Re: Hard Drive Destruct System?

From: Al Dykes (adykes_at_panix.com)
Date: 11/27/04 

Date: 26 Nov 2004 18:31:00 -0500

In article <slrncqfbjh.fl0.ibuprofin@compton.phx.az.us>,
Moe Trin <no.mail.accepted.sorry> wrote:
>In article <1d54b7e4.0411251822.729ae8cc@posting.google.com>, mike3 wrote:
>
>>Would this make a good mechanism to securely destroy a hard drive?
>>
>>1. Crash the heads into the platters with the drive at top speed
>
>Modern (since the early 1990s) hard drives have a very hard and smooth
>surface on the platters. Crashes do occur in real life, and normally
>cause no harm. "Top speed" implies the disk powered up and running
>normally (heads NOT parked). How do you propose to cause the 50+Gs to
>crash the heads while in this condition? Hard drives have gotten a lot
>more resilient since Reynold Johnson's "baloney slicer" system in 1956.
>
>>2. Seek them from one edge of the platters to the other back and forth
>>to \
>> ensure good grinding
>
>See above. Every time you power down a drive, you "crash" the heads
>(admittedly at a slower speed) - how much damage to the heads does that
>cause? To the media?
>
>>3. Pump a concentrated hydrogen/oxygen mix into the hard drive platter
>>chamber
>> while the heads are grinding away.
>
>If you are going to be cracking the case (airflow through the vent is
>meant to equalize pressures, not cause an air exchange), why not open
>it up the rest of the way, and yank the platters so that you can put
>them directly in a fire and melt them.
>
>>Would the heat from the grinding be enough to ignite the H2/O2 mix and
>>ensure that data can't be recovered?
>
>No. Not enough energy in the H2/O2 that would remain in a typical hard
>disk without exploding and blowing open the case. Such an explosion
>might damage the drive to make it unusable, but wouldn't damage the media
>enough to preclude data recovery.
>
>>Would this mechanism be good enough to keep copies of a large company's
>>trade secrets from a competing large company?
>
>No.
>
>1. Go find a recent copy of a hardware book such as Scott Mueller's
>"Upgrading and Repairing PCs" from Que, Roche's "Hardware Bible", and
>learn about hard disk construction.
>
>2. http://www.cs.auckland.ac.nz/~pgut001/pubs/secure_del.html for a
>classic paper on removing information from computer disks.
>
> Old guy
>

An interetesting paper, but it's dated 1996 and has to reflect the
disks that were shipping then. Today's disks have about 100x the data
density (GB/sq inch) and it has to change how much erased data can be
read. I've nver seen this discussed, but I bet it's lots harder to
read data that's been overwritten. 

-- 
a d y k e s @ p a n i x . c o m 
----

Relevant Pages

  • Re: 2GB CF card vs. 4GB Microdrive....
    ... The biggest vulnerability with any hard drive is the delicate gap between the heads and the surfaces of the platter- if the unit has powered down then the heads are already in contact with the platters in a reserved area, and are held against the surface with a force that might surprise you. ... I have not a clue whether a MD has the same insides as other hard drives, nor whether the heads were doing the reseek dance while they bounced about, but my guess is my hammer is ...
    (rec.photo.digital.slr-systems)
  • Re: 80% maximum capacity on a hard drive?
    ... on those days I only know HD heads flying on pure air. ... of head-per-track disks with the entire HDA enclosed in a sealed helium ... the platters, R/W heads, pre-amps, and two He pumps. ... One He pump maintained positive pressure inside the can -- better to have ...
    (comp.sys.ibm.pc.hardware.storage)
  • Re: hard drive repair
    ... WD hard drives has many incompatible type of electronical ... boards and blocks of magnetic heads... ... > I'm trying to repair my crashed 200Gb Western Digital hard disk. ... > platters couldn't be moved around by hand. ...
    (comp.sys.ibm.pc.hardware.storage)
  • Re: Myford lathes vs. modern "equivalents"
    ... data platters from hard drives. ... It needs to consist of a circular, flat head, to ... these fingers has to be adaptable to a number of places around the main ... I have dismantled scrap hard drives in the past and whilst I do not have one ...
    (uk.rec.models.engineering)
  • Re: Why magnetic drums was/are worse than disks ?
    ... I already told you that the maximum number of disk platters isn't decreasing much at all. ... There's nothing about thermal expansion that would 'encourage' reductions in platter count: some desktop drives are being made with fewer platters (and heads) so that they can be manufactured more cheaply - in fact, some drives have been made that use only one side of one platter. ... Split the difference to make a round number, and say that the multi-head assembly length can vary with respect to the 'pencil' length no more than +/- 0.000001 inch, at least if you want track density comparable to a contemporary disk's. ...
    (comp.arch)