Aaagh. The vibratory feeder is turning out to be as difficult as I'd feared.

The plastic vee channel I'm using is so soft that its fundamental resonant mode occurs at pretty low frequencies - too low to make the Lego pieces dance about and separate themselves. Worse, the motor that drives it is so tightly coupled that it too gets locked into the resonances (particularly mode 2) and it becomes impossible to change its speed by simply increasing the current through it. After a lot more experimenting I'm still stuck with mode 2 that includes a dead spot (node) where the vibrations are weak, and the pieces don't move along properly.

I think what I need is a stiffer channel, so that I can drive it at higher frequencies but still below the first harmonic - i.e. the fundamental mode with the uphill end relatively fixed forcing the downhill end to waggle the strongest, no nodes. Cantilever beam, mode 1. Or even operate below first resonance?

Notes to self:

Since frequency is proportional to (K_nL)², might want to choose free-free or fixed-fixed as they have slightly higher fundamental mode frequencies. Both have nodes though. Maybe fixed-fixed, and take of the pieces just before the exit end (where the node is)?

Clearly more to do...