deskthority

Does anyone know how does topre (or other capacitive ou contactless switches in general) implement hysteresis?

Is it done on the firmware? Or do they not have hysteresis at all?

Topre doesn't really have hysteresis. If you look at various force graphs, you can see that the operating position and release positions are in basically the same spot. Compare that to the graph of a buckling spring switch and you can see that the difference is night and day, really.

Topre:


IBM:

I see, thanks!

Topre makes hysteresis by changing threshold of sensor chip TP1684/1683, which has two different thresholds for each down and up stroke. Keyboard controller firmware controls the threshold depending on key state(down or up stroke).
On Topre key sense value changes almost linearly, you would have teribble chatter without hysteresis. I can't refer to any exact numbers or graph, but I won't say Topre hysteresis is very small in comparison with other mechanical switches like alps cherry .

Meanwhile capacitance changes drastically and undetermined state can't retain long due to buckling spring mechanism. I'm not sure IBM capacitive sense internals but they won't need hysteresis in controller firmware and sensor chip unless key matrix is scanned in very fast rate. In other words hysteresis of IBM comes purely from bucking spring mechanically.

This is my research log on geekhack.
https://geekhack.org/index.php?topic=76 ... msg1894368

So there is a separate sensor for each key that is controlled by the firmware?

I would like to add plural of hysteresis is hystereses.

Thanks for the info 002.

It should be easily possible to implement hysteresis. I would assume capacitive keyboards use Schmitt triggers, which can be set to un-trigger at a different level than at which it triggers. The same principle applies in Hall effect and similar keyboards.