[TriLUG] Hard drive recovery

[Brian via TriLUG]

On 4/13/19 6:12 PM, Joseph Mack NA3T via TriLUG wrote:
> 
> I've always been leary of rocking a spinning disk. I don't like moving a 
> working disk. I worry that the platters will gyroscopically tilt and 
> crash into the head(s). I have no idea how much leeway you have to do 
> this. You say gentle. If you say it's OK, I guess it is.

In general, I agree; moving a spinning disk seems risky (at least, ones 
that aren't designed for it, e.g. laptop drives), certainly if the 
motion is linear and not parallel to the drive's rotational plane.  That 
said, I'm confident that gentle rocking (just enough to feel the 
gyroscopic effect) isn't going to cause significant radial displacements 
inside the drive.  The bearings, after all, are designed not to move in 
those directions.  A white paper published by Western Digital [1] 
explaining the decision to move from ball bearings to fluid dynamic 
bearings mentions a "non-repeatable runout" (that is, random 
side-to-side motion within the rotational plane) of 0.01 micro-inches 
(0.00000001 inches); if linear movement is that tight, radial movement 
will be similarly well-resisted.

Don't shake the thing like a maraca, but gentle rocking (it really 
doesn't take much to feel the effect) if you can't hear the motor should 
be harmless.  YMMV, IANAE, TANSTAAFL, etc.

-B

[1] - 
https://documents.westerndigital.com/content/dam/doc-library/en_us/assets/public/western-digital/product/internal-drives/eol/endurastar-series/white-paper-endurastar-j4k100.pdf?_ga=2.87351000.1372394219.1555347520-849722421.1555347520