TRS-80 Model 1 without numpad

MMcM

10 May 2021, 06:07

This keyboard was sold as something else, but the layout is pretty recognizable.
TRS-80-top.jpg
TRS-80-top.jpg (336.11 KiB) Viewed 1430 times
This particular board had it rough before I got it. The PCB was broken off in the upper-left corner. The FFC cable was just yanked off and had contacts twisted up; I just desoldered them and replaced it all with some headers and an improvised cable. The power LED was cut off; I grabbed some random red replacement one from a drawer. The space bar looks to have been cleaned with something abrasive. A bunch of keys have gouges. On the 3 key they are regularly spaced, perhaps from being stacked beneath another PCB.

The Hi-Tek grid is actually in decent shape; only a few of the corners are split.
TRS-80-keys.jpg
TRS-80-keys.jpg (1.14 MiB) Viewed 1430 times
Honestly, I don't think this was ever a nice keyboard to type on.

What is interesting, though, is some of the engineering decisions that went into the design here.

The keyboard did not have its own controller; it was scanned directly by the PC. Moreover, this is done on a common bus, with shared address and data lines. The address is inverted into matrix rows. The matrix columns drive an inverting open-collector buffer. A single keyboard-specific signal enables the buffer.

All of this could have been done on the main PCB, with the keyboard subassembly being a bare matrix, as it is in some other computers of this era. But there's space and the signal quality might be marginally improved.

The matrix itself is also much more complicated than just a straight grid, with extra traces all over the place and a peculiar order for the header signals.
TRS-80-bottom.jpg
TRS-80-bottom.jpg (1.38 MiB) Viewed 1430 times
Why this is becomes evident when looking at the schematic.
TRS-80-schematic.png
TRS-80-schematic.png (432.66 KiB) Viewed 1430 times
Note that the board here does not match that schematic exactly; it's electronically equivalent, but a few of the inverter / buffer pins are different. I suppose that means there was more than one revision of this PCB.

Anyway, the particular row, column assignments of the keys make decoding simpler, possibly avoiding a lookup array (though I don't know that it actually worked without one).

Not that this makes USB conversion any harder or easier; just maybe the wiring a little messier looking.
TRS-80-converter.jpg
TRS-80-converter.jpg (605.65 KiB) Viewed 1430 times

Post Reply

Return to “Workshop”