Drive Train Gear Ratio

[Thorn]

Team 2869 used a mecanum drive train last year and it worked very well. We would like to the same for this year, but with greater speed (and less torque). The gear boxes we had last year gave us a 12.7:1 reduction (AndyMark touchbox mini) and we were able to modify these to bring it down to a ratio of 3.7:1.

We are missing some parts and it will therefore be a number of days before we can actually build the drive train and test it.

Do you think this is too low of a ratio? Will the robot barely be able to get started with a load on it with a gear ratio of only 3.7:1?
We have four motors, each independently controlling one wheel.

[MrForbes]

I think you'll have trouble. That's a 3.5 times reduction in torque. You'll have very little acceleration, and way too much top speed (if you can get it moving), and very little control at low speeds.

[Ether]

Originally Posted by Thorn

3.7:1

5310/3.7 = 1435 wheel rpm

1435/60*2*pi = 150 radians/sec

6" diameter wheel = 3/12 foot radius

150*(3/12) = 37.5 fps free speed.

37.5*80% = 30 fps rough estimated top speed

Could you share your team's thought process that lead to this decision?

[Kevin Sevcik]

Originally Posted by Ether

5310/3.7 = 1435 wheel rpm

1435/60*2*pi = 150 radians/sec

6" diameter wheel = 3/12 foot radius

150*(3/12) = 37.5 fps free speed.

37.5*80% = 30 fps rough estimated top speed

Could you share your team's thought process that lead to this decision?

You're being kind in not pointing out that they'll never actually hit 30fps before they run into the step or a wall. Just a back of the envelope kinetic energy/max power calculation says it'll take them about 2 seconds to hit that speed. And that's assuming max power output the whole way, which is certainly not the case. Also it's entirely too fast to control even with 4" wheels.

Thorn, I'm trying to figure out how on earth you're getting a 3.7:1 ratio out of the TB mini. Andymark only specs down to a 5.95 : 1 ratio. Which is still too fast for 6" wheels, but anyways. I don't see any parts on Andymark that can get that ratio. There's Spinbox gears that would get you to 3:1 or faster. The closest I could find was swapping in a 30 tooth and 34 tooth Vex gear, for a 4:1 ratio...

So, with the note that 3.7:1 is a horribly fast gear ratio, how on earth were you guys going to hit that, anyways?

[Ether]

Originally Posted by Kevin Sevcik

You're being kind in not pointing out that they'll never actually hit 30fps before they run into the step or a wall.

Yeah, that was going to be part of the follow-up dialog once I had some context for their decision.

[Chris Endres]

There is no good reason to go over 13-15 fps in this game. Think about it, any faster and you'd travel from the driver wall to the step in approximately two seconds.

As stated, I'd suggest looking at the JVN design calculator. I use the pull-down version found here: http://www.chiefdelphi.com/media/papers/2815

[JamesCH95]

Originally Posted by Thorn

Team 2869 used a mecanum drive train last year and it worked very well. We would like to the same for this year, but with greater speed (and less torque). The gear boxes we had last year gave us a 12.7:1 reduction (AndyMark touchbox mini) and we were able to modify these to bring it down to a ratio of 3.7:1.

If your drive worked "very well" last year on an open field, why would you gear to a faster top speed (and thus less acceleration) for a much smaller field?

Originally Posted by Thorn

We are missing some parts and it will therefore be a number of days before we can actually build the drive train and test it.

You won't like what you get...

Originally Posted by Thorn

Do you think this is too low of a ratio?

Yes, by at LEAST a factor of 2.5.

Originally Posted by Thorn

Will the robot barely be able to get started with a load on it with a gear ratio of only 3.7:1?
We have four motors, each independently controlling one wheel.

Yes.

I STRONGLY encourage you to look at the JVN motor design calculator. These motors will pull a LOT of current. They will likely overheat and/or blow breakers. You will suffer from poor low-speed control. Your robot will also be very slow as it will not be able to accelerate effectively.

[asid61]

Originally Posted by JamesCH95

If your drive worked "very well" last year on an open field, why would you gear to a faster top speed (and thus less acceleration) for a much smaller field?



You won't like what you get...



Yes, by at LEAST a factor of 2.5.



Yes.

I STRONGLY encourage you to look at the JVN motor design calculator. These motors will pull a LOT of current. They will likely overheat and/or blow breakers. You will suffer from poor low-speed control. Your robot will also be very slow as it will not be able to accelerate effectively.

I'm going to go ahead and agree with most of this. 30fps is way too fast, and even if you go with 4" sheels you're still moving at 20fps, which is crazy fast for 4 cims! Your acceleration will be terrible.
Gear lower. With 4 cims on a tank drive, the maximum effective fps IMO is around 17fps, preferably 16-17. I don't know about mecanum speeds though, but teams have gone up to 15.4fps with mecanum at least.