PROCEEDINGS of the 22 nd International Congress on Acoustics Acoustical Measurements and Instrumentation: Paper ICA2016-859 Broad band air ultrasound reference sound source Angelo Campanella Campanella Associates, USA, a.campanella@att.net Abstract New air ultrasound (AU) reference sound sources are produced to provide a steady and reliable source of air ultrasound for the purpose of calibrating air ultrasound microphones over the frequency range from 10 khz through 400 khz. The air ultrasound field is created by mechanical means. One source produces AU by shear turbulence at the surface of a rotating cylinder. AU field intensity from 10 khz to 100 khz is determined at manufacture with a calibrated microphone located at a fixed point 0.5m from the rotating cylinder. A second source is created to operate up to 400 khz by ultrasound emission from a small jet of escaping compressed air. The AU intensity at a point 8 cm from the jet source is determined at manufacture by the reciprocity calibration technique. Calibration data for both units are presented. Keywords: air ultrasound (AU), reciprocity, calibration, air jet, shear turbulence.
1 Introduction Since 1945 Air Ultrasound (AU) up to 100 khz has been found to be emitted by meadow insects [1], [2]. Previously it was identified that bats use AU to navigate and to find prey. AU usage in more recent years has extended to acoustical modeling, to detecting compressed gas pipeline leaks, for proximity sensing and for studying animal bioacoustics. A calibrated AU microphone is required to support these. Some common ½, ¼ and 1/8 condenser microphones have various degrees of ultrasound sensitivity. Small capacitor microphones as well as low cost MEMS microphones sensitive to AU have been developed recently. In turn, a new stable source of test AU for qualification of these microphones is reported. 2 The need It is essential for ultrasound microphone qualification that a reliable broadband source of AU be available that produces a steady AU field of sufficient intensity as to cause a microphone output signal that is much greater than the preamplifier self-noise so that reliable microphone sensitivity measurements can be made. 3 The first solution In 1997, the RSS101U, an AU source whose field was calibrated to 100 khz was produced to satisfy industrial AU microphone needs. See Figure 1. This unit produces AU by means of shear air turbulence at the surface of a spinning cylinder of fine screen. The test microphone is placed where AU intensity is sufficient to produce good microphone signal-to-noise ratio (S/N), found in the equatorial plane of the spinning cylinder at the specified distance of 0.50 m from the cylinder axis. The SPL levels found there are shown in Figure 2. Several RSS101U units are now in use in USA industries that manufacture gas leak detectors, where gas leak AU emission below 50 khz is typically sensed. The AU emission level of the RSS101U at frequencies above 100 khz is relatively feeble for microphone sensitivity checks. 2
Figure 1 RSS101U http://www.campanellaacoustics.com/rssman.htm#rss101u Figure 2: Table of RSS101U AU source sound level 3
4 The second solution Recent need: There is now a desire to investigate the living AU sounds emitted above 100 khz by small mammals (viz mouse pups) and insects (viz ants). This introduces the need for calibrated broadband AU microphones for use up to 300 khz and above. Means to produce broadband calibration quality AU beyond 100 khz which is both intense and cost effective have been meagre, among which is found the ionophone, [5], [6] which produces audio and AU by a radio frequency (RF) air corona discharge by an RF source that is amplitude modulated (AM) at the desired audio frequency. However, the ancillary air heating and the ozone created by the high voltage air corona discharge must be vented away. This air corona discharge also radiates radio frequency interference (RFI) energy that must be shielded against in nearby electronics systems. The solution: Intense shear for all liquid and gas jets occurs at the boundary between that jet stream and the static medium into which it progresses. This shear produces broadband sonic waves (large jets) and AU waves (tiny jets) which radiate laterally downstream along a conical surface of about 25 degrees from the jet flow axis. (This mechanism is the source of the AU that is most commonly radiated by gas pipeline leaks readily detected by modern Leak Detection apparatus and technology). The RSS101U-H comprises a compressed air jet of about 0.5 mm diameter through which compressed air at 43 psi (2.93A) flows at a rate of less than 1 SCFM (<30 l/m). Figure 3: Air jet ultrasound radiation geometry. 4
All AU microphone output signal measurements, including reciprocity measurement (Figure 4, [3]) from 5 khz to 100 khz were made in 1 khz bins via a special 100 khz Larson-Davis 3200 FFT analyzer. A Kenwood 440S communications receiver was used as a roving 2.5 khz wide bin from 40 khz to 400 khz [4]. UA field SPL values were reduced to 1 khz bin values, db re 20 µpa. Figure 4: Reciprocity Calibration Procedure and data reduction The reciprocity calibration method requires three transducers (viz A, B and C, all ½ condenser microphones polarized with 200 volts): One transducer is is reciprocal ( C here); it can serve as either a microphone or as a sound emitting transducer. Capacitor microphones are capable of being driven by a sine wave oscillator with current I C to emit sound at the sine wave frequency when polarized with, say, 200 volts and is driven with a sine wave voltage that is much smaller than the polarization voltage. All measurements were carried out at a distance of 80 mm from the air jet nozzle. Transducer B serves only as a microphone that produces voltage V BC when exposed to the sound from AU emitter C. A third transducer serves only as an AU emitter, ( A here) driven with a sine wave current I A produce sound that causes voltage V CA when microphone C is exposed to its field and voltage V BA when microphone B is exposed to its field. The sensitivity of microphone B is then computed as the expression for M B in Figure 4. Finally the sensitivity, M D, of ¼ measurement microphone D was determined. Microphone D was then used to measure the AU SPL, P J of the RSS101U-H sound field at distance d. Air attenuation of the AU is calculated as e αd, where α is obtained from the AU air attenuation chart in Figure 7, and d is the distance from the jet nozzle face to the microphone diaphragm in the same units as for α. 5
Figure 5: RSS101U-H AU source sound level Figure 6: Table of RSS101U-H AU source sound level 6
5 Compensation for AU attenuation by test space air humidity AU attenuation due to air humidity (controlled by water vapor, Relative Humidity ) is estimated using Figure 7, computed from ANSI S 1.26 (alternatively ISO 9613-1). The ordinate, A, is in decibels per foot (305 mm). For RSS unit SPL data presented in Figures 2, 5 and 6, AU attenuation by the calibration site air was removed; the values presented are for no air attenuation. For RSS101U and 101U-H application by users, the user should measure the relative humidity in his test site air. The AU attenuation expected in the user test site air is estimated with Figure 7 and applied to data in Figures 5 and 6 to provide the SPL that will be found at the user test microphone or device under test location as follows: V(f)=P J (f)- A*d (1), The value of A is the ordinate found in Figure 7 for the abscissa of the site humidity% at the frequency, f, curve for P J (f) of the datum. The variable, d, is the distance of the microphone sensitive surface from the AU source surface, in feet. 10.0 Air Ultrasound Attenuation @ 20 Degrees C (68 F) per ANSI S1.26 (1995 R 2009) 400kHz 300k 1.0 db per Foot ( +10%) @ 20 Degr C. 40 k 240 k 60 k 200k 80 k 160k 120k 100 khz 20 khz 0.1 Relative Humidity------> Per Cent 0 10 20 30 40 50 60 70 80 90 100 Figure 7: Air absorption of AU, db/foot [db/(305 mm)] 7
6 Practical AU sources Figure 8 shows the RSS101U-H with an integral 60 Hz 115 volts ac air pump (North America) that supplies the necessary compressed air. The air jet orifice piece is the small white object on top of of the filter, in front of the gauge. Spectral level varies from 97 db re 20 µpa @ 75 khz to 57 db @ 400 khz. (Note: An SPL of 94 db re 20 µpa is one (1.0) Pascal RMS of acoustic pressure.) Figure 8: RSS101U-H, 60 Hz 115vac. Figure 9 shows the RSS101U-H(W) for use world wide. The air jet orifice piece is the small white object on the left of the filter regulator - gauge assembly. The RSS101U-H(W) contains the air jet, gauge and filter assembly and an adjustable air pressure regulator that will accept any instrument air supply source (=>3 Atm) via the compressed air tubing extension supplied. Figure 9: RSS101U-H(W)-World use 8
7 Typical AU microphone sensitivity determination Let upper case font symbols be in decibels, and lower case font symbols be the physical values. The RSS101U-H can be used to immediately determine the sensitivity M E (f) in decibels of any air ultrasound microphone, viz microphone E, ( device under test or DUT): Place the DUT at 8cm range and oriented as it will be used in practice. Move the DUT laterally so as to locate the AU intensity maximum. At that position and with the referred DUT orientation to the P J (f) field, measure and record the DUT output voltage in decibels, V E (f) in 1 khz FFT bins over your frequency range of interest. If your FFT bins are not 1 khz wide, then normalize that measured V E (f) value to 1 khz to compute the sensitivity M E (f) of the DUT as: V E (f) = V E (f)-10*log 10 (B), (1) Where V E (f) is the measured output level of the microphone for the user pass band centered at frequency, f, and B is the measured bandwidth in kilohertz of that user pass band, typically between the 3 db-down values of the user measurement filter. The value of v E (f) or V E (f) is determined as v E (f) = p J (f)*m E (f), or m E (f) = v E (f)/p J (f) V E (f) = P J (f) + M E (f), or M E (f) = V E (f)-p J (f) (2a) 2b). Where P J (f) is provided in Figure 6. It is likely that the user will want to normalize the microphone E output spectrum display level to present a flat data spectral response FFT display via, m 0, a constant. Let normalization n(f) be such that m E (f)*n(f)=m 0 =n(f)*v E (f)/p J (f) (3) n(f)=m 0 *p j (f)/v E (f). (4) In decibels, the normalization for a flat spectral display will be: N(f) = M 0 + P J (f) V E (f) (5). M 0 is a constant chosen to size the display vertical scale. P J is found in Figure 6, after correction for air humidity via Figure 7 for a chosen working distance, V E is measured as the microphone output signal voltage in frequency bins, e.g. a Spectrum Analyzer display, when the calibrated DUT is applied in AU field of an AU emitting specimen under study. 9
8 Conclusions Sources of steady and reliable broadband air ultrasound (AU) are described. These can be used to determine the sensitivity of any microphone to AU. These reference sound sources emit AU from 10 khz to 100 khz for the RSS101U and from 10 khz to 400 khz for the RSS101U- H when 115 vac power is applied. The RSS101U-H(W) can be used anywhere in the world when compressed instrument air from an external source is applied. --------- References [1] Dethier, V. G. "The physiology of insect senses", Wiley, NY1963, p 91. [2] Pielemeier, W.H.. JASA 17,4, April, 1946, p 337-8. [3] Rudnick, I. and Stein, M. N., Reciprocity free-field calibration of microphones to 100 k.c. in air, JASA 20 (6), pp 818-825. [4] Campanella, A. J., Finite amplitude distortion of spherically diverging intense sound waves in air, JASA 67 (1). pp 6-14. [5] Oda F. Corona type loudspeaker (Masters Thesis, Pennsylvania State University, 1957). [6] Ackerman, E, Anthony A. and Oda F. Description of a Corona-Type Loudspeaker Used in Biological Research. EXPERIENTIA, Vol.XIV / 10, 1958, pag. 384. (Separatum pp 1-6.) Birkȁuser Verlag, Basel, Schweiz. ~~end~~ 10