Re: Floppy disk density

[mjmahon@aol.com (Michael J. Mahon)]

Pete Rittwage wrote:

>"Paul Guertin" <pg@sff.net> wrote in message
>ld6bo0dsfpd7do52v48qgib11uu9ppft83@4ax.com">news:ld6bo0dsfpd7do52v48qgib11uu9ppft83@4ax.com...
>
>> If you slowed down the motor of a Disk II drive to half its
>> normal speed, could you reliably write twice as much data on
>> a standard, double density floppy disk (assuming a customized
>> RWTS)? There must be a range of allowable speeds for a floppy
>> drive, and I wonder how wide that range is. Has anyone here
>> done any experimenting or read literature on this subject?
>
>The media itself is rated at 5600-5900 bpi (bits per inch) for double
>density disks at 300rpm.  On a Commodore drive using raw GCR (sorry, that's
>my background but I think Apple used the same technology) you can get about
>7700 GCR encoded bytes on a track max depending on the density setting at
>300rpm.
>
>So no, a little more data, sure.  Double, no way.  If it were that easy they
>would have done it 25 years ago.  :)

I'd like to carry your analysis a little further, and be a little more
explicit for purposes of comparison.

Diskette media was originally designed with FM encoding in mind,
which has an encoding efficiency of 50% (alternating clock and data
bits).  Apple GCR, from DOS 3.3 onward, uses 6+2 encoding, in which
6 data bits are encoded in every 8 bits recorded on the disk, resulting
in a 75% encoding efficiency.  This produces a capacity increase of
50% over FM (single density) recording.

You are correct, the maximum reliable linear bit density of the media
is the ultimate capacity limiter.  But each track is a different length.

The inner track has a radius of about 1.375" and the outer track has
a radius of about 2.4".  If the motor speed and the recording bit rate
are constant, as they are for most floppy drives, then the capacity of
the shortest track determines the capacity of all tracks  For a 5.25"
diskette, if the motor speed or bit rate were varied to keep bit density
approximately constant, the capacity of the diskette would increase
by a factor of: 0.5 * ([max radius]/[min radius] + 1), which in this case
works out to about a 37% increase in capacity.

To achieve this, of course, means either varying the motor speed
and keeping the bit rate constant or varying the bit rate and holding
the motor speed constant.  Apple chose the former path for their
800K 3.5" floppies, and more recently has switched to the latter
approach for their "univeral" 3.5" floppy drives.

The variable motor speed option was not available when the Disk ][
was designed, since Apple was a minority player with little leverage.

And using a variable bit rate would have made the elegant Disk ][
Controller much more complex.

It seems that Woz figured that the relatively easy factor of 1.5 that
he got by using GCR instead of FM was enough, and that the extra
1.37 obtainable by a variable bit rate design was not worth the
complexity and hardware cost required.

-michael

Check out parallel computing for 8-bit Apples on my
Home page:  http://members.aol.com/MJMahon/