Dead graphics card... what was the cause?

VBlackie

Commendable
Sep 24, 2016
5
0
1,510
What happened: I was playing rocket league on primary monitor, and in the second one I was watching a full screen tv series (more like listening... but you get the idea). Then boom! Green screen of dead with horizontal stripes on primary monitor and blue screen of dead on the secondary.

What I tried to fix itI restarted my pc and I had no input signal whatsoever. Tried to change monitor and move around cables but nope... I could only see a blackscreen after the asus logo, no way to enter safe mode or anything. I gave up and re installed Windows (I had a partition specially built in case something like that happened, so I only lost the SO and no other harm was done).

Everything seemed fine, Monitor and tv worked so I guessed nothing bad hapened to them. I installed the video drivers (raedon) and when I tried to enter again to windows the screen turned into purple pixels and green lines then black screen of dead and had to restart.

The only workaround was to uninstall video drivers (in windows safe mode I did it), and then I could see again. So the video drivers, when installed, seemed to make the video card go bad. I tried using latest drivers and latest drivers but to no success whatsoever.

Testing
I tried using a different video card on my pc and it worked, unfortunately I couldn't test the new card drivers.
I tried using my video card on 3 different pc's and there was no output signal on any of the pc's I used. Therefore, I assume the video card is the one that is dead
Finally I used the PSU on other 2 pc's and stressed them but they were pretty much fine so I guess there was no damage on it either.

Diagnostics?
So... the graphics card is dead. It was being overclocked and that shutdown had happened before (like other 5 times but it was not all that severe). I thought it was bad drivers or something.
Could the damage have been spread to the MoBo or the psu?
What happened? Bad drivers? Does rocket league have any kind of problem with double fullscreen monitors? Did I asked to much power and it just blowed up?

Are the other pc components "safe of harm"? To what degree?
Any comment is greatly appreciated. Thanks in advance.

Specs

OS: Windows 10 profesional x64
Mobo: Asus M5A97 R2.0
CPU: Fx-8320
Cooling (CPU): Cooler master Seidon 120mm.
Graphics card: HD 7870 2Gb
SSD: Samsung Evo 850, 500gb
HDD: Seagate barracuda 1TB @7200rpm
Memory: 2x- HyperX blu 8Gb ddr3 @ 1600Mhz
Primary monitor: BENQ RL 2445hm
Secondary monitor: Samsung HDTV 32''
 
Solution
No, don't stick it in the oven yet. Wait until we're finished troubleshooting to make baked graphics card for dinner.

When looking at the PCB of the graphics card, check both sides, probably with a magnifying glass, not for desoldered joints, but for slight discolorations. Usually they are pretty subtle, like a slight brown tinge on an otherwise blue PCB. Also, check to see if any of the capacitors or chokes aren't oddly shaped. Capacitors are cylindrical, and chokes are either rectangular or cubes.

You can try reflashing the GPU bios, but I doubt it'll change something. Any idea on what you want to replace it with?

amtseung

Distinguished
Up until a year ago, I was running a heavily overclocked HD7950. Unbeknownst to me, it was a brewing cauldron of problems, because my DIY water cooling solution was beginning to fail. I reattached the absolutely atrocious stock cooling solution as a temporary fix. A few months down the line, I was monitoring temps on my secondary monitor while playing Planetside 2 on my main monitor, when I heard a sound like popcorn popping, and smoke came out of my system. It shut itself off, but it was too late. I yanked the graphics card out, still smoking, and left it outside to cool for an hour, before completely dissecting it. Upon closer inspection, one of the VRMs had popped, exploding out like popcorn. The graphics card, surprisingly, still works to this day, I just couldn't put it under a load over 80%, or the screen goes black and the PC shuts itself off again. I've since unsoldered the popped VRM from the PCB, and I can now load it at a theoretical 100%, but it'll throttle due to power limit once over 80% load.

Anecdotal tales aside, I would take that graphics card and take a screwdriver to it, completely dissecting it while scrutinizing it for burn marks and discoloration.
 

VBlackie

Commendable
Sep 24, 2016
5
0
1,510
@amtseung Thanks, will do that immediately to check for signs of burn even though my graphics card didnt 'popcorned" or sounded to anything simliar. Who knows what I might find?
 

VBlackie

Commendable
Sep 24, 2016
5
0
1,510
At the inside everything seems perfectly fine... Nothing seems off, or any joint that has been dissoldered. Right now I am thinking about flashing the video card BIOS or trying the oven method. Any bets?
 

amtseung

Distinguished
No, don't stick it in the oven yet. Wait until we're finished troubleshooting to make baked graphics card for dinner.

When looking at the PCB of the graphics card, check both sides, probably with a magnifying glass, not for desoldered joints, but for slight discolorations. Usually they are pretty subtle, like a slight brown tinge on an otherwise blue PCB. Also, check to see if any of the capacitors or chokes aren't oddly shaped. Capacitors are cylindrical, and chokes are either rectangular or cubes.

You can try reflashing the GPU bios, but I doubt it'll change something. Any idea on what you want to replace it with?
 
Solution