Solved Random BSODs

[iSupes]

Hi all

I apologize for the very generic title, however these really have been random with no discernible cause.

PC Details:
MSI GT70 0ND Laptop
i7 3610QM
16GB RAM (it's speced with 12GB, I added another 4GB)
nVidia GTX 675M
120GB SSD (2x60GB in RAID) + 700GB HDD
Running wonderfully for about 2 months, last week or so I've had BSODs

BSOD Details:
Like I said, they are random, the only consistent one I can reproduce is after the laptop has been turned off (or put in sleep mode) for at least half an hour. When I boot up it gets past the BIOS, shows the beginning of the Windows animation, then immediately BSODs. This happens between one and 4 times (even booting into Safe Mode or Recovery causes this) and then it boots normally.

This got me thinking it might be something fan or heat detection related as obviously the laptop is still cold when this happens, the one experiment I tried was 'cold' booting, going directly into the BIOS, waiting 10 minutes, then rebooting and there were no BSODs during boot up, probably dumb luck, but mentioning it here regardless.

Other times it seems completely random, playing Wow/GW2, doing work in NetBeans/Office/etc, even just idling.

What I've done so far with no success:
- Removed the extra RAM
- Ran Memtest with and without the extra RAM (no errors)
- Uninstalled VGA/Network/Soundcard drivers
- Updated all drivers available on MSI's website
- Updated the VGA drivers from nVidia's website (this was after trying the latest from MSI's website)
- Updated the laptop firmware and BIOS from MSI's website
- Done a power cycle on the laptop (left battery out for an hour)
- Recalibrated the battery (via a tool from MSI)

I also downloaded WinDbg and had a look at some of the dump files, but I simply don't have the expertise to determine what they are saying.

Reinstalling Windows is my next step, but having discovered this site and the number of successful solves here, I thought I'd try my luck here first

Thanks in advance for any help.
Supes

 

[Arc]

Welcome aboard.

► nVidia Dispaly Driver .... the latest one is problematic.

lmvm nvlddmkm
start             end                 module name
fffff880`0f4f0000 fffff880`101e4000   nvlddmkm T (no symbols)           
    Loaded symbol image file: nvlddmkm.sys
    Image path: \SystemRoot\system32\DRIVERS\nvlddmkm.sys
    Image name: nvlddmkm.sys
    Timestamp:        Tue Oct 02 23:51:13 2012 (506B3099)
    CheckSum:         00CD4613
    ImageSize:        00CF4000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4

• Download and install Driver Fusion.
• Reboot the computer in Advanced Boot Options, safe mode. Search Driver Fusion in your start menu, and remove all components of your nVidia display driver.
• Boot normally now. Go to Drivers - Download NVIDIA Drivers, Drivers > Beta and Legacy; search there with your cards particulars for GeForce 306.23 Driver, dated 13.9.2012 and install it.

► Intel Rapid Storage Technology. As it is crashing, get rig of it.

1. Right click on the My Computer icon> Select manage. It will open the computer management window.
2. Click on the "Device manager" in the left pane.
3. Expand Disc Drives, by clicking the triangle in front of it..
4. Select. Right click > Uninstall them all, one by one.
5. When Done, restart the computer.
6. On restart, default msahci.sys driver will be auto configured. .

► Take care of your Storage devices.

1. Update your SSD's Firmware
2. Reseat the sata and power.
3. Run chkdsk /f/r, following the option two of the tutorial Disk Check
4. Seatool for dos: SeaTools | Seagate download
   Burn it in a blank cd. boot from the CD, click on "Accept", wait for it to finish detecting the drives, then in the upper left corner select "Basic Tests", then select "Long Test" and let it run.

► Uninstall the following programs:

• Start Menu\Programs\MSI
• Start Menu\Programs\MSI\Battery Calibration
• Start Menu\Programs\MSI\BurnRecovery
• Start Menu\Programs\MSI\MSI Software Install

► Free up the startup.

1. Click on the Start button
2. Type “msconfig (without quotes), click the resulting link. It will open the System Configuration window.
3. Select the “Startup” tab.
4. Deselect all items other than the antivirus.
5. Apply > OK
6. Accept the restart.

Let us know the results.

 

[Britton30]

Also it may be a good idea to check your hard drive.
Some hardware tests to run.



Hard drive(s):
Pick one for your HDD brand or use Seatools.
7 Free Hard Drive Testing Software Tools

 

[iSupes]

Update:

Firstly, thanks so much for getting back to me such detail.
Secondly, did everything you said Arc (except update SSD firmware, MSI uses it's own 'Super RAID' config) and so far no BSODs

I have only tried to recreate the first kind of BSOD ('cold' booting) and it's been good so far (done this test twice, once after waiting half an hour and a second time after an hour).
I am going to be doing some stress testing now including a driver verification.

The last test I will be doing is leaving the laptop off over night, if I have no issues tomorrow morning I will most happily mark this as resolved

@Britton, I have run both chkdsk and SeaTools on my disks and there were no problems

 

[Britton30]

OK, did you run the long tests on all drives?

 

[iSupes]

Hi Britton

Yeah, ran it on the HDD and both SSDs.

While I'm posting, ran driver verification and it booted into Windows with no problems and been running multiple applications (games, IDE's, etc) simultaneously since my last post and also, no issues.

Thanks again guys

 

[Arc]

Update:
Firstly, thanks so much for getting back to me such detail.
Secondly, did everything you said Arc (except update SSD firmware, MSI uses it's own 'Super RAID' config) and so far no BSODs

Good to know I also expected so, after rollback of display driver and removal of intel rapid storage.

Inform us how the system works later.

EDIT : Dont run stress tests if those are not needed. I hope you need not to run any stress test right now. If driver verifier is needed at any point of time, we will mention it , then and there.

 

[iSupes]

Update:

I left the laptop off for over 2 hours, booted up and received a BSOD, this occurred 3 times before Windows loaded.
I decided at this point to reload Windows (using a recovery partition) the reload went fine and quicker than expected, I followed your instructions again Arc, just to be sure no problematic drivers were being installed automatically.

Left the laptop off for another half an hour and when I booted I did indeed receive a BSOD. I'm guessing this is actually a hardware issue. The other annoying thing is that these BSODs don't generate dump files (I made sure Windows was setup to generate Kernel memory dumps after reloading Windows). The auto reboots also happened so quickly that I couldn't get a good look at the error (I have disabled auto reboot now and will try to get a pic of the error tomorrow morning).

Not sure if it's pertinent, but these BSODs occurred in the evening, after the temperature had dropped a good 15 degrees (Celsius) so the laptop was a fair bit colder than when I experienced the apparent success this afternoon.

Again, I really want to thank you for the help you've given me so far, it is most appreciated.

 

[Arc]

I am asking for help to a person from whom I learn those techniques, I hope he will post his opinion shortly.

 

[Britton30]

Update:

Firstly, thanks so much for getting back to me such detail.
Secondly, did everything you said Arc (except update SSD firmware, MSI uses it's own 'Super RAID' config) and so far no BSODs

The MSI Super RAID config is the SATA controller on the motherboard. SSD firmware is a chip built in all SSDs. Old firmware may cause BSODs at times. It should be updated from the SSD manufacturer's site.

 

[iSupes]

Update:

Booted the laptop this morning and received a BSOD, the stop code generated was:
STOP: 0x0000007E (0xFFFFFFFFC000000, 0xFFFFF08882EA7E11, 0xFFFFF88003B5582, 0xFFFFF88003B55080)
Just the usual BSOD text along with that, nothing else. Oddly only one BSOD occured before booting normally into Windows.

The SSDs(link) used in this RAID config are not supported by the SanDisk SSD Toolkit (does not even pick them up). I have contacted SanDisk and MSI asking about updated firmware.

 

[Arc]

iSupes, what you exactly need here is to increase your RAM voltage very slightly.

But I am not very good in those stuff, so I cannot suggest you how to do it exactly, and how much to do it.

This situation is called Cold Boot Crash. I asked Dave76 to look at this page, and Britton30 also can guide you in this regard.

But, as a whole your system is getting a bit low voltage.

 kd> !sysinfo cpuspeed
CPUID:        "Intel(R) Core(TM) i7-3610QM CPU @ 2.30GHz"
MaxSpeed:     2300
CurrentSpeed: 2294

 

[iSupes]

Thanks for the update Arc.
The laptop BIOS doesn't support any voltage changing, but I'm still looking into ways of doing this. The only option in the BIOS that might be related was Intel SpeedStep, I've disabled that and will test cold booting again.

I've also downloaded CPU-Z and had a look at the info there, I noticed that the RAM frequency of the extra module I installed differs by about 100MHz from the other 3 (latency is the same for all of them) could this be a possible cause for the cold boot crashing?

 

[Arc]

Might be, I am not very sure. About hardware and overclocking stuff, I dont have a sound knowledge. Let us ask Britton30.

 

[Britton30]

Yes mis-matched RAM can cause issues. I would remove the extra you installed for now and test. Later if it works out you can get a matched 16GB kit.
You're right OEM machines leave hardly any adjustments for a user.

 

[iSupes]

Alright, extra module removed, waiting for it to cool off and will let you guys know how it goes.

Thanks again

 

[Britton30]

When you ran memtest did you let it run for 7 passes?
You're welcome mate.

 

[iSupes]

I let it run overnight, when I returned it had said complete.
I didn't change the config or anything, I assumed it tested everything automatically, should I rerun and change anything?

 

[Britton30]

That sounds like the Windows built-in memory test, memtest 86+ runs till the user stops it.

 

[iSupes]

Hmm, maybe I was just imagining things, will redo a memtest with and without the extra module.

I ran it off a USB stick instead of writing an ISO to CD, that shouldn't make a difference should it?