16 Jan
2023
16 Jan
'23
12:40 p.m.
On Jan 16, 2023, at 11:47 AM, Chuck Guzis via cctalk
<cctalk(a)classiccmp.org> wrote:
The 844 drives date from the early 70s. ...
Can't tell you more about the mechanics of the things--I haven't seen
one of these in many many years. They were the workhorse drive for CDC
large systems for quite some time. We used them on the STAR-100, for
example.
On the CERL PLATO system at U of Illinois, around 1977, we had 20-ish 844-21 drives, and
maybe a few 844-41 as well. Those were roughly the same as the DEC RP04 and RP05 drives,
same pack and track count. Different sectors, though; 322 12-bit words per sector.
Those are 3600 rpm drives, linear voice coil head actuator, dedicated servo surface. The
details of the format was handled in a sort of microcoded bit handling engine, one of two
engines in the programmable controller (7054). I actually have the source code still
around, and the manual for that beast also still exists.
My favorite for wild mechanics is the IBM 1311 we had on an IBM 1629 Mod II. Those have
hydraulic actuators, 100 cylinders. In the controller, the subtractor that would tell the
machinery how many cylinders there are between the current one and the requested one was
an optional feature ("direct seek option") which we didn't have. Without
the option, all seeks would be done by retracting all the way to cylinder zero, then back
out again the number of cylinders corresponding to what the program asked for. You could
really make the drive shake by feeding it a simple program that stepped through the
cylinders; near the end of the loop it would spend a significant fraction of a second
getting from, say, track 98 to 99.
paul