Sign in with
Sign up | Sign in
Your question

PC crashes randomly mostly during a game without a BSOD - Help the newbie :-D

Last response: in Windows 7
Share
May 22, 2014 6:09:32 PM

Hey guys I'm new here and I've always just read through things on this forum but the time has come for me to sign up and actually post a problem.

Out of nowhere I started noticing my PC crashes a bit more than usual. It was normally when I'm playing a game or watching a video (yes i'm thinking it's the graphics card but please read on :-D ). But around the same time I noticed that whenever installing a game or software I'd get a barrage of CRC errors. So I panicked and decided to do a clean install of Windows 7 anyway.

Now that's done the PC definitely doesn't crash randomly anymore but whenever I play counter strike source (i'm old school was just for a bit of fun) it would freeze up and the audio would crash with a loop of the last audible sound like a gunshot and the only thing that could help would be hitting the restart button. Bear in mind there was never a BSOD whatsoever during the crashes though occasionally it would restart by itself and as it's booting up a BSOD would flash for a split second and restart again! I was wondering how would be the best way to investigate this.

I'm willing to accept one of my hardware is definitely faulty but I have no real idea of how to check! Here are my PC details:

MOBO: MSI 870 C-45
Processor: Phenom II x4 955 black edition
Kingston ram: 8gb (2 x 4Gb)
Atapi CD/DVD RW
Maxtor hard drive 250 GB
GPU: XFX Radeon HD5750
Windows 7 Ultimate edition

Anyway if someone could give me some steps to perhaps find if there are any dmp files I can share with you? I used Whocrashed but only to find that it didn't detect any minidmp files. I also ran Malawarebytes just for safe measure but nothing came up ( did a fresh OS install anyway ). I've also run Memtest for my ram and everything came up clear with 2 passes. But whats best to test my hard disk and graphics card?

Hope you geniuses can help!

JK
May 22, 2014 6:23:43 PM

You didn't post you PSU brand and model. Lots of things can cause this: bad PSU, bad GPU, mobo dying, bad HDD etc.
Since memtest is ok i think motherboard an ram are probably ok. This leaves the GPU PSU and HDD.
Use HDDTune for HDD
Run Furmark for GPU (5-10 mins is enough)
Run Furmark + Prime95 at the same time for PSU
m
1
l
May 22, 2014 6:24:18 PM

The most common culprit for a previously perfectly stable system mysteriously starting to crash/reboot without warning is a failing PSU.
m
0
l
Related resources
May 22, 2014 6:25:50 PM

Thanks for the quick reply. Sorry missed out the power supply:

Coolermaster GX550W

I'll get on testing the hardware in turn right now and get back with some updates.
m
0
l
May 22, 2014 6:26:46 PM

HDD with bad sectors can also cause instability. In case all of the above tests work. Run a benchmark of Unigine Valley for a couple of passes. If that fails its most likely the GPU.
m
0
l
May 22, 2014 6:29:28 PM

I had this same problem with my graphics card (different card, mind you). I figured out that the reason that this was happening was because the card was overheating. Could you check your temp while playing a game? A good monitoring software is GPU temp (I believe your card should be supported). If you do this and see that your card is reaching higher temperatures than is recommended for your card then it's definitely a cooling issue. The fan on my card turned out to be completely broken so I replaced it with a fan from an old PSU
m
1
l
May 22, 2014 6:29:48 PM

The CoolerMaster GX (apart from the 450W version) are rated as tier-4 PSUs, which is not catastrophically bad but not particularly good either.
m
0
l
May 22, 2014 6:33:29 PM

Testing the HDD will take a long time, its normal. Don't do a quick test as it will most likely not find any damaged sectors.
m
0
l
May 22, 2014 6:37:03 PM

Sorry I wasn't sure how to post a report up? Screenshot?

Also guys how do I quote more than one reply in one post?
m
0
l
May 22, 2014 6:45:04 PM



That can't be good?
m
0
l
a b $ Windows 7
May 22, 2014 6:47:00 PM

The drive is failing. Back it up and replace ASAP.
m
1
l
May 22, 2014 6:47:31 PM

HDD is dead/dying. Replace it then do the other tests. It is known for failing PSUs to damage HDDs so your PSU might also have problems. Replace HDD then go to step 2 and 3 in my first post.
m
1
l
May 22, 2014 6:49:56 PM

Thanks for the help guys. I'll buy one asap. Don't suppose I could run the benchmarks on the graphics card and test out the PSU whilst I'm waiting for a new Hard drive?
m
0
l
May 22, 2014 6:54:20 PM

You can but results wont be conclusive. The PC might restart because of corrupted data like installing furmark on a bad sector will cause the pc to BSOD when entering furmark or while running it.
m
0
l
May 22, 2014 7:12:23 PM

I gave it a go on Furmark and it scored 920 (15FPS, 60000 ms) I checked to compare and the HD5750 wasn't on the chart but a HD5770 scored 1078 (17FPS). So it's not too far off really. I'll get a hard disk soon anyway and I'm a tad worried the PSU is dying which essentially I've got a get a new hard disk, PSU AND graphics card (was planning the last one anyway).
m
0
l
May 22, 2014 7:18:13 PM

Don't forget to try the last step. Furmark WHILE running prime95 - it will stress the shit out of the psu. If there are any problems with the PSU the system will restart/shutdown (hopefully)
m
0
l
May 22, 2014 7:36:48 PM

emdea22 said:
You can but results wont be conclusive. The PC might restart because of corrupted data like installing furmark on a bad sector will cause the pc to BSOD when entering furmark or while running it.

Most of the time, bad sector simply cause Windows to freeze for several seconds while the HDD retries reading the bad sectors until it succeeds or gives up. If the HDD gives up / times out, Windows usually remains locked up. Sometimes, there are also toaster notifications about delayed writes failing.

Corrupted data rarely makes it out of the HDD because it gets caught by the HDD's PRML encoding on the magnetic media and ECC/FEC block coding applied to the data itself. In most HDD manufacturers' specs, the likelihood of undetected read errors is less than 1 in 10^14. (The likelihood of undetectable errors is less than the likelihood of uncorrectable errors.)
m
1
l
May 22, 2014 7:40:50 PM

InvalidError said:
emdea22 said:
You can but results wont be conclusive. The PC might restart because of corrupted data like installing furmark on a bad sector will cause the pc to BSOD when entering furmark or while running it.

Most of the time, bad sector simply cause Windows to freeze for several seconds while the HDD retries reading the bad sectors until it succeeds or gives up. If the HDD gives up / times out, Windows usually remains locked up. Sometimes, there are also toaster notifications about delayed writes failing.

Corrupted data rarely makes it out of the HDD because it gets caught by the HDD's PRML encoding on the magnetic media and ECC/FEC block coding applied to the data itself. In most HDD manufacturers' specs, the likelihood of undetected read errors is less than 1 in 10^14. (The likelihood of undetectable errors is less than the likelihood of uncorrectable errors.)


Well bad HDDs did cause more than one of my PC's to BSOD and restart during the years.
m
0
l
May 23, 2014 4:25:11 AM

readysetmosh said:
I had this same problem with my graphics card (different card, mind you). I figured out that the reason that this was happening was because the card was overheating. Could you check your temp while playing a game? A good monitoring software is GPU temp (I believe your card should be supported). If you do this and see that your card is reaching higher temperatures than is recommended for your card then it's definitely a cooling issue. The fan on my card turned out to be completely broken so I replaced it with a fan from an old PSU


That's a less likely possibility as I think the fan's loud as hell but works it's ass off anyway. But anyway i'm going to get a r9 290x anyway. Thanks!

InvalidError said:
The CoolerMaster GX (apart from the 450W version) are rated as tier-4 PSUs, which is not catastrophically bad but not particularly good either.


Yup I'm learning a lot the past 10 hours. So I'll probably get a new PSU in a day or two anyway as 550w might struggle with a r9 290x. I had no idea about these tiers until now so I'll do my research and purchase carefully.

InvalidError said:
emdea22 said:
You can but results wont be conclusive. The PC might restart because of corrupted data like installing furmark on a bad sector will cause the pc to BSOD when entering furmark or while running it.

Most of the time, bad sector simply cause Windows to freeze for several seconds while the HDD retries reading the bad sectors until it succeeds or gives up. If the HDD gives up / times out, Windows usually remains locked up. Sometimes, there are also toaster notifications about delayed writes failing.

Corrupted data rarely makes it out of the HDD because it gets caught by the HDD's PRML encoding on the magnetic media and ECC/FEC block coding applied to the data itself. In most HDD manufacturers' specs, the likelihood of undetected read errors is less than 1 in 10^14. (The likelihood of undetectable errors is less than the likelihood of uncorrectable errors.)


I didn't get any problems using Furmark with the faulty HDD so yeh ur right.

So guys I've just bought a Samsung 840 EVO 250 GB which should do for now. Right now I'm really hoping that's what was letting me down. I think next week I'll buy a new PSU anyway perhaps a 800-850w Corsair/EVGA and see how it goes.

Thanks for all your help I hope this solves it all. My rig is coming up to almost 3 years old and when i built it i think my biggest shortcut was the hard disk. I think i expected it wouldn't last long but it slipped my mind over the years. I'll keep you posted!
m
0
l
May 23, 2014 5:26:52 AM

Hi_im_JK said:
Thanks for all your help I hope this solves it all. My rig is coming up to almost 3 years old and when i built it i think my biggest shortcut was the hard disk. I think i expected it wouldn't last long but it slipped my mind over the years. I'll keep you posted!

Three years is still relatively young for a HDD IMO. My living room PC has a 12 years old 60GB HDD in it... I wonder how many power-on hours it has logged so far. I probably have a few that passed the 50k hours mark but no longer use on a regular basis. (Stuffed in external USB/FireWire boxes that I now use only a few times per year... but used to use a lot more often than that.)

The main HDD in my current system is four years old and has logged 30k power-on hours so far with all SMART counters still come out good and 0 reallocated sectors. Same for my older 500GB drive but that one has only logged 12k hours.
m
0
l
May 23, 2014 7:06:46 PM

InvalidError said:
Three years is still relatively young for a HDD IMO. My living room PC has a 12 years old 60GB HDD in it... I wonder how many power-on hours it has logged so far. I probably have a few that passed the 50k hours mark but no longer use on a regular basis. (Stuffed in external USB/FireWire boxes that I now use only a few times per year... but used to use a lot more often than that.)

The main HDD in my current system is four years old and has logged 30k power-on hours so far with all SMART counters still come out good and 0 reallocated sectors. Same for my older 500GB drive but that one has only logged 12k hours.



Even so I genuinely can't remember where I purchased the HDD - all i can say is it wasn't anywhere reputable and it may have even been second hand/refurbished. It was meant to be temporary and it really slipped my mind. anyway I'm not actually experiencing any crashes anymore but I think it's time for a SSD anyway so shall start there then maybe stress the PSU as i never knew i purchased one that was so far down the tiers.
m
0
l
!