21 May
2007
21 May
'07
4:34 a.m.
Patrick Finnegan <pat at computer-refuge.org> skrev:
Pat, what you have is what I would describe as an 11/93 with an
Qniverter. That will not be exactly the same thing as an 11/94.
The KTJ11-B, which is DECs Unibus adapter, have some special signals to
the CPU telling it that the KTJ11 is there, and that changes the
behaviour of the Qbus, according to the manuals.
I haven't said that the wiring of the memory slots in an 11/84 (or
11/94) isn't the same as in a Q/CD backplane. It must (by default) be
atleast very similar to a Q/CD backplane. However, there are some
special signals, and also some signals that change behaviour in an 11/84
and 11/94, according to the manuals. And that will only happen if you
have a KTJ11-B in there.
Johnny
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
On Saturday 19 May 2007 06:35, Johnny Billquist
wrote:
> > "Glen Slick" <glen.slick at gmail.com> skrev:
> > > On 5/17/07, Rod Smallwood <RodSmallwood at
mail.ediconsulting.co.uk>
wrote:
> > >> > OK we have a change
> > >> >
> > >>> > >PMI memory goes in slots 1 and 2, CPU goes in 3
> > >>> > >joe lang
> > >> >
> > >> > Rod
> > >
> > > From my understanding of this after looking at various manuals
this
> > is
true for an 11/73 or 11/83 with an H9872 backplane in a BA23
> > box, but not for an 11/84.
>
> Correct. So if people could stop assuming that an 11/84 have a q-bus,
> we would get a long way towards clearing this up.
> If people don't know about the 11/84 or 11/94, don't write answers
> based on your knowledge of the q-bus based KDJ11 setups.
My "11/94" (which was an upgraded 11/84, and which I got CPU-less)
actually does have a QBUS... it's got an Able Qniverter to hook the
UNIBUS in the chassis to the QBUS from the CPU, and both QBUS and UNIBUS
peripherals.
Does anyone know how common that was? I don't see any evidence pointing
towards it not being designed that way by DEC. The chassis is DEC
badged, with a label on the side indicating that it was upgraded from an
11/84 to be an 11/94 at some point. 
21 May
21 May
12:27 p.m.
On 21/05/2007 09:34, Johnny Billquist wrote:
Patrick Finnegan <pat at computer-refuge.org>
skrev:
Pat, what you have is what I would describe as an
11/93 with an
Qniverter. That will not be exactly the same thing as an 11/94.
The KTJ11-B, which is DECs Unibus adapter, have some special signals to
the CPU telling it that the KTJ11 is there, and that changes the
behaviour of the Qbus, according to the manuals.
Yes, by enabling the PMI, that's all.
I haven't said that the wiring of the memory slots
in an 11/84 (or
11/94) isn't the same as in a Q/CD backplane. It must (by default) be
atleast very similar to a Q/CD backplane.
It's exactly the same. Or to be precise, the A/B slots are normal
22-bit QBus; the relevant C/D signals are fully bussed on their top
surfaces, rather than just connecting the bottom of each slot to the top
of the next one down. That won't matter to PMI memory cards, which have
the fingers on the top surface connected to the fingers on the bottom.
Nor will it matter to a KDJ11-B, which (apart from power) has
connections only on the top surface.
However, there are some
special signals, and also some signals that change behaviour in an 11/84
and 11/94, according to the manuals.
The special signals are the standard PMI signals on C/D (plus one that
tells the processor it's in a Unibus system, IIRC), and the ones that
change are the ones that do so when moderated by the PMI protocol.
And that will only happen if you
have a KTJ11-B in there.
Or something else to assert the PMI :-)
--
Pete Peter Turnbull
Network Manager
University of York
1:19 p.m.
On Monday 21 May 2007 04:34, Johnny Billquist wrote:
Patrick Finnegan <pat at computer-refuge.org>
skrev:
I'd agree, but the case quite clearly has a Digital logo on it, an 11/94
placard, and has a note on the side that it was upgraded to an 11/94
On Saturday 19 May 2007 06:35, Johnny Billquist
wrote:
My "11/94" (which was an upgraded 11/84, and which I got CPU-less)
actually does have a QBUS... it's got an Able Qniverter to hook
the UNIBUS in the chassis to the QBUS from the CPU, and both QBUS
and UNIBUS peripherals.
Does anyone know how common that was? I don't see any evidence
pointing towards it not being designed that way by DEC. The
chassis is DEC badged, with a label on the side indicating that it
was upgraded from an 11/84 to be an 11/94 at some point.
Pat, what you have is what I would describe as an 11/93 with an
Qniverter. That will not be exactly the same thing as an 11/94.
The KTJ11-B, which is DECs Unibus adapter, have some special signals
to the CPU telling it that the KTJ11 is there, and that changes the
behaviour of the Qbus, according to the manuals. from an 11/84. I see no evidence that DEC didn't
sell this machine as
it currently is configured, as an 11/94 (or 11/84).
One more interesting bit about it is the QBUS backplane... the first 3
slots are quad-width Q/CD slots for the memory and CPU, and the rest of
the slots (quite a few, maybe 6 or 8?) are hex-width slots that are Q/Q
serpentine on the left 4 fingers, and CD on the right 2 fingers.
Pat
--
Purdue University Research Computing --- http://www.rcac.purdue.edu/
The Computer Refuge --- http://computer-refuge.org
22 May
22 May
5:03 p.m.
On 21/05/2007 18:19, Patrick Finnegan wrote:
DEC certainly sold upgrade kits to convert an 11/84 to an 11/94; the
backplanes are the same, and the kit included new processor, cables, and
instructions, etc. Plus the badge and a sticker for the backplane to
note the upgrade. I don't have the complete details to hand, but it's
in one of the 11/94 manuals on bitsavers.
On Monday 21 May 2007 04:34, Johnny Billquist wrote:
Pat, what you
have is what I would describe as an 11/93 with an
Qniverter. That will not be exactly the same thing as an 11/94. The
KTJ11-B, which is DECs Unibus adapter, have some special signals
to the CPU telling it that the KTJ11 is there, and that changes the
behaviour of the Qbus, according to the manuals.
I'd agree, but the case quite clearly has a Digital logo on it, an
11/94 placard, and has a note on the side that it was upgraded to an
11/94 from an 11/84. I see no evidence that DEC didn't sell this
machine as it currently is configured, as an 11/94 (or 11/84). One more interesting bit about it is the QBUS
backplane... the first 3
slots are quad-width Q/CD slots for the memory and CPU, and the rest
of the
slots (quite a few, maybe 6 or 8?) are hex-width slots
that are Q/Q
serpentine on the left 4 fingers, and CD on the right 2 fingers.
That is odd. Can you see a number on the backplane? Does this machine
have two backplanes, one entirely QBus and one entirely Unibus? If the
whole of the CPU backplane is QBus, I sort of agree with Johnny that
it's not a "real" 11/94, although software probably can't tell the
difference.
According to my manuals, there are two backplanes for 11/84
systems. The Users and Maintenance Guide for the 11/84-A (and -P)
(EK-1184A-MG-001, June 1988) shows a 13-slot H9277A, part number
70-20650-01. The Users and Maintenance Guides for the 11/84-D and -E
(EK-1184D-MG-001, EK-1184E-MG-001, June 1988) show a 9-slot H9277B, part
number 70-17228-01. But in both cases, the slots below slot 4 are
Unibus. Specifically, for the 9-slot backplane, the slots are listed as
Slot 1 quad width Q22/CD for CPU
Slot 2 quad width Q22/CD for memory
Slot 3 quad width Q22/CD for memory or MLM
Slot 4 hex width special for KTJ11B
Slot 5 hex width Unibus/SPC for hex or quad Unibus device
Slot 6 hex width Unibus/SPC for hex or quad Unibus device
Slot 7 hex width MUD/SPC for hex or quad Unibus device
Slot 8 hex width MUD/SPC for hex or quad Unibus device
Slot 9 hex width Unibus out/SPC for quad Unibus or M7556 MLM
Another difference is that the topmost slot in the H9277A (labelled MDM,
it would be slot 0 rather than slot 1, if they were all numbered) is for
the monitor board. That doesn't seem to be the case in the H9277B,
where the topmost slot is slot 1, for the processor. In both cases,
though, the top slots are quad slots for CPU and memory, and the rest
are hex.
The H9277B is also the backplane used in 11/94-E systems.
--
Pete Peter Turnbull
Network Manager
University of York
11:18 p.m.
On Tuesday 22 May 2007 17:03, Pete Turnbull wrote:
On 21/05/2007 18:19, Patrick Finnegan wrote:
Well, the system has two backplanes.
The system has two labels on the side... the first is "11/84-HX" and the
second is "PDP-11/94E"
The QBUS backplane has 3 quad-width slots and 6 hex-width slots. It's
labelled:
5018955 01 B1
11/84 QBUS BACKPLANE
One more interesting bit about it is the QBUS
backplane... the first
3
slots are quad-width Q/CD slots for the memory and CPU, and the
rest of the
slots (quite a few, maybe 6 or 8?) are hex-width slots that are Q/Q
serpentine on the left 4 fingers, and CD on the right 2 fingers.
That is odd. Can you see a number on the backplane? Does this
machine have two backplanes, one entirely QBus and one entirely
Unibus? If the whole of the CPU backplane is QBus, I sort of agree
with Johnny that it's not a "real" 11/94, although software probably
can't tell the difference. From the labelling, it's definately a DEC
backplane. I can't find any
other (ie Hxxxx) labelling on the backplane.
The UNIBUS backplane is made by ABLE.
According to my manuals, there are two backplanes for
11/84
systems. The Users and Maintenance Guide for the 11/84-A (and -P)
(EK-1184A-MG-001, June 1988) shows a 13-slot H9277A, part number
70-20650-01. The Users and Maintenance Guides for the 11/84-D and -E
(EK-1184D-MG-001, EK-1184E-MG-001, June 1988) show a 9-slot H9277B,
part number 70-17228-01. But in both cases, the slots below slot 4
are Unibus. Specifically, for the 9-slot backplane, the slots are
listed as
Well, the label on the side says it was an 11/84-HX, which isn't any of
the models you listed above...
Pat
--
Purdue University ITAP/RCAC --- http://www.rcac.purdue.edu/
The Computer Refuge --- http://computer-refuge.org
6266
days inactive
6268
days old
4 comments
3 participants
participants (3)
-
bqt＠softjar.se -
pat＠computer-refuge.org -
pete＠dunnington.plus.com