Ravens Forums
General => General Chat => Topic started by: Saxif on May 22, 2009, 04:47:14 pm
-
I have this wierd problem with drivers, I thought the reason Elven Legacy was borked was cos it was buggy, after all I had updated my drivers to the 182.52 version. Then I installed another new game, Demigod and started to play and it crashed in exactly the same way as Elven Lagacy. I checked my drivers and they are still the 7.15.11.8252 version!! How can that be?
I downloaded new drivers, went into safe mode and used Drivesweeper to clean out the old Nvidia stuff, then installed the new drivers from the 'My Downloads' folder but it tells me it installed the old version again, I don't understand. Gonne go and try downloading more new drivers and see if I can get something to work ...
Any ideas on what I did wrong?
Sax.
-
Apparently its the last numbers which tell you the driver number, so I had put the right drivers in.
Still, I am getting exactly the same style of crash on both my new games which would suggest an underlying problem, anyon have any ideas about how to fix something like that?
Sax.
-
Anyone know how to update to the most recent version of Direct X 10? There seem to be a fair few files out there for it.
Sax.
-
Post counting like a mofo!!
I know its not my sound drivers now as I uninstalled them and ran the game, the other sound drivers I had dl'ed didn't work, lucky I had the disc to hand that came with my sound card!
I think it's either my monitor as it is registered as a PnP Monitor and not the actual type but it is an old CRT monitor and doesn't seem to have any official drivers. I am a little worried about dl random files from the net that may be drivers for it, or some basty virus! It could be my mouse also as every time it crashes the mouse pointer gets a special square on the screen of variable colour death.
Sax.
-
It definitely won't be the monitor. Monitors don't need specialised drivers, the generic PnP driver will work just fine.
It probably isn't your mouse. Not much can go wrong with a mouse or its drivers that would crash the whole system. The corrupted mouse pointer is a symptom of something else I fear (more in a moment).
It sounds like your graphics drivers are up to date and so this probably isn't the cause.
Which leads me to think your graphics card might be dying a slow death. Do these crashes tend to happen after a similar sort of time, once the card comes under load and gets hot? One of the symptoms of system crashes due to malfunctioning gfx cards is artifacts on screen, and the corrupted mouse pointer might be just that.
Are you getting crashes with any other games? Try some that will push your gfx card hard. Do you have another gfx card you can swap in and see if the games still crash?
-
Maybe your nto someting Cern. With Demigod it happens always withing the first minute, EXCEPT when I played it on the default settings which included a resolution of 1024 x 768, then I played for about 2 hours striaght, no problems. I tried it at 1600 x 1200 (normal resolution) at 85 hz and 60 hz, crashed within a minute both times, then I went down to 1200 x 968 (or something) and it crashed within a minute or to again. It seems to be the higher resolutions that cause it to crash. Given that the symptom is the same for both games i.e entire screen goes into a mosaix if bright colours, with the mouse pointer being a large square I am guessing it is something wrong with my machine and not the games fault.
I can play (and do) Civ4 for hours with no problems and Dead Space also works fine atm, though I haven't played it that much. Any ideas on checking the temp of the gfx card?
Cheers,
Sax.
edit - tried to run nvidia ntune to work out how hot my gfx card is but it crashed me, a blue screen came up saying that a problem had been detected and that it was closing down to prevent damage to the system. There was loads more but it was gone before I COULD Read it all!.
-
Sadly it does sound very much like your gfx card is on its last legs, especially if when you get a crash the entire screen goes into a mosaic of colours. There's not much that causes this other than gfx card failure (other hardware failures tend to either blue screen or crash to power off).
The reason you can play at 1024x768 but not 1600x1200 is that at the lower resolution the gfx card isn't being pushed very hard. But as soon as it is, it crashes.
The reason you can play Civ4 for hours is the same thing. Civ4 doesn't push the gfx card much, in fact most modern gfx cards will barely break a sweat running Civ4. Same goes for Dead Space.
I'd imagine if you throw something demanding like a FPS, or Oblivion or something like that at your gfx card, even at 1024x768 it will crash.
There's several things that might be going on:
1) The gfx card is overheating due to failure of the card's cooling system or components
2) Components are failing when they come under load
3) Some of the memory on the gfx card is failing (which might explain the crashes at higher resolutions)
Ntune should be able to tell you more, but if you can't run Ntune this isn't a good sign. I have a utility installed by my Nvidia driver called "Nvidia System Monitor". Perhaps if you also have this it might run for you and show the GPU temperature. But if you can't even run Ntune on a cold started system them it might not be overheating which is the problem but rather component failure.
Are you able to put in an old gfx card, or borrow one temporarily, to see if this stops the crashes?
-
I have Nvidia Monitor, that crashes me right away. I'll try tomorrow morning on a cold start system and see if I get anywhere. I might be able to borrow a gfx card from a friend, though he'll end up trying to sell it to me. My pc is still only 7 months old from scan, wonder if it is still in warranty? Usually a year, bet they'll love me, first my HD now my gfc card :(
Sax.
-
Downloaded EFGA Precision and my gfx temp gets to about 72 or 73'C whilst playing Dead Space. The fan speed was set to 40% so I upped it to 50%. Thing is Dead Space is very FPS'ery and I ahve it at max settings and 1600 x 1200, it is certainly more graphically demanding than Elven Legacy and I reckon more so that Demigod but it doesn't crash when playing that?
Any ideas?
Sax.
edit - I read about a dude who was having the same problem as me and it turned out to be his RAM, he had set it to 1.8V and when he turned it to 2.0 it sorted the problems. What do ya think?
edit 2 - I checked my RAM values and it is running at 1.904V, accoridng to this http://www.corsair.com/_datasheets/TWIN2X4096-6400C4DHX.pdf it should be 2.1V, assuming I am reading it right. Do you think it is worth testing? Also how would I actually change the setting, when I was in BIOS there was no obvious way of chaging it.
-
72'C when under load isn't too bad, certainly not bad enough to cause crashes I wouldn't think.
Faulty ram is a possibility, either the ram on the graphics card or maybe your main ram (but if the latter I'd expect problems to manifest in other things too, not just certain games).
I'm not sure about tweaking ram voltages, normally I leave everything as standard. You could try bumping it slightly in the BIOS. I'm not sure how, never done it, but it probably involves unlocking the overclocking features of your BIOS (as adjusting ram voltages are part of the overclocking process). But be aware that this runs its own risks, and if it really isn't the ram that is at fault you run the risk of causing extra problems. Personally I'd leave well alone, I'm sure your motherboard is supplying enough power for the ram - just because something is rated 2.1v doesn't mean it has to run exactly at this.
So Dead Space is working today but was crashing yesterday? Is Ntune still crashing?
Really it becomes a process of elimination. Try swapping out the gfx card, and if you still have problems then try removing one of your memory sticks (assuming you have 2) and seeing what happens (and then swap to the other memory stick on its own).
-
Dead Space never crashed and nor did Civ4, only Elven Legacy and Demigod, of the 2 Demigod crashed way quicker, always under 1 minute.
However!! I installed some Windows updates, it seems it was set to a 3am update time and I also turned my fan speed up from 40% to 55% and played with the side of the case off. I got a full match out of Demigod at 1600 x 1200, it ran at about 71 - 73'C, say about 15 - 20 minutes. I played Dead Space with the case off and fan turned up and it ran about 61-63'C but because I did both these things at once I do not know which one 'appears' to have fixed the problem. The obvious answer is to dial the fan down and put the side of the case back on and see what happens. From what I have read it seems that GPU's should be good up to 100'C, right?
Sax.
edit - There is also a fan in the side of my case I have never bothered plugging in cos I never worked out how and Scan couldn't do it cos they send pc's with lots of crash foam type stuff in. I got a torch and had a look and found a fan port on the mobo, will have to take the sound card out to plug it in but can put the sound card back in after. Also this fan is right next to the gfx card, so should help alleviate heat problems, if there are heat problems. Will do it later, see how it affects temps, idle temp goes up by 3'C with side of case on.
Sax.
-
Yep GPUs can run pretty hot without problems. 100' would be extreme and not good. But figures in the 60-70' range under load are normal and fine.
Are you able to monitor your CPU and motherboard temperatures when under load? If you have an Asus board they have a utility called Asus Probe which runs under Windows and shows CPU and MB temps, fan speeds etc. I would expect other motherboard makers have similar tools.
-
see my above dit! You posted to fast ;p
Not atm, idle cpu is about 45'C accoridng to BIOS.
Sax.
-
45' idle is a little high but nothing serious. Be interesting to see what it is under load, if you can find a utility that can monitor this from Windows when running a game. Also the motherboard temperature.
If things are running better with the side of the case off this does point to a heat problem. However, it may not be that any component is getting hotter than it should be (can't see any evidence of this from what you've said) but more likely some component or other is degrading / failing under normal load temperatures. Even if you manage to fix the problem for now by boosting your case cooling, you're only masking an underlying problem that will rear its head again in time. So, best to get to the bottom of it now while it is still within the 1 year warranty.
I had a problem with stuttering graphics once that happened when my system got hot. Temperatures weren't above the norm for load, but when it got hot I ran into problems. It turned out to be a failing power supply, which wasn't what I expected. I'm not saying that this is your problem here (though can't rule anything out) but just that sometimes the problem isn't the most obvious one.
-
<groan> More pc component detective work :(
How did you go about working out it was your PSU?
EVGA Precision Tool shows GPU temps whilst playing games, just got to find one for CPU but tbh I would be dissapointed if my CPU was to hot cos I have a big ass Zalman copper fan/heat sink bolted onto it, though I guess being so large it could be interupting air flow. Also if the ppl who installed it put it in backwards it would be a problem, as I have front and rear fans that should blow air through the pc over all the components but if the Zalman is blowing the wrong way it'll cause issues. Something else to check I guess.
Cheers Cern,
Sax.
-
How did you go about working out it was your PSU?
Process of elimination really. Swapped the gfx card for an old one, same prob, ran with one mem stick then the other, same problem, removed sound card, same problem and so on. Eventually I dug out a ropey old emergency PSU I had tucked away in a cupboard and tried that, and despite the PSU being horribly under-rated for my system it ran 100% stable so it pointed to the PSU being the problem. Got that replaced and everything back to normal.
As for your fans, I'm sure they're going to be blowing the right way but worth checking of course. The side fan could legitimately run either way, either as a blowhole to extract hot air, or as a method for blowing extra cool air onto the GPU. Which is best depends on the airflow inside your case (I'd probably leave it the way it's already mounted).
It's unlikely to be your CPU overheating if you have an aftermarket cooler and it has been fine up to now (assuming it is still properly fastened down).
-
I played again, this time with the side of the case on and it took longer to crash but it still did when it hit about 85'C. The GPU that is, I dled this http://www.almico.com/sfdownload.php and it tells me all other components are at least 10'C or more cooler than my GPU. Not sure why Demigod makes my GPU work so hard tbh.
Played some more with fan speed at 80% and case off, didn't crash in about 2 hours.
My CPU fan is a sexy bit of kit http://www.scan.co.uk/Products/Zalman-CNPS9500-AT-Intel-LGA775-Aero-Flower-Cooler ;D
Sax.
-
I played again, this time with the side of the case on and it took longer to crash but it still did when it hit about 85'C. The GPU that is, I dled this http://www.almico.com/sfdownload.php and it tells me all other components are at least 10'C or more cooler than my GPU.
Hmm, so does this mean that your CPU and/or motherboard are running at 75'C under load? If so, this is too hot.
-
When my gfx card is running at 75'C after playing for maybe a couple hours, my CPU runs at 40'C on both cores and my mobo is logging 30'C. I'd like to work out a way to keep my gfx card cool, I am sure running my fan on it at high speeds won't do any damage but I need my window open just to keep the room cool enough to stop the pc getting to hot atm which is silly imo.
Sax.
-
Your CPU and MB temps are perfectly fine. GPU temp seems slightly high but not excessive. My GT260 runs at around 63'C when under heavy load.
If your gfx card worked fine with your cooling system for some months and now doesn't I think it is a warning sign that things are beginning to go wrong with it. Yes you can boost your cooling with extra fans or higher fan speeds, but there's probably an underlying problem you need to address. I'd research what the normal operating temp for your card should be and perhaps talk to Scan about why it is falling over under your normal cooling.
-
You pretty much nailed what I intend to do Cern, dunno if you have come across it but I also use Guru 3D for pc advice (don't feel like I am cheating on you or anything ;p). I also find googling problems helps cos whatever has gone wrong has always gone wrong for someone else first. Being bank holiday I'll phone Scan on Tuesday (which incidently I have off as a priviliage day being a Civil Servant :)).
Cheers,
Sax.
-
Ok, a further update, apparently (and obviously when you think about it) fan speed should increase with temperature, my GPU fan speed according to EVGA Precision Tool stays at 40% all the time :( I checked by playing Dead Space for 20 mins then checking out the graphs. This I reckon is part of the problem. I also installed a further side fan that pulls hot air out from around my gfx card, this further helps regulate temps. Whilst playing Dead Space I didn't clock above about 55'C, this is with the case closed up but the extra fan and the fan speed set at 60%.
So I think part fo the problem was the gfx card isn't speeding up the fan when it gets hot and I didn't have a fan next to the gfx card to keep it cool. This is something I will continue to monitor though. Is this a good enough reason to ask for a replacement do you think?
Sax.
edit - It might stay at about 55'C whilst playing Dead Space but Demigod still cranks it up to 80'C :(
-
Ah, if your gfx card fan is stuck at 40% then this could well be the cause of the problems. The fan speed should ramp up under load. If it isn't doing so this could either be an issue with the drivers, or it might be an issue with the card's firmware, or it might be that the fan controller on the card has gone wrong and so it isn't providing feedback to the driver (I would imagine that 40% is the default speed).
If you can manually set the fan speed to something faster via the EVGA Precision Tool then this should prevent the crashes. But it's not ideal as the card should do all this automatically. Unless it is a known driver issue then I would think it would be reasonable grounds for a replacement, as your card wasn't doing this when new, or otherwise you'd had had crashes from new.
-
Ye, kinda what I thought. I have been messing with drivers recently, but afaik I have done the businiess right. Still I'll phone Scan and see what I can get out of them.
Sax.