Cartesian coordinates and decaying circles.

Wolfbrother

Hello all, I'm playing about with the BOE-BOT and found that by trial and error and a bit of acceptance that using PULSOUT commands you can approximate a circle. What I am wondering is has anyone used the basic equation for a circle and coded it for the BOE-BOT. I also then want to modify it and have it decay both linearly and quadratically to approximate different arcs. Has anyone done this and have any advice?

Thanks,

Dave

Beau Schwabe

Wolfbrother,

I assume that you are sending differing pulse widths to the servos to create a slow turn or drift in one direction or the other. The problems that you will see are manufacturing differences between the servos that will create large non-deterministic errors that will combine with varying frictional differences between the wheels and the floor.

Wheel encoding can solve the differences between servos, but you will need to rely on other mechanisms to account for the friction loss.

But to calculate the arc, each wheel basically travels the radius (which can be an imaginary point) of the circle. The larger the circle, the smaller the rotational difference between the wheels. likewise, the smaller the circle, the greater the difference between the wheels. In both circumstances, the outer most wheel from the radius will travel the most distance. To go in a straight line would be a circle defined with an infinite radius.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Post Edited (Beau Schwabe (Parallax)) : 4/14/2009 4:40:38 AM GMT