Phonocardiogram (PCG) is a graphical representation of heart signals, which undergoes analysis to assess the cardiac mechanical function. The primary challenge faced while recording the Phonocardiogram signals is contamination from surrounding noise signals. Consequently, it becomes essential to remove noise from the PCG signal before utilizing it for more sophisticated processing. The present study employs a highly effective adaptive filter architecture deploying the LMS technique to accurately evaluate a clean version of the signal. The filter model, based on Adaptive Noise Cancellers, aims to efficiently denoise and recover the PCG signal. Within the suggested filter configuration, a noisy signal undergoes processing through a single-stage recursive adaptive filter structure. The proposed configuration automatically adjusts the figure of stages to be is iterated and the increment magnitude for each stage. A feedback loop is used to reuse the same adaptive filter ‘n’ number of times. The Single-Tier Recursive LMS-driven adaptive filter model's capability to filter out noise from the PCG signal has been evaluated for its effectiveness. The experimental data used in this study is sourced from the Physionet database. It is strategically interfered with by Gaussian-distributed and pink spectral noise signals with varying input strength of signal relative to noise (SNR) magnitudes to examine optimal adaptive filter structure utilizing the LMS algorithm.