[Wetzel]

I joined 2914 just after kickoff this year. While this was their fourth year as a team, only one student had done FIRST before and many had a very low level of knowledge about hand tools. We had a strategic discussion about the game to figure out how we wanted to play and ended up with a design to get picked rather than rank high. We settled on a priority list of SIMPLE, bridge, defense, feed balls forward, shoot baskets. For us, the shooter was an intentional afterthought.

We initially pursued a pneumatic kicker, and took two weeks to determine (and learn how to use the pneumatics) we couldn’t make it work well and keep it simple. Then we rapidly prototyped a single wheel shooter to chuck the balls to the other half of the field. We built a wood chute and then powered various wheels by various motors, mostly held in place by hand or clamped in place. We did these rough prototypes to get a rough estimate of how the ball responded to different wheels, motors and speeds, then following our SIMPLE mantra chose a COTS gearbox from the available options at AndyMark and BaneBots. We ended up deciding on two RS550 motors into the CIM-U-LATOR based on how far we wanted to be able to chuck the ball, our quick and dirty empirical testing, and the options available. It ended up being a simple lift system to the wheel, and then the ball went under it and above a steel plate. It chucked balls to the other end of the field admirably, fulfilling its primary mission.

From there, we have constantly tweaked the system to improve shooting at all our events, with the target of better than 75% accuracy in autonomous. We went from the ball passing under the wheel to going over it to reverse the spin and 3 major revisions to the shape of the exhaust, with the major changes mostly done to reduce flex and uncontrolled motion. We had a video camera with the ability to capture super slow motion HD video, and that was incredibly helpful to identify what was going on with the shooting and to identify issues. We went from making one shot in autonomous in total at our first event to somewhere around 80% in autonomous at St Louis.

One of the things we realized pretty early on was that the battery made a significant impact on the accuracy of the shot. We first attempted to use a shaft encoder on the shooter axle and to tune a PID loop but were unable to get satisfactory results with that. We got an improvement, but a small one. At St Louis we set up the CAN and drove the shooter motors by voltage rather than PWM and I think that gave us the biggest improvement in having a repeatable shot amoung all the changes we made.

We wouldn’t settle for it is what it is, but remained cognizant of the fact that sometimes better can break good enough.


Design is iterative. Life is iterative. Strive to improve.


Wetzel