Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Reading Dynabyte DSDD 8" floppies

67 views
Skip to first unread message

Ian Justman

unread,
Oct 25, 2012, 3:12:21 PM10/25/12
to
After a very long absence, I'm back.

Anyway, I used to use a Dynabyte DB8/1 many years ago. I no longer have the hardware, but I have several disks I need to get data from. Apparently, in my fits of googling, I've found that Dynabyte did something a little bit different on its DD format. It apparently isn't truly MFM-compliant, or using gaps that a uPD765 can't easily handle. I.e. I can't use Dave Dunfield's ImageDisk to read them. By extension, I somehow doubt my old CompuPro Disk 1A, which I do still have one of in an old Viasyn chassis, will read it either, since it has its Intel counterpart, the i8272. Does anyone know what can read these disks save another Dynabyte? The only other computer I have in my inventory that may come close is an old Ferguson Big Board II (this one uses a WD1797 with an SMC 9216 datasep), but I suspect its data separator, FDC, or both have gone south since it can't even read 3740-formatted disks which otherwise read fine on a PC. If worst comes to worse, does anyone still have a serviceable Dynabyte with DSDD-capable hardware? If so, if willing, I'd love to ship off those disks to have them read and put into a somewhat more usable format, like an enclosed USB thumb drive.

Thanks in advance!

--Ian.

glen herrmannsfeldt

unread,
Oct 25, 2012, 6:35:49 PM10/25/12
to
Ian Justman <ia...@ian-justman.com> wrote:
> After a very long absence, I'm back.
>
> Anyway, I used to use a Dynabyte DB8/1 many years ago.
> I no longer have the hardware, but I have several disks I need to
> get data from. Apparently, in my fits of googling, I've found
> that Dynabyte did something a little bit different on its DD format.
> It apparently isn't truly MFM-compliant, or using gaps that
> a uPD765 can't easily handle.

Gaps shouldn't be much of a problem. You do have to be careful
with gaps on writing, and if the gap is too small on reading
then it will take a whole revolution before reading the
next block. (That is what interleaving is for.)

But if the bit format is different then you need the
right hardware.

Or, something like a catweasel that just reads the magnetic
transitions, and then process them separately.

-- glen

glen herrmannsfeldt

unread,
Oct 25, 2012, 7:20:26 PM10/25/12
to
Ian Justman <ia...@ian-justman.com> wrote:

> Anyway, I used to use a Dynabyte DB8/1 many years ago.
> I no longer have the hardware, but I have several disks
> I need to get data from.

A quick search shows that it should be the WD1781.

It seems, though, that some external logic is needed to
implement DD, and the exact bit format depends on that
logic.

The 1771 was FM only, the later 179x series IBM standard MFM.

-- glen

Ian Justman

unread,
Oct 25, 2012, 11:45:56 PM10/25/12
to
On Thursday, October 25, 2012 4:20:26 PM UTC-7, glen herrmannsfeldt wrote:
I remember reading about the WD1771 back in the day. And I remember examining the boards when I originally got two systems (I have neither anymore that I know of...). Further examination of the newsgroup postings in the past say that the format is apparently M2FM. This machine's design is apparently old enough to be in the middle of the debate of whether to go MFM or M2FM, and apparently, the WD1781 could go either way depending on that external logic you described. I do still have schematics for that board (as well as a few others in the system), but it doesn't say what kind of part it uses for the data separator. It just gives a grid location for the device.

About the only other thing I'm thinking I might have to do is figure out how to scare up the money to get a CatWeasel and go that route. Apparently, it's one of the few devices for PCs out there that can do M2FM and other oddball formats. I just can't justify $100+ for a device I'm only going to use a few times for just a handful of disks.

--Ian.

Ian Justman

unread,
Oct 25, 2012, 11:55:33 PM10/25/12
to
On Thursday, October 25, 2012 8:45:56 PM UTC-7, I wrote:

> I do still have schematics for that board (as well as a few others in the system), but it doesn't say what kind of part it uses for the data separator. It just gives a grid location for the device.

I did find an amazingly-detailed photograph of the board. It looks like it used a LOT of discrete logic to implement mod/demod/data separation before it was shrunken down to eight-pin DIP packages.

--Ian.

glen herrmannsfeldt

unread,
Oct 26, 2012, 1:47:50 AM10/26/12
to
Ian Justman <ia...@ian-justman.com> wrote:

(snip)

> About the only other thing I'm thinking I might have to do
> is figure out how to scare up the money to get a CatWeasel
> and go that route. Apparently, it's one of the few
> devices for PCs out there that can do M2FM and other
> oddball formats.

No, it just measures the time between flux transitions at
some high enough resolution and writes that to the PC disk.

No decoding at all, just the raw bits.

> I just can't justify $100+ for a device I'm only going
> to use a few times for just a handful of disks.

I have an 8 inch drive, but don't have the right cables
to connect it up. Someday...

Otherwise, I believe the M2FM or MFM is independent
of the data separator. The output of the data separator
is data and clock bits, which the rest of the logic
has to decode.

Well, FM is pretty simple, there are always clock bits
except certain bits in address marks, so it knows where
the sector header starts. MFM leaves off the clock bit
except between two zeros, and has different special codes
for address marks. M2FM has a different pattern of
missing clock bits.

-- glen

mon...@vitasoft.org

unread,
Oct 26, 2012, 1:16:06 PM10/26/12
to
Ian, FWIW, I have not yet found a disk format the WD 2793 FDC chip could not read. You would do a complete track read first, examine the "GAP" format to construct a CPM3 parameter table and go from there (quite a bit of work). The ZFDC S100 board could be used to try this. See:-

http://s100computers.com/My%20System%20Pages/ZFDC%20Board/ZFDC.htm

Unfortunately I could not devote the time to this but a number of the boards are out there by now. Perhaps somebody else could try it for you.

Alternatively you could try the earlier WD 17xx chips. There are quite a few in S100 systems (The SD Systems Versafloppy II for example). In general the WD chips have a wider tolerance for unusual formats.
John

Steven Hirsch

unread,
Oct 26, 2012, 6:03:02 PM10/26/12
to
On 10/26/2012 01:16 PM, mon...@vitasoft.org wrote:

> Ian, FWIW, I have not yet found a disk format the WD 2793 FDC chip could
> not read.

John, are you sure that chip can read M2FM? It's a very different animal. I
recently needed to read M2FM diskettes from an old Intel development system
and the only thing able to do the job was my Catweasal controller.

mon...@vitasoft.org

unread,
Oct 27, 2012, 1:07:59 PM10/27/12
to
Could be right Steven. Does anybody know what the controller is in those Dynabyte systems. I have no info on their S100 FDC board. If anybody has one/manual I will post it on the S100Computers "Boards" page
0 new messages