1

Nonlinear Effects in MEMS Capacitive Microphone Design

Sazzadur Chowdhury, M. Ahmadi, W. C. Miller
Electrical and Computer Engineering
University of Windsor
Windsor, Ontario, N9B 3P4
Canada
Abstract
An analytical method is presented that provides a
better approximation for design parameter values of a
MEMS capacitive microphone. The spring softening
effect associated with the nonlinear characteristics of the
electrostatic pressure due to a bias voltage, and the spring
hardening effect associated with nonlinear stretching of
the central region of a uniformly loaded fully clamped
diaphragm with residual stress are both considered. The
method allows a more accurate determination of the
developed electrostatic pressure, maximum diaphragm
deflection and the pull-in voltage. The resulting
electrostatic pressure, pull-in voltage and deflection
profile of the diaphragm are in close agreement with
finite element analysis results.
Key Words: MEMS, capacitive microphone, pull-in
voltage, nonlinearity, diaphragm deformation,
electrostatic force.
1. Introduction
MEMS-based capacitive microphones offer advantages
due to their small size, relatively high sensitivity, batch
fabrication capability, inherently low power consumption
and low noise features. Serious research efforts have been
made to improve the design methodology and
performance of MEMS-based capacitive microphones
[1]-[3]. A MEMS capacitive-type microphone is basically
an electrostatic transducer that depends on electrical
energy in terms of a constant voltage (voltage drive) or
constant charge storage (current drive) to facilitate
monitoring of capacitance change due to an external
mechanical input [4]. The constant voltage driven mode is
relatively simpler to implement in MEMS technology and
is popular in MEMS industry [5]. However, electrostatic
force associated with the bias voltage in the constant
voltage mode is nonlinear due to its inverse square
relationship with the airgap thickness between the
capacitor electrodes. This gives rise to a phenomenon
known as pull-in that reduces the dynamic range of the
diaphragm displacement to one-third of the airgap. If the
bias voltage exceeds this pull-in limit, the diaphragm will
collapse. In addition, the initial deformation of the
diaphragm due to the electrostatic force causes an offset
error in the readout circuit [6]. Applications, such as
hearing aid instruments using this kind of microphone
need additional electrical circuits to minimize this bias
voltage dependent offset error. Furthermore, this initial
diaphragm deformation reduces the airgap thickness, that
in turn reduces the dynamic range of the microphone.
Accurate determination of the pull-in, or the collapse
voltage is critical in the design process. Unfortunately,
the commonly used parallel-plate approximation method
of pull-in voltage determination introduces significant
errors if the diaphragm is fully clamped. In addition, for a
clamped diaphragm, the small deflection model of
diaphragm deformation [7] does not account for
nonlinearities associated with the presence of in-built
residual stress in the diaphragm and predicts
unrealistically high deformation values. Thus, it is
necessary to formulate a method to determine the pull-in
voltage that accounts for the nonlinear nature of the
design parameters. In [8], an empirical method is
provided to approximate pull-in voltage for cantilevers,
fixed-fixed beams and circular diaphragms under
electrostatic actuation. The method can evaluate pull-in
voltage for the mentioned structures within 1% agreement
with finite element analysis results under certain
limitations.
In this paper an analytical solution has been described
to calculate the pull-in voltage and diaphragm deflection
for a clamped square diaphragm under electrostatic
actuation. The method incorporates both the
nonlinearities of the electrostatic force and the large
deflection model for a clamped square diaphragm
deflection. The method can easily be extended to the
cases of cantilevers, fixed-fixed-beams or circular
diaphragms.
2. Design Considerations
The basic structure of a MEMS capacitive microphone
is shown in figure 1. The structure can be viewed as a
parallel plate capacitor consisting of a top diaphragm and
a bottom backplate separated by a small airgap acting as
Proceedings of the International Conference on MEMS, NANO and Smart Systems (ICMENS03)
0-7695-1947-4/03 $17.00 2003 IEEE
2
the dielectric material. When an acoustical sound wave is
incident on the diaphragm, it causes the diaphragm to
deflect and the gap between the diaphragm and the
backplate decreases causing an increase in the
capacitance between them. As the diaphragm vibrates in
accordance with the frequency of the acoustical wave, the
capacitance between the electrodes changing accordingly
due to a variable airgap. If a battery is connected across
the diaphragm and the backplate, following the principle
of energy conservation, electrical charge will flow to and
away from the battery in accordance with diaphragm
vibration. By connecting a suitable charge flow (current)
sensing electrical circuit to the system, a usable voltage
signal representation of the incident acoustical wave can
be obtained.
However, while the supply voltage providing a means
for readout of the change in capacitance due to diaphragm
deflection, the resulting electrostatic attraction force pulls
the capacitor electrodes towards each other and thus
causes the diaphragm to deflect, even in the absence of an
external mechanical pressure. This electrostatic attraction
force is nonlinear and increases with the decreasing gap
width between the electrodes for a fixed voltage.
Furthermore, since the diaphragm is clamped, the
deflection at the central region of the diaphragm is higher
than that at the diaphragm edges. Thus, a non-uniform
force profile is developed. Also, the nonlinearity of the
electrostatic attraction force causes a downshift of the
fundamental resonant frequency of the structure, an effect
known as spring softening [9].
During operation, the capacitor structure is subjected
to four different types of forces: the mechanical input
(acoustical signal), the elastic force generated in the
vibrating diaphragm in response to the deformation, the
damping force generated by the airgap [10], and the
electrostatic force due to the supply voltage as mentioned
earlier. In equilibrium, the damping force can be
neglected. Thus, the perturbing force acting on the
diaphragm is the sum of the electrostatic and the external
mechanical pressure and is counterbalanced by the elastic
force developed in the diaphragm due to its deformation.
Therefore, it is desirable to minimize the effect of the
Figure 1. A conceptual cross-section of a MEMS
capacitive-type microphone
electrostatic force as much as possible. Careful
determination of an optimum operating point for the
supply (bias) voltage is necessary so that; (1) the device
does not collapse due to pull-in; (2) sufficient charge flow
can be established for the chosen readout circuitry; (3) the
spring softening effect does not lower the fundamental
resonant frequency of the structure below the desired
upper cut-off frequency, and; (3) the maximum allowable
deformation due to combined electrostatic and
mechanical forces is within the elastic limit of the
diaphragm material.
As a microphones sensitivity, frequency response and
noise properties, etc. depend critically on the bias voltage,
it is necessary to optimize the bias voltage, considering
the available geometry and fabrication materials, to
ensure safe operation beyond the pull-in limit.
3. Parallel-Plate Approximation
A parallel plate approximation is first considered to
highlight the major aspects of the device's functional
analysis. In this analysis, any fringe field capacitance
associated with the capacitor electrodes is neglected and
the capacitor electrodes and contacts are assumed to be
perfect. It is assumed that the capacitor structure is
situated in a vacuum environment to ensure zero external
mechanical loading of the top electrode. It is also
assumed that the diaphragms restoring force (spring
force) is a linear function of its displacement
A lumped element model of a movable plate capacitor
is shown in figure 2a and an equivalent circuit in shown
in figure 2b. Neglecting the damping effect (as the
structure is assumed to be in vacuum), the equation of
motion of the movable plate due to the electrostatic
attraction force F caused by a constant supply voltage V
can be expressed as:
F kx
dt
x d
m = +
2
2
(1)
where x represents the displacement. The mechanical
elastic force
M
F can be expressed as
kx F
M
= (2)
where the variation of k is assumed to be linear. The
electrostatic attraction force F
E
between the plates due to
the charges on the plates can be found by differentiating
Proceedings of the International Conference on MEMS, NANO and Smart Systems (ICMENS03)
0-7695-1947-4/03 $17.00 2003 IEEE
3
Figure 2. Lumped parameter model of a parallel plate
transducer and its equivalent circuit
the stored energy of the capacitor with respect to the
position of the movable plate and is expressed as:
2
0
2
0 2
) ( 2 2
1
x d
AV
CV
dx
d
F
E

= |
.
|

\
|
=

(3)
where
0
is the permittivity of free space, C is the
capacitance, A is the area of a capacitor plate and
0
d is
the thickness of the airgap. Since at
equilibrium,
E M
F F = , we have:
2
0
2
0
2
2
) ( 2

x d
AV
kx
dt
x d
m

= +

(4)
This equation can be rearranged into a third-order
polynomial. After solving for x and choosing the stable
root, the pull-in voltage can be expressed as [4]:
A
kd
V
PI
0
3
0
27
8

= (5)
The distance where the pull-in occurs is:
3
0
d
x
PI
= (6)
and the pull-in gap is:
3
2
0
d
d
PI
= (7)
The spring constant is given by:
3
0
2
0
8
27
d
AV
k
PI

= (8)
If the voltage is increased beyond this pull-in voltage,
the resulting electrostatic force will cause the movable
plate to collapse onto the fixed plate and the capacitor
will be short-circuited.
By expanding (3) using a Taylor series approximation
about a minimal distance
0
x we get:
..... ) (
) ( 2
) 1 )( 2 (
) ( 2 ) ( 2
0
3
0
2
0
2
0
2
0
2
0
2
0
0
0
+

=
=
=
x x
x d
AV
x d
AV
x d
AV
F
x x
x x E

(9)
After rearranging the terms:
(

= ... 2 1
) ( 2
0 0
0
2
0 0
2
0
x d
x x
x d
AV
F
E

(10)
Substituting
E
F from (10) in (4), we get:
(

= + ... 2 1
) ( 2
0 0
0
2
0 0
2
0
2
2
x d
x x
x d
AV
kx
dt
x d
m

(11)
After rearrangement:
(

=
|
|
.
|

\
|

+
... 2 1
) ( 2
) (
0 0
0
2
0 0
2
0
3
0 0
2
0
2
2
x d
x
x d
AV
x
x d
AV
k
dt
x d
m

(12)
Thus, the electrostatic attraction force effectively
modifies the spring constant of the movable plate and the
term within parenthesis in the left hand side of (12)
represents the effective spring constant at a specified
voltage. The amount of modification is termed as spring
softening and can be expressed as:
3
0 0
2
0
) ( x d
AV
k
soft

=

(13)
Consequently, the resonant frequency of the structure
is shifted from:
m
k
res
= (14)
to
( )
soft res
k k
m
=
1
(15)
For a microphone, the resonant frequency of the
microphone structure must be well above the upper cut-
off frequency of the desired audio frequency range to
avoid audio distortion. Thus the chosen bias voltage must
not soften the spring constant of the diaphragm so much
that the altered fundamental resonant frequency of the
structure falls below the upper cut-off frequency of the
desired audio frequency range. The chosen bias voltage
must also ensure that the deflection of the diaphragm due
to the combined effects of the electrostatic and the
acoustical pressure is within the non-pull-in limit of the
structure.
4. Fully Clamped Square Diaphragm
Analysis
Due to the presence of residual stress and a
significantly large deflection of the diaphragm compared
to its thickness, the developed strain energy in the middle
of the diaphragm causes a stretch of the middle surface.
Consequently, the deflection of the middle surface no
longer depends solely on the external forces and the
rigidity of the diaphragm increases with the deflection.
[7]. This deflection dependent nonlinear behavior of a
fully clamped diaphragm is known as spring hardening
and a large deflection model of analysis must be applied
Proceedings of the International Conference on MEMS, NANO and Smart Systems (ICMENS03)
0-7695-1947-4/03 $17.00 2003 IEEE
4
to determine the deflection. Thus, the analytical solution
for diaphragm deflection from electrostatic forces must
account for this spring hardening effect in addition to the
nonlinear and non-uniform electrostatic forces.
Following the large deflection model, for a fully
clamped square diaphragm with built in residual stress,
the deflection of the midpoint of the diaphragm under a
uniform pressure load P can be expressed as [11]:
3
0
4
2 0
2
1 0
) ( ) ( h
a
tE
v C h
a
t
C h P + =

(16)
where P is the applied uniform pressure, t, the diaphragm
thickness, E, the Youngs modulus, v, the in-plane
Poissons ratio, , the residual stress, a, the half of the
diaphragm side length and
0
h , the deflection of the
diaphragm midpoint. The quantities, C
1
and C
2
are
numerical parameters and their values are given by:
) 1 /( ) 271 . 0 1 ( 994 . 1
and , 45 . 3
2
1
v v C
C
=
=
(17)
The deflection of the diaphragm from mid-side to mid-
side can be calculated from midpoint deflection as:
|
.
|

\
|
|
|
.
|

\
|
+ =
a
x
a
x
h x h
2
cos 401 . 0 1 ) 0 , (
2
2
0

(18)
and diagonal deflection can be calculated as:
|
.
|

\
|

|
|
.
|

\
|
+ + = =
a
x
a
x
a
x
h y x h
2
cos
1611 . 1
2
401 . 0 1 ) (
2
4
4
2
2
0

(19)
In [8], it was shown that for clamped diaphragms, a
fringe field correction is required for the Youngs
modulus as given by:
) 1 /(

2
v E E = (20)
where E

is the effective Youngs modulus.

5. Pull-in Voltage Evaluation
The analysis carried out in Section 3 for a parallel
plate capacitor structure can now be extended to the case
of a fully clamped square diaphragm separated from a
rigid backplate by a small airgap (figure 1). The
deflection of the diaphragm is due to the resultant effect
of the electrostatic, external acoustical, restoring elastic
and air damping forces acting simultaneously.
Figure 3. Linearization of the electrostatic force about
zero displacement
At equilibrium, the damping force can be neglected.
For a parallel plate configuration, the nonlinear
electrostatic force is always uniform. However, for a
rigidly clamped diaphragm under electrostatic force, the
later becomes non-uniform depending on the deformation
profile of the rigidly clamped diaphragm [7] and results in
a lower deflection. Thus, to evaluate the deflection of a
clamped diaphragm by an electrostatic force, it is
necessary to obtain a uniform, linear model of the
electrostatic force to apply in the load deflection equation
(16).
A uniform, linearized model of the electrostatic force
can be obtained from (12) by linearizing the force about
the zero deflection point, i.e. x
0
=0 as shown in figure 3.
Since, prior to any deflection the diaphragm is flat
(x
0
=0), the parallel plate approximation can easily be
applied without any significant error if the airgap
thickness is very small compared to the lateral dimensions
of the diaphragm. Thus, linearizing (12) about the point,
x
0
=0, we get:
2
0
2
0
3
0
2
0
2
2
2d
AV
x
d
AV
k
dt
x d
m

=
|
|
.
|

\
|
+ (21)
Rearranging and neglecting the time dependent term:
kx x
d
AV
d
AV
= +
3
0
2
0
2
0
2
0
2

(22)
The left hand side of (22) describes an approximate,
uniform, linear electrostatic force that can be used to
evaluate the pull-in voltage.
From (22), the effective linearized uniform
electrostatic pressure on the diaphragm due to an applied
bias voltage can be evaluated as:
|
|
.
|

\
|
+ =
3
0
2
0
2
0
2
1
d
x
d
V P
eff
(23)
Substituting x in (23) by the pull-in deflection from (6),
the effective pull-in pressure
PI eff
P

can be evaluated as:

2
0
2
0
6
5
d
V
P
PI
PI eff

(24)
Next, substituting the pull-in deflection
0
3
1
d for
0
h ,
replacing ) (
0
h P by
PI eff
P

, and using the effective

Youngs modulus E

in (16), we obtain:
Proceedings of the International Conference on MEMS, NANO and Smart Systems (ICMENS03)
0-7695-1947-4/03 $17.00 2003 IEEE
5
3
0
4
2
0
2
1
2
0
2
0
3

) (
3 6
5
|
.
|

\
|
+ |
.
|

\
|
=
d
a
E t
v C
d
a
t
C
d
V
PI

(25)
The above equation can now be solved for the pull-in
voltage
PI
V as:
(
(

|
.
|

\
|
+ |
.
|

\
|
=
3
0
4
2
0
2
1
0
2
0
3

) (
3 5
6 d
a
E t
v C
d
a
t
C
d
V
PI

(26)
Equation (26) provides the desired approximation of the
pull-in voltage for a clamped square diaphragm.
6. Illustrative Example
To illustrate the above model of pull-in voltage
evaluation, a microphone structure having device
parameters as given in Table I has been used.
The left hand side of (17) is plotted in figure 4 for a
voltage range of 11 to 17.45 volts. Superimposed in the
figure is a plot of right hand side of (17) using 226.4 N/m
for the spring constant k . At a distance of one-third of
the airgap (1.167 m), the kx graph intersects the 17.45
volts constant voltage curve and thus gives the voltage
where the pull-in occurs.
The intersection of the kx graph with the 12 volts
constant voltage curve shows the stable equilibrium point
at that voltage. The diaphragm deflection at that point is
about 0.41 m. The pull-in voltage calculated using (26)
is 17.453 volts and this is in close agreement with Figure
4. The value of pull-in voltage calculated using equation
(26) is about 2.45 volts higher than that is calculated
using (5).
The midpoint to midpoint deflection profiles of the
diaphragm for different bias voltages is calculated using
(16) to (20) and (23) and are plotted in Figure 5.
Results from 3-D finite element analysis and the
developed analytical model for the diaphragm deflection
for different bias voltages are plotted in Figure 6. The
FEA results show that the pull-in occurs at 17.85 volts
and this is in close agreement with the analytical value.
Figure 7 shows the finite element analysis result of the
diaphragm collapse. The derivations carried out for a
fully clamped square diaphragm, can be extended to any
rectangular or circular diaphragm case by applying
appropriate load -deflection criteria and boundary
conditions.
Table 1. Device Parameters
Parameter Value
Diaphragm thickness, t 0.8 m
Diaphragm side length, a 1.2 mm
Airgap thickness, d
0 3.5 m
Youngs modulus, E 169 GPa
Poissons ratio, v 0.28
Residual stress - 20 MPa.
Figure 4. Electrostatic and elastic forces versus
displacement using parameters given in Table 1
Figure 5. The midpoint to midpoint deflection profiles
of the diaphragm for different bias voltages
Proceedings of the International Conference on MEMS, NANO and Smart Systems (ICMENS03)
0-7695-1947-4/03 $17.00 2003 IEEE
6
Figure 6. Comparison of diaphragm deflection
calculated using the analytical model and FEA
analysis for different bias voltages.
Figure 7. Finite element analysis results showing
diaphragm collapse due to pull-in phenomenon
7. Conclusions
An analytical method is presented to determine the
pull-in voltage of a MEMS capacitive microphone having
a clamped square diaphragm. The method incorporates
the nonlinear and non-uniform nature of the electrostatic
force associated with a clamped diaphragm deformation.
The spring hardening effect associated with deflection of
a clamped diaphragm with built-in residual stress has also
been addressed. The resulting analytical model provides a
more accurate approximation of the pull-in voltage. The
results from the analytical model are in close agreement
with finite element analysis results. The method can be
extended to determine the pull-in voltage for other
microstructures utilizing electrostatic actuation if the
airgap thickness is very small compared to lateral
dimensions.
8. Acknowledgements
This research has been made possible by the interest
and support provided by the Gennum Corporation of
Burlington, Ontario. The authors greatly acknowledge
the additional generous support of the following partners:
Canadian Microelectronics Corporation (CMC),
MICRONET, Natural Sciences and Engineering Research
Council of Canada (NSERC) and the IntelliSense
Corporation, Wilmington, MA, USA.
9. References
[1] C. Gibbons and R. N. Miles, Design of A Biomimetic
Directional Microphone Diaphragm, Proc. of IMECE Intern.l
Mechanical Engineering Congress and Exposition, November
5-10, 2000. Orlando, Florida, 2000, pp. 1-7.
[2] P. C. Hsu, C. H. Mastrangelo and K. D. Wise, A High
Density Polysilicon Diaphragm Condenser Microphone, in
Tech. Digest of. IEEE 11th International Conference on Micro
Electro Mechanical Systems (MEMS), Heidelberg, Germany,
1998, pp. 580-585.
[3] M. Pederson, W. Olthuis and P. Bergveld, High-
Performance Condenser Microphone with Fully Integrated
CMOS Amplifier and DC-DC Voltage Converter, Journal of
Microelectromechanical Systems, Vol. 7, No. 4, pp. 387-394,
Dec. 1998.
[4] S. D. Senturia, Microsystems Design, Kluwer
Academic Publishers, 2000.
[5] J. Pons-Nin, A. Rodriguez and L. M. Castaner, Voltage
and Pull-in in Current Drive of electrostatic Actuators", Journal
of Microelectromechanical Systems, Vol. 11, No. 3, pp. 196-
205, Jun. 2002.
[6] R. Puers and D. Lapadatu, Electrostatic Forces and Their
Effects on Capacitive Mechanical Sensors, Sensors and
Actuators A, Vol. 56, 1996, pp. 203-210.
[7] S. Timoshenko and S. Woinowsky-Krieger, Theory of
Plates and Shells, McGraw-Hill Book Company New York,
1959, pp. 397-428.
[8] P. M. Osterberg and S. D. Senturia, M-TEST: A Test
Chip for MEMS Material Property Measurement Using
Electrostatically Actuated Test Structures, Journal of
Microelectromechanical Systems, Vol. 6, No. 2, pp. 107-118,
June. 1997.
[9] Y. He, J. Machete, C. Gallegos and F. Maser, Accurate
Fully-Coupled Natural Frequency Shift of MEMS Actuators
Due to Voltage Bias and Other External Forces, in Tech.
Digest of. IEEE 12th International Conference on Micro Electro
Mechanical Systems (MEMS), Orlando, Florida, 1999, pp.321-
325.
[10] J. Bergqvist and J. Gobet, Capacitive Microphone with a
Surface Micromachined Backplate Using Electroplating
Technology, Journal of Microelectromechanical Systems, Vol.
3, No. 2, pp. 387-394, Jun. 1994.
[11] D. Maier-Schneider, J. Maibach and E. Obermeier, A
New Analytical Solution for the Load-Deflection of Square
Membranes", Journal of Microelectromechanical Systems, Vol.
4, No. 4, pp. 238-241, Dec. 1995.
Proceedings of the International Conference on MEMS, NANO and Smart Systems (ICMENS03)
0-7695-1947-4/03 $17.00 2003 IEEE