RS-232 Then and Now (Notes and Discussion)

[voidstar78]

For #2 -- I'll stand corrected. I thought I saw a reference of this using "300 baud" on its keyboard, in one of the Kaypro manuals. But reviewing <https://bitsavers.org/pdf/kaypro/1484-D_KayproTechnicalManual_Dec84.pdf>, yea. So I better example of a (possibly?) RS-232 keyboard was maybe the ASCII Keyboard (featured in April 1974 Popular Electronics <https://archive.org/details/popularelectroni05unse_3/mode/2up>). Main point I was aiming for was that there was at least one RS-232 keyboard (maybe the POLY-88's?).

For #3 - I thought RJ45 is a "physical connector with eight pins and eight positions" (8P8C)? Or when pointing to a wall ethernet "outlet", people tend to say "RJ45 port" or "RJ45 wall plate" ? But yea, getting into the more modern and advanced means of data exchange is a whole other topic (hence why in the backup slide section, for future thoughts).

Thanks!

 

[cjs]

For #2 -- I'll stand corrected. I thought I saw a reference of this using "300 baud" on its keyboard, in one of the Kaypro manuals.

Oh, it is indeed using 300 baud! It's just not using RS-232 levels; it's using TTL levels, which is why the keyboard socket connects directly (almost—there are a couple of buffers) to the UART.

To be super-clear on this, as you write on page 18 of your presentation:

• “The standard does not define such elements as the character encoding (i.e. ASCII, EBCDIC, “, etc.
• The standard does not define “the framing of characters (start or stop bits, etc.)”
• The standard does not define “transmission order of bits, or error detection protocols.”

So it's quite possible for me to define a serial communications standard (say, mark=1/space=0, start with space-mark-space, 8 bits of EBCDIC with MSB first, 4 stop bits required) that I can run over RS-232 and yet no "regular" computers will be able to talk to it. Conversely, I can use the Bell standard serial protocol not over RS-232, and a simple set of level converters such as MAX232 ICs will give RS-232 that will work with "regular" computers.

So I better example of a (possibly?) RS-232 keyboard was maybe the ASCII Keyboard (featured in April 1974 Popular Electronics <https://archive.org/details/popularelectroni05unse_3/mode/2up>).

That one is a parallel ASCII keyboard; not serial at all.

I suspect that there were very few, if any, RS-232 keyboards. Given that a keyboard that generates ASCII generally has some TTL (0/5 volt) circuitry on it, and it's connecting directly to a machine also running at TTL levels, changing the 0/5 V to -12/+12 V on the keyboard and then back again at the machine it's connected to really doesn't make a lot of sense.

Main point I was aiming for was that there was at least one RS-232 keyboard (maybe the POLY-88's?).

Nope; according to the manual (p.26 §4.4) it's also a parallel keyboard.

For #3 - I thought RJ45 is a "physical connector with eight pins and eight positions" (8P8C)? Or when pointing to a wall ethernet "outlet", people tend to say "RJ45 port" or "RJ45 wall plate" ?

A lot of people incorrectly call those Ethernet plugs and sockets "RJ45", but it's not. Wikipedia explains this.

 

[voidstar78]

Came across this in DDJ Nov/Dev 1976:




So, early SWTPC systems using an RS-232 keyboard?

 

[Hugo Holden]

A popular IC that IBM used an their serial interfaces was the 75154.

In the Votrax type and talk unit they used a "trick" for the serial interface. While the line drivers were the MC1488, which curiously they powered by +14V and - 5.6v (that I changed to -7.5v because I thought that was too borderline) the trick they used was on receive. They used a CMOS 4049 buffer IC with a series 39k resistor on its input, which converted the +/- 12 or whatever, volts to 5V logic and it did not harm the cmos IC because of its input protection diodes.

It was interesting trying to sort that Votrax serial interface out because it was neither a DTE, nor a DCE, it was a hybrid, I wrote an article about its vagaries:

https://www.worldphaco.com/uploads/VOTRAX_TYPE___TALK_MEETS_HERO_JR.pdf