BSOD on Custom Build With no .dmp files being created

[wisecraker1]

Hello,

I have been experiencing very intermittent issues with my Windows 7 64-Bit PC. It has been crashing with a Blue Screen ~2-3 Times/week for about a month now. The problem is that there are no .dmp files being created even though the system is set to write Kernel Memory Dumps on system failures. All Drivers are up-to-date as far as I know. The only crash file is from December, which is irrelevant to the current issues. My PC has been stable for over a year with the exception of the December crash which was a one-time thing imo. I do not overclock the PC and all bios settings are default with the exception of the RAM profile being set to XMP-1600 (because my ram by default was running 1333Mhz if I remember correctly).

Things I have done thus far:
CMD Prompt: sfc/scannow; chkdsk (C:, F:, H:, Y:, Z all with the toggle /r
- I presumed that perhaps a bad HDD/SSD was to blame as I recently installed the two 3TB HDD's, but they all checked out OK.
Memtest86 for 36 hours, no bad ram
- I recently installed 32GB ram vs 16GB

Attached is the DM log collector file.

Any help on how to get the system to write crash files, or ideas as to what could be causing the issues would be appreciated.

-Thank you,
Wise

 

[wisecraker1]

Update: My PC just crashed again... Twice. I disabled the "Automatic restart on system failure" and wrote down some information about the BSOD. Silly me, I found the .dmp file. It was set to be on a HDD that is NOT the OS drive.

Information about the crash: 0x50, Page_Fault_In_Nonpaged_area
So I presume a driver is the culprit? Also, the .dmp file is 2.5GB, 300MB compressed, I can upload somewhere if need be.

Edit: Second Crash, the .zip folder of the file is attached.
WhoCrashed blamed hardware: 0x124. I don't have any overclocks (aside from my GTX 680 which is factory overclocked).

 

[carwiz]

I was about to say to move your pagefile back to the C: drive. There were numerous "failed to initialize" for the dump file in the log. This can be caused by a "sleeping" drive. It's better to keep it on the system drive because it's always active. Also the single dump is not much help but it occurred while the processor driver (intelppm.sys) was in a wait state and memory was corrupt. There were numerous power state failures in the log too.

I'm leaning toward a weak power supply as the cause. Adding the additional memory may have just put it over the edge. With all the hardware you have, the power supply needs to be top notch. You might try monitoring voltages and temperatures for a while and use the logging feature built into many of the utilities. For Intel processors, I recommend HWiNFO64.

HWiNFO64 Free Version here: HWiNFO, HWiNFO32/64 - Download

 

[Layback Bear]

A PSU Corsair HX Series 650W if working properly should work well on that system.

Maybe it's not working as it should.

The spec requirements for that video card are minimum 550W @ +12V/38amps.
Power supply specs 12V+/52amps

 

[wisecraker1]

I downloaded the HWiNFO software and will monitor the power situation. I don't think it to be a temperature issue as the system never runs hot (always below 65 Celsius for CPU, 70 or below on GPU with high load -- I monitor basic CPU and GPU temps with NVidia System Monitor). And the crashes aren't while doing compute heavy tasks, just light browsing and such (the latest while doing a windows update). Which specific portions of the software should I be closely watching?

Also, Layback, I'm not sure what I should get from your post. Are you saying that you disagree and the PSU should be fine for my system?

-Many thanks,
Wise

 

[Layback Bear]

What I'm saying is.

If the power supply is working properly; which most Corsair do it has enough power for your system.
Now keep in mind any power supply can go bad but most of the time that will show up under a load.
If it was my computer I would not be replacing it at this time but I surely would be watching it.

I would also make sure the power supply fan works.
Some power supply fans come on with a temp sensor but I believe yours will come on just about any time. Just get a flash light and take a look.

Your system will not be using all your hardware at the same time. Example all hard drives will most likely not be running a the same time so a Corstair power supply with 650W/12V+/52amp should work well if in fact it is working properly.

 

[wisecraker1]

Results from HWiNFO64:
CPU
Power: 0.8-1.15v
Temps: 39 Celsius at 0-25% load (light usage), 50-71 Celsius at 50-100% load (video rendering with PremiereProCC)

GPU
Power: Not Shown
Temps: Idle 38 Celsius, moderate usage (limiting fps on games to have 50% usage on GPU) 55-60 Celsius, 100% load produced 75-82 Celsuis (82 with my door closed... it get's hot in this little office of mine... and high ambient temperatures don't help things at all)

But overall, I didn't see any temps that were overly bad. and the VID for the CPU seemed within reason. RAM settings are something I'm unsure of, but should not be a problem as I have ran Memtest extensively.
Gyazo - 81fc6da90e4260d2d0ff946d5d18bec1.png
Is a screenshot of the settings shown in the HWiNFO program.

And my system uses 560W at max load (according to http://extreme.outervision.com/psucalculatorlite.jsp) so 650W should be more than adequate.

I don't know if any of this helps at all.

 

[carwiz]

Please use this method to post images to the forum:
http://www.sevenforums.com/tutorials/9733-screenshots-files-upload-post-seven-forums.html

Not many of us will click on blind links.

Total Watts really doesn't mean much. It's a marketing tool. What counts is the output from the three voltage rails (3.3, 5.0, 12) and what the individual rails supply in Watts. The voltage must NOT vary by more that +/-5% on any one rail. I don't give much creed to those PS calculators. They are general at best. Even the most basic system is around 275-300W. Then start adding on to that.

 

[Layback Bear]

Please do as carwiz requests.


Screenshots and Files - Upload and Post in Seven Forums