Re: [CAD_CAM_EDM_DRO] Hacking Optical Mice as position/distance encoders

Ian Wright wrote:

> Hmmm,
>
> Point taken Jon although I still can't quite figure out how it si done -
> I'll be pulling another mouse to bits tomorrow for sure!! The ones I've been
> into so far have just the one sensor per encoder but it does have three
> wires.

No, they are dual units. Quite likely two phototransistors with no base
leads brought out.

> I had assumed that these were simply normal phototransistors but
> maybe not. However, even if they are not and are two photodiodes back to
> back, they surely can't give real quadrature as I thought that required the
> photosensors to be spaced by half a 'line-width'.

Not back to back, but side by side. And, no, they don't actually have to match
1/4 the full cycle spacing. As long as they are far enough apart to keep the
transitions in the right order, it will count reliably in the right direction. Of
course, the quadrature timing will be way off that way, but who cares for
a mouse?

Anyway, for people playing with mice electronics, you should know this!
Mice are not designed to count every pulse! They try to keep up, but many
or all of them lose pulses every second. Some of them miss all pulses
while sending the results to the serial port. With a mouse, it doesn't
matter, as your hand/eye closes the loop.

Jon