Curse

Fanta · Aug 27, 2006, 05:11 AM // 05:11

OK so here is what happened, i recently just RMA's my 6600GT because of overheating after 5 months, too short imo (i was getting discoloration on screen and when restarting comp all that showed was snow, kinda like a tv with no cable, but it was in polygon form with scattered out random letters).

Anyways, when I got my RMA'd card it worked fine, but within 8 hours of the installation, while in guild wars everything froze, my game, ventrilo, everything. So i had to restart it from the tower, and in doing that my monitor blacks out and the light on it comes up as if VGA isn't even detected to the card.

The next morning all i do is disconnect the power supply and remove the cords in the vid card then replace them and it worked, but the EXACT same thing happened and this time around no matter what i try to monitor does not detect the card; but windows doesn't load either because there is no sound either.

Everything turns on as if it is working fine except the monitor and it doesn't go to the windows login screen *i assume at least*.

I don't think it is the video card though for the fact it's brand new, could something else in my comptuer be causing this and do i need an upgrade somewhere? I called eVGA support and they said i needed a bigger power supply, I kind of don't believe them though because the card needs at least 350W and my power supply is 450W.

Any help at all would be much appreciated, not only can I not play guild wars, but I can't use my computer. Thank You for any help.

one more thing, when using my backup FX 5200 my monitor works, going back to the 6600GT, nothing.

Kaguya · Aug 27, 2006, 09:31 AM // 09:31

Watts on a power supply mean nothing in the end. What really matters is how many ampers the suppy can give to 12V, 5V and 3.33V rails.

What brand and model power supply is it?

Since another vidcard works on the same machine, the problem is related to the 6600GT pretty certainly. Is there another computer where you can plug the 6600 into and test?

Fanta · Aug 27, 2006, 10:19 PM // 22:19

the power supply is by Antec model SL450,
40A @ 5V
25A @ 12V
28A @ 3.3V

I'm gonna try calling a friend up to see if my card works in their's, i'll post when i do so.

and, TY for your help ^^

Mushroom · Aug 28, 2006, 02:31 AM // 02:31

Antec is a pretty good power supply, certainly not a $35 discount model.

The first thing I would check would be the capacitors on the motherboard. Those are the little round things that look like coke cans. Look to see if any of them are buldging, or leaking. If even one is like that, you need a new motherboard.

Bad capacitors on a motherboard will give all kinds of screwy problems, and loss of sound and video are among them.

I would also question the card. eVGA has been having several quality control issues over the past year. Part of the problem seems to be constant overheating. Most of the posts I found on the NVidia website about the card recommend placing a slot fan right next to it, and replacing the chipset cooler if possible. There have been reports of the card hitting over 150 F, and that is absolute death to a Video card. The fact that it seems to happen to you during gaming (when the card is under the largest load) points me in that direction.

Gorebrex · Aug 28, 2006, 02:40 AM // 02:40

Dont forget that it MIGHT be your monitor. If you have a spare, plug that one into your system, see if it works.

Fanta · Aug 28, 2006, 02:59 AM // 02:59

Ok, I haven't tested it on a friends computer yet, but here's what has happened now.

I put in my backup FX5200 so i could get to the bios, when i restored to defaults and tested the 6600GT again it worked. *mind you i had never changed the bios settings, so I don't know what restoring something to what it already was would do*

Also, I noticed when I put in my FX5200, it wouldn't go straight to windows login, I got a message saying something with CMOS was messed up.

Does any of this indicate or pin point the problem?

My card is working fine right now, but i GUARANTEE that it will lock up again in Guild Wars for the fact that I haven't changed anything to prevent the problem.

Kaguya · Aug 28, 2006, 09:15 AM // 09:15

Yeah, Antec is pretty solid PSU manufacturer.

I'd rule out the monitor from the problem list since it works on other vidcard.

Does the CMOS error come on every time you boot, or just first time? CMOS is the chip where BIOS stores it's settings. Usual CMOS errors come from either the battery on motherboard failing (pretty uncommon since the battery has a very, very long lifespan), CPU clocking settings resetting themselves for reason or another (pretty constant after power failures on certain computers here at my workplace).. Thus the BIOS settings might have been completely reset, or partially.

If the CMOS error repeats itself after every boot, easiest fix would be to replace the battery (round silvery one on the mobo) and see if the problem persists. What was the error message?

Numa Pompilius · Aug 28, 2006, 09:52 AM // 09:52

Just FWIW... The most common reason for computers locking up after playing GW for a while is overheating. Either of the CPU or of the graphics card. If it's the CPU you typically get errors and resets, if it's the graphics card you typically get corruption of the graphics just prior to losing graphics all together.

In your case it sounds like your graphics card is overheating.

If you open Display/Advanced/Geforce <name of your card> and enable "classical view" (nVidia has seen fit to hide all info from the user in their 'new highly recommended interface'), you can see the temperature of the core of your graphics card. Both temperatures should be below 80C (176F), above about 100C the card starts glitching, and a little higher it completely cuts out.

Wrt your old card, it almost certainly was a case of dust covering the heat sink - graphics cards have very low closed heat sinks with just a narrow slit where the air from the fan passes, and that very easily gets clogged, effectively stopping the cooling of the card.
The solution is to remove the card, and then carefully removing the dust build up by whatever means feels comfortable to you - me I yesterday carefully used our vacuum cleaner on the heatsink of my 6800GT which had started overheating while playing GW.

Your new card shouldn't have had time to accrue as much dust as your old card, so it may be honest-to-goodness defective, I do not know. Or perhaps you forgot to connect the fan on the card to the PSU? In any case, the first thing you should do is check that the fan (and all other fans in the computer) are working, and the heatsinks not covered in dust.

Fanta · Aug 29, 2006, 05:57 AM // 05:57

In regards to Kaguya, I don't remember the exact CMOS error, but I believe it was something resetting itself. If it does happen again (i think it will, but i hope to god not) I'll write down what it says.

To Numa, 6600GT's have then core temperature sensors in them, but they are disabled from the bios by manufacturers (don't know why) but I have already flashed the card and got the temperature monitor on it. It was running between 55-60C, which I hear is normal.

p.s. Thanks for all your responses and for the help and insite you have given me.