[Coco] Need help with 3.5" drives and NitrOS-9

[Kip Koon]

Hi Robert!
That is an interesting symptom.  I've read many of the emails about the
formatting situation in NitrOS-9, but to be honest, I've never noticed any
strange behavior that may be caused by the type field in the drive
descriptor.  But then I was not looking for this symptom anyway.  I will do
some checks on my system using all my floppy controllers, but once I
filtered through all my old 3.5" drives and got several to run NitrOS-9
successfully, I have had few if any problems as long as I don't keep
changing things around as I tend to do sometimes when experimenting.  I've
done quite a few formats and backups on my 3.5" drives getting NitrOS-9
running on my Cocos using the 720KB 2880 sector OS9 floppy disk format and
I've had no problems using good quality floppies that I remember.  
Incidentally, since you are using MITSUMI model D35M3D floppy drives
(shouldn't that read D359M3D), are you twisting conductors 10-12 to get two
drives running as Drive 0 & 1?  If so, have you done anything to get a third
Mitsumi Model D359M3D drive to run on the same cable?  Could a second twist
of 10-16 accomplish this?  When I was experimenting with PC floppy drive
cables, I noticed that the drive after the twist was drive 2, not 0.  I
thought that was interesting.  Since I have a plethora of Mitsumi drives
myself so far unused, I was also playing around with the idea to create a
little PCB with drive select jumpers for all the 0-3 possible choices.
Well, 0,2 & 3 anyway.  It would be pointless to have a DS1 selection,
wouldn't it.  That way, any floppy drive with no jumpers could be made to
respond as any drive the user wishes.  What do you think?  A double twisty
cable, or PCBs?  The double twisty cable could actually work I think.  Can I
digress.  
Anyway, I will check and see if I have any 3.5" 720KB NitrOS-9 2880 sector
disks getting corrupted on my coco 3 and 1 systems, but at this point I
would say, I think not.  You are truly a wealth of Information!  Take care
my friend.
Kip

-----Original Message-----
From: coco-bounces at maltedmedia.com [mailto:coco-bounces at maltedmedia.com] On
Behalf Of Robert Gault
Sent: Tuesday, December 03, 2013 6:56 PM
To: CoCoList for Color Computer Enthusiasts
Subject: [Coco] Need help with 3.5" drives and NitrOS-9

If anyone has a 3.5" drive attached to a Coco system please respond to this
message.

Currently NitrOS-9 sets all descriptors for 80-track disks as 3.5" types
with typ=$21. When such a disk is formatted, it is stated to be set to
135tpi.

Now if I try to read such a disk on my Coco3  with either a Tandy controller
or a Disto SCII it will read correctly. However all writes to such a disk
corrupt each byte usually forcing the msb high.
If I use dmode to set typ=$20 for such a disk, format makes it a 48tpi disk
and there are still 80 tracks. This type disk both reads and writes
correctly on my system.

Clearly something is wrong. It does not seem to be my controllers and I
don't see how it can be the 3.5" drive. The first question is does anyone
else with 3.5" drives see a similar behavior. The second question is how
should this be fixed?
Is a Coco system just too slow to write 135tpi disks? Is perhaps the 3.5"
drive on my Coco (MITSUMI model D35M3D) not capable of 135tpi writes? But
then how can it read such a disk?

Apparently 3.5" disks can be formatted at 48tpi for 80tracks "double sided"
and perhaps they should be for Coco systems. If that is the case, there
should be some changes to one or all of rbf.d, rb1773desc.asm, and
rb1773.asm in NitrOS-9.

Robert

--
Coco mailing list
Coco at maltedmedia.com
http://five.pairlist.net/mailman/listinfo/coco