• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Asus EAH5670 - Weird artifacts problem

SG81

n00b
Joined
Jan 2, 2007
Messages
11
Alright. I have sort of a weird problem with my video card. And, before I send this off to RMA for the third time, I wanted to check with you guys, first, and see if anyone of you can give me some insight.

Also, the easiest way to diagnose this problem, would be to simply put the card in another system, or install a different one in mine. But, I don't have an extra PCI-E card, and nobody I know has a mobo that supports it. So, maybe, I can see if it's really my system or not.

Firstly, my specs:
Processor - AMD Athlon II X2 240 Regor
Memory - 2GB DDR2
Motherboard - Gigabyte MA785HM-US2H
Power Supply - Antec SmartPower 2.0 500W
Onboard GFX is a Radeon HD 4200 (what I'm using now)
Windows XP SP3
Drivers are up to date and so is the BIOS FW
There's some other stuff that's not really relevant, I think.

System was working fine 5 months, no problems whatsoever except for the occasional freeze here and there, but nothing major. One day, the system would not POST. Took a look inside to see if anything was loose and noticed that the video card fan was at full blast. Re-seated it, powered on the PC- same problem. So, I took it out, switched over to onboard graphics, and it booted up with no problem.

I set up an RMA for the card and sent it off. In the meanwhile, for the 3-4 weeks it took to get back, I was using the built-in video, but I would get a few EA STOP errors, with atkdisp as the offender (note, that's just atkdisp- it wouldn't specifically tell me .dll or .sys or whatever). I don't remember the frequency, exactly, but I seemed to get them more later on. Either that, or I had just started noticing them as something wrong.

Anyway, got the card back and put it in. Windows did its thing, picked the card up, and rebooted. Came back, started doing some stuff, and the the problems started. Hundreds of tiny yellow blocks started scrolling vertically up and down my screen. Eventually, the system would hang, then the screen would blank, then it would come back as normal, with VPU recover from Catalyst would pop up and tell me about the card no longer responding to commands. 'Refreshing' the screen, either by moving windows, scrolling in a webpage, or whatever, would clear the blocks, sometimes with minor corruption. Also, having a program active that has something constantly moving (a webpage with a flash animation, for example), the artifacts would never appear. As long as there was stuff moving (except for a movie in the Media Player), it wouldn't act up. Only when the system is idle do I get the blocks. The appearance of the yellow blocks would continue soon after, and the cycle of blocks, freezing, and VPU recover would go back and forth until, finally, i would get a BSOD with ati2dvag as the problem.

The thing to note here, is that only when the drivers are installed, does this problem happen. At POST, or using DOS programs (for diagnostics), no artifacts are seen. In safe mode, or normal mode while Windows is using their basic video drivers, no artifacts, either.

The next few hours were spent doing all sorts of diagnostics and testing, including driver uninstall and reinstall (using driver sweeper), memory testing, unplugging extra optical drives and hard disks, taking out the one PCI card I have (TV tuner), using different drivers, including older versions of Catalyst, the drivers included with the card, and the ones on ASUS' website, messing with BIOS settings, and, finally, a full format and reinstall of Windows XP. All to no avail. Also, weirdly enough, the card still performed to it's capacity. Just for kicks, I started up Lost Planet with semi-high settings (1680x1050, everything on high except for AA and AAF) and it played flawlessly (no hiccups or graphic oddities) for 2 hours or so.

So, after a quick chat with ASUS support, I'm ready for a 2nd RMA. I sent it in and switched back over to the integrated graphics. Interestingly enough, the built-in video also produced regular BSOD's after an hour or so. It wasn't until I switched to the drivers that came on the motherboard CD, that I was able to get system stability. I even tried to use these drivers with the 5670, but the installed would quit unexpectedly.

Just got the card back from only 3 or so days of repair, which is kind of weird since the website said it takes 10. I'm not sure if this is common practice for second RMAs, but it's like the technician didn't even look at it closely, because the card still has the same problems. I'm about to send it in again, but I'm kind of doubting myself. But, as I said, I don't have access to the equipment I need. I'm very tempted to take it to a 'professional' repair place to just get it tested, or offer someone on CL some cash to test it for me.

The last thing I tried, just to make sure it wasn't a weird XP/driver issue, was install a copy of Windows 7 on a separate partition. Same problem. But, still, the card performed with power, as I was able to play Red Alert 3 on full settings, for a few hours. (I stopped when I had to alt-tab to the browser for a few minutes, after which Windows had told me that the video device was no longer responding or functioning, and promptly shot me into what seemed like 320x240x16).

I tried a more modern game, Black Ops, but it crashed with a 8E BSOD.

I have never experienced a problem like this. If the card were bad, it would have artifacts even in POST, right? Everything I see, leads to a driver issue, but not even a fresh XP and 7 install worked. If it was memory, then those games wouldn't be working. I realize that those games probably aren't using ALL the memory, but it happens when the system is idle and not using any at all? Very weird problem. And, before I start cursing at tech support, I need to make sure it's not on my end.
 
Hmm... ignoring all the obvious issues that I think you've covered (eg, making sure everything's plugged in right), if I were you, I'd try - preferably in this order -

1) RAM testing (which you have, but I'd let Memtest run over night if you haven't, I've found errors in the seventh hour).

2) A new video card, something super cheap ($35 or so) I could pick up at Microcenter or another brick-and-mortar if possible. I'd also consider making it an Nvidia card, just to see if it makes a difference.

3) A new power supply. This one's tricky if you don't happen to have an old one sitting around... but it's what I would do. There should also be a way to test the one you have with some kind instrument, I just have no idea how you'd go about that.

If none of that seems to make a difference... I'd be thinking motherboard.

-HD
 
Hmmm, memory testing... I wonder if the freezing issues with my onboard graphics were not solved by different drivers, but, instead, I'm just lucky that it hasn't froze yet. I'll have to do that one tonight.

I was thinking power supply, too, but the card doesn't use a 12V rail. Does that matter? One of the reasons I got this card was because of the low power consumption- my PSU isn't that beefy.

Ah, the joys of hardware testing. I think I definitely need to find a test PCI-E system, somewhere. That might solve a lot of questions.

Thanks for the replies, guys.
 
Hmmm, memory testing... I wonder if the freezing issues with my onboard graphics were not solved by different drivers, but, instead, I'm just lucky that it hasn't froze yet. I'll have to do that one tonight.

I was thinking power supply, too, but the card doesn't use a 12V rail. Does that matter? One of the reasons I got this card was because of the low power consumption- my PSU isn't that beefy.

Ah, the joys of hardware testing. I think I definitely need to find a test PCI-E system, somewhere. That might solve a lot of questions.

Thanks for the replies, guys.

Well depending on what else your running in the system you are already at 300w just gpu , mobo , ram and cpu.

How many hardrives do you have , how many devices are u using powered by usb ? It starts to add up quickly
 
I have 3 hard drives and 1 DVD combo drive. As for USB, I have a mouse, keyboard and my monitor acts as a bridge, so I can plug in stuff up here, but nothing is plugged in at the moment. I also have an external DVD drive, but it draws its own power with an adapter. The thing is, I went to the bare minimum, 1 hard drive only and it still had the same problem.

You know what's weird? If I have my phone plugged into USB to charge (through the monitor), the PC will boot incredibly slowly. It'll show the BIOS loading screen, and then just hang for 5 minutes while in the memory test. Then, it'll snap back into normal activity and load everything fine. I know it consumes the most power when booting up, but I don't know if that's before or during the BIOS screen. I might just be speculating.

Also, sometimes, the computer won't reboot properly. If Windows does a restart, my PC won't pass POST and it'll just sit there. I'll have to hit the reset switch a couple of times until I hear the beep because the monitor will still be in power-save mode and doesn't really turn back on the fastest. This has happened when it was powering both my old AGP P4 system, and this one.

I'm starting to wonder if it really is my PSU. But, I'm having a little trouble trying to understand that it would cause a card to go bad, and then have it work again after a 'repair, but still play somewhat graphic-intense 3D games with no problem.
 
Got a new, 600W PSU - OCZ600MXSP. Put it in, put the card in, drivers cleaned and re-installed, but same problem. The last thing it did was freeze, then the fan on the card went full blast and the whole PC reset.

Other than the card, could it be anything else? Motherboard?
 
Back
Top