820-00850 PPBUS Short

[hambook]

Hi all, I have an 820-00850 on the bench that's stuck at 4.9V 0.33A on the charger. PPBUS_G3H is shorted to ground on the system side (meter beeps in continuity mode, 0.109 in diode mode on system side with F7000 removed). Measured resistance of all the coils around the CPU and got nothing under 1 Ohm so proceeded to inject 1v into PPBUS_G3H, but nothing gets noticeably warm (used fingers + alcohol to test for hot spots on both sides of the board) and the PSU is pulling under an amp while injecting. Is it safe to go higher in voltage to get it to pull more current, or is there a better way to find the stubborn component?

 

[2informaticos]

For very low values, switch from diode mode to ohm scale.
Post exact resistance to ground at PPBUS_G3H.

The beep in diode mode may be caused by charging filter caps; need to wait 5-10s when measure sometime.
If still beeps after 10s, there is really low value.
Keep in mind, the buzzer beeps until 050-065 (diode mode value), depending of multimeter.

 

[hambook]

Thanks for that info on diode, didn't know that. After about 8 seconds the meter settled on 148 Ohms resistance to ground on PPBUS_G3H

 

[2informaticos]

Did you check at F7000, or L7030?
L7030 is not connected to PPBUS_G3H.

Compare diode mode readings at F7000 with similar board.
Check the reverse way too (black on ground).

 

[hambook]

I got that measurement at F7000. When I was first checking things I removed F7000 and found that with my meter on continuity mode, the side coming from the charger wasn't beeping but the side leading to the system was.

I will compare diode mode readings against a similar board this evening and see what I can find.

 

[hambook]

Have a same model working board on the bench too, used that for measurements to compare. On the working board in diode I get 0.433 with the red on ground and OL with black on ground. On the patient board I get 0.122 for both red and black on ground.

 

[2informaticos]

Check if any big coil gives the same value, in both ways.

 

[hambook]

None of the big coils have the exact same value with with either red or black on ground. L8102 reads 0.102 with red on ground and 0.047 with black on ground. L7430, L7420, L7410, L7210, L7220, and L7270 all read 0.003 with red and black on ground. L7430 and L7420 both read 3.8 ohms in resistance mode. Unfortunately my working test board got picked up today so I don't have a good one to compare against.

 

[2informaticos]

You can try to inject voltage.
Not very low value, nor short, but may have luck and something enter in thermal avalanche.

Just in case, which side of F7000 do you get 0.122 to ground?

 

[hambook]

I will try to inject and see what I can find. I already tried injecting, nothing got warm but I didn't go higher than 1V, will try again 2 or 3 volts and with some thicker cables and see if anything warms up.

The 0.122 to ground is on pin 2 of F7000, system side.

 

[hambook]

Luck was in my favor! It ended up being C6405 and C6406, they just barely heated up when injecting. Removed those and the board boots to the login screen. Will just need to replace from a donor and all should be well. Thanks for the help!

 

[hambook]

Spoke too soon I guess - cleaned the area with alcohol and reassembled, and it worked fine for 1 boot, then when I rebooted the fans were at 100% speed and it was slow. While trying to hold down D and boot into AHT and find out what sensor was causing the issue, the board turned off and now it draws 20v 0.53a and then falls down to 0.02-0.03a where it stays. All 4 ports exhibit this behavior. What the heck happened?

 

[2informaticos]

Any fault history?
The two caps have been damaged by liquid spill?