BSOD Various BCCodes 1a, d1, 24, 7e, 3b

[Brian51]

My nephews gaming computer. A little background; several weeks ago video output quit so nephew called me and I too found NO signal output from any PCI slot (no on board vid) regardless of what video card I tried. Dead pci, called Asus, sent them the card, they sent me a new one. Now I have video output, boots up ok a few times, then the BSODs start, not all the time but too often. Ran memtest86+ on ram twice for at least 8 passes. 0 errors. 600watt PS voltages look good. Just picked up a new PNY geforce gtx750 w/2GB. Gonna try that whilst you contemplate my dilemma. ThanX in advance.
Brian

 

[Golden]

fffff880`05491560  00000000`00000002
fffff880`05491568  fffff880`111a5fcf*** WARNING: Unable to verify timestamp for nvlddmkm.sys
*** ERROR: Module load completed but symbols could not be loaded for [COLOR="Red"][B]nvlddmkm.sys[/B][/COLOR]
 nvlddmkm+0x14afcf
fffff880`05491570  fffffa80`097e90b8

Uninstall the existing video driver, then run through this tutorial to clean out any remnant files:
http://www.sevenforums.com/tutorials/83814-drivers-clean-left-over-files-after-uninstalling.html

Reinstall the lastest driver for that card, following this guide:
http://www.sevenforums.com/tutorials/327414-nvidia-drivers-avoid-problems.html

Now ensure that Windows is fully updated.

If you continue to experience problems, please follow this:
http://www.sevenforums.com/tutorials/330579-graphics-card-problems-diagnose.html

Im hoping its not the PCI-e slot on the motherboard thats the issue.

 

[essenbe]

Brian51, It looks like you've covered your bases pretty good. If you follow Golden's directions, it should eliminate the Drivers, there is not much left but the card, which you've tried several, the PSU or the Motherboard slots. Have you considered swapping cards with another computer. Swapping with a known good card should tell you where to go next. It seems that if a known good card crashes on that computer, you are looking at the PSU or Motherboard. If your new card runs OK on the other computer, it tells you it's not the card.

 

[Brian51]

ThanX essenbe, I have swapped good working cards for comparison. I have followed all of Golden's instructions and am vigorously testing the system with game playing, and repeated shutdowns and restarts. simulating normal usage. All is well sooooo far. Will post again tomorrow.
Brian

 

[essenbe]

Great, I hope your good luck continues. If you want to really test the card, make sure you have a good Temp monitoring program and can set a custom fan profile to keep temps down and run Furmark, Ungine Heaven and Mark 3D Firestrike. Furmark is free, Heaven is too, I think Firestrike is a trial. But those 3 will put your card through the mill, I run them all the time. I just don't trust Speedfan that much.
http://www.sevenforums.com/tutorials/100356-video-card-stress-test-furmark.html
Heaven Benchmark | Unigine: real-time 3D engine (game, simulation, visualization and VR)
3DMark Download v1.4.828

Furmark and Heaven are pure GPU tests, Firestrike is CPU+GPU

We have a benchmark section here.

 

[Brian51]

Well, it didn't work. OK it worked a few times then I shut it down and went and watched a movie on tv. I fired it back up about a half hpur ago to a BSOD BCCode 1a. Thats where this all started. I ran memtest86 overnight twice, no errors. Could it be the video card memory? Or I saw in another post that maybe the video card is using some of the main ram and messing up. I attached a new file.
AM going to bed. Help.
Had another thought; this all really started when we lost ALL signal output capabilities from the PCI slots. Could that have damaged the video card, OR could a defective video card have wiped out the PCI sys.???

 

[essenbe]

First to answer your questions the best I can. I presume anything is possible, but I have never seen a defective video card ruin a PCIe slot, but I have seen what I believe to be the reverse. What you are most likely to find is a PSU taking them both out, from a power spike or an outage. The card has 2GB of memory so I wouldn't think it could need to borrow more unless you are running some very intensive games or a multi monitor set up.

What I would like you to try is running Seatools, Seagate's Hard Drive Diagnostic test. Please run the DOS version of the test, both the short and the long version. http://www.sevenforums.com/tutorials/313457-seatools-dos-windows-how-use.html#Part3

From one of your dump files from today

Use !analyze -v to get detailed debugging information.
BugCheck 1000007E, {ffffffffc0000005, fffff8000335e949, fffff880033689e8, fffff88003368240}
*** WARNING: Unable to verify timestamp for amd_xata.sys
*** ERROR: Module load completed but symbols could not be loaded for amd_xata.sys
[COLOR=red]Probably caused by : amd_xata.sys[/COLOR] ( amd_xata+7de8 )
Followup: MachineOwner

amd_xata
start             end                 module name
fffff880`01085000 fffff880`01093000   amd_xata T (no symbols)           
    Loaded symbol image file: amd_xata.sys
    Image path: \SystemRoot\system32\DRIVERS\amd_xata.sys
    Image name: amd_xata.sys
    Timestamp:        [COLOR=red]Wed Apr 11 09:40:27 2012[/COLOR] (4F8597DB)
    CheckSum:         0000E9AF
    ImageSize:        0000E000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4

That is the AMD Sata driver.

I may also suggest you see if you can get an updated sata driver from the Motherboard's web site, Motherboards - M5A99X EVO R2.0 - ASUS but I don't believe there is one there.

 

[Brian51]

OK, I have done many more tests with included results; newest dump, cpuid hwmonitor, cpuid asus cpu-z, and 3dmark results. So, enssenbe, golden, or anybody, what do you gleen from these reports.
Will do seagate tests now.
link to 3dmark page
NVIDIA GeForce GT 630 video card benchmark result - AMD FX-8120,ASUSTeK COMPUTER INC. M5A99X EVO R2.0

 

[Brian51]

Almost forgot, I ran hwmonitor while I ran 3dmark tests.

 

[essenbe]

Ok, it would help me more if you posted screenshots of HWmonitor and CPUz as well as GPUz sensor tab. It is easier to read. The GPUz sensor log test is fine, but with all of these it would mean more if I knew what you were doing at the time of these logs.

 

[Brian51]

NVIDIA GeForce GT 630 video card benchmark result - AMD FX-8120,ASUSTeK COMPUTER INC. M5A99X EVO R2.0 new test ran today with monitor programs on. screen shots at idle right after bootup, then after running 3dmark tests.
I am going to put in a brand new PNY GTX 750 w/2GB now and run machine

 

[essenbe]

Can you tell me, I can't find it, what the max GPU temp was? Also, I'm not sure about software readings of the voltages, but, if you will look, your +12V, +5V and +3,3V are all at the very bottom of the tolerable range. ATX Standards are +/- 5%. If you will look at the minimums, you will see they fall off to the very bottom of the standards range. Software readings are notoriously inaccurate. I use software too, but I use it only as a guideline. Those voltages would scare me.

 

[Brian51]

On the hwmonitor after the testing the graphics card high temp says 79deg C
Should have switched gpuz to semsors tab, sorry
Just got results from 3dmark primary test with new gtx750 card have to switch computers to send you info

 

[essenbe]

OK, no problem, just let me know.

 

[Brian51]

NVIDIA GeForce GTX 750 video card benchmark result - AMD FX-8120,ASUSTeK COMPUTER INC. M5A99X EVO R2.0 new 3dmark results with new gtx750 card
I have a new 550w PSU here, I will change it out and test again.

 

[essenbe]

Just so you know, almost all of the Mark 3D tests are CPU+GPU tests, the scores depend on both. If you want to check the GPU, run the Ungine tests, they are GPU only tests and tend to stress the GPU a little more. I do use both for system stability but use Heaven for testing the GPU.

 

[Brian51]

Ran Heaven with new PSU and video card. The main voltages look better, but some of the readings give me pause! They are too identical to the last PSU. Maybe its just me. I am going to run this system this way on/off waiting that BSOD. If I get none, tomorrow I will drop the old gtx630 back in and test. again
file:///C:/Users/Chas/Desktop/Brian%20save/Unigine_Heaven_Benchmark_4.0_20141215_1315.html

 

[essenbe]

Yea, those voltages don't look much better than the other one. It could be the program too.

 

[Brian51]

OK, with the new PSU and Video card I have gotten 2 BSOD , 3B SYSTEM SERVICES EXCEPTION ( I got another 3b yesterday in safe mode when I started GPU-z, before I switched the PSU & vid) and I got an bccode 'a' ........
The only other thing I have seen is a little corrosion on the power connector for the sata drive
I included the dump. My nephew did tell me last night that it did crap out (bsod) several times BEFORE the troulbe with the video output and subsequent MB replacement.

 

[essenbe]

Here are what I found from the 3 dumps from today. I know you said you have run memtest86+ a couple of times, but this is what the dump file says.

Use !analyze -v to get detailed debugging information.
BugCheck A, {28, 2, 0, fffff8000331ad80}
[COLOR=red]Probably caused by : memory_corruption[/COLOR] ( nt!MiFindNodeOrParent+0 )
Followup: MachineOwner

Use !analyze -v to get detailed debugging information.
BugCheck 3B, {c0000005, fffff880111c9fe2, fffff880092ab520, 0}
*** WARNING: Unable to verify timestamp for nvlddmkm.sys
*** ERROR: Module load completed but symbols could not be loaded for nvlddmkm.sys
[COLOR=red]Probably caused by : nvlddmkm.sys[/COLOR] ( nvlddmkm+1c5fe2 )
Followup: MachineOwner

Use !analyze -v to get detailed debugging information.
BugCheck 3B, {c0000005, fffff96000087dd4, fffff88002a10ed0, 0}
[COLOR=red]Probably caused by : memory_corruption[/COLOR]
Followup: memory_corruption

From a strict BSOD perspective, the memory corruption is more concerning than anything else. If there is a ram problem, it can cause a BSOD at any time doing any thing and whatever driver is most in use at the time will usually get the blame. So, from my perspective, if there is a memory problem, nothing in the dump files can be reliable. Let me show you. The second one today that blamed the Nvidia driver, when I dig down into the files, this is what it shows as the cause.

fffff880`092aa590  fffff880`00c87800Unable to load image \SystemRoot\system32\DRIVERS\amd_sata.sys, Win32 error 0n2
*** WARNING: Unable to verify timestamp for amd_sata.sys
*** [COLOR=red]ERROR: Module load completed but symbols could not be loaded for amd_sata.sys
 amd_sata[/COLOR]+0x1800

fffff880`092aa018  fffff880`04bea29eUnable to load image \SystemRoot\System32\Drivers\dump_amd_sata.sys, Win32 error 0n2
*** WARNING: Unable to verify timestamp for dump_amd_sata.sys
***[COLOR=red] ERROR: Module load completed but symbols could not be loaded for dump_amd_sata.sys[/COLOR]

Those are the causes of that dump and no where does it mention the Nvidia driver that was originally listed. There were some references to the Nvidia Display Driver, the NTFS file system the AMD sata, which are all mostly related to the hard drive. In all fairness the last one, dump_AMD_SATA is simply a way for the system to make room for the dump file to be written.