Skrying said:
Temps are okay. CPU heats 50C at load, stock heatsink. GPU hits 85C at load. PSU feels normal as far as heat. All fans are clean, everything heat wise seems fine.
I'm gonna try that PSU on Monday. And try a different card if I have an AGP laying around.
One thing I've noticed. Since I unhooked the secondary HD, when the screen gets the barcodes I can simply open a new window and they go away, there is no more locking up. Not sure what this means yet, any clue?
That GPU temp looks a bit high, but you say you have problems even when it's relatively "idle" on a 2D desktop, right? I would suspect the video card if it weren't for the fact that you are also having problems with the HDD.
It might be interesting to use Coolbits or some other NV tweaking program to underclock the GPU and VRAM to see if this helps solve the problem.
This difference of behavior when you unhook the secondary HDD suggests that it is involved in the crashes. I completely forgot to ask you if you had looked at the event logs for any hints. I forget the exact error name/code, but HDD resets should be reported there (someting like: IDE0 timeout).
It's dangeorus to guess without having all the information or the system in front of me to see it for myself. It could be the PSU and it already took the HDD and GPU with it (or at least still causing them to temporarily malfunction). It could just be two independent problems with the secondary HDD and the video card and your PSU is fine.
A detailed history of the progress of the problems might reveal more.
Here's the thing: I don't see why a secondary HDD failure would cause video corruption. It doesn't seem to be, because this problem still persists with that drive removed. But...it no longer hard locks, like you say, with that drive detached. This, to me, would suggest a common point of failure and that would be the PSU (it could also be the mainboard, CPU, RAM, but this you seem to have verfied independently).
One more thing: that click you hear (like the "seek on old drives") is most likely a HDD reset of the actuator. It may also be performing a spin-down/spin-up and this means that it would be the most power demanding state the HDD can be in. If something is wrong, this is the time it will "burn". This plays into the theory and events you see with the non-hard lock-up when the drive is removed. You are opening a window, drives are scanned, drive fails, drive cycles to re-attempt a read, power draw is increased, "bar codes" appear on screen (???). Perhaps that is the wrong order.
The more I think about it, the more I feel you should not be testing this system unless you are also willing to lose your primary HDD. If you continue, at least make sure you have adequate cooling (perhaps remove the side panel on the case).
Just out of curiosity, when you unlocked your video card's extra pipelines, how did you do this and did you reverse the process exactly or is there a possibility you disabled some other pipelines when going back?
PS. I am mainly rambling now (I am so curious that my brain is running all tangents and permutations). Be wary of using any advice in here.