Solved BSOD ntoskrnl.exe with different bug check strings every time

[bcware]

BSODs occur randomly, but most often when the computer is first started cold in the morning. It generally happens several times in a row and usually within 30 seconds of reaching the desktop; other times windows does not make it to the desktop at all. This HAS occurred in safe-mode with networking turned on as well.

The bug check codes and strings are all over the place; I'm getting a mix of everything. BlueScreenView is showing ntoskrnl.exe in red for every crash even if it was not listed as the cause.

After troubleshooting for a while in safe mode I will randomly get a stable boot and the computer will run all day long (16+ hours playing games and running Microsoft Office with no issue).

Temperatures seem very low; I have some pretty beastly cooling for my setup and crashes aren't occurring more often under load or more often as the computer runs for a longer period of time. Ambient temperatures are 65 - 70 F.

I ran memtest86 v4.3.7 for 4.5 hours with no errors.

I rolled my graphics card driver back to the previous version I was using before the crashes started; no dice. BUT there have been many recent drivers released. I could continue to roll it back.

I have run Malware Bytes, AVG Free, and Kaspersky TDSSKiller scans; nothing found.

I'm sure not all of my drivers are up to date, but I wouldn't even know where to start because there are so many.

Thank you for the help in advance. I appreciate it.

 

[bcware]

I will literally mail you a cookie if you help me fix this.

 

[axe0]

Please run the dm log collector again, there are some important files missing.

 

[bcware]

Done. I ran it as administrator and hit "grab all." It finished without errors and I zipped it once instructed. The new attachment has been uploaded to the first post up top.

Oh, and the specs in my sig are for this machine if that is of any benefit; I included all of the model numbers as well.

 

[axe0]

Uninstall AVG while troubleshooting and use Microsoft Security Essentials as alternative
http://www.sevenforums.com/tutorials/343899-avg-uninstall-completely.html
Microsoft Security Essentials - Microsoft Windows

fffff880`0a372798  fffff880`043a9cf5Unable to load image \SystemRoot\system32\DRIVERS\avgidsdrivera.sys, Win32 error 0n2
*** WARNING: Unable to verify timestamp for avgidsdrivera.sys
*** ERROR: Module load completed but symbols could not be loaded for avgidsdrivera.sys
 avgidsdrivera+0x20cf5

Update your USB network adapter through the manufacturer of the USB if possible.
Always keep USB network adapter drivers up to date, because they sometimes could cause network problems.

fffff880`02f8c808  fffff880`06608a01Unable to load image \SystemRoot\system32\DRIVERS\athurx.sys, Win32 error 0n2
*** WARNING: Unable to verify timestamp for athurx.sys
*** ERROR: Module load completed but symbols could not be loaded for athurx.sys
 athurx+0x4a01
2: kd> lmvm athurx
start             end                 module name
fffff880`06604000 fffff880`067e2000   athurx   T (no symbols)           
    Loaded symbol image file: athurx.sys
    Image path: \SystemRoot\system32\DRIVERS\athurx.sys
    Image name: athurx.sys
    Timestamp:        [COLOR="Red"]Wed Apr 20 12:07:45 2011[/COLOR] (4DAEB071)
    CheckSum:         001E1C78
    ImageSize:        001DE000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4

For the 0x4E and

MODULE_NAME: memory_corruption

IMAGE_NAME:  memory_corruption

FOLLOWUP_NAME:  memory_corruption

MemTest86+ is different from MemTest86 and MemTest86+ is better then MemTest86 with DDR3 what most users have.

Diagnostic Test

 RAM TEST

Run MemTest86+ to analyse your RAM. MemTest86+ - Test RAM - Windows 7 Forums

   Note

MemTest86+ needs to be run for at least 8 passes for conclusive results. Set it running before you go to bed and leave it overnight. We're looking for zero errors here. Even a single error will indicate RAM failure.

 

[bcware]

Update:

I did a fresh install of windows. I started BSODing like crazy on startup, so I immediately ran memtest86. Within 1 minute I found many errors (like 38 or 48 I believe). Great, so it's the ram, right? Well... I'm having a hard time recreating the error.

At this point I am running memtest86 with 1 stick of ram in at a time and moving them through all of the different slot positions. I have reached the point where each stick has undergone 25+ total passes; generally 6 - 12 passes in each of the 4 different positions on the motherboard. Each stick is separated and I am recording results accurately. So far I have been unable to find an error. I am continuing to run the tests even now.

Should I still be using memtest86+ if I have DDR2 ram? It was mentioned that memtest86+ is better for DDR3, but that isn't what I have. Also, I had trouble making a bootable USB of memtest86+, which is why I went with memtest86 initially.

Any suggestions?

 

[axe0]

What are the model and make of the sticks?

 

[bcware]

G.skill 4 x 2.0 GB (f2-6400cl4d-4gbpi-b)

So, two of these kits for a total of 8.0 GB

I also did another quick 4 passes with all memory installed and did not get an error. After all of my testing of the sticks one at a time I had completed as much as 25 passes in a single run without errors. I only got errors the one time with all of the memory installed and it was within 1 minutes or so of testing.

I'm running the system now to see if it will crash more on me.

 

[axe0]

When you test your memory with MemTest86+, straight 8 complete passes are needed for conclusive results. Less is simply a waste of time.

Please try MemTest86+ on a different bootable USB with all the modules.

To answer your question, MemTest86+ supports DDR2.

 

[bcware]

Thank you. I will do this and report back with the results.

 

[bcware]

Update:

memtest86+ 8 passes completed without any errors

Observations:

Is it possible that the first error was a fluke? Could the ram have been installed poorly (not seated properly) or might dust/debris contribute to the error? So far I have had good stability since running all of these tests and pulling the ram dozens of times but it is still too soon to say (2 days of up time so far). I have attached some information gathered from cpu-z. I was wondering if voltage or timing might be an issue; the bios is currently set to auto detect all settings. It all looks correct to me, however, but I am no expert.

My ram specs: G.SKILL PI Black 4GB (2 x 2GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Dual Channel Kit Desktop Memory Model F2-6400CL4D-4GBPI-B - Newegg.com

***NOTE*** both cpu-z and memtest86+ report the ram model number wrong; I have confirmed the above link is correct by checking the sticker on the sticks.

 

[axe0]

Update:

memtest86+ 8 passes completed without any errors

Observations:

Is it possible that the first error was a fluke? Could the ram have been installed poorly (not seated properly) or might dust/debris contribute to the error? So far I have had good stability since running all of these tests and pulling the ram dozens of times but it is still too soon to say (2 days of up time so far). I have attached some information gathered from cpu-z. I was wondering if voltage or timing might be an issue; the bios is currently set to auto detect all settings. It all looks correct to me, however, but I am no expert.

I wouldn't know if the first error was a fluke, it could be due to poor seated sticks or somewhere in the connection some dust.

Yes to both other questions.

 

[bcware]

I thought I would follow up since the problem has been "resolved."

Following my last update/post windows was stable for around 20 days and then suddenly I started BSODing frequently with random codes. Over time this got worse and the computer would refuse to run through those initial bios checks when power is first started (memory, cpu, drives, etc).

It seems like it was hardware related, so I simply replaced the motherboard AND memory. I needed the upgrade anyway and the price was right. I will mark this as resolved as well. Thank you for the help.