annoying freezing issue

Sometimes Windows (XP Pro SP2) likes to freeze, and when I hard restart or shut down (press that shiny round button) it stops responding (black screen) after the Windows logo loading thing. Then I have to wait a while before booting it back up, and eventually it freezes again. As a superstition I unplug it and flip the PSU switch before waiting. Sometimes it freezes while I'm gaming, sometimes when I'm just trying to browse the Internet. Sometimes it'll even get to the point where the cursor is on the screen but the login screen doesn't show up when booting. In any case, the screen freezes, but any current audio continues normally and I can still move the cursor. I can't click anything because nothing responds.

I used to be getting BSODs from my wifi card, so I disabled it since I wasn't going to be using it (and no drivers seemed to work, so I gave up). My wifi card is listed as a USB device in Everest, yet it's mobo integrated, and it sometimes gives me an IRQ_NOT_LESS_OR_EQUAL, which means IRQ conflicts, right? Weird. Then I started getting usbport.sys and STOP 0x0000007F BSODs (not both in the same screen), and once in a while ones about ctaud2k.sys (Creative SB Audigy 4 drivers, updated from the Creative site). I still have no idea what that's about.

Now it just freezes. It's also telling me there's an unknown PCI device present. The only PCI devices I have are the sound card and modem, and those are detected, drivers updated, everything's fine. The little Safely Remove Hardware icon in the taskbar has some random "USB Mass Storage Device" listed as plugged in, yet the only USB things plugged in are my keyboard and mouse. I click "remove" for that, and after a couple minutes of the window hanging it says this nonexistent device can be unplugged.
Any idea what either of these mean?

The GPU and RAM are fine, I checked that out, tried switching RAM, did tests and all that. I just can't press that Restart button or I have to shut down and unplug for a few minutes before hoping it'll boot fine.
I tried a 30-minute stress test using nVIDIA's monitoring program right after turning it on for the first time that day one afternoon, and it got about halfway through before everything froze. The mobo got to about 43C, the CPU about 54C, and the GPU gets to about 74C/63C (I have a 7950GX2, one of those SLI-in-one-slot dealies). Do any of those seem too high? I turned the GPU fans all the way up after this test ("166.67%/65C" according to the nVIDIA panel thing) and it hasn't helped, and I do notice at least the top GPU (the one more towards the middle of the case) is quite hot to the touch, if that means anything. I also checked that the fans were facing in the right directions.

Anyone have any idea what's going on? Driver issues, maybe? Is there a possibility of overheating? Also, I highly doubt it's this, but a failing or insufficient PSU would be more like nothing working at all, right, not even the cursor?
Sometimes it'll go for hours, sometimes 20 minutes before freezing. I just can't shut it down or restart using the case buttons because it won't boot into Windows (well, I can boot into Vista, but it takes ten minutes and everything runs insanely slow). A normal restart/shutdown functions properly. If it does it again I'll try opening the side panel and facing a "normal" large fan into the case and see if that helps, even though I'm not so sure it's overheating. I'll probably end up reformatting eventually (probably won't help), but this never happened until I got home. Maybe because the case is in a slightly smaller area, but the back of the case where two of three exhaust case fans are (radiator fan for the Mini R80 and PSU fan) is still about a foot from the wall.


Also, as a random question, why does the Safely Remove Hardware thing have my internal hard drives listed? I don't dare trying to click "remove" for them, but I have no idea why they're listed in the first place if they're not USB or other external devices. Maybe it lists SATA devices, too?

 

I would suspect the PSU, but I think it is some kind of driver conflict, not sure how to fix it.

For your question about drives, yes, some motherboards list SATA drives as a USB device. They are supposed to be hot swapable. Don't worry, It won't let you stop your boot drive.

E

 

Yeah, I was thinking it might be a PSU issue, but it's still weird that everything was working perfectly fine until I brought it home from school last week. Well, other than the BSODs. I think I'll reinstall the original sound card drivers and see if that helps any, since that's the only other error I had that I have the CD for (the 12-in-1 card reader is my guess for usbport.sys).
I don't have another similar PSU to test with, and I'm not too keen on buying another until I'm sure it's the problem. It's getting incredibly annoying, and if a reformat doesn't help (after reinstalling sound card drivers and I'm too lazy to go through all the drivers), I might have to, and it would be one with more wattage (I have a 535W which should be fine, but who knows).
The rails are listed as fine according to ASUS PC Probe II, but they numbers seem to be a little higher than what should be maximum. Is this just an inaccuracy, or is the PSU suddenly working too hard? Is there any method of testing stability besides checking the rails with a multimeter?

Any ideas about the mysterious unknown PCI device or random "USB Mass Storage Device"?

 

Oops. I think I broke it.

It happened again, so I turned it off and unplugged it. Then I flipped that little red switch in the back that changes the VAC from 115 to 230. After waiting a few minutes I plugged it back in and tried turning it back on. It turned on for about a second before shutting off. So I waited about half an hour, thinking maybe the CPU (somehow) overheated and needed to cool down. Nope, still won't turn on.
No smell, smoke, or sounds, but when I flip the PSU power switch my desk lamp doesn't flicker like it usually does. Tomorrow I'll be taking it out and testing it in another system to confirm its demise. I'm sure it was powerful enough because, funnily enough, the ASUS mobo manual listed that a 500W works for a system nearly identical to mine (half the RAM and a slightly more powerful CPU).
If it was about to poop anyway, I'm glad I put the damn thing out of its misery.

I hope for the sake of computing that the PSU didn't take anything with it. When I get the new one I'll cross my fingers and hope everything works.

 

That sucks! I hope nothing else got zapped. I would think 500W or so would be enough for most systems. Just don't get a cheap no name brand thing thinking you are getting a good deal. I learned that the hard way! Get a nice one with a good efficiency rating.

E

 

Any brands anyone can recommend? I pretty much know the specs (>500W, ATX12V V2.0+ -compliant, SLi ready, I'm assuming the same pin types as the old one?). Is Enermax good, but I just had bad luck with this one?

 

I think this is the one I have.

http://www.newegg.com/Product/Product.aspx?Item=N82E16817341001

E

 

Heh, looks like that one has a fan on the top. On my case there's no hole on the top there, and my PSU has an intake on the bottom and exhaust on the back.
Thanks for the suggestion, though!

 

Look at the screw holes carefully. I think in that picture, it is upside down to show off the fan! I think this is a picture of the top of the PSU. http://www.newegg.com/Product/ShowI...+Supply+100+-+240+V+CB/CE/CSA/UL/FCC+-+Retail
But I could be wrong.

E

 

I did realize that (plus I guess PSUs can go either way), but I don't know how much of a difference the temp will be with one less exhaust and one more intake of already warm air. Other 2-fan setups with the same pin setups are hard to come by, so I might have to get one with a single exhaust facing the back.

 

You lost me there. Not sure what you mean.

E

 

My current (dead) PSU has two fans, one on the bottom (draws air into the PSU from inside the case) and one on the back where the plug and stuff is (spits air out). The one you showed has one fan, which I'm assuming is exhaust (blows air out of the PSU). Since the fan isn't on the back, and I can't have it face-up, it would be facing in, which means it would be blowing warm air out of the PSU and into the case.
I'm just wondering how much this would increase the temp inside the case and possibly of other components. Since it's hard to find non-Enermax (because now I'm biased ) 2-fan PSUs with the same main connector, PCI-E, and other pin arrangements, I might have to get one with a fan on the back where the plug and stuff is, so it blows air out of the case instead of into it.
Hope that explains things a little better.

 

The fan in the OCZ PSU is an intake fan, that takes air from your case and blows it out the back of the case. So you can count it as a restricted air flow exhaust fan. I say restricted because of all the stuff in the PSU.

E

 

Heh, cool. I guess that goes for other one-fans as well.

Turns out I can get a free replacement! For one I hope nothing else got grilled, and second it better not turn out like the first.

Thanks again!

 

Free is always good! Good luck with it! While I prefer a multimeter, this program should give you your voltages off the new PSU. http://majorgeeks.com/download.php?det=4181

E

 

Oh yeah, about that, turns out my dad had a multimeter at work. We confirmed the PSU was dead by checking the molex connector that hooked up to the fan on the side panel. Considering it wasn't doing a thing when I pressed the power button we were pretty sure it was dead anyway.
I'm already using Everest, mostly for checking temps, but thanks anyway.