Digital sound reconstruction using arrays of CMOS-MEMS microspeakers

We report on the development of a direct digital method for producing sound using an array of CMOS-MEMS microspeakers. In a previous article [1], we described the theory behind digital sound reconstruction, characterized the acoustic response of a single CMOS-MEMS membrane, and showed experimental results involving a 3-bit (7 speaker) array composed of individual microspeaker chips. In this work, we report on the design, fabrication, and operation of an 8-bit (255 speaker) array integrated on a single chip. With this new device, we have been able to further study sound reconstruction and understand the sources of harmonic distortion and waveform quality that are essential to creating a complete digital sound system.