Multiple BSOD on new build (partial)

[Hanzilol]

I've reviewed many previous threads with similar problems and still have the same issue after trying the suggestions. I recently bought upgrades for my wife's PC. These included new mobo (asrock 970m Pro3), new RAM (Crucial Ballistix Sport 8GB DDR3-1600), new CPU (AMD FX8320E) and a new SSD (PNY 240GB SATA III 6Gb/s 2.5" Internal Solid State Drive - SSD7CS1211-240-RB),.

I put it all together today with an HIS 7770, thermaltake toughpower qfan 500w PSU, and two HDDs (which were already installed and functioning well). I booted windows from a USB stick, wiped the drive with the OS on it (the other was just storage), and installed a fresh W7 ultimate on the SSD. I logged on, started installing drivers, got most of the way through, and started getting BSOD about every 5-10 minutes. I changed SATA cables on the SSD, ran chkdsk (no errors), moved RAM to another slot, then reinstalled windows. Same problem.

I'm currently running memtest, at 50% on first pass with no errors, but I plan to let it run for a while.

Could it be that the PSU is not enough to power all 3 drives (it seems that it should be)? Is it likely to be my memory that is bad? Could it be a faulty mobo?

The BSOD messages that I caught were Page fault in nonpaged area and irql not less or equal. Seems like there was another that said something about a driver trying to access something. Any ideas?

Update: I also wanted to note that while I was in windows, I noticed that my drives kept popping up (similar to how flash drives pop up when you plug them in). I unplugged one of the drives, found a case fan with a bent plug, fixed that. I'm coasting right now, but I'm not sure whether or not it'll BSOD on me again.

Update 2: Just BSOD'd again. Happened right after I installed the display drivers. Not sure if that is relevant.

Update 3: Got it running for a bit, downloaded all my software, opened WoW, BS'd at the character selection screen. This time it was "system_service_exception".

Update 4: Here are some DMP files, not sure how to convert them to txt: https://drive.google.com/open?id=0B6jahBT99YNSdlAzcVR0UUFvaWc
https://drive.google.com/open?id=0B6jahBT99YNSb2RyOGNLN3pXUVk

Update 5: Ran memtest overnight. Checked this morning, no errors, only ran 2 passes. Considering underclocking the ram to 1333 when I get home to see if that gets it stable.

 

[undefined512]

At this point, are you getting a BSOD after leaving your system idle after booting? If it only happens when under heavy loads, your problem could be a faulty power supply.

In case your problem isn't a faulty component, I'd reseat the RAM, reseat the graphics card, and check all the power and data cable connections once again.

 

[Hanzilol]

At this point, are you getting a BSOD after leaving your system idle after booting? If it only happens when under heavy loads, your problem could be a faulty power supply.

In case your problem isn't a faulty component, I'd reseat the RAM, reseat the graphics card, and check all the power and data cable connections once again.

It seems to be when it's under heavy load, but I just booted up several times and got BSOD as windows was loading. I just posted the dmp files in an update. I reseated everything. I noticed that there's a slider on this pcie slot, rather than the usual tab that secures the card. This wasn't locked. That could possibly be the problem. It's working right now, but I'm not sure for how long. The PSU was perfectly functional with the previous build, so I doubt it's the PSU itself.

 

[Hanzilol]

Is it possible that I should underclock the ram to 1333? The board is only rated for 1600 as over clock, so I'm assuming it's auto ocing it.

 

[marko55]

Most BSODs and random/frequent freezes are RAM related. If you've only got one RAM stick then t-shooting is pretty tough as you don't have another DIMM to try to prove the other stick is bad.

Easiest way to snuff it out is to boot to one RAM stick only and run something like Prime 95 for a while. When you run prime's stress test, select "custom", uncheck "Run FFTs in-place" so you can specify how much RAM to put under load. Make sure that number is enough to bring your overall RAM utilization to a good 95%. So if you're running idle after a fresh boot, and windows alone (and background processes) are using say 1.5GB of RAM, and you've got a single 8GB stick installed, enter 6400 in the memory box. This should bring your memory utilization close to 100% and ensure that all parts of the RAM are utilized. If you still crash/BSOD, then put a different stick in and see if you have the same problem. If it happens with two different sticks, its most likely not RAM related.

Memtest is a decent tool but I've had memory issues that it didn't sniff out and had to find it a different way, like above.

Also, did you upgrade the BIOS? ASrock boards are infamous for having wonkey memory issues, even with sticks that are on their QVLs. BIOS upgrades typically stabilize them.

 

[Hanzilol]

Most BSODs and random/frequent freezes are RAM related. If you've only got one RAM stick then t-shooting is pretty tough as you don't have another DIMM to try to prove the other stick is bad.

Easiest way to snuff it out is to boot to one RAM stick only and run something like Prime 95 for a while. When you run prime's stress test, select "custom", uncheck "Run FFTs in-place" so you can specify how much RAM to put under load. Make sure that number is enough to bring your overall RAM utilization to a good 95%. So if you're running idle after a fresh boot, and windows alone (and background processes) are using say 1.5GB of RAM, and you've got a single 8GB stick installed, enter 6400 in the memory box. This should bring your memory utilization close to 100% and ensure that all parts of the RAM are utilized. If you still crash/BSOD, then put a different stick in and see if you have the same problem. If it happens with two different sticks, its most likely not RAM related.

Memtest is a decent tool but I've had memory issues that it didn't sniff out and had to find it a different way, like above.

Also, did you upgrade the BIOS? ASrock boards are infamous for having wonkey memory issues, even with sticks that are on their QVLs. BIOS upgrades typically stabilize them.

I did upgrade BIOS. I'll try another RAM set when I get home and see if that solves the problem.

 

[johnbl]

overall, update the motherboard drivers to match the windows 7, and apply all of the windows 7 patches released after sp1.
and see if you can produce a bugcheck after the updates. There are just too many old bugs in the old code. Update and see if you can still get a bugcheck.

(the bugcheck shows a attempt to access a memory address that looks like it should be valid but was not. Normally in this case I see a really wrong memory address indicating corruption) Do the updates and post the next bugcheck and the debugger should work better and maybe I can see the problems cause.

-looks like the bugcheck is coming out of a 32 bit subsystem.

---------
remove \SystemRoot\system32\DRIVERS\AsrAppCharger.sys Tue May 10 01:28:46 2011
and maybe \SystemRoot\system32\DRIVERS\usbfilter.sys Tue Dec 13 23:44:02 2011

apply the motherboard driver updates:
http://www.asrock.com/mb/AMD/970M%20Pro3/?cat=Download&os=Win764

these updates are required when you updated the bios. Mainly because the BIOS has the new spec changes in 2012 and 2013 and you have a old version of windows 7 sp1 from 2010 (with drivers that existed before the spec changes)



machine info:
BIOS Version P1.50
BIOS Starting Address Segment f000
BIOS Release Date 01/11/2016
Manufacturer ASRock
Product 970M Pro3
Processor Version AMD FX-8320E Eight-Core Processor
Processor Voltage 8bh - 1.1V
External Clock 200MHz
Max Speed 3200MHz
Current Speed 3200MHz

 

[Hanzilol]

I replaced the RAM and it works now, except I get a few random freezes here and there. The system usually comes out of it, but it's very annoying, and lasts like 2-3 min. Could it be that the system was installed with bad RAM? Should I just reinstall windows or should I try applying updates before that?