FWIW...we ran 4.3 (locally hacked up, of course) on our 11/750s at
GaTech in the late 80s and it was really very pleasant. A lot of
good research work got done on those boxes (as well as a ton of
nethack).
KJ
"Bob Armstrong" <bob at jfcl.com> wrote:
>> Dave McGuire wrote:
>> I think 5MB is correct. And don't they use the same memory array
>> boards as the PDP-11/70?
>
> Yes, the 725/730 is limited to 5Mb, although it's tricky to find enough
> slots for this if you've got any peripherals. I have a 11/730 that's
> fortunate to also have the BA11-K UNIBUS expander box, so I've got room for
> 5Mb, but that was an unusual configuration.
>
> The memory boards used in the 730 are the same as the 11/750 1Mb memory
> boards.
Fun. I didn't know the 11/730 used the same memory bus.
The 11/750 and PDP-11/70 MK-11 box uses the same memory bus, but
unfortunately the MK-11 don't support any larger arrays than 256 KB
memory boards.
Johnny
Google is failing me on this one. A while back, I picked up some random boards on ePay. The boards all have the name TRIAD, and they include a passive backplane, what appears to be a central controller with an FPGA and a bunch of RAM, and several boards that each contain two Z80s and ten 8530 serial communication chips. One of the oddest things in the silk screening is the identification of two of the ribbon connectors across the top of these latter boards: "LAMB/DROID, ODD HALF" and "EVEN HALF". (Ironically, I just finished a really good science fiction story regarding a sheep, entitled "The Android's Dream"....)
The only hints I've found are that this may be the guts of a PBX or other telecom unit. Does anyone have any insight on this? I'm asking because there are a lot of cool chips on these boards, but before I render them down for spare parts I want to be sure I'm not destroying anything significant. It sure looks like some sort of distributed processing architecture, but I have no idea what it was processing. Thanks for any ideas -- Ian
I recieved an email from someone looking to sell an IBM PS2 8573-061 (he
saw my name mentioned in a newpaper article about collecting old computers
and contacted me). Asking about the price, here's what he said:
> I did a little research on pricing this morning and I see the same unit at
> Computer Fusion Inc. for $425 and on eBay for $490. I'd be willing to sell
> mine for half price to get it out of my garage.
>
> I'm thinking of putting it on eBay. Let me know.
If anyone is interested, say so and I'll pass on his email address.
The person lives in Boyton Beach, Florida and works in Boca Raton (home of
the IBM PC).
-spc
Hi folks,
A few of you might remember the Jupiter Ace I had that took a 9V spike to
the expansion slot -- and my futile attempts at repairing it. It's been well
over four years since I sent it to a listmember who offered to repair it,
and
all attempts to get the board or the spares I sent with it have failed. At
this point, I haven't seen hide nor hair of him in months, although he's
apparently still updating his website...
So basically I'm the proud (?) owner of the two sections of case, most of
the snap-rivets (apparently Maplin sell these, so I might be able to
complete
the set), the rubber keyboard membrane, documentation, demo tape and a
(supposedly working) 16K RAM pack.
I've heard rumours of spare, unassembled Jupiter Ace PCBs kicking around;
does anyone happen to have one up for grabs? It looks like the only way I'm
going to get this thing up and running again would be to build a new
mainboard
>from scratch, thus that's what I'm planning to do. About the only parts I
need
are a Z80 CPU and a couple of 74LS logic chips.
Only other alternative would be to get a few PCBs made up for it, then
build one from scratch on that -- the catch being I haven't had any luck
creating a decent track master from the scans of the PCB that I have....
Can anyone assist me in my Quest to Build a Working Jupiter Ace (tm)?
Thanks,
--
Phil.
classiccmp at philpem.me.uk
<http://www.classiccmp.org/mailman/listinfo/cctalk>
http://www.philpem.me.uk/
-----REPLY-----
Hi Phil,
If it were me, I would just take the information from the wikipedia page and
build your own replacement PCB. The Jupiter Ace appears to be a pretty
straight forward Z80 SBC so just build a prototype board and then get KiCAD
and design your own PCB. With some luck and patience you probably can
design a board that fits right in the original case.
You might even be able to recoup your costs of the PCB manufacturing run by
selling PCBs or even fully manufactured SBCs to other Jupiter Ace
enthusiasts. I'd guess they'd be thrilled to see some new hardware
available. Just a suggestion.
Thanks and have a nice day!
Andrew Lynch
Hi! I am building some prototypes for the N8VEM home brew computing project
and using the new ECB Prototyping Board. It works great and it allows
wire-wrap and/or point-to-point soldering construction. I am making an ECB
serial board as part of another larger development.
I have some wire-wrap supplies, sockets, and tools but I can see I am fairly
quickly using them. If anyone has any old wire-wrap stuff they don't need
anymore and would like to sell for a nominal amount, or just find a good
home for where you know they'll get used as intended, please contact me.
Any supplies offered will be used as intended not scrapped or sold.
I will gladly pay for shipping and/or nominal price. Obviously I would
rather use old stock than buy up a bunch of new stuff since I will bet there
are lots of people with old wire-wrap stuff they don't use anymore.
If anyone can help out I certainly would appreciate it. Thanks and have a
nice day!
Andrew Lynch
> > I wish that I knew a good way in HTML to express offsets from beginning
> > of a document.
>
> Are you taking about an artibrary HTML document that you may not have
> control over? Or just for HTML documents you control? If the latter, then
> there are two methods---one works for all browsers, and the other for more
> modern ones.
Just thinking of the PDP-10 archives, it would be real nice if there
were a standard way to let someone reference into the middle of a document
with a standardized content-based offset reference.
For example, someone who is talking about historical LISP's wants
to reference something in the MACLISP reference manual pulled from
a 1970's tape on my site. The original document was not, of course, in
HTML; it was just HTML rendered for web presentation.
They can link to the entire HTML-rendered
version of the document, but it's hundreds of screens long. It would
be nice if there were a way to automatically and context-sensitively
reference something in the middle of the document - and in a way
such that future renderings of the same document still had the same
tags, despite moving to different presentation technologies.
Since many of the original documents were formatted to be presented
on line printers, or on terminal screens, the concepts of "page" and
"line" usually exist. Some of them have originals in RUNOFF-type
sources, where there may be some kind of context-referencing system
in place, but the details of the kind vary from document to document.
So it looks like the best way is to let others index in by original
page number, or original line numbers.
A future project would probably be a way of turning all the different
variants of RUNOFF (and the variants span 3 decades at least, 4 or 5
decades if you count those who still used it in the 90's and 2000's)
into HTML with good invariant content-based subreferences.
Tim.
Hi folks,
I just found out that someone called Hans Pufal has already designes a
PDP-8 on an FPGA device.
I've been doing the same this year. My first approach ran on a Xilinx
Spartan-3 (200K). It passed all basic tests and was able to run Chekmo.
The system could be clocked with about 80MHz using Xilinx ISE's free XST
compiler and mapping tools.
After realizing that I would like to have a complete 8/e implementation
(with more memory), I redesigned the whole thing a lot.
The new design seemed more reasonable to me - but gets worse synthesis
results. I still expect about 50-70 MHz in the end. The current version
is not yet debugged.
All memory is kept in FPGA onchip block rams. RAM is capsuled into a
pdp8_memory module that can be altered in size (several versions exist)
and adapted to different FPGA architectures.
I have a kind of SoC-OMNIBUS which supports 2 to n cycle IO operations.
Currently I only have a TTY implementation for it. The TTY seems to be
100% compatible to the original. RK8E or similar are planned.
Cycle times:
* Memory reference 2
* Operate 1
* Jump 1
* ISZ 3
Indirection, auto-index each add one cycle.
My core will have a front panel interface that allows attachement of
several flavors of front panel logic, including the original
functionality and perhaps some more (I hate not being able to directly
manipulate AC, for example!).
I would like to know if there is any "public" interest in my project. I
appreciate every help or ideas to merge my project into one of those
nice front panel projects I've seen on the web.
And now, please comment!
Best wishes,
Philipp :-)
--
http://www.hachti.de
