Mechanical beam steering — mirrors on motors — works but is slow and bulky. Solid-state alternatives use pixelated phased arrays: many tiny emitters, each controlled independently, synthesizing a beam direction from interference. This works but requires complex fabrication and hundreds or thousands of control channels.
Acousto-optic beam steering is a third path. A sound wave propagating through a crystal creates a periodic density variation — a diffraction grating that moves. Light passing through the crystal diffracts off this moving grating, and the angle of diffraction depends on the sound frequency. Change the frequency, change the beam direction. One transducer, one control signal, continuous steering.
The problem: efficiency. The acousto-optic interaction in a single pass through the crystal is weak. Most of the light goes straight through undiffracted. The conventional fix — a longer interaction region — makes the device bigger and slower.
The unconventional fix: wrap the light path in a ring resonator. The ring stores the optical power, circulating it past the acoustic region multiple times. Each pass adds more diffracted light. The resonance enhancement boosts the steering efficiency to 20% over an 18-degree field of view — comparable to phased arrays but with a single acoustic transducer.
The same lithium niobate platform that generates the acoustic waves also provides electro-optic control. By dynamically locking the ring's resonance to a chirped laser frequency, the device can perform FMCW ranging — LiDAR — in addition to beam steering. Two material properties (piezoelectric for sound, electro-optic for tuning) enable two functions (steering and ranging) on one chip.
The general principle: when a single-pass effect is too weak, recirculation can substitute for a stronger interaction. The ring doesn't change the physics of the acousto-optic effect. It changes how many times the light experiences it.