Wearable multisensor devices enable continuous cardiac monitoring, but high‑fidelity audio transmission poses challenges due to high data‑rate requirements and low‑power Bluetooth constraints. This thesis presents the development and implementation of an embedded audio driver for real‑time cardiac auscultation on an STM32U545 microcontroller. The system integrates a MEMS microphone through an I²S interface using DMA and double‑buffering for low‑latency and deterministic acquisition. A digital IIR filter, designed in MATLAB and implemented on the microcontroller using a CMSIS‑DSP biquad cascade structure, preprocesses phonocardiographic (PCG) signals to enhance clinically relevant frequency bands (20–400 Hz for S1/S2 sounds and murmurs) with controlled gain while attenuating undesirable noise components.