4WD vs. Chain Drive

[Ratchet1]

Right now we are trying to figure out whether we should use a 4WD system with motors on all 4 wheels, or a chain system with 2 motors. Any suggestions?

[Bob Steele]

Either would work for you. Just remember that you need to power any wheels that in contact with the floor to give yourself the best acceleration at possible.

You can, of course do 4WD with two motors and chains to all wheels.
In this game I would really doubt if you will need 1 motor per wheel because the friction is so low that you can't apply much torque to the wheels anyway.

Two motors is lighter...

Good luck!!

Robert Steele

[billbo911]

Originally Posted by Ratchet1

Right now we are trying to figure out whether we should use a 4WD system with motors on all 4 wheels, or a chain system with 2 motors. Any suggestions?

Please search these forums. There are several discussions regarding the advantages of both.

No for my opinion (Warning, Warning, opinion on the loose)

This year's game definitely takes some prototyping and mathematical analysis to make many decisions, the drive train is a perfect example. For the approach we are taking this year, we have decided to not try to go with lots of power for the drive train. Instead, we are minimizing torque by limiting the gear reduction, using just two motors and doing a little coding magic.

[Ratchet1]

ok. Thanks for the help. And i did look, and couldnt find anything on it, thats why i started a new thread.

[Jared Russell]

2 CIM motors with the gear ratios from the kit transmissions provide more than enough power for this years' surface.

4 motors would make you go no faster nor push harder (it would allow you to use higher gear ratios, I suppose).

The difference comes down to what is easier to mount in your frame.

[EricLeifermann]

If you go the 4WD route it could give you the opportunity to monitor each wheel and reduce power if it starts slipping too much...

[MrForbes]

If you go with the chain, 2 motor, 4wd design you could monitor the speed of both wheels on each side, and reduce power if they slip too much. No need to have the wheels driven independently to do this, the wheels on each side of the robot will slip about the same amount (in the forward direction).

[Daniel_LaFleur]

Originally Posted by squirrel

No need to have the wheels driven independently to do this, the wheels on each side of the robot will slip about the same amount (in the forward direction).

Your assumption, here, is that the weight of the robot is roughly equal on all wheels. This assumption may, or may not, be true, and will give different results with different CoM.

[MrForbes]

No, my assumption here is that you are not turning too sharply.

If the wheels are connected by a chain, they will turn the same speed...doesn't matter how much weight is on each one.

Might be that you were confused by my brief wording. What I mean is that on each side of the robot, the two wheels on that side will slip about the same amount, so you can implement slip control on a "per side" basis, instead of a "per wheel" basis.

Hope this makes it more clear!

[Daniel_LaFleur]

Originally Posted by squirrel

No, my assumption here is that you are not turning too sharply.

If the wheels are connected by a chain, they will turn the same speed...doesn't matter how much weight is on each one.

Might be that you were confused by my brief wording. What I mean is that on each side of the robot, the two wheels on that side will slip about the same amount, so you can implement slip control on a "per side" basis, instead of a "per wheel" basis.

Hope this makes it more clear!

The side will slip when the wheel with the least amount of traction breaks free because once that happends all of the torque from the slipping wheel will go through the chain to the non-slipping wheel and that will cause it to slip as well.

Thus 4 independently driven wheels will be slightly better ... although with traction control you may be able to compensate for the differences.

[MrForbes]

Originally Posted by Daniel_LaFleur

The side will slip when the wheel with the least amount of traction breaks free because once that happends all of the torque from the slipping wheel will go through the chain to the non-slipping wheel and that will cause it to slip as well.

All of the torque? or the difference in torque between static and dynamic friction?

Thus 4 independently driven wheels will be slightly better ... although with traction control you may be able to compensate for the differences.

I agree that you can get a bit better traction with independent 4wd, but I also think you can do a lot with per side traction control.

In our case, we're using up all the motors on the ball handling stuff so it's a non-issue, we only have one CIM and one tiny Mabuchi motor (and the window motors) left for each side of the drivetrain.

[JesseK]

Hmm, interesting points above made me think of something.... Arguably, you could take 1 FP + 1 Banebots 545 through their own AM Planetaries into a toughbox and have that drive one side, with an identical setup on the other side. Estimations for worst-case here are that each motor pulls 5 amps when you're pushing. This is only doable because of the coefficient of friction though. So long as the chains drive all wheels that touch the ground, it would work ok. I'd argue that acceleration with this setup is very close to that of 1 CIM per side.

It's downside is that it adds much more complexity than needed and is somewhat expensive for what you get.

[Daniel_LaFleur]

Originally Posted by JesseK

Hmm, interesting points above made me think of something.... Arguably, you could take 1 FP + 1 Banebots 545 through their own AM Planetaries into a toughbox and have that drive one side, with an identical setup on the other side. Estimations for worst-case here are that each motor pulls 5 amps when you're pushing. This is only doable because of the coefficient of friction though. So long as the chains drive all wheels that touch the ground, it would work ok. I'd argue that acceleration with this setup is very close to that of 1 CIM per side.

It's downside is that it adds much more complexity than needed and is somewhat expensive for what you get.

Also remember that the CoF on the carpet edges is different then the Regolith