820-3332-A SHORT ON PPBUS_G3H SYSTEM SIDE

[2click]

Hi, I have this 820-3332 and it wasn't showing video but it was turning on because the customer used it via Teamviewer. Anyway I opened it and it was dirty as hell... began to read about this motherboard and found that I should change the solder on U8900. Did that, turned it on, and that generated a short on SYSTEM SIDE. I´ve been looking to find the short, and found that L7630 has continuity to ground. What is the best way to find the short using the multimeter? when I connected it to my PSU it used .280 ma. Thanks for your help!

 

[2informaticos]

"change the solder on U8900 "
Did you remove U8900 and place it back again?
You should only resolder it with leaded solder, without moving the chip.
Probably some short between pins, or some pin left unsoldered.
Is also possible to have a moved/knocked off component close to U8900.

However, the short on PPBUS_G3H should be generated by Q8930, not Q7630.
Compare exact ohm value on PPBUS_G3H with L7630 and L8930/60.

 

[2click]

First I did change the solder on the U8900, turned it on, but still had no image. Then it just turned off and I started seeing .280 aha at the PSU. I removed the F7040 and F7041 to see where the short was... I thought maybe by removing the U8900 the short might go away, but still there...
I removed the short on PPBUS_G3H, but the short on L7630 is still there, after checking various componentes, I decided to remove C1682 and C1683, the short is still there on PIN 1 of each. Does that means that the CPU is on short? Did I did something wrong to make it short? I never injected voltage I just did the solder change on U8900 then, turn it on, and then it went on short... Please Help!!

 

[2informaticos]

"I thought maybe by removing the U8900 the short might go away "
Did you power on the board without U8900???
In such case, Q8930 could get burnt one section and pulls down PPBUS_G3H to GPU VCORE.

L7630 can have low value to ground as normal.
"Compare exact ohm value on PPBUS_G3H with L7630 and L8930/60. "
???

 

[2click]

I haven't turned it on without any component, I did put everything back and plug it in and have .129 mha consumption. Im sorry the values from PIN 2 From F7040:
L7630= 1.3 kohms
L8930= 20 kohms
L8960= 18 kohms

what I did is try to find the short using diode mode with the continuity beep. That way, I saw what was connected to L7630, and started following the paths... but never turned it on without a component.

 

[2informaticos]

I don't understand how did you measure; cannot have kohms to ground on the mentioned points.
Use diode mode, red probe connected to ground; touch test point with black probe.

 

[2click]

Red probe on ground, black probe on pins.... diode mode " -->|- " (is that right?)
L7630= 0.004 v
L8930= 0.007 v
L8960= 0.007 v

 

[2informaticos]

Forget the "v", when you post diode mode readings.

L8930/60 look good.
And probably L7630 too; I don't have such board to confirm that 100%.
You didn't post the value on PPBUS_G3H (L7030 to ground)...

 

[2click]

Red Probe Ground black probe pins on L7030
PIN 1 = 310
PIN 2 = 320

 

[2informaticos]

Doesn't look to have any short then.
Which voltages do you get on the board now?
First be sure all the solders are good at U8900...

 

[2click]

I will be checking that, but now I have a consumption of .180 mha on my PSU....

 

[2informaticos]

Use original 85W Magsafe charger and ignore that consumption indication.
We need voltage level for signals and power rails of the board in question...

 

[2click]

I connected it to its original Magsafe 85w, its not turning on, when the battery is disconnected when plugin in the magsafe it goes orange-green-orange... when the battery is connected the magsafe goes green-orange like it should... should I measure the voltages using the original magsafe charger?

 

[2click]

Is it normal to have continuity to ground on this line: PPCPUVCCIO_S0_REG ??

 

[2informaticos]

Yes, use 85W Magsafe charger for test.

CPU power lines have low value to ground as normal.
Multimeter's buzzer beeps until 050-075, depending of model; not all beeps means short!

 

[2click]

So I have pretty much every 5v, 3v, 3.3v, rails... I am missing one that is created by the U7600:
PPCPUVCCIO_S0_REG

Voltages around U7600
13 = 5
14 = 5
3 = 3.3
6 = 0
10 = 0
4 = .5
7 = 0
9 = 0
12 = 4.74
11 = 0
15 = 0
8 = 0
2 = 0

 

[2informaticos]

I'm not sure if you check all coils.
Do you have voltage at L7510/20/30, but not L7630???
Compare ohm value (not diode mode) to ground at L7630 with working board.
Be sure all surrounding components and traces are good; pay atention to R/C7630.

 

[2click]

Do you have voltage at L7510/20/30, but not L7630???
I don´t have voltage on any of those...
Compare ohm value (not diode mode) to ground at L7630 with working board.
Working board 20 ohms, my board 1.2 ohms
Be sure all surrounding components and traces are good; pay atention to R/C7630.
Everything looks good...

 

[2informaticos]

"Working board 20 ohms, my board 1.2 ohms "
Looks like dead CPU ...