820-00239 drawing 5v from charger short on ppdcin_g3h

Gannons3

New member
I have a 13 inch touchbar that is drawing 5v from the charger, I have a short to ground on ppdcin_g3h, I have removed all the cd3215s all the caps on that line and even all the transistors (Q3200 ect) short remains, its a 1 ohm short and I cannot find it for the life of me
 

2informaticos

Administrator
Staff member
You should remove R7020 first.
First suspects are Q7030, U7000, or one of the caps from PPDCIN_G3H_CHGR line.
Check resistance to ground at L7030 and F7000.
 

Gannons3

New member
removed r7020, q7030 and u7000 also removed all caps on ppdcin_g3h_chgr still have 1 ohm to ground on ppdcin_g3h
1 ohm to ground on f7000
16k ohms to ground on pin 1 L7030 (tba_phase1)
456k ohms to ground on pin 2 of L7030 (tba_phase2)
 
Last edited:

2informaticos

Administrator
Staff member
"removed all caps on ppdcin_g3h_chgr"
Removed why, if the short is before R7020???

"16k ohms to ground on pin 1 L7030 (tba_phase1)
456k ohms to ground on pin 2 of L7030 (tba_phase2)"
Impossible!
Unless L7030 is open, which I doubt.
Check in diode mode, red probe to ground. Only switch to ohm mode, for very low value (i.e. less than 010).

Please, post accurate information, or our replies will not serve for anything...

"1 ohm to ground on f7000"
This is bigger problem, as it can be caused by shorted MOSFET in one CPU power spupply.
Check if any big coil has the same value.
Next check between PPBUS_G3H and that coil; ohm mode this time.
If you get less than 1 ohm, there is a shorted high-side MOSFET on that coil.
If get 2 ohms, check other coils.

On PPDCIN_G3H you can try to inject tension.
Start with 1V and go up, maintaining ampers bellow current limit of your bench power supply.
 

Gannons3

New member
I was removing caps before i took off r7020 that was my bad
L7030 is around 500k ohms to ground on both sides im not sure why I got that messed up reading the first time
im getting about 1.5 ohms to ground on f7000, all big coils around the cpu are between 2-5 ohms to ground
I dont have a bench power supply so im unable to inject voltage
 

2informaticos

Administrator
Staff member
If you plan to stay in this business, get 30V/10A adjustable power supply.
You don't need expensive one; digital model not needed.

Do NOT inject on the PPBUS_G3H before doing above mentioned tests...
 

Gannons3

New member
im getting 73k ohms between ppbus_g3h and the coil.
is there anything else i can check? or is my next step having to inject voltage?
 

2informaticos

Administrator
Staff member
You can safely inject voltage on PPDCIN_G3H at this moment.
Always use the same method.
"Start with 1V and go up, maintaining ampers bellow current limit of your bench power supply."

No reason to go for PPBUS_G3H if you cannot solve PPDCIN_G3H problem first...
 

Gannons3

New member
i injected at pin 8 of q3200 on ppdcin_g3h, i used the alcohol method and the only place getting hot was the ppbus_g3h cap CB701
 

2informaticos

Administrator
Staff member
Looks like both power rails meet inside the board; very strange.
Did you check CB701 out of board?
Does the short disappear from any rail, after removed it?
 

Gannons3

New member
yeah that cap is gone theres a litteral hole in the board under where the ppbus_g3h pad was, it goes all the way through to the other side, right on the other side of the board is another cap on that same line, i can see through the board now with both caps off. Both lines are still shorted to ground 1.7 ohms to ground
 
Last edited:

Gannons3

New member
also now with both caps removed its still drawing 5a at 2v, but Ive covered the whole board in alcohol and i didnt find a single hot spot anywhere
 

2informaticos

Administrator
Staff member
The last option you have, try to clean internal border of that hole.
Use a tiny drill and polish the interior.
Pray that no signal line pass in that area.
 

Gannons3

New member
i tried drilling the hole, inside the bigger hole i can see where the layers are visibly black and there is cavity, in order to drill out the cavity the hole would impact surrounding components. this board is dead, thanks for the help with everything
 
Top