custom optical mouse to trackball conversion?

[Tweaked]

Has anyone attempted to use the bottom side of a new optical mouse to make a custom or home made track ball?

To me it seems much easier than the mods that use the old style ball mouse with optic sensors.  What I have in mind would be basically mount the mouse upside down and backwards under the control panel so that the track ball sits just off its surface supported by 3- sealed ball bearings.  This could easily be held in place by the control panel face if beveled correctly and would not require much if any mods to the mouse.  My major concern would be if it could track fast (or slow enough depending on the speed of the mouse) enough along the surface of the track ball.

[u_rebelscum]

Has anyone attempted to use the bottom side of a new optical mouse to make a custom or home made track ball?

To me it seems much easier than the mods that use the old style ball mouse with optic sensors.  What I have in mind would be basically mount the mouse upside down and backwards under the control panel so that the track ball sits just off its surface supported by 3- sealed ball bearings.  This could easily be held in place by the control panel face if beveled correctly and would not require much if any mods to the mouse.  My major concern would be if it could track fast (or slow enough depending on the speed of the mouse) enough along the surface of the track ball.

Yes, might find them using search.  Came down to ball round, table/mousepad flat; ball very uniform, table/mousepad fuzzy that optical sensors can use.  One or the other might be okay, but with both, you need to perfectly position the ball to exactly the distance the sensors are designed for and use a not shiny polished ball.

Others have tried a gave up.  With the newer tech, maybe you can do it, but it's not as easy as you think.

[Bgnome]

also, if you are mounting an optical mouse sensor upside down, you will have an axis problem.  it might be easier to start with an optical trackball..

[Tweaked]

As for the axis theory, actually I tested my theory using a Coke can.  Set it flat on the sensor with the mouse wire facing the user. and one you roll the can forward like you would the track ball the correct axis goes the correct direction and I did the same thing for left/right axis.  It seem to work.  It tracked fine during this initial testing...  However I did not have a  polished or smooth ball available.

I guess I'll give it a go when I get home, won't be for a little while as I am currently in Iraq.  I am just getting some ideas together to finish a project I started planning prior to my deployment.

Thanks

[u_rebelscum]

Check that backwards axis again.  One or the other is backwards.

[Trimoor]

I posted this idea about a year ago.
The axis thing tuns out to be normal, no reversal needed. (mame can do that anyway though)

The mouse I tried doesn't like my white imperal though.
It registers nothing.

[jswhale]

I have an optical mouse working with a spinner I made out of an old HD.  I had to use a design stuck to the spinning plate (actually one of the HD platters!) to ensure it was reading correctly.  I think you'll have the same problem with a trackball since it's harder ensure that the optics can pick up movement accurately - particularly with the faster speeds you will find with a trackball.