This invention relates to a means and method for the normalization and calibration of the microphones in sound-intensity probes.
Sound-Intensity Probes
The sound-intensity vector is the time average of sound-power flow per unit area expressed in spectral form.
The sound-intensity probe that is currently in greatest use consists of two microphones that measure a single component of the vector along a line joining the microphone centers. Usually the measurement is made in a direction perpendicular to a surface, such as a hypothetical surface enclosing a sound source or the surface of the source itself. Such probes are described in
Recently a new acoustic instrument, the acoustic vector probe (AVP) was developed by
The AVP is used principally for locating and quantifying sound sources, as described in
In order for these two types of probe to measure sound intensity accurately, the microphones have to be corrected so that their response is substantially identical over the frequency range of the measurement. This is particularly important for AVPs because, to determine the direction of a sound source accurately, the probe has to be omnidirectional, i. e. with a sensitivity that has no directional bias.
Composite sound-intensity probes having a common coordinate system and measurement point can be constructed, consisting of nested arrangements of either the two-microphone probe or the AVP. These arrangements increase the frequency range of the measurement by extending measurement accuracy for higher and lower frequencies. As before, the microphones in these probes have to have a response that is substantially identical to achieve the required accuracy.
Currently microphones used for sound-intensity measurement are assumed to have a flat response over the frequency range of the measurement. The response is generally depicted on a decibel scale where deviation from flatness appears less significant. Using the flatness assumption, microphones are calibrated and phase-matched at a single frequency, typically about 250 Hz. The calibration and phase-matching are then considered to apply over the appropriate frequency range, as described in Reference 1 and in
The present invention includes and utilizes an apparatus and method for making the microphones of a sound-intensity probe, or of a composite of such probes, have a substantially identical response with a standard comparison microphone, by determining the transfer functions between the microphones of the probe and the comparison microphone. The purpose is to improve the accuracy of sound-intensity measurement, particularly in determining the direction of a sound source.
The apparatus includes a normalizer-calibrator tube with a loudspeaker at one end and a fixture at the other end that holds the microphones of the probe, along with the comparison microphone. The comparison microphone is stable and has known acoustical characteristics provided by the manufacturer. The microphones are all flush with the fixture's inner surface where they are simultaneously exposed to plane waves proceeding down the normalizer-calibrator tube from the speaker. In general the speaker emits pseudo-random white noise or other broadband time-invariant or stationary signals. Standing-wave sinusoids in the normalizer-calibrator tube are absorbed by quarter-wave attenuators protruding from the side of the tube. The attenuators are a series of narrow tubes with openings flush with the wall of the normalizer-calibrator tube and with the outer ends closed. The attenuators decrease in length from a maximum that is essentially half the length of the normalizer-calibrator tube down to a small minimum length, thereby absorbing standing waves from the lowest possible frequency up to high frequencies. The attenuators protrude in two banks. One protrudes to maxima at the ends of the normalizer-calibrator tube and decreases to a small minimum at the center. This absorbs the even standing-wave sinusoids. The other protrudes to a maximum length at the center of the normalizer-calibrator tube and decreases to a small minimum length at the ends. This absorbs the odd standing-wave sinusoids.
The microphones in the probes are preferably small electret microphones such as the FG series available from Knowles Electronics LLC, of Ithaca Ill. The Knowles microphones are omnidirectional and small, having outer diameters less than 2.6 mm with similar body lengths. Despite their small size they have a sensitivity of about 22 mV/Pa, which is comparable to the sensitivity of larger microphones. A standard condenser microphone with known acoustical characteristics is used as a stable comparison microphone for normalization and calibration of the microphones in the probes.
There are two types of sound-intensity probes. One is a side-by-side arrangement of two microphones that are inserted together with the comparison microphone in the fixture at the end of the normalizer-calibrator tube. The other probe is an acoustic vector probe (AVP) with four microphones in the regular tetrahedral arrangement pointing in pairs in opposite directions. The microphones of the AVP are inserted in the fixture, one pair at a time, forming a line on either side of the comparison microphone. The comparison microphone passes through the center of the probe and is located centrally in the fixture.
Each type of sound-intensity probe can be combined with the same type of probe to form a composite probe that extends the frequency range of the sound-intensity measurement. Composite probes have a common orientation and measurement point. There are two types of composite probe, one with at least two nested arrangements of side-by-side two-microphone probes and the other with least two nested arrangements of AVPs. The constituent probes are chosen to cover different parts of the acoustic frequency range. The fixture in the end of the normalizer-calibrator tube can hold at least four microphones of a composite probe, together with the comparison microphone
The normalizer-calibrator system is used to determine the transfer function between each microphone of a sound-intensity probe and the comparison microphone. When measuring sound intensity, the spectral form of the sound pressure measured at each microphone in a probe is multiplied by the corresponding transfer function. This makes the microphones have substantially the same response as the comparison microphone. In this way the responses of all the microphones in the probe appear identical and the probe is essentially omnidirectional. The sound-intensity vector can then be calibrated using the known acoustical characteristics of the comparison microphone to provide accurate measurements of sound intensity.
In the Drawings:
a) and (b) show plan and elevation views of how the microphones of probe 50 can be inserted into the fixture 76 in relation to the comparison microphone C.
The use of transfer functions in the normalization and calibration procedure can be described mathematically as follows. Standard DFT (digital Fourier transform) techniques are performed in the microprocessor to determine the transfer function H1C(f) between microphone 1 (for example) and the comparison microphone C, as follows
H1C(f)=G1C(f)/G11(f) (1)
where G1C(f) is the cross-spectrum between the signal at microphone 1 and the calibration microphone C, given by
G1C(f)=FpC(f)·Fp1(f)* (2)
and G11(f) is the auto-spectrum of the signal at microphone 1 given by
G11(f)=Fp1(f)·Fp1(f)* (3)
where the asterisks denote the complex conjugate. To make the signal Fp1(f) at microphone 1 look like the signal FpC(f) at the calibration microphone C, it is multiplied by the transfer function in Equation (1) to give
Fp1C(f)=Fp1(f)·H1C(f) (4)
The process is repeated for microphone 2 using relations corresponding to Equations (1) through (4) with 2 substituted for 1, as follows
H2C(f)=G2C(f)/G22(f) (5)
where
G2C(f)=FpC(f)·Fp2(f)* (6)
and
G22(f)=Fp2(f)·Fp2(f)* (7)
To make Fp2(f) look like FpC(f), Fp2(f) is multiplied by the transfer function in Equation (5) to give
Fp2C(f)=Fp2(f)·H2C(f) (8)
For the four-microphone AVP, transfer functions for microphones 3 and 4 are obtained in the same way by reversing the vector probe and rotating through ninety degrees so that the tubes 60 are inserted into the fixture 76 placing microphones 3 and 4 in the same plane and in line with the comparison microphone C. In this way all four microphones in the probe can be made to look like the comparison microphone C, making the sensitivity of the probe omnidirectional and calibrating the individual microphones using the known acoustical characteristics of the comparison microphone. Similar procedures can be used for the microphones of the composite probes 150 and 200. The transfer functions are stored in the signal processor for later use in measurements with the probes. Calibrations based on the known acoustical characteristics of the comparison microphone are applied in the digital signal processor for accuracy in the measurements.
While the invention has been described by reference to certain preferred embodiments, it should be understood that numerous changes could be made within the spirit and scope of the inventive concepts described. Accordingly it is intended that the invention not be limited to the disclosed embodiments, but that it have the full scope permitted by the language of the following claims.
THIS APPLICATION IS A CONTINUATION-IN-PART OF U.S. patent application ENTITLED “ACOUSTIC MEASUREMENT METHOD AND APPARATUS” Ser. No. 10/396,541, FILED 2003, Mar. 25, now U.S. Pat. No. 7,058,184 AND OF CONTINUATION-IN PART ENTITLED “SOUND SOURCE LOCATION AND QUANTIFICATION USING VECTOR PROBES” Ser. No. 10/746,763 FILED 2003, Dec. 26, now U.S. Pat. No. 7,054,228 BY ROBERT HICKLING THE PRESENT INVENTOR.
| Number | Name | Date | Kind |
|---|---|---|---|
| 1356359 | Heck | Oct 1920 | A |
| 1924091 | Eisenhour | Aug 1933 | A |
| 4236040 | Chung | Nov 1980 | A |
| 4421957 | Wallace, Jr. | Dec 1983 | A |
| 4463453 | Cohen et al. | Jul 1984 | A |
| 5125260 | Hedeen | Jun 1992 | A |
| 5172597 | Hedeen | Dec 1992 | A |
| 5363452 | Anderson | Nov 1994 | A |
| 5567863 | Larson et al. | Oct 1996 | A |
| 7068801 | Stinson et al. | Jun 2006 | B1 |
| Number | Date | Country |
|---|---|---|
| 2864199 | Dec 2003 | FR |
| Number | Date | Country | |
|---|---|---|---|
| 20070223730 A1 | Sep 2007 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10746763 | Dec 2003 | US |
| Child | 11415910 | US | |
| Parent | 10396541 | Mar 2003 | US |
| Child | 10746763 | US |