While I cannot help you with the catweasel northstar
project, I do know
that at least the linux cwtool toolkit is missing a number of disk
definitions which are very useful and common.
It does NOT normally support:
SSDD 5.25 8 sectors/track 160k
SSDD 5.25 9 sectors/track 180k
DSDD 5.25 8 sectors/track 320k
SSHD 3.5 18 sectors/track 720k <- very unusual format
SSHD 5.25 15 sectors/track 600k
[snip]
Thanks Jonathan,
Those definitions you mentioned are all soft sector MFM encoded formats, I
believe. I suspect cw2dmk will already decode those since they have the
necessary sync marks, etc. Have you tried cw2dmk? It can extract dmk
images from any FM, MFM, or RX02 encoded soft sector disk.
It is good thing you are doing to add capability to the Linux cw driver for
more formats. Thanks for doing that. The more Catweasel support of any
kind, the better it is for everyone.
Unfortunately, writing software for the Catweasel to make images of
NorthStar hard sector format is an entirely different creature altogether.
The NorthStar controller has its own unique sync sequence which makes the
disks incapable of being read by NEC 765 or derivative FDCs. As a result,
it requires entirely custom software to decode. Thank you to those
individuals who have been of invaluable help to me in getting this far with
the project (you know who you are!).
Here is a bit of technical background on the problem from the NorthStar
floppy disk controller. The NorthStar double density sync sequence is 32
$00's followed by $FB, another $FB, 512 bytes of data, and then a one byte
checksum. The single density sync is similar. The NorthStar timing
sequences is based on the ten sector holes and one index hole. I found the
index hole because it is between two sector holes. The sequence is
something like this:
Gap 1 24 mS
Gap 2 24 mS
Gap 3 12 mS <- index hole passed by
Gap 4 12 mS <- start of sector 0
Gap 5 24 mS
.
.
.
Gap 11 24 mS
My prototype code is currently two parts: the first is a modified utility to
make a raw image, and the second is another utility which extracts the data
from the bit sequence. I can make images with my code
but it is only in the
"proof of concept" stage. It is way too crude for a
general release and
that is why I am asking for some volunteers to help (re)write it and test it
to bring the code up to some reasonable level of decency before any release.
Here is a sample dump from the imaging tool of a disk. It is the NorthStar
DOS 5.0 Double Density boot disk. It is the one which comes with the Dave
Dunfield NorthStar Horizon simulator and tool set. I created an actual boot
floppy on my Horizon and used my Catweasel station to create the image.
http://www.geocities.com/lynchaj/out3.zip
OP: Sending the original request message 3 times was
unnecessary.
I entirely agree and I apologize for sending multiple posts. Believe me, it
was not my plan for three postings. I first posted Thursday last week but
my messages did not show up in the archive until today for some unknown
reason. I read the list by reading the archive of messages and I thought my
messages were going into the bit bucket. I suspect the cctech mailing list
was having some sort of problem. Sorry for the repeats.
Thanks!
Andrew Lynch