new problems

Minato

Minato

Minato

Associate
Joined
29 Aug 2008
Posts
1,113
Location
London
Well, those of you who saw my epic 12 page thread to get my pc up and running will know i've had quite a few troubles with it. After finally getting fed up with this one and not being able to wish it away i'm posting again.

Basically, whenever i try to play any game, it plays fine, normally, but after a while it just goes black for a couple of seconds, then i go to the desktop with the game minimized. I can't get back into the game unless i wait for around 5 mins. this normally happens in CoD4 and CoD:WaW, but with CoD4 it happens after around 1-3 mins, CoD:WaW takes around 5-10 mins. Crysis just crashes at launch straight away. Call of Juarez took around 20+ mins for this to happen. Aion has only given me this once after quite a while of play.

I've never had this with any programs, just games.

It started when i started having this INFURIATING nvlddmkm.sys BSOD's. after ages of hassle, and a near abandonment of pc gaming, i did a complete fresh install of windows, i installed some old nvidia drivers (the ones on my gfx card cd) and it ran ok for a bit. then i started recieving nvlddmkm.sys has stopped working and recovered, so basically the same problem except without the BSOD's.
Now, somehow, and i don't know how, that's stopped happening, but the above problem has been happening for a while. I seriously prefer gaming on my beautiful pc, rather than my ps3.

any help, greatly appreciated.
 
Hi Minato, sorry to hear you are having more problems :(

My suggestion would be to monitor CPU and GPU temps during the games and before crashes (if at all possible), that would help to rule out component overheating.

Do you have anything running overclocked at the moment?

beefybarn said:
Whats are the components you are running?
Click to expand...

If memory serves: i7 920 D0, Gigabyte UD3R, Kingston DDR3 6GB, GTX 285, Corsair TX 650W. How close am I?
 
cmndr_andi said:
Hi Minato, sorry to hear you are having more problems :(

My suggestion would be to monitor CPU and GPU temps during the games and before crashes (if at all possible), that would help to rule out component overheating.

Do you have anything running overclocked at the moment?



If memory serves: i7 920 D0, Gigabyte UD3R, Kingston DDR3 6GB, GTX 285, Corsair TX 650W. How close am I?
Click to expand...

perfect, and thanks for the suggestions. I do have my old 9600gt, but i don't think it would serve as an accurate representation because i probably wouldn't be able to play the games at the same settings, but if my memory serves i didn't have a single fault with that card.

edit: no overclocks, i told myself i would only oc after i knew all was stable...which never really happened =).

do you guys know of any good programs to moniter temps for a bfg gtx 285?
 
Last edited:
downloaded them and gave it a shot, idle seemed to be around 65 degrees, when i ran fumark, with the furry doughnut thing, according to the graph at the bottom it went to 90 degrees. i only ran it for a few minutes though, but no crash it seems.

Edit: the fan started spinning madly, and the gpu was making audible noise. this hasn't happened when playing the games, so i don't think it reached these levels of heat?
 
Do you have EVGA Precision to increase the fan speed of your GTX 285? If it's a cooling issue, it might help you. 65C idle sounds pretty toasty to me.
 
I just did some reading on GTX 285 temps. It seems that you should be looking at around 50C idle, so I suppose it could be a cooling issue (especially when it is kicked into high gear during games). Personally, my GTX 295v2 (which uses the same gt200 chips) idles at around 51.
 
Yeah 65C idle sounds pretty high for idle, and 90c+ after a few minutes furmark doesnt sound to good either.
I run furmark for 30min and didnt go above 76c with my 4870 which idles at 44C.
Definetly sounds like a GPU problem, try running furmark a little longer and see if it gives you the same crash as you get in games then you will know for sure. Good luck.
 
well, after running it for 10 mins straight it didn't seem to go much higher, deviating between 90 and 91 degrees. HOWEVER, when i exited the program, the GPU-Z program was telling me that the GPU 1 was 79 degrees, and the other GPU reading was 69 degrees. I had been looking at the line in FuMark called GPU Z Temperature, etc.

but no crash happened, which makes me think that it could possibly be a windows 7 issue or a driver issue?
 
I would try underclocking your cpu and lowering voltage and try then ... would help determine if it is overheating
 
Cool it

I went through a similar hell with nvlddmkm.sys and feel your pain mate.
I completely fixed this issue by doing some work on my case cooling.

I bought 3 new case fans (slight overkill maybe) and created a good airflow over my heatsink and 4870.

I also cranked my built in 120mm case fans up to medium from low.
I upped my 4870 cooler to 40% and last which was defo overkill, put a little heat extractor PCI card in below my 4870.

I then monitored temperatures.
Unsurprisingly they were fairly artic after that lot and the nvlddmkm.sys was history, that was 6 months back.

Make a cheeky little investment and get those temps down. Should work for you.
 
thanks for the advice mate, it seems to me that i've escaped the horrible nvlddmkm problems, just slightly concerned with this minimizing thing.

one example was i was playing aion, reached level 10 and when the cutscene for the fight happened, i was able to watch the start, then it minimized and i was unable to restore it till near the end of the entire fight scene. both annoying and wierd.

CoD games are similar in the way that when they minimize, cod4 will normally say server disconnected when i manage to restore it, and cod5 will probably be back to normal, except i will have missed a bit of the game.

EDIT: If this is a software issue, i might consider having a 75 or so gb partition of windows xp, for gaming, and the other part windows 7 for normal use. however again, this would be inexpressably frustrating needing to always reboot, i prefer just being able to let things work. one advantage, i must say, for consoles over pc is that you can just stick the disc in and play, no errors or annoying problems.
 
Hey Minato

Reading mooky32’s resolution to his problems, coupled with the fact that the weather has dropped the temperature a few degrees recently, this may explain your sudden nvlddmkm.sys disappearance (if it was heat related).

However, it may be that you have 2 totally unrelated problems and coincidently one has been cured by the drop in external temps (a possible example could be: you have a temp problem that’s been causing crashes but, over time, this in turn has corrupted a file/s and is causing additional random problems). Hopefully, you just have the one underlying problem that’s expressing itself in different guises dependent on the surrounding ambient temperature.

If you haven’t got a spare gfx card to test(?) - I would remove the side panel from the case and, if you have one, direct a desk fan onto the internal gubbins. Then test the games that you’ve been having problems with, run furmark and prime 95 and monitor all temps. This won’t be a conclusive test but if it is a heat related problem it may drop temps enough for the troublesome component to remain in tolerance and keep the system stable.

If by some minor miracle it does then at least you’ll have a solid foundation to troubleshoot from. If it doesn’t, unfortunately, you still can’t rule out temps as a possible problem but some components would be less likely to be suspects after the above test.

Additional: What OS and drivers are you using? If you're using windows 7, for example, have you let win 7 install all the drivers or have you downloaded and installed the manufactures latest drivers? And, if you are using win 7, are you sure it hasn’t updated a driver without your knowledge?

Basically it may be worth uninstalling/re-installing all your core drivers as it’s relatively quick and painless to do and is a good thing to rule out.
 
Last edited:
There's someone with similar problems on another well known site. They're RMAing the GPU having tried all known combinations of drivers, OS and PSU.
 
The fact nvlddmkm.sys is not giving you so much hassle is good news but I would still not rule out temperature.

Plec's idea is spot on, side panel off and some desk fan goodness straight to the heart of the matter!
Give it a go at least to rule out an area of investigation.

Other than that it sound like a graphics card issue. If you can beg, steal or borrow another graphics card I would go down that route.
You may even want to grab a cheapo one to confirm the fault.

Wishing you luck as I know what a pain this scenario is
 
i don't know why, but i've been playing Cod4 for 10 mins now and nothing...i just turned the computer on so all will be quite cool...fingers crossed eh?
 
Good to hear.

If you think it is a graphics card overheating problem you may want to try:

Easy: Give the GPU manufacturer a ring/email and tell them your problem, they may offer to replace the card.

Medium: Remove the stock heatsink, remove thermal paste (there is usually way too much and it isn't the best stuff) and apply the Thermal paste you used for your cpu. You should get a drop in temps of a few degrees.

Hard: Add an after market GPU cooler such as this. Will significantly reduce temps and keep the card much quieter.
 
well just now played for about 1 min, this time as i scoped in on cod4 and held my breath, my screen went into powesave, came out again, and i was at the desktop unable to resume. however, i did have the heartbeat constantly going which was quite cool.
 
You must log in or register to reply here.
Back
Top Bottom