Hello Melkuth,
We concur with the info and comments from the other contributor -
overheating !
Aurora IT Systems are one of Europe's foremost Data Recovery and
Forensic labs and are actually located in Sweden. We see many incoming
Maxtor HDD's each and every day of the week. LaCie external drives
have a similar failure rate (any guesses which drive they usually use
?)
You requested some advice; here is at least some background to your
problem:
The failure is almost always due to overheating of the fluid bearing,
causing either vibration of the platter (the actual disk on which your
data is recorded) or a complete motor stop. Either case requires most
advanced data recovery techniques:
IF the platter vibrates, the outer edges of the platter can actually
touch the read head. The read head should be floating on a microscopic
cushion of air and should never come in physical contact with data
stored on modern HDD's (that's Hard Disk Drives) When it happens,
the energy released causes an equal and opposite movement of the tip of
the servo arm supporting the read (and write) heads; since the servo
arm is usually constructed of a thin strip of non-magnetic stainless
steel, it now functions as an excellent spring . . .
The read head now returns to the media, crashes into the surface again
and bounces several times, rather like throwing a flat pebble across a
smooth water surface.
You think that's bad news . . . it gets much worse.
The above mentioned head collision with the highly polished magnetic
media will almost always caused some particles to be released; these
particles are referred to as debris. This debris now rotates within
the, previously surgically clean, hermetic enclosure of your HDD. Since
the air within the HDD is rotating just slightly slower than the actual
platter and motor combination, it is a very short time before the
rotating cloud of debris collides into the read head, in practice about
10 milliseconds after the first crash.
If the actual crash occurred near the center of the platters, much of
the debris will miss the read heads due to the fact that centrifugal
force within the vortex will move many particles outwards. Note that I
use the term 'much', not 'all'
Many of the particles that now collide with the head will stick to the
previously polished surface of the read heads, very much like pollen
attached to the rear legs of a honey bee.
// some excellent microscopic images of this effect can be seen on our
very large web site - aurora.se //
Unfortunately these particles are both very hard and also very sharp,
they instantly escalate the first crash site into more serious
concentric scratches, which in turn produced more debris etc etc . . .
In many cases much data can be successfully recovered after
considerable effort, changing read heads and cleaning the entire media
surface.
(I hope that all this is not too technical).
The other scenario: total motor stop
Even worse; the platters ( just look on the aurora site if you're
curious ) are assembled using an industrial robot, and at the moment of
production have no particular positional or angular relation to each
other. Once the drive is factory formatted, a great deal of technical
data is stored on specially used zones on the HDD, this area has
several names; service area and factory zone among others. Access to
this region of the HDD is paramount to ANY recovery attempt.
From this moment onwards the position of each platter relative to other
platters is most critical. We work with microscopes equipped with laser
markers and nano manipulators to index the current reference of each
platter before disassembling the drive to its component parts. The
entire drive is then rebuilt using a donator or drive having almost
exactly similar characteristics, even the magnet and servo arm must be
exactly replaced.
Well, that's just a part of the story, to succeed you also will need
about 20+ years of experience and preferable at least a good degree in
physics and electronics. You will also require head lifters,
nano-manipulators, optical microscopes and, in our case, an array of
Tektronix analysis and test equipment. Lastly you will require
facilities with air showers, surgically clean lint free protective
clothing, environmental control and breathing filters and obviously,
clean working areas with at least class 100 status.
If you are a student, we would be happy to take a look at your drive
without charges, but to be very honest with you, the cost of recovery
in the above scenarios (even taking account of our generous price
reduction for private persons) is usually prohibitive.
Under NO circumstances whatsoever should you or anyone else open the
drive, nor should you even attempt to restart the drive unless you are
prepared to loose your data forever.
Contact us if we can be of any help,
Tony Kvarnstrom
CTO
Aurora IT Systems AB
SWEDEN
