meh. this was an experiment -- actually, the repeat of an experiment i did about 20 years ago -- a taut wire as the resonant element in an electronic feedback loop. as kludgey as the setup looks, it's mechanically very stable. too stable, in fact, as the previous experiment had all sorts of interesting characteristics that this experiment revealed as a likely due to a lack of stiffness in the original -- which was a frame made from crap wood.
singing wire test 1
the stiffness inherent in this setup -- steel wire fastened to a building rafter, and anchored with a fat machinist vise, didn't allow for all the wild resonances of the original (which tended to deform, just sitting there).
the electronics is simple; an automotive "brick amp" (about 35 watts RMS) driving a small coil of wire placed on a U-shaped piece of steel that forms a magnetic circuit with the steel wire. the impedance of the driver is awful, near-zero, hence it gets very hot in operation (doesn't matter, just a test hack).
the pickup is a similar magnetic device, with more and finer windings. it is biased with a small neodymium magnet (just stuck on the side) that provides a magnetic field for the moving wire to modulate and induce the signal voltage that drives the brick amp, visible here.
oh right -- sound. since this was just an experiment, i was more concerned with the physical oscillator, but i did dangle a speaker across the driver output. the sound is just audible in some of the videos.