|
Seems like the problem was solved by replacing the PSU.
I recently installed an extra HD for storage, which has been working fine for a couple of weeks but three times this week my PC randomly restarted without any blue screens or any other indications that something was wrong (logged as kernel power 43 errors without any additional error codes).
Since the HD was a recent addition to my system I decided it would be my first suspect. I thought maybe my PSU was failing and the additional power draw of another HD would cause it to fail randomly but that doesn't really make sense, does it? HDs don't need much power and my PC was restarting on low load while I could play games at full load without a problem.
Then I tried to manually shut off the HD from inside the Windows Device Manager and every time I tried, it resulted in my PC restarting without a BSOD. Could it be possible that at the times of the random restarts Windows was trying to shut off the HD because it wasn't in use (I just stored some shit on it a few weeks ago) in order to conserve power and therefore causing the restart? And what would cause a restart on a secondary HD shutting off anyway?
I have now completely deinstalled the new HD to see if the random restart problem persists but if it turns out that the new HD is the problem, what can I do?
|
Well, return the HD and get a new one? Or if you think it's just the shutting down/going on sleep issue that causes the problem change your power settings so the HD just stays awake even if it is not being used.
|
On March 23 2014 01:22 Jonoman92 wrote: Well, return the HD and get a new one? Or if you think it's just the shutting down/going on sleep issue that causes the problem change your power settings so the HD just stays awake even if it is not being used. Well, how do I know it's not some other problem than the HD being faulty? Why would my PC crash when I turn a HD off?
And currently my power settings apparently turn off a HD if it's not been used for 20 minutes but I don't really understand why the problem has only started this week while I've had the HD for a couple of weeks and it only happened three times. My PC is on a lot and also idle a lot. Why doesn't it happen more often? I guess something must be accessing the HD but there's only some static data stored on the HD, so what would that be?
|
Okay, I removed the HD a week ago and didn't get the crash again. However, yesterday and today it crashed three times again, so it couldn't have been the new HD. Now what?
Going to try updating my GPU's drivers. Then I want to check my RAM, is there a way to do that without Memtest86 as I don't have a USB or CD burner at my disposal right now?
|
Windows has something like memtest86 built in. Type "memory problem" into the Start menu. You'll see a single search result. Use that. It wants to restart the PC and will then do the memory test outside of Windows just like memtest86.
|
Make sure the drive firmware is up to date, and maybe try disabling SATA link power management.
|
I updated my GPU's drivers, didn't solve the problem.
Windows has something like memtest86 built in. Type "memory problem" into the Start menu. You'll see a single search result. Use that. It wants to restart the PC and will then do the memory test outside of Windows just like memtest86. The Windows memory diagnostic tool didn't find anything wrong. How reliable is it? It only ran for a couple of minutes before it finished, hardly seems like enough time to find a rarer problem.
Make sure the drive firmware is up to date, and maybe try disabling SATA link power management. I took the new HD out of my PC and the problem still persists.
So what should I do next. It looks like a random crash without a BSOD could be just about everything's fault, it's kinda disheartening Basically the only 'clue' I have is that it seems like it crashes at random. It happens during load and it happens while nothing is going on. Sometimes it doesn't happen for days.
If anyone has like a priority list of what the causes for this could be that I could check off, that would be great. I don't want to buy a new PSU (and sadly I don't have a spare one lying around) before I'm kinda sure nothing else is the problem.
PSU is a ~2.3 year old 500W Corsair CX500 powering a i5-2400, HD6870 and 8 gig of 1333MHz Kingston (I think) RAM all seated on a ASUS P8H67.
|
There's a way to make that Windows memory test tool run longer than the short test you've seen but I don't remember how to do that. You have to hit some key? It might be mentioned on screen at some point.
You are right that it can be caused by anything. When something like this happened to me, I just updated anything I could find. I have no idea which update fixed it.
There's the BIOS for the board, the graphics card has a BIOS, drivers for network controller, audio controller, Intel chipset drivers, drivers for the USB 3.0 controller. If software is the cause, it has to be something that runs inside the Windows kernel as normal programs can't freeze the PC. That can be Intel drivers, graphics card drivers, perhaps also network drivers. I have no idea about USB and audio drivers.
|
There's a way to make that Windows memory test tool run longer than the short test you've seen but I don't remember how to do that. You have to hit some key? It might be mentioned on screen at some point. I'll try to get an empty USB drive and run memtest.
Running Prime95 on blend mode didn't immediately explode my PC but I stopped it after a minute because CPU temps reached 85C. I want to run it a bit longer but I am afraid I'll melt the CPU (i5-2400, stock cooler). Note: my PC doesn't overheat while running anything else. Idle temps are ~30C and load temps during demanding games are 65-75C (still not ideal but I've ran nothing since the problem started that would get above 70C).
I've also configured Windows to make a memory dump upon crash, hopefully that contains useful information.
And if someone would change the title to reflect the actual problem (something like "random PC restarts, kernel power 41 error"), I would be grateful
|
Okay, now what.
PC crashed and restarted itself. Attempting to log into Windows resulted in a restart before the desktop was loaded. This happened three consecutive times. The third time the PC attempted to restart, I heard a strange noise and the system shut itself off before any bios screen even appeared.
Am I correct in guessing that my PSU blew itself (and hopefully not any other components) up? Would it be a good idea to try and turn the PC on again or would that carry a risk of frying everything?
Update: Tried booting the PC ~30 minutes later and it worked. Writing this from said PC. So, apparently the PSU hasn't died yet, but maybe it's because it cooled itself and will start failing again soon. The system just shutting itself off after the last restart before anything happened, what does it mean?
|
Maybe some of the following things will help to find out what's happening:
What does your windows event viewer (WIN-X choose event viewer) say under critical? post error messages and error codes at the time of the crash(es). Is it always the same error?
Turn off automatic reboot when there is a BSOD (WIN-X choose system -> advanced system settings -> startup and restore settings). If you have this already and don't see a BSOD it is a hardware problem. Could be PSU, GPU or motherboard.
Do memory dumps get written? If yes analyse these dumps with http://www.nirsoft.net/utils/blue_screen_view.html (thx to Ropid for mentioning this program. Didn't know about it before.) Post a screenshot of this program with the crash memory dump loaded.
[Edit] Oh btw WIN-X works on Windows 8.1 don't know which version you have. If you have a different one you got to find these settings yourself.
|
What does your windows event viewer (WIN-X choose event viewer) say under critical? post error messages and error codes at the time of the crash(es). Is it always the same error? It's only the generic error: + Show Spoiler +EventData
BugcheckCode 0 BugcheckParameter1 0x0 BugcheckParameter2 0x0 BugcheckParameter3 0x0 BugcheckParameter4 0x0 SleepInProgress false PowerButtonTimestamp 0
.
Turn off automatic reboot when there is a BSOD (WIN-X choose system -> advanced system settings -> startup and restore settings). If you have this already and don't see a BSOD it is a hardware problem. Could be PSU, GPU or motherboard. I haven't been seeing BSOD's in the first place, it just instantly crashes and restarts. Yesterday it restarted three times in a row while attempting to log into Windows and during the fourth restart it made a weird noise and shut itself off completely (before bios screens appeared).
I enabled memory dumps yesterday already but none of the crashes have registered any dumps. Does this signify problems with the power supply?
|
Mmh, it's not the GPU because memory dumps would still get written if it were the GPU. In my opinion it's either the PSU or the motherboard. If you have automatic reboot after BSOD turned off you can also exclude any drive errors because you would see see the blue screen anyway. Have you turned automatic reboot off? It's on by default.
|
So, I procured another PSU, let's see if the problem persists. I turned off automatic reboot now so if it's going to crash again we'll see if it's going to show a BSOD.
edit: has been performing for a day without a crash edit2: looks like it was the PSU starting to fail, haven't had a crash in days
|
|
|
|