Basic Drive Train Drawings

[Sachiel7]

Well, If you think I was saying that every bot should be 4 wheel from the name, that's not what I meant.We just call this standard drive as opposed to our multi drive. Anyway, by no means do I think every team should be 4 wheel. I'm just giving an example of our system.
I mustv'e seen sparky dancin
Once you get used to building 4wd systems, It's not any different from 2 wheel, really. But I guess that really depends on your preference.
The wheel in the back eliminates all possible turning anomalies that might occur. In that design, we took the 2 CIMs, ran them through a gearbox to the wheels. They were 8" pneumatics, I think...
Anyway, the reason for using 1 caster instead of 2 is the way the drive force is executed. When turning, if you have 2 casters, they each will want to point a separate way, if they're behind the drive wheel on the opposite side of the frame. This can cause problems because you'r introducing opposing forces in your drive direction.
Here's an experiment: Find one of those office chairs that have small casters at the bottom. If you find one with an even number of wheels, let me know. The chair I'm sitting in right now at my computer has 5. And, IF it does have and even number of wheels, are they layed out in a perfect square like pattern? Chances are that you will have a very hard time finding this type of office chair. If you do, try sitting on the front edge of it, pushing yourself forward with your feet. Now turn, and try to go straight. You probably won't be able to do it without a major arch in your straighting out. The Reason:Triangular force.
My Chair's casters are set up as thus:
\_/
/|\
Now the angling in that little drawing is way off, but to the point, there are no 2 casters directly across from each other. Theres always 1 across from a pair of 2, forming a triangle.
Basically, the force of both wheels (the drive wheels) gets put on the one caster, and it finds a sort of median in the force and pivots to suit that position. It works pretty good.
You could try this with the edubot. Make 2 little casters (make sure the wheels isn't centered around the pivot point) and try some different configurations. You'll find that single centered casters tend to work more. They absorb both forces of pull.
Anyway, as far as frame stability is concerned, there were no problems at all. Since the caster could pivot around, it usually held up the side against the opposing drive force. It also helps to have your weight more centered on the bot with that design, so if it does tip, it should re-level itself up again.
I really liked Sparky 4 this year. Your self-righting device has inspired several of our teammates (and me too ) to brainstorm ideas for devices that aren't always game-related (self righting devices included)
I'll see if I can dig up some of the pictures from MD so you can see the bot.