Stumped by BSOD

If anyone could help me figure this one out, I'd much appreciate it.

Intermittent BSOD crashes with no common program or activity I can observe. Growing more frequent recently. No other problems.

BSOD says:
The MSDN page tells me the second param is the I/O status code, and that I have to look it up in Ntstatus.h. Ntstatus.h says 0xB5 means STATUS_IO_TIMEOUT.

Machine passes both memtest86+ and chkdsk. I've taken it apart and reseated/recabled everything twice.

Sooooo.... what's going on here?

My best guesses:

• HD is failing slooowly? (Maybe something not on the platter, hence the OK chkdsk??)
• ACPI.sys is corrupted resulting in nonsense page request?
• System bus is failing slowly, causing ACPI.sys to make nonsense page request?

Possible short-term work-arounds:

• Disable ACPI in BIOS
• Disable paging in windows

If anyone can help me track down the cause of this problem or suggest a short-term work-around, you'll have my undying gratitude.

what is your system configuration? specifically the hard drive? Also how many devices are plugged in to your PC (usb firewire E-SATA etc)? If you are getting I/O timeouts there could be a device hooked to your PC which is causing that.

You could try clearing/resetting the BIOS. Some timings could have gotten out of whack and I've seen lots of cases where power spikes (lightening, etc.) have modified BIOS settings.

And, of course, turn any overclocks you have, back to defaults. (I know that's not likely the case, but you'd be surprised at the number of times I've run into cases where people have OC'd and forgot or ignored it.)

Beyond that, it sounds like it's probably a hardware failure of some sort, which could be hard to pin point (or not).

I/O timeouts ? smells like hard disk issue ?

P.S.: Even if I ran a scandisk , the scandisk reported that the 1.5TB 7200.11 Seagate is fine and no errors found, but then it was giving a BSOD , I RMA the hard disk and chaing into another brand now. Make sure you are not getting this because of the hard disk, maybe you can try another HD ?

I'd recommend trying a burn-in test for the drive if you haven't already (This'll be difficult if you're running low on disk space) to see if you can generate any errors at all with it.

Other then that, you could try reinstalling acpi.sys just to cover your bases to be sure it's not corrupted. You could also use another HDD controller if you have that option to see if it's consistent across more then one controller.

Out of curiosity, what components are you running?

1. Not my machine, I'm just tasked with fixing it.

2. Seems like my last round of reseated/recabled everything worked. No crashes reported for 3 days now. Keep your fingers crossed.