deskthority

I picked up this little terminal keyboard because it had the look of being from before everything was standardized or so probably with something interesting going on. The case is all metal: the top is aluminum and the bottom is steel. It wasn't a total surprise that I'd never heard of Bridge Electronics Company. I figured the keyboard was an OEM, but the PCB also has their name and offers a little more information, an address. That's enough to get more names, such as Bridge Data Co. and James M. Yates. And a phone number (800) 325-3282(DATA). In those days this would be much more durable than anything else. They show up in search as offering online real-time market information on Tymshare and CompuServe. And survived into this century in bankruptcy. By then, James Milne Yates (Wash. U -- that is, WUSTL -- '63) would have been around retirement age.

I'm not sure whether this keyboard went with a regular terminal that they sold to those without other means of accessing online services or with one somehow specially adapted to viewing their data. Or maybe a T20 was a terminal that connected to some kind of intermediate like a PDP-11.

Anyway, the switches are Hi-Tek high profile and the keycaps are okay double shot. The PCB, BEC-4, looks to have been hand-soldered, confirming some kind of low volume product. I figured it would be serial ASCII, having no special keys and only four signals. Rather than an MCS-48, though, it has a National Semiconductor MM57499 Serial Keyboard Interface. Given a suitably wired switch matrix, you get 1200 baud TTL with no custom programming required at all. I think I picked up a few of those from some surplus place -- cause you never know -- some time ago, but they aren't right now where I thought they should be. I don't think I have any other keyboards using one.

The serial input signal can be used to drive a shift register and thereby some LEDs. And, indeed, this PCB has space for a chip like a 74C164, seemingly wired per the example in the data sheet, and holes for eight LEDs.

But the keyboard only has two LEDs and at the bottom instead. The red one, power on, is lit all the time. The yellow one, caps lock on, is connected to the fourth wire of the 4P4C RJ connector. Which I guess seemed easier to the designer.

The key mapping has a couple of peculiarities. First, there is a little macro facility. Ctrl-Esc captures a few keys and Ctrl-; transmits them again each time.

Second, some of the shifts are reversed. Specifically, ^ ~, [ {, \ |, and ] }. No shift gives the top legend and shift the bottom one. And it isn't that this chip was specialized in some funny way. That's what the datasheet says. I don't have a good theory for why you'd want this. Space Cadet has unshifted braces. And Hyper7 has separate caret and tilde keys. But I don't think I've ever seen keys with the now standard legends the other way round, so I think you just have to know that shift means the bottom one for these four keys. If I find the SKIs I think I have, I might check whether they were always like that.

As for converting to USB for a typing test, for ASCII terminals with no special keys I usually emulate a serial port rather than a keyboard, since it's not going to have enough for GUI control anyway. For parallel ASCII, that needs an MCU. But here it's just a question of converting TTL serial to USB. Why not keep with the pre-fab switch grid, with minimal layout required, and pre-fab controller, with no programming required, and use a pre-fab serial converter? Like an FTDI. Of, even better, a CP2102, which is less than a buck on AliExpress. Since I couldn't monitor for the special Caps Lock changed codes, I just wired that fourth signal up to DTR. So now it's the online LED instead.

Ooo, this is nice. I think I’ll add it to my list of modern twist builds.