I get a lot of emails and comments concerning the Magnetic Table CV Controller that I built and demonstrated here:
I wanted to add a few notes on the construction and process of building one for those who wish to build one for themselves.
I built it out of spare parts that I had lying around, plus a cheap wooden box that I bought at the craft store, then stained it and made it look older with some raw umber and shoe polish. Originally, I just fastened an aluminum rod on the end of the joystick for the pendulum arm, but it was a little more wobbly than I wanted it to be and it easily loosened to the point where the subtle changes in position would not transfer to the joystick motion. I eventually took the joystick apart, removed the "stick" and machined the end of a long aluminum rod to match the end of what was inside of the joystick so it became one long continuous shaft. Most of this work was done using a lathe, a file and scotch brite for the final smoothing of the end.
The arm that holds the pendulum above the table was something that I had salvaged out of a discarded store display from a temp job that I worked at years ago. It was sitting around my shop for years before I put it to use. It needs to be adjustable for a few different reasons: 1) if you want stronger magnets, the easiest way to make this happen is to stack them. When you stack them, you need to adjust the height of the arm to compensate for the height of the magnets. 2) At one time, I had not used it for a few months and when I went to use it, I noticed that the magnet at the end of the pendulum had broke free of the epoxy glue and the magnetic attraction over time had been enough to break that bond. Since then, when it is not in use, I adjust the arm so the pendulum rests on the table and no additional forces are trying to pull the magnet free of the epoxy bond. I have been asked where one could obtain the parts to make a similar arm. The closest thing that I have found is available at Edmond Optics: http://www.edmundoptics.com/optomechanics/optical-breadboard-components-laboratory-tables/posts-post-holders/articulating-arm-mounting-systems/1369
There is a post on a forum here: http://www.muffwiggler.com/forum/viewtopic.php?t=76458
where synthesizer enthusiasts have talked about building their own version of this. The discussion came up of having the table be curved and maybe a wok could be used. This is a great idea and though it would certainly eliminate the need to stack the magnets at different heights for the outer extensions of the pendulum, it would drastically cut down on the ability to see the center positions of the pendulum and might not be as visually satisfying. Honestly, stacking magnets is not a big deal.
One last note/comment on the electronics inside of the pendulum. I simply have a 9V battery going into a couple of LDO linear regulators with input and output capacitors for each X and Y position. The regulators make the position consistent and repeatable as the battery dies. Realistically, I change the magnet position so often, that consistency and repeatability are not really practical issues. Just running the 9V battery through the joystick attenuators would be sufficient. You could also easily modify this to add oscillators, filters, etc into the box. I left mine as a CV controller to modulate external oscillators, filters, etc. The knobs on mine strictly limit the range of the joystick itself by adding series resistors in the voltage divider. One could also process the CV through an opamp and make the knobs control offset and amount.