Design and analysis of low-acoustic noise motor drivers

Design and analysis of low-acoustic noise motor drivers with random pulse-width modulated inverter based on programmed PWM is presented in this paper. In motor drive applications, the high switching frequencies often results in acoustic noises, which make operator feel uncomfortable. The programmed PWM are used to solve nonlinear equation eliminating harmonics with initial value obtained by equal area PWM to obtain the switching angle. More than 100 sets of switching patterns eliminating different harmonics are solved by the proposed method to randomly sample at every period to generate PWM waveform. Random probability function is used to obtain the harmonic spectrum distribution by random sampling method to determine the output voltage waveforms. For each cycle, the PWM switching patterns are programmed to generate random selected harmonics to spread out the side-band harmonics reducing annoying acoustic noises. The random selected harmonics are calculated by using Gaussian distribution to minimize the acoustic noise and harmonic losses. The voltage spectrum is analyzed using fast-Fourier transformation (FFT) by TI TMS320F2407 which is used to determine the PWM switching pattern of the next cycle to minimize the concentrated sideband acoustic noises.