Quote:
Originally posted by Frank(Aflak)
You see, with three wheel killough(kiwi), if you spin all three wheels you get a robot that is turning . . and you won't get all three motors adding forward power. With four wheel, you can have all four motors putting all their power into forward push. Whether or not you are powerful at that point depends on your traction, and polyurethane is a great sticky thing, and your gearratios, which are entirely up to the team.
So you see. You get to make the same choice between power and speed you do when designing a tank drive, but you get super maneuverability thrown it! No extra charge! Well, a little for the wheels, but not too much!
|
I direct you to my earlier post
here. Omniwheels allow lateral motion but prohibit perpendicular motion (to the drive direction). Therefore, only a limited amount of torque would be exerted. It's simple trigonometry. Only anywhere from 50% to 70% of the total torque of all 4 motors can be exerted in any one direction. If you look at the numbers, with a killough platform you get 57% of the total torque (all 3 motors). Of course, with the four wheel variant you have more motors contributing to torque, but it's not more efficient by any significant margins.
Quote:
Originally posted by Frank(Aflak)
In motion, some energy is lost to the omniwheels inefficiency, but if you buy 'real' omni wheels, as linked above by myself, you get nearly the same efficiency as say, a tank drive going forwards.
|
I don't see where you're coming from. How would you lose efficiency? You might experience some lateral friction since the rollers don't have individual bearings, but I don't see how you would experience any significant power loss. Other than that, omniwheels are omniwheels, as long as they permit lateral motion. I don't see what makes those omniwheels 'real'.
As for which, crab drive or an omniwheel design, is better, I don't know. It depends on your design strategy. If your strategy (or the game) requires torque over mobility, then a tank/crab drive is better, because all 4 motors contribute 100% of the available torque to the direction of motion. If mobility is considered to be advantageous over torque, then an omni-wheel design might be better. However, an omni-wheel design does present a programming challenge, or opportunity, depending how you look at it. One idea I had for last year was a saucer-shaped wedge robot and use a digital compass and programming to create relativistic motion (from the drivers POV). However, even though local magnetic interference can be compensated for (your own motors), other robots would be a challenge. You could do the same thing with an accelerometer, but the error would add up, and toward the end the robot may not go in the desired direction.