Seemingly Random Lock-ups

[jsbailey]

Hi all,

I've had a recurring problem for the last few months that I've just ignored but I finally have decided to attempt to determine the problem and fix. Any help you can offer, either in how to determine what is causing the problem or general solutions to hardware problems, is greatly appreciated.

The Problem:

My computer will randomly freeze/lock up and can only be recovered via hard reset. I have noticed that it tends to happen most frequently when playing a game on my main screen and try to do something on my second monitor (i.e. typing in skype, web browsing, etc.). My suspected culprits are video card and/or RAM.

System Specs

• Main monitor (Samsung XL 2370) center, second (Samsung E2420L) to my left.
• Windows 7 Home Premium
• i7 2600k
• 16GB G.SKILL Ripjaws Series DDR3 1600
• EVGA GTX 570
• Gigabyte Z77X-UD5H
• Corsair HX850
• 128 GB Crucial RealSSD C300 (Windows Drive)
• 120 GB OCZ Vertex 3 (Games Drive)
• 1TB WD Black (Storage Drive)

Thanks in advance for any help. Let me know if you need any additional info.

[denis280]

did you check the temp.

[jsbailey]

Is there a way I can see temps at time of crash, other than having a monitor on screen when it happens? I highly doubt that is the culprit, as I'm not doing any OC and have a Noctua NH-D14 on the processor, but I can try to see what they're at for the next one.

I downloaded a program called WhoCrashed and this was the report it created:

On Sat 5/26/2012 12:52:55 AM GMT your computer crashed
crash dump file: CWindows\Minidump\052512-10623-01.dmp
This was probably caused by the following module: win32k.sys (0xFFFFF960000A000E)
Bugcheck code: 0x50 (0xFFFFDB00E9CB48C0, 0x0, 0xFFFFF960000A000E, 0x7)
Error: PAGE_FAULT_IN_NONPAGED_AREA
file path: CWindows\system32\win32k.sys
product: Microsoft® Windows® Operating System
company: Microsoft Corporation
description: Multi-User Win32 Driver
Bug check description: This indicates that invalid system memory has been referenced.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
The crash took place in a standard Microsoft module. Your system configuration may be incorrect. Possibly this problem is caused by another driver on your system that cannot be identified at this time.

On Sat 5/26/2012 12:52:55 AM GMT your computer crashed
crash dump file: CWindows\memory.dmp
This was probably caused by the following module: win32k.sys (win32k!EngBitBlt+0xAF7E)
Bugcheck code: 0x50 (0xFFFFDB00E9CB48C0, 0x0, 0xFFFFF960000A000E, 0x7)
Error: PAGE_FAULT_IN_NONPAGED_AREA
file path: CWindows\system32\win32k.sys
product: Microsoft® Windows® Operating System
company: Microsoft Corporation
description: Multi-User Win32 Driver
Bug check description: This indicates that invalid system memory has been referenced.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
The crash took place in a standard Microsoft module. Your system configuration may be incorrect. Possibly this problem is caused by another driver on your system that cannot be identified at this time.

On Fri 5/25/2012 10:38:34 PM GMT your computer crashed
crash dump file: CWindows\Minidump\052512-6162-01.dmp
This was probably caused by the following module: tcpip.sys (tcpip+0x3A58B)
Bugcheck code: 0xD1 (0x1, 0x2, 0x1, 0xFFFFF88001C3C58B)
Error: DRIVER_IRQL_NOT_LESS_OR_EQUAL
file path: CWindows\system32\drivers\tcpip.sys
product: Microsoft® Windows® Operating System
company: Microsoft Corporation
description: TCP/IP Driver
Bug check description: This indicates that a kernel-mode driver attempted to access pageable memory at a process IRQL that was too high.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
The crash took place in a standard Microsoft module. Your system configuration may be incorrect. Possibly this problem is caused by another driver on your system that cannot be identified at this time.

[jolancer]

as denis280 pointed out it may just be over heating... If it has no airfilters and never cleaned it may just be dust balls inside inhibiting airflow through the Heatsinks or PSU. cleaning it would fix it IF that is the issue. also what PSU are you using?

if its not that dirty or doesnt fix the issue... you maybeable to recreate it to isolate were its coming from.. sorry i'm not up on the latest stress testing apps but these are a couple i use to use.

i use to use Prime95 for CPU stress testing, dont know if theres other versions out now but if its still single threaded as it use to be.. you just need to create multiple shortcuts for the proagam on your desktop and then rightclick it go to /properties and -A1 -A2 etc to the Target lines as example bellow, use as many instances as you have CPU Threads to fully load it. and in Prime just go to /options/Torture test

> 1st Instance Shortcut Target: "CProgram Files\Prime95\Prime95.exe" -A1
> 2nd Instance Shortcut Target: "CProgram Files\Prime95\Prime95.exe" -A2

memtest86 - http://www.memtest.org/ will test your Ram maytake some time but just let it run

Im sure theres tones of apps now for GPU stress testing, Im not up on any though Someone esle can probably recomend them.. but for example the best ones use to be those that have all the modern latest hardest to run DX filters enabled. For example back in the day that was High Res HDR based rendering back when DX9 was popular that ap was rthdribl.exe

[Dave3000]

You say you are not overclocking, but still you are running outside the specifications of your i7-2600. Set your RAM to 1333. I say this because i7 2600 and all other SB CPU's only support up to DDR3-1333. Maybe the memory controller in your i7-2600 can't handle DDR3-1600.

[jsbailey]

Sorry to necro the thread, but I did some of the things mentioned and made some updates to my system and still encountered the error last night. Wanted to see if any of the updates gave anyone any new ideas:

I reformatted my old SSD's (Vertex 3 & C300) and added 2x 3TB HDD's (Raid 1) and a 256GB Vertex 4 as my new primary drive. After installing Windows, all my games, and all the windows updates, I installed the old SSD's and HDD's and started playing a game. It was during playing I encountered the hard lock-up again (screen freezes doing whatever I was doing and I get an ear-splitting continuous noise from the speakers until I reset).

I had sort of hoped that the problem was something with my Vertex 3 (old OS drive), although I knew it was a long shot. I also took some compressed air to the system (it definitely needed it) but that also does not seem to have solved it. I highly doubt that it's a heat issue. I'm not feeling any warm air from the system and at this time of year especially, it's very cool in my house.

I think my next move is the memtest. Someone suggested I do each stick individually, is that going to make any difference?

Also, I just wanted to note that I say I'm not OC'ing even with my RAM at 1600 for a few reasons. I don't really see it as OC'ing (although technically you are correct) because all I did was turn the XMP mode on, which is supported by the mobo and the RAM, and I was under the impressed the 2600k supported it too. That's a possibility, I suppose, but it seems unlikely to me. But, what do I know, that's why I'm here

[Smoove910]

I'm thinking the ram is a good place to start... try downclocking to see if it helps. You could also remove all but one stick and start testing with Memtest. I seem to remember something about 1600mhz ram requiring 1.65v to run properly since it was essentially rebadged/overclocked 1333 ram. I could be wrong, but maybe try giving some more juice if you are running it at 1.5. Don't know the exact specs of your ram, so don't know for sure.

[jsbailey]

So I finally got around the running the memtest last night. Unfortunately, because I've got a Noctua NH-D14 sitting on the CPU, if I want to remove/change memory, I've got to remove that, make my change, and reinstall the fan. For this reason, I ran memtest with my memory in it's current configuration, and this is the result:

http://img19.imageshack.us/img19/467...1207220743.jpg

How beneficial would it be at this point to go through the task of testing each individually? At this point, I'm starting to suspect GPU or PSU over RAM, but I'm sure someone more technically literate than myself has better thoughts?

[AnonymouseUser]

One pass of Memtest is not usually enough to consider RAM fully tested. Looping overnight is recommended, and I've had RAM that ran 3 or 4 passes before giving just one or two errors which is enough to consider it bad RAM.

Originally Posted by jsbailey

It was during playing I encountered the hard lock-up again (screen freezes doing whatever I was doing and I get an ear-splitting continuous noise from the speakers until I reset).

Regarding this statement, I would suspect your onboard audio. Make sure you have the latest drivers installed, and see if there are any compatibility issues with that particular game. You may even want to disable onboard audio in the BIOS and try to replicate the crash in game again (yes, no sound). If it crashes without sound then it will rule out the audio.

[jsbailey]

Originally Posted by AnonymouseUser

One pass of Memtest is not usually enough to consider RAM fully tested. Looping overnight is recommended, and I've had RAM that ran 3 or 4 passes before giving just one or two errors which is enough to consider it bad RAM.

Regarding this statement, I would suspect your onboard audio. Make sure you have the latest drivers installed, and see if there are any compatibility issues with that particular game. You may even want to disable onboard audio in the BIOS and try to replicate the crash in game again (yes, no sound). If it crashes without sound then it will rule out the audio.

Thanks for the info. I'll let the memtest run overnight tonight and see if it reveals any further information.

Let me give some additional info on the audio portion of the crashes. It happens independent of what game I'm playing but dependent of what audio device is my current default device. I have a Logitech G35 headset that I'll use (USB) when I'm playing and talking with people using a various VoIP client. If I'm not, then I'll switch my default device to my speakers (using onboard audio) and take the headset off. Whenever a crash occurs, if there was any sound currently playing (game, music, etc.), I will get the noise I made a horrible attempt at describing on whichever device is set to the system default device. To my credit, the noise is indescribable and probably the worst sound I've ever encountered. It makes nails on chalkboard or watching TV static on full blast look easy. I'll check for driver updates on both ends. Hope some of this helps.

[jsbailey]

Ran memtest for 8 hours. 3 passes, no errors. Should it go longer?

Also, I backed my RAM down to 1333 a few days ago, and have since encountered the error. >.< This is getting frustrating.

[AnonymouseUser]

Originally Posted by jsbailey

Ran memtest for 8 hours. 3 passes, no errors. Should it go longer?

Also, I backed my RAM down to 1333 a few days ago, and have since encountered the error. >.< This is getting frustrating.

That should be enough.

Since the game crashes with both onboard and USB sound, it's probably not audio, so let's focus on video.

With a temp monitoring program (eg, CPUID's Hardware Monitor) running on your secondary monitor, play a game for a few minutes. Watch the GPU temps to see if they rise and how hot it gets before it crashes (if at all). Switch to a program on your secondary monitor like you usually do when it crashes, and see if any of the temps rise. Make note of the temps just before it crashes, if at all.

[jsbailey]

Did some gaming tonight, nothing intense just LoL. Max recorded GPU temp was 56C(132F) and voltage ranged from 0.913-0.975. Max CPU temp was 42C(107F). Max mobo 35C(95F). Didn't experience a crash.

Something to note is that when I do crash, everything locks up and I have to hard reset, so I can't switch to the second monitor and look at things, so if I don't have HWMonitor on top on the second monitor I can't see the temps.

I'm not 100% certain that the noise element happens when my default output is the headset, I'm trying to recall from memory as I haven't used the headset much lately. I'm fairly certain, but my memory is horrible.

[mlc]

1) I didn't see mention of the psu being used? what's it rated for on the 12v rails? That graphics card requires a lot of juice on the 12v rails i think (... and make sure you have both 6 pin psu connectors plugged into that card, assuming it has 2

2) Assuming the psu is sufficiently rated, etc... did you rule out a dual monitor related issue and try running with just one monitor?

3) did you check to ensure you have the latest graphics card and mobo/chipset drivers installed? I would think if it's locking up. .that it could be a driver problem, assuming you've ruled out the gpu temp as an issue...

Mike

[jsbailey]

Originally Posted by mlc

1) I didn't see mention of the psu being used? what's it rated for on the 12v rails? That graphics card requires a lot of juice on the 12v rails i think (... and make sure you have both 6 pin psu connectors plugged into that card, assuming it has 2

2) Assuming the psu is sufficiently rated, etc... did you rule out a dual monitor related issue and try running with just one monitor?

3) did you check to ensure you have the latest graphics card and mobo/chipset drivers installed? I would think if it's locking up. .that it could be a driver problem, assuming you've ruled out the gpu temp as an issue...

Mike

1-Corsair HX850

2-I have not ruled this out, but I could certainly try. Would hate not to be able to run two monitors though.

3-All drivers were updated when I reformatted a few weeks ago.

[mlc]

psu looks solid...

I understand your concern about it not wanting to be an issue using dual displays.. but if you can rule out that the dual display usage is triggering the problem, that may help to narrow down where the problem lies...

1) try removing one monitor to see if you can reproduce the problem...

2) check with EVGA to see if there are any firmware updates available to address this.. If the card is still under warranty, it may be worth pursuing this problem with them as well.. as the GTX-570 isn't exactly a trouble free design.. so there may be an RMA in your future...

3) I would also check to see if the bios for your motherboard is current..