UC Irvine

In recent years, the consumer electronics market for battery-powered devices such as smartphones and tablets has been rapidly expanding. The requirements for audio CODEC in these portable devices have extended from merely supporting voice calls to high-fidelity music playback. As a result, audio driver performance has become one of the most important differentiating factors among products from different suppliers. There are three basic performance metrics that are typically used to benchmark audio modules: the maximum delivered output power, the audio fidelity measured in terms of dynamic range, THD+N, and finally the battery life. Maximizing all three of these performance metrics has proven to be an exceptionally hard task as portrayed by the research publications.

This work presents an attempt to push all three of these metrics together and provide an acceptable balance which is achieved by selecting the right topology. Conventionally, headphone drivers are designed using a linear amplifier topology for many reasons- most prominently- to achieve a superior THD+N and PSRR requirement which in the past was essentially the only key performance metric needed. This came at the expense of realizing mediocre power efficiency targets, thereby wasting battery life. This picture changed dramatically over the last decade with smartphones and other portable devices becoming the first choice of the young generation. These devices are extremely power hungry due to the unlimited functions and features they provide and therefore battery life has come to the spotlight as a key resource that need to be preserved. As a result, in this work a headphone driver is based on a switching topology that is able to deliver more than 230mW of power (or equivalently 2Vrms) to a 16Ω load while achieving better than -98dB of THD+N , more than 108dB of SNR, and about 108dB PSRR while still maintaining a peak power efficiency of more than 84%.