Apollo 7121

MMcM

30 May 2021, 04:27

This is not a particularly nice keyboard to type on.
Apollo-case.jpg
Apollo-case.jpg (368.1 KiB) Viewed 1029 times
It uses a Micro Switch Silent Tactile membrane and is 106ST13-2E-J.
106ST13-2E-J.jpg
106ST13-2E-J.jpg (205.61 KiB) Viewed 1029 times
This particular board has some scratches.
Apollo-back.jpg
Apollo-back.jpg (290.15 KiB) Viewed 1029 times
There were also some broken keys that needed to be glued and even worse some that apparently already had something like that done to them and aren't straight any more.

On boot, it is 1200 baud, even parity, serial ASCII. As described in the Technical Reference, some function keys also send up transitions. But all the shifting is done inside the keyboard.

If that were all, it would end up in that pile with only a TTL serial to USB converter. But, as it turns out, this is only a compatibility mode 0 for use with older computers. It also has a mode 1 where every key sends both down and up transitions. CAPS LOCK is still handled in the keyboard, its up / down code effectively giving the on / off state of the LED. But other than that, there is no translation in this mode.

That is enough for a straightforward converter to USB keyboard.
Apollo-converter.jpg
Apollo-converter.jpg (514.4 KiB) Viewed 1029 times
And I admit that the left-hand navpad with its various kinds of arrow icons still has a cool tail-end of '80s feel from before everything became the same.

It also has a mouse port, which also uses a DE-9 connector, although it isn't a serial mouse. It would be straightforward to include support for this in the converter, but I don't hold out much hope of ever coming across an old Apollo mouse.

How the keyboard accomplishes all this is at once simple and surprising.
Spoiler:

The PCB, SC-15312, has two microcontrollers, both an MCS-48 8049-compatible and an MCS-51 8031.
Apollo-inside.jpg
Apollo-inside.jpg (875.61 KiB) Viewed 1029 times
Each one has its own EPROM, running the presumably closely related SD-03861 and SD-03863, and its own latch.

It's not quite as crazy as running two separate encoder programs and switching between then. The Honeywell membrane encoder SC23181 is the same chip as in the ST-serial APL keyboard converted a while ago. I still need to scope this chip in action, but I believe it is synchronous serial with pin 15 input, pin 27 clock, and pin 28 output. These go to 38(P27), 39(T1), and 36(P25) on SC-15189's 8048. Here they go to 36(P25), 39(T1), and 35(P24) on the µPD80C39. So essentially the same circuit. And it looks like that then talks to the 8031, which has a simple UART that in turn drives the serial output after a buffer, transceiver, and switch.

Without desoldering the ROMs to get the programs off, it's hard to say what the exact division of labor between the two programs is.

Post Reply

Return to “Workshop”