06-23-2010 11:34 PM
I have an HP d4100e desktop computer running Win XP Pro. I usually leave the computer on all day long, but turn it off overnight, and re-start it each morning. The machine runs happily all day long, no glitches, no high temps, no issues. Until recently.
For the past month or two, occasionally (maybe once every 10 starts or so), the computer will shut down of its own accord. A hard shutdown, requiring a complete restart (which involves removing power from the computer by turning off the UPS, waiting a few minutes until the green light on the power supply turns off, and re-applying power). Sometimes the shutdown occurs during the boot process, or within seconds after booting, and sometimes the computer will run for several minutes before shutting down. In each case, after a restart, the machine runs happily from then on (all day long, and I suspect it could run 24/7 if I chose not to turn it off at night).
I've been trying to analyze what's happening, and I'm guessing at a couple of possible scenarios.
1. The power supply could be failing, and the startup and/or running loads might be bringing down;
2. The CPU fan might be sticking on startup occasionally, and the CPU might be shutting the machine down due to overtemp (could this happen within seconds of turning on the machine?).
1. Any other likely scenarios that might explain the problem?
2. How can I test the power supply without actually replacing it?
Any help gratefully appreciated.
06-23-2010 11:51 PM
There's are chances that the shutdown issue has to do with processor overheat, what i would suggest you is to access cmos setup in bios and monitor the processor temps from there. Once you're within the bios setting, use the down arrow key to look for the following either hardware monitor or health status depending which bios you're using, if the temps falls within 70-80 then is indeed heating up.
06-24-2010 12:14 AM
Thanks for the prompt reply, Dan.
After reading it, I immediately re-booted into the BIOS and got the following readings:
CPU temperature: 38 degrees C
CPU fan speed: 1844 RPM
System fan speed: 1369 RPM
I exited the BIOS and allowed the computer to fully boot.
Then, using Everest, I got the following readings:
CPU temperature: 40 degrees C
Motherboard temperature: 39 degrees C
CPU fan speed: 1848 RPM
Chassis fan speed: 1369 RPM
Note that these are readings after the computer has been running all day. The BIOS and Everest readings compare well with each other, and appear to be stable. The temperatures are reasonable.
But inorder to properly test your theory, I would need to be monitoring the temperature at the actual time of failure, which is unpredictable. Also, it's possible that the shutdown will occur before I even have time to access either the BIOS or Everest.
One possibility for testing this theory is to run the computer with the case open, so that I can view the CPU fan and verify that it's operating properly. I may do this, but I'm reluctant to run the machine for very long with the case open, because the cooling air flow patterns designed into the case would be disrupted, and might itself cause overheating (just thinking out loud - not even sure if that theory is correct).
Thanks again for your response.
06-24-2010 02:43 PM
Try this procedure:
Unplug the PC and open it up. Clean out all the dust. Carefully remove and replace all the cables going to the motherboard one at a time. Do the same for the memory dimms and the video and sound cards if you have any. Plug your PC back in a give it a go.
06-24-2010 09:29 PM
I'll try that next. Although, coincidentally, I just ran Memtest to check out the memory - passed with no problems.
I don't expect to find much dust (I periodically peek at the innards), but I rarely disconnect/reconnect connectors, so I'll give that a shot. The Vid card was pulled to access the memory slots when I upgraded the memory a few months ago, so probably no issue there. The only other card is a modem that's no longer being used. But LOTS of MB connectors to re-seat.
Thanks again for the suggestions.
06-25-2010 03:48 AM
" inorder to properly test your theory, I would need to be monitoring the temperature at the actual time of failure, which is unpredictable. Also, it's possible that the shutdown will occur before I even have time to access either the BIOS or Everest."
Not necessary, if the issue is indeed due to processor temps over heating, all you need to do is just monitor a few seconds or if you're afraid, maybe say 10 mins (monitoring within bios setting) and if the temps is within it's proper range like 38-50++ is pretty normal i would say
Since is not processor temps related, we go on to the next step, by the way is there any graphic card on your rig? If there is, try swapping it or if there's onboard video, just use those for the time being. Also i assume you did not OC or crank your rig be it through bios or tightening the ram timing? Have you try running your rig with just a piece of ram as in swapping one by one to see how it goes. Though you mention when you run memtest and it come out clean.
Just a thought is this a recent fresh install of windows? Have you install anything lately that causes the shutdown? Other than the above suggestion, i'll suggest you to download the harddisk utilty from it's manufacturer site and run a test on the drive itself. The other thing is try looking for any burn capacitor on the motherboard itself. Also you might wanna swap the psu to ensure is not due to it
Lastly have you try booting into safe mode and see whether the issue occurs, if it doesn't next go to event viewer and look for prompt or clues
06-25-2010 11:08 AM - last edited on 02-20-2017 03:20 PM by OscarFuentes
If the problem is not solid then it's going to be hard to find.
Your observation that the PC shuts off after during boot up tends to point to a hardware problem and not software. Boot into the bios (setup) and try some diagnostics.
I am leaning toward a power supply issue.
06-27-2010 01:57 PM
OK, here's what I've done since my last posting:
I opened the case, verified all fans (case, CPU, PS, graphic card) running on startup (which means nothing, since I haven't had the problem again since I originally posted).
I verified that all connectors and memory sticks were properly seated (did not actually pull them from the MB). All were tight.
Blew out all dust (there was none that I could see).
Dan, I have not OC'd this computer or messed with the original BIOS settings. I do have a graphic card, but nothing to replace it with. And it has been working perfectly, as far as I can tell. I haven't tried the on-board video.
Dave, the computer has shut off during boot, immediately (several seconds) after boot (Windows running), and several minutes after boot. Doesn't appear to be any consistancy to it, except that if it runs after boot for 10 minutes, it'll run forever.
Also, when the problem occurs, it always restarts with no problem, and runs without any further issues. It could be a PS issue (I haven't tried the steps in the link you provided yet), but it's unlikely that I'll be able to catch the problem by troubleshooting the PS if it's truly random. If it continues (or gets worse), I'll probably just replace the PS with one that has a higher rating than the OEM (they're certainly cheap enough).
Meanwhile, while I wait for it to happen again, the computer is quite usable.
Again thanks to you both for your suggestions.
06-27-2010 07:11 PM
You might try booting into the bios (setup) and run some diagnostics.
Use a flashlight and inspect the mothrboard for heat damage. Look at the capacitors and see if any are flaired or burst at the top.
06-27-2010 10:39 PM
As mentioned above, I've already run memtest (downloaded and burned to a CD, as you've suggested). No problems.
Although I didn't specifically mention it above, when I opened the computer to check the connectors and fans, and blow out the dust, I also inspected the MB. No visible damage, and the capacitors look fine.
As for booting into the BIOS, what sort of diagnostics are available in the BIOS? I don't recall any, but I'll have a look.