PC Started Crashing in the Past 2 Days

[peternm22]

Hey Everyone,

Hopefully someone can help me diagnose this problem. I've had this computer since June, and no problems until the last couple of days.

Specs:

Windows 7 Home Premium 64bit
Motherboard: Gigabyte GA-MA785GM-US2H
CPU: AMD Phenom II X4 945
4GB RAM
OS Drive: 80GB Intel SSD
Secondary Drive: Western Digital WD15EARS


Yesterday, the computer was sitting idle with a single Firefox window open. I was doing something else, but I noticed the computer monitor suddenly go black. I don't have any sleep mode stuff currently enabled, so I was worried. I couldn't get the computer to respond, so I had to hit the power button on the front of the case.

I can chkdisk on both drives, no errors. Also ran a virus scan, and malware scan, nothing. I checked the event log and noticed an error 41. Google didn't show anything that I could understand, so I hoped it was just a one off occurrence. Then it happened again tonight when I was simply writing an e-mail.

I grabbed this from the event viewer, but I don't know if it means anything:

- <Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
- <System>
<Provider Name="Microsoft-Windows-Kernel-Power" Guid="{331C3B3A-2005-44C2-AC5E-77220C37D6B4}" />
<EventID>41</EventID>
<Version>2</Version>
<Level>1</Level>
<Task>63</Task>
<Opcode>0</Opcode>
<Keywords>0x8000000000000002</Keywords>
<TimeCreated SystemTime="2010-11-16T05:56:26.142415100Z" />
<EventRecordID>76199</EventRecordID>
<Correlation />
<Execution ProcessID="4" ThreadID="8" />
<Channel>System</Channel>
<Computer>Peter-PC</Computer>
<Security UserID="S-1-5-18" />
</System>
- <EventData>
<Data Name="BugcheckCode">0</Data>
<Data Name="BugcheckParameter1">0x0</Data>
<Data Name="BugcheckParameter2">0x0</Data>
<Data Name="BugcheckParameter3">0x0</Data>
<Data Name="BugcheckParameter4">0x0</Data>
<Data Name="SleepInProgress">false</Data>
<Data Name="PowerButtonTimestamp">0</Data>
</EventData>
</Event>

I haven't changed any hardware in this machine since getting the computer. I've been having trouble over the last month with Flash videos online, with hardware acceleration enabled in Flash, Firefox would hang and max out one of the cores on my CPU when trying to play them. Trying to end the task in task manager never worked with the hanged application (nicely done Microsoft), so I had to force the computer to reboot. I upgraded the graphics driver to the latest on ATI's website a week ago, and I haven't had any issues with Flash videos since. So I'm not sure if this is related or not, since computer crashing didn't appear right away. The two crashes in the last two days both occurred when I wasn't viewing any video.

Does anyone know how to narrow this down some more? I've tried looking through the event viewer but I can't make sense of most of it, and the only "Critical" item that appeared was the Event 41.

I would like be able to narrow this down to some driver, or RAM, or something...

Any help would be appreciated.

-Peter

[JetreL]

Here is the knowledge base on the error you posted.

Link [microsoft.com]

This is a general error encompassing a good deal of possibilities. If this continues to happen I would check your power supply, any drivers you recently installed, etc. The reason you had to hit the power button was the computer was "turned off". Hope this helps!!

[ccsmoke]

Noticed you have a SSD. Was wondering if you turned off defrag, indexing and such?

Good guide:
http://www.ocztechnologyforum.com/fo...ks-Utilities-*

Not to familiar with Intel's ssd trim software, but I believe there is some sort of tool that will scan the drive and see how many dead blocks there are and life remaining of the drive.

I would also suggest running prime 95 on the computer for 48 hours.

http://www.mersenne.org/freesoft/

[peternm22]

The SSD is properly configured. Indexing, defrag, prefetch, etc is all off.

I ran memtest 86+ last night for 9 hours, and didn't have a single error in any of the 10 passes. So I think that rules out memory.

[ccsmoke]

9 hours is just not long enough IMO. I have my gaming rig, when I first built it, actually pass the 24 hour mark but got a rounding error after that. Generally you will get an error right away but that is not always the case.

Everything look ok under device manager? What is the brand and size of the powersupply? This is the tool that I was referring to http://downloadcenter.intel.com/Deta...&DwnldID=18455

[peternm22]

I've used that Intel tool before, I'll give it another go when I get home today to see if anything is popping up now.

From what I've read, It seemed like 10 passes was pretty good for memtest 86+.

I'm going to revert back to my old video card drivers (the ones the PC came with). Do I need to do something special to ensure that all of the drivers are wiped out before installing the older ones?

My power supply is a Seasonic X650, which according to most reviews is pretty top of the line.

When the PC crashed, it didn't actually shut it down on it's own. The screen just went black, and nothing would bring it back. All of the lights and case fans were still going.

I understand that sometimes Windows will create a dump file if it crashes, does anyone know where to find it? Or how to parse it?

Thanks.

[peternm22]

Ran all of the diagnostic tests in the Intel SSD Toolbox, everything passed.

Ran Western Digital Lifeguard diagnostics on the 1.5TB drive, it passed.

Will try reinstalling older video card drives next...

[peternm22]

New video drivers wiped, old ones reinstalled. Also reinstalled all the chipset drivers.

Now I wait to see if it crashes again...

[ccsmoke]

Just to throw this out there...I had a flash update around that timeframe, I noticed that you had firefox open when issues were caused. One of the things you can try is running firefox in safe mode and ensuring there is no plug-ins running.

http://kb.mozillazine.org/Issues_related_to_plugins

[peternm22]

I saw something kind of weird in the event viewer related to the Foxit PDF reader's Firefox plugin, with a time stamp about 10 minutes before each crash (despite the fact I had not recently viewed a PDF).

I uninstalled Foxit for the time being. If I get another crash while Firefox is running, I'll try the safe mode thing.

Thanks for all the help.

[peternm22]

Well after several weeks of no problems, I had another crash tonight.

I use the PC as my SageTV server, and my HD300 stopped responding. I checked the computer and the screen was simply black again and I had to hit the reset button on the computer.

I think the Event 41 in the event viewer is simply the act of me having to physically press the button on the computer for it to restart.

Again, does anyone know how to check dump files etc to try to narrow this down to a particular driver or something? I'm completely stumped.

[mistergq]

I had a similar problem last year. Replaced part by part until I finally ended up replacing the motherboard and cpu.

Good luck trying to find the problem. I remember how frustrating this was. Something not as obviously, have you checked to make sure all the capacitors on the motherboard look normal?

[peternm22]

I'll check, but my computer is only about 6 months old so I don't think the capacitors can go that quickly. I thought most of the capacitor plague problems had been solved.

Odd thing happened when I had to reboot last time after the crash, my Hauppauge 2250 had disappeared. Sage couldn't see it, and it wasn't in the device manager. Another reboot rectified that problem.

I've also had a few recordings from the 2250 with odd glitches in them in recent weeks. Basically, the file duration is reported as being lower than it actually is (5 minutes, instead of 60 minutes for example). SageTV sees this as the end of the file and stops playback. VLC reports the same duration. Running the files through VideoReDo's quick stream fix solves this. I have never had this problem with the 2250 before.

This leads me to believe the 2250 may be the cause. I know the last time I had the crash, the 2250 was recording. I'm unsure if that was the case for the first couple of crashes though. Could a faulty 2250 actually crash the system this way?

I'll keep an eye on it, but it all seems so sporadic I can't get a feel for what is causing the problem.

Thanks.

[davephan]

You can isolate the problem between software and hardware very quickly by recovering your boot drive back to a point in time with an image, taken at a point in time when everything was working normally. If the problems do not re-occur, then the problems are fixed. If the problems re-occurs, then you need to start looking at the hardware - power supply, system board, memory, etc.

If you don't have a image of your boot drive, then you are stuck with a painful and time consuming scratch rebuild. If that is the case, you might do the scratch rebuild on a different hard drive in case there may be drivers saved on the old boot drive that you may need. If the scratch rebuild fixes the problem, then take an image of the system. Add more software and configuration changes, then take another image. With periodic images during the re-build process and after the system build is completed, you will never have to go back to a scratch rebuild again.

If you still have the problems after the scratch rebuild, then you will need to look at hardware problems.

Dave

[peternm22]

I do have system images, but since the crashes seem to happen so far apart (it was about a 3 week gap last time) it will be hard to tell if the problem is fixed or not. If the crashes were easily reproducible, I would absolutely use an older system image to see if it fixes it.

Doesn't Windows 7 have some type of dump logs or something? With my old XP system a few years back, I was getting random reboots with it. I go through the dump files, and find that it was my network card driver causing the problem. I updated the driver, all fixed.

I find it hard to believe that Windows 7 wouldn't at least have the same, if not better, logging functionality for crashes. I tried Googling for the answer, but not much luck finding anything.

Thanks for the suggestion though.

[gplasky]

Not sure what you googled. If you want to look at the memory dump on Windows 7 download and install the debugging tools.

windbg

Gerry

[SHS]

Oh this a dead giveway

When the PC crashed, it didn't actually shut it down on it's own. The screen just went black, and nothing would bring it back. All of the lights and case fans were still going.

That sound lot like a bad memory problem I had very same problem with my crucial ballistix 1066Mhz with random off wall crash about 6 months back with my gaming system which turn out to bad memory.
A good tool to have round BlueScreenView http://www.nirsoft.net/utils/blue_screen_view.html

[peternm22]

Thanks Gerry, I'll have a go with those.

SHS, I tried BlueScreenView but I don't think it's appropriate for this since I never actually get a blue screen of death, the computer just goes to a black screen with no info or anything displayed. BlueScreenView is also reading "0 Crashes".

Do I need to enable something in Windows 7 so it creates dump files? Or is the fact that i don't even get a BSOD mean that the dump files won't show anything in this case?

Thanks.

[SHS]

Do I need to enable something in Windows 7 so it creates dump files? Or is the fact that i don't even get a BSOD mean that the dump files won't show anything in this case?

By defaults it all ready on
Any way see this page
http://pcsupport.about.com/od/window...-windows-7.htm
Your problem more likely to do with memory like main system memory, cpu and possable the video card memory.
Dowmload any video benchmark tool run in loop mode or demo loop mode with case panel on and let run for a few hours if still runing then chanes are it the main system memory so pull the case panel check system memory see is it getting hot if fell way to hot more then likely that where the problem is.

[peternm22]

Just had another lock up (first time in about a month).

This one was a bit different, instead of a black screen the screen simply "froze" on the desktop. Only thing running was the SageTV service recording a single show on the R5000 (there goes my theory about a broken 2250). Time on the computer was frozen at 6:25, and I didn't come to the computer until about 5 minutes after.

All of the times in the event viewer correspond to when I physically turned the power off/on at 6:30. Not a single event near 6:25. I believe the kernel power event mentioned earlier in the thread simply corresponds to the physical shutting off of the computer, not the actually crash. No memory dumps or anything of that nature were created (same as the previous crashes)

I haven't tried any of SHS's benchmarking tool tests, simply because the system isn't really do much of anything when the crashes happen. It's not as if the system is being overworked.

Found this hotfix on Microsoft's site: http://support.microsoft.com/kb/2265716/

Installed it, hopefully this will fix it.... seems to have helped others with similar problems. If anyone else has any ideas, I'm open to them.