[TriLUG] SOLVED OT: the optimum location for inner track/cylinder on rotating media?

Joseph Mack NA3T jmack at wm7d.net
Fri Jul 30 17:49:54 EDT 2010


On Thu, 13 Aug 2009, Joseph Mack NA3T wrote:

> On Wed, 12 Aug 2009, Joseph Mack NA3T wrote:
>
>> The test of the theory (that the maximum is not used to placate users) 
>> would be to measure r for a device that the users can't see, like a floppy 
>> or hard disk. These should all have r=R/2. I can't find any floppies to 
>> disassemble. and I don't have any dead hard disks
>
> for the floppy disk r=R/2
> for the hard disk   r=R/3
>
> the floppy fits the theory, while the hard disk doesn't. This is disturbing.
>
> I now see that all disks have a region at the center of r=R/3 that isn't 
> used. The vinyl record has a label, The cd doesn't use the center section at 
> all. The hard and flopyy disks have a large hub (for the drive mechanism?). 
> In that case the max area allowed for the constant linear velocity type 
> platter is pi*R^2(1-1/9) rather than pi*R^2. This doesn't affect my theory 
> that the optimum location for the inner cylinder on constant angular velocity 
> platters is r=R/2.

this is an old thread.

My theory was that at constant angular velocity (and hence 
constant bits/cylinder), a hard disk would maximize storage 
capacity by having the innermost cylinder at half the radius 
of the disk. However inspection of a harddisk shows the 
innermost cylinder at R/3.

It turns out that harddisks (at least since the early 
1990's)

http://en.wikipedia.org/wiki/Zone_bit_recording

don't have constant bits/cylinder. They use ZBR (zone bit 
recording). The outer cylinders have twice the number of 
bits and the onboard hardware has to handle twice the data 
rate when the heads are on the outer cylinders. Presumably 
for the outer tracks, a "cylinder" as presented to the OS 
(if such a thing as a cylinder is used any more), is half of 
a physical cylinder.

The question then becomes the optimum location of the inner 
cylinder if the disk can have cylinders of 1x bits and 2x 
bits.

Joe
-- 
Joseph Mack NA3T EME(B,D), FM05lw North Carolina
jmack (at) wm7d (dot) net - azimuthal equidistant map
generator at http://www.wm7d.net/azproj.shtml
Homepage http://www.austintek.com/ It's GNU/Linux!



More information about the TriLUG mailing list