There was a problem with a friend’s computer. Lots of ‘valuable, sensitive’ stuff on it. (I’ve no idea what). She asked me to take a look.
There’d been a power failure immediately preceding the onset of the problems. May be related to the source of the problem?
After trying safe mode w/o success (worked briefly then restarted and/or generated an error), I decided to use the recovery console to run chkdsk /r. (maybe a good idea or maybe not)
As I couldn’t boot the HD I sought to boot from the win2k setup CD (and subsequently with diskettes for good measure). No dice.
It’d reach one of several (apparently) randomly determined points, then restart.
So then I took one of my spare HDs (identical to hers btw) and tried to install win2k to it on her machine. No dice. At various different points during the process the computer would either reveal one of several different blue screens (prob’ly a half dozen different ones- mostly stop errors) and/or restart.
Since then, in pursuit of this, (formatting and installing win2k on the spare HD (identical to the one that had been previously working in the machine)) I have disconnected every non-necessary item attached to the machine and tried again. No dice.
Then I proceeded to swap out the questionable necessary components (RAM sticks, cables, power supply, CDROM, 3.5 floppy) one by one with known good ones and try again. I tried all of this with the BIOS on her settings, default settings and optimal settings (as well as a few other sets of settings). No luck.
Then I tried it all again with only known good devices. No luck.
Now bear in mind that her Hd is sitting on the table in the living room and uninvolved. This all involves items that worked on other machines and still do work on other machines. I tested all of these known good components on another machine after trying all of this just to make sure that they were, in fact, known to be good. The test format and install I’ve tried since on another machine with all of these same components went as it should.
Am I right in my conclusion that there must be something wrong with the hardware?
In this instance, is it correct that software problems shouldn’t impact the format and installation of the OS on an identical HD?
if yes then
Since I’ve substituted functioning, compatable (and in some instance indentical) components, am I also correct that in concluding that the problem is most likely in one of the components I have not substituted?
There’s only the motherboard and the CPU left, AFAICT.
if yes then
How much effort is required to get the HD to function as it was in another computer?
Is it a more efficient option to just reinstall win2k and then reinstall the apps?
Would an indentical MB and CPU make things any simpler/easier?
I don’t happen to have a known-good, compatable CPU to try in place of what’s currently in there so I haven’t been able to try a substitution. Neither do I have an identical MB to try.
So, given what I’ve done, have I reached good conclusions?
If I have, what’s to be done?
If the CPU or MB are bad, they’ll have to be replaced anyway (but she’d rather not if they’re not the issue) so if it’d make things easier as far as having a fully functional computer again then that’s what we’ll do.
I think that I’m right and that either the CPU or MB are busted. However, there are an uknown number of variables involved (unknown to me anyway) with this process. Have I missed something that’d render my dioagnosis invalid?
Probably a silly question, but is the Hard Disk type set to ‘auto’ in the CMOS settings? - I had a computer once where the settings were explicitly (but wrongly) entered and although it let me format the drive and install windows, it would just suddenly crash (it would make horrible buzzing and clicking noises from the hard drive too, but with a modern, quiet hard drive, these might not be all that noticeable).
I manually ran the BIOS’s ‘auto-detect hard drive’ function and trusted its settings. I didn’t try to manually readjust settings as I figured that the BIOS knew what it was doing and I don’t.
I can try this though.
But since I used first the settings that had been on there from the original HD, wouldn’t the setting be the same for an identical HD? Unless something changed the BIOS settings between when she went to bed and when she got up the next morning (the power outage?).
I’ve no idea as to the age of the mobo.
The drive that I’ve tried to use is identical to the one that was previously installed and working.
Since the previous drive, my friend’s, was working, is it safe to assume that the mb is capable of an accurate auto-detect?
If it’s not, how do I determine the appropriate manual settings for the drive?
By “Since the previous drive, my friend’s, was working, is it safe to assume that the mb is capable of an accurate auto-detect?” I meant
"Since the previous drive, my friend’s, was working a few days ago before the problem occurred, is it safe to assume that the mb is capable of an accurate auto-detect?
I suppose it is possible that the CMOS battery failed and it forgot the hard drive settings and auto-detected the incorrectly, but I fear that this might be just a red herring, so in no way do I insist on this idea or even particularly favour it.
The hard drive should have a label saying how many heads, cylinders and sectors it has and these settings are the ones you’d enter in the CMOS setup if you were configuring the drive entry manually.
Certainly looks that way; you seem to have eliminated almost everything else and it has to be a hardware or firmware problem.
It might be worth turning the case upside down and shaking out any debris and dust, plus blowing out the rest with a can of compressed air (vacuuming the inside of the PC can be a bad idea because vacuum cleaners can genereate ferocious static charges as air and particles of fuzzy dust rush along the inside of the plastic), then unclip the processor and reinsert it (there’s just the slightest possibility that one of the pins/blades is corroded slightly, but enough to make the connection dodgy).
But it sounds like it might be time for an upgrade.
There are only a few reasons a bare mobo would be giving bluescreens on setup with a known good hard disk, cdrom drive and installation CD.
It’s broke
The HD is connected or configured incorrectly.
The BIOS is non-ACPI compliant and needs to be upgraded. Alternately, you can load the setup program in compatability mode by pressing F7 when it indicates you can press F6 to load SCSI drivers.
It doesn’t seem to have been mentioned yet, but why not install the suspect HD in another machine as a secondary HD? Since the perhaps-broken disk is not the primary, the OS should boot as normal, then you’ll be able (assuming less than massive corruption here) to access the “valuable sensitive” stuff at your leasure.
That’s what* I * said.
However, there’re apparently important applications that she can’t re-install because she doesn’t have the disks or what-have-you.
If possible, she needs to have all of the programs running. I discussed it with her last night and she doesn’t even know what all programs are on there. She didn’t set it up and has a partner who uses it remotely. Her partner is off traipsing the globe in parts unknown at the moment. Due to other snafus she won’t be able to contact him until he reaches the US, or at least an America.
I’ve bit my tongue about the need for a good back up of important info, because I just lost a whole wedding album. I’m sure that I have a good excuse, but I just don’t know what it is yet. :smack:
This might be silly, but have you checked for heat issues? Run the case while its open, check the motherboard and CPU temps in the BIOS before trying to load the OS. Does the CPU fan appear to be failing while you are booting up? Are the CPU temperatures extremely high >75C? If so, the CPU fan could have problems or the thermal material between the processor and the heat sink could have flaked off. It does appear that you have done a proper and thorough job troubleshooting this problem.
I will try that.
I’ve got to take the CPU out one way or another anyway, so why not?
I’m not sure if the BIOS has a CPU monitoring capacity. I hope so though.
It didn’t occur to me because I’m use to mine that’ll just scream and shutdown.
If there’s no hardware monitor in the BIOS, I’ll just take the CPU out, examine it and then regrease and re-install.
I don’t know if I have a spare CPU fan that’ll fit hers though. But a CPU fan is a cheap check.
Thanks. I hope this works.
WARNING It’s been a few years since I last attempted this on a W2K box as I’ve since gone to XP and 2003 but it should be possible.
Get a fresh working computer, bang in the harddrive, fire it up and cross your fingers and see what happens. It may successfully reinstall the new hardware and whatnot – it’s probably going to be a bit stuffed tho’.
What I’ve done in the past is put in the harddrive, don’t even attempt to start it but stick in a bootable W2K CD and boot from that then one of the rescue/recover options will basically reinstall the OS for the new hardware but preserve a lot of the programs and data. I can’t remember the exact options and I know there are degrees of repair it can do (which vary in risk of data loss).
As I said it’s been a while but I did manage to the OS running on a seperate box mostly OK. Having said that it was never as good as a clean install and I nuked it later once I’d recovered all I needed and started again.
Also note that you might not get access to encrypted files and non-standard programs will very likely not work.
IIRC it was a different MB/CPU, hence why the need to reinstall.
If you get an identical MB/CPU then windows’ll probably not care.
Even if you just replace/upgrade the CPU (and not the MB) it should probably work OK.
It’s the motherboard drivers that require the rescue/reinstall stuff (particularly with on-board sound/graphics or bizarre drivers for hard drives etc).
This is what I suspected.
But I’m very computer ignorant. Hence my quest for aproval of my methods. I don’t know what all I don’t know about computers.
