Computer Hope
Hardware => Hardware => Topic started by: Darusame on December 07, 2008, 10:04:40 PM
-
I'm using Windows XP Professional with SP3. My video card is a NVIDIA 9800GT, and it's been having BSODs related to nv4_disp.dll since I got it. The most common solution is to reinstall the drivers, and I've done it multiple times.
I tried doing it again today, and my problem got worse - in addition to nv4_disp.dll BSODs, i've started getting lock ups, hang ups, and random restarts ingame. There are two potential causes that I can see: Last night, I automatically fine tuned my computer with the nTune program (performance in Nvidia control panel). It overclocked my [PCI-E something clock speed] 17% to around 2900. I also used Driver Cleaner to clean off my nvidia drivers, then I reinstalled my video drivers again. I've since clocked my computer's clock speed to what it was originally. However, that did not solve my problem, as the lock ups, hang ups, and random restarts still occurred.
I then tried following the Driver Cleaner's instructions more precisely. First, I uninstalled my Nvidia Drivers from Add / Remove Programs (as I did before). When my computer restarted, it claimed it detected both a video controller and an Other PCI Bridge Device, but I choose not to automatically update drivers for either. At this point, I think my computer may have started to get random restarts even outside of playing games. Driver Cleaner's instructions then prompted me to start the computer in Safe Mode. I did this through msconfig. However, here's where my current problem is at:
Although I can get to the login screen and I get to the prompt that asks if I want to run in safe mode or go to system restore, my computer always restarts there. I've tried holding F8 and choosing last good configuration or perhaps another type of safe mode. Although I can get to the Windows Advanced Options Menu http://www.computerhope.com/issues/chsafe.htm , as seen there, my computer does not register my pressing of the up and down keys. Thus, I am stuck - I can't get into the Windows OS past the login screen, and I can't choose to start up the computer in any alternate way.
-
This is what I'd try...
Turn off computer. Open the case, clean it well with compressed air.
Reseat RAM, and video card. You may clean connectors with an eraser.
-
I'm using Windows XP Professional with SP3. My video card is a NVIDIA 9800GT, and it's been having BSODs related to nv4_disp.dll since I got it. The most common solution is to reinstall the drivers, and I've done it multiple times.
I tried doing it again today, and my problem got worse - in addition to nv4_disp.dll BSODs, i've started getting lock ups, hang ups, and random restarts ingame. There are two potential causes that I can see: Last night, I automatically fine tuned my computer with the nTune program (performance in Nvidia control panel). It overclocked my [PCI-E something clock speed] 17% to around 2900.
This.
Your video card is overheating!
Clock that thing back to its defaults so you don't fry the dang thing!
-
I've since clocked my computer's clock speed to what it was originally
-
Is it possible that my video card has already been damaged, and that my increased rate of errors is a consequence of that? And if it has been damaged, would it be covered by warranty? I think I recall nTune giving some warning, although I didn't pay enough attention to it.
I'll probably try cleaning up my computer's insides tomorrow if some other possible solution doesn't pop up. I'm getting pretty desperate.
-
As killerb255 said, it looked like overheating from the very beginning, and yes, the card may be damaged, but don't start crying yet :)
Do the cleaning, and we'll see.
-
Do you think it was overheating even from when I was having nv4_disp.dll? I was running a heat monitor a couple weeks ago, and the computer held a stable 72-73 C while running Call of Duty 4. However, I still had suspicion that it would spike up in heat shortly before crashing.
...And when you say to clean the connectors with an eraser, do you mean a literal eraser (as in the stuff used to erase pencil marks), or something else I'm unaware of?
-
This is what I'd try...
Turn off computer. Open the case, clean it well with compressed air.
Reseat RAM, and video card. You may clean connectors with an eraser.
Dont use an eraser if you attempt to do this. Its well noted that it will leave behind parts of it.
-
Ah, okay. So I should just take the ram and video card out and put it back in? Is this process analogous to taking out an old cartridge from Super Nintendo and blowing it?
-
Ah, okay. So I should just take the ram and video card out and put it back in? Is this process analogous to taking out an old cartridge from Super Nintendo and blowing it?
Not unless you leave the pc on and try it (might pay off to earth yourself before hand though). That's why they built a cartridge lock into the SNES.
-
thats not a good thing ether since. when you blow your atucally removeing dust yes but your allowing the parts to age faster cause of small amounts of water an what not.
-
however reseting it should be okay or if you run an electric clear on the contacts.
-
Oh, I was just using the Super Nintendo example as an analogy. ;p
What exactly is an "electric clear on the contacts?"
-
Dont use an eraser if you attempt to do this. Its well noted that it will leave behind parts of it.
Total nonsense. You can safely use pencil eraser.
he computer held a stable 72-73 C while running Call of Duty 4
Absolutely too high, unless it's laptop.
What processor is it?
-
Here's a rundown of my parts:
Windows XP Professional with SP3
ABIT AN8 Ultra Motherboard
AMD Athlon Processor 3200+
600 Watt PSU
2 X 512MB DDR1 RAM
Nvidia Geforce 9800GT 512MB
-
There are different versions of AMD Athlon Processor 3200+
What's its speed?
-
CPU AMD 64 |3200+ ATHLON 64 939P RT
This is what I gathered looking through my old receipt - if you can decipher what it means.
-
Athlon 64 3200+ Socket 939 max temp is listed at 65° C (!), so no more COD until the problem is solved. At 72-73 C you're killing your CPU.
Start with turning the computer off. Open the case, and clean it well inside, using a can of compressed air (do NOT use vacuum cleaner!).
While cleaning fans, hold the blades with a pencil.
When done, keep the case off, start computer, and see, if all fans are running.
Go to Windows, and check temps again.
Post the results back.
Temporarily, you may use house fan, pointed toward the open case to keep temps down.
If nothing helps, removing old thermal paste, and applying new one, may help.
-
I have a clarification:
I ran the heat monitor out of the Nvidia control panel - that's where I saw my 72-73C. Wasn't that number based off the video card's heat? Or am I gravely mistaken?
Also, as aside, I had a previous malfunction with an underpowered PSU where the number (which I got from the Nvidia control panel) spiked from 65 - 83Cish when I started up a game.
-
Also, are there any alternatives to using a can of air? I won't be able to get one for another week or two.
-
are there any alternatives to using a can of air?
Not, that I'm aware of. It's not my business, but I'm wondering about your location, so it'll take 2 weeks to get to the store.
At least, open the case, and see how it looks inside, and if all fans are running.
Also, let see what your temps are...
Download, and install SpeedFan: http://www.almico.com/sfdownload.php
Post your computer temperatures:
(http://209.85.48.8/228/109/upload/p4151509.gif)
-
I'm at college until near Christmas, and I don't have access to a car. And I already have a can of air - it's just at home. =/
And being that I can not get past the login screen (yet), my computer can not download and run speedfan.
-
Did you?
At least, open the case, and see how it looks inside, and if all fans are running.
Are you using USB, or wireless keyboard?
-
I use a USB keyboard. The strangest thing to me is that the keyboard will work everywhere but the menu where I select Safe Mode, Last Good Config, etc.
It works when I press F8 to get in there and it works when I press Enter to get into my account.
-
If we want to do anything before Christmas, you have to keep answering ALL my questions
For the third time:
Did you?
Quote
At least, open the case, and see how it looks inside, and if all fans are running.
Try PS/2 keyboard.
-
Oh, I'm very sorry - I did, and they're all working. And in terms of using a PS/2 keyboard - I don't have one, sorry.
-
Is computer dirty inside.
You'll have to borrow PS/2 keyboard, or you're gonna be getting stuck.
-
Yeah, the computer is dirty inside - I'll definitely clean it up once I get access to a can of air. And unfortunately, I don't think I have access to a PS/2 keyboard right now either.
I guess I'll wait until I get those things and report back. Thanks for the help!
-
No problem. Keep us posted.
-
The University Bookstore probably carries compressed air...
-
Oh, i never thought to check there - I'll look tomorrow. Thanks.
-
No problem...and Welcome Aboard !
-
Alot of local stores will carry it as well look around some have better prices walmart,staples maybe its kind of a hit or miss.
-
Hi everybody.
I just went through my computer fans with a can of air, and it did not solve the issue.
I won't have access to a PS\2 keyboard until Sunday, and I have not tried using a house fan yet (although I may soon).
Does anyone have any other suggestions until I try those two things?
Also, what parts of my computer do you guys think may be damaged / crashing my computer (from heat overload or whatever it is)?
-
Your initial temps were already high...cleaning doesn't always solve it right away...
DLoad and install SpeedFan and post your temps for now.
Try using the machine as little as possible til we narrow this down unless you want to be buying a new machine.
-
The OP won't be able to use SpeedFan until he can actually get into the OS. Right now, he's stuck on the "F8" menu and his keyboard's not registering key presses. Not much to do until he gets a PS/2 keyboard...
-
Alright, I'll wait for that keyboard.
Is the PS\2 Keyboard sure to work, by the way? Is that why I havent been able to do anything on the F8 Menu?
-
I was just attempting to move the process along killerb...
I didn't re-read the entire thread to see what steps that were suggested were ignored.
But good observation on your end.
-
Err, does the BIOS register Keyboard buttons?
Make sure to Enable USB Legacy devices (should be under chipset or advanced)
then your usb k/b should work in the F8 menu
-
Using the PS\2 keyboard, I selected Last Good Configuration and got in to my OS. From there, I restored to a day before I tried to automatically tune with the Nvidia program.
I also installed speedfan. Here are the temperatures:
(http://farm4.static.flickr.com/3229/3128521158_121d5cd51d_o.jpg)
It seems parts of my computer are overheating =/
I know that the GPU is the Graphics Card, but what's the Core?
What can I do? Upgrade power supply, get more fans (although I don't know if there's space), or get a new case?
What negative effects will result if I continue running my computer at this rate?
-
Could you make a picture of the interior of your case? Easier to give advise that way.
-
(http://farm4.static.flickr.com/3234/3128570306_e4d46757c8_o.jpg)
This is the best picture I could get with my cell phone - is that sufficient?
-
Not really, no.
If you can't get a decent picture then just describe the kind of thermal paste used, the kind of cooling installed how many casefans you have.
I assume you've cleaned any dust that may have been present out of your fans by now?
-
Those temperatures are good.
Core in this case is your graphics card's core temperature, which is why it's the same as GPU.
-
In comparisment to the rest of the system, the videocard is too warm so there must a lack of airflow or a bad cooler in that general area. That should be fixed. The rest of the system is A-O.K.
Assuming you gave it the time to warm up first and not immediately made the screenshot after it had been cooled down several hours before turning it on.
-
Those temperatures are fairly normal for a 9800GT with a stock cooler.
The fan by default is at 29%, my 8800GT was idling at around 60C before I increased the fan speed and that was with an intake and exhaust fan, both 120mm.
-
Those temperatures are fairly normal for a 9800GT with a stock cooler.
The fan by default is at 29%, my 8800GT was idling at around 60C before I increased the fan speed and that was with an intake and exhaust fan, both 120mm.
Then I'd like to know what temperature it runs at when it's in use.
-
As in, while it has a game running in the background? I'll fire up Call of Duty 4 now.
-
That's the issue Raptor, they can get to dangerous levels very easily.
I increased the fan speed on my card and it's never above 68C (Crysis for a few hours), automatic fan control via Rivatuner so it's not too loud either. I'd never try gaming with the fan on automatic, controlled by the hardware as default - the fan, from what I've heard, only kicks up to 100% when the card gets to about 85C.
Slightly off topic, but I'm just saying that 63C idle is about normal.
-
(http://farm4.static.flickr.com/3225/3128674786_74c6f2795d_o.jpg)
I started up Call of Duty 4 Multiplayer, joined an empty server, then alt tabbed.
I took this picture after about 1-2min, the temperature stayed stable.
-
Well, it's not high enough to get worried about but if you got some extra cash, you might want to consider investing in a higher quality videocard cooler. Or a PCI-exhaust fan cooler if the stock cooler doesn't blow air directly out of the case.
The cooler the better. Other than that, everything's just peachy.
-
That's good to know.
Would extra cooling improve performance or reduce bluescreens, by any chance?
And would upgrading anything else reduce the heat load on the graphics card (e.g., power supply, motherboard, ram, and processor)?
I'm so glad my computer's back and stable. :)
It's been a long two weeks - thanks for the help, guys.
EDIT:
Oh, and now my GPU's going at about 71-76C at the same settings (with 1-4 other people in the server) - is it possible that it might hit dangerous temperatures when I join servers that have lots of people in them? (It's hard to tell since it immediately drops temp as I alt tab, so my max temps could be even higher)
Oh, and how about rivatuner? Can it automatically tune the GPU fan so that it's running at full speed all the time?
-
Oh, and now my GPU's going at about 71-76C at the same settings (with 1-4 other people in the server) - is it possible that it might hit dangerous temperatures when I join servers that have lots of people in them?
You'll have to run a little test to see what level the temperature rises when you play for an hour or so on a busy server.
Some cards seems to be better able to withstand high temperatures than others, mainly those that come with passive cooling (IE: no fan). According to my Nvidia control panel the videocards won't shut down until they reach 130c... It is in no way recommended to play at a temperature that can boil water!
Anyhow, to answer your other questions;
Would extra cooling improve performance or reduce bluescreens, by any chance?
Yes and no. Mostly a blue screen means a conflict somewhere, not necessarily a heat problem. Although the conflict can be caused by a heat problem I suppose.
What overheating will mainly do is cause slow downs and artifacting. Sometimes even leading to a shutdown if the temperature reaches a critical level.
And would upgrading anything else reduce the heat load on the graphics card (e.g., power supply, motherboard, ram, and processor)?
Having a good airflow is a must but since all your other parts are quite well cooled I'd say your GPU might just need something more powerful to get the temperature down. The rest is apparently cooled well enough as far as we can tell from a screenshot.
Hope this helps. In the end, it's your decision. The cooler components run, the longer their life expectancy and the better their performance.
-
My computer gets random occasional slowdowns while I'm play COD4 (not any other game, so far), so I don't think that overheating is too big of a factor (for now).
The bluescreens I get are typically associated with not reinstalling patches correctly, although I've amended that issue and still get them. I was just making sure it wasn't a heating issue.
I'll definitely look into get something that specifically cools my graphics card - where do I start and what are these parts called specifically? I didn't even know there were parts that specifically cooled down GPU heat.
Are there different varieties? Which would be the best for just getting my card to normal temperatures? I'm not interested in overclocking (especially after this incident).
-
Zalman and Cooler Master are two brands that develop high quality (and silent) coolers. Another name that jumps to mind is arctic cooling. These coolers are mainly referred to as "VGA coolers".
Not all coolers fit on all videocards, so make sure that you check for compatability first. If you are interested in a cooler, make certain that you read at least one review on it. Installation is often easy. Zalman even offers an online step-by-step flash movie. Most of these coolers come with thermal paste so you do not need to go out and buy that seperately.
-
I think a better solution is to control the current cooling system better. It involves a lot less time and money.
http://www.overclock.net/nvidia/273072-how-set-auto-fan-profile-most.html <-- that may help you.
I'm going to get a guide on how to use Rivatuner like this sometime soon, I'll step it up a little.
However, if you do decide on a new cooling system, my choice would be the Zalman GV1000 or Arctic Cooling Accelero (that needs a specific revision as there are several, check before buying).
-
I'm still trying to grasp your link, Calum, as the terminology is a tad confusing, but:
What's the harm in having fan speed at 100% all the time? Does it drain too much energy / too unnecessary?
-
I'm still trying to grasp your link, Calum, as the terminology is a tad confusing, but:
What's the harm in having fan speed at 100% all the time? Does it drain too much energy / too unnecessary?
It's noisy. Other than that, nothing I can think of.
I don't really get the terminology either. Maybe Calum can explain.
-
If you follow the first few steps, i.e. install Rivatuner (link (http://www.guru3d.com/rivatuner/)) then the rest should be clearer.
They're all options within the windows it refers to, in Rivatuner.
You can ignore the whole formula thing, it doesn't make much sense to me either.
Start with the settings the poster references using for his 8800GTX (use the key he provides to help) and play around, see what you can get from it.
I agree that if you're a first time user that is a little confusing, sorry and I will get a clearer guide done soon.
-
That's the issue Raptor, they can get to dangerous levels very easily.
I increased the fan speed on my card and it's never above 68C (Crysis for a few hours), automatic fan control via Rivatuner so it's not too loud either. I'd never try gaming with the fan on automatic, controlled by the hardware as default - the fan, from what I've heard, only kicks up to 100% when the card gets to about 85C.
Slightly off topic, but I'm just saying that 63C idle is about normal.
I just configured with the guide's settings, and I'm about to restart to see the difference.
However, I'm confused about what you said earlier Calum -
you say that your temp never goes over 68C with Crysis on, then you say "I'd never try gaming with the fan on automatic, controlled by the hardware as default."
Does that mean you'd never try gaming with the default fan settings, or the other way around? Is there something I don't know about (sorry if I'm a tad paranoid, but I'm still a little confused about this whole fan thing).
-
Did you do that test yet?
-
Oh, I have not. =/
I haven't really had a chance to sit down and play a good hour of anything, especially since I'm still contending with intermittent BSODs and the busy winter schedule. I do plan on playing the Witcher for an extended amount of time tonight or tomorrow - I'll report the temperature when I get it.
Also, my computer has stayed a steady 52-55C (not running games) after changing to the settings in the guide. For reference, I had previously held a stable 63ish degrees.
-
After over an hour of the Witcher, my computer was at about 60C. I also didn't get any BSODs (yet)!
So how would I go about lowering my temperatures even more? Which variable would I mess with? I'm still a tad reserved in terms of fiddling with rivatuner.
-
Also, I just discovered that I have a 500W PSU, not a 600. Is the PSU possibly bottlenecking (or causing nv4disp.dll BSODs)? Or perhaps, causing some parts to generate more heat than necessary?
-
Also, I just discovered that I have a 500W PSU, not a 600. Is the PSU possibly bottlenecking (or causing nv4disp.dll BSODs)? Or perhaps, causing some parts to generate more heat than necessary?
Try a wattage calculator on the web to see if you meet the requirements. What kind of PSU do you have?
As for your 60c, yes, that is very acceptable.
-
I mean I'd never try gaming without Rivatuner controlling the fans.
So, I'd never game with the default fan settings, as in without any software managing the speed.
Does that make more sense?
Sorry for the confusion.
-
I use a fan controller for my two videocards.
-
A fan controller is another option of course.
Having the fan controlled automatically is more convenient though, no need to watch the temperature and keep adjusting the fan to compensate.