Jonathan N. Little wrote:
nospam wrote:
In article , Good Guy
wrote:

I would try removing from the machine and putting in a USB enclosure and
try using File explorer to recover as much as possible.

the clicking is a bad sign and isn't going to magically stop if it's in
a usb enclosure.

Agreed. That is the sound of a bad controller. Data recovery is VERY
expensive. Years ago a local business loss a drive after and send drive
to OnTrack and was nearly 6K for 3GB of data and this was 15 years ago.


I agree 15 years is a very long time but Hard Drives do lasts longer if
enough care is taken not to misuse them.


Platters can lose data from just sitting over the years, but that is not
OPs issue. Definite sign of bad controller. Options a

1) You have backup right? (best option)
2) Data recovery (very expensive option)
3) Get another drive of the same model that works and swap platters.
(not easy on many aspects).

I have done #3. First it will be hard to find another 15 year old drive.
I was lucky to have several on hand. It requires that you CAREFULLY
disassemble both drives. Requires micro torx drivers and great care.
Hardest part is removing the the head arm without scratching your
platters. You have to get it out of the way to swap platters. The rare
earth magnets are VERY strong. Then you have to connect it up, cross
fingers and if lucky start dumping data as fast as you can. I do not
bother putting cover back on just trying to dump data to good drive as
fast as possible. That said, odds of success a small, but if #1 & #2 are
not possible what do you have to lose?

The price of data recovery has come down over the years.

The price is proportional to the amount of effort put into it.
In some cases, recovery is achieved without disassembly, in
which case the price charged still gives a very healthy profit.

If the drive responds and gives the ID string, that proves that
quite a bit of the drive is working.

The controller only contains a bootstrap code, and the controller
has to move the heads out and read the Service Area to become
fully functional. The firmware (for ATA command set) is loaded
on the Service Area, as well as data structures for sector
sparing and so on.

If the drive will ID itself, that means the SA reads OK.

The drive has the option of running an internal SMART short
test, which could cause a series of accesses not under
the users control. Attempts to recalibrate, banging the
heads against the stops, might have actually been triggered
by some early SMART activity.

If a head was bad (say, one of eight heads), the drive
might stumble during the SMART test, but succeed when reading
the SA. There's no particular reason the SA has to take up a
full cylinder of space and require all heads to be functional
to get it loaded.

Before doing anything, I'd be running the exact make and model
number through Google, to see what's known about fault modes.

Some of the older drives (my two dead Maxtor 40GB drives) were
*******s when it came time to fail. From first symptoms, to
non-responsive (dead), only took 24 hours or less, which means your
"execution" during that interval had to be flawless if you wanted
your data back for free. And checking the history of a drive,
may indicate how they perform during that time.

Sometimes even the data recovery sites, keep a web page for
specific models with "noteworthy" failure modes. Like the
drives where the motor controller burns out (giving the
impression the motor controller chip simply wasn't
designed properly in the first place).

If the spindle is stuck (FDB motor lost lubricant and the
spindle seized up), the motor controller is uses to "pulse"
the spindle to try to free it up. This gives a characteristic
noise. And the pattern applied also doesn't try to burn out
the motor controller - the motor controller is designed
for a current flow profile, in a sense being current
limited. It'll try for some number of seconds, before
giving up.

Paul