# Can a robot be too fast?

Is there a limit where a robot becomes too fast to control, and how can you deal with that? I was thinking that you could press a button to reduce the speed for when you need more precision, but there could be better ways.

Too fast for the operator to control is not an issue for Vex but one can easily gear the drivetrain too fast so that the motors stall and the breakers overheat and trip, removing power from the motors. This is very bad.

Many teams use cubic functions. What this does is instead of linearly mapping joystick inputs to motor speeds (ie. a reading of 50 from the joystick results in the motors running at 50), it narrows down the range of medium speeds and instead allows the driver to make very fine adjustments while driving slowly. This is useful since most robots are usually either moving at full speed, or at very low speeds in order to make small corrections.

This website describes cubic functions in more detail as well as teaches how to implement one:
http://nakkiran.org/personal/robotc-tips/cubic-map/

As mentioned above, cubic functions are very useful for having good fine control without compromising speed or using extra buttons, however it is mostly used in FRC, as drivetrains in VEX do not typically have a great enough range of speeds to justify it.

There is no rule limiting the speed of your robot, however if it is intentionally mashing other teams robots, that would be a severe violation. However, because of the weight of robots, as well as the VEX motor limitations, you would find it extremely difficult reach dangerous speeds. Even with a 6 motor drivetrain, speed in excess of 3 feet/second (roughly 1:2 with 4" wheels), would simply trip the motor’s internal thermal breakers.

As we learned painfully this year, it is far more important to have a reliable drive than to be extremely fast and die 45 seconds into a match.

I can’t see driving too fast being a problem in vex. Especially if you build your bot in a somewhat standard way, and only use 4 to 6 motors on drive. If you don’t do the cubic method, as suggested before, you could also use one of the buttons on the top of the controller. For example, when the button is down, power motors at a third of the controller input.

Too fast in a striaght line - no
Too fast to contorl a turn - yeah, but software and a gyro can help
Too much acceleration to pop a wheelie and tip you over - yes, absolutely (maybe in a turn too but I have not seen that yet)
Too much acceleration or poor weight distribution to spin the wheels - yes

It’s not always the top speed velocity but the acceleration and jerk that gets you in trouble. (Jerk is a lesser used term but is the next derivative down and gets you spikes when there are sudden starts/stops).

Having some software to ease you in and out of situations is usually only done in programming skills (e.g. slowing down to a target using P or PID controls) but it is very useful in driver control too sometimes. Otherwise you rely completely on the skill of the driver.

A problem with a robot being fast is that when you make it go faster, like with gears, you will lose strength. It won’t hold as much, and if there is too much weight on it, it will burn out the motor. So you must find that perfect balance, which is nearly impossible to find. Toss Up is more of a speed game. You do not need that much power to lift any type of ball, large or Bucky. I’m still curious where they got the name BuckyBall.

The technical term is Buckminsterfullerenes, hence the abbreviation. Because I don’t even know how to pronounce that.

If you want some information on the origin of the word or chemical structure, here you go.

There have been lots of great VRC robots over the years, but this match features two of the fastest of all time (on the blue alliance). The key to good control is effective software and practice, practice, practice.

Those being one of your teams do you happen to remember the ratio?

I was a mentor for those teams – both robots weighed about 8 pounds and were driving 2.75" wheels at 1:3 (or 3:1 from a different point of view), for a theoretical top speed of about 3.6 feet per second. They were powered by six 3-wire motors. I believe 575 was slightly lighter and slightly faster, and could cross a VEX field in about two seconds. 418 could pull cubes from the loader, which 575 could not.

Both robots scored 60 points in Robot Skills during the season (the highest score of the year at an official tournament), and 575 scored 42 points in Programming Skills at one tournament (plus 37 at Worlds). In my opinion, speed, programming and practicing are extremely valuable.

About three years ago my team built an extremely fast vex drive, something like 5-6 feet per second (fps). This was not for a competition robot, just a fun exercise to build a super fast robot. It’s been awhile and the details are kind of sketchy, but I remember that it involved 10-12 393 motors so not practical on a competition robot.

As far as the VRC is concerned there is a practical limit that relies on two factors:

[LIST=1]
*]Motor Power: you have to accelerate the 10-15 lb robot in enough time to not trip the breakers, which for 4 motor drives usually comes means a max speed of somewhere between 2-3 fps if I remember correctly. Naturally adding more motors will allow a greater acceleration and a higher max speed.
*]Travel Distance: Related to the motor power, it takes a certain time to reach top speed and over this time the robot will travel a distance. The higher the speed the further the distance the robot will travel. For very high speeds a vex robot can not reach max speed in the space of the field. For instance my super fast robot took half the length of the hallway (about 2 full fields length to reach max speed)
[/LIST]

Assuming that your mechanical design allows for a high max speed, it is certainly possible to control a robot with these speeds, with the correct programming (stepped output, a fast mode button, PID etc.).

A few years ago in gateway our A team’s world championship winning robot was so fast that the driver had to press a clutch button to reduce the motor power percentage in order to do small precise movements.

How is it the gearing a drive train towards speed, even simply in 1:2 ratio could throw the thermal switch in the motors? The motor isn’t really doing that much work is it?

Last year, our robot was quite heavy and most of the weight was concentrated above our 4-wheel, direct drive, short, mecanum base. We had a lot of problems with the motors geared internally for speed (1:1.6) (especially with strafing) but even at 1:1 with the weight concentrated above the wheels, strafing was somewhat of an issue.

Part of the problem could have also been that we loaded up one of the cortex’s internal breakers with all 4 of the drive motors and the other breaker with all 4 of the lift motors. Regardless, we are now firm believers in seriously analyzing what your motors can handle. Look at the bright side of choosing a conservative ratio, you get more torque than you were originally planning for :).

With a small, light, robot, a 4-motor drive geared 1:2 should work pretty well. The problem is that when you have a larger robot (especially one made of steel), or you strain your drive trying to push another robot, the motor is actually doing a lot of work. Remember that a 1:2 speed ratio divides the torque of your drive train in half, so with a 4-motor drive geared 1:2, you have the same torque as a 2-motor drive geared 1:1.

Seconded. We’re at 2:1 now, no trouble. Having a light robot helps.