[quinxorin]

Responding to the previous few posts, here's the actual calculations:

CIM starting stall torque (the maximum amount of torque it can exert before stalling and not moving):

343.4 ounce-inches.

Converting to foot-lbs:
(343.4/16)/12 = 1.778 ft-lbs.

Force exerted by each wheel (assuming the wheel is 8in), therefore, is:
1.778 x 3 (4in, the radius, is 1/3 of 12in) = 5.366 lbs-force.

Apparently, each wheel exerts 5.366 lbs of force pushing forward.

Now, let's look at the friction properties of the wheels. Because the wheels are perpendicular, they will have to overcome each other's friction in order to move. At least 2 of the wheels will have to slide at all times.

We can assume that high-traction treads have a coefficient of friction (μ) of at least 1 on carpet, though it is probably much greater (think 1.7-2). I will use 1 as the estimate. An object (i.e. the wheel) is capable of sliding if the following inequality is true: F(force) x μ > weight. The robot, including batteries, bumpers, and the minibot, will weigh 169.2 lbs. Therefore:

5.366 x 2 (there's two wheels) x 1 > 169.2
10.731 >169.2

That inequality is definitely false. Therefore, your robot will not move. Sorry, but it's the sad truth. And that doesn't even include the actual force it would take to move the robot, just to beat the friction of your other wheels.

A lot of people have been posting like me, that the drive train simply won't work. But none have posted a recommendation. If you want traction and not speed, here's mine:

build a six-wheel drive train, wide type. Use traction wheels on one pair of opposing corners and omnis on the other (otherwise your robot won't turn well). Use a gearbox with a ratio of at least 12.75:1 (standard AndyMark Toughbox). If you REALLY, REALLY want traction, put belts on it. I don't recommend that, though.

Here's the lesson to learn from this: You can have traction or you can have speed. You can't have both.