This week I made a capacitative position sensor for the x and y axes of a gantry on the multifab. This was based of the the step response capacitative sensor Neil described in class. I made a board that had two inputs. I then tried doing the split triangle configuration, but it was very noisy. I couldn't get an arithmetic combination that I was able to map to an linear location. So instead I ended up using one tx/rx per axis. I was ok, but very noisy, so I gave up on this triangle setup. Just wasn't cutting it.
Instead, I used two rectangular pieces, one on top of the other. The one attached to the moving gantry was smaller, and a long strip of copper foil was placed below the driving belt. On the x axis, I was able to use the adhesive of the copper to tape it to the belt. The longer strip I left on the adhesive cover so as to limit effects of the case protecting the belt.
The y axis was a little more complicated. For some reason, although it was the same board and code as the x axis, it was really noisy. I couldn't isolate what was the cause, but I believe it could be due to the alligator clips. The capacitance values were not only more variable, but also higher. To try and eliminate effects of the belt cover, I placed the strips on the outside. Taped the alligator clip with a copper foil to the directly to the gantry so it would sweep over the long copper foil as it moved.
The setup of the x and y capacitative position sensors.
The x axis was variable, but consistent. The trend can be seen as the gantry moves along the x axis.
The y axis was a failure. No real trends could be seen as it moves.