Sign in with
Sign up | Sign in
Your question
Solved

Random Freezes and crashes, GTX 460 (fixed) heat problems now...

Last response: in Graphics & Displays
Share
December 2, 2010 7:20:55 PM

My problems began last night while playing bad company 2. The screen would lock up and freeze, while the sound would do a "crackly" loop. So I shut down the computer and restart it, and now I am getting stuck on my logon screen. The computer will restart fine in safe-mode, however locks up as soon as i type in my password when starting normally. I assume that there is some sort of service that is locking it up.
So right now, I have all services and startup items disabled and it starts fine.. I am going one by one through each service and re-enabling it to try to find the service causing problems..
I was wondering if any of you have experienced this before, and if so, what service was causing the issue?

Last night I was able to get back to my desktop twice without disabling any services. Everything appeared normal, however when I got back in the game, I'd be able to play for anywhere from 5-15 minutes, then I'd get the freeze and trouble logging back into windows after that.


System Specs:

ASUS P7P55D-E Pro - LGA 1156 - Intel P55 - DDR3 - USB 3.0 SATA 6Gb/s - ATX Motherboard
Bios Version: Factory (0806 I think?)

EVGA GeForce GTX460 768MB DDR5 PCI-Express 2.0 Graphics Card 768-P3-1360-TR
Running Driver: 260.99_desktop_win7_winvista_64bit_english_whql

Corsair XMS3 4 GB PC3-12800 1600Mhz Dual Channel Core i3 i5
i7 DDR3 Memory Kit CMX4GX3M2A1600C9
Running manual timings at 9-9-9-24

Intel Core i5 750 Processor 2.66 GHz 8 MB LGA1156 CPU I5-750BOX

Western Digital Caviar Blue 500 GB Bulk/OEM Hard Drive 3.5 Inch, 16 MB Cache, 7200 RPM SATA II WD5000AAKS
Recently did a clean install to try to fix this issue

XFX 750W Black Edition (P1-750B-CAG9)

I have done some research and found this thread:
http://www.evga.com/forums/tm.aspx?m=586183
as well as others on the evga forum about people experiencing random freezes with their gtx 400 series cards.

I am also only running with a stock heatsink and fan, and have been for the past 4 months. I was told that this could damage my card and I should get a cpu cooler ASAP.
My CPU temp in BIOS is 47.5 C
My Mobo temp in BIOS is 29 C

I also have a thread started at http://www.evga.com/forums/tm.aspx?high=&m=716637&mpage... with a few more details.

Any ideas to what is causing these random crashes?
a b U Graphics card
December 2, 2010 8:19:03 PM

Is this only happening when your gaming? Otherwise it is stable?

You might want to check on your graphics card temp too, see if it is overheating.

First guess, I would reinstall the nvidia drivers. If the heat is fine and the driver update did nothing, then... I'd have to troubleshoot more based on exactly when the instability happens.
m
0
l
Related resources
December 2, 2010 8:26:22 PM

This started happening when I was gaming..
Upon restart, it started happening all over the place, randomly.
Sometimes it wouldn't get past the post, sometimes it wouldnt get past the welcome screen, and sometimes it would lock up on desktop.

It seems to work ok in safe mode, so I disabled all services and started enabling them one by one.
I have mostly everything enabled now except the special graphic services for windows 7.
what program can I use to check graphics card temps?
I just did a clean install of drivers to version 260.99
m
0
l
December 2, 2010 8:56:04 PM

i was able to play 1 round in bad company 2 before it crashed..
all I noticed is the processor got up to 81-82 degrees..
Would running it this hot for hours on end this weekend cause issues that would be affecting me now?
m
0
l
a c 592 U Graphics card
December 2, 2010 9:30:06 PM

Your CPU is getting too hot. Consider taking off the heatsink and reinstalling it to ensure a good seal. For a more advanced technique: take off the heatsink, clean the CPU and heatsink surfaces with isopropyl alcohol, apply new thermal paste, then reinstall the heatsink. Finally, as a last resort, get a new heatsink if the first two suggestions don't work.
m
0
l
December 2, 2010 9:38:13 PM

17seconds said:
Your CPU is getting too hot. Consider taking off the heatsink and reinstalling it to ensure a good seal. For a more advanced technique: take off the heatsink, clean the CPU and heatsink surfaces with isopropyl alcohol, apply new thermal paste, then reinstall the heatsink. Finally, as a last resort, get a new heatsink if the first two suggestions don't work.


Cool, I had my girlfriend pick up some thermal paste.
So the best way is to dis-assemble, clean surfaces with isopropyl alcohol, apply new past, and reinstall?
How much thermal paste should I apply? and what technique is best for applying?
Thanks
m
0
l
a c 592 U Graphics card
December 2, 2010 10:01:59 PM

You would be wise to look that up on the web. There are various techniques. The key is to use a very small amount. You can put a pea-sized amount in the very center of the CPU and let the pressure of the heatsink spread it out. You can also lightly spread the paste over the entire surface of the CPU. Some people draw lines with the paste on the CPU. The type of heatsink does make a difference for the technique. My preference is to spread the paste out thinly and evenly over the entire surface, that way you make sure it covers all the way to the edges.
m
0
l
December 2, 2010 10:13:34 PM

here's a link to an article that shows how to properly apply thermal paste depending on heatsink design. it also tackles on how to properly apply TIM on a design similar to Intel's rather anemic stock solution.

http://benchmarkreviews.com/index.php?option=com_content&task=view&id=170&Itemid=1&limit=1&limitstart=3

you almost have the exact same build as the one i built for a friend, except his VGA is a bit less powerful (9800GT 1GB) and only has a 650W PSU. and his CPU barely goes past 59*C on load. and that's with the stock cooler. but then again, he has a Thermaltake V3 midtower case, so heat simply goes up the 2 120mm fans on top of his case.

hope that helped. :) 
m
0
l
December 2, 2010 10:30:43 PM

thank you very much for the link!

I have a Haf 922 case, which has a huge fan on the front and top, and a huge air intake on the side.. the case should not be the issue.

also to be of note: it froze again, yet i was still able to open and close the CD drive.. Seems like if it was completely frozen, it wouldn't respond when I push the button to open it..
m
0
l
a c 214 U Graphics card
December 2, 2010 10:39:59 PM

Try taking the card out and run with the onboard graphics if you have it.See if the problem persists
m
0
l
December 2, 2010 10:42:18 PM

I don't have onboard graphics with my Asus P7P55D-E Pro
m
0
l
a b U Graphics card
December 2, 2010 10:42:58 PM

There is no reason a stock 750 should go to 82c while gaming.
You need to fix this, the heatsink prongs have to be unseated or something else very wrong.
m
0
l
December 2, 2010 10:59:32 PM

hmm. if you have an older PCIe video card lying around, you can try that first.

of course, it could also be the RAM.
m
0
l
December 2, 2010 11:07:57 PM

all my extra cards are so old they arent even compatable..
I wish I had an extra so I could rule out the vid card
m
0
l
December 2, 2010 11:31:20 PM

ouch.

well, you can try re-seating the RAM first. see if that helps.

m
0
l
a b U Graphics card
December 3, 2010 1:10:57 AM

As everyone has mentioned your CPU shouldn't get to 80C. I have the same CPU and even when gaming, like a 4v4 SC2 game with max units, my CPU maybe gets to 65C.

For sure try re-applying the heatsink/fan to the CPU. Once you've done that, you need to check to make sure your CPU fan is spinning properly.

Also, another tip, go into your BIOS and go into setup. Once there, you can change your fan settings to automatic for your CPU fan and also your other case fans. The manual for your MB shows how to find it if you are confused.
m
0
l
a b U Graphics card
December 3, 2010 1:16:57 AM

Also, just to give you an idea, here is a screenshot of my system. You can see the fan speeds and processor temperature. This is with setting my bios fan speed to auto.

m
0
l
a c 172 U Graphics card
December 3, 2010 1:19:36 AM

I didn't see any thing about a after market cooler so make sure that every push pin is properly snapped into place is you are using such a cooler. Also replace the stock compound that intel uses and they really haven't changed much in that respect in the days of the Pentium 2. Dries up all the same and performance is terrible and is easily out performed by even the cheapest compound out there.
m
0
l
December 3, 2010 1:34:34 AM

ok, so I took all the parts out of my motherboard, took out the processor, applied new thermal paste, reset everything (Processor, heatsink, ram and video card), reconnected all of the wires. Wow, doing that improved my CPU temps BIGTIME!
I now idle at about 35 C
However, my freezing problem still persists. It seems that when I have minimal services running that it prolongs the freezing problem, but it will still lock up. It is VERY hard to disable and re-enable the services to find if it is one of those, because it doesn't seem to be any specific service because it's so random..
I really don't think it's a memory issue, because I ran memtest overnight and it ran 8 1/2 passes with 0 errors. But it ended up freezing on the 9th pass at 50%... So I dont know what that means exactly..
m
0
l
a c 172 U Graphics card
December 3, 2010 2:11:18 AM

Run memtest and if you don't know and no one here offers up how to load it then look it up for better help. It basically test your ram to see if it is faulty and bad ram could be the cause of your problems.
m
0
l
December 3, 2010 2:43:24 AM

I did end up running memtest. see last sentence in my last post :-P
m
0
l
a b U Graphics card
December 3, 2010 12:40:13 PM

Still monitor your CPU/GPU temp before it crashes. Do you have dual monitors? If so, keep HWMonitor open and load your intensive game until it crashes. Keep an eye, not on the game, but on your temps so you can quickly write them down. It may be that while it is cooler in general, it is still overheating when pushed.

How long have you had this graphics card installed? How long have you had your computer? If you could give some background on when this started as a comparison to when you started using the comp and if you installed new hardware, that would help a lot.
m
0
l
a c 172 U Graphics card
December 3, 2010 12:41:58 PM

If I or others haven't suggested start looking for hotspots in your system. The card and the board.
m
0
l
December 3, 2010 2:20:42 PM

My Cpu temps got to about 68 while stress testing with 3dmark06
My GPU temps got to 43 I think..

I have had the graphics card installed long as my computer has been built which was at the end of July.
Everything was PERFECT after first building, and I never ran into any problems until now, where I get the random freezes...
I haven't installed any new hardware.
But I did take out my graphics card, blew it out with air and reset it before problems began.. I also blew out my heatsink and psu... The thing that is so weird is that Right now I've been running stable for about 5 minutes with no crash.. sometimes it's 5 seconds, sometimes it's 20-30 minutes.. I don't know what would cause this crap...

What are hotspots to look for? Spots in the case where there's no air circulating?
m
0
l
December 3, 2010 2:51:25 PM

So I don't know if I'm just being foolish or not, but I just ordered a new GTX460. I will wait to send the old one back until I get the new one and can verify that it is in-fact the card.. If it isn't, then I have a SLi setup :-).. And then I can probably rule that it's the memory..
Right now the system has been stable about 30 minutes.. So I don't know what to think about it all
m
0
l
a b U Graphics card
December 3, 2010 4:09:04 PM

Run a scan disk for any errrors. I can imagine that all those abrupt shutdowns could have caused some issues on the HDD level.

Maybe you've fixed it with the heatsink and the last shutdown was a system error and not a overheating. Post again if it crashes or when you get the new GFX card.
m
0
l
December 3, 2010 8:56:27 PM

I ran a scan disk with no errors, but it froze before getting to the bios post screen after restarting from it.. It runs stable enough between crashes for me to type these posts.. but sometimes i get 1/2 through typing my reply and it freezes.. COMPLETELY random...
m
0
l
a c 592 U Graphics card
December 3, 2010 9:17:52 PM

Maybe your are having a problem with your power supply?

The other thing, are you running Windows 7? Win 7 is particularly demanding on your hard drive. It does a lot of background tasks, indexing, prefetch, superfetch that will reveal flaws in an otherwise good hard drive. I had a similar problem that manifested in excessive hard drive activity, but also resulted in random freezes. My first remedy was to turn off indexing and superfetch. Scandisk could not detect any errors. Ultimately, when I replaced my hard drive, all the problems disappeared and everything worked a lot better and faster.
m
0
l
December 3, 2010 9:24:43 PM

17seconds said:
Maybe your are having a problem with your power supply?

The other thing, are you running Windows 7? Win 7 is particularly demanding on your hard drive. It does a lot of background tasks, indexing, prefetch, superfetch that will reveal flaws in an otherwise good hard drive. I had a similar problem that manifested in excessive hard drive activity, but also resulted in random freezes. My first remedy was to turn off indexing and superfetch. Scandisk could not detect any errors. Ultimately, when I replaced my hard drive, all the problems disappeared and everything worked a lot better and faster.


that's an interesting thought. I am running windows 7 64bit. However like a week before I started getting crashes I was optimizing my PC and disabled prefetch, superfetch and indexing. I also currently have them turned off in services for now.
m
0
l
December 4, 2010 2:41:49 AM

my latest memtest revealed an unbelievable number of errors..


where do i go from here?
m
0
l
a c 172 U Graphics card
December 4, 2010 2:43:45 AM

I KNEW IT!!!!!

m
0
l
December 4, 2010 2:45:42 AM

so what do i do? RMA the RAM?
m
0
l
a c 172 U Graphics card
December 4, 2010 2:47:04 AM

Give then either a call or email with their support tech with either the company that made the ram or ware you bought your ram.
m
0
l
December 4, 2010 2:49:07 AM

its Corsair XMS3.. i bought it though amazon.. I'll call them tomorrow first thing (hopefully they'll be open since it's a weekend)
m
0
l
a c 172 U Graphics card
December 4, 2010 2:53:02 AM

I have done some shopping with amazon my self. So long that you have the original amazon box the box that the ram came in you will likely get a refund. I once got a double refund lol. :) 
m
0
l
December 4, 2010 2:55:21 AM

***.. I'd be hard pressed to find either...
m
0
l
a c 172 U Graphics card
December 4, 2010 3:02:02 AM

Call the manufacturer and if they ask for a prof of purchase then send a digital copy of the receipt. Even if you lost the physical receipt then you can use screen capture for the one from amazon and save it a png or jpeg and forward that to them so long that it doesn't compromise any private information. They should honor their warranty and ship you a replacement kit once you have sent in your old one. Some will in advance send as replacement kit ahead of time that you can reuse the package to send them the defective kits.
m
0
l
December 4, 2010 3:18:47 AM

So I took out each stick to try to isolate which one was causing issues.
I ran the test on 1 stick, it had 48 errors.
I ran the test on stick 2, it had 0 errors..

So I put only the stick in with no errors, and it still freezes on me..
m
0
l
a c 172 U Graphics card
December 4, 2010 3:21:58 AM

Chances are that memtest is missing something or there may be an external issue. It is possible that something may be damaging your ram. Did you overclock at one point or having unusual voltages or have suffered power spikes or brawn outs?
m
0
l
a c 592 U Graphics card
December 4, 2010 3:26:19 AM

c0d3ster said:
that's an interesting thought. I am running windows 7 64bit. However like a week before I started getting crashes I was optimizing my PC and disabled prefetch, superfetch and indexing. I also currently have them turned off in services for now.

Sounds like bad memory is the latest avenue you are pursuing, but I wanted to mention that disabling those services did not fix my random freezing problem. The only to fix it was installing a new hard drive. The main symptom was that I noticed near constant hard drive activity when the machine was sitting idle.
m
0
l
December 4, 2010 2:46:20 PM

well I just adjusted the timings on my ram to 8-8-8-24.
I also adjusted it to be set at 1333mhz.

I am running a memtest right now, and it made it all the way through the test with 0 errors on the first pass of all the tests. I will let it run 2 -3 more passes to see if it yields anything, but maybe this fixed it.. I will test for freezes after the memtest is done..
m
0
l
a b U Graphics card
December 4, 2010 2:53:36 PM

Did you set the ramVoltage to 1.66 ?
By default most m/b's set it at 1.5, my ram - m/b can barely get to the setup screen at that voltage for the ocz 1.65 v ram I have. If that was not set , it lowers the clock speed ability of the ram, also timings.
If you having freezing , it would not hurt to set to 9-9-9 -31
To see if something stops the freezing. ram errors
m
0
l
December 4, 2010 3:12:01 PM

so you suggest to bump the voltage to 1.66, and then set the timings to 9-9-9-31?

currently I have it set at 1.65 with timings of 8-8-8-24.
It just made it through the second pass of memtest with no errors..

I am relatively new to what RAM timings and all this voltage stuff means and how it is related. Lets see if I have this correct. If I up the voltage to 1.66, would it allow the RAM to run a faster timing of 9-9-9-31?
m
0
l
a b U Graphics card
December 4, 2010 3:22:47 PM

9-9-9- 31 is 'slower'
more conservative.
Thats why I found it strange you went from 9 in the upper posts down to 8.
Faster timing give slightly ever so performance, but not if there is errors. For testing, eliminating potential problems, put it up to 9 9 9 31 it might even allow that other stick to pass.
m
0
l
December 4, 2010 3:29:41 PM

notty22 said:
9-9-9- 31 is 'slower'
more conservative.
Thats why I found it strange you went from 9 in the upper posts down to 8.
Faster timing give slightly ever so performance, but not if there is errors. For testing, eliminating potential problems, put it up to 9 9 9 31 it might even allow that other stick to pass.


haha, that's funny, because I set it to 8-8-8-24 thinking that it was "slower"

I will test for desktop freeze and get back to you..
m
0
l
December 4, 2010 4:39:02 PM

So it appears as though everything is running stable...

Unless I try loading a game, or youtube, or anything that requires some sort of video processing power.

I am able to browse the web without crashing and I can listen to internet radio without crashing.. If I leave my internet radio running, and then try to open youtube, it freezes, yet the internet radio keeps playing.

Does this seem like a video card issue?
m
0
l
a c 592 U Graphics card
December 4, 2010 7:25:19 PM

Whatever you do, do not increase the RAM voltage above 1.65. A Core i5/i7 has 1.65 as the upper limit. Above that is possible, but you risk damaging your CPU. For now, it's best to run at stock speed, voltage, and timings.
m
0
l
!