8

I have a home built Windows 7 x64 computer with a Core 2 Quad CPU. It has recently taken to suddenly shutting down/turning off at unusual intervals. I should note that it powers down suddenly, and does not do a blue screen or shut down. It seems to be most common when playing media. I have tried running SpeedFan, and the CPU temp seems to hover around 49C. The computer has been running fine for over 1 year. Could heat be my problem, and how should I address it?


Steps Taken: (Problem Resolved)

  1. Completely cleaned the inside.
  2. Reseated the heatsink.
  3. Ran memtest 86, which found one memory error, which was followed sometime later by the computer switching off (which leads me to believe it's not the memory).
  4. Installed a brand new power supply.
  5. Inspected all fans to ensure they are running.
  6. Switched out memory sticks. This is not the problem, unless they're all bad.
  7. Reseated hard drive cable.
  8. Tried with a different video card.
  9. Tried the CPU in a friend's motherboard. Same exact problem occurred there.
  10. Applied for a warranty replacement.
  11. Replaced the CPU, seems to be working now.

Just as a side note, the computer is still running fine, it was definitely the CPU.

7
  • 1
    Is it just turning off as if someone flipped the power switch, are you getting a BSOD, or is it causing a Windows "proper" shutdown as if you told the computer to shut down in the start menu?
    – wag2639
    May 23, 2010 at 22:06
  • @wag The former, as in the power switch.
    – C. Ross
    May 24, 2010 at 13:34
  • @C. Ross - Why would you think that the memory error isn't the problem? You should have zero memory errors in a properly working system.
    – MDMarra
    May 27, 2010 at 22:57
  • @MarkM I've tried all the memory separately with the same results. I'm fairly certain ALL the memory isn't bad.
    – C. Ross
    May 28, 2010 at 1:26
  • @C. Ross, that would point to a motherboard or PSU issue then, and since you have ruled out the PSU, it seems the motherboard is your most likely culprit.
    – MDMarra
    May 28, 2010 at 1:35

6 Answers 6

11

Replace the power supply and stress test your computer.

Reseat (or replace) your heatsink and reapply thermal paste.

Run +memtest on the RAM.

Check all cabling; reseat the SATA/IDE cables to your hard drives.

Clean out the dust as best you can.

--

49 isn't that high of a temp. What heatsink are you using?

5
  • Couldn't agree with you more
    – anon31097
    May 21, 2010 at 2:50
  • Stock Intel Heatsink that came with the CPU. Arctic Silver paste. Already removed all the dust. My GPU is running quite a bit hotter (70C). I tried to run memtest, but it died before I could get it up, I'll try again tonight.
    – C. Ross
    May 21, 2010 at 17:05
  • I'm really leaning toward the PSU, and I have one on order. Went ahead and ordered a heatsink upgrade just to be safe (they're cheap).
    – C. Ross
    May 24, 2010 at 13:38
  • 2
    I've found nearly all spontaneous shutdowns to be from heat issues. Have had a couple of loose cards, hard drives, and one bad memory stick cause it too. Oddly, have never seen a PSU cause it, but many other people claim it as a top cause. May 24, 2010 at 14:42
  • +1. Although none of this helped, it's good to see you know what you are talking about, and have made good recommendations.
    – Myersguy
    May 28, 2010 at 3:01
2

You don't need to start swapping hardware until you have eliminated software as the cause. Check temps first. If they are good I would start with simple tests, like booting to a Ubuntu CD to see if your computer shuts down with a clean OS running. If that works then test your RAM with +memtest. If you have a memory failure pull your RAM out and run +memtest again, testing each stick individually. If you still have not found a problem by now you can start swapping stuff, like your GPUs, PSU, and other stuff.

1
  • 2
    Spontaneous reboots most of the time are PSU or mem errors. The only way the computer would spontaneously reboot via software would be through updating or a virus. Assuming the user isn't an idiot this isn't a software problem.
    – anon31097
    May 21, 2010 at 3:00
2

If you are using nvidea 64-bit drivers, make sure they are the latest version. I also had several machines rebooting/hanging/crashing at random.

3
  • I had that problem, and I updated them a while back. The display driver hasn't crashed in over a month.
    – C. Ross
    May 21, 2010 at 16:38
  • I believe driver issues usually result in blue screens.
    – wag2639
    May 23, 2010 at 22:07
  • He said it was crashing before booting into windows in some cases. Probably not the video drivers.
    – Myersguy
    May 27, 2010 at 22:58
2

In keeping what mand0 said, I suspect its your power supply.

However, as best practices go, I would recommend the following:

  1. Disconnect all unnecessary peripherals (use onboard graphics if possible) and to see if its your mobo, processor, ram, psu, or hdd
  2. Part of 1 really, use only 1 ram stick, but try different sticks and in different slots, and for good measure, use an air spray dust cleaner
  3. Make sure theres no dirt or loose cable hanging around shorting your case unexpectedly
  4. Stress test using Knoppix or similar (a usb method might be better but harder to set up)
  5. If theres still a problem, then try swapping hardware
0

Here are a few suggestions:

  1. Try getting realtemp instead of Speedfan. Speedfan often only logs the temperatures of your CPU, and not the individual cores. Realtemp tends to be more accurate at reading core temps, and that is what you need to be watching.

  2. When you re-seated your CPU heatsink, did you clean off the surfaces and re-apply thermal paste? It only makes matters worse if you simply move the heatsink. Use Rubbing Alcohol to clean off the surfaces of both the CPU and the heatsink, and then re-apply the thermal paste.

  3. Reset your CMOS back to default settings. Considering you used Arctic Cooling Thermal Paste, I assume you may have done some overclocking. Perhaps a setting has pushed the system to instability, and the easiest way of resolving it is to reset back to default. To do this, simply boot into your Bios (Usually DEL key upon startup) and find the "Reset CMOS settings", or "Restore CMOS defaults" button (or something to that extent). You can also see if there is a reset CMOS button on your motherboard.

*4. If you are running Vista/7, get your install CD, boot to it, and run the startup recovery tool. If you are no longer to enter your OS, and keep restarting, it sounds like a possible windows file corruption. This has happened to me.

*5 if #4 doesn't help, and you have a spare Hard Drive available, try installing windows and booting to it. If that is successful, then it is clear that you either have a broken drive, or a corrupt windows. Either run off of this second drive, or back up your files onto this drive, and then format the corrupt one.

1
  • 1. Can't get to the OS anymore. 2. Cleaned both surfaces, tinted the heatsink and re-applied fresh thermal paste. 3. No overclocking, and no bios changes at all. I'll try it though.
    – C. Ross
    May 28, 2010 at 1:28
0

I have been having the "unexpected shutdown problem" for some time for about a month on a new HP.

Since my last computer that apparently got a virus, I wanted to be protected with troubleshooting and virus protection programs.

The HP came preloaded with Norton anti-virus. I also installed Web-root anti-virus and a program called Fix-It.

After trying many fixes and going through many system shut-downs for weeks, and the shut-downs are usually after the computer goes into sleep mode. I decided that My troubleshooting and virus programs may be conflicting and shutting down the computer.

I removed "Fix-It" and the computer has run without shutting for several hours now.

I think that was the fix it needed. Having more than one virus and troubleshooting program makes them confict and turn off the computer.

I hope this helps.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .