We're actually pretty well off, there; we have:
- MS11 Maintenance Manual (DEC-11-HMSAA-D-D)
- MS11 MOS Memory Troubleshooting Guide (DEC-11-HMSTS-A-D)
- MS11-B Engineering Drawings
There's also a little bit about the MS11-C (not covered in the documents
above) in EK-11045-MM-007.
There is a later rev of the MS11 manuals, which does cover the MS11-C; Chuck
McManis had both of them:
EK-MS11A-MM-006 MS11-A,B,C memory systems maintenance manual
EK-MS11A-OP-001 MS11-A,B,C memory systems users's manual
but when I contacted him about them a while back, he wasn't sure if he still
had them, or if he'd given them to Al K to scan and put up (he was on the road
at the time, so couldn't check if he still had them).
Al, are they in your queue somewhere? (No rush to do them, if so; I just want
to make sure we know where they are.)
If the board is a cache, how does it get filled? It
would have to listen
to the UNIBUS the memory is on. ... Note that there has to be a signal
from FastBus ... which tells the CPU if the MS11 has a given address or
not ... so the cache board could use that line to tell the CPU whether
or not the location in question is in the cache.
Studying the MS11 Maint Manual, the MS11 controller has access to the full
address and data from both the CPU (FastBus) and UNIBUS B. (The FastBus
actually has two uni-directional data busses; in and out.) So all that info,
this hypothetical cache board can get from the slot it is plugged into
(assuming the cache is plugged into one of the controller slots), over its
connector pins.
The connectors on the back of the card, and two small boards, must be for
listening to UNIBUS A (in configurations in which the two UNIBI aren't joined
together)? (I'm too lazy to check the slot numbers are see what they actually
are.)
And there is indeed a signal which the MS11 uses to tell the CPU it has the
location the CPU is asking for, so it's theoretically possible to build a cache
card that plugs into a FastBus slot.
Noel