I'd like to get some opinions on this issue.
Approximately two weeks ago my PC suddenly shut down, as if a loss of power. Lights, Fans all off. It then re-started. I thought, hell, my power supply is going south.
However, over the next days I noticed a pattern. The PC would shut down within 2 to 4 minutes after first boot up after its been off some time (ie over night), but after it reboots it stays up for as long as I'm using it (up to 6 hours or more at times). I've got no crashes, no lock ups, no Fatal Exceptions or Blue Screens. I've had it under some load processing video, I've left it idle, it goes into sleep mode and comes out just fine. I would have expected some outages the longer the machine was on, or at least randomness in the problem.
I used a DOS 6.22 boot disk in order to bypass Windows and once again, on the first boot up, while just sitting at an A:> prompt she shut down and re-booted.
I went into the BIOS and looked at a setting for AC Power failure mode. It can be set to Power On after an AC outage, remain Powered Off or return to the Same State it was when the outage occurred. I was surprised to see that it was set to Power Off. That should have meant that after the power went out, the system should stay off. So it appears that the BIOS is not seeing this as an AC outage.
On first startup the other day I went back into the BIOS, and this time the system did not shut down. I'm sure I left it sitting in the BIOS for 10-15 minutes (well past the 2-4 minute period it normally fails at). I've only been able to do this once so far. I'll try again the next couple of days to see if this pattern holds.
I'm considering reflashing the BIOS, though it is the most current version and I haven't made any changes or updates to it in quite some time and I generally don't update firmware unless required.
+ Reply to Thread
Results 1 to 12 of 12
A few long-shot ideas, hopefully not insulting to your expertise:
o Heat problem? Maybe your CPU fan (or maybe PS fan) isn't working well at times and your system shuts down because of CPU heat buildup. After the reboot, for some reason the fan works a little better and no more reboots.
o Loose connector/cable somewhere? I just had a computer hang during bootup at "checking IDE drives" all of a sudden. Since I'd had the case open a few days before, I thought something might have been loosened, although the computer had booted since then. Sure enough, the power cable to the HD wasn't making good connection. Reconnecting it solved the problem. I'm guessing the heat-up/cool-down cycles when on/off may have caused it to work loose. This leads me to the next thought:
o General maintenance, hardware: Re-seat everything that might be loose, such as memory, boards, fan connectors, etc., etc. Over time, crud and oxidation can build up on connectors causing increased resistance until the point something will just no longer work. I've solved a few odd problems over the years by doing this. Didn't always know which component was the problem, but if the computer worked, I didn't care.
Good luck. If you determine the cause and solution, please let us know.
lots of other things it "might be". dying power supply. bad ram. bad ram bios settings. bad capacitor on motherboard. i'd start looking by unplugging and removing everything until it's just the cpu ram and videocard. see if it will stay on in bios with just that minimum. while in the bios the cpu runs 100% so you'll find out if it is a heat issue. watch the temps in the bios screen that shows them. if it works there then start adding back components until it starts failing again.
 while in the bios you can also watch for voltage fluctuations that might point to bad p.s. or capacitor.--
"a lot of people are better dead" - prisoner KSC2-303
I believe that this is a heat issue. If the heat gets to high, the processor will shutdown to protect itself - no warning and no OS intervention. Most every other failure mode would flag an error somehow.
Intel has a temperature monitoring program that will live in your icon tray that will give you a realtime check on the cpu temp. I believe that it is called coretemp, but just use Google to locate it. I last found it on Intel's main site.ICBM target coordinates:
26° 14' 10.16"N -- 80° 16' 0.91"W
No, it ain't Heat Related. If it was, it would keep re-booting, not just the first time. 2-4 minutes is not usually long enough, anyway.
Really, really odd. I would try another power supply. Check if there are any fan speed errors, if your BIOS reports them.
Thanks for your comments! I just got home and booted up to the BIOS, then had to leave the room. When I got back the PC had re-booted, so the problem also occurs when just sitting in BIOS. I'm suspecting its could be the motherboard. I vaguely recall a long time ago getting a message on post. I can't remember what it was but I recall thinking "WTF? I don't overclock". It could be a clocking issue. I believe if the clocking becomes unstable the MB will shut down. I don't overclock BTW. Anyway, I can only reproduce the problem about once per day so it may take me some time to narrow it all down. I think I'll run some memory and CPU tests this weekend, as well as check to make sure all the fans are in fact working at start up. It it will be checking each component after that. If I come across anything interesting I'll post back. Thanks
see if the bios has a "safe defaults" option you can chose and then reboot. also the default ram timings may be too fast for the voltage setting, maybe try bumping it up to the next higher voltage. it can happen even if you don't o'c.--
"a lot of people are better dead" - prisoner KSC2-303
I'm considering reflashing the BIOS
I have seen this. It's a common symptom of motherboad failure. Typically a capacitor near the cpu is bulging and when you attempt to turn the pc on the fan jumps to a high setting and won't do much else. Twenty or thirty reboots and the pc will probably boot up normally and then when restarted the cycle will begin again.
Lowellriggsiam has a good point; it is easy to check. Open the case and check the caps for any bulging tops. They should all be flat as a pancake, any that look like they are being pushed upwards are bad. Especially if you see brown "goo" leaking out the top or side is a sure sign of failure.
ASUS brand motherboards will give a overclocking has failed message when the computer hasn't started properly.If I'd known I was going to live this long, I'd have taken better care of myself.
Worked on the machine over the weekend... and may have some positive results, but only time will tell.
Yes, it is an ASUS board (P5K-E). I've dealt with capacitor issues in the past (ie DELL GX270). I did check the M/B for telltale signs, and all looked OK. Doesn't mean they all are OK, but no component (capacitor or otherwise) looked deformed, discoloured etc.
Heat is not an issue. The software monitor values seemed too low to me but when I checked in BIOS all temps were surprisingly low after regular system use (motherboard at 28C, CPU at 29C). All Fans start up properly, air flow was very good, dust was minimal, though the front air intake filter was a bit dirty so I cleaned it. VCore voltage was 1.12 to 1.22.
Saturday I decided to do two things. I yanked the RAM and re-booted and got the right no RAM beep code. I add the RAM back, then set the BIOS to "Default" settings.
On Sunday I booted up and.... no shutdown. I'll see what happens throughout the week.