5 ms·
Reverse-engineering an unusual IBM modem board from 1965
- kens 5y agoAuthor here if anyone has questions about vintage IBM hardware :-)
- Animats 5y ago600 baud? FSK at 600 baud is marginal. It would be interesting to see the receive side.
- kens 5y agoLooking at the waveforms, you don't get much of one frequency before it switches to the next at 600 baud, so receiving seems challenging to me. Unfortunately, I don't have any information about the receiving circuity. This particular configuration is limited to 8.25 miles, though.
- mjevans 5y agoSynchronization between bits is probably the hardest part; offhand I don't know how to begin looking at that. Within a given bit, given the properties described of the waveforms (one notably less than half the frequency of the other) as well as the tolerances involved; maybe something edge or center-crossing focused. Possibly use the peek or zero-crossing as a switch on a charge accumulation counter and measure with a comparison amplifier of any kind.
- NovemberWhiskey 5y agoThe Bell 202 modem standard is FSK at 1200 baud (albeit introduced later in the 70s) and the standard actually continued to be popular into the 2000s for communication between process automation systems. 1200 baud AFSK is also the baseband encoding for the popular APRS amateur radio standard.
- Animats 5y agoAh, the FSK frequencies were moved further apart. Bell 103A: 1,270 Hz and 1,070 Hz. 600 baud with that narrow a shift would be iffy, especially before phase-locked loops. Bell 202: 1200 Hz and 2200 Hz. With a full 1KHz shift, 1200 baud could work.
- jonsen 5y agoConsidering the relative simplicity of this transmitter circuitry the receiver would probably have been comparably simple. Could be two resonance bridges (two capacitors and four resistors each) tuned to the two reference frequencies respectively. A simple rectifying of the bridge outputs would give a differential pulse sequence which would be fed to a comparator. From there it would be bit detection and synchronization as is done in a typical UART.
- Animats 5y agoThe problem is getting much out of a simple filter after just a few cycles.
- jasonjayr 5y agoWas this unit capable of demodulating the waveform? Or was is a one-way transmitter?
- kragen 5y agoThis is the mo-. The -dem is another board, or more.
- jes 5y agoI enjoyed the article. Acoustic couplers were common back in the day as well. I used an acoustic coupler a couple of years before I was able to buy my first Hayes modem, which was a 300 baud modem. The coupler was connected to a Zenith Z-19 24x80 CRT-based data terminal. The standard telephone handset was a friction fit into the two earmuffs on the coupler. To use the coupler, you dialed the remote computer by hand (rotary dial phones worked as well as touch-tone phones) and when you heard the whistling tone from the remote end, you shoved the handset into the earmuffs on the coupler. A few seconds later, you’d get a (sometimes garbled) login: prompt.
- jcun4128 5y agoI watched a video by Cathode Ray Dude on this pretty neat "AT&T's '60s Modem That Won't Die" at 19:20 He demoed it transmitting data on a computer
- ahartmetz 5y agoI have some doubts about "it's heavy so it's probably oil-filled". There are two common materials in transformers that are much denser than oil: iron and copper. There may or may not be oil in addition.
- BBC-vs-neolibs 5y agoMaybe the heuristic was, "something this heavy must be power dense, hence may need oil cooling".
- h2odragon 5y agoDunno about "oil" either, but pre-70's phone network stuff was weird. Bell was cagey about what could connect and how; i think partially because they had trouble keeping nasty transients off their wires. This transformer very likely didn't connect directly to a phone network; I would guess. It wouldn't surprise me if this was capable of being connected to nearly any lash up "intercom" type line too, which would be a place to have a robust output.
- jpm_sd 5y agoVery common, frequently toxic https://en.wikipedia.org/wiki/Transformer_oil https://en.wikipedia.org/wiki/Transformer_oil
- guenthert 5y agoVery common in power transformers as it is used for cooling (heat transfer). In an isolation transformer used in a modem cooling seems unnecessary.
- zw123456 5y agoBack before the "Carterfone" decision https://en.wikipedia.org/wiki/Carterfone https://en.wikipedia.org/wiki/Carterfone in 1968, the phone company had a rule that you were not allowed to connect anything to their network of your own. You could only use equipment they provided that you leased from them. Hence, the acoustic coupler was used to connect your fax or computer to the phone network. The exception were special services private lines, since this device is circa 1965, it was probably intended for that. Before the internet, big companies would set up big private networks, typically Bisync or later for IBM SNA. Those private line networks were like a big conference line, sometimes referred as shout down circuits. That name came from the original use which, a common example would be a bunch of junk yards would each have a shout down and they would yell into it things like "we need a 59 Chevy Radiator" and then a guy at another junk yard would yell back "this is Bob and we have one" or whatever. These circuits were basically a giant conference bridge but not switched, it was permanently wired. These circuits later got used by computer companies by putting a "master " modem at the head end which would send out a message here is some dat for Terminal 123, which all of the modems would receive and then modem 123 would respond with some data, the protocols used were polselect or bisync but there were others. I am guessing, based on the date, that is what that modem would have been used for. Some of those goofy lines on RS232 ( or RS422) like DCD, RTS, DTR etc were used to control the modem and tell it when it was OK to raise it's carrier and reply, otherwise it would just listen for it to be called on.