The directivity of a loudspeaker is gated by the frequencies of the waves it is required to emit. Tones audible to the human ear have relatively long wavelengths, and as directivity is proportional to the ratio of the size of the speaker to the length of the wave, a practical loudspeaker will emit sound essentially omnidirectionally. One novel solution to this problem is through the utilization of ultrasonic speakers, emitting frequencies inaudible to the human ear. Audible frequencies can piggyback on the ultrasound passed to the speakers, attached in much the same fashion as an AM radio transmitter. Due to the nature of nonlinear acoustics, audible sound is emitted from the ultrasound itself, allowing for highly directed beams of audio.
Achieving a sharply directed beam begs a further question: how can the direction of the beam be changed? If a listener using an ultrasound-to-sound speaker moves from the audio beam, the speaker must be manually rotated to allow for uninterrupted listening. This problem may be solved through the use of an ultrasound-to-sound phased array, which allows for electronic steering of the audio beam. Phased arrays consist of elements that emit with a calculated phase difference between each other, resulting in a directed beam along the phase-aligned angle. This phase difference may be produced in a number of ways, many of which were investigated and compared over the course of the research process.
The research presented here poses a tapped analog delay scheme as a potential implementation for phased ultrasound-to-sound speakers.
Simon Fink, ’17 Physics & Engineering Brookfield, VT Derin Sherman