Intel 432 floppy flux images for decoding

Eric Smith spacewar at gmail.com
Tue Jul 26 20:18:46 CDT 2016


On Tue, Jul 26, 2016 at 8:53 AM, Dennis Boone <drb at msu.edu> wrote:
> A collection of intel floppies containing software related to the 432
> processor was identified and purchased by Nigel Williams.  I made flux
> images of the diskettes some time ago, but have failed to assemble the
> attention span to decode most of them.

Thanks *very* much for doing this!

> These are 8" diskettes, written in several different formats.
> There are FM and M2FM, and possibly MFM, floppies.

M2FM would be Intel MDS-720/SBC-202 double-density format, which is
unique to Intel. Although the SBC-202 was sold for general use, so it
*could* appear in other Multibus systems, I've never actually heard of
it being used in anything other than an Intel MDS 800, Series II, or
Series III development system. These used ISIS-II. I've written some
Python code to extract files from ISIS-II disk images (raw sector
format, not flux transition)

MFM was never used on the MDS 800, Series II, or Series III. Any MFM
disks are probably for an 8086-based system (e.g., Intel System
86/330) running iRMX-86 and being used as an I/O processor for the
432.

Standard IBM 3740 format FM disks could be either ISIS II on an MDS,
or iRMX-86 for an I/O processor.

Three flux image files were created from each
> diskette, at three sampling rates: 25, 50 and 100 MHz.  (This is
> certainly overkill, but I wasn't sure what would end up working best.)
> Both sides were imaged, and the reader software was told to take two
> passes over each track.  The FM diskettes are SSSD, with 128 byte
> sectors.  (The FM ones are already mostly decoded.)

I applaud your degree of concern for ensuring that there will be
usable data. That said, I can assure you that you are correct about
overkill, and that 25 MHz is more than enough to accurately image any
normal floppy disk, up to high-density 3.5-inch (so-called "1.44M").

I am not able to resolve your server's IP address. Al was kind enough
to package everything up for me in a tarball, which I'm downloading
now. I'll look over it soon.

Last year I started hacking up code to read M2FM from flux
transitions, but I didn't finish it. Now I've got some motivation to
hack on it some more!      :-)

For those not familiar with the iAPX 432, Intel's first 32-bit
microprocessor from 1981, the software for it was:

* OPL-432 and Object Builder: a dialect of Smalltalk that runs only on
the 432/100 eval board (and that's the ONLY software that runs on the
432/100, everything else needs a 432/6xx system)
* CDS-432 Cross-Development System: Ada compiler, linker, and related
utitlities that run on a VAX, either under VMS or BSD 4.x
* iMAX-432 operating system, runs on 432/6xx, supplied as object
modules and Ada package specifications for use with CDS-432, and I/O
processor code for 8086
* DEBUG-432: debugger, runs on Series III MDS under ISIS II, loads and
executes code image
* 432/6xx diagnostics, runs on Series III MDS under ISIS II


More information about the cctalk mailing list