820-00165 PPBUS-G3H short to ground

Jay

Member
Got a light on the charger but no power. PPBUS_G3H not present. I have 8.6V incoming on F7140 but 0V coming out.
Replaced F7140 and put the board back in the Mac, connected battery and smelt something burning. Upon checking F7140 I had 8.6V going in, 0V coming out. Not sure if it was F7140 that I smelled.

Checked for short to ground on pin 2 of F7140 and got one.

Now I'm injecting power to PPBUS_G3H (on pad 2 of F7140) but nothing gets hot or even warm. Went up to 8V and 2A but nothing.

What would the experts recommend I do/check next?
 

larossmann

Administrator
Staff member
a few possibilities

1) cpu mosfet is shorted and voltage is going to the CPU, cpu still has heatsink on

2) it is one of the caps near the jtag or near the fuse itself

3) it is in the board and you are screwed.
 

Jay

Member
Louis with the good news special! ;)
as this is my first time injecting a board, is there a rule as to how high I can go on V/A? Figured 8.6V was at that point anyway so it’s safe but Amps wise I don’t know.

Set to 8V 2A it only outputs 0.022V, is that enough to warm anything?

Will check cpu without heatsink and the mosfets next. Thanks!
 

Jay

Member
Without heatsink, CPU does not get warm. Checked caps near jtag, don't get warm. No caps near the fuse itself except the ones feeding power into the fuse and those checked out.
 

jadao

New member
You need to increase power (watts) for something to heat up.

You get that by fixing Amps at max and you control gradually voltage.

Example
Max Amp at 1V, Max Amp at 2V, Max Amp at 3V and so on.

Keep in mind this power rail doesn't go more than 8.6V. It probably can handle a little higher but better to stop at the power rail spec.

Now particularly with this power rail there is another risk of toasting cpu. If it's a shorted mosfet cpu, this means it will let pass 8V to cpu!

If nothing heats up, maybe internal short or direct metal short connector that will not let you feel any heat. it's rare to blow a ppbus_g3h fuse, because the fuse is rated 24v 8A it means 192 Watts.
Same as sucking 24A at 8V! If it was a chip or cap at fault I think you will notice it completely burned.
 

Jay

Member
Thanks for the explanation jadao. So Amps = always at max, then I start raising the Voltage (just repeating to be sure). Will try that.
 

jadao

New member
Yes it's one method of finding shorts among others.

When you put Amps at maximum, let say 5A, it doesn't mean 5A are delivered into the circuit, it means 5A are available for the circuit if it needs it, now you will gradually increase the voltage 0.5V, 1V, 1.5V and so on until you feel something heating without exceeding the limit of the power rail shown in the schematic.
When you gradually increase the voltage you should see and monitor amps display on the PS, it will increase depending on what the circuit is taking (we don't know how much current it will take so we just set it to max and control the voltage). Usually you don't need to reach the maximum amp to find the short except when nothing heats up and you are desperate and think the short is internal to the board.

It's just a matter of power Watts=Amps X voltage. For axample, if your shorted component on a 12V power rail start to heat at 3V, 1.5A (3*1.5=4.5Watts) someone else could have the same result with the same shorted component starting to heat with 2.25V and 2A (2.25*2=4.5Watts)

Now for normal operation, if you have a known working device at 12Vdc 0.5A, of course you just set the voltage to 12Vdc and limit the current for good practice to 0.5A but even if you put 5A, if the device is not faulty (shorts) it will not take more than 0.5A.

I really hope I didn't bring more confusion :confused:
 

smiba

New member
Small correction, you need to increase Amperage. Wattage will not heat things up, its the amparage that will.
Always use a as low voltage as possible and only increase the voltage if you can't increase the amps anymore due to the voltage stabilizing
 

smiba

New member
I probably should've made a full reply about this, but I just wanted to add that you should always start with the lowest voltage possible and going above ~1.75V on PPBUS lines can be risky in case its leaking away into the GPU or CPU (Which is probably good for 1/5th of the PPBUS shorts). Otherwise good explanation!
 

jadao

New member
By increasing Current (Amps) you are increasing the power (watts)!

By increasing the voltage (Volts) you are increasing the power (watts)!

Of course it's a matter of power dissipation, of heating energy resulting both from voltage AND amperage.

When you are looking to find a short you are looking to feel the heat so you are looking to create a power dissipation!

Now, how to create this power dissipation that will let you find your short?
By setting a voltage and a current.
 

dukefawks

Administrator
When you started this post there was 8.6V present. Remove F7140 again and see if it comes back, else something got fucked.
 

Jay

Member
Had to replace F7140 again, now get 8.6V on both sides. The unit now boots as long as the ribbon cable between board and DC board is not connected. With it, no boot, without it, boot. Searching for another cable to test but in the mean time someone may have seen this issue before?

EDIT: Other cable didn't help.
 
Last edited:

jadao

New member
Do you have another I/O board to test with?
When the board doesn't boot with I/O board cable plugged in, what is pp3v42_g3h and ppbus_g3h?
Inspect J9500 connector, maybe a signal is pulled down when you plug the cable... Inspect also connector on the I/O board but better to test with a known working board.
 

Jay

Member
Tested with other I/O board, same issue. PPBUS_G3H = 8.6V. PP3V42_G3H = 3.4V. (same values with or without cable) and connectors look good on both ends.
 

Jay

Member
It is but only with the ribbon cable connected. Without the cable the unit boots but magsafe shows no light.
 
Top