On Fri, 9 May 2003, Nick Garnett wrote:
> Fred asks if we have considered substituting 1.2M 5.25"
> drives. Hmmmm. I didn't know there were such drives.
> I know of DSDD 360K 5.25 drives and 1.44M 3.5 drives,
> even 720K 5.25 drives, but not 1.2M 5.25 drives.
20 years ago, IBM brought forth onto the microcomputer world a "new"
computer, with "advanced technology", called the "IBM AT". I doubt that
they sold more than a few million of them with the 1.2M drive. Counting
other manufacturers, there were probably less of them sold than
MacDonald's hamburgers. The "AT" used a 80286 microprocessor, which could
handle up to 16M of RAM, although for half a dozen years there was little
if any software available to make use of any of its special features.
But the most interesting part of the "AT" was the floppy drive. A number
of years prior to the "AT", several drive manufacturers (Mitsubishi was
the first one that I'm aware of) created a 5.25" disk drive that was
INTENDED to be usable as a replacement for 8" drives. Well before the AT,
Amlyn produced a 5.25" 1.2M drive with a 5 disk changer built in; now THAT
is an unusual drive!
Unlike the previous (360K DSDD) drives which turned at 300 RPM, 1.2M
drives turned at 360RPM (like an 8" drive). And when used with 600
Oerstedt diskettes ("High density"), its electrical interface was almost
indistinguishable from that of a "standard" 8" drive, such as a Shugart
SA8xx, or a Tandon 848.
The power supply for it was 5V and 12V, exactly like a "standard" (SA400)
5.25" drive, as opposed to the distinctly NON-standardized 8" drive power
requirements.
The interface cable for it was 34 pin, like a "standard" (SA400) 5.25",
rather than the 50 pin that was the de-facto standard for 8" drives.
It had 80 tracks per side, at 96 tracks per inch, instead of the 77 tracks
of a "standard" 8" drive.
The format that IBM used was 512 bytes per sector, with 15 sectors per
track. That gave a capacity of 1228800 bytes, or 1.121875 "honest"
megabytes (where a K is 1024 and a M is 1024 K (1048576)), or 1.228
Million bytes, or 1.2 IBM SLEAZY BOGUS "mega"bytes (where a "megabyte" is
creatively redefined to be 1000*1024). But the format was largely a
software issue, and the machine could do any format that the floppy disk
controller was capable of, (see, for example:
http://www.xenosoft.com/fmts.html#8 ) It could mimic almost any 8" MFM
format. With different FDCs, it is capable of any format that an 8" drive
could do with the same controller.
The ORIGINAL intent of the drive, according to a few drive manufacturers,
was that by changing the power connector and interface cable, it could be
used as a drop-in replacement for "standard" 8" drives.
There is even an unsubstantiated rumor that when MICROS~1 was working on
the OS mods for the new machines that would use it (without having seen
the machine yet), that the programmers at MICROS~1 thought that IBM was
planning to release a machine with 8" drives!
But IBM used the 1.2M drive as the primary floppy for those machines, and
even modified their disk controller design to incorporate a 300K bits per
second data transfer rate to permit the "new" drive to be able to read
disks from 360K drives. It could also write such disks, although with a
few complications. [You prob'ly don't want to hear my rants about THAT]
Some other manufacturers built 1.2M drives that had both a 360 RPM AND a
300RPM motor speed capability as an alternate way to incompletely solve
the interchange issues.
IBM made the 1.2M drive their "standard" 5.25" drive from then on, and
even started stamping an asterisk on the front of any further 360K drives
that they made in order to label them as being different.
NOTE: It was intended as a substitute for "STANDARD" 8" drives. In the
words of the late George Morrow: "Standards are a wonderful thing;
everybody should have a unique one of their own." As has been pointed out
by a few people, there existed numerous NON-STANDARD drives. The 1.2M is
NOT very suitable for use in replacement of 8" drives where only one brand
and model of 8" drive could be made to work, such as hard sectored with
PERIMETER holes, drives with non-standard electrical or mechanical specs,
drives where the drive circuitry was implemented elsewhere in the machine,
etc. It is quite easy, in 8", 5.25", and even 3", 3.25", and 3.5" drives
to find examples where NO other drive can replace the one used.
> On the general subject of using a substitute drive. I
> have actually been considering if this could be done. I
> have not suggested doing that because I need to find out
> the feasibility. I would hope we could at least get into
> the 1990s with a 1.44M 3.5 in floppy drive.
Probably, but anybody who can't replace an 8" drive with a 1.2M 5.25",
will NOT be able to replace it with a 1.4M 3.5" (all of the changes needed
for a 5.25" 1.2M would also be needed for a 3.5" 1.4M, plus a number of
other changes). The BEST way to change over from an 8" drive to a 3.5"
1.4M is to FIRST change over to a 5.25" 1.2M, and get THAT working, THEN
make the change to the 3.5" 1.4M.
> These replies will probably be helpful to someone with a
> hardware background (which I'm not). They point out the
> complexities of substituting one type of drive
> for another. This level of complexity is about what I
> expected.
Mostly, with some exceptions, those were documenting the existence of
"NON-STANDARD" 8" drives, for which ANY change, even to a different make
or model of 8" drive would be a challenge.
You need to start by identifying the exact machine, and exactly what drive
it currently uses. If it "doesn't matter" what brand and model of
8" drive, then it also won't matter whether it is an 8" drive or a
5.25" 1.2M.
There has been some discussion here recently about how to identify whether
a 5.25" drive is a 360K v a 1.2M. There have been proposals involving
latch design, color, LED, weight, etc. The best way is to identify it by
model number, and then confirm experimentally. Some of the many model
numbers to look for include: Teac 55G, Shugart/Matsushita 475, and
Mitsubishi 4854. Going rate for used ones is about a dime a dozen.
--
Grumpy Ol' Fred cisin_at_xenosoft.com
Received on Fri May 09 2003 - 14:36:00 BST