Dr. James Marsters, TTY deaf service developer [telecom]

Excerpt from the New York Times:

Dr. James Marsters passed away. He and two deaf colleagues broke that barrier for themselves and tens of thousands of other hearing-impaired people in 1964 when they converted an old, bulky, clacking Teletype machine into a device that could relay a typewritten conversation through a telephone line. It was the first example of what became commonly known as a TTY and is now, in a greatly updated and compact version, called a text telephone.

When they introduced their device, the partners met strong resistance from AT&T, which then had virtual control over the nation?s telephone system and prohibited direct connections to its network.

For full article please see:

formatting link
On another newsgroupo several people said AT&T didn't offer any service the deaf community could use so Dr. Marsters developed his system.

I'm don't think that's correct. For many years both AT&T and Western Union offered switched teleprinter services that anyone could get (TWX and Telex). AT&T also would rent Teletypes for use on its voice network for people to call computer time sharing services.

Also, I believe AT&T would rent anyone a modem for connection to privately owned equipment. Normally that would be a computer, but there were other machines, such as process control and telemetry reporting, that were connected as well. The modem acted as the interface to protect the network. All machines of course were owned by the user, AT&T did not rent out any computers. So if someone had an old refurbished Teletype AT&T would rent them a modem. (Also, private line subscribers could use their own modems.)

What I think Dr. Marsters' developed was a _low-cost_ system for deaf people to use instead of renting a Bell System modem, teleprinter, or a Western Union Telex connection. Since AT&T's and WU's services were intended for busineses, the prices weren't cheap.

***** Moderator's Note *****

TWX and TELEX were, as you point out, too expensive for ordinary users, but there were other barriers to deaf/hearing-impaired users having access to the telephone network, and the inventors had to overcome all of them.

  1. TWX and TELEX machines worked on closed networks; nobody without a TWX or TELEX machine could communicate with them, so even those few Deaf/Hearing-impaired users who could afford to have access couldn't depend on being able to reach another Deaf/Hearing-impaired user.

  1. The only dial-up data services available for connections via the PSTN was Tymnet (Telenet wasn't established until 1974): it was geared to business users, and it didn't have any capability for placing calls to destinations on the PSTN or to manually-controlled nodes.

  1. The modems available for use on ordinary POTS lines were "Full-duplex" designs, which wouldn't operate unless they were constantly connected with the same type of modem. That meant that they wouldn't work with common telephone devices such as answering machines, and that they couldn't be used in situations where only one person on the call was Deaf or Hearing-impaired, which is a very common occurence, especially with "Late deafened" adults who have lost their hearing but are still able to speak intelligibly.

Marsters, Weitbrecht, and Saks solved these problems with a combination of innovation and elbow-grease:

  1. They chose to use the older Model 15/19 and Model 28 Teletype machines which ham operators were also using. These machines were nearing the end of their commercial life, and were relatively inexspensive.

  1. They designed and built a half-duplex modem which is compatible with telephone answering machines: it could be used to record an outgoing message which would be printed on a caller's machine, and also to leave a message for later printing by the recipient.

  1. They included automatic Transmit/Receive switching, which allowed for calls to be made where one party was using a TDD and the other was speaking, as in the case of late-deafened adults.

In a sense, the TDD was a "bridge" device, which made it possible for Deaf/Hearing-impaired users to make use of the PSTN during the time that the Internet was developing. Since it's now routine for most major companies to have a web presence, and for them to offer customer service and bill payments online, there's less of a need for a device which will operate on the PSTN, but this is a (surprisingly sharp) double-edged sword.

  1. Many Deaf/Hearing-impaired users depend on the TDD as their only means of access to government services such as the registry of motor vehicles or their local tax assessor, but at the same time that these organizations have adapted to the Internet, they have tended to step back from TDD services and upkeep.

  1. Dependence on Email and web-sites for communications between Deaf/Hearing-impaired users has reduced demand for TDD devices and training, which has pushed prices up and re-introduced the problem of not knowing what system a given user has available.

  2. The TDD gave Deaf/Hearing-impaired citizens a small compensation for the marginalizatoin they endured in many aspects of their lives. Since TDD devices were not designed to communicate with computers, government agencies and utilities (which were mandated to offerr TDD lines) did, as a result, sometimes give a higher standard of customer service to TDD users than to the hearing public. TDD users could usually get through to a government agency much more quickly than those using voice, and they were also spared the dubious benefits of voice-response systems.

Dr. Marsters and his team changed the world for deaf and hearin-impaired telephone users. I'm sad to see him go. Bill Horne

Reply to
hancock4
Loading thread data ...

Yes, a "service" existed. It was *NOT* 'practical' for the deaf community to use it. The reason? The usual one, money.

*snicker* Got any idea how much that _cost_, relative to a standard 'voice grade' phone line? And, of course the Teletypes/modems that AT&T provided for time-share service dial-up, did "originate" mode _only_. you could _not_ directly connect two such machines over the 'voice' network'. That meant a deaf person _could_not_ call another deaf person directly with such a device. The two people had to 'co-ordinate' their calls, so that they both called 'some other place', that had 'answer' modems, and could shuffle the bits back and forth between them.
*NOT* entirely true. Lots of places had limited dial-in- capability direct to _their_ mainframes. Tymenet was unique in that it was a 'national' front end to anybody who got an X.25 'server' connection from them. "anyone" could dial the 'nearby' Tymenet number, and _then_ decide which host system to connect to. Beat the h*ll out of having to make a long-distance call to the city where the particular 'host' was located. It _also_ got you whatever bandwidth you could get on that 'nearby' call -- generally =far= better than one could get going through the voice LD network.

What Tymenet, etc., did was really the 'precursor' of the communications "revolution". The only element in the 'variable cost' of a customer "'net" connection was the amount of data they transmitted/received. *DISTANCE* was not a cost item. The server across the country (or in _another_ country) didn't cost more to access than the one next door -- as far as the 'communication' cost went, that is. What the server operator charged, how much data you passed, and _how_fast_ you passed it, _did_ all factor into your total bill.

***** Moderator's Note *****

You're right; I should have said "it didn't have the capability for placing calls to manually-controlled destinations on the PSTN".

Bill Horne

Reply to
Robert Bonomi

Thanks for your comments. There are several things I don't understand.

Yes, TWX and Telex operated on closed networks and they were expensive. But if they weren't expensive, why couldn't the deaf community simply become an active part of that network? In addition to being able to communicate with each other, they could also communicate to the many businesses that were on-line. A deaf user still had to have a teleprinter and modem no matter what carried the call.

I'm confused about this. By 1968 and probably earlier (1965?) computer time sharing services were available through dial up terminals on the PSTN. Anyone who had a terminal (usually a Teletype

33 or 35 rented from Bell, admittedly not cheap) could _also_ call any another person with a teletype at regular PSTN rates, which for a residential local call was often free and untimed. So here, too, aside from price, the deaf community get service. (Back in school sometimes we called other schools and conversed via the Teletypes, Early from of teen texting.)

There was mention that the TDD service used "simplex" instead of half or full duplex. But again I don't understand. Presumably all modems out there were full or half duplex, and on the PSTN, 110 baud (Teletype 33/35 speed). I don't think there were many answering machines back in the 1960s.

Undoubtedly far cheaper than renting a Bell machine. But were there problems with maintenance? The mechanical guts of a teleprinter were complex and needed periodic maintenance. (On a teletype, when a key was depressed it lines up various notched levers which determine the bit pattern to be sent. On receipt, electro-magnets line up the levels per the bit pattern and the proper key level drops down. All this had to work precisely otherwise the bit pattern would be out of sync with the proper lever, each bit being only 1/50 of a second long).

So you're saying the modem "handshaking" didn't require something on the other end; the originating modem would transmit without a return carrier? And answering machines could record and play back modem tones?

The Teletype 33 ASR (automatic send receive) could automatically answer and allow the other party to leave a printed message. Earlier models might have had that capability. Indeed, on Telex and TWX that was the common practice. But they were costly.

I don't understand. Did the party who could hear be able to decipher TDD tones?

One of my co-workers was very active in the deaf community and demonstrated his TDD device to me, and he was very familiar with all of this. It used an acoustic coupler so it could be used from any telephone and was very compact, displaying messages on an LED strip. Unfortunately, he passed away suddenly, less than a month before his retirement.

Thanks again for the explantion.

[snip]

***** Moderator's Note *****

I'll answer your questions in a separate post.

Bill Horne

Reply to
hancock4

We don't get to say "If they weren't expensive": anything that comes after that supposition is just speculation. My mother used to say "If wishes were horses, beggars would ride".

Well, aside from price, I'd be visiting the Mare Imbrium on my next vacation.

Computer time-sharing services might have been available (I'll defer to other readers on the timeline for that: I worked as a Teletype repairman for an educational time sharing system in 1976, although Ward Christensen didn't start CBBS until 1978), but even if they were, their price limited them to business users, and they weren't available for free to anyone but students, and only during school hours. You seem to be saying that these were viable alternatives for every deaf and hearing-impaired person who had a large trust fund sufficient to deploy Model 33's at every location they might want to communicate with. That's not a reflection of the deaf & hearing/speech-impaired community.

Television and radio use simplex transmission: it's a one-way-only system where the listeners and viewers can only receive, and the radio and tv stations can only send. The only simplex telephone modems I know of were one-way devices used for telemetry.

TDD's use half-duplex transmission: either end can send, but not at the same time. It's the same paradigm as a CB radio, i.e., "Push to talk" - only one guy gets to talk at a time.

The PSTN doesn't care what speed asynchronous modems run at (within the limits of the channel bandwidth, of course), since the speed is determined by the devices which are attached to the modems. TWX machines ran at 110 Baud, but that speed was determined by the Model

33 or 35 Teletype machines in use for the system. I used to own an Anderson-Jacobson 831 terminal, and it connected to the PSTN with an accoustic coupler and worked at 134.5 Baud.

Presumably, all modems in common use were full or half-duplex devices, but the only half-duplex models I'm aware of (The Bell 202 series) were made for synchronous transmission of data between controllers on IBM or compatible SNA networks, and they weren't available to the general public.

There may not have been many answering machines available back in the

60's, but they _were_ available, and Dr. Marsters (who was, don't forget, a dentist who needed to keep in touch with his office) knew that deaf and hearing/speech-impaired people had to be able to use them with a TDD: after all, if you're being called by a deaf or hearing/speech-impaired person, how is someone going to take a message?

Actually, they were a lot more reliable than the new models that replaced them: the Model 32 Teletype was only rated for 1500 shaft hours before major overhaul. OTOH, the Model 15 RO was the standard terminal for news reports in both the AP and UPI networks well into the 70's.

The mechanisms were complex, but maintenance needs were minimal: TDD use, even in high-traffic settings like Gallaudet University, wasn't anywhere near the 24/7 duty cycle that the Models 15/19 and 28 were designed for, and routine lubrication could be handled by end users.

Do tell ;-).

For 60 speed machines, it's 1/45th of a second, i.e., 22 ms per bit, but I digress.

Exactly.

As I said, costly meant "not viable". End of story.

Um, no, the party who could hear would listen to the phone while the deaf/hearing-impaired person on the other end was talking.

I sometimes use the same sort of machine, but the accoustic cups are designed for the round earpiece and microphone found on "500" sets, so they don't work as well with trimline or similarly shaped phones. I prefer a direct connect model.

Sorry to hear about your friend. Look at the bright side: he gets to listen to the angels sing now.

You're welcome.

Reply to
Bill Horne

I can fill in some details about the timesharing services. They began appearing late 1966. I couldn't get enough time on the EDL's IBM 1130 so I started calling around for "trial" accounts with Tymshare, ITT, some PDP-10 company (Call-A-Computer ?), and a few others. TTY ASR33s were the terminals used by all of them and 110 baud acoustic modems were used over the PSTN. I attempted to "sneak" in as much EDL work as I could on the trial accounts and no one seemed to mind since I also provided a lot of feedback to the companies, so it was kind of a win-win for all.

But Bill's correct about the cost. The EDL ended up contracting with Tymshare for "overflow" computing work and there was one month the bill exceeded $10,000 -- this was 1967 or early 1968 -- and no one thought that amount was out of line since the work had to be done and the US Govt was basically footing the bill.

By that time I saw the value of timesharing vs. batch processing and when Tymshare made an offer, I couldn't refuse. I ended up doing their EE and other engineering programs, eventually compilers, editors and database systems, police information system (POINT), and participated in the development of the first commercial RDMBS (MIDAS or Magnum, depending on the computer platform) before joining another computer startup [company] in 1972.

Heh, I found my old Tymshare business cards. The Palo Alto location was

1968, the Cupertino was 1969 when everything moved to larger facilities:
Reply to
Thad Floryan

I was pleased to know all three of the men mentioned in the article, though I knew Bob Weitbrecht a lot better than the other two.

Bob got him ham radio license W6NRM as a youngster. He was profoundly deaf, but was able to hear or feel vibrations just enough to copy Morse code with headphones. For many years that was his main social outlet. I like to compare it with that famous cartoon captioned, "On the Internet nobody knows you're a dog", thinking "In ham radio (Morse code) nobody knows you're deaf."

Amateur radioteletype came into being circa 1950, principally in and around New York City, using a lot of obsolete Model 12 machines that had been surplused by the police department there. At the time there was a lot of commercial and military radioteletype on the air; but amateurs were handicapped by FCC regulations and by the difficulty of obtaining machines. Bob, on the West Coast, and then at Yerkes Observatory in Wisconsin, was an early adopter and promoter of RTTY as it is called. Much of his professional life was connected with instrumentation for astronomy; hence the Yerkes job. He was a major technical contributor to RTTY, and also traveled a lot to visit with other RTTY enthusiasts or to help hams get on the air with RTTY. So much of his social life revolved around other hams and communication via RTTY.

After he left Yerkes and returned to California he continued his RTTY activity. He also had a telephone put in; one of his friends had an amateur station remotely located from his home that could be controlled and operated over a telephone line. All this was strictly disallowed by the telephone company "no foreign attachments" rule at the time; but the guy who did it was an expert (a former Bell Labs man, in fact) so there was no way the telephone company would know of this activity. Bob also started using the telephone to communicate with ham friends in the local area, using Morse code and TTY signaling.

Bob became acquainted with Jim Marsters around 1960. This opened up a world of new ideas to Bob. Jim was a deaf person who lived life to the fullest: a Porsche driver and an airplane pilot, among other things. Bob hoped to get his new friend into amateur radio so they could converse by RTTY across the distance between Redwood City and Pasadena. At that time an amateur license required proficiency in Morse code, and that turned out to be one thing that was beyond what Jim could achieve with his degree of deafness. So Bob proposed that they communicate over the telephone, in spite of the cost of long-distance calls at that time period.

Radioteletype uses frequency-shift keying, in which one frequency is transmitted for mark and a slightly different frequency is transmitted for space. This is practically necessary to get good signal quality. In his land line work Bob had experimented with single-tone space-only transmission. The advantage of this is that the spacing tone is present only during transmission of a character, so that no send-receive switch is necessary. The much more complicated Bell System modems used two pairs of frequency shifted tones to achieve full duplex operation. This also made a send-receive switch unnecessary, but required that the modem distinguish between the calling and called party.

The single-tone space-only modem worked fine around the local area. When Bob and Jim tried it on a long-distance call the transmission was troubled by echoes. (digress to standard telephony lecture on the causes of echoes on long circuits and the need for echo suppressors. I don't know for sure, but my guess is that the Redwood City to Pasadena connection was too short to require the telephone company to put in echo suppressors, but not short enough that the echoes were not troublesome to the TTY signals)

Eventually Bob hit on the idea of transmitting a tone during mark intervals that would essentially drown out the echoes. By having this tone come on as soon as a key was pressed on the machine, and having it go away soon after the end of a character, he preserved the feature of not requiring a send-receive switch. Bob received a patent on this kind of modem, number 3,507,997 issued in 1970.

Another part of the system was the use of old Baudot Teletype machines, as the radio amateurs were using. Someone mentioned the problem of maintaining that complicated machinery. This was something the amateurs were already coping with. There were simplified instruction books and articles, including one produced by the deaf community, on Teletype repair. Also there were lots of hams or friends of hams who had learned Teletype maintenance with the telephone company or Western Union or in the military and were willing to help out. So maintenance of the machines was not a big problem for the deaf any more than it was for the hams.

The use of Baudot was part of the reason the Bell System and the government resisted helping the deaf people. Official policy was to speed the adoption of ASCII and the phase-out of Baudot.

As Bob became busier with his work for the deaf he became somewhat less visible in amateur radio, although he continued to operate and to develop equipment. Many younger hams seem to think that amateur RTTY practically began with Irv Hoff, who wrote a series of articles on RTTY and was generally looked on as an authority during his lifetime; they are nearly oblivious to all the extensive work that was done before Irv started publishing.

There is a whole book devoted to these matters. "A Phone of Our Own : the Deaf Insurrection Against Ma Bell" by Harry Lang, published by Gallaudet University Press.

None of the above is meant to diminish the contributions of Jim Marsters and Andy Saks (and countless others who were inspired by their work). As Teletyping took hold among the deaf there were interesting activities all over the country. Before the advent of home computer BBS systems there was something, in St. Louis as I recall, that allowed TTY users to dial in and receive the news, which was being stored in punched paper tape.

Reply to
Jim Haynes

We were using computer timesharing services, via PSTN dial up, back in

1968, which were an outgrowth of the Dartmouth GE system, which began a few years earlier. When the commerical services began I don't know but by 1968 they were out there. By 1970 a number of competing time sharing services were available. HP had its 2000A time sharing system which was popular (offered by small companies) and pretty good. Our school system had its own (which was not so good, but connect time was free). The systems were available nearly 24/7 except for maintenance times.

Bill Gates got his start on such systems, and he would work all night in the computer center.

No, I understand price was the issue. The people in the other newsgroup said there was NO such service available at ANY price until Dr. Marsters came along. I wanted the be clear about the distinction between "not available at all" and "not affordable by everyday people".

The modems on our teletypes had a tiny switch for half or full duplex. Certain computers wanted half, certain wanted full. If you used full on a half duplex connection, depressing a key would transmit the character but nothing would be printed.

Yes, the Teletype models 32 and 33 were intended for light duty service. But we used them extensively and they held up very well. On this site's archives there is an article in the WUTJ about their mechanical designs.

Good point. The newer "K" handsets, which are more angular, probably don't fit well either. Many handsets today have all sorts of odd shapes.

I understand it now.

Reply to
hancock4

I worked for a company which used a timesharing computer. We were under strict orders to prepare everything offline on paper tape first, run it, then do all debugging offline, so as to minimize the connect time charges. We had a 300 baud terminal, acoustical coupler. Printing was much quieter than a Teletype, but the paper tape punch at

300 baud was horribly noisy.

There was a rich kid at school who rented his own Teletype for a few months.

Reply to
hancock4

I guess you missed this paragraph in the original post:

"What I think Dr. Marsters' developed was a _low-cost_ system for deaf people to use instead of renting a Bell System modem, teleprinter, or a Western Union Telex connection. Since AT&T's and WU's services were intended for busineses, the prices weren't cheap. "

Not correct. The Teletypes AT&T provided for time-sharing access via PSTN were ASR, automatic send receive. You [could] call another teletype, and if [it was] unattended, it would automatically answer, and you could leave a typed out message. If it was attended you could have a conversation.

A deaf person defniitely COULD call other people directly with such devices, either to leave a message or have a conversation. There was the rental on the teletype ($100/month), but the phone call was regular residential rates.

I believe Telex also offered the option of leaving messages or having a conversation. But Telex charged by connect time (not much), and most users 'batched' their messages via paper tape to minimize connect time.

Further, you could own your own non-ASR Teletype and modem and call other Teletypes, but they'd require someone to be there.

We used to do this years ago. Early form of teen texting.

Reply to
hancock4

I don't recall the actual rates, but $15/hour connect rings a bell. But that wasn't what contributed to that one month's $10,000 bill, it was the CPU charge that kept the meter spinning at high speed.

Assuming it was just me (it wasn't), $15 * 8 hours/day * 31 days is $3,720. Most of the programs I was running I already had running on the IBM 1130 and I just copied them over (Fortran) to Tymshare whose SDS-930/-940s were able to compute faster (but at a cost).

New programs were actually cost-effective to develop online given interpretive compilers which eventually became the salient features of many of the timesharing companies over and above pre-packaged apps.

Media compatiblity was nearly non-existent except for 8-level paper tape and I still have my reader and several boxes full of some of the old programs on tape since I had a TTY ASR33 at home in the late 1960s (followed by its CRT counterpart, a Datapoint 3300, then a DataMedia DT-80, then real (UNIX) computers such as AT&T 3B1 and 3B2, Convergents, Suns, etc.).

Reply to
Thad Floryan

That's a _neat_ trick, regardless of the capability of the terminal device, when the modem supplied were "originate only" devices.. Two modems, both in 'originate' mode, simply cannot talk to each other.

Telex and TWX were both closed networks. Had to have your own dedicated line for that service. And you couldn't reach anybody who didn't have their own dedicated line for the same network you used.

Those networks worked *very* differently from the PSTN. On Telex and TWX, the 'C.O. equivalent' played the 'host' side of the connection to both sets of customer gear.

NOT over the PSTN, prior to the "Carterphone decision", or at least not _legally_. If you used your own modem, you had to have a "DAA", which required _manual_ intervention, both for sending and receiving.

***** Moderator's Note *****

AFAIK, TWX calls were connected directly from the originating to the terminating machine, without the CO providing anything but a path for the modem tones. Of course, on calls from 4-row to 3-row machines, there was an intermediate step, but I'm sure that 4-row to 4-row was a direct connection. Each 4-row machine could originate _or_ answer calls, switching modes to use the correct set of modem tones.

Prior to its sale to Western Union, TWX _DID_ use the PSTN. The TWX offices were simply set up with translations which prevented calls between 4-row TWX numbers and POTS numbers, but TWX calls _were_ routed via PSTN trunks. ISTR that 3-row was converted from manual to DDD at some point, and that 3-row numbers could be directly dialed from POTS numbers, but it's been a while.

Bill Horne

Reply to
Robert Bonomi

The modems supplied were NOT "orginate only". Where did you get that idea?

The modems on teletypes _were_ originate and answer. As mentioned, two-way conversations were common.

A major feature of TWX and Telex was the leaving of messages and two- way conversation.

Now, there were "receive-only" teletypes, such as wire service printers. Perhaps the modems associated with them (if they weren't on a current-loop line) were receive only.

Telex and TWX could talk to each other; this was established later on.

In the waning days of Telex and early days of services like Compuserve, you might (not sure) have even been able to send a message from Compuserve to them. I know you could send WU Mailgrams; and Compuserve had other connections.

No, it merely provided the communications path. Look up the WUTJ on this website.

Or, as commonly done, an acoustical coupler.

Also, use of a DAA required no attention whatsoever; it merely acted as a filter on the line and was completely transparent to the user. The "manual intervention", if needed, was merely lifting a lever upon receipt of the carrier sound. Often the phones used for dial-up were the two-line models that had a rotating knob and pull-knob. When connecting, the knob was pulled up. (Such phones were often used for a variety of special wiring other than the two-line function).

***** Moderator's Note *****

The modems used for Teletypes on press wires were, indeed, one-way: a perfect example of Simplex transmission which I did not mention earlier.

TELEX and TWX could talk to each other through gateways, after WU inherited TWX from Bell.

WU had an Easylink service which offered TELEX interconnection: I used it to send an international TELEX in 1986. I don't know if it could be reached via Compuserve.

I think the point the previous poster made about Data Access Arrangements was that they required manual activation. However, I can't help but wonder if the Bell companies leased out their remaining inventory of TWX machines for "TTY" service after WU took over the "official" TWX network, and if that was so, they would probably have been capable of auto-answer.

Bill Horne

Reply to
hancock4

I'll put in this comment, before it gets any deeper. The Bell modems, 101 and 103 series, were full duplex AFSK using a different pair of tones in each direction. Hence a modem had to know if it was originating or answering to know which pair of tones to use for which purpose; and the modems were complicated because of having to be able to transmit or receive using either tone pair.

There were strapping options in the modems to select which tone pair to use for originate, and which tone of a pair was mark or space. Hence by strapping there were eight possible mutually-incompatible services using the same modems. One of these was used for TWX, and another for DataPhone. Others were to have been used for things like WADS and one called WADS-prime. Although TWX mostly used the PSTN it was charged at a lower rate than voice calls. Hence TWX machines did not have a handset so you couldn't talk over the connection. DataPhone used the PSTN at the voice rates, so it did include a telephone handset that could be used for conversation between operators.

With acoustic couplers, or with third-party hardwired modems after Carterfone, if the intent was to use the machine as a computer terminal only then the modem could be originate-only, since it never had to answer incoming calls. The modem needed to transmit on only one tone pair and receive on the other.

The original Dial TWX modems, the 101 series, were pretty monstrous, being about six inches thick and a foot wide and maybe 18 inches long. The Teletype machines designed for use with these modems had 99 wires running between the Teletype and the modem. This was necessary because most of a telephone set, the dial and speaker or earphone, the control buttons, etc. were built into the Teletype machine. When the 103 modems with built-in telephones came out, some of the Bell operating companies cut costs by buying much simpler private-line machines from Teletype and using the 103 modems, with strapping appropriate for TWX or DataPhone as needed.

***** Moderator's Note *****

I'm puzzled by your post: I'm not saying you're wrong, but I am confused. I infer that the modems were wired for different tone pairs when used on TWX vs. DataPhone, and I have not experienced that in my usage.

When I worked at Back Bay Toll in Boston, we had a 35ASR which was used for company reporting. It was compatible with TWX machines - I know this because I once sent a TWX to a real TWX machine by plugging into the TWX circuit of a WU customer, and it worked fine.

However, the machine was also compatible with the common modems used for Bulletin Boards, and I know _that_ because I used it to log into Ward Christensen's CBBS. It was, of course, hard-wired for local echo, so every character that I typed printed twice, but it _did_ work.

Please provide any URL's or other information about the ways the modems were wired differently for TWX or DataPhone. TIA.

Bill Horne

Reply to
Jim Haynes

On this discussion we also have to include the specific time frame because equipment changed noticeably over time. Over pioneer pieces of equipment gain new functions over time*.

I'm speculating here: Bill, you mentioned going into a "TWX circuit of a WU customer" and also using your teletype for BBS access. This sounds like it was toward the late 1970s. Was this after WU acquired TWX? Perhaps TWX had changed by that point. Or, perhaps by that point the modems were a later generation and more flexible, esp on the

35 which was the heavy duty model. Or perhaps your machine at the toll center was more sophisticated than a customer's machine.

When did Hayes come out with the AT command set? Who invented modems that could automatically dial and do other functions once done by hand? (Like when we set the speaker volume, we turned a knob on it.)

***** Moderator's Note *****

This was about 1973 or 74: WU had taken over TWX, but N.E.T. was still maintaining the TWX (WADS) office at Franklin Street in Boston (a #5 Xbar), so it was "just after" they sold the TWX network.

Bill Horne

Reply to
hancock4

Bell 103 has seperate originate and answer tones. TDD does not. TDD is not Bell 103, not even a little bit like Bell 103. TDD isn't even really FSK... it's ASK with a keeper tone.

--scott

Reply to
Scott Dorsey

Well, now, maybe they never made use of those features - we need someone with a BSP for 103 modems from back in the 1960s. I just remember around Teletype Corp. they were talking about WADS and WADS-prime before those services failed to gain FCC approval. I was under the impression that they had deliberately made TWX and DataPhone incompatible, but maybe that was only talked about and was not done.

I forget, too, how it was that 4-row TWX and 3-row TWX machines were prevented from calling each other directly, but I guess that was something about the area codes.

Reply to
Jim Haynes

Late 70s.

The 'Bell 801 automatic calling unit' handled that, _external_ to the modem itself.

Existed more than a decade before Hayes.

Reply to
Robert Bonomi

It did, but the control of the dialer was separate from the data path. Hayes' important innovation was to combine the data and control channel so any old crummy microcomputer with a minimal serial port could autodial.

R's, John

Reply to
John Levine

Yup. And that separation is a 'big win', in terms of security and access controls. :)

It was also possible to 'share' one 801 across multiple modems on multiple lines -- where you had multiple 'dial-out' lines, that is.

Hayes "contribution" was to do it *without* using any of the so-called 'modem control' signals present in a full RS-232 interface.

With the _big_ cost differential between a minimal 3-wire terminal port, and a 'full RS-232' port, one could successfully sell a modem that needed only a 3-wire port for a significantly higher price than a 'full RS-232' modem sold for, and still *SAVE* the customer money, overall. (combined cost of modem and the serial port for the computer)

Other people, before Hayes, had used 'in-band' signalling (and still do) for conveying dialing instructions to the modem. However they require manipulation of some of the additional signals in a 'full' RS-232 interface to switch between the mode where commands are sent and one where all data is passed through 'transparently'.

Hayes' _key_ innovation, and the one for which they got a patent -- the only patentable feature of their basic system -- was the means of escaping from "_transparent_ data transmission" back to "command mode". i.e., the mandatory 'guard time' surrounding the magic character sequence ('+++', by default).

Reply to
Robert Bonomi

Actually, at the time there were a bunch of modems (including the Anchor Signalman II) which used the DTR line to control the hookswitch. Doing this allowed software on the computer to toggle the DTR line and pulse-dial the modem. Then it would wait for the CD line to go high when the modem finally detected a carrier and locked up.

--scott

Reply to
Scott Dorsey

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.