Hard Disk Failure

Ben Scott dragonhawk at gmail.com
Tue Feb 14 10:16:00 EST 2006


On 2/14/06, Christopher Chisholm <christopher.chisholm at syamsoftware.com> wrote:
> An old trick is to put the drive in a freezer overnight.  While I really
> can't explain it, sometimes you'll get a few minutes of time with the
> drive when you take it out.

  There's a couple of things that trick can address.  One is stiction.
 Sometimes bearings will start to seize or heads will stick to
platters.  Rapid cooling/heating can sometimes cause enough
contraction/expansion to free things up for a little while.  The other
thing is that marginal electronics components will sometimes start
working again if you cool them down.

  Another old trick was (seriously) to gently hit the drive on the
side with a hammer.  Stiction again.  That was generally considered a
last resort, though, for obvious reasons.  :)

> Another option (one I've never tried and probably is not a good idea) is
> to buy another of the same model drive and try swapping the platters
> from the bad drive to the good drive.

  Eeesh.  Yah, I wouldn't try that.  If one bit of dust gets between
the heads, you could make things a lot worse.  A lot safer would be to
just swap the PCBs.  If it's an electronics failure (as opposed to a
mechanical or media problem), that will fix it.  But the drives
generally have to be *very* similar.

> I believe this is what the drive recovery people do.

  They have a lot of tricks; that is one of them.  They use clean rooms, though.

  One demo I read about used a special rig where they mounted an
individual platter on a gimble.  They used it when platters were
physically bent, to move it around a fixed read head.

-- Ben



More information about the gnhlug-discuss mailing list