linear punch or flywheel?


OK so since roughly half of the game points rely on the flags(15 points) I think that a launcher is pretty important. The problem were having is choosing between a flywheel or a linear punch. I mean its not like NBN where linear punches did so well because it could stay consistent weather the balls were new or not. This year the balls aren’t foam so I imagine a flywheel would be more accurate than in NBN. I would say that top priority would be space, being that we have to have a launcher, ball intake, lift and a claw.


Actually puncher is more accurate than flywheel. But flywheel is easier to change the range of the ball.

So you will need to decide and prioritise your requirements.


Agreed, punchers are far more accurate but take into account on how you plan to load the ball when you field. Flywheels are much easier to load than punchers, so each has its downfalls.


Catapult maybe?


In my observation, flywheels uses more motors? I’ve seen lots of punchers with 1 / 2 motors but most flywheels have 2 / 3 motors.


If we’re running on v5 (which have ~2.5x the power) you should be off okay using just one motor for a flywheel. It all comes down to how its executed


Correct me if I’m wrong, but doesn’t every flywheel setup require at least 3 motors? 1 motor to intake from the floor, 1 to actually push the ball into the flywheel, and 1 for the flywheel. One major benefit that I see with the puncher is that the balls can be passively fed into it, since the puncher isn’t constantly running.


I feel like linear punch accuracy vs flywheel accuracy becomes insignificant if robots line up as close as possible right in front of the flags.

One problem with punchers is that if you want to shoot both flags, you’ll either need to change the angle of the puncher, create two punchers, or line up shots in two different places.


@Wiredcat Robotics Oh yeah. I forgot to mention that we’re using V4 because of the cost…
But true, v5 probably will handle a flywheel with 1 motor.


You don’t need to spend so many motors on your intake (esp. if you’re using V5) because it doesn’t take that much torque to move the balls. Since you’re limited to 2 balls, you won’t really need an extra mechanism to feed balls to the flywheel, that could probably be part of your intake


Robots I helped build in NBN used 1 motor for intake, and could’ve been done with 1 motor on the flywheel. If it’s a continuous feed it’ll be fine.


I was thinking that you could probably run your elevator and intake on on motor being there so close. they would just mean that they would move at the same time which isn’t necessarily a bad thing.


and also I think I’ve come up with w solution for what launcher as well. I was thinking with a flywheel, you could probably use encoders or take advantage of the recovery time after you shoot a ball to have it where you shoot two at once and the first hits the top and the second hits the middle flag.


I don’t see how a puncher will be strong enough to hit the flag, because the ball is plastic.


That would depend on what elastics or latex tubing being used and how far you stretch em.


Having messed around with the balls a bit, I could see a puncher hitting them, but (obviously) it’d need pretty good elastics to do effectively


If a puncher and flywheel (or catapult, trebuchet, slingshot, etc.) get the ball moving at the same speed and angle, the flag isn’t going to care what launched the ball. The question is just if you can get comparable launch speeds and angles.


Puncher could equal ball launching with only one motor:

With ratchets, ball intake with a motor going positive and puncher elastics with the same motor negative. Puncher is mechanically linked to lift arm for vertical angle, and drive base provides horizontal angle. Mount a vision sensor to the puncher arm and code to center the flag in the image using drive and lift motors. With a flat enough trajectory you should be pretty accurate. Otherwise, you would need to position the arm with the flag lower in the image based on its size in the image.

5 or 7 motors left for other systems.


You don’t need to use size for targeting, and I would recommend against it. If you just get a value for the height (preferably of the middle) of the color block on the screen, that’s enough to figure out how far away it is given than you know the heights of your vision sensor and the flag. I wrote the calculations in a recent thread somewhere. Size will work, but it’s definitely harder.


@callen I am suggesting the vision sensor tilts along with the puncher mechanism such that its aimed along the flight path of the ball. In that situation, the flag’s (vertical) image size should be a good indicator of distance, right?