IBM Selectric parallel/serial interface
Curtis Rempel
name) in the circa '78-'81 era that allowed you to convert an IBM Selectric
typewriter into a printer? I even remember an article in one of the
hobbyist magazines at the time that described how to build such a device
that would do the ASCII to EBCDIC translation. I think there were two
flavors of this idea - one was a box with a bunch of solenoids that would
actually sit on top of the keyboard of the typewriter and mimic a real
person pressing the keys (can you imagine such a thing?). The second flavor
was I believe for IBM's own modified Selectric typewriter that they used in
banks and such as hardcopy terminals. I don't know the model number of this
unit but it did have some sort of interface, parallel or serial, that spoke
EBCDIC.
The reason I ask these questions is that I am interested in acquiring a unit
of the latter flavor for purely hobbyist use to add to my collection.
Would anybody know anything about these units and where they could still be
found (certainly not in active service but perhaps at surplus shops and the
like)?
Thanks,
Curtis
email: curtis...@remove-this-part-if-you-are-not-a-spammer.shaw.wave.ca
Michael Black
little. They were kind of expensive back when the idea of hooking them
to a computer was useful because they were still cheaper than printers,
and sue had better print quality.
But the idea of hooking a Selectric to a computer was fairly generic.
Don Lancaster had a small section, apparently theoretical, on the topic in
his "TV Teletype Cookbook". In the early Bytes and Kilobauds there were a
number of articles about doing it yourself. I seem to recall a similar
article in Radio Electronics. There were a number of commercial
implementations of the idea, though I don't hve any specific names. The
point being that that if you are looking for the idea, you don't have to
just look for a certain manufacturer.
I don't remember the one with solenoids over the keys, but I can imagine
that there was such a solution. The Selectrics were expensive, and in some
cases rented, so a unit that fit over the keyboard and didn't modify the
typewriter would have a certain appeal. Who would have thought back then
that there were better things coming, and Selectrics would one day be
found at garage sales for virtually nothing, or even simply found in the
garbage?
If you can find one of the articles, you could simply build an interface
yourself.
Michael
cjt&trefoil
>
> Does anybody remember a kit that came out by a company (can't
remember the
> name) in the circa '78-'81 era that allowed you to convert an IBM
Selectric
> typewriter into a printer? I even remember an article in one of the
> hobbyist magazines at the time that described how to build such a
device
> that would do the ASCII to EBCDIC translation.
I vaguely remember that.
>I think there were two
> flavors of this idea - one was a box with a bunch of solenoids that
would
> actually sit on top of the keyboard of the typewriter and mimic a
real
> person pressing the keys (can you imagine such a thing?). The
second flavor
> was I believe for IBM's own modified Selectric typewriter that they
used in
> banks and such as hardcopy terminals. I don't know the model number
of this
> unit but it did have some sort of interface, parallel or serial,
that spoke
> EBCDIC.
I don't know the model number of those, but I know I spent one heck of
a lot of hours sitting at one in the early 70's, because it was the
only terminal we had that could access timesharing. They were slow
even then, and the print quality wasn't as good as on our secretarial
typewriters (the terminal version was built into a desk, BTW).
>
> The reason I ask these questions is that I am interested in
acquiring a unit
> of the latter flavor for purely hobbyist use to add to my collection.
>
> Would anybody know anything about these units and where they could
still be
> found (certainly not in active service but perhaps at surplus shops
and the
> like)?
>
I haven't seen one in many years. I think they disappeared around the
same time as did keypunches.
> Thanks,
>
> Curtis
>
> email:
curtis...@remove-this-part-if-you-are-not-a-spammer.shaw.wave.ca
JMHO.
kirto
On Wed, 2 Sep 1998, Curtis Rempel wrote:
> Does anybody remember a kit that came out by a company (can't remember the
> name) in the circa '78-'81 era that allowed you to convert an IBM Selectric
> typewriter into a printer? I even remember an article in one of the
> hobbyist magazines at the time that described how to build such a device
> that would do the ASCII to EBCDIC translation. I think there were two
> flavors of this idea - one was a box with a bunch of solenoids that would
> actually sit on top of the keyboard of the typewriter and mimic a real
> person pressing the keys (can you imagine such a thing?). The second flavor
> was I believe for IBM's own modified Selectric typewriter that they used in
> banks and such as hardcopy terminals. I don't know the model number of this
> unit but it did have some sort of interface, parallel or serial, that spoke
> EBCDIC.
Three flavors known to me:
Wang Undercarriage--late 60's--output only
An add-on to standard IBM Selectrics that fit beneath the typewriter and
connected to the bars normally actuated by the keys. Sold as an output
writer for the Wang 370 and 380 calculators. This is the Wang in Lowell,
Mass, formerly of Tewksbury, MA but I doubt that anyone there remembers
these devices.
Keypusher--early 80's--output only
External sets of solenoids, one for each key, that sit atop the keyboard
and operate the keys. There were 2 or 3 suppliers, at least one of which
was Japanese. Code translation happens immediately as the signal from the
source (8-bit ASCII) connects to the address lines of a ROM, the data
lines of which provide the drive signal for the keypusher solenoid
electronics.
Commercial terminals--???--Input and Output
Made by GE, I think, possibly called terminet. Widely used by airlines and
car rental companies. Often painted red.
--Kirt
Joe Morris
>Does anybody remember a kit that came out by a company (can't remember the
>name) in the circa '78-'81 era that allowed you to convert an IBM Selectric
>typewriter into a printer? I even remember an article in one of the
>hobbyist magazines at the time that described how to build such a device
>that would do the ASCII to EBCDIC translation. I think there were two
>flavors of this idea - one was a box with a bunch of solenoids that would
>actually sit on top of the keyboard of the typewriter and mimic a real
>person pressing the keys (can you imagine such a thing?).
Sure. I used one back in the mid 1960s to generate output from a nuclear
analyzer; it was a small unit with solenoids for the ten digits,
period, tab, and carriage return, and was designed to fit over the
keyboard of a narrow-carriage non-correcting office Selectric. For
its time this gave me a way to get clean printed output (which was
then hand-plotted -- sigh) without having to buy a budget-breaking
dedicated printer...and the typewriter could be used *as* a typewriter
by unhooking the solenoid box, an action that took a couple of seconds
at the most.
Everyone referred to the box as the "Magic Fingers" but I can't recall
whether that was a brand name or merely an informal generic term.
> The second flavor
>was I believe for IBM's own modified Selectric typewriter that they used in
>banks and such as hardcopy terminals. I don't know the model number of this
>unit but it did have some sort of interface, parallel or serial, that spoke
>EBCDIC.
The boxes shipped by IBM as data output devices were designed for that
purpose and weren't modified office units. The internal guts of the
mechanism were the same -- if one broke the typewriter repairmen were
often better than the CEs at repairing them -- but the frames were built
at the factory with the solenoids and other stuff (such as the keyboard
latch that locked the keyboard until the interface or host were ready
to accept data) in place. Several vendors, however, took standard office
units and kludged up the linkages required to provide terminal support.
The mechanical unit supported neither ASCII nor EBCDIC; it expected its
own encoding pattern which must include shift orders as distinct
characters in the data stream, and required some type of external
electronics to translate data to and from its selector patterns. (If
you have access to a copy, you can find discussions of the data streams
used in documentation of the IBM 2740 or 2741 remote terminals.) I
don't recall ever seeing a Selectric-based terminal that used ASCII
or EBCDIC, but that doesn't mean that there weren't any.
Joe Morris
George R. Gonzalez
Joe Morris wrote in message <6smcta$8...@top.mitre.org>...
>"Curtis Rempel" <curtis...@shaw.wave.ca> writes:
>
>>Does anybody remember a kit that came out by a company (can't remember
the
>>name) in the circa '78-'81 era that allowed you to convert an IBM
Selectric
>>typewriter into a printer
There was a device advertised in those early days. It seems unlikely it
would have a solenoid for each keytop, but you never know.
I did once hook up a Selectric that had IBM official solenoids driving
the mechanism. This was used as a console printer on the IBM 1130 IIRC,
and also sold as an APL terminal (2746?).
We eventually got it working with a PDP/8 thru a parallel interface.
It was just about the most MISERABLE, KLUDGY device you could imagine.
The
tech manual was quite explicit about its many failings. You had to
actuate a solenoid to set/clear the shift key. But if it was already
shifted, nothing would happen, and you wouldnt get the "operation
complete" handshake signal. So you had to first check another signal
which told you if the device was already in shift-mode. But this signal
was delayed a bit, so you couldnt rely on it to be correct. You also
had to software debounce and time-out several other signals that were not
too reliable. Repeat this kind of classy kludge for about 6 more
signals.
And oh yes, it didnt accept ASCII or EBCDIC. You had to translate each
character into a bunch of "rotate ball left", "tilt up", "tilt up HARD",
"rotate hard", signals.
We did eventually get it to print. But after the thrill of watching
letter-quality printing come out at 13.4 characters per second wore off
(at about page 3), nobody used that printer. Part of the problem was we
did not have anything near a full ASCII character set. The typical IBM
printballs of the time had a very Selectric-oriented character set. IIRC
there even wasnt a separate digit "1", you had to print a lower-case
letter L. Reminds me of some other typewriter where you had to
overstrike the apostrophe and period to get an exclamation mark!
The device driver code wouldnt fit into a standard device driver code
page or even two, so it had to be driven by a stand-alone program. DEC
assumed you'd want to drive some halfway rational device that could be
controlled by 128 or at most 256 instructions.
Charlie Gibbs
(Joe Morris) writes:
>The mechanical unit supported neither ASCII nor EBCDIC; it expected
>its own encoding pattern which must include shift orders as distinct
>characters in the data stream, and required some type of external
>electronics to translate data to and from its selector patterns. (If
>you have access to a copy, you can find discussions of the data streams
>used in documentation of the IBM 2740 or 2741 remote terminals.) I
>don't recall ever seeing a Selectric-based terminal that used ASCII
>or EBCDIC, but that doesn't mean that there weren't any.
The 2741-based machines I had direct contact with had paper tape
punches (and possibly readers). They used "correspondence code",
which had no relationship to either ASCII or EBCDIC. Being the
local paper tape wizard at the time, it fell to me to write the
programs to read the tapes produced by these machines. I remember
being totally grossed out when I figured out that the punches in
the paper tape were a bit-for-bit encoding of the tilt and rotate
codes for the typeball. (But hey, that's what the TRanslate
instruction is for...)
--
cgi...@sky.bus.com (Charlie Gibbs)
Remove the first period after the "at" sign to reply.
Joe Morris
I don't recall seeing a 2741 with paper tape devices attached; that
would be an oddity since there was *very* little circuitry in the
box. A 2740 would be only slightly less improbable; might you be
thinking of the 1050 terminal system, which did have support for
devices such as paper tape?
(Of course, it's not impossible that someone grafted a paper tape
punch onto a 2741. I've seen odder lashups, such as a communications
adapter grafted onto an IBM 024 keypunch...)
"Correspondence" was one of two standard configurations of the
keyboard; the other was "EBCDIC"; the primary difference was
the set of shifted characters on the top row of the keyboard. Both
of them required translations from EBCDIC (or ASCII if you were trying
to run them from a non-IBM system) into the PTTC tilt/rotate codes,
with appropriate insertion (transmit) or recognition (receive) of
shift codes.
A kludge? Of course, but at a time when the "standard" communications
device was the TTY33 with smudged uppercase letters on yellow newsprint
at all of 10 characters per second, the Selectric-based terminals like the
2740 and 2741 provided letter-quality output at the blazing speed
of just under 15 cps.
Joe Morris
Curtis Rempel
certainly has generated some interesting discussion! My memory has been
jarred somewhat from a few postings and I seem to remember that it was in
fact not ASCII or EBCDIC but rather BCD that these things used. I think I
may still have one of those early copies of whichever magazine published a
do-it-yourself kit sitting in a box in the garage somewhere. After my
basement is finished (where I'll be setting up some of my old equipment), I
think I'll do some "excavation" for that particular article in mind. If I
find it, I'll be sure to post a followup.
Regards,
Curtis Rempel
Joe Morris wrote in message <6sn3t6$l...@top.mitre.org>...
Carl R. Friend
>
> [The Selectric arrangement a] kludge? Of course, but at a time when
> the "standard" communications device was the TTY33 with smudged
> uppercase letters on yellow newsprint at all of 10 characters per
> second, the Selectric-based terminals like the 2740 and 2741 provided
> letter-quality output at the blazing speed of just under 15 cps.
If I'm recalling things correctly here (which may be questionable
at best), the Flexowriter (from Friden, pre-dating the Selectric by
many years) was capable of letter-quality print at around 15 cps on
a good day. This, in the late '50s.
Flexowriters were commonly used on early computers; both the PDP-1
and pb250 (as examples) used them. They, too, were parallel devices.
The difference was that the Flexowriter was designed from the outset
as an "office automation" device whereas the Selectric was, first and
foremost, a typewriter.
Bizarre bit, the Flexo, though. It's definitely more of a type-
writer than a terminal (look inside one)....
--
______________________________________________________________________
| | |
| Carl Richard Friend (UNIX Sysadmin) | West Boylston |
| Minicomputer Collector / Enthusiast | Massachusetts, USA |
| mailto:carl....@stoneweb.com | |
| http://www.ultranet.com/~engelbrt/carl/museum | ICBM: N42:22 W71:47 |
|________________________________________________|_____________________|
danny burstein
but would print out form letters with mix-and-match addresses.
It was desk sized. No, not desk-top, but the size of an office desk. Or
possibly two in length... (I was smaller then...)
Afraid I no longer have any recollection as to who made it or how it
worked but I'd be curious if this jogs anyone's memory.
danny 'i'd hate to tell you what types of letters it would send out to my
family' burstein
[material below left for reference]
--
_____________________________________________________
Knowledge may be power, but communications is the key
dan...@panix.com
[to foil spammers, my address has been double rot-13 encoded]
Charlie Gibbs
(Joe Morris) writes:
>I don't recall seeing a 2741 with paper tape devices attached; that
>would be an oddity since there was *very* little circuitry in the
>box. A 2740 would be only slightly less improbable; might you be
>thinking of the 1050 terminal system, which did have support for
>devices such as paper tape?
I guess I shouldn't have said "2741" - curse these failing memory
cells. But it was definitely a Selectric mechanism in a beige
case that was extended off to the side for the paper tape unit.
The 1050 was a modular system, wasn't it? This was all in one
box. Maybe we could call it an ASR Selectric. Or maybe I'll
roast in hell for even thinking such a thing.
William Hamblen
wrote:
>"Correspondence" was one of two standard configurations of the
>keyboard; the other was "EBCDIC"; the primary difference was
>the set of shifted characters on the top row of the keyboard. Both
>of them required translations from EBCDIC (or ASCII if you were trying
>to run them from a non-IBM system) into the PTTC tilt/rotate codes,
>with appropriate insertion (transmit) or recognition (receive) of
>shift codes.
Don Lancaster in his TV Typewriter Cookbook called them "bail codes"
which might have been the official name.
Twenty years ago you used to be able to get surplus paper tape word
processors based on a Selectric mechanism. I made an interface to
connect one of these machines to an SWTPC 6800 parallel port and it
worked pretty well for a while until one of the tapes that wiggled the
type ball broke. I replaced that machine with a Digital Group dot
matrix printer with the print head driven directly from the parallel
port.
You could get an ASCII character set type ball from IBM for the
Selectric.
Dennis Ritchie
IBM 2741 and 1050, though certainly not a list of
possible variants. Particularly with the 1050,
it was obvious that options were possible.
Memories: both were indeed based on the Selectric
mechanism, and both were mounted in large consoles.
The 1050 in particular was in an L-shaped desk-height
cabinet, with the electronics in the part that stuck
out to your right. I had one of these at home for
a couple of years and it took a couple of strong
movers to get it in (and out) of the basement.
The 2741 was smaller, its desk was rectangular
and the electronics were less obtrusive.
Both terminals were determinedly half-duplex.
2741s certainly needed to be ordered with the
"reverse break" option that made it possible for
the computer to listen for a "break" signal from the
user; otherwise, if you started something that
printed endlessly, there was no way to stop it
short of hanging up.
These were used to connect to Multics and also to
early Unix up to about 1972. In Multics and Unix,
we used the 938 type-ball, which mapped best (not
perfectly) to ASCII; I think this was called the
BCD encoding. As other point out, other type-balls
had the characters in completely different places:
the Correspondence encoding.
For whatever strange reason, the only fonts available
on the BCD balls were fairly ugly, and the ones
in the Correspondence coding were nice for writing
letters. Thus Unix at this date supported the terminals
for ordinary bidirectional I/O with the BCD encoding only,
but had a terminal mode that would (on output) translate
to Correspondence code.
One of the few bits of our detritus from this period that
survives in machine-readable form is a fragment of
the terminal driver of 1972. So I have the tables that map
between the characters received from the serial device
into ASCII, and have attached them. The first table
is the ASCII value from looking up the 6-bit value
from the 2741 keyboard in lower-case, the second
in upper-case. Not shown here are the codes that shifted
between cases and other control-type things.
(As a side note, these terminals had the virtue
of making their "new-line" code the ordinary way of
getting to a new line, just like typewriters. They
did have an "index" function to advance the paper without
moving to the margin.)
Dennis
---
inl938:
.byte ' , '-, '@, '&, '8, 'q, 'y, 'h
.byte '4, 'm, 'u, 'd, 00, 00, 00, 00
.byte '2, 'k, 's, 'b, '0, '?, '?, '?
.byte '6, 'o, 'w, 'f, 00, 10, 00, 00
.byte '1, 'j, '/, 'a, '9, 'r, 'z, 'i
.byte '5, 'n, 'v, 'e, 00, 12, 12, 11
.byte '3, 'l, 't, 'c, '#, '$, ',, '.
.byte '7, 'p, 'x, 'g, 00, 00, 00, 00
inh938:
.byte ' , '_, '\\, '+, '*, 'Q, 'Y, 'H
.byte ':, 'M, 'U, 'D, 00, 00, 00, 00
.byte '<, 'K, 'S, 'B, '), '?, '?, '?
.byte '', 'O, 'W, 'F, 00, 10, 00, 00
.byte '=, 'J, '?, 'A, '(, 'R, 'Z, 'I
.byte '%, 'N, 'V, 'E, 00, 12, 12, 11
.byte ';, 'L, 'T, 'C, '", '!, '|, '^
.byte '>, 'P, 'X, 'G, 00, 00, 00, 00
Scott Peterson
>My elementary school back in 1964 had a thingee which was big and huge,
>but would print out form letters with mix-and-match addresses.
>
>It was desk sized. No, not desk-top, but the size of an office desk. Or
>possibly two in length... (I was smaller then...)
>
>Afraid I no longer have any recollection as to who made it or how it
>worked but I'd be curious if this jogs anyone's memory.
>
I remember seeing pictures of an IBM Office Products Selectric based
system like this. It used magnetic media about the size and shape of
the 80-column card to store things.
Scott Peterson
jmfb...@aol.com
"George R. Gonzalez" <g...@foundsys.com> wrote:
>
>Joe Morris wrote in message <6smcta$8...@top.mitre.org>...
>>"Curtis Rempel" <curtis...@shaw.wave.ca> writes:
>>
>>>Does anybody remember a kit that came out by a company (can't remember
>the
>>>name) in the circa '78-'81 era that allowed you to convert an IBM
>Selectric
>>>typewriter into a printer
>
>There was a device advertised in those early days. It seems unlikely it
>would have a solenoid for each keytop, but you never know.
>
>I did once hook up a Selectric that had IBM official solenoids driving
>the mechanism. This was used as a console printer on the IBM 1130 IIRC,
>and also sold as an APL terminal (2746?).
>We eventually got it working with a PDP/8 thru a parallel interface.
>
>It was just about the most MISERABLE, KLUDGY device you could imagine.
Yup. But it had the best print quality at the time I used it (1971-2).
>The
>tech manual was quite explicit about its many failings. You had to
>actuate a solenoid to set/clear the shift key. But if it was already
>shifted, nothing would happen, and you wouldnt get the "operation
>complete" handshake signal. So you had to first check another signal
>which told you if the device was already in shift-mode. But this signal
>was delayed a bit, so you couldnt rely on it to be correct. You also
>had to software debounce and time-out several other signals that were not
>too reliable. Repeat this kind of classy kludge for about 6 more
>signals.
>
>And oh yes, it didnt accept ASCII or EBCDIC. You had to translate each
>character into a bunch of "rotate ball left", "tilt up", "tilt up HARD",
>"rotate hard", signals.
>
>We did eventually get it to print. But after the thrill of watching
>letter-quality printing come out at 13.4 characters per second wore off
>(at about page 3), nobody used that printer. Part of the problem was we
>did not have anything near a full ASCII character set. The typical IBM
>printballs of the time had a very Selectric-oriented character set. IIRC
>there even wasnt a separate digit "1", you had to print a lower-case
>letter L. Reminds me of some other typewriter where you had to
>overstrike the apostrophe and period to get an exclamation mark!
We had printballs made special. In addition to distinguishing the
digit 1 from the letter l, we also added square brackets, backarrow,
maybe some others. I do remember having to decide what characters
I was willing to give up (but I don't remember which ones I did).
>
>The device driver code wouldnt fit into a standard device driver code
>page or even two, so it had to be driven by a stand-alone program. DEC
>assumed you'd want to drive some halfway rational device that could be
>controlled by 128 or at most 256 instructions.
We ran that "device driver interface" software on a PDP-10 running
TOPS10. It was one of the important steps I used to get all of DEC's
documentation on-line. Print quality was a huge issue when
producing galleys suitable for bookmaking.
/BAH
Sigh! - Subtract a hundred and four for e-mail.
Tom Van Vleck
> These were used to connect to Multics and also to
> early Unix up to about 1972. In Multics and Unix,
> we used the 938 type-ball, which mapped best (not
> perfectly) to ASCII; I think this was called the
> BCD encoding. As other point out, other type-balls
> had the characters in completely different places:
> the Correspondence encoding.
The correspondence ball was the 963. Multics had a few
commands that you could issue when not logged in: two of
these were "963" and "938". The idea was that if you
dialed up to the system and it typed a garbage greeting,
you would type the 963 command (numbers were the same
encoding on both typeballs) to switch mappings. I seem to
remember that if the user issued the "kigub" command, we
siwtched mappings and then did a login.. some strange word
like that anyways. The system also typed a special line
at dialup that read "(garbage) type 963 before logging in"
or "type 938 before logging in (garbage)" depending on
which typeball was mounted.
I had a 2741 at home for a while too. See
http://www.best.com/~thvv/terminals.html
for my recollections of 30 years of logging in from home.
Sam Yorko
A Mag-Card Typewriter, and if anybody has access to one, I'd love to
print out another copy of my senior project (done in 1980).
Sam
Isaac Wingfield
Morris) wrote:
--snippage--
> The mechanical unit supported neither ASCII nor EBCDIC; it expected its
> own encoding pattern which must include shift orders as distinct
> characters in the data stream, and required some type of external
> electronics to translate data to and from its selector patterns. (If
> you have access to a copy, you can find discussions of the data streams
> used in documentation of the IBM 2740 or 2741 remote terminals.) I
> don't recall ever seeing a Selectric-based terminal that used ASCII
> or EBCDIC, but that doesn't mean that there weren't any.
Well, there was the one I built for my S-100 system. I bought a
Selectric-based terminal that must've had several thousands of miles of
paper run through it: the nickel-plate was worn off the letters on the ball
and there was a *crater* at every impact location on the platen.
After an interesting (and *not* expensive) foray into IBM's replacement
parts facility in LA, the thing was working, but a pain-in-the-butt to
drive. I wrote a software ASCII-to-tilt level-rotate translator/driver for
it, but was not happy with it (the software-based timing didn't always work
and it made printing errors), so I built a small state machine to do the
same thing.
I don't remember the details, but it took serial ASCII as input and used an
EPROM-based sequencer to drive the mechanism. The whole thing was only
about three by four inches, and fit inside the printer's case. An
interesting complication is that the selectric mechanism is not
synchronous; different characters take differing amounts of time to print.
I used the printer for years; even after I got a dot-matrix jobber for
graphics, the selectric was my "prettyprinter".
The printing quality was excellent, especially with a carbon ribbon. The
keyboard had a completely mechanical two-key rollover, and was totally
impossible to "jam"; you simply could not start a print op until the
previous one had gotten out of the way. Driven from a 'puter, it was,
however, possible to overlap successive print ops, and this was a very good
idea, as it avoided wear and tear on the main shaft's clutch.
The selectric may well be the *last, great* mechanical device ever
designed. In the typewriter version, the only electrical part in it was the
motor. NO solenoids, switches (except power switch), or other similar
things. Terminal versions, of course, had solenoids for ball control and
switches for status. Tilt and rotate were not binary; AFAIR they were
1-2-2-5 or something similar, and involved negative as well as positive
tilts and rotates.
Isaac
Joe Morris
>The 1050 was a modular system, wasn't it? This was all in one
>box. Maybe we could call it an ASR Selectric. Or maybe I'll
>roast in hell for even thinking such a thing.
Or maybe you're thinking of the CMCST (Communicating Magnetic Card
Selectric Typewriter) which was effectively a 2741 with an attached
magnetic card reader? I don't recall the machine type for the
CMCST, but it wasn't considered to be a member of the 2741 family
even though it could be used as one.
Or the MTST (Magnetic Tape Selectric Typewriter)?
Joe Morris
Danny Lingman
If anyone wants it, it's theirs for the cost of shipping (or they can
come to Ottawa to pick it up)
Dan.
Douglas W. Jones,201H MLH,3193350740,3193382879
> My elementary school back in 1964 had a thingee which was big and huge,
> but would print out form letters with mix-and-match addresses.
>
> It was desk sized. No, not desk-top, but the size of an office desk. Or
> possibly two in length... (I was smaller then...)
Sounds like a Flexowriter, with auxiliaries.
The basic Flexowriter was a large heavy electric typewriter with paper
tape reader punch attached on one side. Built like a tank, and heavy
enough that two people should be on-hand when one is to be lifted.
It came with a desk that contained paper-tape dispensor and take-up
reels, plus a bit of power distribution wiring, and there was a pedestal
mounted relay box with a second paper-tape reader.
If you wanted to do the "mailmerge" function to make form letters, you
needed both paper-tape readers. Use one for the form letter, mounted
as a continuous loop of tape, and use the other for the mailing list. Codes
punched on the tapes would switch the read function back and forth between
the tapes to merge the mailing list into the text of the form letters.
The thing worked very well. I have one, minus the desk, but with the
auxiliary reader. I have yet to debug it; I suspect that some of the
capacitors have shorted and burnt out some of the selenium rectifier
stacks.
All active components were, of course, relays. There are lots of them!
Doug Jones
jo...@cs.uiowa.edu
Lisa or Jeff
word processor, then switched the to Mag Card. A very popular machine
in the 1970s.
The Selectric also had a composing machine for typesetting.
Tom Van Vleck
>The 1050 was a modular system, wasn't it? This was all in one
>box. Maybe we could call it an ASR Selectric. Or maybe I'll
>roast in hell for even thinking such a thing.
The 1050 was as Dennis describes, an L-shaped desk with builtin
Selectric; but you could get optional units on their own casters
that plugged into the electronics box. The one we had on the 8th
floor at Project MAC had a card reader, and I think I remember
seeing brochures for a paper tape reader. On the vertical face
of the typewriter, above the keys, there was a long row of
toggle switches, and room to install more. The switches let you
switch input and output, ignore EOT codes, all kinds of stuff.
It was clear that the device was extremely configurable and that
IBM expected that it would be used in many different environments.
The 1050 seemed more like a Flexowriter than an ASR to me.
Matt Ackeret
>>did not have anything near a full ASCII character set. The typical IBM
>>printballs of the time had a very Selectric-oriented character set. IIRC
>>there even wasnt a separate digit "1", you had to print a lower-case
>>letter L.
Well, the "Introduction to your Apple //e" program (named something like
that, it's in my pile of 5.25" floppies somewhere) definitely instructs people
not to use a lower case l instead of a 1. So I guess that was pretty
ingrained in people.
--
mat...@area.com
D. Peschel
>Well, the "Introduction to your Apple //e" program (named something like
>that, it's in my pile of 5.25" floppies somewhere) definitely instructs people
>not to use a lower case l instead of a 1. So I guess that was pretty
>ingrained in people.
I suppose the manual also made the same point. It may even be in Macintosh
manuals -- I'm not sure.
Now, DEC took the opposite approach -- their EVE text editor for VMS will
recognize 'l' as a digit.
-- Derek
William Hamblen
>Well, the "Introduction to your Apple //e" program (named something like
>that, it's in my pile of 5.25" floppies somewhere) definitely instructs people
>not to use a lower case l instead of a 1. So I guess that was pretty
>ingrained in people.
Manual typewriters did without the numeric one and you struck the
lower case ell instead. My old Royal portable is built that way.
Typing reflexes definitely were to hit the lower case ell for a one.
You could get Selectrics with numeric one keys.
Edward Rice
Sam Yorko <JOATno...@compuserve.com> wrote:
> > I remember seeing pictures of an IBM Office Products Selectric based
> > system like this. It used magnetic media about the size and shape of
> > the 80-column card to store things.
> >
> > Scott Peterson
>
> A Mag-Card Typewriter, and if anybody has access to one, I'd love to
> print out another copy of my senior project (done in 1980).
Minor correction: it was an MCST, a Mag Card Selectric Typewriter. IBM
also had the earlier (and more successful, I believe) MTST, Magnetic Tape
Selectric Typewriter.
Nick Spalding
> Minor correction: it was an MCST, a Mag Card Selectric Typewriter. IBM
> also had the earlier (and more successful, I believe) MTST, Magnetic Tape
> Selectric Typewriter.
Was it not with the MTST that the phrase Word Processing entered the
language?
Dave Garland
>dan...@panix.com (danny burstein) wrote:
>
>>It was desk sized. No, not desk-top, but the size of an office desk. Or
>>possibly two in length... (I was smaller then...)
>>
>
>I remember seeing pictures of an IBM Office Products Selectric based
>system like this. It used magnetic media about the size and shape of
>the 80-column card to store things.
>
The IBM Mag Card wasn't that big. (Physically, it was a big Selectric
+ a desk-high box that was maybe 10 inches wide and 18 inches deep.)
And, IIRC, came out maybe 1968. Was there a bigger predecessor?
-Dave
Ethan Dicks
>In article <6smr9l$a29$1...@news1.tc.umn.edu>,
> "George R. Gonzalez" <g...@foundsys.com> wrote:
>>
>>Joe Morris wrote in message <6smcta$8...@top.mitre.org>...
>>>"Curtis Rempel" <curtis...@shaw.wave.ca> writes:
>>>
>>>>Does anybody remember a kit that came out by a company (can't remember
>>the
>>>>name) in the circa '78-'81 era that allowed you to convert an IBM
>>Selectric
>>>>typewriter into a printer
>>
>>There was a device advertised in those early days. It seems unlikely it
>>would have a solenoid for each keytop, but you never know.
>>
There was such a device. I lusted after it in 1978. My mother was a
typist with a home office. We had two Selectrics at home. I wanted to
get one of those one-solonoid-per-key boards, but it was well over $100.
I would have driven it from my 32K PET.
ISTR that it was advertised in Creative Computing and/or BYTE, the two
magazines I was reading at the time.
-ethan
--
Ethan Dicks http://www.infinet.com/~erd/
(dicks) at (math) . (ohio-state) . (edu) sellto: postmaster@[127.0.0.1]
harvestbot fodder: pres...@whitehouse.gov fcc...@fcc.gov root@[127.0.0.1]
David L
"Selectronic" -- a big black box that was the interface, where you plugged
a serial line in, and a cable or two that plugged somewheres in the guts of
the Selectic itself amd drove it. I think we brought the selectric with it
and it was specially modified somewhat but I'm not certain of that -- this
was in 1978.
I worked fairly well actually, we had it hooked up to our Altair 8800a
and used it for word processing output!
jagua...@gmail.com
> Does anybody remember a kit that came out by a company (can't remember the
> name) in the circa '78-'81 era that allowed you to convert an IBM Selectric
> hobbyist magazines at the time that described how to build such a device
> that would do the ASCII to EBCDIC translation. I think there were two
> flavors of this idea - one was a box with a bunch of solenoids that would
> actually sit on top of the keyboard of the typewriter and mimic a real
> person pressing the keys (can you imagine such a thing?). The second flavor
> was I believe for IBM's own modified Selectric typewriter that they used in
> banks and such as hardcopy terminals. I don't know the model number of this
> unit but it did have some sort of interface, parallel or serial, that spoke
> EBCDIC.
>
> The reason I ask these questions is that I am interested in acquiring a unit
> of the latter flavor for purely hobbyist use to add to my collection.
>
> Would anybody know anything about these units and where they could still be
> found (certainly not in active service but perhaps at surplus shops and the
> like)?
>
> Thanks,
>
> Curtis
>
> email: curtis...@remove-this-part-if-you-are-not-a-spammer.shaw.wave.ca
jagua...@gmail.com
I was always interested in electronics but got into building homes in calvert county maryland As fate would have it my company contracted with a person whos business was repairing IBM Selectric typewriters We discussed his business and my interest in electronics well he said he used to get IBMs in from time to time with a metal bamd that seperated the bottom cover from the typewriter this band made allowed a metal frame with several solenoids that grabbed the different mechanical functions to cause the typewriter to function without pushing the keys He gave me several of these bands with several of the metal frames with the solenoids Suprising enuf there was an article in an electronics magazine oh how to convert an IBM seclectrc to a printer using the exzact metal solenoid frame assembly and band the electronic interface was easy to build using i beleieve the parallel port It alson included the code modification to your operating system at the time Which i was using CPM with a heathkit all in one computer My wifes sister was dating a guy going to college in pa and he used helped me with the insertion of the code into the CPM operating system
It worked great most of the time but still had some bug that would cause the CPM system to crash What i could remember it would print documents without any mistakes but sometimes would crash the operating system but reloading the operating system would allow you to do it again we never found out why
i dont not know what i did with these frames but it was very satisfying to see the selectric print
jagua...@gmail.com
jagua...@gmail.com
I remember the article and build the interface in conjunction to a metal frame with solenoids the mounted between the Selectric and bottom cover A one inch metal band allowed for the installation I was using CPM operating system with a code modification to CPM to get the Selectric to print email me and ill give you my cell number on more about this
David Wade
> On Wednesday, September 2, 1998 at 3:00:00 AM UTC-4, Curtis Rempel wrote:
>> Does anybody remember a kit that came out by a company (can't remember the
>> name) in the circa '78-'81 era that allowed you to convert an IBM Selectric
>> typewriter into a printer? I even remember an article in one of the
>> hobbyist magazines at the time that described how to build such a device
>> that would do the ASCII to EBCDIC translation. I think there were two
>> flavors of this idea - one was a box with a bunch of solenoids that would
>> actually sit on top of the keyboard of the typewriter and mimic a real
>> person pressing the keys (can you imagine such a thing?). The second flavor
>> was I believe for IBM's own modified Selectric typewriter that they used in
>> banks and such as hardcopy terminals. I don't know the model number of this
>> unit but it did have some sort of interface, parallel or serial, that spoke
>> EBCDIC.
model" was the 2741, but they came in other guises. Some were printer
only, others were proper terminals.
>>
>> The reason I ask these questions is that I am interested in acquiring a unit
>> of the latter flavor for purely hobbyist use to add to my collection.
>>
>> Would anybody know anything about these units and where they could still be
>> found (certainly not in active service but perhaps at surplus shops and the
>> like)?
>>
been one on E-Bay UK in the last two years. I just bought it.
>> Thanks,
>>
>> Curtis
>>
>
> jagua...@gmail.com
> I remember the article and build the interface in conjunction to a metal frame with solenoids the mounted between the Selectric and bottom cover A one inch metal band allowed for the installation I was using CPM operating system with a code modification to CPM to get the Selectric to print email me and ill give you my cell number on more about this
>
G4UGM
Peter Flass
gadget that sat on top of a selectric keyboard and pushed the keys to
generate printout. (see wikipedia article)
--
Pete
Jon Elson
>> The reason I ask these questions is that I am interested in acquiring a
>> unit of the latter flavor for purely hobbyist use to add to my
>> collection.
>>
>> Would anybody know anything about these units and where they could still
>> be found (certainly not in active service but perhaps at surplus shops
>> and the like)?
paper tape reader, an I/O Selectric, and an electrical interface. I didn't
have much use for the paper tape or the keyboard, but made up an interface
to use it as a printer. It was pretty noisy, but worked OK. It had a bunch
of discrete transistor logic, and I think it spoke EIA code (Pre-ASCII) and
converted that to the I/O selectric's required code. I don't believe it was
EBCDIC, but raw ball rotation and tilt codes.
Jon
Charlie Gibbs
I have a code card for it right in front of me. The tape was punched in
what was known as "correspondence code"; when I took a close look at it,
it was indeed just tilt and rotation codes, which I translated to EBCDIC.
--
/~\ cgi...@kltpzyxm.invalid (Charlie Gibbs)
\ / I'm really at ac.dekanfrus if you read it the right way.
X Top-posted messages will probably be ignored. See RFC1855.
/ \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign!
Quadibloc
> The idea is enticing these days, now that Selectrics can be had for so
> little.
It's less enticing when you find out that the reason they can be had for so
little nowadays is because you can't get the elements for them any more.
John Savard
Michael Black
That's funny. I saw the new reply to the old thread, and I was thinking
"did I post in that thread?". I couldn't remember, but it was the sort of
thing I would have. By 1998, it was all in the past, three years later I
had a $15 used laser printer that was way better than a Selectric. 18
years after the thread, why is the guy who resurrected this thread talking
about Selectrics? There were endless articles forty years ago, but that's
when letter quality printers were expensive and scarce.
Michael