In July 2014 I mentioned that I did some reverse-engineering of a Quay
900 computer system, based on the Quay 90F/MPS single-board Z80
computer and two CDC/MPI 9406 double-sided 8-inch floppy drives. Back
then I wrote a simple monitor ROM and a floppy disk formatter program,
but at the time I didn't have any double-sided 8-inch media on hand,
so the formatter was written for single-sided use only.
I just updated it for double-sided use and put the updated source code at:
http://www.brouhaha.com/~eric/retrocomputing/quay/quay900/floppy_formatter/
I still haven't tracked down any documentation on the -022 variant
(full P/N 77618022) of the 9406 drive. The service manual on Bitsavers
lists that variant in the table documenting mechanical hardware of
variants, but it does not list the PWA (circuit board) p/n 77615454 at
all. This variant seems to have a Shugart-compatible pinout, perhaps
what is known as a "9406-3", while all of the variants in the service
manual do not have a Shugart-compatible pinout. I don't have the time
or energy to reverse-engineeer the PWA.
A mystery has arisen in that I tried to autodetect double-sided media,
since most Shugart-compatible 8-inch double-sided drives will assert
the TWOSIDED/ signal on pin 14 if they detect a two-sided medium,
based on which index sensor is active. The uPD765/8272 will provide an
indication of that after seek-class commands in ST3, but I found that
the bit is always reading as zero regardless of whether I use a
single- or double-sided disk.
Inspecting the signal on pin 14 reveals very unexpected behavior. Pin
14 is always either unasserted, or toggling at about 976 Hz. It does
seem correlated with whether I have a double-side disk inserted,
except that it does that regardless of whether a drive is selected. If
I put a single-sided disk in one drive, and a double-sided in the
other, pin 14 toggles all the time. This really makes me wonder just
what this particular 9406 variant is trying to do with pin 14.