Have you considered doing this for acoustic delay lines?
Actually, I was thinking about making an emulator that would use
.WAV files instead of .T64 or whatever.
>too. However, I could generalize and say they were all DOS-based,
>written in Pascal or assembler, don't come with source code, have
>poor documentation, etc. and I want to roll my own in straight portable
C.
>I'd rather make it general to handle old S-100 tapes, C-64 tapes, etc.
>instead of just hard-coding one flavor. It should be ready in
>the year 2010.
>
Don't bet on it. I have a book that predicted quality software by
1990. Seriously, though, I don't think collectors are in danger of
making their hobby worthless by saving too much. I doubt you will see
an "extra" C-64 in 2010. They will all be in the hands of collectors
or landfills.
>rescue old data. Sure, QuickCams are nearly disposable now. Cheap
>$1,500 PCs include them these days. Five years from now, they'll be
>embedded in cheap monitors. Ten years from now, they'll be in cereal
>boxes. Unless the hobby of collecting computer junk is adopted by
>Hollywood stars, I humbly suggest that it will be at least *more
difficult*
>for you to find spares for your original equipment than it will be
>for me to find something that could deliver a bitmap by looking
>at my punched card. :-)
>
>- John
>Jefferson Computer Museum <http://www.threedee.com/jcm>
>
>
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
>> If not the schematics themselves, then how about a database indicating who
>> has what? Users could submit a list of what they have, and others could
>> search it for needed contacts.
That's a pretty good idea... maybe store machines, related hardware,
whether the user has manuals for the machine or not, other notes etc...
It's a shame though about the copyright situation; I try to store binary
images of everything that I can in the hope that I can recreate boot
disks etc. in the event of problems, but it would be nice to be able to
share those freely with others...
(I suppose the danger with this, as with anything on the 'net, is that
we end up with thousands of people all trying to do the same thing and
end up with the current situation of knowing that the information is
probably there *somewhere* but spending years finding it...)
cheers
Jules
>
On Apr 22, 16:08, Philip.Belben(a)powertech.co.uk wrote:
> Subject: Re[2]: PDP 11/23 help needed
> Seth and Pete were discussing the PDP11-23...
> ISTRT the F11 processor is settable between 18 and 22 bit
I forgot -- there's a bit in SSR3 that sets 18/22. You're quite correct about
that.
--
Pete Peter Turnbull
Dept. of Computer Science
University of York
On Apr 22, 16:08, Philip.Belben(a)powertech.co.uk wrote:
> Subject: Re[2]: PDP 11/23 help needed
> Seth and Pete were discussing the PDP11-23...
>
> >> 2) Same as above, but for the M8044-DB boards. I could put one
> >> of these in with the M8047's to get a full 64Kword of RAM, yes?
> >> Does anyone know what the DIP-switch settings for these boards
> >> are?
> >
> > Yes, but I'm not sure why you say "full" and 64Kword" together :-)
> > 32KW (64KB) is the limit for 16-bit addressing, or 128KW (256KB) for 18-bit
> > addressing. Ignoring the I/O page, that is.
>
> Um... Am I way out here? Doesn't the 23 support 22 bit addressing? And
> I never before heard of a 16 bit Qbus! ISTRT the F11 processor is
> settable between 18 and 22 bit (128KW, 256KB and 2MW, 4MB respectively).
> The 18 bit setting is used in the 23 on 18 bit Qbuses and in the 24 on
> unibuses. The 22 bit setting is used on 22 bit Qbuses, but you need
> extra hardware to use it in the 24 (i.e. on unibus).
An 11/23 only supports 18/22-bit addressing if it has the MMU chip fitted,
which
almost all do, though it was, strictly speaking, an option (at least, for most
of the 23's life).
Early KDF11-A's (Rev.A) only support 18-bit, most (Rev.C) support 22-bit
addressing -- iff they have the MMU. Otherwise, they can only access 16-bits
of
address space. I don't recall any setting to change that, you just only get 18
bits in an 18-bit backplane. If you try to access beyond that range, you get a
bus error. ISTR that the ODT still works (always 18-bit) without the MMU,
though. I can't easily check as the only one I have running ATM is a 22-bit
system.
A number of KDF11-A's were fitted to 11/03's as upgrades, and those had 18-bit
backplanes. The 11/03, however, only had a 16-bit address range, as did early
core, MOS RAM, and ROM boards. I wasn't referring to the bus as 16-bit, but to
the address range.
For slightly different reasons, you can't use an MSV11-D (or several other
options) in a 22-bit system. It will fit in the backplane, and work fine, but
it effectively turns the whole system into 18-bit, because it doesn't decode
BAL18-21, and therefore responds to sixteen blocks of addresses.
For similar reasons, the RKV11-D is not often used in 18-bit or 22-bit systems,
since it can only perform DMA with 16-bit addressing (you can access the
registers in an 18- or 22-bit system, of course, because it responds to the
BBS7 signal). There are lots of other I/O options with similar restrictions.
--
Pete Peter Turnbull
Dept. of Computer Science
University of York
Hi all,
are there any good sites out there containing collections of schematics
for old machines? I occasionally come across sites with a few schematics
/ info for specific machines, but has anyone collected stuff together
for several different machines into one place?
(It's something I keep on meaning to do myself - would there be much
interest in this? I was thinking more along the lines of some of the
more obscure hardware out there though, as there's probably plenty of
places to get details on common machines such as the popular 8-bit
micros of the early 80's. Would be nice to have copies of ROM/Disk
images where possible too...)
cheers
Jules
On Apr 22, 14:25, Julian Richardson wrote:
> >> There is the problem of copyrights and permission. Not as easy as you'd
> >> think as the copyright live past the companies demise so you have to
> >> track where or who still holds it.
>
> I think roms/disks for some machines are no longer copyrighted though,
> no? (or at least freely distributable) - I think Acron's BBC machines
> fall under this category.
That's VERY manufacturer-specific, and doesn't apply to Acorn code. As Allison
pointed out, you need some dispensation, either individually or to the public
at large.
> There seemed to be a lot of people a few years
> back who took to putting rom/disk images up on public sites until
> somebody complained - presumably there's been a crackdown on that sort
> of thing recently though...
Well, Motorola had a go at somebody a year or two ago, as I recall, and ISTR
that ended after a bit of a row with the website in question being closed.
OTOH, there's a site with old Sun boot roms and a note to the effect that if
anyone from Sun objects, the site will remove them immediately on official
request.
--
Pete Peter Turnbull
Dept. of Computer Science
University of York
On Apr 22, 0:16, Seth J. Morabito wrote:
> 1) The M8047-CA boards need to be wire-wrapped to assign their
> address vectors -- they're combination MOS RAM and Async EIA,
> and I have no docs for them. Can anyone guide me to some info,
> or tell me how to jumper one of them to be console serial
> port, and the other to be next in line on the bus?
> The wire-wrap pins have absolutely no markings on them, not
> even any single-letter or number labels, so this one could
> require ASCII-art to describe :)
There are umpteen pages of link settings in the manual. I'm not keen to type
all that in ATM, but if no-one else answers in a few days, I might scan the
pages and accidentally store them on my website (copyright? wassat?)
> 2) Same as above, but for the M8044-DB boards. I could put one
> of these in with the M8047's to get a full 64Kword of RAM, yes?
> Does anyone know what the DIP-switch settings for these boards
> are?
Yes, but I'm not sure why you say "full" and 64Kword" together :-)
32KW (64KB) is the limit for 16-bit addressing, or 128KW (256KB) for 18-bit
addressing. Ignoring the I/O page, that is.
The MSV11-D addresses memeory on any 4KW boundary, set by the switches.
SW1-5 = A13, SW1-1 = A17. Right under Sw1-5 are 2 wrapped links (6 posts) that
set the memory *size* which you shouldn't need to change. There are 3 posts
labelled 6,5,7 which set parity/no-parity; they should be jumpered 5-7 to set
no parity for the -Dx. There are 3 posts labeled 2,1,3 near the B
edge-connector which enable/disable use of the bottom 2K of Bank7; 1-3 to
disable that. Lastly, there are two sets of power jumpers just above the notch
between the connectors; these are used to set battery/no-battery option.
> 3) I'd love to have the RK05 controller in there, in the hopes
> that someday I'll have an RK05 to play with. Just like the
> above... How do I jumper it, and where (physically) in the
> Bus should I put it?
Usually after the other IO/memory options.
> 4) Actually, that raises a good question. All of these boards
> are single-height (1/2 the width of the Q-bus backplane).
> I know there is some special physical layout the boards should
> use when they populate the backplane, but what is it?
> My best (probably wrong) guess right now is:
> CPU in row 1, slot 1 (is that left or right?),
> M8047's in row 2, slots 1 and 2,
> M8044 in row 3, slot 1,
> M7269 in row 4, slot 1, DSD controller in row 4, slot 2.
>
> Does that make any sense? Should the CPU only live in the
> first row, not RAM? I seem to remember something like this
> from the darkest depths of my mind, but I don't remember
> for sure.
The processor should go in slot 1 (top) because that's the only one with
connections to the RUN signal (used for the front panel light), though it will
work elsewhere apart from that. Except in BA23/BA123 cabinets with H
backplanes (see below).
Normally you'd put the memory next, then the I/O, starting with the options
that need the best CPU response (which is often the SLUs, not the disks).
As to which side, that depends on the type of backplane. There are two main
types; "serpentine" (aka "zigzag") which have Qbus in both A-B and C-D slots
(A-B are the left side as you look into the card cage from outside, with slot 1
at the top), and "straight", which have QBus in the A-B slots, and C-D
interconnect in, surprise surprise, C-D. In serpentine backplanes, the slots
are wired in the order 1A/B, 1C/D, 2C/D, 2A/B, 3A/B, 3C/D....
There are variations in microPDP-11 backplanes, where the first 3 (BA23
cabinet, H9278-A backplane) or 4 (BA123 cabinet) slots are wired straight (Qbus
in A-B, interconnect in C-D) and the rest are serpentine. That's to allow PMI
memory in the top, and lots of dual-width options below.
H9273, H9276 are straight. H9270 and H9275 are serpentine. You probably have
an H9273, if it's an early 11/23. Later ones had H9276's (22-bit instead of
18-bit). There should be a label somewhere on it. That means you probably
want all the cards in the left side, and none on the right.
There are also odd ones like the various H9281-x which are only dual-wide, and
the DDV-11 backplane which is hex wide, with Qbus in A/B and C/D, and
interconnect in E/F.
> 5) OK, simple question, one I've wondered about but never bothered
> getting answered because I felt like a complete idiot moron
> asking it: Does the QBUS need to be terminated by a special
> card in any way, in order to work?
Usually, yes. H9275 and H9278 backplanes have built-in terminator options.
Others need a terminator card, such as a BDV11. The normal termination is a
nominal 120 ohms, usually as 180/390-ohm resistor packs.
> 6) What's the pin-out on the M8047 EIA ports? They're 9-pin Berg
> connectors, and I need to build a cable for them to connect
> either to 9-pin or 25-pin PC-style serial in order to set up
> any kind of console terminal.
Looking at the back of the card, components uppermost:
____________________
| |
| 9 7 5 3 1 |
| |
| 10 8 6 4 2 |
|___________________|
1 UART clock in/out (depending on a link, not all SLUs have this)
2 signal ground
3 transmit +
4 transmit -
5 signal ground
6 index - no pin
7 receive -
8 receive +
9 signal ground
10 some SLUs have +12V here, from a current-limiting resistor or a fuse, to
supply a converter. I use it with a 1K series resistor to provide a
pseudo-DTR signal for some terminals.
For normal RS232, link 7-9, and use 3 as XMIT, 8 as RCV, 5 as SG, and ignore
pin 4.
> 7) Anyone know what the Data Systems Design board is? It has
> "RX" stensiled onto the board near the jumper block, among
> other things like "BOOT", so I assume it's some sort of RX01
> or RX50 controller or some such.
Much more likely to be RX01 or RX02 than RX50. What kind of connector does it
have? If 34-pin, it might be RX50 but could be like the Baydel units that use
a 34-pin connector but are actually an SA800 interface. If 50-pin, probably an
SA800/801 interface for 8" drives.
> WHEW, that's _too_ many questions. Anyone who can tackle one of them
> gets my respect, and you may award yourself one cookie.
I prefer flapjacks :-)
> I'd like to piece this system together and get it working to the
> point where I can play with it and at least fiddle with the monitor
> again, playing with Octal. And I'd dearly love to put it in a
> proper DEC desk-side rack with an RK05, but that comes later...
--
Pete Peter Turnbull
Dept. of Computer Science
University of York
"Doug Coward" <dcoward(a)pressstart.com> wrote:
>Now I just have to order some new paper tape.
>It came with 3 rolls but the tape is so old that it breaks when an entire
>row of holes are punched and the sprocket keeps ripping the sprocket
>holes even with no tension on the tape.
I just got 21 rolls of original yellow Teletype 1" tape from
someone on the RTTY mailing list for the cost of shipping, and I
promised to share the wealth, so ...
- John
Jefferson Computer Museum <http://www.threedee.com/jcm>
>> No.
the simple answers are always the best :)
>
>> There is the problem of copyrights and permission. Not as easy as you'd
>> think as the copyright live past the companies demise so you have to
>> track where or who still holds it.
I think roms/disks for some machines are no longer copyrighted though,
no? (or at least freely distributable) - I think Acron's BBC machines
fall under this category. There seemed to be a lot of people a few years
back who took to putting rom/disk images up on public sites until
somebody complained - presumably there's been a crackdown on that sort
of thing recently though...
(always seems a pity that the information is out there... somewhere...
but almost impossible to get to!!)
ta
Jules
>
>
From: Max Eskin [mailto:maxeskin@hotmail.com]
<< I'm curious about the various home/small business computer standards.
I know about the PC standard *sigh*. There was also the MSX standard
which involved a Z80 and 64K ram, I think. What other ones were there?
>>
Quite a few. from memory....the S-100, probably the earliest micro bus
to become popular. A standard business configuration for 8-bit would be
a Z-80, 64K RAM, 5.25" floppies, serial ports to a CRT terminal and
printer, and CP/M-80, running applications written in CBASIC. For
16-bit systems it would be an 80286, 1MB RAM, hard drive in the 40MB
range, 4 to 16 serial ports to CRT terminals and printers, maybe even a
modem, running MP/M or Concurrent DOS, again with applications written
in BASIC. A lot of single board Z80, 8086, and 80286 systems (like
Altos) built this same basic configuration, but without an expansion
bus.
The SS-50, a competitor for the S-100 but using Motorola 6800 CPU, never
really caught on. VME, an industrial bus still alive today, lots of
different CPUs supported over the years, but it started with an Intel
8080. Motorola came up with a competing bus that looked very similar to
S-100 except it used 86 pin bus, called Exor, something like that, never
caught on either, though Motorola did their best to promote it for a
while.
There were some holdovers from the mini makers. Q-bus, DEC's bus for
the PDP-11 micros and later the MicroVAX. A typical PDP would be a
PDP-11/03 CPU, 64K RAM, 2.5MB hard drive (RK05), 4 port serial to a
VT-100 terminal, and a DECwriter II for a printer. Business apps were
written in FORTRAN, DIBOL, or BASIC, usually running on the RT-11
operating system. BTW the Q-bus wasn't strictly DEC proprietary, DEC
sold PDP-11 CPU chips for a while, Plessey made their own PDP-11 systems
with them. DEC later dropped CPU sales when they found out the other
manufacturers were just taking away DEC customers, not expanding into
new markets. Sort of like MAC clones... <s>
VAX Q-bus configurations are still around in sizeable numbers (we have
some in the office). The machines aren't fast, but very reliable (as in
years per hardware failure). It's not unheard of for VMS systems to run
for a year or more between boots. A common setup in the late 80's was
the MicroVAX II (roughly equivalent to a 386), a 32-bit CPU with up to
16MB of RAM. An 8 slot chassis, holding the CPU board, 2 memory boards,
disk controller (RQDX3, MFM), tape controller (TK50), Ethernet
controller, and an 8-port serial card (DHV11). Drives would be either
70MB or 160MB. The tape drive could hold about 95MB. It ran VMS (still
in use today), or less commonly Ultrix (DEC Unix). Several languages
are available, including C, FORTRAN, COBOL, and BASIC. DEC has
discontinued new VAX Q-bus machines, and new VAXes in general, but they
have large stocks of VAX CPU chips, enough to last a few more years.
Intersil made the IM6100, a microprocessor version of the 12-bit PDP-8,
but I don't recall if a bus was ever associated with it. Data General
had a bus for the NOVA minis and I believe Fairchild made a micro
version based on the 9440 microprocessor (NOVA instruction set). Texas
Instruments made the 9900, a micro version of their minis, and it had
some kind of proprietary TI bus.
Jack Peacock
