Chuck Guzis wrote:
On 11 May 2007 at 19:59, jim wrote:
MSC if I recall correctly morphed into Xebec, who
was one of the first
if not the first supplieres ot controllers to IBM for the XT.
<snip>
Sometime between this unit and the XT controller, I think they went
to programmable bit-slice components for their controllers (8X300?
2901?). This one's interesting in that, with the exception of the
8291, it's all discrete logic; no big ROMs or LSI.
The controller that Microdata had was modified from the controller design
that ran Microdata branded drives directly, but was also all 7400 logic with
a 74181 alu type cpu sequencer. I have schematics for both versions of this
that hopefully will get scanned and posted on bitsavers sometime soon.
Microdata's cpus were both ALU type cpus while I was there in the late
70's and the controllers all had 1 or 2 in their sequencers as well, in
their
disk controllers, and in their terminals. Tape and com controllers were
more state machine type devices, and the printer, card reader, and
paper tape read punch boards were just parallel access interfaces
with the cpu doing the work.
Was the back end of this controller anything resembling sasi, sequential
or parallel register interface? I think the MSC controllers into the
microdata minis were all serial interfaced on the back end, since
the controller cabling and logic had RF serial / deserializer logic
that had to remain, since they talked to both microdata drives which
were serial RF as well as the MSC controllers w/o any extra
connectors to go around the serial logic.
you moved a jumper or two, but other than that, the MSC controller
just made more heads and cylinders show up to the software, and
the seeks were also passed as pulse trains, not as parallel
info (like SMD).
Xebec picked up and modifed the SASI interface and blessed us
with a fouled up early SASI command set which survives at the
heart of IDE of course. I'm not aware of a "Xebec" or "MSC"
design that survived, but just as you say a lot of custom adaptations.
Jim