[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Why can't Gs 800k drives read 720k IBM disks?
In article <3562@swuts.sbc.com>, <JB3819@JB3819.SBC.COM> wrote:
>
> The electrical encoding scheme used to write bits onto the media is
> enough different that heads set up for GCC can NOT read MFM and vice
> versa.
Rubbish. The heads don't care one whit what type of data is passing
underneath - they just sense changes in the magnetic field.
Even the drive doesn't usually care what data format is used. The
main restriction on accessing GCR (not GCC) disks with MFM drives and
vice versa is that the disk controller cannot handle the other format.
Some drives have limitations that may impose further restrictions.
For example, the Disk ][ is a single sided drive and only supports 35
tracks (some go to 38 or 39, but very few can access 40), so it would
be impossible to read an entire IBM 360k disk in a Disk ][, even if
the controller could handle MFM.
Apart from the number of tracks and single/double sided issues, the
head's recording and reading signal strengths might be wrong (e.g.
when dealing with high density disks), or the drive rotation speed
might be wrong (e.g. when dealing with double density 3.5" disks).
> Whether it is 720K or 800K is a different question and your
> (a) and (b) may apply. So, if you use a controller card that'll
> handle an IBM drive (MFM) you can, within physical limits of the drive,
> write whatever you want on it, including 800K ProDOS formatted data.
That is just a file system issue - you couldn't do 800k, because the
controller is not cable of writing 800k to an MFM disk with 80 tracks,
two sides and double density - there isn't enough space on the tracks.
You could do an MFM 720k ProDOS disk, given the right software.
> However, because of the encoding differences between GCC and MFM, an
> Apple drive (GCC) can't even read the BITS on the diskette.
Flying summary of GCR:
Bit cell width is 4 microseconds on a 5.25" disk, 2 microseconds on a
3.5" disk. Sector headers are encoded using "4-and-4", where every
second bit is a 1 - each data byte requires two disk bytes to be
written. Data is written using a software encoding scheme which
results in 1.33 disk bytes for each data byte (6-and-2 encoding). The
disk bytes that can be written match the following rules:
- Each byte has bit 7 set (a requirement of the controller).
- There can be no more than one pair of adjacent zero bits, and must not
be more than two adjacent zero bits.
- There must be at least one pair of adjacent one bits (software
requirement).
Note: in both GCR and MFM, a "1" bit is represented on the disk by a
flux reversal.
Flying summary of FM (the basis of MFM):
Bit cell width is 4 microseconds on 5.25" disks. Data bits are
separated by clock bits (1) - there is a potential bit (clock or data)
every 4 microseconds. Each data byte requires 2 disk bytes (this is
the equivalent of 4-and-4 encoding in GCR).
Flying summary of MFM:
Bit cell width is 4 microseconds on 5.25" disks, but bits may be
positioned with 2 microsecond resolution. A simple encoding scheme is
used to skip clock bits where they aren't needed. A clock bit is only
inserted when writing two consecutive zero bits, and is out of phase
with the data bits.
data cDcDcDcDc
1001 010010010
1011 010001010
1101 010100010
1111 010101010
Each digit represents a 2 microsecond half-bit cell (alternating clock
and data positions). As you can see, the 1 bits are never closer than 2
cells (4 microseconds). The only clock bit is between the two zero
data bits. The longest gap on the disk is slightly under four
microseconds. MFM has twice the storage capacity of FM, with the same
disk media.
On a double density 3.5" disk, the same system is used (there are
twice as many tracks, hence 720k instead of 360k). A high density
3.5" disk has bit cells that are half as wide (1 microsecond
resolution, 2 microsecond minimum space between bits).
[I think I have one of the bit timing details wrong - some of the
numbers don't add up. If I'm right about all of the above, then an
GCR 3.5" disk should be able to fit nearly twice as much as an MFM
3.5" disk, assuming the same rotation speed.]
--
David Empson
dempson@actrix.gen.nz
Snail mail: P.O. Box 27-103, Wellington, New Zealand