Loud Crash but no BSOD = Graphic Card?

[play]

Starting this thread in case someone else has the same problem (or someone knows more than I do).

About a week ago, in the middle of watching a movie, the computer froze on a frame, with a buzzing sound at top volume. No way to quit the application, the computer had crashed. But no BSOD.
After rebooting, same crash. But it happened after five minutes.
Reboot, re-crash, without even starting a movie. (So unrelated to the player.)

The next day, on cold boot, the computer would run for about an hour. Then crash (but no BSOD). If while watching a video, with the same horrible noise.

To start troubleshooting, as a random starting point, I swapped out the graphic card (EVGA GeForce 9500GT) and put one in from four years ago. So far, today, the computer has been running for ten hours without crashing. Fingers crossed.

Does this sound like symptoms anyone else has seen?

Wishing you all a happy 3rd of Jan.

ASUS P8P67 Deluxe / unknown case / Intel i7-2600K / Corsair 750HX
16GB Corsair Dominator / Noctua NH-D14 / 128GB Crucial m4 SSD
2TB Seagate Barracuda Green / EVGA nVidia GeForce 9500GT
ASUS DRW-24B3ST 24X / Asus VH242H 23.6" / Logitech X-230
Windows 7 Home Premium SP1 64-bit OEM edition
Virtual OS: OSX Lion in VMWare, Linux Mint 11 (Katya) in VMWare

 

[palladium]

How do you mean "crash"? Did the screen go blank, or did the computer stopped responding...? One starting point is to browse through event viewer and look for any warnings/errors that correspond to the time & date of the crash. You should also configure Windows to not automatically restart during a failure (right click my computer -> Advanced system settings ->Startup and Recovery ->uncheck automatically restart and make sure you choose either small dump or kernel dump under debugging information). If you don't do this sometimes Windows may just restart when it BSOD-ed (but it won't show you that you had a BSOD).

Having said that if you swapped out the gfx and its still working fine, sounds like the card may be giving you trouble. Are the temps OK? If you have evga's lifetime warranty you can try RMA-ing the card.....

Also I'm assuming you are not overclocking anything...if you are then you should reset all to default first.

 

[notty22]

Take a look at the card you took out. Look for bulging or exploded capacitors.
They are the little aluminum looking cans on the card.

 

[play]

Thanks palladium, thanks notty22.

Look for bulging or exploded capacitors

Ah, that would have been exciting. But no, they look fine.

did the computer stopped responding

Correct, no response.

I'm assuming you are not overclocking

Correct.

browse through event viewer

Yes, I did that yesterday. Though I didn't do anything with the information.
There was one critical error related to not properly shutting down. (Event 41). That made sense to me as I had to flip the switch.
The other one was (according to my notes from yesterday) "Error: Event ID 10; Source: WMI". Just googled it, thanks for nudging me. Sounds related, people one the threads mentioning freezes, and one mentioning an NVIDIA problem.

Thank you for sharing your troubleshooting skills.
We watched a movie on the old card tonight, no trouble.

Will report back after a while to update on the status.

Wishing you both a fun day.

 

[BathroomFeeling]

I used to get this. I solved it by removing the card, blowing on the slot, then reseating it again. I test by playing Super Mario Bros 3. Everything's cool so far, so you should try it. Good luck!

 

[taltamir]

What you are describing is actually pretty common crash type, and indeed suggests a probelm with the video card.
You already solved it by swapping out the video card, now try to RMA the broken one. If it is out of warranty, trash it.

 

[play]

Hi BathroomFeeling, hi taltamir.
It's great to get some confirmation that the symptoms describe a problem with the video card. And a relief.
Thank you.

Wishing you both a beautiful day.

 

[play]

Hi all!

Sadly, after swapping out the graphic card, the big freeze happened again. It usually happens after a movie plays for a while, or when the computer has been on for a while. The temps look okay (35 motherboard, 50 cpu). But since it seems to happen more on hot days and when the system has been running, still wondering about temp.

I tried all three slots for the graphic card. Same problem.

What else could be causing this symptom?
SSD? RAM? Motherboard?

After putting all this time in picking the components and building the system, and only enjoying it for three months, it is stressful to have it fail already.

 

[birthdaymonkey]

My first guess would be RAM. It seems to just "go bad" more than other parts. If you have another set, try that. If you don't, try just running with one stick at a time. You could also try Memtest86+, which involves downloading it, making a boot disk, and running it overnight to search for errors.

 

[play]

Hi birthdaymonkey,

Thank you so much for your very helpful message!
I was really glad and relieved to have clear direction for the next troubleshooting task.

I downloaded Memtest86+ (thank you for mentioning that wonderful utility to me) and ran two passes... But no errors. So the RAM is probably not the culprit. In a way, I'm glad to hear that as two sticks are lodged under the Noctua cooler.

Wondering in which direction to go next.
Worried about a possible motherboard of CPU fault... Are there utilities like Memtest86+ to test them?
Or could a hard drive problem (SSD?) cause these symptoms?

Wishing you all a beautiful weekend.

 

[BathroomFeeling]

Did you update all your drivers? Like motherboard, chipset, sound, video, etc..?

 

[BD231]

Some fed up neighbor brought me a seemingly dead computer not to long ago, said after a little while of use it would just freeze. Being that it runs fine for a while I assumed something was just overheating and I was right, northbridge thermal paste had dried up on the motherboard. Make sure all heatsinks are checked and not getting to hot. It's a new board so I doubt any paste dried but best to be safe and check for any thermal issues with the sinks on your motherboard.

Remove the SSD as the boot drive, boot from mechanical and see if maybe you've got a faulty SSD on your hands.

NVidia drivers are definitely known to cause hard locks so be sure and do a clean install of the drivers . I notice some guys with i5's/i7's are having low voltage issues at idle clock speeds, wherein the cpu just isn't getting enough power when it down clocks. These are all just suggestions though, good luck and if I think of anything I'll be sure and post it.

 

[birthdaymonkey]

If the RAM is truly OK (and passing Memtest86+ isn't always a 100% guarantee of that), then the next most likely suspects are the SSD and motherboard. As the SSD is more easily isolated, do as the previous poster suggested and take it out of the equation. Try running off a standard hard drive for a while and see if you still get the lockup.

I'd still suggest running with just one stick of memory, or even just two sticks. Using four DIMMs is sometimes less reliable (more load on the memory controller... idk exactly why) than using one or two.

Something else you could do as a troubleshooting step is running a Linux live CD (or flash drive)... take Windows out of the equation. Just go about your normal activities without any drives hooked up (web surfing, etc.), and see if the crash happens. If not, that increases the chance that it's a software rather than a hardware problem.

If none of this stops the crashes, then you'll have to find a way of getting another motherboard to try.

Troubleshooting is all about isolating things. Make the system as simple as possible and add stuff back in till the problem surfaces, or vice-versa (take stuff out till it stops).

HTH... good luck!

 

[play]

Thank you all so much for sharing your troubleshooting insights!

After implementing some of the steps, the machine is now running and has been playing videos in a loop all afternoon. If it crashes again, will try the next steps.

To document the process, here are the steps I tried from your last recommendations:

Phase 1
- Checked fins of heatsink for excessive heat. (None detected.)
- Updated BIOS from version 1606 (11/4/26) to 2103
- (As a follow-up, had to tweak BIOS screens to fit the RAM spec)
- Checked chipset in CPU-Z, ran chipset updater, checked in CPU-Z again: no change.

At this stage, the computer FROZE again.

Phase 2 (half an hour later)
- Updated GEFORCE driver from 280.26 to 285.62

This seemed to do it. So far so good!
If it turns out to be the Nvidia cards... Well, I'm waiting for a cheap Radeon card (as this is not a gaming rig), and I can't wait to switch.
// HIS Radeon HD 5450 Silence 1 GB

What is mysterious to me is that the machine didn't freeze for the three first months after building it. Why then would the driver cause a problem only after three months? This puzzles me and makes me wonder if something else might be going on.
If so, we will know soon enough.

Thank you all a million times for your kind guidance through this difficult time.

Wishing you a fun day,

 

[Lean L]

Ive had this problem with a build I had in the past. It drove me insane since the crashes were always random and occured when the computer was not even being used.

Long story short, it was the nvidia chipset for the motherboard. I'm actually glad nvidia cant make chipsets anymore. That computer has caused me some amount of pain.