22 Jul
2018
22 Jul
'18
10:25 a.m.
s?ndag 22 juli 2018 skrev Paul Birkel <pbirkel at gmail.com>:
26 bits (or 13 bits) doesn't make any sense on a
16-bit machine; makes
more sense as a high-speed I/O buffer.
One can note that it is actually two different types of 1k chips. 16 chips
are 94L415 and 10 chips are 93415. As far as I understand the L is the
slower variant.
This could mean that 16 bit data is in the L chips while the faster chips
are used for a 10 bit cache tag. Maybe 8 address bits plus some valid bit
and possibly a dirty bit?
The switch is marked ON/OFF which could simply cache on/off. The
handwritten label on the board says that it is not in use and should sit in
slot 21.
And of course those two I/O connectors don't belong on a cache.
Those IO connectors are connected to two double height boards in 26 /27 AB.
They are also made by ACT and contain a few TTL chips.
So it pulls out some signals out of both Unibuses but 20 + 10 signals at
most it not much of a complete bus so I wonder what kind of signals go
there.
While odd to use slot 21 (Fastbus) for something other
than memory I don't
know why a fast memory-mapped I/O channel couldn't go there.
Also note all of the signals employed on tabs C-D-E-F?
It may not even employ the Fastbus; just talk to Unibus B.
Unfortunately there's not much documentation for the MS11.
It seems likely that A-B isn't anything like the usual Unibus signals, and
who knows where the Fastbus signals are routed.
On my backplane D-E are essentially unused whereas A-B-F are busiest.
I see the marking "copyrighted 1976", which is rather earlier than the ACT
/ ABLE documentation online.
From Bitsavers see the ABLE documents for the SCAT/45:
Able_Computer_Product_Summary.pdf - page 3
Able_Computer_Product_Brochures_1982.pdf - pages 16-17
The PN 10003 doesn't seem to match anything documented from ACT, however
it's consistent with them.
The original QBus Univerter is PN 10001, and is dated 1976.
There are some documentation to get with the machine so the manual for the
board might turn up.
What are the pair of DIP24 ICs on the lower-left?
Fairchild 9308 Dual 4 bit latches.
-----Original Message-----
From: cctalk [mailto:cctalk-bounces at classiccmp.org] On Behalf Of Paul
Anderson via cctalk
Sent: Saturday, July 21, 2018 6:54 PM
To: Mattis Lind; General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Strange third party board in PDP-11/45
I think it's Applied Computer Technologies, and I think they made cache
and several other options. They were popular back in the day. I have a
bunch of their boards here.
Paul
On Sat, Jul 21, 2018 at 1:37 PM, Mattis Lind via cctalk <
cctalk at classiccmp.org> wrote:
This board was sitting in slot 21 of the
backplane in a 11/45
https://i.imgur.com/ZYWZQCo.jpg
What kind of board is this?
It has 26 bipolar RAMS. Fairchild 93415 1kbit SRAM.
The manufacturer might be ACT whatever that is.
My guess is that it is some kind of cache board? It is connected to both
unibuses in the machine.
Better ideas? Documentation?
/Mattis
