Vol.2 No.4 2010

32/68

Research paper : How the reliable environmental noise measurement is ensured (R. Horiuchi)−262−Synthesiology - English edition Vol.2 No.4 (2010) also technically supported the tests for domestic calibration service providers. Constituent organizations were assessed for their calibration procedures and uncertainty evaluation methods by a third party to verify their measurement capability objectively.Development of individual elemental technologies and their integration by the traceability system resulted in an international scheme of providing the reliable environmental noise measurement results for end-users.3 Microphone as national standard3.1 Primary calibration of pressure sensitivityThe microphone as a national standard (laboratory standard microphone) has two requisites: the stability of pressure sensitivityTerm 3 and the established method for precise absolute calibration (primary calibration) of the pressure sensitivity. The coupler reciprocity method[15] is used for primary calibration of pressure sensitivities on laboratory standard microphones. As shown in Fig. 4, the sound wave is kept within the acoustic coupler (a cavity with small volume) during calibration. Other calibration methods were not adopted because one cannot cover the whole audible frequency range (20 Hz to 20 kHz)[16] and the others have a drawback of having larger uncertainties[14,17].NMIJ/AIST introduced digital signal processing technique into the coupler reciprocity method and improved the calibration uncertainty from the point of view of noise reduction, attenuator calibration and cross-talk minimization[18].Firstly, digital signal processing by using an FFT analyzerTerm 4 was adopted as a new noise reduction technique because analogue signal processing by a traditional filter with narrow bandwidth had a problem in measurement repeatability. Synchronous average method by using a built-in signal source of the FFT analyzer improved measurement repeatability from 0.02 dB to 0.007 dB in the main frequency range and reduced measurement time by half.Secondly, small output voltage (0.1 mV to 0.8 mV) which was attenuated to adjust a signal level could not be directly measured with small uncertainty to that point[19]. If the ratio of attenuation is pre-determined, small uncertainty can be ensured with only the measurement of large input voltage applied to the attenuator. Thus, the advanced calibration method for attenuators was developed by using the FFT analyzer and it decreased the uncertainty from 0.01 dB to 0.001 dB.Lastly, measurement circuits were re-designed to minimize the cross-talk. It is one of the uncertainty factors, meaning that a signal is wrongly mixed due to bypassing the unexpected paths. Severe measures were taken as in the high frequency circuits and the uncertainty caused by the cross-talk was decreased from 0.01 dB to 0.001 dB.3.2 Instability of pressure sensitivityHowever, deviation of the pressure sensitivity still remained after the calibration system was improved. The author supposed that the cause of instability might be inherent in the microphone to be calibrated and considered various causes to verify this hypothesis. Theoretical analysis on frequency characteristics of the sensitivity deviation revealed that the microphone is deformed when contacting surfaces between the microphone and the coupler are sealed with grease and that the change of microphone’s acoustic characteristics results in instability of the pressure sensitivity[20].Such deviations were remarkable to a specific type of domestic laboratory standard microphones and thus this type was not adopted as an acoustic standard[21]. At present, the uncertainty due to instability of the microphone sensitivity is 0.012 dB for LS1P microphones and 0.008 dB for LS2aP, respectively (Refer to Fig. 1 for the difference between LS1P and LS2aP microphones).3.3 Measurement uncertainty of pressure sensitivityAs a result of improvement, of the uncertainty components reformable at the state of the art, the influence could be minimized to a negligibly small level. Remaining uncertainty components are the instability of the microphone sensitivity and the internal volume of the coupler. Volume uncertainty of the coupler used for LS1P microphones (its internal volume is approximately 20 cm3) is 0.008 dB and that for LS2aP (1 cm3) is 0.015 dB, respectively. In the main frequency range, the uncertainty (95 % level of confidence) of the pressure sensitivity for both LS1P and LS2aP microphones is evaluated to be 0.04 dB. This uncertainty is half of what it was before [22].Fig. 5 Anechoic chamber of NMIJ/AIST.Many sound absorbing wedges made of glass wool protrude from the inside wall of the anechoic chamber. Plywood was put on the wire meshed floor for a better view.

※このページを正しく表示するにはFlashPlayer9以上が必要です