We attempted to improve the speed drop when the ball goes through the flywheel by gearing down the motors. We hoped that the increased torque would create a smaller drop and planned to increase the power to compensate. To gears the motors more for torque, we replaced the high speed gears in the motors with high torque gears. If the relationship between speed and power were linear, we would be able to increase the power to 120 and be at approximately the right speed. However, we discovered that the relationship between speed and power is not linear and we could not increase the power to get more speed. The speed given by a motor power maxes out at about 80.

A graph of the speed of a motor given a certain power.

A graph of the speed of a motor given a certain power.

We then tried replacing the external gears (after replacing the internal high speed gears) to gear it towards torque, but less. The balls were making it into the net, but only at 80 power, so there was no way to get a higher speed for the spike or an adjustment function. We then replaced the gears with the original gears. During the process, some spacers were moved to reduce drag (they were rubbing against the gears) and the bar was raised because it compressed the ball and increased the flywheel’s work.

Another thing that we tried to reduce the speed drop was spreading apart the wheels. When we did this, the ball was not compressed enough and didn’t go very far.


A graph of the drops in speed as balls go through the flywheel

The last thing we worked on was the PID. However, while attempting to tune it, we noticed that the drops when the ball went through were very different. When calibrating the speeds (which was done before tuning to see how bad the difference was) we saw that the squishiest ball needed 2.9 RPS speed to reach the net, while the hardest ball only needed 2.15 RPS. All the other balls were spread out in between.

Before we continue work on the PID, we need to try to reduce the difference between hard and squishy balls and also to reduce drag so the maximum speed is not the speed needed to launch a squishy ball (since there is no room to add any speed to bring the speed up).

We also noticed that the bar holding both sides of the spinner together is one hole short. This made the robot asymmetrical and could possibly warp the shafts. In addition, it is affecting the outside supports.

Share this post: