Computer shuts off at random

Discussion in 'Hardware' started by TheLinkster, Oct 17, 2004.

  1. TheLinkster

    TheLinkster Private E-2

    Since I installed UT2004 I've been having problems with the computer powering off. I have to use the power button on the back to get it going again. It's been getting worse the last few days and I'm out of ideas.

    Some history: installed UT2004 and it began to power off in the game. Got Prime95 and memtest (memtest.org) to check for memory problems. Prime95 shut off the computer as well. Got all the latest patches, driver updates etc and no difference. Got MBM5 to check for heat problems and found that it would shut off consistently at 67C, which is odd since the COP on the MB is set to 85C. Swapped the CPU fan for a Themaltake Volcano and it made the problem worse. Went back to original fan and system was better. Yes, I installed the heat sink correctly. At the moment, the computer is stable until you do something with it, like burn a CD, enter a form on a web page, play music, etc. I've use HJT, AdAware and Spybot to hunt for problems. Currently SpyBot and online virus scans will shut the system down (tried Symantec and Panda).

    I'm stumped. This is my girlfriend's computer and was built up by her teenager cousin. Any ideas as to what could be causing this? Specs are below:

    WinXP Pro SP2
    ASUS A7V8X-MX MB
    350W Power supply (don't know the make. Came with the case)
    AMD Athlon XP 2400+
    512 MB DDR SDRAM
    ATI RADEON 9200SE (latest drivers from ATI installed)
    SoundMAX integrated audio (latest drivers ASUS installed)
    Maxtor and Seagate HDs
    ASUS 50x CDR
    Iomega ZIP 4x CDRW
     
  2. BeerMonkey

    BeerMonkey Master Sergeant

    92000 Se 256mb 128-bit?
     
  3. TheLinkster

    TheLinkster Private E-2

    Sorry, I should have been clearer. It's an ATI RADEON 9200SE AGP (0x5964) with 128 MB and Bios BK-ATI VER008.015.028.000
     
  4. Doby

    Doby Sergeant

    Hi,

    This is most likely a hardeware problem that just showed up or was there and undetected all along.
    When you ran memtest did you leave it run for at least 3 passes? If it showed any errors then thats one problem.
    I assume that 67c is the cpu temp and that is way to high
    When you swapped to the thermaltake that should have made it better. Did you apply new thermal paste? Peferably a very thin layer of artic silver5.
    What rpm is the fan running at?
    When you do something with the computer like burn a cd it creates more heat and might trigger the bios to shut it off because of overheat.

    You may need to add more case fans
    The power supply is also questionable, a name brand at least 400w is always best with todays newer boards, those that come with cases can be flaky and cause these kinds of problems.

    Rick
     
  5. TheLinkster

    TheLinkster Private E-2

    Thanks for the feedback. I let memtest run for 4 passes and it found no errors. I did use Arctic Silver on both the Thermaltake and when I put the original fan back in. I did notice that the Volcano kept the CPU temp down around 50C, but it shut itself off more often and the noise from the fan running at 6800 rpm was quite loud. I ended up putting the stock fan back in as it's much quieter and wasn't causing a shutdown to occur as often. The BIOS says the fan is running at 3600 RPM and I don't see a way to adjust that.

    I've checked temps and 67C for an AMD is not too hot. That temp would cook a P4 but AMD runs hotter than Intel. The CPU typically runs around 57C, but when stressed it will go higher.

    Earlier today I noticed a faulty power connection on the power bar, so I moved the plug and it's not shutting down as often (it seems), but it does still shut off.
     
  6. Doby

    Doby Sergeant

    The max CPU temp that AMD publishes in the Tech Docs is measured directly on the CPU die itself. The thermal diode is part of the Mobo and "reads" the radiant temp from the CPU. The BIOS/software readouts use the thermal diode and these temps are typically 10-15C lower than the actual on die temp. Most AMD CPUs have a max operating temp between 75C-85C measured directly on the CPU die, which is the point where the CPU will start to self-destruct. If a CPU was running at 75C on die temp, the thermal diode and BIOS/software readout would be about 60C, which is [content edited] HOT !

    In my experience, CPUs run at temps above 52C indicated by BIOS/software readouts can start to experience system hangs or crashes. Some folks have not experience problems with BIOS/software temps up to 60C. I personally would add fans or do whatever is required to keep the max temp under full load below 52C to insure 100% stability,
     
  7. TheLinkster

    TheLinkster Private E-2

    That's good information to know! I had no idea of the temp difference in readouts. I'll do something about the cooling problem and see what happens.

    Thanks a bunch!
     
  8. Kodo

    Kodo SNATCHSQUATCH

    Try some arctic silver 5 for thermal paste. It brought my temps down a total of 8 degrees.
     

MajorGeeks.Com Menu

Downloads All In One Tweaks \ Android \ Anti-Malware \ Anti-Virus \ Appearance \ Backup \ Browsers \ CD\DVD\Blu-Ray \ Covert Ops \ Drive Utilities \ Drivers \ Graphics \ Internet Tools \ Multimedia \ Networking \ Office Tools \ PC Games \ System Tools \ Mac/Apple/Ipad Downloads

Other News: Top Downloads \ News (Tech) \ Off Base (Other Websites News) \ Way Off Base (Offbeat Stories and Pics)

Social: Facebook \ YouTube \ Twitter \ Tumblr \ Pintrest \ RSS Feeds