XFX R9 380 No Ram Voltage.

[saronno]

Hi, I am trying to repair this R9 380 by XFX.

Fans works, no video .... but it's not the gpu ... there is no voltage
on the ram.

Looking at the board, there are two APW8722 near the two PCI-e power connector.

Checking the datasheet, there are two interesting pins: BOOT and COMP.
If COMP is pulled down, the IC shut down.
The BOOT honestly I am not sure, but it should give a voltage bump at
the beginning to initiate the voltage regulation.
VCC is presente. The two MOSFET connected to the IC are good.
There are not evident shorts checking here and there with the multimeter .. and also checking with a finger if something is hot results in nothing obvious.

Any idea? I was guessing the APW8722 was faulty but since both of them have the same behaviour I am inclined to think there is something else who stop them to regulated the voltage, maybe pulling down COMP ...

At the beginning I also guessed there was something wrong with the bios but I check the chip with my ch341 and the bios was ok ... but there is no voltage to the chip when you star the board !!

[SMDFlea]

Check the voltage on the gate of 2N7002 (Q3 in the Typical Application Circuit in the datasheet) when you power up .

[saronno]

.

[saronno]

Finally I have some spare time to keep analyzing this board.
It seems the problem has nothing to do with the APW8722.
I check the pci-e voltage and both of them are presente (12v and 3.3v).
But GS7103 near the eprom has a very low VCC. So I find what component
is responsible for generating VCC .... and it's a adjustable voltage regulator (bm1117).
But the Vin there is very low too.
Anyway there are no low-impedence paths on the Vin and Vout line .... but nearby there is this little component who looks very suspicious.
One pin is attached to bm1117 Vin the other one is attached to 12v line (present).

Since the Vin should be more or less in that range (20v max) and Vout should be in the range 3.3v-5.0v (GS7103 VCC) I think think the problem is that little component ...
it's a fuse or something else (it's not open and it's not short ....)

Are you able to identify it?

[piernov]

I see a 0 ohm resistor on a picture found on the web. Seems reasonable. On the board it says "R401" so it's supposed to be a resistor.
What you have on your board doesn't look like a 0 ohm resistor at all, and looks broken. Did someone work on this before and put the wrong component?

[saronno]

Originally Posted by piernov

I see a 0 ohm resistor on a picture found on the web. Seems reasonable. On the board it says "R401" so it's supposed to be a resistor.
What you have on your board doesn't look like a 0 ohm resistor at all, and looks broken. Did someone work on this before and put the wrong component?

Exactly .. I agree with you but it would be better to have a confirmation before making any damage as you said.

[saronno]

This is from a radeon rx470 ....
http://nekosoftpc.blogspot.com/2019/...-rx470-8g.html



It look more or less the same ... a 0 ohm resistor is a good guess.

[saronno]

Done. Put a 0 ohm resistor in the card ... everything fine.
The only problem is ... it will be enough? It is smaller than the original ... who
knows ... time to test and see.

[DrakeFrst]

I'd strongly advise to give it a good run and test for a day or two and if everything is fine to not overuse it, don't let it run for days straight and don't mine on it obviously. I've had a similar issue with my older gfx card.