On 07/05/2012 09:00 PM, Paul_Koning at Dell.com wrote:
On Jul 5, 2012, at 7:35 PM, Johnny Billquist wrote:
...
Oh, and I saw some additional discussions about the CIS option for the F11. Unless I remember wrong, and understood things wrong, it is not just a simple PROM. It's a coprocessor, which have logic in addition to whatever microcode the coprocessor need.
So no way of just duplicating a PROM to get a CIS.
Not unless someone could dig up the details. If it *is* just a PROM, even if it's an oddball bipolar model with different interface specs, it should be easy to reproduce that with a small FPGA. Any reasonable modern FPGA is way faster than a 1980s era bipolar ROM, so modeling the interface timing should be easy. And level conversions should be doable.
So a sort of daughterboard? That's a good idea.
I'm pretty sure the F11's manuals are out there, but I haven't gone
digging for them. From that, and possibly even just from the KDF11-Bx
schematics, one should be able to glean the important parts of how to
interface to the microinstruction and microaddress buses.
Then all we'd need is a copy of the actual code.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
On Jul 5, 2012, at 9:03 PM, Steve Davidson wrote:
...
The node PLUTO:: is a PDP-11/23+ with both CIS and FPU running RSTS/E.
You can run an emulated one using SimH if you do not have the physical
hardware. I have both running. You would be hard pressed to know if
the system running (and connected to HECnet) was the real one or not!
COBOL was the reason that CIS was designed for the 11/23+. Other
languages can also make use of it. One that comes to mind is DIBOL.
I don't know if Dibol uses it. I do know that the RSTS kernel does for memcpy.
paul
On Jul 5, 2012, at 7:35 PM, Johnny Billquist wrote:
...
Oh, and I saw some additional discussions about the CIS option for the F11. Unless I remember wrong, and understood things wrong, it is not just a simple PROM. It's a coprocessor, which have logic in addition to whatever microcode the coprocessor need.
So no way of just duplicating a PROM to get a CIS.
Not unless someone could dig up the details. If it *is* just a PROM, even if it's an oddball bipolar model with different interface specs, it should be easy to reproduce that with a small FPGA. Any reasonable modern FPGA is way faster than a 1980s era bipolar ROM, so modeling the interface timing should be easy. And level conversions should be doable.
paul
On 2012-07-06 01:42, Boyanich, Alastair wrote:
Oh, and I saw some additional discussions about the CIS option for the
F11. Unless I remember wrong, and understood things wrong, it is not
just a simple PROM. It's a coprocessor, which have logic in addition
to
whatever microcode the coprocessor need.
So no way of just duplicating a PROM to get a CIS.
Ah ha! Game over then. Pity. Shall have to keep poking about for bits
for the 23Plus then :)
Yeah. Well, had it been that easy, the market would have been flooded with CIS options. :-)
Johnny
Oh, and I saw some additional discussions about the CIS option for the
F11. Unless I remember wrong, and understood things wrong, it is not
just a simple PROM. It's a coprocessor, which have logic in addition
to
whatever microcode the coprocessor need.
So no way of just duplicating a PROM to get a CIS.
Ah ha! Game over then. Pity. Shall have to keep poking about for bits
for the 23Plus then :)
Thanks!
Al.
On 2012-07-05 23:43, Boyanich, Alastair wrote:
G'day Johnny,
Okay, then "someone on the internet is wrong" (tm) .. will accept your
version of it. :)
The internet is wrong most of the time, if you just start digging into enough details. :-)
Was going off this site:
http://www.cpu-galaxy.at/Boards/Boards_CPU/Boards_CPU.htm ..and a
couple of others that've probably sourced it.
Yep, the information there is wrong. No CIS for the J11. The only CPUs that ever got the CIS was the F11 and the 11/44.
(Now I'm sure someone is going to point out the 11/74 CIS option, but it was never sold, even though I've heard that at least one prototype was done.)
Okay, I'll check the 23Plus when I get home. Might get lucky there and
be able to have a play with it there. On topic, seems to work in the
quick test (sadly had work to do at home) I did last night which was a
couple of memory fills so that's promising.
Right. So no CIS for 83/94, but there is for 23Plus.
Right.
Thanks for the straightening out :)
No problem.
Oh, and I saw some additional discussions about the CIS option for the F11. Unless I remember wrong, and understood things wrong, it is not just a simple PROM. It's a coprocessor, which have logic in addition to whatever microcode the coprocessor need.
So no way of just duplicating a PROM to get a CIS.
Johnny
Al.
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On
Behalf Of Johnny Billquist
Sent: Thursday, 5 July 2012 5:45 PM
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] pdp-11/83-84
On 2012-07-05 09:40, Johnny Billquist wrote:
On 2012-07-05 08:22, Boyanich, Alastair wrote:
G'day $ALL.
Quick question. I've just snaffled an M8190-AE with what I think is
the
21-21853-06 FPU as an upgrade for either the 11/70 or 11/23Plus at
home.
I've seen reference to CIS microcode for these boards about the
place
and am extremely keen to have a go at it with macro. I'm not
"actually"
sure if it has it in the PROMS, but being handy with a EPROM
programmer
hopefully I can find it. Does anyone have any reading material on
how to
figure out if the CIS microcode is installed ?
This CIS option is not a PROM, even though it might sound like that.
It's more like a co-processor.
If the CIS options is installed, you can easily see that,
physically, by
the large chips next to the CPU. If it's not there, you have some
empty
sockets...
Oh! And Darn! I should have checked proper. The M8190 is indeed the
11/83 and 11/84 CPUs. That's a J11 CPU. There is *no* CIS for this
CPU.
I thought you were talking about the 11/23+ and 11/24 in my first
reply.
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 07/05/2012 05:48 PM, Boyanich, Alastair wrote:
Thanks, Yeah I was more than a bit hot for it also. Looks like it came
out of Ford Aerospace and Comms Corp (P.S. FAAC Best acronym next to our
old one of FACOM which was changed when someone heard a .jp accent
saying "FAKKUM" to one of the clients *G*). No clue how it ended up in
Adelaide unless they were doing defence contracting.
Yeah really! But, a great score either way.
You CAN, however, get a CIS option (if you can find it!) for your
11/23! I too would love to mess with CIS a bit; I've never done so.
I
have two CIS option board sets for my 11/44s but have not yet
installed
them.
Nice! (tm) .. got a photo's of the 11/44 option?
Not yet, but keep after me about it and I will photograph the boards.
Really wondering about
the /23Plus. If the option for it is just PROM's that should be pretty
easy to reproduce.
It won't be at all easy to reproduce. They are MICROMs. The F11 CIS
option is one of those multi-chip carriers, but for CIS it is two
40-pin-DIPs wide, and contains six MICROM chips. These are bipolar
mask-programmed ROMs with an (I think?) undocumented interface. I would
be more than a little surprised if anyway, even someone of technical
means, could reproduce that.
Since you have an 11/23+, you are in luck; that has one more 40-pin
socket than the dual-width 11/23 CPU board. It can accommodate both the
CIS MICROM option and the floating point MICROM option. On the
dual-width 11/23 boards, you have to remove the floating point MICROM.
You could remove the MMU chip to make room for the double-wide CIS
chip carrier, but that'd be pointless because, if memory serves, the
floating point registers reside within the MMU chip.
The 11/24 has yet one more socket, so it has no such limitations.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
G'day Dave,
Nice!
Thanks, Yeah I was more than a bit hot for it also. Looks like it came
out of Ford Aerospace and Comms Corp (P.S. FAAC Best acronym next to our
old one of FACOM which was changed when someone heard a .jp accent
saying "FAKKUM" to one of the clients *G*). No clue how it ended up in
Adelaide unless they were doing defence contracting.
Huh? That's a J-11 based board, and I'm pretty sure the CIS option
for the J-11, while planned, was never actually shipped. It would
have
been two additional chip carriers on the bottom side of the big white
J11 multi-carrier chip.
The only ROMs on a KDJ11-Bx board contain the boot/diag code, not
microcode, and conversely, the only microcode on a KDJ11-Bx (or any
J11-based board for that matter) is in the microcode ROM, which is a
part of the "control" chip, which is one the two chips on the J11
multi-chip carrier itself. No microaddress lines or any other
internal
state machine stuff is brought out to the pins of a J11.
(sorry!)
You CAN, however, get a CIS option (if you can find it!) for your
11/23! I too would love to mess with CIS a bit; I've never done so.
I
have two CIS option board sets for my 11/44s but have not yet
installed
them.
Nice! (tm) .. got a photo's of the 11/44 option? Really wondering about
the /23Plus. If the option for it is just PROM's that should be pretty
easy to reproduce.
Al.
G'day Johnny,
Okay, then "someone on the internet is wrong" (tm) .. will accept your
version of it. :)
Was going off this site:
http://www.cpu-galaxy.at/Boards/Boards_CPU/Boards_CPU.htm ..and a
couple of others that've probably sourced it.
Okay, I'll check the 23Plus when I get home. Might get lucky there and
be able to have a play with it there. On topic, seems to work in the
quick test (sadly had work to do at home) I did last night which was a
couple of memory fills so that's promising.
Right. So no CIS for 83/94, but there is for 23Plus.
Thanks for the straightening out :)
Al.
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On
Behalf Of Johnny Billquist
Sent: Thursday, 5 July 2012 5:45 PM
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] pdp-11/83-84
On 2012-07-05 09:40, Johnny Billquist wrote:
On 2012-07-05 08:22, Boyanich, Alastair wrote:
G'day $ALL.
Quick question. I've just snaffled an M8190-AE with what I think is
the
21-21853-06 FPU as an upgrade for either the 11/70 or 11/23Plus at
home.
I've seen reference to CIS microcode for these boards about the
place
and am extremely keen to have a go at it with macro. I'm not
"actually"
sure if it has it in the PROMS, but being handy with a EPROM
programmer
hopefully I can find it. Does anyone have any reading material on
how to
figure out if the CIS microcode is installed ?
This CIS option is not a PROM, even though it might sound like that.
It's more like a co-processor.
If the CIS options is installed, you can easily see that,
physically, by
the large chips next to the CPU. If it's not there, you have some
empty
sockets...
Oh! And Darn! I should have checked proper. The M8190 is indeed the
11/83 and 11/84 CPUs. That's a J11 CPU. There is *no* CIS for this
CPU.
I thought you were talking about the 11/23+ and 11/24 in my first
reply.
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 07/05/2012 02:22 AM, Boyanich, Alastair wrote:
Quick question. I've just snaffled an M8190-AE with what I think is the
21-21853-06 FPU as an upgrade for either the 11/70 or 11/23Plus at home.
Nice!
I've seen reference to CIS microcode for these boards about the place
and am extremely keen to have a go at it with macro. I'm not "actually"
sure if it has it in the PROMS, but being handy with a EPROM programmer
hopefully I can find it. Does anyone have any reading material on how to
figure out if the CIS microcode is installed ?
Huh? That's a J-11 based board, and I'm pretty sure the CIS option
for the J-11, while planned, was never actually shipped. It would have
been two additional chip carriers on the bottom side of the big white
J11 multi-carrier chip.
The only ROMs on a KDJ11-Bx board contain the boot/diag code, not
microcode, and conversely, the only microcode on a KDJ11-Bx (or any
J11-based board for that matter) is in the microcode ROM, which is a
part of the "control" chip, which is one the two chips on the J11
multi-chip carrier itself. No microaddress lines or any other internal
state machine stuff is brought out to the pins of a J11.
(sorry!)
You CAN, however, get a CIS option (if you can find it!) for your
11/23! I too would love to mess with CIS a bit; I've never done so. I
have two CIS option board sets for my 11/44s but have not yet installed
them.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
