Date : Sun, 18 Feb 1990 14:05:24 EST
From : dg%lakart.UUCP@XAIT.Xerox.COM (David Goodenough)
Subject: 10 512-byte sectors/track (was RE: Kaypro...)
slsw2@cc.usu.edu (Roger Ivie) writes:
> dbraun@cadev5.intel.com (Doug Braun) writes:
>> How is it that this format has 10 sectors per track, while IBM PCs
>> have only 9 (and the old ones could only get 8 in)?
>
> If you omit the soft index in the standard format you get enough space
> to just squeeze in an extra sector IF your drive is within about 0.5%
> of being on speed.
>
> DEC did this for the RX50 (^&*#@&$ piece of trash). Unfortunately, the
> drive is speced at 1.5% speed variation, so you are not guaranteed that
> you can format a disk on the damn thing.
This is only part of the story, the other half concerns the FD controller
chip. Kaypros use the WD 1793, whereas the PC uses the Intel 765. Basically,
the 765 is a piece of junk, in that it is far less fault tolerant than the
1793, hence the 1793 can operate reliably with a shorter intersector gap,
thus cramming more information onto a disk.
To cite another example, My Televideo uses a 1793 to get 18 * 256 bytes
on one track, whereas the 765 can get only (I think) 16 * 256, I'd have
to check the documentation.
To take this to it's logical conclusion, in some experiments with variable
sector sizes, I was able to get the equivalent of 11 * 512 byte sectors on
a disk, and do so reliably. My main problem was figuring the sector
interleave since I did this with 5 * 1K (like an Osborne), and an extra
512 byte sector - after the 5 * 1K are put on, there's about 640 or so bytes
left over, which is plenty of space to snuggle a 512 byte sector into. One
day I'll get it worked out, because the concept of 440K on a 5.25 DS DD
floppy is too attractive to pass up :-)
--
dg@lakart.UUCP - David Goodenough +---+
IHS | +-+-+
..... !harvard!xait!lakart!dg +-+-+ |
AKA: dg%lakart.uucp@xait.xerox.com +---+