Crashing Problem, buzzing sound.

ted · 13 Nov 2013

Hi guys,

sorry in advance, it seems like this question has been asked to death, but I've found a ton of different answers and i'm just a little overwhelmed as to what to do.

so, a little background, built this computer about 4 years ago. it didn't have too many issues, except that for a while it would suddenly shut off while i was playing some more graphically demanding games. turned out this was overheating, and it was especially bad since i was in texas at the time and it was a really hot summer. after reapplying the thermal grease on the cpu, i haven't had the problem since.

then, for a good while (a few years) no major issues,

anyways, lately these days i've been having this problem where the computer just freezes- and a buzzing sound comes out of the speakers. it shows the last frame, and i'm sure the buzzing is the last sound being looped over and over again. no mouse movement, nothing-- have to do a hard shutdown.

at first the issue seemed to only appear when i played games that strained the computer more. but the issue seems to have gotten worse. once it crashed while i was just using an internet browser. another time it crashed in the "welcome" screen. at one point, i used ccleaner, and defragged, and it seemed to mitigate the issue, but it's gotten worse again.

anyways i'm just wondering, how can i isolate the problem? what steps should i take? would a system restore help? i'm really hoping it's not a hardware problem, but a driver thing or something. would it make sense, that it'd be hardware? i mean it's worked for 3 years, why would it start now?

here are my specs:
Windows 7 Home Premium
Service Pack 1
rating: 5.9
Intel Core 2 Quad CPU q9550, 2.83 GHz, 2.17GHz
RAM: 4gb (3.25 usable)
32 bit
(the hard drive is around 1 tb i think, but it's nothing special, no solid state. the motherboard is from Gigabyte)

also, sorry if this is the wrong section! i was going to post in the BSOD one, but i never get a BSOD from the issue.

thanks,
ted

TVeblen · 13 Nov 2013

Hi Ted. Welcome to the forum.

Yes, hardware can crap out pretty much anytime, but it is more likely when the hardware is new, and when it gets old. So unless you know of a cause and effect symptom regarding drivers - you installed a new driver (or new component) and then the problems started - then you must consider hardware failure as a likely possibility.

Symptoms like yours have a couple of likely suspects: Overheating, RAM, and the Video Card. You could do some tests of these to start with.

For overheating try this: Take the sides off the case and set up a house fan to blow air directly at the case. Run the PC like that for a while. See if the symptoms go away.

For RAM, run a diagnostic. Memtest86+ - Advanced Memory Diagnostic Tool is the best. Start the test before you go to bed and check the results in the morning. The more passes it does the better the test. No errors is the only acceptable result.

For the video card try a clean reinstall:

Go to the ATI or nVidia website and download the most current known good driver and save it to an easy to find location.
Now go into Windows Explorer and in the C: drive find the nVidia or Program Files > ATI folder (inside will be Drivers > your driver version) and delete it (the whole folder).
Go into Start > Control Panel > Remove a Program and uninstall all programs for the video card. For nVidia, do PhysX and the Vision drivers first, then the main driver. The control panel will uninstall with the driver. For ATI, select “ATI Catalyst Install Manager” and click on “Change”, then “Uninstall All Components”. Restart the computer.
When it reaches the desktop Windows will find new hardware and will install it's own WDDM1.1 driver. Let it. You will be asked to restart. Do it.
Once back on the desktop you can now install the nVidia/ATI driver package for your card.

Both ATI and nVidia now have "Clean Install" options in their driver packages. These work well but skip the reinstallation of the generic Windows driver. In a diagnostic situation this is important. You can run the PC on just the Windows generic driver for a while to see if it makes any difference.

Let's start with that basic stuff and see what shakes loose.

rvcjew · 15 Nov 2013

Just so know I've had issues where bad sticks pass memtest with flying colors, do you have any errors in your event log? Also if you can afford it I would start by replacing the ram to save you the headache of it possibly being bad and maybe upgrade to 8gb in the process, what is your motherboard model?

Edit: looks like your using 32bit windows and 4gb is all your going to get.

ted · 30 Nov 2013

hey guys, sorry for not replying for a million years, that was very uncool of me. i know youve probably been anxiously awaiting a reply haha.

anyways, so i updated my nvidia drivers, and the problem seemed to go away. a few weeks later, the problem came up again-- check the nvidia website, drivers were outdated again, so updated again- problem fixed again! so it seems to only happen when the nvidia drivers are out of date.

is this weird?

TVeblen · 01 Dec 2013

Yes. But I believe that this can be a case of wishful thinking.
The problem seems to go away but then it happens again could just mean that the duration between events just happens to fall between some actions you take. It is like a sports fan that notices that their team scores every time they let the dog out thinking that letting the dog out makes their team score.

But if your observations turn out to be correct, then it indicates a defective video card.
The best test for determining if the card is defective is to run the system with a different video card. Easier said than done. When I needed to do this in the past (when I did not have spare cards lying around) I bought a "cheap" $50 card to test with. If the system ran well with a $50 card but not the $250 one then I could be pretty sure that the big card was defective.

Did you perform the other 2 tests? They don't cost anything but your time.

Memtest is not perfect - it is not the be all and end all of RAM testing, but it is the very best initial test for performance issues and picks up the most basic problems - the ones that account for the majority of RAM failures.
There are other RAM tests to perform, but always start with Memtest.

ted · 01 Dec 2013

ah, well, i think you're right, it just crashed again last night.

i opened the case, used a little fan on it-- it didn't seem to make difference. i also ran memtest and it said the ram was okay. so if it is the videocard, is the only way to fix it to buy a new videocard? i'll have to see if i can get my hands on a cheapo one to test it out.

TVeblen · 01 Dec 2013

Unfortunately, yes. There is very little that the average user can do, unless you are an electrical engineer. You can look for anything obvious and do a few checks:

Poorly fitting power connectors
Remove and re-seat the card (PC power disconnected - proper static precautions)
While the card is out, use compressed air to blow out the casing, especially around the fan.
Look for (or just blow out) any dust or hair in the PCI-E slot

Not much beyond that without a solder gun.

EDIT: Poor power could contribute to it. If you have a multitester you could check the 12v feed from your power supply for steady power.