Re: What is the most used ROM in the brown 1541?

Hello,

* On Thu, Jun 18, 2015 at 08:11:06PM +0200 Ruud@Baltissen.org wrote:
 
> I'm still busy with commenting the sources of the various ROMs and I 
> noticed that in version 901229-03 the number of bytes in the header 
> gap was increased from 8 to 9.
[...]
> This meant that 9 is also the value 
> used in 901229-05, -06AA and the 1541-II. And this is the value that 
> was already present in the CBM 4040. 

No, it is not. ;)

For brevity reasons, I will use the term 1540 for the 2031, 1540 and the
1541 with -01 and -02 ROM, and I will use the term 1541 for the 1541
with -03 ROM and later.


The 4040 used 8 bytes. Anyway, it used 8 bytes before conversion to GCR,
which was done in hardware. Thus, the gap was 8 * 10 bit = 80 bit long.

The 1540 just reused the value 8. Unfortunately, here, it is 8 *real*
bytes, resulting in a gap of 8 * 8 bit = 64 bit.

This could result in problems: After the last byte of the header, the
drives start writing a SYNC mark after this gap. On the 1540, the gap is
64 bit, on the 4040, it is 80 bit.

Now, if the 4040 writes on a 1540 formatted disk, it is already in the
SYNC when it starts writing the SYNC. It has to change from READ to
WRITE, which could result in "unclear" bits that could be read as "0".
If this happens, the reading drive has already read 80 - 64 = 16 bit of "1"
(resulting in SYNC), then a 0 bit - thus, for the reading drive, the
SYNC has just ended. Unfortunately, no data follows. Instead, there is a
SYNC. Thus, the drive has a read error.

The other way around, it is not the problem. If there is the 80 bit GAP
of the 4040, and the 1540 writes on this disk, it just starts writing
its SYNC too early - no big deal.



Why did the change from 8 to 9 bytes in the 1541 change the rules?
Because now, the 1541 GAP is 72 bit long. Thus, the 1541 only writes
80 - 72 = 8 bit too late on a 4040 disk. As 8 bit do NOT form a SYNC yet
on reading, there is no risk for an erroneous SYNC mark.


Why didn't Commodore use 10 byte, which would result in a GAP of 80 bit,
like in the 4040? Most probably, it was for compatibility reasons with
the 1540. As disks with 8 byte GAP already existed, the same problem
would result if the 1541 writes on 1540 disk. With the 9 byte GAP, the
problem does not exist.



Regards,
Spiro.

-- 
Spiro R. Trikaliotis
http://www.trikaliotis.net/

       Message was sent through the cbm-hackers mailing list