Audiosonic Acoustic Detection of Air Corona Discharge based on Fast Fourier Transform

In this paper, acoustic emission in audiosonic range of air corona discharge was detected by using fast Fourier transform (FFT). A microphone was employed to record the acoustic signal of corona that was generated by using a needle-plane electrode with air gap. A high frequency current transformer (HFCT) sensor was used to acquire the partial discharge (PD) current pulses. Furthermore, the PD current pulses were correlated to the recorded acoustic signal. The given ac voltage stress was increased until the light spot appeared on the tip of the needle. The results showed that the acoustic emission of corona could easily be detected by using FFT. Based on the frequency spectrum of acoustic signal, the presence of corona was seen on the high frequency, whereas the noise was found on the low frequency. The high frequency amplitudes created a unique pattern like hilly shapes. Hence, the high pass filter that was applied to the acoustic signal was effective to separate the acoustic signal of corona from the noise. The higher voltage stress, the more visible PD current pulse and the greater high frequency amplitude of acoustic frequency spectrum. However, the occurrence of corona was firstly known by its PD current pulse rather than its acoustic signal.