Procedia

Chemistry
Procedia Chemistry 1 (2009) 875–878

www.elsevier.com/locate/procedia

Proceedings of the Eurosensors XXIII conference

Design of a high sensitivity FET integrated MEMS microphone

Malhi Charanjeet kaura*, Rudra Pratapa, Navakanta Bhatb
a
CranesSci MEMS Lab, Mechanical Engg. Dept., Indian Institute of Science, Bangalore, Karnataka, India
b
VLSI devices Lab, Electrical and Communication Engg. Dept., Indian Institute of Science, Bangalore, Karnataka, India

Abstract

FET based MEMS microphones comprise of a flexible diaphragm that works as the moving gate of the transistor. The integrated
electromechanical transducer can be made more sensitive to external sound pressure either by increasing the mechanical or the
electrical sensitivities. We propose a method of increasing the overall sensitivity of the microphone by increasing its electrical
sensitivity. The proposed microphone uses the transistor biased in the sub-threshold region where the drain current depends
exponentially on the difference between the gate-to-source voltage and the threshold voltage. The device is made more sensitive
without adding any complexity in the mechanical design of the diaphragm.

Keywords: FET (Field Effect Transistor); Microphone; subthreshold; suspended gate; sensitivity

1. Introduction

High sensitivity is one of the most desirable performance characteristics of a microphone. The overall sensitivity
of an electromechanical transducer, such as a microphone is a product of two sensitivities, viz., mechanical and
electrical sensitivity. The state-of-the-art [1-3] comprises of methods to increase mechanical sensitivity by making
the diaphragm more compliant by providing slots or corrugations on its periphery. The mechanical sensitivity can
also be increased if the air gap between the diaphragm and back-plate is reduced but the reduced air gap leads to
higher damping due to movement of air in and out of the gap when the membrane vibrates which leads to reduced
sensitivity [3]. On the other hand, the increase in electrical sensitivity of capacitive type microphones is limited by
the bias voltage applied between the two electrodes [4].

In this work, we propose suspended gate (SG) architecture for a MEMS microphone where the diaphragm
behaves as the moving gate of the transistor. SG-MOSFET is already being used for various applications like
MEMS resonant gate transistor [5], RF MEMS switch [6-8], pressure sensor and microphone [3,9]. In the published
literature on SG-MOSFET based sensing, the transistor has been biased either in above threshold region or switched
from weak to strong inversion regions of operation. The novelty in our proposed microphone lies in the fact that we
bias the transistor in the sub-threshold region [10]. We exploit the exponential variation of the drain current with the
difference between the gate-to-source voltage V , and the threshold voltage V ,  ie., V ‐  V to make it

*
Corresponding author. Tel.: +91-(0)80-2293-3224; fax: +91-(0)80-2350-8659.
E-mail address: cjeet@mecheng.iisc.ernet.in.

1876-6196/09 © 2009 Published by Elsevier B.V. Open access under CC BY-NC-ND license.
doi:10.1016/j.proche.2009.07.218
876 M. Charanjeet kaur et al. / Procedia Chemistry 1 (2009) 875–878

extremely sensitive to very small deflections of the diaphragm. The proposed concept increases the electrical
sensitivity and hence the overall sensitivity of the microphone without making any design changes in the diaphragm.
This circumvents the complexity in fabrication if the corrugated diaphragms are to be fabricated to increase
mechanical sensitivity. Further, the advantage of FET based active sensing lies in reduced parasitic and simplified
signal conditioning as compared to the capacitive type passive sensing [9].

2. Working principle and construction of the SGFET microphone

The structure is similar to that described in [6,9] with the difference that an Silicon-On-Insulator (SOI) substrate
is used instead of a bulk one. SOI substrate is used to exploit the steep sub-threshold slope characteristics [11].
Figures 1(a) and 1(b) show the 3D structure and the equivalent capacitor circuit of the n-channel SGFET MEMS
microphone.
The structure comprises of an elastic diaphragm which deflects due to the incident sound pressure, thereby
changing and the air gap between the moving gate and the fixed substrate. The change in the gap changes the drain
current which forms the sensor response. For a constant V V , when the gap changes, the V changes and this
leads to a change in drain current. The back gate voltage is assumed zero and the source is assumed to be grounded
in the analysis.
The sub-threshold drain current for a moving gate MOSFET built on an SOI substrate can be written as [12]:

⁄
 μ 1                                                                                                              (1)         

where, is the mobility, is the width and  is the length of the channel, is the effective depletion capacitance
formed by and , is the effective gate
capacitance formed by and ,  is the
thermal voltage, is the drain-to-source voltage
and 1 .

 is proportional to as seen from

eqn. (1). Therefore, for a constant V V , keeping
all other factors constant, change in air gap leads to a
change in V and and that leads to an exponential
change in drain current. This feature also provides
for an inbuilt amplification in the sensor response.
The proposed method makes the microphone
sensitive to very small deflections, (of the order of
sub-nanometers), of the diaphragm. The sensor is always operated in fully depleted condition by ensuring that the
silicon film thickness is lesser than the maximum depletion width.

Fig.1(a) 3D structure of the proposed n-channel SGFET MEMS microphone; (b) The equivalent capacitor circuit

3. Simulation methodology and electro-mechanical behavior of the SGFET microphone

Lumped parameter modeling is an effective tool to model the dynamic performance of a multi-domain device like
a microphone [13]. The mass and compliance of the vibrating element are represented as an inductor and
capacitor respectively. The acoustic impedance of the air in contact with the vibrating diaphragm is represented
by a radiative resistance  and mass  . The compliance of the air in the back-chamber is modeled as [14].
Figure 2 shows an equivalent circuit model of the mechanical domain of the proposed microphone. The mechanical
impedance can be written as [9]:

        (2) 
 M. Charanjeet kaur et al. / Procedia Chemistry 1 (2009) 875–878 877

The mechanical sensitivity (S ) is defined as the change in the gap with the change in the external stimulus, ie.,
pressure [15] and can be written as:

S (3)

where,   , d is the original air gap, is the deflection of the diaphragm and is the sound pressure. The
deflection of a forced spring mass damper system can be represented as:

(4)

where, is the area of the diaphragm. The electrical sensitivity ( ) can be defined as the ratio of the change in the
drain current to the change in the air gap and can be written as:
⁄
S (5)

The overall open circuit sensitivity can be defined as:

⁄
S S  S (6)

Equation (6) shows that the overall normalized sensitivity can be increased
by increasing the active area of the diaphragm, decreasing the mechanical
impedance of the microphone system and/or by increasing the electrical
sensitivity. The electrical sensitivity can be tuned by various factors like the
silicon film thickness, front and back gate oxide thicknesses, air gap and the
doping concentration.
Fig.2 Equivalent circuit for the mechanical domain

4. Simulation Results

We now discuss the gain in sensitivity obtained because of biasing the microphone in the sub-threshold instead of
the above threshold region. When the diaphragm deflects due to sound pressure, the gap reduces and consequently
V   hanges. Hence for biasing the microphone at a fixed bias voltage  , the value of  is calculated at the
maximum deflection of the gate and the bias voltage is fixed at 40 mV less than the threshold voltage.

Fig.3(a) Percent change in drain current with percent change in displacement for sub-threshold and below threshold region. The relevant
dimensions are: w/l =1, air gap= 98 nm, Tsi =10 nm, Tox= 10 nm, Tbox= 1µm, Na= 5x1023 /m3, membrane size = 500µm, membrane thickness
=1µm; Fig. 3(b) Simulated frequency response of the SGFET microphone; Fig. 3(c) Simulated frequency response of the SGFET microphone in
absolute units (Pa-1)

Figure 3(a) shows the percentage change in the drain current for a device biased above and below the threshold
regions respectively. The percentage change is calculated by taking the original gap and corresponding drain
currents (above and below threshold) as the reference values. It can be seen that for a 2% change in the air gap,
there is a 560 % and 8614 % percent change in the drain current for above and below threshold biased devices.
878 M. Charanjeet kaur et al. / Procedia Chemistry 1 (2009) 875–878

Figure 3(b) shows the frequency response of the SGFET microphone simulated using eqn. (6). It can be noted from
figure 3(c) that the normalized sensitivity is higher for sub-threshold biased microphone as compared to the
microphone biased in the saturation region. For eg., the normalized sensitivity for below threshold region
microphone is 1.7 while that for the above threshold region microphone is 0.009  . Thus, the change
in the drain current, with reference to the current at the original gap, for the same change in the pressure, is higher
for the proposed microphone. It can be seen that the designed microphone shows a flat frequency response in the
audible range of frequencies (ie. 20 Hz to 20 KHz). The resonance is pushed far beyond the audible frequency range
and is in the range of a few hundreds of KHz. The designed microphone can be operated in the stiffness dominated
region where the stiffness of the diaphragm dominates the stiffness of the air in the cavity.

The simulations were carried for air gaps varying from 100 to 195 nm, silicon thickness varying from 10 to 20 nm,
doping concentration varying from 1x1023 /m3 to 5x1023 /m3, oxide thickness varying from 1 to 10 nm and buried
oxide thickness varying from 1 to 2 µm. These simulations lead to the following observations.
• Higher silicon thickness, larger air gap and higher doping concentration lead to higher threshold voltage.
• Doping concentration, silicon thickness and initial air gap are the most significant design parameters
affecting the normalized sensitivity as they affect the threshold voltage the most.
• The normalized sensitivity can be increased as the bias voltage (   is moved more closer to the threshold
voltage.

5. Conclusions

A sub-threshold biased SGFET based MEMS microphone is designed and its performance is simulated to determine
its sensitivity and frequency response. The proposed design shows an order of magnitude higher change in drain
current when compared to that obtained by biasing of microphone by existing methods. The electrical sensitivity can
be tuned by various factors like doping concentration, silicon thickness, and initial air gap.

References
1. Scheeper, P.R., Olthuis W., Bergveld P.,”The Design, Fabrication, and Testing of corrugated Silicon Nitride Diaphragms”, J.M
icroelectromechanical Systems, 3, No. 1, 1994, 36-42.
2. Wang, W.J., Lin, R.M., Li, X., Guo, D.G., “Study of single deeply corrugated diaphragms for high-sensitivity microphones", J. Micromech.
Microengg. , 13, 2003, 184-9.
3. Graf, E., Kronast, W., Duhring, S, Muller, B., Stoffel, A., “Silicon membrane condenser microphone with integrated field-effect
transistor”, Sensors and actuators A, 37-8, 1993, 708-11.
4. Li, X., Lin, R., Kek, H., Miao, J., Zou, Q., “Sensitivity improved silicon condenser microphone with a novel single deeply corrugated
diaphragm”, Sensors and actuators A, 92, 2001, 257-62.
5. Newell, W.E., “Novel circuit aspects of the resonant gate transistor”, International Solid State Circuit Conference, Feb 1996.
6. Nicolas Abele, ”Design & fabrication of suspended gate MOSFETS for MEMS resonator, switch & Memory applications”, Ph.D. Thesis,
EPFL, 2007.
7. Ionescu, A.M., Pott, V., Fritschi, R., Banerjee, K., Declercq, M.J., Renaud, R., et.al., “Modeling and Design of a Low-Voltage SOI
suspended-Gate MOSFET(SG-MOSFET) with a Metal-Over-Gate Architecture”, International Symposium on Quality Electronic Design, 2002.
8. Akarvardar, K., Eggimann, C., Tsamados, D., Chauhan, Y.R., Wan, G.C., Ionescu, A.M., “Analytical Modeling of the Suspended-Gate
FET and Design Insights for Low-Power Logic”, IEEE transactions on Electron Devices, 55, No. 1, Jan 2008.
9. Kuhnel, W., “Silicon Condenser Microphone with Integrated Field-effect Transistor”, Sensors and Actuators A, 25-27, 1991, 521-25.
10. Bhat, N., Pratap, R., Malhi, C., “A sub-threshold elastic deflection FET sensor for sensing pressure/force, a method and system thereof”,
International Appl. No. PCT/IN 2008/000387, Patent pending.
11. Colinge, J.P., Silicon-On-Insulator technology: Materials to VLSI.Boston/Dordrecht/London: Kluwer academic publishers, 1991.
12. Grotjohn, T., Hoefflinger, B., “A parametric short-channel MOS Transistor model for subthreshold and strong inversion current”, IEEE
transactions on electron devices, 31, No. 2, 1984.
13. S. S. Mohite, V. R. Sonti and R. Pratap, "A Comparative Study of the Equivalent Circuits for MEMS Capacitive Microphones and a
Critical Evaluation of the Equivalent Parameters", Advances in Vibration Engineering, 4, No. 4, 337-349, 2005.
14. Hsu, P.C., Mastrangelo, C.H., Wise, K.D., ”A high sensitivity polysilicon diaphragm condenser microphone”, Conference Record IEEE
11th International Workshop on MEMS, 580-85, 1998.
15. Scheeper, P.R., Van der Donk A.G.H., Olthuis, W., Bergveld, P., “A review of silicon microphones”, Sensors and actuators A, 44, 1994,
1-11.