View Single Post
Old 06-13-2011, 03:32 AM   #6793
Darth Avlectus
@Darth Avlectus
I'd buy that for a dollar
Darth Avlectus's Avatar
Join Date: Dec 2006
Location: My pervert mansion
Posts: 4,397
Current Game: A dirty old man.
@TiE: I've not built my own computer yet (might be in the cards soon though). I'm afraid all I can offer you is my best guess based on past experiences fidgeting with hardware and micro/mini processors.

Originally Posted by TiE23 View Post
Y'know. That was a point I didn't cover at first. The way it's designed, the heat sink is only attached to the four screws surrounding the GPU die. I felt like I was going to snap something if I handled it by the heat sink, it would creak a lot.
Ehh, that kind of scares me. At the very least I would maybe tighten it down a bit more. And make absolutely sure the pressure is distributed evenly and firmly.

Now, yes. Since Speedfan tends to lock my system up (I have no idea why), I only use it when I'm directly interested in the temperature of the card. So I didn't discover the super high temps until after a day or so of it being in the case. That's when I decided to install the ASUS cooling utilities... heh. Seems like the coolers on this card sort of demanded it.
Just tossing ideas here, not sure if I'm being much help. Additional considerations couldn't hurt, though.

Probably asking some real dumb questions: what about your power supply? Are you near its maximum output with that graphics card onboard? My best guess relating purely to power and overheat would be on overdraw of current. Even within "working" ratings I'd still only consider 75% of full wattage capacity to be the absolute maximum safety margin for a supply.

Even if your supply isn't stressing out, an under-powered device might act erratic. Damage from that is possible, however frightfully unlikely in a computer I would think because there's usually safety features preventing that.

Otherwise I'd say comparability issues. I'm stumped.

Perhaps some other modification you might have made to your system prior?
*can't help, sorry*<snip>

So, I went to a restore point (seriously, I've only recently discovered the fool-proof-ed-ness of restore points, they work ****ing wonders) back in the middle of May (my system back-up was turned off the last two weeks). I was then back on my computer with January 2011 nVidia drivers and I'm back up to square one. For awhile I was fearing that this serious amount of mucking around with drivers would require an OS reset, but the system restore saved me there. But now I'm fearful of updating the drivers... I think the combination of new drivers and a damaged card nearly broke my system. I fear that updating to the new June drivers will mess up even my 460 at the moment.
As an engineer told me once: change only one thing at a time and test it each time. It may be tedious but you'll be glad you did. Cuts down confusion of troubleshooting when something stops working.

Eh, what ever. I'll try to carefully update drivers in the next day or so and I'll see how that goes.

But now... for the replaced card I should be getting soon. ... I've got Arctic 5 thermal paste, would it be worth pulling the cooling unit from it to replace the thermal paste that comes with it? I highly doubt that it would void any sort of warranty, but would it be worth the time?
Ehh, no. If you're getting a fresh new one, I'd say leave that alone. I was making the suggestion in the case of the old one if you weren't getting a replacement. See, I'm tenaciously persistent and tend to try *everything* I can (within reason of course) before I spend more money.

I'd try to figure out what is causing issues with the speedfan, though. And if possible I would do it *before* installing the replacement video card. Again, any modificaitons that might possibly be causing it? Maybe incompatibilities with your existing hardware if not your software?

Making sure that heatsink isn't creaky might serve you well, if you're comfortable gently slowly and incrementally adjusting screws to make sure of that. Then again as expensive as that thing is, it may already be full up optimum so I might think twice about that and only do it if I am next to positive something isn't kosher about it. If you experience something similar and warranty has run out, where the thing refuses to work as it did before, then you might consider the possibility the factory stock thermal paste may have dried up.

Still, screws and not a clamp or something? Really? Sounds chintzy.

When the new card comes I'll definitely make sure that the temperatures stay good... there is a good probability that the card cooked itself / I cooked the card -- within the first few minutes of using it. If my memory serves me right, I believe the first game I played on it was Crysis 2, and that ran just fine for about 90 minutes. Then when I went on to BFBC2, it crashed the game twice and would not support any hardware rendering, as in, when I tried to start up Mirror's Edge, it crashed to desktop immediately. I never had a hard lock or BSOD with the card. Only crashes to desktop... would an overheating video card do that?
I don't know. My only suggestion besides asking the company that made the part, is making only one modification at a time to simplify troubleshooting.

Best of luck.
Darth Avlectus is offline   you may: