted/partial dumps, read/write errors, "no chip found", etc. Really aggravating. Better to just take 3 minutes to solder wires or pull the chip. Just a suggestion.

 

[gcain]

Thanks for the heads up.
I did experience something similar with a clothes-peg-adapter I built a few years back now. Maybe pulling them off is a cleaner solution.
Once an EFI chip (13"Retina board) cracked in half when it came off and it turned into a nightmare. So I've been reluctant to pull them off unless I have to ever since. Of course you could say it was already compromised which is what lead me to remove and flash it in the first place.

 

[dukefawks]

You have 2 boards that have the same? The most common case is liquid or someone stabbed a component off. BIOS is last ditch.....