To overcome the design and implementation issues as mentioned in the previous article. The finger control mechanisum has been redesigned.

Instead of using control wires to keep the motors in the finger base, the motors are to be embeded within the finger itself. Allowing for simple control of finger anghle using the same potentiometers for feedback and bevel gears for transfering the motor power onto the axis of movement.

The difficulty with controlling the robot with the wire method, as used currently, is the difficulty in controlling multiple degrees of freedom in a series setout while keeping all control acutation inline and not obstructing other axes.

This has caused difficulty with the current design, as when one degree of freedom is controlled the actuation works well, as shown previously. However, when multiple axes are controlled the finger tends to lock up, suffering from a mix of tollerance/ design issues with the actuator location and lacking power within the motors to overcome these inefficiencies.

The probable solution to this will follow in another article.

This video demonstrates the PID control of a single joint, using the pot feedback to return the finger to the vertical position, manual force is used as a disturbance.

The electronics controlling the robotic finger is based around closed loop PID control using potentiometer feedback. The motor controller is currently an L298N dual H-bridge module, controlled by a Teensy 3.2.