3662 short to ground on ppbus_g3h

[Patricia]

Hi to you all

I have a 3662 with a short to ground on PPBUS:G3H
I removed the F7140 to see on what side the short is and with my bad luck it is on the system side ....
PP3V42 is ok
Any tips on how to proceed ,,,,
The board has no liquid damage and according to the owner it just died .... no knockoff charger used .
It takes 0.245 A with the power supply and has green and orange light .
After close visual inspection I couldnt find any corrosion or anything that could direct me in the right way
I have not done anything else , want to consult with you first
Nothing gets hot according to the Flir ...

Thanks in advance for any help

 

[2informaticos]

Post exact ohm value.
Compare with ohm value readings on all big coils.
Common failure is one of FDMF6808 chips.

Do NOT inject voltage yet...

 

[Patricia]

I will have to change my posts to "Hi 2 informaticos " you are the only one replying
Sorry I couldnt post back yesterday because the option to reply wasnt there .
I removed the Phase one FDMF just to see if any change but when I measured the pins I noticed that I have a short on PPVIN_S5_HS_COMPUTING_ISNS as well
Exact ohm value on ppbus_g3H is 0.4Ohm......
I really do have bad luck with these 15 retina boards ....
All the big coils have 0.3 or 0.4 Ohms
Thank you for your answer

 

[2informaticos]

Sounds like dead CPU.
There is very low probability to still be alive.
Start removing FDMF6808 chips one by one.
Each time check diode mode at PPBUS_G3H.
When it goes high, you've just found the bad chip; can be more than one.

0.4 ohm at CPU VCORE is sign of dead CPU.
If you get aprox 1-1.5ohm, after bad FDMF6808 has been removed, you can try replace it with new FDMF6708
Buy it from reliable store, NOT chinese seller.
Then PRAY-PRAY-PRAY, power on, PRAY again and tell us if the God heard you...

 

[Patricia]

Thanks 2informaticos
yeah I have been reading other posts and usually bad FDMF6808 means dead cpu ....
I have removed the first one already and will do the other two as well ....
I knew it .... as I said with my luck with all retina 15 boards
of all Louis videos I have seen I have never seen him replacing one FDMF..... is it just me or is he extremely lucky with his retina 15 boards
I will order new ones from Digikey , I have found them very nice with support and fast shipping , mouser switzerland dont have them in stock .
As soon as I get them I will start praying .... will post the results
Thanks again

 

[Patricia]

Just a small update ... removed all 3 FDMF and short STILL there

 

[2informaticos]

Post exact ohm value (200 ohm scale) at CPU coils.
CPU could be good, but FDMF chips must be replaced anyway.

Any other big coil has the same value as PPBUS_G3H to ground?

 

[Patricia]

All 3 coils have the same value 35.6 ohms to ground (L7310 , L7320, L7330) PPBUS_G3H has 0.4 ohms to ground and the other coils are in the kiloohms (around 300 kOhms)....
I am beginning to beleive the Jessa conclusion "Its never the coil"

 

[2informaticos]

Of course never is the coil.
But same value on a coil like PPBUS_G3H signs the shorted MOSFET.
That's the reason we do this test.

35 ohm at CPU coils sounds really good.
If you had 0 ohm before, that means at least one FDMF chip was shorted.
If now stil short on PPBUS_G3H, could be a filter cap.
Should be safe to inject voltage now.
Start with 1V and slowly go high.
Keep ampers setting at max and check for any hot component.
Remove CPU heatsink; more easy to detect the heat on the board this way.

I recommend to be sure first, no solders remain between FDMF pads on the board.