Yesterday my computer locked up in the middle of something and I was forced to do a hard reboot. At first everything looked okay, but during the POST the boot up came to the choice of start WIN XP normally or start it in one of 3 safe modes.
When I choose start normally, to process continues to the black XP splash screen and then I catch a brief glimpse of some BSOD blue and the whole thing starts over.
If I choose to start in safe mode, I see the list of drivers that are being installed and then I get the quick sight of BSOD blue and the reboot starts again.
My first thought was that there is a physical problem with the hard drive, but am not certain about that.
I think that I want to use the XP disc to see if I can repair the install, but my disc is in another city for some strange reason and I won’t be able to get it until Monday at the earliest.
Is there anything else that could cause these symptoms? I imagine that I have overlooked something, but what?
There is no option on the boot option screen called “turn off automatic restart on system failure”? With that turned off you can actually read the blue screen.
Last known Good Configuration gets to the black XP splash screen, you know with that blue dot going back and forth. The next thing the the quick glimpse of the BSOD. When the system locked up, it was in the middle of loading a new level in Call of Duty 2.
Not that I can see in the POST, but now that you mention it, I recall seeing something similar to that in the BIOS. I’ll look at switching that and seeing what happens.
Another thing that I failed to mention in the OP was that the computer also locked up while I was futzing with a new monitor. I just bought a 47" LCD TV and can’t place it in my bedroom yet, so I figured it might be cool to see what it looks like as a monitor . Turns out I have to sit too close, but I did check the video card and it is still firmly seated in the mobo. I wonder if that could have anything to do with it.
You could try to boot to something other than the hard drive to see if it’s a software or a hardware problem. Dig out that old DOS floppy or use a Linux CDROM.
Still not enough to pin anything definitive on, but based on what you’ve posted so far and my own experiences with such things, a few of the more common possibilities are:
Video card is in some way damaged; possibly one or more RAM chips. It would still function and allow the computer to POST and even display basic graphics, but as soon as anything requiring larger amounts of VRAM be used it can throw an exception and bluescreen you to death. There are other possible ways a video card can be damaged with similar symptoms, too. If you have access to another video card you can use, try booting with it installed after you’ve removed yours.
Main RAM is potentially damaged. This one is hard to diagnose because it depends on how far down the chain of chips the fault lies, but if it’s fairly close to the beginning of the chain that there’s a bad chip, you’ll error out much quicker and more consistently. If you have more than one stick, remove all but one and boot from it. If it works, you’ve found the problem. If not, swap it with another; same deal. If it doesn’t boot with any of the sticks, it’s less likely to be the RAM. If you only have one stick, try and see if you can borrow someone else’s (same type) and see if that works.
Power supply is no longer able to provide consistent clean voltage to the motherboard, which can result in varying levels of instability, including lockups, bluescreens and trashed data if it doesn’t just shut off altogether. If you have access to an alternate power supply, try booting with that.
There is a bad sector spaning one or more critical Windows system files, or one or more critical Windows system files has become corrupted, which is causing it to error out when Windows tries to load it. If it’s a bad sector, you’re pretty much lost until you either replace the hard drive or format it and allow the formatting program to mark the bad sectors. If it’s corruption, a repair install is your only hope.
There may be other potential problems that could cause that, but those are off the top of my head.
I’ve just tried to boot the machine with BIOS set to fail safe mode, and still no luck. Since it is a new video card, I’d be pissed if that is the problem but it is an energy hog - it uses two 6 pin energy connectors. I do still have the the previous video card so I’ll give that a shot.
I swapped out the video cards and still had a problem.
I did find an evaluation copy of XP and stuck that in to use the repair feature. I did re-write the boot sector -no luck. Next I ran a CHKDSK and it came back with the message that one or more sectors were damaged. Lucky me I bought a HDD last night. I’ll slave the broken one to the new one and install XP on it. That should at least allow me to recover some data and not mess up what may still be okay.
I have had luck booting off CD containing Norton System Commander and doing either Winsock repair or Disk Scan/Repair with the Norton Tools when I started to get BSOD at Windows Boot (even safe mode).
You might try booting from XP CD and doing a scan/repair of disk before reinstall of Windows etc.
It requires two six-pin connectors? As in, two +12v connectors? I can understand two of the standard 4-pin +5v connectors – most modern video cards require one +12v molex, but will survive on two +5v molexes using a +12v adapter. I didn’t know there were any video cards power hungry enough to require 24v. (Two high end cards running in SLI, sure – one 12v connector each – but single-slot double decker cards like the GeForce 7950GX2, which I currently run, still only require one +12v.)
How powerful is your current power supply? And what brand? And what video card? I’m beginning to suspect a power supply fault at this point, one that either caused some garbage data to be written to clusters on the hard drive that were occupied by some critical Windows files, which has resulted in your inability to boot.
Mentioning that you have just swapped video cards brings to mind possible driver problems. Did the fault occur as soon as you swapped cards?
Have you tried booting to safe mode? If this works, go to Computer Properties, click the Advanced tab and System Settings, look for an option around System Failure called “Automatically Restart”. Untick this and when it next BSODs, you can copy down the bugcheck codes.
Either google or post them here and we’ll be able to help. If you can’t do safe mode, try booting with logging enabled. That will produce c:\bootlog.txt. Look at a Windows repair/bootdisk which will boot you into a shell. You can then read the file. Again, post it (or as much as possible here) and we can find out whats faulting.
Yeah. It is a GeForce 8800 GTX. The manufacturer says a 450 watt PSU is needed. I think mine is just about that big, but it ran fine. Here is a picture of one, you can see the two connectors in the top right corner. The card is massive.
Because of this problem, I have ordered components for my new computer from Newegg. I can’t wait. I am going overboard with a Core 2 Extreme processor, SLI 680 mobo and 4 GB of RAM. And a 700 watt PSU for when I add that second GPU.
As for the ‘old’ computer, is only 2 years old -it shouldn’t have this problem - it is a failed hard drive. I found an old trial version of XP and popped that in to use the repair function. I rewrote the boot sector, but that didn’t work. I did a chkdsk scan and found that one or more sectors are damaged. Following along with the scan it seems that one platter in the HDD is screwed badly, but the rest are okay. I put in the new HDD that I bought on Thursday, just in case, loaded the trial disk of XP and voila! No problem.
Not a driver problem at all. The new video card had been in the machine for about 2 weeks and had the most current drivers I could find for it. As I mentioned before, I could not boot into safe mode. I could see the list of drivers load and then back to the POST.
Ah – the 8800. I’d forgotten they were pigs on power. All the same, I still want one.
Never mind that – you’ll need that for the first GPU. 700W should be fine for SLI on those though – just make sure it has enough +12v connectors.
I’m still thinking that your PSU was probably struggling to maintain enough power to keep everything running smoothly and coolly, but managed to flake out enough to cause some corruption. I’ve had this happen myself – I had a 500W Thermaltake PurePower die on me a little over a year after buying it (it’s currently in the RMA process). In so doing it irrevocably corrupted my Windows installation such that I couldn’t boot into Windows anymore. I bought a temporary PSU, a cheap 460W and reinstalled Windows without a hitch. The PSU mostly died inside two weeks; it worked, but would overheat within 15 minutes and shut down. Now I’m running an OCZ 700W and it’s cool and running smooth.
And yes, hard drives can die long before their time. Also currently in the RMA process is a 500GB hard drive I bought a year ago. Thing was starting to seriously flake out on me. I bought an exact replacement and it works great. I’ll just RAID the replacement with my current one when it gets back.
. Turns out I have to sit too close, but I did check the video card and it is still firmly seated in the mobo. I wonder if that could have anything to do with it.
