Fully optical spinner or trackball

[clhug]

First, I know all mice and trackball encoders are "optical" in a sense, but bear with me...

This came out of the old "SlikStik Tornado Spinner News" thread from a month or so ago.  While it had some debatable discussion about specific products, embedded in it was a good discussion on physical resolution of the encoder for a spinner.  It talked about how an Arkanoid spinner had very high resolution because of gearing, but that causes a "rough" spinner.  I suggested a belt driven idea that would provide a gearing type situation but smoother operation.

I recently thought of this additional idea.  Has anyone tried to build an encoder for a spinner or trackball based on the newer purely optical mouse technology where there is no moving ball or rollers but instead looks at the reflection of light off the surface that the mouse is sitting on?

I envision still having an encoder wheel, but instead of being spoked that light shines through, the wheel would be solid with some kind of surface, and the emitter/reciever unit of the mouse would be mounted over the wheel.  As the wheel turns, it looks to the emitter/receiver like the mouse is moving over a surface, only the surface is moving instead of the mouse.

I have no idea if this could really work but it seems to me if it could, you could achieve much higher physical resolution since it's not dependent on how close or how small the physical spokes on the encoder wheel are, but instead on how pinpoint accrurate the opticals are in the mouse emitter/receiver.  I've always heard that "optical" mice have a much higher resolution than old "mechanical" mice but I don't have any idea how much higher, or if that marketing hype is even really true at all.

Just wanted to throw this out to see what others thought of the feasibility of such a thing.

[woodygjw]

Did a search and found an old thread, and it doesn't seem to work well...

http://forum.arcadecontrols.com/index.php?topic=11907.0

woody

[clhug]

Cool, thanks for the pointer.  I went ahead and read through those threads just now.  They had a slightly different idea than I have, but I think they're all right that the key to making it work is the right surface for the optics to reflect off of, and the proper distance from that surface to the optics.

Most of those other messages were about trying to use a single mouse's optics to run both directions of a trackball.  Theoretically I agree that should work, but practically I see how they're probably running into the problem of the distance, and the surface of the trackball not being optimal.

What I'd do is still keep the axes separate.  Even though a single optic can do both directions, I'd still keep it separate.  I'd still have the trackball (or a spinner) spin encoder wheels for each direction.  But I'd replace the traditional spoked encoder wheel with a solid wheel and some type of surface (like a paper sticker or something more permanent) on top of the wheel.  The exact composition (color and pattern) of the surface and distance of mounting the optics from the surface of the wheel would have to be determined by experimentation, but once figured out you could permanently mount the optics at the proper distance.

Also, others seemed to be trying this as a way to save money.  I'm not thinking about money at all.  In fact, it could be that it would be more expensive in the long run for a trackball because I'd essentially be using 2 entire mice equivalents (4 axes) to drive only one trackball (2 axes) (or a single 2 axes mouse to drive a single axis spinner).  For me it's purely about resolution or sensitivity.  There was one guy in those other threads who kind of got it to work and even said that it was much more sensitive than the regular encoders (mouse moved much farther with the same movement of the ball than just through the normal encoders), so I'm inclinded to think I'm on the right track.  Just seems the key is figuring out the optimal color and pattern to make that wheel and the proper distance to mount the optics from the wheel.

[MinerAl]

http://forum.arcadecontrols.com/index.php?topic=5232.msg34552#msg34552

Looks like it has been tried with a spinner.

Seems like this solution, if it could be made reliable, could be the answer to the up/down problem.  If the pattern on the drum was tall enough you could slide the drum up and down without needing to move the optics.  I guess that Randy tried it with older optical technology.

I tried it with a multicolored track ball and an old apple optical mouse.  It worked, but you had to reverse one axis in the applications for it to track correctly.  It too was sketchy though.

Might be a different story with the new high resolution lazer mice though.

[Kremmit]

There was this one a couple months back:

http://forum.arcadecontrols.com/index.php?topic=50654.0

Seems to me that a guy could adapt this method to make a variable-sensetivity spinner.  Instead of using a little old bolt head, use a larger disc.  Mount the optics such that you can adjust how far from center they sit.  The closer to the center of the disc the optics are, the less surface will travel past the optics for each rotation of the disc.  Move the optics further out, and more surface passes per rotation.  Voila!  Just slide the optics out for Arkanoid, and back in again for Tempest.

[clhug]

VERY COOL!  Thanks much for that pointer.  That's about EXACTLY what I was thinking of.  Yep, my thought would have been to put a big washer on it, but if it works for him with just the bolt, that's great too.  You can always turn down sensitivity in the software.

I guess I can move any further discussion on this to that other thread.