GTX 570 - Stops PC from powering up

[dicky96]

So I have this GTX 570 The original fault was resistance to ground from Vram of just 10 ohms. Vcore was 1 ohm.

I traced that problem to the VRM controller IC for Vram (had a hole blown in it) and also the two VRAM phases had high side mosfet short to gate

So I removed the Mosfets and controller and Vram now had 60 ohm to ground, which seems reasonable enough

Understanding that I could have a faulty GPU and/or RAM due to the high side mosfet short, I decided to replace the APW7088 control IC plus the Highside (K03B9) and low Side (K0382) mosfets anyway

I checked all resistances were OK, still had 60 ohm Vram to ground, no short from 12V to Vram - so I powered it up

I found I had Vram of 1.5V.... for about 5 seconds then the PC powered down

Now I can't get the PC to power up with the graphics card fitted. I don't see any shorts - Vcore is reading 0.8-1R

I also noticed if I remove graphics card from the PCIe slot but leave the 12V 6 pin cables connected to it, then the PC will power up

If I plug the graphics card into the PCIe without the 12v cables attached the PC will not power up. The CPU fan spins for less than a second then stops)

If I put another graphics card in the PC powers up

I checked the 12V and 5V from the GTX 570 PCIe connector to ground. I have 60R (12V pins) and around 190R (3.3V pin)

So there are no shorts that I can see - but it still stops the PC from powering up

[dicky96]

Re: GTX 570 - Stops PC from powering up

You can see what happened - here

It looks like the resistance on 3.3V and 12V PCIe did change after the PC shut down - but still no shorts, the resistance on 12V PCIe went down to 60R and the resistance on PCIe 3.3V went up
https://youtu.be/vk1bga4dyWc (short 4 mins video)

[piernov]

Re: GTX 570 - Stops PC from powering up

So I removed the Mosfets and controller and Vram now had 60 ohm to ground, which seems reasonable enough

I checked the 12V and 5V from the GTX 570 PCIe connector to ground. I have 60R (12V pins) and around 190R (3.3V pin)

Sounds like a hint to what's gone wrong…

[dicky96]

Re: GTX 570 - Stops PC from powering up

Yeah it does - like something changed since the PC powered down - but neither 12V or 3.3V from PCIe has a particularly low resistance AKA a short

Unless you are suggesting the 60R I now see on PCIe 12V and the 60R I could previously see on Vram are not a coincidence? In which case I may know where you are going with this.......

I must recheck that tomorrow though as I took the last resistance readings and posted the OP quickly just before I left the workshop. I have the original resistance readings from the video (there is more of it than I linked here)

[dicky96]

Re: GTX 570 - Stops PC from powering up

It's like you supposed Piernov. So the High side Mosfets on Vram went short again after a few seconds - hence the 60R now being visible from 12V PCIe to ground

The APW7088 has not visibly blown this time but it doesn't necessarily mean it is still any good. It could be that the K03B9 Mosfets I got from Alexpress were fakes but seeing as the original High side Mosfets had gone short (and taken the APW7088 with them) it would seem more likely there is an underlying reason for this

I could replace the K03B9 again, and even the APW7088, a I bought several of each as they were cheap. The 60R I see on Vram to ground would not explain a lot of current flowing at 1.5V (like it would only account for 25mA), though I guess the resistance of the load is much lower once the GPU and RAM are running?

My guess is that the GPU/RAM are drawing more current than they should - hence the VRM went pop after some seconds.

Is there any point in going further with this repair? I could try replacing the FETs and then powering the VRM from my bench PSU with current limit - to see how much current is drawn on 12V, and assuming an efficiency of say 90% for a synchronous buck regulator it should give me a good idea of the current on Vram 1.5V. But apart from being quite interesting would this tell me anything really, and if Vram a drawing a low current what would that say about why the FETs went pop?

Also has anyone here got one of these and how useful are they as a fault diagnostic tool? I'm seriously considering buying one

https://eu.evga.com/products/product...100-UV-0600-BR

https://www.ebay.co.uk/itm/133501345...oAAOSwYBFfQ0eQ

Rich

[jeffwae]

Re: GTX 570 - Stops PC from powering up

you must find schematic board of this 570 1st , then do on viewboard, and you can find bad componens,you can remove and place new part