More 11/44 basic newbie help request

From: Tony Duell <ard_at_p850ug1.demon.co.uk>
Date: Tue Feb 11 18:31:00 2003

> Eric wrote...
> > Get the manuals from Al's site.
>
> You know... I think I've just been told "RTFM" *grin*
> S'ok, was well deserved :)
>
> After perusing the manuals for the 11/44 and the RL02's, and the RL11
> (looking for docs on the 8 port mux now).... Most of my questions are
> answered there (big suprise). Not trying to start any holy wars or anything,
> but to my "newbie with unibus" mind, the design of having to change
> backplane wiring for certain cards seems to be rather... ummmm silly
> (euphamism{tm}), at least as compared to other systems of the period.

The Unibus got one thing very right. It allowed multiple devices to share
BR (interrupt) and the NPR (DMA) lines. Anyone who's battled with the ISA
bus will certainly appreciate this.

The various devices on each line are assigned priorities by how 'far'
they are from the CPU. The device nearest to the CPU (strictly the
arbiter) is the highest, etc.

The way this works is very simple. A device asserts the appropriate
request line (this line is bussed to all devices). The arbiter asserts
the appropriate grant line, which is not bussed. Instead it's connected
to the first device. The first device then passes the grant on to the
second, and so on. Devices actually asserting the request line (there
could be more than one) don't pass on the grant, so that devices further
away from the CPU than the first device trying to cause an interrupt
never see the grant.


This works very well, but it has one problem. You can't have empty slots
in the backplane. Instead you have to insert grant continuity cards to
link the grant-in to grant-out lines on unused slots, so as to pass on
the grant.

There are 5 grants on the Unibus, 4 for interrupts (BG) and one for DMA
(NPG). Originally, there were no single-card DMA devices -- DMA devices
were complete backplanes that you connected into the unibus chain.
Therefore the original grant cards only handled the BG lines.

Later on, single-card DMA devices became possible, and the NPG line was
connected to pins CA1 and CB1 on the normal SPC (Small Peripheral
Controller) slot. Hwever, there's a problem in that the older grant cards
didn't connect these 2 pins together. Worse than that, nor did any of the
older SPC I/O cards.

So there had to be another way of linking them together on an empty slot.
And that's the wire-wrap jumper I mentioned.There is a later dual-height
grant continuity card that does connect the NPG lines as well, but you
still have to know how to do the wire-wrap version since any older
non-DMA cards (DL11, PC11, etc) require it.

>
> Regarding the unibus terminator card...M9302 - everyone says "the last slot
> of the unibus". Does this mean physically the last slot in the backplane, or

Physically the last (leftmost) slot. The other slots may well be wired
slightly differently)!. [OK, I know there are exceptions to this, the
VT11 being the obvious one, where the terminator goes somewhere totally
different, but let's keep this simple for the moment).

> "logically the last slot of the unibus" meaning after the last card? I'm
> guessing the former, because it sounds like the last slot of the backplane
> has different wiring, and hence the need for lots-o-grant-cards.

Yes.

>
> The manual states that for each open SPC slot, you have to use a grant card.
> But it's a little unclear as to "SPC" slot. In the "main" backplane, there
> is a hex SPC slot and a quad? SPC slot after the memory, the AB spots on the
> 2nd SPC slot being for the 9202 jumper card. This gives rise to a few
> questions: If the last 2 slots in the "main" backplane are for SPC, are the
> remaining slots (in the 2nd backplane, AFTER the 9202 jumper) SPC slots? I

Yes. SPC = Small Peripheral Controller. 'Normal' backplanes, meaning ones
that weren't designed for specific multi-card peripherals (like the
RK11-D, RK611) are all SPC slots.

> assume you need a grant card in the hex SPC slot if nothing is there. But
> what about the quad SPC slot that also has the 9202 in it - does that slot
> need a grant card too? And about the 1st slot in the 2nd backplane that has
> the other side of the 9202 in it - that needs a grant card as well? I'm
> assuming it does...

Yes. The SPC part is strictly the front 4 connectors only. The rear 2
connecotrs -- A and B -- are either MUD (Modified Unibus Device, used for
some memory cards) or plain Unibus in/out, or even something stranger.
The front 4 connectors of the last slot and first slot on your backplanes
are part of the grant chain and need the continuity cards (and NPG jumpers)

>
> So in the example where you have no cards at all in the 2nd backplane except
> the initial 9202 and the final 9302 at the other end.. you must have a grant
> in every one of those open slots? If so, ummm I need a bunch of grant cards!

Yep, except of course you'd not do that. You'd pull the M920 jumper and
put the terminator in the last slot of the CPU backplane.

-tony
Received on Tue Feb 11 2003 - 18:31:00 GMT