The Black Ghosts

We considered the mechanical design a great success, considering the timeframe we had to work with. We succeeded in making an array of fins that could be spaced very close together, provided a great deal of torque, and could be extended to as many fins as desired. We consider our design an elegant solution to the problem, given the resources and timeframe we had to work with.

However, we noticed that during our tests, the coils would slam into the nuts if they were given too much power. The wires coming from the coils had been glued in such a way that they were sandwiched between the fin and the bolts when this happened. Eventually, the coating on the wires wore down and the wires tended to short out when they hit the bolts on the side, causing our fins to behave erratically. Because of this problem we were reluctant to put the fin array in water, even though we had built a structure that allowed the fin array to slide up and down a rail as it hung suspended upside down. We had hoped to test the array in water, but this last minute problem prevented us from doing so. This problem could be quickly fixed in a redesign by gluing the wires in a place where they would not be sandwiched.

On the day following class presentations in the mechatronics lab, our fin suffered a fall while in transit to Tech for another demo.  Fortunately, the mechanics of the fin was undamaged, but some of the connections between the coils and the wiring harness appeared to be damaged.  When we tried use the fin, we observed that Fin 4 had no power, and Fin 3 had no connection to the sense coil, while Fins 1 and 2 remained operational as they had the day before.  This kind of damage could be fixed by re-soldering the connections, and adding a physical synch on the ribbon cable that would take any stress on the wires off of the soldered connections.

The amount of time spent trying to use the Handy Board for digital feedback and control was a bit of a disappointment, but for some drift in the integrator circuits our analog circuitry worked very well.  Digital feedback would avoid this kind of error, but might introduce some significant processing constraints if polling the sense coils were required. 

Special thanks goes to Michael Epstein and Professor Michael Peshkin, who created the first prototypes for this project, for their help and guidance in making this project a success.