Conference PaperPDF Available


D. T. Martin1, K. Kadirvel1, T. Nishida1, and M. Sheplak2
1Department of Electrical and Computer Engineering
2Department of Mechanical and Aerospace Engineering
Interdisciplinary Microsystems Group, University of Florida, Gainesville, FL
This paper presents improved results for a micromachined
dual-backplate condenser microphone designed for aeroacoustic
applications. Previous microphone characterization yielded an
unacceptably high noise floor. This paper presents improvements
to the noise floor through the use of a low-noise voltage amplifier
and improved packaging. With bias voltages of ±9.3 V, an average
sensitivity of 166
V/Pa and a noise floor of 22.7 dB/Hz for a 1
Hz bin centered at 1 kHz are obtained. The noise floor is reduced
to a level below that of other MEMS-based aeroacoustic
microphones while maintaining sufficient bandwidth and
maximum pressure.
The goal of this research is to develop a microphone for
aeroacoustic testing that equals the performance of conventional
commercial measurement microphones. The applications for such
a microphone include wind-tunnel testing using arrays consisting
of a large number of microphones [1]. To be suitable for
aeroacoustic measurements, a microphone must operate linearly up
to a sound pressure level of 160 dB and have a bandwidth
extending to 100 kHz. Furthermore, the noise floor should be as
low as possible. The Brüel and Kjær 4138 1/8th inch condenser
microphone meets these requirements with a noise floor of 18
dB/Hz for a 1 Hz bin at 1 kHz [2]. However, due to the high cost
of these microphones, a low-cost MEMS microphone of adequate
performance would enhance these measurements by allowing a
greater number of microphones to be used.
There have been many MEMS microphones developed in the
past; however, most of these have focused on audio applications
[3]. While there has been limited development in the area of
aeroacoustic microphones, their overall performance is far from
that of the B&K 4138 condenser microphone [4-7]. Several
previous MEMS microphones have sufficient bandwidth and
maximum pressure [4-6]; however, the noise floor of these devices
is too high.
A dual-backplate capacitive microphone has been designed to
meet the specifications for aeroacoustic measurements. Previously
reported results for this microphone yielded an unacceptably high
41 dB/Hz noise floor [8]. This was obtained using a charge
amplifier located far from the microphone, resulting in high
parasitic capacitance. This paper presents results obtained with a
low-noise voltage amplifier with packaging designed to reduce the
parasitic capacitance.
The dual-backplate capacitive microphone consists of three
circular conducting plates. These plates are separated by two air
gaps. Each layer is comprised of conductive doped-polysilicon,
thus forming two capacitors. The outer plates are perforated with
holes to allow the incident pressure to deflect the diaphragm. A
photograph of the microphone is shown in Figure 1 and a
schematic cross-section is shown in Figure 2.
Figure 1: Photograph of the dual-backplate capacitive
The operation of this type of device is similar to that of the
common single-backplate microphone. As the incident pressure
impinges on the microphone, it deflects the diaphragm. This
changes the values of the two capacitors causing one to increase
while the other decreases. Bias voltages are applied to the two
backplates, and the output is the voltage on the diaphragm.
Figure 2: Cross-section of the dual-backplate capacitive
microphone showing the key elements.
The microphone is fabricated using the SUMMiT V process
at Sandia National Laboratories [9]. The 5-layer, planarized
polysilicon process is well suited for this microphone. The
conductive polysilicon is used to realize the three electrodes, and
the use of chemical mechanical polishing (CMP) in the process
results in uniform air gap spacing. This feature is also a major
drawback of the process; the use of CMP results in high variability
in the air gap thicknesses, which is manifested as a variability in
device sensitivity [8]. A summary of the microphone physical
specifications is given in Table 1.
A schematic of the microphone and interface circuit is shown
in Figure 3. The microphone, represented by C1 and C2, is biased
with symmetric voltages ±VB, and the diaphragm electrode is
connected to the voltage amplifier. Due to the small capacitance of
this microphone, it is essential to minimize the capacitance load
0-9640024-7-7/HH2008/$20©2008TRF Solid-State Sensors, Actuators, and Microsystems Workshop
Hilton Head Island, South Carolina, June 1-5, 2008
presented by the parasitic capacitance, Cp, and the amplifier input
capacitance, Ci. As these capacitances increase, the sensitivity of
the microphone is reduced [3].
Table 1: Microphone physical specifications.
Property Nominal value
Diaphragm diameter 460
Diaphragm thickness 2.25
Air gap spacing 2.0
Top capacitance 867 fF
Bottom capacitance 632 fF
Figure 3: Circuit schematic showing the microphone and buffer
To achieve minimal attenuation due to these capacitances, an
amplifier with a very low input capacitance is used. The SiSonic
Microphone Amplifier, courtesy of Knowles Electronics, is used
for the interface circuitry. This amplifier has an input capacitance
of approximately 0.3 pF [10].
The microphone package was designed to contain the
amplifier and the MEMS die in close proximity to minimize
parasitic capacitance introduced by cabling and electrical
interconnect. This hybrid package was also designed to minimize
interference with the acoustic field by mounting the microphone
flush with the front surface of the package. A photograph of the
assembled microphone package is shown in Figure 4. In sequence,
the microphone is first mounted in a printed circuit board (PCB).
The amplifiers are mounted on the reverse side of the circuit board.
This preserves the flush surface, while minimizing the parasitic
capacitance. This PCB is then secured in a Lucite block designed
to be compatible with the acoustic test setup. Photographs of the
front- and rear-views of the microphone package are shown in
Figure 5. This shows the flush-mounted microphone die, which
contains four microphones as well as several test structures. Gold
wirebonds connect the MEMS die to the circuit board. These bond
wires are covered in a protective epoxy.
The final package has a frontal surface area of 0.75 in. x 0.75
in. This is considerably larger than an individual microphone.
However, each die contains four microphones, thus the package
contains four amplifiers. A smaller package footprint is possible
for a re-designed die containing only one amplifier.
Previous results reported for the microphone employed a
charge amplifier located off-package [8]. In this configuration, a
sensitivity of 390
V/Pa was achieved with a noise floor of 41
dB/Hz at 1 kHz. The resonant frequency of the microphone was
determined to be 178 kHz. The dominant source of noise for this
initial characterization was due to the charge amplifier interface
circuit. The results presented in this section show a significantly
improved noise floor with a modest reduction in sensitivity.
Figure 4: Photograph of the packaged microphone.
Figure 5: (Left) Front view of the microphone package showing
the flush mounted MEMS die. (Right) Back view of the PCB
showing four covered, mounted amplifier die.
A total of seven microphones are characterized with the
improved circuitry. Experimentation is performed to determine the
microphone linearity and frequency response. These acoustic tests
are performed using a plane wave tube [11] with a maximum
frequency of 20 kHz, as shown in Figure 6. The noise floor of the
microphone is estimated by measuring the output of the
microphone/amplifier system without an acoustic input. Bias
voltages of ±2 V are used unless otherwise specified. The
experimental setups are described in detail by Martin et al. [8].
Figure 6: Experimental setup for acoustic experiments.
Figure 7 shows the output of the microphones for varying
incident pressure. The amplitude of a 1 kHz tone is varied from
approximately 40 dB to 160 dB re. 20
Pa. The output voltage of
each microphone is plotted vs. the incident pressure. The dashed
lines indicate the predicted output voltage. The average sensitivity
is 166
V/Pa with a variation of ±20
V/Pa between the seven
Figure 7: The output voltage of seven microphones vs. incident
pressure at 1 kHz.
The microphone output saturates near 160 dB; at this pressure,
the microphone output voltage exceeds the input voltage of the
amplifier. To quantify the distortion generated by the microphone,
the bias voltage is reduced to 2 V to lower the sensitivity. This
eliminates any distortion generated by the amplifier for the range
of sound pressure levels tested. The total harmonic distortion of
the seven microphones is given in Figure 8. The microphones
show an average distortion of 4% at 164 dB.
Figure 8: Total harmonic distortion of the seven microphones
versus incident pressure at 1 kHz.
The frequency response of the seven microphones is
measured from 300 Hz to 25 kHz. The magnitude response is
plotted in Figure 9 and the phase response is plotted in Figure 10.
The upper frequency for plane-waves in the acoustic waveguide is
20 kHz. Below this frequency, the device under test and a
reference microphone are exposed to the same incident pressure.
However, above this frequency, the two microphones are sensing a
different pressure. Therefore, the magnitude response above 20
kHz is a qualitative measurement demonstrating microphone
sensitivity up to this frequency. This measurement confirms the
close sensitivity matching of the seven microphones.
The phase response shown in Figure 10 is plotted from 300
Hz to 20 kHz. For the majority of the frequency range, the phase is
matched to within ±1º. The dip near 10 kHz corresponds to a
reduced output of the acoustic driver. The cut-on of higher modes
in the plane wave tube near 20 kHz is evident in the phase
Figure 9: Magnitude response plotted over the range of 300 Hz to
25 kHz.
Figure 10: Phase response of the seven microphones plotted over
the range of 300 Hz to 20 kHz.
The final measurements performed on the microphones with
the improved circuitry determined the noise. The packaged
microphones were placed in a Faraday cage to reduce the effects of
electromagnetic interference [12]. With no acoustic input to the
microphone, the output noise voltage power spectral density (PSD)
of the amplifier was measured with a spectrum analyzer over the
range of 10 Hz to 100 kHz. Dividing the square root of the PSD by
the microphone sensitivity yields the input referred noise, plotted
in Figure 11. For a 1 Hz bin centered at 1 kHz, the average noise
level of the seven microphones is 22.7 dB/Hz. The input referred
noise for the seven microphones falls within ±1 dB. Table 2 shows
the average noise floor in several equivalent units including a
minimum detectable force and capacitance change. The average
A-weighted noise figure for the seven microphones is 60.4 dBA.
Figure 11: Input referred noise spectrum of the seven
These results demonstrate the importance of the interface
circuitry on the performance of the capacitive MEMS microphone.
The lower noise voltage amplifier and hybrid package reduced the
noise floor by 18 dB. This improvement was due to a large
reduction in the amplifier noise. The acoustic noise generated by
the microphone structure was less significant than the electronic
noise [8].
Table 2: The average minimum detectable signal of the seven
microphones expressed in several equivalent units at 1 kHz in a 1
Hz bin. Average Minimum
detectable Signal Value
Pressure 22.7 dB
Pressure 273 μPa
Force 15.1 pN
Displacement 75.4 fm
Capacitance 17.9 zF
The performance of the dual-backplate capacitive microphone
is compared to other MEMS microphones, as well as the Brüel and
Kjær 4138 condenser microphone in Table 3. The other
microphones are sufficient in two specifications listed; however,
all fall short of the B&K 4138 in at least one. The present work
exceeds the performance of the other MEMS microphones and
compares most favorably to the commercial microphone. It is
within 4 dB of the 4138’s maximum pressure and 3 dB of the noise
floor. However, the diameter of the dual-backplate microphone is
significantly smaller. With the appropriate packaging, this device
has the potential to enable microphone placement in locations
prohibited by the size of the B&K 4138 microphone.
Table 3: Comparison of the present work to other aeroacoustic
Microphone Diameter Max
Pressure fmax Noise
Present Work 460 μm 164 dB 178 kHz 22.7 dB1
B&K 4138. [2] 3.2 mm
168 dB 140 kHz 18 dB1
Arnold et al. [4] 1.0 mm
160 dB 100 kHz 52 dB1
Scheeper et al. [5] 3.9 mm 141 dB 20 kHz 23 dBA
Horowitz et al. [6] 1.8 mm 169 dB 59 kHz 35.7 dB1
Pedersen [7] 360 μm 140 dB 75 kHz 22 dB1
1 Noise figure at 1 kHz in a 1 Hz bin
An instrumentation-grade dual-backplate capacitive micro-
phone has been designed for aeroacoustic measurements.
Improvements in the packaging and the use of a low-noise voltage
amplifier significantly improved the device performance over the
initial characterization results. The noise floor has been reduced
while maintaining adequate bandwidth and maximum pressure.
The device characterization demonstrates performance that
exceeds that of existing MEMS aeroacoustic microphones. In
addition, the dual-backplate capacitive microphone compares
favorably to the Brüel and Kjær 4138 condenser. For applications
not requiring measurements at the extreme ends of the 4138's
dynamic range, the designed MEMS microphone is suitable.
Financial support for this project was provided by the
National Science Foundation grant #ECS-0097636 and Sandia
National Laboratories. The authors gratefully acknowledge Dr.
Pete Loeppert, from Knowles Electronics, for providing the
SiSonic microphone amplifier used as part of this work.
[1] T. J. Mueller, Aeroacoustic Measurements. Berlin, Germany:
Springer-Verlag, 2002, pp. 158–179.
[2] Brüel and Kjær, Product Data, Condenser Microphone
Cartridges—Types 4133 to 4181, p. 6. [Online]. Available:
[3] P. R. Scheeper, A. G. H. van der Donk, W. Olthuis, and P.
Bergveld, “A review of silicon microphones,” Sens. Actuators
A, Phys., vol. 44, no. 1, pp. 1–11, Jul. 1994.
[4] D. P. Arnold, S. Gururaj, S. Bhardwaj, T. Nishida, and M.
Sheplak, “A piezoresistive microphone for aeroacoustic
measurements,” in Proc. ASME IMECE, 2001, pp. 281–288.
[5] P. R. Scheeper, B. Nordstrand, B. L. J. O. Gullov, T. Clausen,
L. Midjord, and T. Storgaard-Larsen, “A new measurement
microphone based on MEMS technology,” J. Micro-
electromech. Syst., vol. 12, no. 6, pp. 880–891, Dec. 2003.
[6] S. Horowitz, T. Nishida, L. Cattafesta, and M. Sheplak,
“Development of a micromachined piezoelectric microphone
for aeroacoustic applications J. Acoust. Soc. Amer, Vol. 122,
No. 6, 2007, pp. 3428-3436.
[7] M. Pedersen, “Development of microelectromechanical
systems capacitive microphone for high-frequency
applications," in Proc. of 151st Meeting of the Acoust. Soc.
Amer. Providence, RI, 2006.
[8] D. T. Martin, J. Liu, K. Kadirvel, R. M. Fox, M. Sheplak, and
T. Nishida, “A Micromachined Dual-Backplate Capacitive
Microphone for Aeroacoustic Measurements," Journal of
Microelectromechanical Systems, Vol. 16, No. 6, 2007, pp.
[9] J. J. Sniegowski and M. S. Rodgers, “Multi-layer
enhancement to polysilicon surface-micromachining
technology,” in Proc. IEDM Tech. Dig., 1997, pp. 903–906.
[10] P. V. Loeppert and S. B. Lee, “SiSonic—The first
commercialized MEMS microphone,” in Proc. Solid-State
Sensor Actuator Workshop, Hilton Head Island, SC, 2006, pp.
[11] T. Schultz, L. Cattafesta, and M. Sheplak, “Modal
decomposition method for acoustic impedance testing square
ducts,” J. Acoust. Soc. Amer., vol. 120, no. 6, pp. 3750–3758,
Dec. 2006.
[12] R. Dieme, G. Bosman, M. Sheplak, and T. Nishida, “Source
of excess noise in silicon piezoresistive microphones,” J.
Acoust. Soc. Amer., vol. 119, no. 5, pp. 2710–2720, 2006.
This paper describes the development of a micro- machined microphone for aircraft fuselage arrays that are utilized by aeroacousticians to help identify aircraft noise sources and/or assess the effectiveness of noise-reduction technologies. The developed microphone utilizes piezoelectric transduction via an integrated aluminum nitride layer in a thin-film composite diaphragm. A theoretical lumped element model and an associated noise model of the complete microphone system are developed and utilized in a formal design-optimization process. Optimal designs were fabricated using a variant of the film bulk acoustic resonator process at Avago Technologies. The experimental characterization of one design is presented here, and measured performance was in line with sponsor specifications, including a sensitivity of -39 μV/Pa, a minimum detectable pressure of 40.4 dB, a confirmed bandwidth up to 20 kHz, a 129.5-kHz resonant frequency, and a 3% distortion limit approaching 172 dB. With this performance-in addition to its small size-this microphone is shown to be a viable enabling technology for low-cost, high-resolution fuselage array measurements.
Full-text available
Accurate duct acoustic propagation models are required to predict and reduce aircraft engine noise. These models ultimately rely on measurements of the acoustic impedance to characterize candidate engine nacelle liners. This research effort increases the frequency range of normal-incidence acoustic impedance testing in square ducts by extending the standard two-microphone method (TMM), which is limited to plane wave propagation, to include higher-order modes. The modal decomposition method (MDM) presented includes four normal modes in the model of the sound field, thus increasing the bandwidth from 6.7 to 13.5 kHz for a 25.4 mm square waveguide. The MDM characterizes the test specimen for normal- and oblique-incident acoustic impedance and mode scattering coefficients. The MDM is first formulated and then applied to the measurement of the reflection coefficient matrix for a ceramic tubular specimen. The experimental results are consistent with results from the TMM for the same specimen to within the 95% confidence intervals for the TMM. The MDM results show a series of resonances for the ceramic tubular material exhibiting a monotonic decrease in the resonant peaks of the acoustic resistance with increasing frequency, resembling a rigidly-terminated viscous tube, and also evidence of mode scattering is visible at the higher frequencies.
Full-text available
This paper describes the design, fabrication, and characterization of a bulk-micromachined piezoelectric microphone for aeroacoustic applications. Microphone design was accomplished through a combination of piezoelectric composite plate theory and lumped element modeling. The device consists of a 1.80-mm-diam, 3-microm-thick, silicon diaphragm with a 267-nm-thick ring of piezoelectric material placed near the boundary of the diaphragm to maximize sensitivity. The microphone was fabricated by combining a sol-gel lead zirconate-titanate deposition process on a silicon-on-insulator wafer with deep-reactive ion etching for the diaphragm release. Experimental characterization indicates a sensitivity of 1.66 microVPa, dynamic range greater than six orders of magnitude (35.7-169 dB, re 20 microPa), a capacitance of 10.8 nF, and a resonant frequency of 59.0 kHz.
Full-text available
This paper presents the development of a micro-machined dual-backplate capacitive microphone for aeroacoustic measurements. The device theory, fabrication, and characterization are discussed. The microphone is fabricated using the five-layer planarized-polysilicon SUMMiT V process at Sandia National Laboratories. The microphone consists of a 0.46-mm-diameter 2.25-mum-thick circular diaphragm and two circular backplates. The diaphragm is separated from each backplate by a 2-mum air gap. Experimental characterization of the microphone shows a sensitivity of 390 muV/Pa. The dynamic range of the microphone interfaced with a charge amplifier extends from the noise floor of 41 dB/ radicHz up to 164 dB and the resonant frequency is 178 kHz.
This paper presents the characterization of a piezoresistive silicon microphone. The microphone was designed for aeroacoustic measurements where small size, high dynamic range, large frequency bandwidth, and low power requirements are desired. It consists of four dielectrically-isolated, single-crystal silicon piezoresistors mounted on the top surface of a circular, tensile silicon nitride diaphragm. Multiple devices were characterized in terms of linearity, frequency response, drift, noise, and power. The sensors exhibit an average sensitivity of 0.6 μV/PȧV, a flat frequency response, minimal drift, and consume 15 mW of power when operated at 3 V. They show a linear response up to 750 dB SPL and a 52 dB SPL noise floor (1 Hz bin centered at 1 kHz).
In this paper the development is described of a new high‐frequency capacitive microphone based on the Knowles SiSonic MEMSmicrophonetechnology. The intended application for these microphones is aero‐acoustic testing, which requires a transducer bandwidth of 80–100 kHz and the ability to measuresound pressures to 150 dB SPL without excessive distortion. The MEMSmicrophone structure was modified to meet these requirements and still be manufactured using the Knowles SiSonic production process. In addition, the microphone package was modified to eliminate the front port resonance, at approximately 12 kHz, associated with the standard SiSonic device. Specifically, the microphone package was modified with a lid that has lower acoustic inertance, thereby increasing the resonance frequency of the Helmholtz resonator formed by the front port and the microphone interior cavity. Details of the design process and the theoretical behavior of the microphones are presented, and it is shown that a microphone sensitivity of 0.5–1 mV/Pa can be achieved for a 100‐kHz bandwidth. Preliminary results from fabricated microphones indicate a sensitivity around 0.5 mV/Pa and an A‐weighted noise level of approximately 61.5‐dB SPL. A more detailed analysis of the microphone noise spectrum and frequency response is presented.
This paper provides evidence for electrical 1/f noise as the dominant source of excess noise in piezoresistive microelectromechanical systems (MEMS) microphones. In piezoresistors, the fundamental noise sources may be divided into frequency independent thermal noise and frequency dependent 1/f excess noise dominating at low frequencies. Noise power spectra are presented for both commercial and research-prototype MEMS piezoresistive microphones as a function of applied voltage bias for both free and blocked membranes. The contributions of various mechanical and electrical noise sources are compared using a lumped noise equivalent circuit of the piezoresistive microphone. The bias dependence and membrane independence of the output noise indicate that the primary source of the excess noise is electrical in origin. 0 2006 Acoustical Society of America.
Conference Paper
A multi-level polysilicon surface-micromachining technology consisting of 5 layers of polysilicon is presented. Surface topography and film mechanical stress are the major impediments encountered in the development of a multilayer surface-micromachining process. However, excellent mechanical film characteristics have been obtained through the use of chemical-mechanical polishing for planarization of topography and by proper sequencing of film deposition with thermal anneals. Examples of operating microactuators, geared power-transfer mechanisms, and optical elements demonstrate the mechanical advantages of construction with 5 polysilicon layers
Silicon micromachining has successfully been applied to fabricate piezoelectric, piezoresistive and capactive microphones. The use of silicon has allowed the fabrication of microphones with integrated electronic circuitry and the development of the new FET microphone. The introduction of lithographic techniques has resulted in microphones with very small (1 mm2) diaphragms and with specially shaped backplates. The application of corrugated diaphragms seems a promising future development for silicon microphones. It is concluded from a noise consideration that the FET microphone shows a high noise level, which is mainly due to the small sensor capacitance. From this noise consideration, it can be shown that integration of a capacitive microphone and a preamplifier will result in a further reduction of the noise.