I suspect one of my 3 hard drives is going bad on my old Abit build.
I've been hearing some noise for a while and yesterday when I turned
the pc off there was a fairly loud SMACK just before the fan lights
went out. Since then everything seems to work ok and HDTune seems to
think their Health is OK. Is there anything I can be looking for to
try and id the failing drive?

--- PyGate Linux v1.5.14
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On Wed, 4/22/2026 1:50 PM, crasso@nycap.rr.com wrote:

I suspect one of my 3 hard drives is going bad on my old Abit build.
I've been hearing some noise for a while and yesterday when I turned
the pc off there was a fairly loud SMACK just before the fan lights
went out. Since then everything seems to work ok and HDTune seems to
think their Health is OK. Is there anything I can be looking for to
try and id the failing drive?

If this is a 0.8" high hard drive, instead of a plastic landing
ramp off to the side, it may be landing the heads in a landing
zone near the hub. Presumably there is a "stop" at the
arm bearing, that prevents the arm from attempting to "climb the hub"
which would destroy it. The voice coil would not de-energize
unless the transition to that location had completed.

This is my dead ST 3250310AS 250GB drive, showing it still has a
bumper of some sort, on the voice coil end. That's so it doesn't hit
the stop too hard. If the rubber comes off, then you'll get an
acoustic effect. I included the cover in the shot, to show there
are seven T-9 Torx screws, one screw fits into the center of
the voice coil bearing.

[Picture] landing-zone-drive-bumper-on-arm.jpg

https://postimg.cc/9DNRN5jv

https://imgur.com/a/FwnlXGq

The white filter-pak in the upper left, the surface is clean
and white, so the surface of the platter did not fail. Yet the
evidence is, the surface is damaged, based on numerous CRC errors.
We have seen pictures of previous generation drives, where the
plateup was "rust colored", where the filter pak was filthy
and dark colored from debris. And yet such drives had been
functioning right up to the end... Makes you wonder just how
small the debris flakes are. Must be beyond-microscopic.

I have two of those drives. The opened one is a total failure.
The running one has gone read-only and has no spares left. I
tested a data recovery software which is "better than ddrescue",
and within 24 hours of read attempts, the drive was completely
scanned and had 100 detected CRC errors (which could not be improved
by many many read attempts).

You can see in the picture, there isn't a lot in there that can make a
noise.

On a modern drive, there is a plastic landing ramp off to the side.
I don't know if the arm can fall off the ramp and down beside the platters
or not. I don't know if there is clearance for that. I don't have any
landing ramp drives suited to picture taking. Everything above 250GB
is still running (even if more of these landing zone drives are
slowly on their way out, the 500GB ones). I think the secret to
the landing zone drives, is to leave them spinning 24 hours a day.

*******

You should run the drive model numbers through Google, to see
if any of the drives happen to be "notorious ones" :-) I've found
field reports (of symptoms) to be quite prescient. If someone
describes a certain noise a model makes, well damn if I'm
not hearing that too.

Paul

--- PyGate Linux v1.5.14
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On Wed, 22 Apr 2026 15:46:07 -0400, Paul <nospam@needed.invalid>
wrote:

On Wed, 4/22/2026 1:50 PM, crasso@nycap.rr.com wrote:

I suspect one of my 3 hard drives is going bad on my old Abit build.
I've been hearing some noise for a while and yesterday when I turned
the pc off there was a fairly loud SMACK just before the fan lights
went out. Since then everything seems to work ok and HDTune seems to
think their Health is OK. Is there anything I can be looking for to
try and id the failing drive?

If this is a 0.8" high hard drive, instead of a plastic landing
ramp off to the side, it may be landing the heads in a landing
zone near the hub. Presumably there is a "stop" at the
arm bearing, that prevents the arm from attempting to "climb the hub"
which would destroy it. The voice coil would not de-energize
unless the transition to that location had completed.

This is my dead ST 3250310AS 250GB drive, showing it still has a
bumper of some sort, on the voice coil end. That's so it doesn't hit
the stop too hard. If the rubber comes off, then you'll get an
acoustic effect. I included the cover in the shot, to show there
are seven T-9 Torx screws, one screw fits into the center of
the voice coil bearing.

[Picture] landing-zone-drive-bumper-on-arm.jpg

https://postimg.cc/9DNRN5jv

https://imgur.com/a/FwnlXGq

The white filter-pak in the upper left, the surface is clean
and white, so the surface of the platter did not fail. Yet the
evidence is, the surface is damaged, based on numerous CRC errors.
We have seen pictures of previous generation drives, where the
plateup was "rust colored", where the filter pak was filthy
and dark colored from debris. And yet such drives had been
functioning right up to the end... Makes you wonder just how
small the debris flakes are. Must be beyond-microscopic.

I have two of those drives. The opened one is a total failure.
The running one has gone read-only and has no spares left. I
tested a data recovery software which is "better than ddrescue",
and within 24 hours of read attempts, the drive was completely
scanned and had 100 detected CRC errors (which could not be improved
by many many read attempts).

You can see in the picture, there isn't a lot in there that can make a
noise.

On a modern drive, there is a plastic landing ramp off to the side.
I don't know if the arm can fall off the ramp and down beside the platters
or not. I don't know if there is clearance for that. I don't have any
landing ramp drives suited to picture taking. Everything above 250GB
is still running (even if more of these landing zone drives are
slowly on their way out, the 500GB ones). I think the secret to
the landing zone drives, is to leave them spinning 24 hours a day.

*******

You should run the drive model numbers through Google, to see
if any of the drives happen to be "notorious ones" :-) I've found
field reports (of symptoms) to be quite prescient. If someone
describes a certain noise a model makes, well damn if I'm
not hearing that too.

Paul

1. I dithered for awhile, vainly watching HDTune and holding my
fingers on the drives when they shut down. Couldn't find a clue. So...
I threw money at it. Ordered a 2T Western Digital hard drive off
Amazon on the theory that the boot drive, with partitions to boot xp,
win7 and win 10 would be the most troublesome to replace if it REALLY
did blow up on me. The #2 drive, a 1T with 4 data partitions, I backed
up and will do so monthly now. #3 drive is just a 1T backup drive,
that won't cripple me if it blows. New drive arrived without being
either GPT or MBR and DiscManagement let me choose MBR. 3 hours of
Macrium cloning and it was ready to boot. Only real pain (fingers
crossed) was chasing the scrambled drive letters Windows picks out.
Now I guess I wait for a disk to fail, unless it really was that boot
drive. I couldn't see anything physically wrong with any of the ones I
was shuttling around.

Thanks again for the suggestions.

--- PyGate Linux v1.5.14
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)