This keyboard was sold as a wooden-cased creation from the early hobbyist computer days.
Maxi-Switch-216004-case.jpg (426.13 KiB) Viewed 1759 times
Which means that the seller had a somewhat inflated idea of its worth. Though not as bad as the one asking 35 G's for a Hi-Tek grid missing two keycaps because it resembles part of a complete working system with flawless provenance, and even though theirs has none. But more importantly, there was some flexibility in this one's price, so here we are.
So, what is it?
First off, none of it is wood. The case is sheet steel. The bottom is thicker and black. And the top has wood-grained vinyl contact paper on it. Which gives it kind of a '70s low-volume light-industrial vibe.
Second, it isn't for a hobbyist system. One must remember, of course, that those ASCII keyboards that got repurposed, often from surplus, already had a market.
Spoiler:
It's the programmer module for a One-Up Inc early LED sign system. I can't find a good picture, but I think you can recognize it here.
One-Up-advert.jpg (28.53 KiB) Viewed 1759 times
And this is confirmed by the silkscreen on the smaller PCB.
One-Up-top.jpg (437.33 KiB) Viewed 1759 times
One-Up-bottom.jpg (373.5 KiB) Viewed 1759 times
Once again, some poor worker had to spend their days defacing the 74-series TTL chips. And, once again, one can immediately work out what they are from what the device does. For instance, all the parallel data lines go through alternating pins of a pair of chips. So those are hex inverters, as inverting twice was a common trick for cleaning up signal edges or converting to open collector.
I bet the rest is for turning certain key presses into separate signals.
Anyway, that isn't part of the keyboard proper and there is no sign of the sign module, so I am going to ignore it. Just as one might the driver card for a terminal when the CRT and HV circuitry has gotten separated.
The main PCB is Maxi-Switch part 216004 with linear switches.
Maxi-Switch-216004-top.jpg (380.54 KiB) Viewed 1759 times
Maxi-Switch-216004-bottom.jpg (441.65 KiB) Viewed 1759 times
The encoder is an SMC 3603-024, which looks to have the same pinout as the KR3600 that was used by The Keyboard Company, among others, so there's a somewhat real hobbyist connection. The ASCII mapping (3600 bits worth) does not seem to be custom. Other than that, there's just a one-shot for the strobe signal with passives for timing and a 5-to-12V converter to get the -12V the encoder needs.
The switch is for selecting whether bit 6 of the output comes from bit 6 or bit 9. The encoder offers both bit shifting and alphabetic shifting and this chooses which. So, it's similar to the switch mod on the TKC encoder board that let it output lowercase if you had the right ROM to handle it.
Apple-II-switch-mod.jpg (610.26 KiB) Viewed 1759 times
Conversion just means a virtual serial port; all there is is ASCII. But what about the cabling? The longer ribbon cable can't be taken out of the case without cutting it and that's against the rules (that I follow). So, instead, I replaced the daughter card, with two IDC headers, by a straight header between the two cables.
Maxi-Switch-216004-splice.jpg (430.9 KiB) Viewed 1759 times
A Teensy for the other end.
Maxi-Switch-216004-converter.jpg (467.57 KiB) Viewed 1759 times
Impressive stuff. I also have a "mystery" sign programmer board that I'd basically given up on converting. It's easily the most 80s board I've seen with an angular black plastic case and an LCD and card slot.
It even has ICs covered in grind marks for some reason...
Any idea why they did that? To keep people from reverse engineering it or servicing it out of warrenty?