Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
Loud fan noise from computer
My computer wasn't so loud but today its quiet loud even if im not doing anything. I kind of ignored and while i was gaming i noticed a significant fps drop. When i restarted the game the fps was back to normal but it got me worried about the systems stability and the fan noise. Is the loud fan noise from the cpu? I looked inside the case through my side panel window and I see that some cables are touching video card. Is that okay?
Comments
Sounds like your fans are spinning up to full speed because its overheating. Turn it off, unplug the computer, discharge any remaining power by pushing the start button a few times, open the computer and see if its dirty.
Also if your videocard has a sticker on its fan, check to see if it fell off and is stuck in the fan blade (hate those things, haha).
Velika: City of Wheels: Among the mortal races, the humans were the only one that never built cities or great empires; a curse laid upon them by their creator, Gidd, forced them to wander as nomads for twenty centuries...
It sounds like something could be overheating. If you listen carefully, you should be able to figure out which fan is causing the noise, which will help you isolate the cause.
Alternatively, you could do a stress test to see directly if something is overheating. For the processor, download SpeedFan and Prime95. Run SpeedFan, then run Prime95, and have it do a stress test, with one thread for each core. Watch the temperatures on SpeedFan. If anything goes over 90 C, then shut it down immediately, as that's really bad. If you get the same sort of noise as before, then you've probably found the problem. Otherwise, report the highest CPU temperature that Prime95 can cause.
For the video card, download FurMark and run it. You can do a stress test with FurMark and see how hot the GPU gets. If that gives you the same noise as before, then you've found the problem. If it goes over 100 C, then shut it down immediately, as that's really bad. Otherwise, report the highest GPU temperature that FurMark can cause.
Now, FurMark or Prime95 might well make your system get somewhat noisier, as those are stress tests. But if your processor peaks at 70 C and your video card at 80 C, then that's normal and safe.
I assume this is the new computer that you just built last week? That shouldn't have had time to have dust accumulate yet.
There's also the possibility that a fan is hitting a cord, which makes a horrible noise. Moving the cord fixes it.
Modern PC's have multiple cooling fans. Mine has four case cooling ones, a power supply one, a CPU one and two graphics card ones. The only way you'll be able to diagnose which one of yours is the noisy one is to take the side panel off and investigate. However, the graphics card and CPU ones are very important. If they are failing, some quite expensive damage can be done to the components they protect. The fact that you are asking this question tends to suggest you are not that conversant with the inner workings of a PC. You should probably take it to a repair shop to have it checked out ASAP.
You can use the tube from a roll of paper towels as a rough form of stethoscope.Point it at each fan and have a listen.
Quizzical for the prime95 software, don't I have to run it for 24 hours for the tes? Also I opened the case and nothing seems to have a problem. Also, I tried running tropics demo( a graphic benchmark tester) and it freezes in the beginning. The sound contiunues to come on . I tried this like 3 times and it all freezed so i had to reboot.
Instead of Speedfan, does CPU thermometer work? RIght now I turned on prime95 for liek a minute and it says all 4 cores are 100% and the temperature reads 58.6, it seems to slowly creep up
I think its staying around 58 degrees quizzical. So im guessing its a problem with the gpu. Should i stop the prime95 stress test? Im also downloading furmark right now. I hope to god that i didn't any bad parts.
So i just opened furmark. Do I do a burn in test? A benchmark?
If it only hits 58 C after a few minutes of Prime 95, even with a separate thread on each core, then the processor temperature is fine.
What Prime95 really does is test to see whether certain very large numbers are prime (hence the name). The algorithm that does this happens to push the processor very, very hard, which is one reason it is used as a stress test.
Another useful stress test property of the program is that it is a very long string of pseudo-random computations, with a known final result. If at any point, a single bit gets flipped wrongly, then the final result is nearly guaranteed to be wrong. That lets the program flag that, while the computer didn't crash, the processor isn't perfectly stable, and being able to find that within hours is useful information to someone who is overclocking the processor. If a processor makes an error in one computation out of a quadrillion, then Prime95 is good enough to pick that out if you leave it running overnight. If you're not overclocking, then you don't particularly care about this feature of the program.
The other feature of the program, which was the original purpose, is testing large numbers to see if they're prime. Whenever there is an announcement that someone has discovered a new largest known prime number that is millions of digits long, it was found using Prime95.
You could try FurMark to see if the GPU is overheating. You might want to first make sure that you're not blocking any air vents in the video card, and thus giving it a reason why it should overheat. If the video card doesn't overheat or at least get really noisy as you were complaining of, then it could just be a bad fan somewhere.
Sorry for posting so much but i want to give information as I'm doing it. Well I ran furmark and did a burn-in test. I saw the temperatures and they slowly rose to about 86 degrees when it froze again. So the gpu was the problem. What is wrong with my gpu?
I think the best way to resolve this issue is to remind everyone what machine we helped you build recently:D i remember seeing your request for help but dont really remember what thread we gave all the advice in:D
once we know the parts, we may be better able to diagnose the probable cause of the problem. as i remember it, your machine is new so dust is out of the question. include what games you are playing and what software have you installed recently that may have a negative impact in the system. (namely the softwares you installed before the problem started)
Hey psyclum thanks for helping out last time. My computer specs are
CPU : AMD phenom2 955 3.2
Ram: Crucial 4096MB PC10600 DDR3 1333MHz (2x2048)
PS: XFX PRO650W Core Edition Power Supply 80+ Bronze
Mobo:ASUS M5A87 AMD 870 Socket AM3+ Motherboard
Harddrive: WD Caviar Black 1TB SATA HD 7200/64MB/SATA-6G
Optical Drive: ASUS 24X DVDRW SATA OEM
Graphic card: XFX Radeon HD 6870 1GB GDDR5 PCIe, Dual DVI & HDMI
Sound card: Asus Xonar DS 7.1 PCI Sound Card
Oh also, I noticed that my gpu is directly above my sound card. Should that be a problem. There is no where else to plug in my sound card so if that is a problem how should i fix it? ( Probably not though right?)
Oh yea i forgot about the softwares. I play starcraft 2 and world of warcraft for games. The things I downloaded in my computer are the stress teseting applications like Prime95, Cpu Thermometer, Tropics gpu benchmarking, Furmarks.
If the video card took a few minutes to hit 86 C, then that's perfectly normal, and not overheating. If it hit 86 C within a few seconds and then froze, then I'd worry more. If it's taking a few minutes for FurMark to make the system freeze at 86 C, then that could be a heat or power problem of some sort, but the GPU itself isn't overheating. It's possible that you had the bad luck to get a defective video card or power supply, and may need to RMA it. I'd think the video card would be the more likely culprit there.
You could try pulling the sound card out and seeing if that makes any difference. Try uninstalling your sound drivers, removing the sound card, and seeing if FurMark still crashes the system. It might be that the sound card drivers are the real problem here, and everything else is fine.
Also, I hope you've installed Catalyst 11.6 or whatever the latest is, and not whatever video drivers came with the card. Sometimes video cards ship with early drivers that may be unstable.
In the ATI catalyst control center. There are options where I can adjust clock speeds I think i changed that around. I turned back to the factory settings and im going to test it again.
The gpu seems to be fine after i changed the settings back to the factory settings. XP Quizzical thanks again for all your help as well as other people in here too. You guys are really helpful. Also for the graphic card driver, do i have to uninstall the currernt gpu driver in safe in order to install the new driver from the website?
I agree, from reading everything in this thread so far, it's almost certainly the video card.
If I were you, as new as it is, there's no harm in just RMA'ing it and not even fooling around with it.
Aside from that, shut the computer down, and unplug and pull the video card out of the computer. Spin the fan with your fingers, make sure it doesn't have anything hanging up in there (a dust bunny, a gum wrapper, anything else), or otherwise feel as if it has any resistance to it when you spin it with your hand. Make sure all the vents are clear and plug it back in.
When you boot up again, make sure you absolutely aren't doing any kind of overclock on it. If you have no idea what I'm talking about with that, you aren't overclocking it. If you do, go back to stock clocks with whatever program or BIOS tweaks you used to overclock.
Re-run Furmark and see where the crash happens again. If it keeps crashing, and you've checked to make sure the fan is ok and the vents are clear, your only options from that point are to RMA the card, or get an aftermarket video cooler and replace it yourself (and hope that the card isn't damaged aside from the heatsink/fan assembly).
Nope, you can install on top of the old driver.
That sounds like it's nothing more than an unstable overclock. The higher you set the clock speeds, the lower the temperatures it takes to crash it. That's actually true further than you might think; at sufficiently high clock speeds, the GPU might be stable under -100 C (note the minus sign) and crash above that. For the record, the stock speeds on a Radeon HD 6870 are 900 MHz core and 1050 MHz memory.
not to sound like an arse, but specifically what do you mean by "not doing anything"
are you sure you didnt have something running in the background like one of the benchmark tests?:D if you were playing around inside catalyst and overclocked the card, it can very well explain the fan going crazy. do keep in mind that the heatsink on stock cards arent really made for too much overclocking. if you decided that you wanted to max out the clock on your gpu... you could have done some damage:D
just set the card back to factory and hope for the best:D
My computer is super quiet now
Could take something as simple as using a vaccum to remove lint from the vent holes on the side. Fans wear out some time and are cheap and easy to install generally. Is your comp shoved back into a small spot where no air can reach it? Is it setting in a very hot spot in the room? Put a room fan on it a while.
Man Of Constant Sorrow
I bet you anything you set it to manul fan speed and cranked it up. Past about 25% those things start sounding like a hair dryer. Not hurting anything if that is what you did, just crank the speed down until the noise is at a tolerable level.
"Gypsies, tramps, and thieves, we were called by the Admin of the site . . . "
He overclocked it too far without realizing what the clock speeds did. Card should be ok, I don't think the Catalyst drivers will let you melt the card down like some BIOS tweaks will, but it definitely will let you drive the card beyond stability.
Excessive fan speed could be from taking manual control (Catalyst will let you do that), or it could just be a symptom of the overly aggressive clock settings too.
Either way, as long as taking it back to default stock speeds fixed everything, looks like a cheap lesson learned.