Digital Implementation of Reference Signal in Active Noise Control System for Blade Passing Frequency Noise of Vacuum Motor

Abstract

This paper deals with active noise control problem for blade passing frequency noise of vacuum motor. In order to make the system inexpensive, we propose a digital synthesis method of a rectangular reference signal with 25% duty ratio whose fundamental period is approximately 1/8 to rotational period by using a counter and interrupt processing on digital hardware (proposed method 1). Furthermore, an anti-aliasing filter is integrated into the previous method in order to reduce aliasing components in sampled reference signal (proposed method 2). These methods are examined by control experiments with inexpensive hardware (dsPIC): First for simplicity of experiment, only the 1st blade passing frequency noise is considered as control target, and an open-loop control system is examined where the controller's coefficients are tuned manually so that error microphone output is minimized. The resultant coefficients are strongly dependent on the vacuum motor's load, which indicates that adaptive control algorithm which uses error microphone information is necessary to consider load change; Second, an adaptive control system is examined based on the standard filtered-x LMS algorithm. The similar performance to the open-loop control is obtained regardless of load change, which indicates that the optimal coefficients are automatically obtained by adaptive control; Finally, the 1st and 2nd blade passing frequency noises are both considered as practical situation. By comparing experimental results for both methods, it is shown that the aliasing components in error signal by proposed method 1 is remarkably attenuated by proposed method 2, which concludes that the proposed method is available to realize inexpensive active noise control system for blade passing frequency noise of vacuum motor.