• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

0.95v mod - bricked card?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

RothFTP

New Member
Joined
Oct 1, 2017
Hello everyone,

THE SUCCESSFUL PART
I attempted the notorious 0.95v rail mod on my Sapphire R9 290, which seemed to work ok, initially. The PC booted successfully, got into Windows, run some games and furmark for a while.

OVERCLOCK FAIL
Attempted to overclock. In MSI Afterburner I maximized AUX voltage (+100) and Memory clock (1625) and raised the voltage to +200mv, raised the core to the point, where furmark froze. The rail measurement was being a steady 0.97v to that point. Then I tried adjusting the variable resistor which I soldered for the mod. After some adjustment I got a frozen screen, I restarted and since then I was getting frozen, blue, or black screens in windows and even distorted image in bios.
The rail measurements were fluctuating between 0.97v to 2v+, after starting Windows normally, steady 0.97v in safe mode. For as long as the measurement was lower than 1v it was ok, but when it was reaching 2v, blue/frozen/black screen.

UPDATE 1 - BLACK SCREEN ALWAYS
1) I disassembled the card and removed the soldered wires. No screen at all this time - all black, always. Windows do boot, though (typical startup sound effects of connecting devices play)
2) I resoldered the VR, tried adjusting it, still no screen.

UPDATE 2 - WORKED IN THE x4 PCI-E SLOT
1) I tried connecting the card to the other PCI Express slot (4x) and it worked normally. Run games, furmark everything ok, except of course for the lower performance, due to the slot bandwidth.
2) Tried connecting the card to the x16 slot again, got to windows, then got a black screen. Since then there is always black screen, even during post, windows boot.

NO ADAPTER FOUND
1) Tried adjusting the variable resistor - no effect, always black screen.
2) Used another card in the x16 PCI-E slot and the R9 290 in the x4 one. Atiflash -i now shows "no adapter found".

Any way of diagnosing what went wrong and a possible way to fix the issue?

Thank you in advance

DSC09870 (Medium).JPG
DSC09872 (Medium).JPG
DSC09875 (Medium).JPG
DSC09876 (Medium).JPG
DSC09877 (Medium).JPG
 
Last edited:
DSC09879 (Medium).JPG
DSC09884 (Medium).JPG

The aftermath...
Other than ugly, is there something that could be to blame?
 
Your last pic says it all.
3 legs appear to be shorted together. That component is junk.
Probably the entire card now as well.
 
Before installing back the card I checked those legs with each other for continuity and seemingly there was none (the polymeter didn't beep). Besides, being as it is above, the card worked fine today for hours in the x4 pci-e slot. Not sure why after I plugged it in the x16 one it doesn't work in either of the slots, though.
Is there any way to confirm that the component is working or not?
Is it also possible to solder another similar component from another video card? I have two 3 video cards lying around (2x 8800gt and an old ati something) and 3 dead motherboards.

- - - Updated - - -

Paste on the ram... yikes... its a mess, that card.

I'm using the kryographics Hawaii R9 290X/290 waterblock. I am supposed to use paste on the rams.
I'm using Arctic MX4. Thankfully it's not conductive, so other than aesthetics it should have no effect on the function of the card.
Right now looks is the least I'm concerned with.
 
Last edited:
I'm using the kryographics Hawaii R9 290X/290 waterblock. I am supposed to use paste on the rams.
I'm using Arctic MX4. Thankfully it's not conductive, so other than aesthetics it should have no effect on the function of the card.
Right now looks is the least I'm concerned with.


Wow, that's interesting. I've never seen a waterblock manufacturer use thermal grease instead of pads on the mem.
 
There are some resistors in that last pic that don't look real happy, either.

I made a mess :( The soldering iron I used was enormous for this project plus my hands were shaking from stress not to damage anything, and so I damaged everything.
The resistors on the right of this 8-pin component (chip?) measure 195/146 ohms, depending on which polymeter pin I attach to each side of the resistor. The lone resistor under the chip shows 1648/548 ohms.
 
The near end of C46A looks chunked. I'm tempted to declare your card kaput. If you have the appropriate replacement parts, and a smaller soldering iron, you have nothing to lose trying to repair it. Just don't spend a lot of your GPU fund on a new iron. Seriously. But if you get it up and running PLEASE make a thread. It would be epic. :thup:

Edit: And no covfefe tis time.... :D
 
Which one is this C46A? How can I know check if it's working and under which criteria am I supposed to choose a replacement? I have three spare gpus and three mobos that I can scavenge parts from.
I don't understand, though how did it work yesterday in the x4 pci-e slot and now nothing. The components were looking as they do now and the card worked flawlessly for 3 hours, playing games and running furmark. With that in mind I can't give up on it just yet.

PS: just looked up "covfefe" - the more you know...
 
C46A shows the same ohm measurement (195/145 ohms) as the others on the right of it, does that mean it's not destroyed? Is it possible to replace the IC with another one from those 8800GTs I have, or from those mainboards? I can see quite a few similar sized 8-legged ICs. How can I know which one is fit as a replacement?
Another thing; in my attempt before to make sure that the IC's legs are not shorting with each other, I digged between them with the multi's pin, scratching the surface of the pcb. Now there's an obvious copper mass underneath. Something to do for that?
 
Any idea what could be done for the scratched pcb between the IC's legs?
 
After looking at that pic half a dozen times. I almost think one of those two legs on the low side of that eight pin are compromised . That could explain why it kind of worked and then just failed. I would look there first. If you have a fine tip just a bit of heat and a tad of solder there might reconnect them. Put some flux on it first but stay away from that bare copper
 
After looking at that pic half a dozen times. I almost think one of those two legs on the low side of that eight pin are compromised . That could explain why it kind of worked and then just failed. I would look there first. If you have a fine tip just a bit of heat and a tad of solder there might reconnect them. Put some flux on it first but stay away from that bare copper

Me thinks there's too much crud/dirt in that area that needs to be cleaned out too.
 
Back