Heat issues?

[jax_spades]

Okay.
So a buddy of mine just got a new graphics card.
ATI Radeon x1800xl.
And all of the sudden, it's beginning to freeze up on him, and just splashes garbage onto the screen momentarily, goes black, resets, and brings up an error message. The game is still running though.
His card temp apparently is at 65C.
Is it overheating, or what?
Any help would be great.

 

[raven_87]

@ 65C? Heat problem - neg. Unless thats idle temp.

Might be something else...

 

[dmantech]

Do you have a PSU that could handle it? If not, that may be the problem.

 

[waylander]

the first thing you should check is the driver, make sure it is the most up to date.

 

[Omid]

What wattage is the PSU, this is a very heavy card power wise!

 

[jax_spades]

He does have the latest drivers. He updated them when it first started happening.

It idles usually at around 55-60C. I'm still new to the whole hardware stuff, so let me know.

http://www.supermetroidclassic.com/screeny.JPG (image of his temps and some other settings)

His PSU wattage is 450. Thanks everyone.

 

[CrimsonJustice]

EDIT: Added system specs below.

Hi guys, I'm the one with the problem here and I'll give more specifics.

My card ( ATI Radeon X1800XL) arrived 2 days ago, and I installed it properly. Everything seemed to be working fine with the default drivers provided with the card. I even spent a few hours playing WoW with absolutely no problems whatsoever.

The following day I went to play again and for most of the morning it ran fine, up until it crashed for the fisrt time. Thinking it was a small driver error, I updated the software with the newest from ATI's site ( With the Catalyst 6.7 package ). Another couple hours later it crashed again, and then in a smaller space of time yet another crash occurred.

Now by crash, I don't mean system crash or anything. what happens specifically is this:
The video in the game freezes for a minute.
Then, a blank screen is produced and my monitor informs me there is no signal for a split second. Occasionally junk will appear on the screen momentarily as well.
Afterwards, Catalyst informs me that there was an error and that my card has been reset. The game is still running and I send the error report.

I recently updated my mobo drivers and cleaned up the spagetti in my case by moving wiring around a bit for more air flow, but without any luck. I went maybe 15 minutes in-game before it crashed and reset, which is a significantly shorter amount of time than before.

The current temperatures that the card seems to run are this:
Current: 50 deg Celcius
Last Desktop Session: 58 deg Celcius average
Last Gaming Session: Went up to around 65 deg Celcius max

I have requested help from ATI as well, but their first response was to update everything and then try minor fixes which I do not have (No bios setting for video caching-AGP only? And SMARTGART does not show up in the Catalyst control panel to disable Fastswitching ).

I will be mailing them back as well.

System Specifications:
Mobo: ECS RX480-A, Socket 939
CPU: AMD Athlon 64 3800+
RAM: 2GB PC3200 Dual channel
OS: Windows XP Home SP2

Average system temp: 30-40 deg Celcius
Average CPU temp: 40-50 deg Celcius
( These I can only view in the BIOS as I don't have any active monitoring facilities )

 

[CrimsonJustice]

No responses? *bump*
ATI hasn't responed either...

I also opened the case and used a household fan just blowing on the graphics card. For the time I tested it, I had no problems, but just the other day when I went to try it without the open case and fan, it crashed within an hour.

Somehow, with the open case, ATI's temperature readings were basically unchanged. Under load it went from 65-70 Celcius but it didn't crash, which is the same temps that it said it had before. Is the default monitoring in the Catalyst Control Panel inaccurate? Do I just need more cooling then?

 

[TheGreatGrapeApe]

Doesn't sound like heat for either, sounds like a power issue on both.

The 450W PSU of the first poster sounds like the problem unless it's by a bulletproof company like PC Power&Cooling.

Yours sounds the same.

Those temps aren't high and the reset sounds like a power reset (done by the PSU to avoid frying stuff, but don't keep doing it because it does stress a system).

If you have everest check your PSU health especially while gaming and especially after crashing check the temps.

 

[CrimsonJustice]

Just so you know...it's the same problem, jax was referring to me.

I'll give everest a try later then. Though it seems kind of odd for a PSU failure since everything else continues running. Would the power reset be somehow specific to the graphics card? It receives power through a molex chain, in which case, wouldn't other devices on the chain fail due to the reset?

Still doesn't explain how it has successfully run without crashes twice using the open case and household fan. Though admittedly extensive testing has not been performed in such a case and the crash seems to be relatively random.

 

[rexter]

I would say that maybe power supply on both but jax_spades image has nvidia running and ati on the background. Wouldn’t that affect the driver also? I'm not but these guys here might know more about that.

If haven’t done it yet, and you previously uninstalled nvidia card from you system then you should uninstall the driver using drivers uninstaller from NVIDIA to clean the driver properly.

 

[CrimsonJustice]

Again, there is only 1 computer in question here, the image posted by jax_spades is a shot of my desktop. Yeah it's got nvidia running in the background...for an nvidia card no less. I have an old Geforce4 PCI card that I'm using temporarily while the X1800 XL is acting up. I suitably swap between them as required. Pretty sure that's not going to be the problem, as I installed the nvidia card following problems with the X1800 XL.

 

[waylander]

Which exact psu is it?

This might be a case of both heat and power, what case with how many case fans do you have?

 

[rexter]

Sorry about that CrimesonJustice. But like the guys here I would still think that it's the power supply. Given your discription.

 

[CrimsonJustice]

The exact PSU is unknown to me at the moment, but I can get specifics tonight. I can say that it came with the case, so it would likely be some generic brand.

I don't know the specific case details either, I bought it locally. As for fans, I bleive there are two, though there may only be one and I'll have to check tonight.

I was thinking of trying the ATI Tray Tools and setting the fan speed, as well as maybe picking up a slot cooler and an extra fan for the case. I'll probably get an extra fan anyway, even if heat isn't the problem.

Still I would like to know (because I'm not well versed in this, any experts?) how the power supply would cause the error exactly while leaving other devices on the molex chain intact (cd and hard drive are on the same chain I believe, one of them to be sure). Is it just dumbing down the power supply, causing the video card to have an insufficient amount of power to run in a heavy load environment?

Also, since the card has a (12 volt?) additional plug that it has attached via a molex adaptor, does it draw most power from this while still drawing some from the motherboard or what? Maybe these last questions are a bit more technical than normal, but I'd like to know either way.

 

[rexter]

There's another one more way to check if its power issue. Connect 1 rail of power line just of the video card. And the next rail to power your HD, and the rest. And then try to disable as much hardware as you can just for testing, like your case fan.

Then try to find a crack for that game that does not let you require the cd/dvd drive to play the game. So you can disconnect the optical drive to see if that’ll we give enough power for the video card.

 

[CrimsonJustice]

Sounds reasonable, I'll give it a go tonight. It doesn't need the cd/dvd or anything, So I'll be able to hook up just the hard drive and the graphics card. As for the fan...it probably won't make much of a difference but I'll do that too.

Supposing it fails still, would there be any reason to doubt the PSU at that point? I mean one would think the PSU would be capable to run the simple setup alone, even for a generic brand.

 

[TheGreatGrapeApe]

I'll give everest a try later then. Though it seems kind of odd for a PSU failure since everything else continues running.

It's not a failure so much as dirty power, it will basically reset the VPU if it's not getting enough clean power.

Still doesn't explain how it has successfully run without crashes twice using the open case and household fan.

Well since PSUs provide less power the hotter they themselves get, could be even the PSU needed better clean cool air. But you're asking people to trouble shoot from 1000s of miles away, really you should be taking it to a service centre if these quick fixes people offer don't fix it.

Though admittedly extensive testing has not been performed in such a case and the crash seems to be relatively random.

And until you check them thoroughly, you'll be basically poking at the next best possible cause, until you elimintae them, at which point you're going to have to have it checked more throroughly by someone else with better tools.

These suggestions are based on , here's what you're displaying as a problem it matches this scenario most of all, but without you testing then it's just a shot in the dark. But if the PSU is 'unknown' and is a small Wattage rating then it's a likely suspect.

 

[TheGreatGrapeApe]

Again, there is only 1 computer in question here, the image posted by jax_spades is a shot of my desktop. Yeah it's got nvidia running in the background...for an nvidia card no less. I have an old Geforce4 PCI card that I'm using temporarily while the X1800 XL is acting up. I suitably swap between them as required. Pretty sure that's not going to be the problem, as I installed the nvidia card following problems with the X1800 XL.

REMOVE nVIDIA drivers!

You never run the drivers of the two companies on one rig and expect gaming stability. Run that PCI card as a standard PCI card using just the basic Windows drivers. Even for multi-monitor setups it's best to run one driver for your primary and then let windows run the second card.

Get DriverCleanerPro and blast out all remnants of nV. Sticky old drivers cause massive stability issues for both ATi and nV.

 

[rexter]

As to my knowledge I don't think Everest is still downloadable since they drop the support for home version. Unless he already have it.

I use the Lavalys Everest Home Edition my self but I think PC Wizard just as good. I wonder do you know any other utility tool software for free?

He can try the trials from Lavalys, Ultimate Edition but I'm not sure if it has less or more options since I haven't tried it my self.

 

[TheGreatGrapeApe]

As to my knowledge I don't think Everest is still downloadable since they drop the support for home version. Unless he already have it.

*In my best Agent Mulder voice*
The file is out there.

It might not tell him what he needs, but it will give him more information without having to use true diagnostic tools.

 

[Gary_Busey]

Great site:

http://www.oldversion.com/

 

[CrimsonJustice]

Well thanks for the replies guys. I'll give a couple of things a try tonight which will hopefully solve or at least narrow down the problem.

 

[TheGreatGrapeApe]

Great site:

http://www.oldversion.com/

Yep, just the kinda site I was talking about. SweEet, add another to my favourites list. 8)

ALSO if he wants the full corp version the are other way too. :twisted:

 

[rexter]

Thanks, That is a nice site.