抄録
The effects of the structure, the distribution of sensitivity and phase of each microphone unit and the dimensinonal errors on the performance of an arc-arrayed microphone system are theoretically investigated. The arc-arrayed microphone system is considered to be useful for detecting and investigating sound sources in the noise control engineering. The microphone system is composed of 101 miniature microphones arrayed on a circle at the interval of 2cm. The radius of the circle is 5m and the length of the array is 2m. First, the effect of the weighting function on the spatial distribution of sensitivity is investigated with uniform weighting, Hamming weighting, the composite weighting with the former two and Blackman weighting. The Hamming weighting is considered to be the best in the view point of practical use. Next, the effects of the distribution of sensitivity of microphone units on the spatial distribution of system sensitivity are investigated. The deviation of sensitivity in magnitude does not affect so much on the peaks of the side robes but the deviation in phase does. The deviation of the sensitivity is measured for 150 miniature microphone. The standard deviation in sensitivity is about 1 dB in magnitude and that in phase is negligible. Comparing the result of measurement with the theoretical investigation, it is concluded that the commercialized miniature microphones can be used for the unit of the arc-arrayed microphone system. The effects of the dimensional accuracy of the array are then investigated. The deviation of the distance between every microphone unit gives little effect on the shape of the side robe. But, the accuracy is strictly required in the distance from the focul point to every microphone unit, especially in the high frequency range. To suppress the magnitude of side robes less than -40 dB comparing with that of the main robe, the deviation should be 1/40 of the wave length.
