Drop Down Menu

Wednesday, February 12, 2014

Consistency is Key (Long Post Alert)

Our #1 goal this week was to got our robot shooting consistently. And we made it happen through some hard work. Thor can shoot from 18ft. away from the goal to about 8 ft. from the goal. This is going to be a long post. Haha. Sorry!

Back tracking a little bit, last year, our robot was consistent on shooting... When it was lined up. Haha. The problem was that it would take us a frisbee shot or two to get lined up. So we'd have 4 frisbees and typically use one of them to line up, then shoot 3. That was awesome work by us! We had the same aiming problems and consistency with Rebound Rumble in 2012. What we've done in the past is not good enough frankly. Especially when we only have one ball to shoot.

We need it consistent.

To this end, we've been working REALLY HARD on incorporating some sensors. We tried some limit switches and they worked. The problem was that mounting them in the place we wanted was very tough. Plus, we want 3 positions (down, up, shoot) so limit switches make that difficult to code and operate.

We moved on to trying a gyro. We've never used a gyro before. We mounted it on the gripper arms which is kind of against the purpose of it. A gyro measures the angle position (from a reset point). So the usual application of a gyro is to mount the gyro flat and have a robot drive straight during autonomous (the gyro can adjust for turns). We tried something different by mounting it on it's side and sensing it then.

Lots of problems with the gyro. The reset happens every time the robot turns on so we would have to have the grippers in the right spot every time we turn it on. When the gyro is on for a long time, it "drifts," which means it slowly reads an angle measure gain, even when it's stationary. So every time we hit a ball, we would have to reset the grippers. Window motors are not the most accurate motor ever (especially with a ball in them) so hitting the same angle every time was a challenge. The programming did not go very well (through no fault of our own, we were definitely pushing our abilities).

We thought about a potentiometer but we're at the stage in the build season where we didn't want to wait for a package and we didn't have time to learn something totally new and mount it.

So here's our solution:

We call them "Stingers." They're mounted on the back of the hammer. When the hammer goes up, they shove the ball to a consistent position that we can kick from. Don't worry about the frame perimeter, we will start autonomous with the hammer up. That negates the frame perimeter issue. We will post a youtube video later and some gifs.

Sorry for the long post but we wanted to give our programmer, Karl, some time on the blog. He's been super helpful and awesome all year and especially this week. His willingness to try stuff for a long time, research, and work, then throw it all away to try something new is great. GO KARL!

No comments:

Post a Comment