Computer turns self off

By Nerdwing
Jul 31, 2010
Post New Reply
  1. Hello there.

    I've had a problem lately with my computer turning itself off at seemingly random intervals. I've had it suggested to me that my problem may be Mobo, heat-related, GPU, or funky RAM. Having looked at each of these issues in detail, I cant single out a possible cause for my situation. I can go between five minutes and two days without a turn-off, its truly random.

    My CPU: AMD Athlon 64 X2 Processor 3800+

    my Video card: Nvidia 7950 GT

    2 Gigs of ram in four 512 MB sticks.

    My psu is a Dynex 400-Watt ATX

    All drivers up to date.

    I was told by a friend that it sounds like graphics card failure. However, I have noticed no artifacts or other hints that my video card may be failing.

    I investigated each ram stick, and reseated it.

    My GPU is running at a modest 68 celsius under load.

    And overall, my case temperatures are well within safety shutoff limits. The shutoffs seem very random in nature as mentioned, and occur during anything from opening a Youtube video to a crash during Starcraft 2 installation.

    I'll be glad to give any more information as needed, I've had this problem for about two weeks now. Thank you in advance :)
  2. Archean

    Archean TechSpot Paladin Posts: 5,681   +86

    Have you had a look at the Event Viewer Logs? May be they can offer some clues.
  3. Nerdwing

    Nerdwing TS Rookie Topic Starter

    Yessir I have. I just experienced a crash upon closing Starcraft 2 and before fully returning to desktop. The computer just turned itself off.. It occurred at roughly 11:11 PM, but the Event viewer logs noted nothing peculiar at all having taken place when I checked them after reboot at 11:12 PM. Its quite baffling :(
  4. Archean

    Archean TechSpot Paladin Posts: 5,681   +86

    Please download OCCT, and stress test your graphic card / and its memory with it, to ensure it is not the cause of all this.
  5. Nerdwing

    Nerdwing TS Rookie Topic Starter

    Thank you very much for the link, this seems like a useful program. However, when I try and test my GPU Memory I am immediatley met with a "CUDAMemTest.exe has encountered a problem and needs to close" window.

    Could this error in itself be indicative of the problem?
  6. Archean

    Archean TechSpot Paladin Posts: 5,681   +86

    Probably it may be hardware issue, if you have another graphic card can you try it and see how your system behaves with it.
  7. Nerdwing

    Nerdwing TS Rookie Topic Starter

    The only alternative I have is a Geforce 6100, which is my integrated graphics card.

    I'm going to try and find a blank CD somewhere around the house with which to use as a boot-disk for Memtest86. Maybe that will show something irregular.
  8. Archean

    Archean TechSpot Paladin Posts: 5,681   +86

    If the onboard graphic solution gives you no errors/Bsods, then there are few other avenues to check. 7950GT required 18A on 12V rails (with minimum 400W PSU); so one possibility which comes to mind is degradation of your PSU. I am unsure about quality of Dyex PSUs, I hope someone with bit information about these will have a look at this thread and guide you in this regard. If you haven't cleaned out your system in a while please do that as well to get rid of any dust inside the chaises.
  9. Nerdwing

    Nerdwing TS Rookie Topic Starter

    My PSU is brand new, and I experienced this problem originally the very day I switched the old one out. Upon replacing the new one with the old one again, I still experience the problem. I am wondering if physical damage may be a problem, as I am not very dexterous and may have unintentionally damaged something without immediately knowing.
  10. Archean

    Archean TechSpot Paladin Posts: 5,681   +86

    Have you tried working through the problem by switching to onboard graphics and pulling out your nVidia card ? If, so what are the results?
  11. Leeky

    Leeky TS Evangelist Posts: 3,797   +116

    If your PSU is new, I would highly recommend you follow Archean's advice, and remove your dedicated graphics card, and try running your PC on its integrated one for a couple of days. If the problem goes away you know the graphics card in one way or another is a problem.

    Since you said you'd changed the power supply, does your graphics card have a power connector on it? If it does, are you sure you've refitted the power connector having fitted your new PSU?
  12. Nerdwing

    Nerdwing TS Rookie Topic Starter

    I've switched to the Integrated graphics card just now (Geforce 6100) and still experience the problem apparently. I just had a shutoff while running League of Legends after about the same amount of time as with the dedicated card.
  13. Archean

    Archean TechSpot Paladin Posts: 5,681   +86

    Hmm, what the temperatures of your CPU and moterhboard zones? Is your casing have good airflow?

    Also closely inspect your motherboard/RAM, for any dust, or carbon, or any other unusual stuff like bulging capacitors etc.

    I am assuming here all the required motherboard power connectors are perfectly seated.
  14. Leeky

    Leeky TS Evangelist Posts: 3,797   +116

    Following on from this, is the ATX 12v 4 pin square connector near the top of the motherboard connected and fully seated home?
  15. Nerdwing

    Nerdwing TS Rookie Topic Starter

    The only 4 pin square connector that my PSU provides is plugged into the lower part of the motherboard, adjacent to the silver battery (CMOS I think its called? Not entirely sure). It is properly seated to my knowledge.

    Using the CPUID HWmonitor.exe, my CPU temperatures read at 30C and 33C as I type this. I'm not sure how to translate the motherboard temperatures, but currently my "TMPIN0, TMPIN1, and TMPIN2" are running at 26C, 28C, and 51C respectively if thats of any use.

    EDIT: I realized that there is indeed a second 4 pin adapter branching off my main 20 pin adapter. It is currently seated directly adjacent to the 20 pin, entirely filling the slot provided. I'd like to note however that my previous PSU didnt have that additional 4 pin adapter, so the 20 pin was the sole occupier of that slot on my motherboard.

    SECOND EDIT: Using HWInfo32.exe, I notice that for the "Motherboard" entry I have nothing there. Its an empty field. Could this be a problem?

Similar Topics

Add New Comment

You need to be a member to leave a comment. Join thousands of tech enthusiasts and participate.
TechSpot Account You may also...