A Photonic Microphone in a Popular Transistor Outline Package Based on Self-Mixing Interference

Abstract

High-performance and small form-factor acoustic sensing is critical for speech-based human–machine interaction and industrial monitoring. In this article, we propose a photonic microphone based on self-mixing interference (SMI), which optically transduces the motion of a compliant membrane to deliver an output signal that is linearly correlated with acoustic pressure, thereby enabling direct sound readout. The device features fully integrated photonic and mechanical components in a popular transistor outline (TO) package, leading to an ultracompact geometry compared to other existing photonic sensors. Owing to its optical interference mechanism, this microphone exhibits an exceptional sensitivity of −15.10 dB (ref: 1 V/Pa) and achieves a minimum detectable pressure (MDP) of 6.45 dB (ref: $20~\mu $ Pa) under standard free-field test conditions at 1 kHz. Within the frequency range of everyday speech (60–2000 Hz), the microphone exhibits a response deviation of only ±0.85 dB, demonstrating excellent performance in weak signal detection and high-fidelity sound cloning. These results underscore its promising potential, surpassing most of the existing capacitive and piezoelectric microphone chips.