Power Supply killing HDDs?

[shaked]

Alright, here's my dilemma: I have been having some big trouble with my hard drives. I had one (WD1200JB) die on me right before school started. RMA'd it, and I had the same exact problem with the new one. They work ok (can access anything) but after being on for a bit they both would just start spinning up, as if they would when you turn them on. This leads to the computer freezing. I highly doubt the problem is heat, seeing as it's never over 30C in my room (it's the middle of winter in New England) and I have 5 80mm fans pushing air all over the place. Could my power supply be bad? Could it be other components? I've memtest86'd the machine before, and it went all night and it's prime stable as well. Does anyone have any ideas? Also, my slave won't take WinXP. I don't want to format it because it's just information, but every time i get Windows loaded on it, it corrupts files or something. I guess that drive is shot too. I need some serious help here. I'm on my laptop and I really miss my desktop. As for specs:

Athlon 2600+
2x512 Corsair XMS PC3500
Asus A7N8X Deluxe
WD1200JB
WD2500JB
Thermaltake Silent Purepower 480W

I have yet to check the rails on the PSU, but is that my issue? I want to get this problem fixed, but it's just annoying me to a point where i want to throw the whole thing out and start over. ANY help would be greatly appreciated. I'm sure I left out some details so feel free to inquire further if necessary. Thanks.

Rich

 

[LiLithTecH]

Sounds more like a problem with POWER MANAGEMENT than with the PSU.

If you use System Standby or Hibernation, the DEFAULT Power settings turn
the drive OFF after a specific time. You may want to change that to 'NEVER'
in the power settings.

As for the Slave Drive not acceptting WinXP, how do you have the drives configured?
(MASTER-SLAVE, CABLE SELECT(csel), etc...)

 

[resStealth]

can someone please tell me what you mean by checking the rails on the psu nad WHAT IN GODS NAME IS CABLE SELECT? i only know master/slave

 

[shaked]

The power settings are the same as they have been. I can double check them, but I'll be in the middle of doing something (talking on AIM for example) and it will just start ticking. The drive will tick maybe 6 or 7 times, then the computer freezes. The first drive was fine for over a year before this happened, the RMA'd one lasted 6 months.

As far as the slave, it will load WinXP fine (sometimes), but after I start updating and what not, I'll reboot and I'll get an error saying a file is corrupt. If i attempt to fix it using the XP cd, it just reinstalls and then I get even MORE problems (Windows won't even load after it copies files from cd). I'm starting to think it isn't the drives, which is why I asked about the PSU. I always check my jumpers for my drives. Something else has to be causing it. I'm going to RMA the drive (AGAIN), but I don't feel like wasting everyone's time. I'm ready to scrap the whole computer at this point to be honest.

Rich

 

[fishmonger12]

Originally posted by: resStealth
WHAT IN GODS NAME IS CABLE SELECT?

easy. cable selects which one it wants to be master and which one it wants to be slave.

have you tried Western Digital's drive checking utility? are you overclocking? sometimes overclocking is known to corrupt data on the hard drives.

 

[montag451]

Depends on which connector on the cable the drive is.

Both drives would need to be set to CS, and the drive furthest away on the cable from the mobo would be the MASTER.
The middle drive would be SLAVE.

Best though, to use either MASTER/SLAVE jumper settings.
If it is the only drive on a cable, see if it has 'SINGLE DRIVE' setting - if it has, then select that.

 

[Britboy]

I had the exact same problems you describe with my athlon a few months ago. It turned out to be a bad RAM chip. It was the last thing I tried, I'd already reinstalled the OS, new PSU, new harddrives, new HD cables etc etc

 

[montag451]

I don't suppose you have a voltmeter handy do you?

 

[Blazer]

might want to have a good electrician check the outlet that you use for your sys,may be a weak/open ground/ or overvoltage

 

[rocketPack]

I would be more akin to agreeing with Britboy about the memory. Memory problems can lead to all kinds of wierd stuff, for example one time I was working on a computer that was on and I accidently drug a piece of conductive material across a chip on a stick of ram and the computer rebooted. Seems odd that a short across on ram chip could reboot the entire machine, but it did. I've also seen many many times where formatting and reinstalling windows to a drive did no good, replaced the motherboard with no luck, hard drive, everything.... then finally the ram, and that was it. I would recommend at least taking your memory out and blowing out the slots (could be a dust short) and then perhaps replacing them in different configurations. Or replacing them with memory from a working system and see if the system has the same problems.

And to be honest, memory test programs don't often know what's going on. I did a 45 minute test on the ram that produced the errors in windows and random problems, and all tests came back as PASSED. Go figure. I think for those tests to fail the memory has to be so far gone it wont even boot.

Anyway, the clicking of the drive does sound like more of a mechanical issue but some IDE controllers do strange things, we have one machine that powers off the drives during shutdown, then powers them back on, and then off again. All the while there is a lot of audible clicking going on. That's an IBM drive though, they always sound like a piece of heavy machinery.

I'd RMA that WD for good and get a Seagate personally, why pay the same (or sometimes more) money for a drive with a shorter warranty? Seagate owns.

 

[VirtualLarry]

Originally posted by: rocketPack
Anyway, the clicking of the drive does sound like more of a mechanical issue but some IDE controllers do strange things, we have one machine that powers off the drives during shutdown, then powers them back on, and then off again. All the while there is a lot of audible clicking going on. That's an IBM drive though, they always sound like a piece of heavy machinery.

Well, the "ticking" generally means that the drive is attempting to re-read the sector a few times before giving up and returning an error, you can hear the repeated seek noises when that happens usually. Unless that happens right after a spin-down/spin-up-sounding event, that means that the HD has reset itself and is running through its own internal POST re-calibration/testing routines. (If it sounds just like it does when it normally first powers up - it usually makes a few ticking sounds then too.) If that happens to be the case - check your power + data cable connectors, and consider replacing the IDE cable. I've had issues like that before, either the molex connector on the HD is slightly loose or broken and doesn't make good contact, sometimes it could get bumped and then would randomly spin-down/spin-up/restart on me. Likewise, with a damaged IDE cable that the RESET or similar line was getting triggered somehow.

I can't say for certain what the cause is, it could simply just be bum luck in getting two bad drives. But check those first. Also, yes, if your PSU is unstable/underpowered, that could be the reason too - if the drive doesn't get enough power, it could enter the shutdown/spindown phase, and then when it gets enough power (PWRGOOD signal transition on the HD PCBA), it spins up again and resets itself.

Do you get any error events in the system Event Log? What are they? What brand chipset/IDE controller are they attached to?

 

[shaked]

Okay, I just changed the IDE cable and am running the computer now. First thing, Windows did the "disk check" and fixed what I guess were errors. Second thing, data on my slave is corrupted (ie: some songs dont play, so I assume it needs to be scanned and fixed). So that's two drives with immediate problems. I'm using an alternate cable, so if it holds out through the night we'll see if that was indeed the problem. Otherwise, I'm really still just searching for issues and nothing is coming to mind.

Oh yeah, Larry, no there are no error logs. When the HDD freezes, it literally freezes everything and there's no time for documentation.
Rich