Evident HD failure (my first)

[Muse]

I bought this drive along with a 200 GB and 120 GB 2½ years ago. They are all in this, my main desktop system. Fortunately, the 160 GB is the least critical since I used only for backup and temporary DVR stuff. I was ripping a DVD with DVDShrink and when I put in the DVD-RW disk to write to I got a couple weird error messages about being unable to write. Head scratchers and I figured it was my old Windows 2000 OS going a bit crazy again (due for a reinistall any day). However, I see now that the drive has disappeared from Explorer, isn't shown in Device Manager or Disk Management and it's not showing up in the BIOS under IDE devices. Is there any thing I should do before calling Seagate support for an RMA?

I've had probably close to a dozen HDD's and never had an actual failure. I RMA'd one that was having bad sectors early on. This thing gave me no warning at all. A little over a year ago all the data on the drive was lost but I think that was because Windows 2000 in combination with my MB IDE controller didn't support that big a drive. It had been on my Promise IDE card, which did support it but I put it on the MB controller instead and the data loss occurred, I think when the data on the drive reached past a certain limit (about 120 GB, IIRC). When that happened, the drive did show up in Device Manager, etc., it just didn't show up as "healthy" and formatted. At that point, Seagate assumed the drive was OK and I reformatted it and used a registry hack to force Windows 2000 to accept the drive so I don't get complete data loss again.

 

[Muse]

So, I reboot yesterday several times, there's no sign of this 160 GB HD, which has two partitions, one NTFS the other FAT32. No sign of it in the BIOS.

This morning I boot Windows 2000 and it's there! I'm busily copying the data on it I don't want to lose, there's not much of it as I say. Gotta leave for a while, but I'll attend to this later. Guess I should run Seatools? I suppose I can expect a complete failure in the near future?

 

[Jiggz]

You're absolutely on the right track. Most likely the hdd is on its last leg.

 

[Captante]

Originally posted by: Muse
So, I reboot yesterday several times, there's no sign of this 160 GB HD, which has two partitions, one NTFS the other FAT32. No sign of it in the BIOS.

This morning I boot Windows 2000 and it's there! I'm busily copying the data on it I don't want to lose, there's not much of it as I say. Gotta leave for a while, but I'll attend to this later. Guess I should run Seatools? I suppose I can expect a complete failure in the near future?

Sounds like a good idea becuse theres no telling how long it has left... I'd say recover all the data you can & just RMA it to Seagate.

 

[Muse]

So I leave the machine running for 6 hours, come back and turn on the monitor and see a BSOD! Turn off the box a few hours, reboot and the drive is back. Compute an hour or two and still see the drive in Explorer. However, it's a drive that in general doesn't get accessed unless I'm doing ripping or HTPC stuff. I click on the drive and hourglass.... hourglass,.... then BSOD! I figure the drive is dieing. I could try it in my other desktop system or save myself the trouble and just send it to Seagate.

 

[memories2002]

Run a "chkdsk X: /r"

 

[j00fek]

you need to run the diag software to get a code and start the rma process on seagate.com

that will also tell you too if the drive is dead

 

[Muse]

Originally posted by: memories2002
Run a "chkdsk X: /r"

There are two partitions, H: and I:. Do I drop to a CMD prompt in Windows and run it there with X = H or I?

 

[Boyo]

H then I.

 

[Muse]

Originally posted by: Boyo
H then I.

If the HD is failing, I don't think it will make it through the test without a BSOD. Will chkdsk report hardware errors? Before the drive heats up enough to fail again?

I have a mind to move the disk to my other desktop system and see how it performs there instead of trying to run chkdsk on it in the current system. If it fails in the other system (different MB, controller -- I think I have a Promise IDE controller in there, but can run it on the MB or Promise PCI card, whichever is most convenient), different cable, different RAM) I think I can simply assume the HD is failing and send it to Seagate. Maybe it will test OK in the 2nd system and I can start looking for other causes in my main desktop system:

Virus
bad RAM
bad cable
IDE controller problems on the MB
?

BTW, the BSOD I saw yesterday (first one) said this:

STOP: 0x00000077 (0xC000000E, 0xC000000E, 0x00000000, 0x01C11000)
KERNEL_STACK_INPAGE_ERROR

 

[ForumMaster]

what you should do is run seatools to check the drive. you could also check it in your other rig but it does seem like your drive is dieing.

BTW, do you have S.M.A.R.T. enabled in the BIOS? try looking at the S.M.A.R.T. data.

 

[Muse]

Originally posted by: ForumMaster
what you should do is run seatools to check the drive. you could also check it in your other rig but it does seem like your drive is dieing.

BTW, do you have S.M.A.R.T. enabled in the BIOS? try looking at the S.M.A.R.T. data.

There is still data on the drive. I think it isn't important data (backup or backed up), but I'm not compleTexttely sure. Like I said, I sort of doubt I can complete chkdsk or probably Seatools either on it if it continues to crash after heating up. Trying it in the other box would tell more more in that event. I could pretty much rule out anything other than drive failure if it crashes in the other box too, I figure.

I don't know about S.M.A.R.T. Don't know if my MB supports it or if I have it enabled. Would it be enabled by default? I'll go in there and try to find it. What screen would it be under? I'll have a looksee now. I won't boot Windows, just go in the BIOS. Thanks!

Edit: The specs on my mainboard don't appear to mention S.M.A.R.T.:

KT3 Ultra2

 

[tkistre]

If you download Seagate's Diagnostic Software, it will run in DOS, so no BSOD. You will have to put the software on Floppy or CD and boot from which ever you use.

Depending on what exactly is wrong with the drive, it isn't uncommon to see the drive one day and not see it the next time you boot up. The fact that it is doing this IS your sign that something is wrong. Like "Captante" said above, back up your data while it's working, if it still is. Good luck!

 

[Muse]

Originally posted by: tkistre
If you download Seagate's Diagnostic Software, it will run in DOS, so no BSOD. You will have to put the software on Floppy or CD and boot from which ever you use.

Depending on what exactly is wrong with the drive, it isn't uncommon to see the drive one day and not see it the next time you boot up. The fact that it is doing this IS your sign that something is wrong. Like "Captante" said above, back up your data while it's working, if it still is. Good luck!

Thanks. I kinda think there's no data on it I don't have elsewhere. It was a backup drive and for DVR/HTPC (HDTV timeshifting and recording).

So, I downloaded the latest Seatools and burned to a CDRW, but neither of my desktops will boot from it for some reason. I have a Seatools CD that I wrote in 2006 when I had another problem (coincidentally, I think, with the same drive). That one I can boot to and I'm running a "Full" diagnostic right now. I'm using my other desktop to write this. I removed the drive and put it in my seldom-used desktop system, this time running on the Promise controller card, not the mainboard IDE channels. Anyway, it hasn't failed yet, and I'll have to see how it does on the tests. I figure it might pass the tests and still be dieing. There's other things I can do including putting one of my unused 80 GB Western Digital drives on the channel I had the 160 GB Seagate on and see if there are problems.

 

[Muse]

So, I have the 160 GB drive in my second system and booted from a Seagate Seatools CD made Feb. 2006. I'm running Full Diagnostics on the drive and it's been about 15 hours. There's no way in hell to know how long it will last except that a number shown keeps incrementing. It says this presently:

Sectors... 2385920


So, does this mean it will finish by April?

There's no other information given. There's a scrollable window below that but it's blank. Maybe any errors found are supposed to show up in that. The utility is not so hot. I'm getting no mouse support in it. Maybe I would if I hooked up a PS2 mouse to the box. Currently I'm using a KVM with the two desktop systems, can toggle the monitor to check out what's happening. Yesterday evening after running Seatools for about 5 hours I hit some keys to see if I could get the cursor onto something interesting. Basically I just hit the spacebar a bunch of times. I saw absolutely no response in the program. I came back a while later to find that the program had terminated. It thought the test had been cancelled. So, I started the test all over again and have left well enough alone. It's 15 hours now, like I say. Maybe I should move the drive back to where it was in my main system and test there. First I'm going to attach a spare 80 GB HD on that channel (Primary/Slave) and see what transpires.

 

[Muse]

I swapped out the HD for another and got another crash, so I figure there's nothing wrong with that first HD. So, I figure it's either the IDE cable or the IDE channel on the MB. I swapped out the IDE cable and now the system does not respond in any way when I press the power button. Please help!

 

[btcomm1]

Ok, couple of things. Did you try the old IDE cable back the way it was? Second is are you sure there is power plugged into every device that is connected though IDE? Also when you tried the new drive did you remember to set the jumpers to the correct settings on the new drive? What if you disconnect all drives from the motherboard does the computer repsond? Do you get any video? If it's not even powering on anymore maybe it's your power supply or possibly the motherboard that was going bad.

 

[Muse]

Originally posted by: btcomm1
Ok, couple of things. Did you try the old IDE cable back the way it was? Second is are you sure there is power plugged into every device that is connected though IDE? Also when you tried the new drive did you remember to set the jumpers to the correct settings on the new drive? What if you disconnect all drives from the motherboard does the computer repsond? Do you get any video? If it's not even powering on anymore maybe it's your power supply or possibly the motherboard that was going bad.

Have not tried the old IDE cable. I have others, actually.

I have no way of knowing if "there is power plugged into every device that is connected though IDE." The system gives absolutely no response to the power button. The last response I got to that power button on the ATX case was when I pressed it to shut off the system from the blue screen when the system crashed with the substitute HD in there. I think strongly now that pressing that power button then was a mistake. Well, maybe those things are engineered so there's no additional strain on the power button under any circumstances, so maybe I'm thinking wrong there. Anyway, I have not forgotten what was told me by the guy who sold me my first system. He was an EE and he'd built the system from parts, a 486DX33 proprietary local bus machine. He said I should turn the system on and off by toggling a power strip and not press the on/off button on the computer case in order to prevent eventual failure of that button. I don't know how to troubleshoot that button other than maybe finding a replacement for it or using another ATX case. I do have another ATX case, but right now it's hosting my only working desktop. I'm using my laptop to write this message.

The jumpers on the new drive were set OK. It worked in the system, just like the other HD did until the system warmed up for a couple of hours. It's only after serious warming up and/or grinding that the system crashes.

I figure your last sentence has it -- the PSU or MB gone bad, or possibly that on/off button. I figure my next move is to swap PSU's between my two ATX desktop systems. If my working system works with the PSU from the other, I know that PSU is OK. If it doesn't work I can figure the PSU has died. If the problem system powers on again with the other PSU I know the switch and MB are OK and I can take it from there. It's a little hard for me to imagine that there's only one problem now, but maybe there is. Maybe the MB is gone. I might have mashed it when I switched the IDE cables.

 

[btcomm1]

If you want to test the power button theory you can disconnect the power switch wire and try to start the computer with a screw driver, you just touch the 2 power pins with a screw driver at the same time and it should start to boot up.

 

[Muse]

Originally posted by: btcomm1
If you want to test the power button theory you can disconnect the power switch wire and try to start the computer with a screw driver, you just touch the 2 power pins with a screw driver at the same time and it should start to boot up.

Yeah, my theory at the moment is that it's the power switch. One side of it looks to me like something broke off. It's not like the other side where a nub presses upward. I think the nub on one side broke off. There are 4 wires, though. I don't know what to do with them. There's two pairs, one pair on either side. One pair is red and white, the other pair is green (I'm pretty sure it's green, but I'm "total green blind" so I can't swear, but it "looks" green to me ) and white.

Edit: Yep, it's the switch! I traced the wires back to the MB, pulled off the connector and shorted the two pins with a screwdriver and it powered right up! Well, that EE was right who sold me my first system (which he custom made). He told me to use a powerstrip to turn on the computer so I wouldn't suffer this problem some day. That was 15 years ago... Well, that might work for an AT case, but I don't think it works for an ATX case where the button is a momentary connection switch like a door bell.

Now I still have to figure out why I was getting BSOD's the last week!!

 

[Muse]

It was definitely the power button breaking that stopped the system from starting up. So, that's a separate problem and I'll work that out tomorrow.

I figure there's probably nothing wrong with that HD, since the system crashed (Blue Screen) with a different HD in its place. I put the first HD back on the same channel with a different cable and so far no crashes. So, I figure that somehow the cable was at fault. Will monitor the situation. Meantime, I only use that drive for unimportant data, and that's always been the case. Now, if it had been one of the other two HDs in the system, I really would have been freaking out. Not horrible, but my boot drive has one partition with data that's largely not backed up. The other drive has the most important stuff backed up, but there's some data there that isn't backed up. Anyway, I think I can deal with this. If it starts crashing again I have to figure it's the Primary/Slave channel that's flaky and I'll just stop using it and put an IDE card in the box (I have an extra).