My team is toying with drivetrain ideas .. .

[Frank(Aflak)]

you be the judge. This is the most radical (that we are considering)


You have two wheels in front. They are your drive wheels. They are powered through a differential . . . . .

You have one 'wheel' in back. like monsieurcoffee's teams omniwheel from the gallery:

http://www.chiefdelphi.com/forums/pi...&quiet=verbose

the wheel is perpendicular to your direction of motion . . or whatever, its just sideways. Ours would probably not by that robust as it is not going to support very much load.



Anyway, the pluses we see with this drivetrain are as follows:
1- low energy turning. We only need to spin a smaller motor to turn, unlike a tank tread bot where you need to use the same power to turn as you do to go full speed.

2- need for only one transmission . . Your wheels are not driven independantly, so only one transmission/gearbox is needed.

3- your axis of rotation will be the midpoint between your front wheels when you are stopped. with some designs we are playing with this will allow us to spin aroung a box to acheive a better angle without even touching our main drive.

4- did i say you only need one transmission? cheaper.



possible drawbacks

1- you need to build/aquire a differential. If you aquire it and it fails, diffiult/impossible to replace/repair.

2- the three wheel design may be prone to tipping, especially when on the ramp.

3- an omni wheel, as monsuier(however thats spelled) coffee calls it, will need to be machined. . . . we would probably use rollerblade wheels or something, tho, it would need to be able to get across those little 'lips' on the start of the ramp easily.

So, did I just describe something you already have been using?

is it feasible?

should we do it?



I really agree with my physics teacher: he saw us turning our edubot (4 wheel drive version) by reversing one side . . and laughed at how inefficient that was.