Computer won't boot

[CrazyAL]

I am working on a computer. It is an AMD based machine. I recently upgraded to Win2K from WinME. Over the last year while running WinME, the machine would blue screen fairly regularly. I figured the Win2K would correct this. WRONG! I had the machine running OK for a week, but then kept getting error messages. So I formatted the HD and did a fresh install. This worked for a couple days, but then I started getting Registry Corruption Messages, and the machine will no longer boot. I think it is futile to perform another fresh install at this point without addressing the root problem (I think it must be a hardware issue). I have run the memtest88 memory checker several times (over a 24 hour period ran continuously) any the memory tests fine. I have a WD 40GB hard drive, and have just finished running the diagnostics programs on it, and they have found no errors on the drive. It has a AMD TBird 1200 with a ECSK7VMA and 256 MB SDRAM (PC133). It has a 250W Deer PSU. What could the problem be? I have tested everything but the PSU, but I haven't had any reboots, or other typical PSU problems. I am at my wits end here. Any help would be greatly appreciated.

Thanks.

 

[Cybordolphin]

Are you running overclocked?

 

[CrazyAL]

No overclocking. Just stock everything.

 

[CrazyAL]

Bump for some wisdom...

 

[KF]

To me it seems the only likely way the registry will be corrupted is if it is corrupted on its way to be written to the HD. In my experience when I have gotten that error, Window has been right, unlike nearly every other error message Windows delivers. OCing the bus definately can cause data corruption to the HD. Since you do not OC, I would get another HD cable. Since it took a week at one point to get an error, the error rate is very low, and checking for the problem really being solved will be very time consuming. The low error rate can be explained by the fact that the ultra DMA modes use error detection and retries. What you are going to have to do is run your system for couple of weeks to get a reasonable assurance the problem has been corrected.

Several years ago, when HDs were smaller and before the newer ultra DMA modes, I tracked a problem to an error that only occured once in about every 200MB. It turned out to be the HD cable. The only way I could detect it is to compare every byte of a 200MB file copy.

This type of error used to be a symptom of a HD failing, but nowadays people seem to find that one day the dang thing just will not work at all, instead of the problem creeping up on you. Make sure you have SMART turned on if available.