6

Was it possible to send data over the phone-line without a modem?

Clarification

  • Was it possible to send data over the phone-line without a modem, either by connecting computers directly to the phone line, or, by connecting cassette-recorders/players directly to the phone line
  • It must not require the use of a commercially available modem, but allow only a bit of extra wiring and maybe some resistors, and, or, audio-speaker and microphone .
  • This is for the scenario where two people may want to share a program or data, and they don't want to waste money on the modem .
  • I'm assuming the answer is no, or else people would not have bothered buying modems.

UPDATE
I suspect that either it may have worked for some computers and phone-lines, or, that it may have been easy for the computer manufacturers to design the specifications so it would function well without a modem, however, I suspect it was strongly discouraged in order to avoid damage to computers or to phone-lines :

  • Only commercially available modems were legally certified to be connected to the phone-line, not the computers themselves, to prevent people damaging the line .
  • And connecting a computer straight to the phone line may risk damaging not only the line, but also the computer .

Parameters:

  • It must not require the use of a commercially available modem, and only a bit of extra wiring and maybe some resistors .
  • Initially, this question is not about the use of any external memory devices like cassette-recorders etc.
  • Or if required, this question is about computers that had access to cassette-tape-recorders, not disk-drives, but information about disk-drive machines may also be useful.
  • For the common affordable home-computers in the years approximately 1978 - 1984 (a few years before, or after, is ok)
  • NOTE - By plugging the cable straight from the transmitting computer and into the telephone, the cable that the transmitting computer normally used to save to cassette-tape, to the microphone-jack of the telephone (did that exist?).
  • Or, by attaching an audio-speaker to the cable that the transmitting computer normally used to save to cassette-tape, and then placing that audio-speaker next to the telephone-microphone.
  • (Theoretically, could you just have left a program on a cassette-tape-answering-machine, unless there were copyright problems in those days, or data quality loss because of the cassette tape)
questiontype
  • 1,061
  • 7
  • 12
  • 16
    One person reads a program off the screen, and the other person types it into a keyboard? – supercat Feb 15 '19 at 03:42
  • 6
    Don't know if it was done over phone by anyone, but at least in 2 countries (UK and Sweden) public radio or television sent prigrams over the air. – UncleBod Feb 15 '19 at 04:13
  • @UncleBod That happened in Serbia too, according to Voja Antonic – Omar and Lorraine Feb 15 '19 at 08:32
  • 1
    Frequency Shift Keying (FSK) was used in early modems and also in cassette tape data storage, so it seems plausible. You would just need the right adapter. – snips-n-snails Feb 15 '19 at 10:21
  • 1
    @UncleBod It happened in next to all European countries. Including the Eastern. – Raffzahn Feb 15 '19 at 10:33
  • 3
    Even in the @supercat case, one might argue that the person reading the program and speaking into the phone is "modulating" the information, and the person listening on the other phone and typing in the program is "demodulating" the information. This requires a very broad definition of what is meant by modulating and demodulating. Something like "changing to a form suitable for transmission over an acoustic voice link" and back. – Walter Mitty Feb 15 '19 at 12:42
  • 1
    Programs transmitted over the air are, in one sense, modulated. The M in AM or FM stands for modulation. A radio transmitter conveys an acoustic signal over a carrier wave by modulating the carrier wave in some fashion. A radio receiver decouples the carrier from acoustic signal by demodulating, and recovering the original acoustic signal, to some level of approximation. – Walter Mitty Feb 15 '19 at 13:21
  • 3
    I think this is basically asking if computers can, without special equipment, modulate/demodulate data suitably for transmission down a POTS line? Which is clear enough I suppose. This seems relatively on-topic, so I'll reserve my close vote for now. It may be unclear because the OP isn't clear themselves on what a modem actually does for us? –  Feb 15 '19 at 14:28
  • (Reminds me of the HP system I knew about back in the mid-80s that had a facility for using dedicated phone lines to allow developers to connect to the serial terminal that normally would be RS-485 [or that weird proprietary HP one, I don't remember] with minimal hardware involved. You just connected a terminal of some sort and shouted down the line and the terminal would eventually answer.) –  Feb 15 '19 at 14:33
  • 1
    Just FAX the program over. This is a joke. First off, there were only a few hundred fax machines in 1978. Second, a FAX machine contains a built in modem. – Walter Mitty Feb 15 '19 at 16:43
  • @jdv: If one person were to hold a telephone phone mouthpiece next to an Apple ][ speaker while another held a microphone up to a telephone earpiece, and fed the microphone signal through a preamp into an Apple ][ cassette input, it would probably be possible to send data, slowly, without any other hardware. I don't know of any machines of that era with a built-in microphone suitable for the task. – supercat Feb 15 '19 at 16:50
  • @supercat, sure. I mean, this is sort of the same as retro systems that can grok sound files made from data cassettes audio output. It's that modulating to make the beautiful music and the demodulating to turn that music into information that is key. Reminds me of the many sound and image based transmission protocols that gaming systems used in the past (maybe the Timex computer watch interface). Or the unconfirmed warnings a few years back about malware jumping the air-gap with sound. –  Feb 15 '19 at 16:55
  • @jdv:One of the key features of a telephone modem compared with ad hoc approaches is that telephone modems generally include the ability to filter out their own transmissions and operate independently of the attached computer's CPU. Even if one tried to connect two apples as described, with each having a microphone and speaker, they would be limited to transmitting at discrete times, and neither side would be able to do anything useful while it was sending or receiving data. One might be able to design an interactive game if one computer were the "master", and... – supercat Feb 15 '19 at 17:00
  • @supercat I'm sure we are in agreement. A modem type device makes it very easy to originate, answer, and otherwise fold, spindle and mutilate data with minimal help from the host system(s). –  Feb 15 '19 at 17:02
  • ...knew how long the slave would take to do anything before reverting to looking for communications from the master, and when the slave was expected to send data back. Having a designated master is important to allow for recovery if a particular transmission fails. If the slave goes into a "wait for input until it is received, and then send a reply" mode and the master goes into a repeated "send data and look for reply" mode, the units will re-establish communication. If the slave tried to send data unsolicited, that could make it miss the master's transmissions. – supercat Feb 15 '19 at 17:03
  • @jdv: I remember back in the day wondering why cassette interfaces weren't used for connection to phone lines; I think it's probably helpful to note that such a thing would be possible, but be sufficiently inferior to a modem as to justify the extra cost of the latter. – supercat Feb 15 '19 at 17:05
  • @supercat it can come down to good, fast, or cheap; choose any two. A modem device provides a lot of value for the purpose of reliable intercommunication over POTS. But mod/demod can be abused in all sorts of cheap ways for fun and profit. Now I'm imagining a near-ultrasonic broadcast aloha protocol to coordinate killer 80s robots. I mean, I didn't have to make the killer robots, but once you set up the island fortress and hire all the henchmen, the rest just flows naturally. But now I'm monologuing! –  Feb 15 '19 at 17:09
  • @supercat Actually, the issue you describe of master/slave is quite similar to collision detection in an original Ethernet network. Wait for quiet. Start sending. Too noisy because someone else started about the same time? Stop and wait a random amount of time and try again, using the modem that you have built but just refuse to call a modem :-) – manassehkatz-Moving 2 Codidact Feb 15 '19 at 18:02
  • @manassehkatz: Assigning fixed master/slave relationships eliminates the possibility of collisions and the need to add random waits to avoid them. The problem with such relationships is that any slave-to-slave communications will at minimum need to be coordinated with the master, if not routed through it, so they don't work well in true peer-to-peer systems. – supercat Feb 15 '19 at 18:40
  • @supercat Agreed. And I have worked with many systems of all types (as have so many others). The main point is that modem (whether official or unofficial) communication is little different from Ethernet or RS485 or WiFi or a bunch of people at a meeting - you have to figure out how to either make each system take a turn or figure out how to handle the inevitable conflicts. – manassehkatz-Moving 2 Codidact Feb 15 '19 at 18:51
  • I looked up the history of FAX machines, and I was surprised to learn that the first such machine predated the telephone. – Walter Mitty Feb 15 '19 at 20:59
  • 3
    With these updates, an even deeper hole is dug ... ready to bury this. Serious, legality needs also to clarify the area the claim is made to. And claims about damaging the computer will need an near endless list of computers and parts involved to make any useful criteria. A bit more and Australia will be reached. – Raffzahn Feb 15 '19 at 22:14
  • With "tape output connected to phone line", you've got one way communication. So you're basically hoping the data make it through unscathed. There is no opportunity for the receiver to NAK the sender. If you can also connect a microphone, you can maybe use it half-duplex mode with a simple ACK/NAK scheme, but you're going to have to invent a way to do line turnaround (have the computer stop sending and start listening). – dave Feb 16 '19 at 03:03
  • 1
    I think you should have just asked "Could you use a retro computer's built-in cassette tape interface to connect two computers over the phone line, without any modem peripheral attached to each"? – Brian H May 31 '19 at 15:07

10 Answers10

30

TL;DR If it acts like a modem, it IS a modem.

What is a modem?

a combined device for modulation and demodulation, for example, between the digital data of a computer and the analog signal of a telephone line.

What are you trying to accomplish? send digital data (a program) from one computer to another over the analog signal of a telephone line. So whether it is an older modem with an acoustic coupler or a Hayes Smartmodem or anything in between, including something you just come up with and experiment with (violating telephone company rules along the way), you are using a modem when you transfer digital data over an analog phone line. An analog tape interface, as was the case with all microcomputer tape drives of that era (mainframe tape drives were different) does the same thing - convert digital data to audio tones. Connect that tape interface to a phone line and it is now a modem. DTMF created from a computer is a fancy (and slow) way of converting digital data to audio tones - send it over a phone line to be decoded by a computer and you have created a modem.

Community
  • 1
manassehkatz-Moving 2 Codidact
  • 14,358
  • 2
  • 37
  • 65
  • 1
    So, it would be correct to call the tape interface a modem then? – UncleBod Feb 15 '19 at 07:38
  • @UncleBod no. they're tape interfaces. there's no analog phone line involved, is it? – Tommylee2k Feb 15 '19 at 09:51
  • 3
    @Tommylee2k It wouldn't take much to adapt an RJ11 jack to 3.5mm phono connector. – snips-n-snails Feb 15 '19 at 10:08
  • @traal that is what I'm thinking also. If the frequency used is not too high, it would be perfectly possible to send it over a phone line, wich seems to be what the question is about. – UncleBod Feb 15 '19 at 11:02
  • 6
    "you are using a modem when you transfer digital data over an analog phone line" says it. you're spinning in circles, using a self-made modem still is using a modem. – Tommylee2k Feb 15 '19 at 11:50
  • In modern usage, the word "modem" is taken to mean "equipment which modulates and demodulates signals for transmission over a communications channel", and tapes are not considered channels in this context, while phone lines (including wireless) and radio/satellite links are. Technically, the modulation/demodulation done by modems isn't materially different from the modulation/demodulation you would do with an audio tape recorder, but by common convention we don't use the word "modem" to describe tape devices -- it would be more confusing than useful. – Ken Gober Feb 15 '19 at 14:57
  • The takeaway here is that the signalling, or baud, rate is different for different signal paths. The rates supported by copper going to a tape head causing metal oxides to line up in a special way is going to be different that mechanically switched POTS lines from a decade ago. Something is in the business of staying within that signalling rate or there is no successful transmission because the signal will not be able to be discerned from noise. –  Feb 15 '19 at 14:59
  • 2
    @KenGober The circuitry connecting a computer to an analog (== voice grade in the stated time frame) tape drive is fundamentally the same as a similar vintage modem - convert digital data to a bunch of audible tones. But it only becomes a modem, as I've already stated, once you connect it to a phone line - which is what the OP is talking about doing. – manassehkatz-Moving 2 Codidact Feb 15 '19 at 15:08
  • 1
    @jdv Most (classic homecomputer) tape formats use frequencies way within the guaranteed range of POTS. Wasn't there a somewhat related question recently? Ahh, got it: https://retrocomputing.stackexchange.com/a/8543/6659 – Raffzahn Feb 15 '19 at 15:08
  • @jdv But the bit rate of a mainframe (or more modern microcomputer tape drive - e.g., QIC) is quite different from what you get by turning an analog voice-grade tape drive into a computer storage device. See TRS-80 tape drive for an example - it was really a "modem" in every way except for connecting to a phone line - speed was 250 bits/second - close to the typical 300bps of a modem in those days. – manassehkatz-Moving 2 Codidact Feb 15 '19 at 15:11
  • 1
    @manassehkatz I'm pretty sure my statement agrees with yours. The KIM-1 cassette interface also relied on software modulation I recall. I'm backing up the statement that, at the end of the day, each end of a connection need to agree on some sort of modulation and demodulation. Special hardware for POTS also had other jobs that we often associate only with "modems". But it does make sense to use these terms in context of this specific Q&A. But I'm reluctant to get caught up in nomenclature, other than agreeing the word "modem" means something specific in different contexts. –  Feb 15 '19 at 15:14
  • @KenGober: The term "modem" typically refers to devices that combine the "mo" and "dem" functions. Many systems had a tape-drive "mo" function and a tape drive "dem" function, but they wouldn't be used together like the halves of a modem. – supercat Feb 15 '19 at 16:48
  • 5
    A tape drive interface connected to a phone line would be a half duplex modem. – manassehkatz-Moving 2 Codidact Feb 15 '19 at 16:53
  • 3
    @manassehkatz: A typical tape drive interface doesn't actually modulate or demodulate anything itself, but instead transmits pulses generated directly by code that modulates data, or makes pulses available to code that can decode them into data. The combination of the interface and the code to drive it could be a half-duplex modem, but the hardware itself wouldn't be. – supercat Feb 15 '19 at 19:01
  • 6
    @supercat Then the computer itself is the modem and so the question contradicts itself. – snips-n-snails Feb 15 '19 at 19:24
  • @traal: In general, a general-purpose computer running a program to perform a task would not be described as a device to perform that task, but rather a device which can be used to perform the task. I don't think one would say, for example, that an Apple //e running "Visicalc" is a spreadsheet. – supercat Feb 15 '19 at 20:51
  • @supercat My cell phone is a general-purpose computer but I also call it my "camera". – snips-n-snails Feb 15 '19 at 22:05
  • Supposedly the Commodore Datasette stored bits directly onto the tape without modulating them into audio tones first. (For reasons that escape me, Wikipedia says this made it slower than contemporary systems with audio modulation). – hmakholm left over Monica Aug 09 '19 at 12:53
14

"Hello, Jeff?"

"Yes?"

"I hear you got your new Radio Shack Model 3?"

"Yes!"

"Is it on?"

"Yup."

"type the number ten and a space, then type print quote hello world quote and press enter"

"Okay now what?"

"Type run and press enter"

"Wow!"

Sorry, I couldn't resist.

Bill Hileman
  • 1,206
  • 7
  • 12
4

Yes, it's possible. For example, systems like the X10 fr home automation can be controlled via a phone line using DTMF (touch tone) dialing. While such systems are often designed to be used by humans, there have been instances of direct machine-to-machine communication too.

For example, the user could call the X10 receiver connected to an IBM compatible PC running the DOS based X10 software, and use DTMF tones to reprogram the system, altering timers and controlling devices, i.e. sending programming.

Unfortunately these systems are so old that I can't find any good references on the net. We are talking early DOS days. The best I can do at the moment is this LGR video with a demonstration of X10 control over the phone.

Note: The DFTM encoder and DTMF decoder devices are not modems. Modem is short for "modulator-demodulator", i.e. a device that incorporates both modulation and demodulation for two way communication. The X10 system is one way, one end is a pure modulator and the other is a pure demodulator, thus neither are modems.

user
  • 15,213
  • 3
  • 35
  • 69
  • 10
    DTMF is a modulation scheme that can represent 16 symbols and the decoder at the far end is a demodulator. That makes the whole setup a modem, albeit a simple one. – Blrfl Feb 15 '19 at 13:34
  • 3
    I was going to write something similar at first, saying that you basically can't send any digital data over a POTS line without modulation. But the question does say a "modem", and a DTMF decoder is not generally considered to be a modem. It was designed for human-machine interaction, so calling it a modem is a bit like saying someone typing in a program from a book makes the keyboard a data storage device. – user Feb 15 '19 at 14:10
  • 5
    @user Of course it's a modem (or at lease the MO part). It modulates user interaction (data) into sound - which sound frequencies then get turned into electric frequencies. So perfect example for a modem - and just because the frequency points used are the same as for in band signaling doesn't change the fact - beside, that's exactly the standards way phone systems have used for data exchange between them (no soudn involed) since the 70s. – Raffzahn Feb 15 '19 at 14:59
  • @user If you program a computer with AI-like capabilities to use voice synthesis and voice recognition to communicate with people. And then you have it call into a bank or utility company or whatever that has an automated voice response system as the gateway before getting to real people. And the two systems interact in a functional way. Are those modems? Arguably, yes. Not designed as modems, but modems nonetheless - converting digital data to analog values for transmission over a phone line. – manassehkatz-Moving 2 Codidact Feb 15 '19 at 15:35
  • 1
    @manassehkatz Yep. Have Alexa talk to Siri over the phone... – tofro Feb 15 '19 at 17:52
  • Well... It's not a "modem" in the common sense of the word, i.e. one of those boxes that makes the weird whistling noises and connects you to the internet or a BBS. – user Feb 15 '19 at 20:23
  • @Raffzahn it's definitely not a modem. Modem standards for modulator-demodulator, and in the DTMF system each end only does either modulation or demodulation. I.e. to be a modem a single device has to incorporate both functions. – user May 31 '19 at 11:02
  • 1
    And that differs in what part from my comment? And no, a modem doesn't have to incooperate both parts. That's only useful for bidirectional communication, isn't it? Beside, I'd say, trying hard to be negative isn't a good idea and wast of time. – Raffzahn May 31 '19 at 12:15
  • @Raffzahn has you say in your comment, it is at best half a modem. – user May 31 '19 at 12:17
4

Not an answer just points to be connected...

  1. Phone system means analogue dial up network for voice (sound) communication.

  2. Sound is transmitted between terminals (phones) as electric signal

  3. Phone systems are designed to transfer electric signals with frequencies of 300 to 3000 Hz (*1)

  4. There is no galvanic (direct) connection.

  5. Audio transformers are used to separate lines and systems.

  6. Their frequency response is designed to support this range

  7. Frequencies outside the range (300..3000Hz) will be dampened.

  8. Filters may be applied in addition to guarantee these limits.

  9. Phone systems do only offer a/this single (user side) channel with these properties.

  10. System signaling may be in or out of band (time or frequency diverted) but is not available for user side content transfer.

Derivates points:

  1. No DC signal can be transmitted.

  2. No level (voltage) based signal can be transmitted.

  3. No current signal can be transmitted.

  4. Only signals made up from frequencies within the limits can be transported.

In more than 100 years of application of these systems, many special to purpose devices have been created and named by their inventors and/or users. So, Question: What would be the basic term for a devices that transforms user signals into a collection of frequencies for transmission and back?

*1 - Overall simplification, some systems may use slightly different. These limits are not just for interoperability between systems, but also to ensure higher level of transportation (like multiplexing on lines between interchanges)

manassehkatz-Moving 2 Codidact
  • 14,358
  • 2
  • 37
  • 65
Raffzahn
  • 222,541
  • 22
  • 631
  • 918
  • Not upvoting because "not an answer". But valid points that support my premise quite well, thank you. – manassehkatz-Moving 2 Codidact Feb 15 '19 at 14:43
  • 1
    @manassehkatz :)) Ouch, according to one source, I should be devastated now about the missed points. Let's check tomorrow if I survived the resulting hangover:)) I do not think the Quesion is really asking for an answer, it seams more like a pile of made up rules to prove some point I fail to see. That's why I restricted myself to list a few of the basics involve (And to be honest, I didn't vote for you as well (I almost did), as the answer (albeit perfectly good) written trips into the missleading parts of the question - at least in part.) – Raffzahn Feb 15 '19 at 15:10
  • 1
    Thank you, that was the freqency bandwith I wanted to find. So, if the cassette interface did keep itself within those limits (300 to 3000 Hz) the audio could in theory be sent over the phone line without any more change to the signal. I assume that is what the question is about. – UncleBod Feb 15 '19 at 15:39
  • 1
    @UncleBod But then you have just turned a tape interface + extra wires into (TA DA!) a MODEM! – manassehkatz-Moving 2 Codidact Feb 15 '19 at 15:59
  • Isn't the frequency range 300-3400 Hz? (see eg Voiceband on wikipedia) – JeanPierre Feb 16 '19 at 08:30
  • @JeanPierre As said, it depends on your country/phone network (and the time looking at). POTS in the US for example was defined as 300..3300 - except where not :)) In Europ it was usually defined as 300..3400 within a 4000Hz channel. Reality all around the world usually ended up with smaller bands, depending on age and technology used. – Raffzahn Feb 16 '19 at 08:42
2

In 1985, Your Computer magazine had a software download service called TELSOFT. Audio files of (I presume) recorded MODEM tones were played down the phone line in a continuous loop. It was supposed to be used with a MODEM and special software that you typed in, but I did not have a MODEM for my ZX Spectrum so I tried receiving it directly through the Spectrum's 'EAR' socket.

Back then a toll call from New Zealand to the UK cost ~$10 per minute, and I wasn't sure how well my program would work, so I recorded the audio on a tape recorder connected to the telephone's receiver insert. I could then play the recording back as many times as I needed until I got the software working properly.

After a few hours of debugging the software I finally managed to get a good data dump, only to discover that it was for a Commodore 64! I had forgotten that magazines took 3 months to get from the UK to New Zealand, so of course they were transmitting files 2 months ahead of what was in my magazine. :(

Bruce Abbott
  • 6,673
  • 15
  • 31
  • 1
    That's a "software modem" - a program acoustically coupled to the phone, which MOdulates and DEModulates digital signals to and from "sounds" that can actually be transmitted by the telephone. It doesn't really matter if you buy a "modem in a box" or use an existing computer as a modem - it's still a modem. (Which is the problem with this question). – dirkt May 31 '19 at 11:40
  • No, it's not a MODEM because it only does DEModulation, not MOdulation. A MODEM has to be capable of doing both. – Bruce Abbott Jun 01 '19 at 00:42
  • It's only half a modem, then. But adding the other part is not hard. (And that wasn't the point - the point was that it doesn't matter if you do it in software or in hardware, and it doesn't matter if you do it on a dedicated device or a general-purpose computer - it's the same thing. In fact the modems you could buy at later stages are "computers" with a microcontroller and hardware to offload some or most of the modulating/demodulating work from the main CPU, and leave the CPU to deal with somewhat higher protocol levels.) – dirkt Jun 01 '19 at 04:30
  • Half a MODEM isn't a MODEM, no matter how much you try to redefine it as one. A tape player plugged into the ear socket of a ZX Spectrum is not a MODEM, and no amount of software can turn it into one. The OP specifically asked about using a tape player with "only a bit of extra wiring" instead of a commercially available modem, which is precisely what I implemented. Saying 'no you can't do that because then it is a MODEM' is not only wrong, it's silly. – Bruce Abbott Jun 01 '19 at 09:49
1

When I was a student in late 90's, I did something like this as a research / hobbyist project. The hardest part was the hardware interface to the phone line which had to be completely galvanically separated from the computer, and that also was tasked with amplifying the analog signal coming from the phone line so that it could be reliably detected by the built-in comparator chip.

As for transmitting data, one could use a Manchester code or simple frequency modulation but the transfer rates would be limited to something like up to 600 baud, I guess.

DmytroL
  • 2,162
  • 1
  • 12
  • 20
  • 2
    You built a modem. Homemade non standard but still a modem. – manassehkatz-Moving 2 Codidact May 31 '19 at 13:25
  • 2
    @manassehkatz depends on what you call a modem :) The external hardware I built did not actually perform any modulation or demodulation - it was merely an analog adapter to connect the computer's tape recorder port to a phone line. To be precise, of course it was a bit more complex than "only a bit of extra wiring and maybe some resistors", but the additional complexity was purely intended to achieve full galvanic isolation and avoid any damage. – DmytroL May 31 '19 at 13:52
  • 1
    If you only recorded an audio signal to tape using the computer as a controller, with no interpretation/decoding of the signal at all* * then arguably this wasn't a modem. But if you did all that and then played the tape back and decoded that signal, you have basically made a delayed-action modem. In the end, it sounds like the goal was the same - take a digital signal and transform it to analog and back to digital in order to transfer data from one computer to another via a phone line. That's a modem. – manassehkatz-Moving 2 Codidact May 31 '19 at 13:59
1

It was quite common to communicate to a computer without a modem, teletype interfaces worked over the phone lines and did not used modems. They switched up to 20 mA on up to a 48V phone line. The fact that old phone lines were DC-coupled (your handset was powered from the local exchange) was one of the reasons the phone companies tried to keep strict control over all equipment you could attach to the phone socket, making it illegal to connect uncertified devices and in many cases they insisted on renting you the equipment. Up through to the late 1970s the typical console of a computer was a teletype machine.

I used an IBM System/7 in 1973 to receive data from rotary phones in a factory. Touch tone phones did not exist yet, and they put dials along a Ford production line so that workers would dial in the number of an approaching chassis so we could send them a print out (teletype, outgoing lines only) of the options to be added.

The S/7 had a number of options for digital and analog I/O in factory and laboratory work, and in this case I was using a module which could report open/closed circuit which was a variation on the teletype interface which could be programmed with different timings. The dialing mechanism was much slower, if I remember correctly the pulse rate was about 10 per second and you just counted from 1 to (1)0, then a longer pause indicated the end of the digit.

Early time sharing systems used teletypes over the phone system, essentially just generalizing the adjacent console mechanism to support multiple phone lines each of which had a teletype at the other end. I also used those in 1973 to run APL on a time-sharing service, remotely in another city.

I also used signalling like this in early 1981 to connect a Victor 9000 to a brand new IBM PC (we had them talking the same day the PC was unboxed) and although the outputs were only toggling 12V the circuits were most likely compatible with the telephone voltages, as that was a common signalling standard at the time. I suspect some folks may have programmed PCs to work with teletypes including answering the phone lines. Modems were still expensive and rare, while many teletypes were in use, almost every medium business had a teletype machine (like more recently they would have a fax machine).

Tanj
  • 11
  • 2
  • Was it really possible to attach a current-loop mode teletype directly to a public phone line? I had assumed that the phone companies would put a stop to that. If this was possible, which teletypes for example were certified to be connected directly to the phone line? What about the equipment on the host side, was that also directly connected? – dirkt Aug 25 '19 at 06:35
  • Looking it up, of course you could connect teletypes to the Telex-network or similar services, but those were normally separated from the (voice) phone line network. However, the Bell Teletypewrite Exchange Service shared line with the (voice) phone network (POTS), but was "set up with a special Class of Service to prevent connections from POTS to TWX and vice versa. " Given that the point of a modem is to transfer data over POTS, I'd assume that Telex-like connections don't count for this question. – dirkt Aug 25 '19 at 07:03
  • Another way of connecting a computer to the Telex network was to punch the word processor to papertape. The computer output via Telex machine over telephone lines generated papertape for a remote system.

    Lawyers needed to reliably transmit documents sent by Telex without typos. A solution was to punch Baudot code from CP/M and Apple II systems. While working for a S100 system reseller I built interfaces and wrote code so that the punch was effectively a printer. Don't remember what model of papertape punch was used and certainly never kept any of the company's circuit diagrams or code.

     – PDP11 Sep 11 '23 at 05:04
0

As the question specifically states that this is about common home computers of the early 1980s, the answer has to be no.

To interface with the analogue telephone system, a modem is required (internal or external). Provision of an internal modem would add to the cost (both in bill-of-materials and in testing and regulatory compliance), and networking was such a minority interest at the time that this would not be justified by the demand. On many systems, even a standard serial interface is an optional extra.

The nearest you could get without a proper modem would be to abuse a cassette interface to drive an acoustic coupler (but even between a pair of like systems, you'd need to write your own driver, as the bandwidth of a telephone line is much narrower than that of cassette tape).

It turns out that the lack of provision of internal modems was probably a blessing in disguise - separate modems connected by RS-232 cables allowed innovation by modem manufacturers to advance transfer speeds without significant changes to the home computers driving them.

Toby Speight
  • 1,611
  • 14
  • 31
0

The kind of circuitry needed/used to drive a tape recorder that only offers audio connections, for purposes of storing digital data from a computer or similar device ... IS A MODEM. Even if is not a (purpose built) telephone modem (which is usually implied when people speak of modems these days).

rackandboneman
  • 5,710
  • 18
  • 23
-1

A direct electrical connection between a phone line and a computer would immediately destroy the computer, because the phone line carries a 50V DC supply to power telephones connected to it.

It's possible to get over that hurdle, of course, by inserting an isolating audio transformer or an acoustic coupler into the connection. You must then contend with the fact that a phone line will typically carry only signals in the 300-3000 Hz range, so you must arrange for the signals to be emitted from the sending computer to fall into that range. This is, in fact, precisely what a modem does.

If you do not have a modem designed for use on telephone lines, but you do have a tape-cassette interface on both machines and a suitable electrical-isolation device as noted above, then it is likely that the tape-cassette interface will work sufficiently well as a modem. Not only is the problem of storing signals on tape reasonably similar to the problem of sending signals over a phone line, but in many home computers the very same techniques were used. The 1200/2400Hz FSK signal used by the Kansas City standard is nearly identical to that used by half-duplex TTY-modems of the period (Bell 202 or ITU V.23).

If you do not have suitable hardware, then it is possible to encode the data in ways that are easy to read out loud, or send by hand in Morse Code, requiring only a proficient operator at the far end to transcribe these voice or CW signals back into the computer.

For example, 5-bit digits can be represented using Base32 encoding and printed out in four-digit groups, two such groups conveying five 8-bit bytes; Base32 is specifically designed to resist transcription errors by humans, as similar-looking letters and numerals are mapped to the same meaning. Similarities in the pronunciation of letters (or differences between national conventions for same) should be disambiguated by using the NATO phonetic alphabet, if reading the digits out loud.

Chromatix
  • 16,791
  • 1
  • 49
  • 69
  • You are aware that many older computers (particularly those using valves rather than transistors) had circuitry well in excess of 50 volts, right? As normal I/O, without any damage. – Toby Speight Aug 09 '19 at 12:42
  • @TobySpeight Among the parameters in the question is "For the common affordable home-computers in the years approximately 1978 - 1984 (a few years before, or after, is ok)" - so that means NMOS and CMOS chips running at about 5V levels, which would not withstand 50V for even a moment. But most of them did have a tape interface. – Chromatix Aug 09 '19 at 16:05
  • Ah, sorry, I missed that bit. – Toby Speight Aug 09 '19 at 16:07
  • 3
    I'm confused over why this is getting so many downvotes. If you do so, please leave a comment explaining why. – Chromatix Aug 12 '19 at 05:09
  • I didn't down vote this answer but do have reservations over the opening statements. For example, 20mA style interfaces for computer systems even predate the IBM PC. It is possible to connect a computer to a 48V line without it being immediately destroyed. Therefore your opening statement in Bold characters about immediate destruction would instantly attract downvotes. My guess is a number of readers would not be able to get past that bold, initial statement in your answer. As I said, I didn't downvote your answer and this is just my best guess as to the thoughts of others. – PDP11 Sep 11 '23 at 05:20