Consequences of Removing the Shoulder Support Spring

By Anuhya, Trey, Leo, Gabriel, and Vance

We learned the consequences of making small changes to build without fully knowing the outcome

A small change can make a huge difference: we have learned that we really should have put the shoulder spring back in place after we rebuilt Taubot. The Crane arm fails to maintain its shoulder angle, and this is an issue which has only started since we built Tau2. We're probably hitting the thermal cutoff for the motor port, which causes the control hub to stop sending power to the shoulder motor and the whole arm crashes to the ground.

This also means that its stalling the motor and it pulls a huge amount of Amps. We also have horrendous battery management, which leads to short battery life and permanent battery damage. The Crane being held out fully extended further damages it.

A lot of Tau2 is an inherited, yet modified, design. We aren't fully aware of the considerations that caused a spring to be in Tombot, the original robot we based part of our design on. We also changed the gear ratio of the shoulder motor so we could power through lifting the shoulder at extremes, and we're pulling many more electrons to compensate for something which was built-in initially.

The original reason was a design goal - whenever you make an arm, you want as little motor force as possible dedicated to maintain the arm against gravity. The motor should just have been changing the position, not maintaining it. Real world cranes use counterweights or hydraulics to fix this problem, something which we don't conveniently have access to.

The spring was meant to reduce the burden of maintaining the crane's weight against gravity. It wasn't perfect by a long shot, but it definitely did help. Our batteries would also last longer in matches and have a longer life.

Thankfully, we still have access to the original CAM we used for the log spiral, and a spring could easily be reintroduced into our robot if we messed with some of the wires. It's a high priority and needs to be done as soon as possible.

However, hardware changes like these are incredibly annoying for coders. We recently added a custom transfer plate too, and it was very helpful to increase transfer efficiency. It also changed the angle of the initial transfer plate to the flipper gripper, which added 3 hours of extra coding time to retune transfer and grasping the cones so it didn't get knocked off. It will be better once fully tuned, but it doesn't change the fact that small changes in physical build are incredibly annoying for the coding team. Adding the shoulder support spring will also have this kind of cost, but it's crucial for further success.