Recycle Rush: Speed, Torque, or Both

[billbo911]

Now that we have all had a week+ to contemplate our strategies and designs, there is an aspect of this game that I would like to open a discussion on.

The field that we play on this year is about half of what we have used in the past. Granted there are only 3 robots on it, and there are a lot of game pieces, but over all, the area we drive in has been greatly reduced.

So the topic I would like to discuss is, "What do you aim for with your drive train design?"

With the field being smaller, is top speed a consideration?
With the tight quarters, do you place an emphasis on torque to help with maneuverability?
Is too much torque going to be a problem?
Will a good balance of both be a better choice?
Single speed or multispeed gearbox?
2 CIM, 4 CIM, 6 CIM, 4 CIM + 2 mini's, all mini's?
Holonomic, Mechanum, Swerve, WCD, 4 WD, KOP?

All of these can also be answered by "other" of course and I'm sure there are other considerations I have not mentioned so feel free to bring them up.

As you walk through your answers, please give a bit of your reasoning behind your choices.

[AdamHeard]

I think 10 fps is the magic number for 99% of teams this year (independent of drive type).

[Michael Hill]

Originally Posted by AdamHeard

I think 10 fps is the magic number for 99% of teams this year (independent of drive type).

Agreed. We figured the typical max distance a team will want to travel is on the order of 11 ft. If you gear for much faster, you won't even achieve your max speed. Also, having too much speed is pretty dangerous when there are totes that are being stacked all around you that you don't want to knock down, so it also comes down to controllability. If you gear for ~10 ft/s, then you'll get to your destination in about 1.5-2 seconds (accounting for acceleration/deceleration), which is, IMO, a pretty decent target.

[Ether]

Imagine you have 2 robots which are identical in every way, except that Robot A is geared for 8 fps, and Robot B is geared for 16fps but has the motor voltage limited to 50%.

Which robot has better fine control of small slow-speed motions?

[Gdeaver]

The word this year is precision.

[markmcgary]

Originally Posted by Ether

Imagine you have 2 robots which are identical in every way, except that Robot A is geared for 8 fps, and Robot B is geared for 16fps but has the motor voltage limited to 50%.

Which robot has better fine control of small slow-speed motions?

I'll bite.

Robot B.

[AdamHeard]

Originally Posted by markmcgary

I'll bite.

Robot B.

Why?

[markmcgary]

Originally Posted by AdamHeard

Why?

"better fine control of small slow-speed motions"

It seems to me that the high gear ratio, for higher speed, would help with control at low power and low speed because the drive would be 'less responsive' at low power, thus helping with fine control. (And, I thought I'd be different.)

[Jared]

Originally Posted by Ether

Imagine you have 2 robots which are identical in every way, except that Robot A is geared for 8 fps, and Robot B is geared for 16fps but has the motor voltage limited to 50%.

Which robot has better fine control of small slow-speed motions?

I'd compare it to parking a car. Robot A is my car in first gear, and Robot B is my car in 2nd/3rd gear, but I can only push the gas pedal half way down.

It would be harder to deal with the inertia of the car if it were stuck in second gear and had less torque. You'd try to move some small amount, but it wouldn't react until you applied lots of power, and once it started moving, it would be harder to stop.

I believe that the robot geared for 8 feet per second would be much more maneuverable and well suited for precise control, especially with skilled drivers. Lower gear ratio means more torque, and more torque means faster response. It's easier (to a certain point) if the robot responds faster. I don't believe 8 fps is too twitchy.

[Dunngeon]

Originally Posted by Jared

I'd compare it to parking a car. Robot A is my car in first gear, and Robot B is my car in 2nd/3rd gear, but I can only push the gas pedal half way down.

It would be harder to deal with the inertia of the car if it were stuck in second gear and had less torque. You'd try to move some small amount, but it wouldn't react until you applied lots of power, and once it started moving, it would be harder to stop.

I believe that the robot geared for 8 feet per second would be much more maneuverable and well suited for precise control, especially with skilled drivers. Lower gear ratio means more torque, and more torque means faster response. It's easier (to a certain point) if the robot responds faster. I don't believe 8 fps is too twitchy.

This was my line of thought too. Our 17fps bot last year had its turning speed reduced to 30%. There was a space on the control pad where the motor did nothing. Then all of a sudden it would break loose and start accelerating. Robot A would be more controllable

EDIT: Accidentally mixed up the robots

[philso]

Originally Posted by AdamHeard

Why?

The PWM is usually generated using a clock signal in the microprocessor so it has a finite resolution in time. That means there is a minimum time increment to the on time and off time of the PWM signal. The controller in both robots would have the same number of PWM steps but Robot B is only using half of those steps so each step constitutes a larger percentage of the range of control. Thus Robot A would have finer control.

[AdamHeard]

Originally Posted by philso

The PWM is usually generated using a clock signal in the microprocessor so it has a finite resolution in time. That means there is a minimum time increment to the on time and off time of the PWM signal. The controller in both robots would have the same number of PWM steps but Robot B is only using half of those steps so each step constitutes a larger percentage of the range of control.

So you're voting for B or A?

[asid61]

Originally Posted by Gdeaver

The word this year is precision.

It depends a little, I think. For example, roller wheel pickups greatly reduce the amount of precision required compared to a fixed forklift, presumably more so if they're actuated thin and wide actively.
OC you will probably need enough precision to get totes onto the step without falling.

[dellagd]

Originally Posted by Ether

Imagine you have 2 robots which are identical in every way, except that Robot A is geared for 8 fps, and Robot B is geared for 16fps but has the motor voltage limited to 50%.

Which robot has better fine control of small slow-speed motions?

Since robot A is geared lower, it will have finer control on slow speed motions for two reasons.

First, since you are utilizing a pulse width for your transmission of the speed signal, and these microcontrollers only have so much fidelity in their creation of an arbitrary length pulse, being able to utilize 0 o 100% power in robot A instead only half of your available discrete 'steps' of width will give you better control of the motor speed.

Second (and most importantly), in order to start moving from a stand still, there are static frictions that must be overcome in the gearbox and in the wheel-carpet system (if your are turning). These static frictional will be overcome and turn into kinetic ones once a certain torque is applied, and since that torque happens at a 50% lower motor voltage level on robot A than it does on robot B, robot A will have much finer control. In a high speed drive, if you go to make a small adjustment, odds are you have to apply a high enough motor voltage that once you do begin to move, you have already overshot your target. Once the static forces are overcome, it is possible to slow down slightly depending on the difference in the coefficients form static to kinetic, but once again robot A will be able to withstand motion at a lower speed, which I believe would be roughly 50% lower than robot B.

Those are my thoughts, and as for the balance of speed vs torque, for those reasons I think it is still important to balance torque and speed this year correctly. Just because there is no defense doesn't mean it is a good decision to gear your robot as fast as possible.

[markmcgary]

Originally Posted by markmcgary

I'll bite.

Robot B.

Further reading and learning on this and other threads compels me to switch sides. I'm liking Robot A. (And, we need to buy some appropriate cluster and output gears.)