gpu glitches after using gpu fan adapter cable and cablemod extensions

Discussion in 'Troubleshooting Help' started by shN88, Nov 6, 2018.

  1. shN88

    shN88 Member

    Joined:
    Dec 30, 2014
    Messages:
    43
    hi OCAU,

    so i have a 1080ti with an AIO stuck to it using a G10 and kraken x61. the system was working fine until today when i installed a gpu fan adapter cable, pwm y splitter and cablemod extensions for my power cables.

    my plan was to use the gpu fan adapter to plug my two 140mm radiator fans into the gpu so it can change speed based on gpu temps. i also added cablemod extensions to make the pc look nicer and neater.

    once i installed everything my screen was black, after multiple restarts, sometimes i could see the bios splash screen and my windows 10 loading screen but it would either freeze/restart/screen go black when it hit the user login page. I plugged my displayport cable into my mobo directly and no issues. I unplugged the radiator fans from the gpu (using adapter and splitter) and plugged them back into the psu/mobo. still the same issue.

    i plugged my displayport cable back into my gpu and this time it randomly boots up and im able to get into windows. however ive noticed some graphic glitches. i opened hwinfo to see if my temps were ok and they are. then my screen momentarily froze, then glitched a bit. i have taken photos of the glitches. after it glitched the GPU power values in HWinfo is greyed out and not moving.

    purple squiggles on the NZXT Cam App.

    [​IMG]


    [​IMG]


    does anyone know whats going on? as soon as i booted in and saw those glitches i made this post. im not using the gpu fan adapters or splitters, but i am using the cablemods.

    did i rek my card? is there something wrong with the cablemods? did plugging my radiator fans into my gpu fry something?

    will update thread as i try to isolate the problem
     
  2. OP
    OP
    shN88

    shN88 Member

    Joined:
    Dec 30, 2014
    Messages:
    43
    tried to run timespy, same purple squiggles appear before freezing and failing.

    in windows it runs ok then occasionally i see artifacts, screen momentarily freezes and refreshes.

    Also, ive noticed that in HWinfo, it says Performance Limit Utilization is 'yes'. when i am sitting idle in windows. is this normal?

    ive attached a screenshot of my sensors. does anything seem out of place?

    [​IMG]
     
    Last edited: Nov 6, 2018
  3. zwof

    zwof Member

    Joined:
    Oct 30, 2006
    Messages:
    15
    Location:
    Sunshine Coast QLD
    Have you tried removing the cablemod extensions and restoring the cables to how they were before the issues started?
    Make sure all the cables to/from the GPU are plugged in correctly? Unplug everything from the card, check the pins/plugs, and then plug it all back in securely?

    Isolate anything that has changed, check and verify connections.
     
    Last edited: Nov 6, 2018
  4. OP
    OP
    shN88

    shN88 Member

    Joined:
    Dec 30, 2014
    Messages:
    43
    ive tried without the gpu cablemods and i am getting the same problem.

    ive attached a screenshot of hwinfo recordings. the green box is my current test which fails right after it goes under load. (dark green). the blue box is a previous benchmark ive done recording with hwinfo. the dark blue is when it is under load. the difference seems to be that my GPU power doesnt go past 65.466watts and my FPS is stuck at 10fps.

    [​IMG]
     
  5. OP
    OP
    shN88

    shN88 Member

    Joined:
    Dec 30, 2014
    Messages:
    43
    so i took off the AIO and put back on the stock cooler. it works. at this point im assuming that the pump is dead. im a little puzzled though because it was sitting 36-37deg at idle. i would have thought it would be hotter.
     
  6. zwof

    zwof Member

    Joined:
    Oct 30, 2006
    Messages:
    15
    Location:
    Sunshine Coast QLD
    Sounds like there wasn't good contact between the AIO and GPU causing either core, memory or VRM to overheat. Are you using additional heatsinks on the mem/vrm?
    The temps you are getting are indeed unusual, is the temperatures you've listed the liquid temp in the X61 or the actual GPU temp?
    The G10 doesn't list the 1080Ti as being compatible (only the 1080), the G12 does though. Maybe the G10 doesn't make contact correctly with the different size 1080Ti die (though it does seem to work according to Google)?
    Again, speculation as I've never used a G10/G12.
     
    Last edited: Nov 7, 2018 at 1:16 PM
  7. OP
    OP
    shN88

    shN88 Member

    Joined:
    Dec 30, 2014
    Messages:
    43
    maybe it was poor contact. I reseated it 3 times and all had the same problem so i find it hard to believe. however today, after checking that the AIO pumping is indeed working, I put it back on the 1080ti and now its working fine again. I've no idea what the issue was.

    edit: no vrm heatsink, jayztwocents did a video on the g10/g12 bracket and the single fan is sufficient enough to show improved cooling over stock. so i dont think its the VRM heating. 1080ti fits fine with my current setup, ive read other people online with the same setup no issues. im thinking now that maybe i overtightened it.
     
    Last edited: Nov 7, 2018 at 2:09 PM

Share This Page