@ James, not really, you just make the mount fixed (as I’d imagine it already is), get rotary encoder (along with an angular rate gyro if you like), and you know the mount’s relation to NSEW, you know the speed of the rotation, and you know the actual angle as well as which sensors along the edge are actually being activated the most. (I think that might work).

@Finger

A digital compass would probably be cheaper & easier to implement, as long as the case doesn’t have too much metal

Why ue wind to charge it when theres a perfectly good solar power unit??? plus having it spin would cause all the sounds play at once!

I’m surprised the arduino-haters haven’t flown in yet…

This is an awesome project!

Powering the device by wind demands a certain amount of breeze above a particular speed over time in order to keep the system charged up. If wind flux (aeolian flux? surely there’s a real term for this!) is less than solar flux, then solar is better.

Given that these devices will generally be used in areas which are relatively sheltered, the breeze will likely be sporadic, light and turbulent, and not so great for wind power.

For a related device, look up ‘aeolian harp’. The guy at ixian.ca has made one (see here: http://ixian.ca/gallery/thumbnails.php?album=25) and many other impressive bits of engineering.