15 Jul
2009
15 Jul
'09
8:36 p.m.
-----------Original Message:
Date: Wed, 15 Jul 2009 10:02:24 -0500
From: Chris Elmquist <chrise at pobox.com>
Subject: Micropolis 1015-2 floppy drive
Does anyone have specs or even better original documents for a Micropolis
1015-2 (aka mod II) 5.25" floppy drive? I have two connected to
a Micropolis designed floppy controller in an old S-100 system.
My memory tells me these were single sided, double density, 77 track,
100 tpi drives. But I may be confusing them with other Micropolis models
that were used in this particular S-100 system around the same time.
The S-100 system was built by Multi-Tech Systems Inc (the modem guys)
here in St. Paul, MN around 1980,81. It is a model MT-500 and is a
clone of an R2E design that came to market around the same time.
I wrote portions of the PROM and CP/M BIOS for this thing when I was
in highschool. I, of course, saved many pieces of the things and am
preparing to bring one of them back online soon.
Chris
--
Chris Elmquist
---------Reply:
Your memory's pretty good; they are indeed SS/DD 100TPI; there's
also a 1016-2 which looks like pretty well the same drive.
Vector Graphic used these drives with a Micropolis controller; I have
two VG systems that use them (with hard sector disks, FWIW).
The bare mechs were also used in Commodore 8050 dual disk drives
(as well as Tandon and MPI) but with custom electronics.
You can find a maintenance manual (26MB) at:
www.vector-archive.org
in the VG Docs CD folder.
The Micropolis controller manual is in the Manuals folder.
Good luck!
mike
****************************************
15 Jul
15 Jul
9:52 p.m.
On Wednesday (07/15/2009 at 11:36PM -0400), M H Stein wrote:
Your memory's pretty good; they are indeed SS/DD 100TPI; there's
also a 1016-2 which looks like pretty well the same drive.
Vector Graphic used these drives with a Micropolis controller; I have
two VG systems that use them (with hard sector disks, FWIW).
The bare mechs were also used in Commodore 8050 dual disk drives
(as well as Tandon and MPI) but with custom electronics.
You can find a maintenance manual (26MB) at:
www.vector-archive.org
in the VG Docs CD folder.
The Micropolis controller manual is in the Manuals folder.
Good luck!
Very cool! Thank you. I've grabbed some very useful stuff. Much
appreciated.
I've had some success building a 10-hole hard sector emulator using 3.5"
drives and so the 16-hole Micropolis stuff is the next mission to emulate
on 3.5"-- but in the meantime, I'll use the 1015-2 assuming they are
still workable and then also have a line on some Tandon 100-4M which
are also 100 tpi drives.
Thanks again!
Chris
--
Chris Elmquist
10:21 p.m.
On 15 Jul 2009 at 23:36, M H Stein wrote:
-----------Original Message:
Date: Wed, 15 Jul 2009 10:02:24 -0500
From: Chris Elmquist <chrise at pobox.com>
Subject: Micropolis 1015-2 floppy drive
Does anyone have specs or even better original documents for a
Micropolis 1015-2 (aka mod II) 5.25" floppy drive? I have two
connected to a Micropolis designed floppy controller in an old S-100
system.
I'll check--we used them and their blue-handled predecessors at
Durango between ca. 1977-79.
I've still got a 1015-II and a 1016-II sitting on the shelf--heavy
buggers, with a very slow track-to-track seek. But the way they're
built, they'll still be here when the cockroaches rise to replace
man. They're basically the same drive, with some very minor
mechanical modifications for the drive-insertion microswitch--the PCB
is the same.
The older 1014? 1012? drives didn't have the "spin the motor during
insertion" circuity and they mangled diskette center hubs
mercilessly. I didn't see any point in keeping them around so I
dumped them some years back.
I may have some docs on them, but it'll take a lot of digging. But
yeah, single-sided, 100 tpi. ISTR that track-to-track is at least 15
msec. One notable thing is that they have the Micropolis pinout--
e.g. READY on pin 6. I think the 1015's added an edge connector
where the earlier ones had only a header for the signal cable
connection.
ISTR VG used them in a hard-sector application, but we used them as
soft-sector GCR (we used a WD1781 as the basis of the controller--
anyone have any of those?) at a somewhat elevated data rate (the
drive PCB was slightly tweaked). 12 sectors of 512 bytes per track.
I wrote all of the floppy driver code, so I still have a pretty good
memory of them.
Eventually we replaced them with Tandon TM-100-4M drives; Micropolis
was having a miserable time getting the double-sided version of the
1015/1016 to work and there was the seek-time problem, as well as
their non-competitive pricing and we ran out of patience. We put our
own PCBs and tach circuits on the Tandons--the standard ones didn't
cut the mustard. We also briefly used some 100 tpi drives from
CDC/MPI that were pretty awful mechanically.
I still have the sample 1115 drive that Micropolis dropped off.
Still the slow leadscrew positioner, but with buffered seek and a MOS
technology processor on board. These were DSDD and could be had in
96 or 100 tpi versions and had a strange mechanical arrangement--
reminds me of an old Fokker airplane engine--when you open the latch
on the drive, the positioner assembly, stepper motor, leadscrew and
all, swings out of the way. Cast body, not steel plate, like the
older drives. A curious drive, but still seeks slowly, even if it is
buffered. It was probably a good idea for Micropolis to get out of
the floppy business after that.
Wonder why I didn't receive Chris' original message? cctech is
supposed to be posted to cctalk, is it not?
--Chuck
16 Jul
16 Jul
5:01 a.m.
On Wednesday (07/15/2009 at 10:21PM -0700), Chuck Guzis wrote:
On 15 Jul 2009 at 23:36, M H Stein wrote:
Mike Stein pointed me at some VG documentation online which solved
the near term problem... so I am set for now. Thanks.
-----------Original Message:
Date: Wed, 15 Jul 2009 10:02:24 -0500
From: Chris Elmquist <chrise at pobox.com>
Subject: Micropolis 1015-2 floppy drive
Does anyone have specs or even better original documents for a
Micropolis 1015-2 (aka mod II) 5.25" floppy drive? I have two
connected to a Micropolis designed floppy controller in an old S-100
system.
I'll check--we used them and their blue-handled predecessors at
Durango between ca. 1977-79. ISTR VG used them in a hard-sector application, but we
used them as
soft-sector GCR (we used a WD1781 as the basis of the controller--
anyone have any of those?) at a somewhat elevated data rate (the
drive PCB was slightly tweaked). 12 sectors of 512 bytes per track.
I wrote all of the floppy driver code, so I still have a pretty good
memory of them.
Yes... the Micropolis controller design, which VG and Multi-Tech clearly
used as a reference design, is a 16-hole hard sector thing.
Multi-Tech also had a GCR controller but I don't have one of those. It
was a board chock full of parts including a Z80 I think. It was a PITA
to align and very expensive so I opt'd not to obtain one as a parting
gift from Multi-Tech.
Eventually we replaced them with Tandon TM-100-4M
drives; Micropolis
was having a miserable time getting the double-sided version of the
1015/1016 to work and there was the seek-time problem, as well as
their non-competitive pricing and we ran out of patience. We put our
own PCBs and tach circuits on the Tandons--the standard ones didn't
cut the mustard. We also briefly used some 100 tpi drives from
CDC/MPI that were pretty awful mechanically.
Yup. We played with all of those too. The frusteration due to all the
different track densities that were in play at the time is starting to
come back to me. I know it caused endless confusion for customers who
had these floppies that all looked the same-- same exact part numbers on
them-- but couldn't be exchanged between machines because some had 48tpi
drives, some 96 and some 100. All because sales people always wanted the
customer to buy the latest and largest storage but never considered any
backward compatibility with the machines they had bought the week before.
Too much fun.
I still have the sample 1115 drive that Micropolis
dropped off.
Still the slow leadscrew positioner, but with buffered seek and a MOS
technology processor on board. These were DSDD and could be had in
96 or 100 tpi versions and had a strange mechanical arrangement--
reminds me of an old Fokker airplane engine--when you open the latch
on the drive, the positioner assembly, stepper motor, leadscrew and
all, swings out of the way. Cast body, not steel plate, like the
older drives. A curious drive, but still seeks slowly, even if it is
buffered. It was probably a good idea for Micropolis to get out of
the floppy business after that.
Wonder why I didn't receive Chris' original message? cctech is
supposed to be posted to cctalk, is it not?
I sure have a lot of trouble posting to these lists. What I see here is
that traffic coming from 'cctalk' has a Reply-To: field that sends it to
'cctech'. If I forget to manually edit that out, then my replies will
go to a different list than they originated from. I do know that my
original on this thread went to 'cctalk' and not 'cctech'. Should I
have sent to both? I assumed that was frowned upon.
Chris
--
Chris Elmquist
17 Jul
17 Jul
11:34 a.m.
On 16 Jul 2009 at 7:01, Chris Elmquist wrote:
Yes... the Micropolis controller design, which VG and
Multi-Tech
clearly used as a reference design, is a 16-hole hard sector thing.
Multi-Tech also had a GCR controller but I don't
have one of those.
It was a board chock full of parts including a Z80 I think. It was a
PITA to align and very expensive so I opt'd not to obtain one as a
parting gift from Multi-Tech.
I think I have a couple of Multi-Tech diskette samples in my files.
The GCR design was also used by Future Data in 8" format for their
CAD workstations.
Does anyone have any documentation for the Micropolis demo system? I
can recall seeing it back then, but I ddn't pay much attention to it
and didn't save any of the paper.
In any case, ours was complicated--the controller for our system
took up an entire Multibus-sized board. In retrospect, that was
probably thought not to be too bad, when viewed in the light of the
floppy controller that was being supplied with the Intel MDS-800--two
boards with Intel 3000-series bit-slice logic on one.
The design for ours was done by an engineer we hijacked from Sperry
ISS who did work on big-iron drives The WD1781 was buggy (had a
tendency to hang) and WD got tired of making it. On the next system
design, we used 96 tpi drives and MFM. Customers were disappointed
in getting 800K (10x512 tracks) instead of 980K (12x512) on their
floppies, but it was much cheaper, required no adjustment and could
be tucked into a corner of a board. By then, they were using hard
disks anyway.
We found that Micropolis was very sloppy when it came to head azimuth
on drives from the factory and it took some table-pounding for them
to get their act in shape.
--Chuck
1:56 p.m.
Multi-Tech
also had a GCR controller but I don't have one of those.
It was a board chock full of parts including a Z80 I think. It was a
PITA to align and very expensive so I opt'd not to obtain one as a
parting gift from Multi-Tech.
I think I have a couple of Multi-Tech diskette samples in my files.
The GCR design was also used by Future Data in 8" format for their
CAD workstations.
3:48 p.m.
On 17 Jul 2009 at 21:56, Tony Duell wrote:
I have a schematics book for it (about the only manual
I do have), but
I think it only covers the 'main' unit. I don't think I have any info
on the disk controller board in the disk drive unit (and from what I
remember when I looked inside, it's not simple, there was certainly a
processor (Z80?) in there).
I can't help you much. About all I recall is that I received a
package of 8" HS diskettes from a defense contractor technical guy
who said he was told that they contained "source code" for some
project he was very interested in. He didn't know what kind of
source code, nor the system involved in producing it--he had only the
project contract information--and he wasn't at liberty to disclose
that.
Fortunately, GCR has a very distinctive pattern when flux changes are
plotted on a histogram. It helped that the disk was hard-sectored,
as I could pretty much depend on some sort of sector ID at the
beginning of each sector, probably also containing a track number.
As this turned out to be a simple 5-to-4 group code, after I'd
figured out what the sector headers were, I had most of the puzzle
and was getting some ASCII data out of the raw bit-crunching.
One of the first strings was "FUTURE DATA" and that rang a distant
bell.
About all I have to add is that the customer was fairly happy with
what he got--however, one disk went missing over the years and the
source code with it, so the project wasn't a complete success. I did
retrieve some Intel-format HEX binaries, so he could at least fill in
the gaps with some reverse assembly and go on to develop whatever new
weapon of mass destruction he was bent on creating.
--Chuck
5637
days inactive
5638
days old
6 comments
4 participants
participants (4)
-
ard＠p850ug1.demon.co.uk -
cclist＠sydex.com -
chrise＠pobox.com -
dm561＠torfree.net