Micromachines 11 00862

micromachines
Article
Microfabrication Process-Driven Design, FEM
Analysis and System Modeling of 3-DoF Drive Mode
and 2-DoF Sense Mode Thermally Stable
Non-Resonant MEMS Gyroscope
Syed Ali Raza Bukhari 1 , Muhammad Mubasher Saleem 1,2, *, Umar Shahbaz Khan 1,2 ,
Amir Hamza 1,2 , Javaid Iqbal 1,2 and Rana Iqtidar Shakoor 2,3
1 Department of Mechatronics Engineering, National University of Sciences and Technology,
Islamabad 44000, Pakistan; saraza.mts19ceme@mts.ceme.edu.pk (S.A.R.B.);
u.shahbaz@ceme.nust.edu.pk (U.S.K.); a.hamza@ceme.nust.edu.pk (A.H.); j.iqbal@ceme.nust.edu.pk (J.I.)
2 National Centre of Robotics and Automation(NCRA), Islamabad 44000, Pakistan;
rana.iqtidar@mail.au.edu.pk
3 Department of Mechatronics Engineering, Air University, Islamabad 44000, Pakistan
* Correspondence: mubasher.saleem@ceme.nust.edu.pk; Tel.: +92-51-54444470

Received: 6 July 2020; Accepted: 14 September 2020; Published: 17 September 2020
Abstract: This paper presents microfabrication process-driven design of a multi-degree of

freedom (multi-DoF) non-resonant electrostatic microelectromechanical systems (MEMS) gyroscope
by considering the design constraints of commercially available low-cost and widely-used
silicon-on-insulator multi-user MEMS processes (SOIMUMPs), with silicon as a structural material.
The proposed design consists of a 3-DoF drive mode oscillator with the concept of addition of a collider
mass which transmits energy from the drive mass to the passive sense mass. In the sense direction,
2-DoF sense mode oscillator is used to achieve dynamically-amplified displacement in the sense
mass. A detailed analytical model for the dynamic response of MEMS gyroscope is presented and
performance characteristics are validated through finite element method (FEM)-based simulations.
The effect of operating air pressure and temperature variations on the air damping and resulting
dynamic response is analyzed. The thermal stability of the design and corresponding effect on
the mechanical and capacitive sensitivity, for an operating temperature range of −40 ◦ C to 100 ◦ C,
is presented. The results showed that the proposed design is thermally stable, robust to environmental
variations, and process tolerances with a wide operational bandwidth and high sensitivity. Moreover,
a system-level model of the proposed gyroscope and its integration with the sensor electronics is
presented to estimate the voltage sensitivity under the constraints of the readout electronic circuit.
Keywords: MEMS gyroscope; multi-degree of freedom (multi-DoF); non-resonant; microfabrication;

thermal stability; finite element method (FEM); system modeling; robustness
1. Introduction
The gyroscopes detect rotation of an object and are an integral part of Inertial Measurement
Units (IMUs) which enable the development of self-contained and high precision Inertial Navigation
Systems (INSs). In comparison to traditional mechanical gyroscopes with rotating parts, ring-laser
and fibre-optic gyroscopes, the microelectromechanical systems (MEMS) gyroscopes offer small size,
are lightweight and low-cost owing to batch fabrication, low power consumption and integration with
integrated circuit (IC) technology.
The working of vibratory MEMS gyroscopes is dependent on the Coriolis effect which requires
oscillations of the proof mass to be sustained in the drive axis and a sensing mechanism for the
Micromachines 2020, 11, 862; doi:10.3390/mi11090862 www.mdpi.com/journal/micromachines

Micromachines 2020, 11, 862 2 of 26
measurement of displacement of the proof mass corresponding to an input rotation. In MEMS

gyroscopes, the oscillations in the proof mass and sensing of its displacement are achieved by using
different transduction mechanisms including electrostatic [1], electrothermal [2], piezoelectric [3],
piezoresistive [4] and optical [5]. Among these transduction mechanisms, the electrostatic is most
commonly used due its fast response time, low power consumption¸ ease of fabrication and integration
with the drive and readout electronics [6].
Based on the working principle, the MEMS gyroscopes are divided into two main categories
including resonant and non-resonant gyroscopes. In resonant MEMS gyroscopes, the device is operated
at resonance and both the drive and sense mode resonant frequency values are generally matched
which leads to high mechanical sensitivity [7]. The reliability of resonant MEMS gyroscopes is of
major concern since a mismatch between the drive and sense mode resonant frequencies occurs due
to microfabrication process tolerances and fluctuations in the operating temperature and pressure
conditions [8,9]. Thus, resonant MEMS gyroscope requires strict frequency control using feedback
electronics [10] or electrostatic frequency tuning [11].
The non-resonant MEMS gyroscopes make a multi-degree of freedom (multi-DoF) system with
masses coupled to each other through mechanical suspension in either the drive or sense mode.
The non-resonant MEMS gyroscope works in the flat region between the resonant peaks of a multi-DoF
system and offers inherent robustness against fabrication process and environmental variations.
The main challenges in the design of multi-DoF MEMS gyroscope involve increasing both mechanical
sensitivity and operational bandwidth. The earliest design approach, based on the two-mass
system in the drive mode utilizing the dynamic vibration absorber concept, was presented in [12].
This approach was used and built upon by different researchers for the design of non-resonant
MEMS gyroscopes [13–15]. The main drawback of this two-mass system-based approach is that the
operational bandwidth and mechanical gain of the MEMS gyroscope can only be increased by
optimizing the mass and resonant frequency ratio of masses that are coupled together. To overcome
this issue, a modified vibration absorber design concept has been proposed in [16] with the addition of
a grounded mechanical spring that is attached to the absorber mass. The proposed design concept
overcomes the requirement of having very small absorber mass but does not resolve the issue of
variation in the stiffness of mechanical springs due to fabrication imperfections or high operating
temperature values. A multi-mass-based system with the addition of collider mass between the drive
and sense mass has been presented in [17] to minimize the effect of operating conditions and to achieve
higher dynamic amplification. The collider mass in this design approach transmits the dynamic energy
from the drive mass to the final oscillating sense mass. Previously, the authors have presented the
implementation of the collider mass concept, for the transfer of energy from drive mass to sense mass,
for a 3-DoF drive and 1-DoF sense mode electrothermally actuated MEMS gyroscope [18]. The MEMS
gyroscope was optimized considering the microfabrication process constraints of MetalMUMPs
process with 20 µm thick structural layer of nickel. In comparison to our previous work, in this
paper, we present an electrostatic multi-DoF MEMS gyroscope, based on the collider mass concept
presented in [17], with 3-DoF drive and 2-DoF sense mode. The proposed design is optimized using
commercially available SOIMUMPs microfabrication process with 25 µm thick structural layer of
silicon. The 2-DoF sense mode and the fact that the silicon is more thermally stable in comparison to
nickel, allows to achieve higher robustness, in the operating temperature range of −40 ◦ C to 100 ◦ C,
for the proposed MEMS gyroscope. Table 1 summarizes the multi-DoF-based non-resonant MEMS
gyroscopes presented in the literature along with their material, structural layer thickness, device size,
sensitivity and operational bandwidth. The sensitivity values vary in units owing to the fact that in
the literature either mechanical, capacitive or voltage sensitivity values are reported for a respective
MEMS gyroscope design. Most of these designs utilize a multi-DoF design concept in either the drive
or sense mode, but not in both.
Table 1. Multi-degree of freedom (multi-DoF) non-resonant microelectromechanical systems (MEMS)

gyroscopes presented in the literature.
Structural
Layer
References Configuration Device Size Sensitivity Bandwidth
Thickness
Material
(µm)
2 DoF drive,
Acar et al. [13] Polysilicon 2 0.7 × 0.5 mm2 0.72 × 10−3 µm/◦ /s 23 Hz
2 DoF sense
Schofield et al. 1 DoF drive,
Silicon 75 - 2.34 µV/(◦ /s) 600 Hz
[14] 2 DoF sense
1 DoF drive,
Trusov et al. [15] Silicon 50 3 × 3 mm2 56 µV/(◦ /s) 250 Hz
2 DoF sense
1 DoF drive,
Sahin et al. [16] Silicon 100 131 µV/(◦ /s) 1 kHz
2 DoF sense
3 DoF drive, 0.565 fF/rad/s, 0.052
Saqib et al. [18] Nickel 20 3.2 × 3 mm2 1.2 kHz
1 DoF sense × 10−3 µm/◦ /s
2 DoF drive, 0.045 fF/rad/s,
Saleem et al. [19] Nickel 20 2.4 × 1.6 mm2 1.71 kHz
1 DoF sense 0.0105 × 10−3 µm/◦ /s
1 DoF drive,
Verma et al. [20] Nickel 9 2 × 1.9 mm2 - 100 Hz
2 DoF sense
1 DoF drive,
Acar et al. [21] Polysilicon 100 4 × 4 mm2 0.0308 mV/◦ /s 50 Hz
2 DoF sense
2 DoF drive,
Riaz et al. [22] Nickel 20 2.2 × 2.6 mm2 - 1.4 kHz
1 DoF sense
2 DoF drive,
Verma et al. [23] Nickel 8 2.13 × 1.93 mm2 - 1 kHz
1 DoF sense
The operation of a MEMS gyroscope is strongly dependent on the operating temperature [24].
The high-temperature values may lead to stiffness degradation of the mechanical springs, variations in
the air damping force and misalignment of the sensing parallel plates in electrostatic MEMS devices
due to thermal expansion. Generally, for the MEMS gyroscopes, different temperature compensation
techniques are implemented after the fabrication for reliable operation [25,26], which results in added
complexity. Thus, both the selection of thermally stable material for the fabrication of MEMS gyroscope
and structural design optimization considering the temperature effects must be considered at the
design level.
The electrical sensitivity of a MEMS gyroscope is not only dependent on the structural design,
but also on the readout electronics. For the MEMS gyroscopes, that work on the electrostatic actuation
and sensing principle, it becomes very important to analyze the device performance in integration
with the sensing electronics at the design phase. Most of the works presented in the literature on the
MEMS gyroscope design generally lack this aspect. In this paper, we present the design of a multi-DoF
non-resonant electrostatic MEMS gyroscope with 3-DoF drive and 2-DoF sense mode oscillators to
achieve both wide operational bandwidth and high mechanical sensitivity. The design is optimized
considering the commercially-available SOIMUMPs fabrication process with a structural layer of
25 µm thick silicon material. The effect of temperature and pressure variations on the air damping
and dynamic behavior of the MEMS gyroscope is analyzed in detail. Moreover, the integration of the
proposed MEMS gyroscope with readout electronics is discussed in detail.
2. Structural Design of Multi-Degree of Freedom (Multi-DoF) Microelectromechanical Systems

(MEMS) Gyroscope
Figure 1 shows the proposed multi-DoF MEMS gyroscope design with a 3-DoF drive mode
and 2-DoF sense mode oscillators. The drive and sense masses are connected to each other using
serpentine-shaped mechanical springs. These springs allow the displacement of the masses in the
desired direction only, thus minimizing the cross-axis displacement. The oscillations in the drive
axis (x-axis) are obtained through a comb drive-based electrostatic actuator which is attached to the
outer mass m1 . The combination of mass m3a and m3b along with the decoupling frame of mass m f ,
acts as a final vibration absorber m3 . For an applied actuation voltage to the electrostatic actuator,
a dynamically-amplified displacement is achieved in the mass m3 while the mass m2 transfers the
dynamic energy from mass m1 to m3 . For an input rotation in the z-axis, a Coriolis force is generated
in the orthogonal axis to drive and rotation axes i.e., y-axis. This leads to a displacement in the
2-DoF sense mode oscillator consisting of a mass m3a and m3b , which are connected to each other
through mechanical springs. The mass m3b acts as a vibration absorber for the mass m3a in the sense
direction, resulting in a dynamically-amplified displacement in the mass m3b with respect to mass m3a
corresponding to the Coriolis force. The displacement in the mass m3b is sensed using parallel plates,
which are arranged in a gap–antigap-based differential configuration [27]. The small gap (d1 ) is kept as
3 µm while the large gap (d2 ) value is 9 µm. The change in capacitance in the sensing parallel plates
has an inverse relationship to the initial air gap between the plates. The minimum air gap allowed by
the microfabrication process is 2 µm. However, to meet the microfabrication process constraints and to
minimize the Brownian noise and air damping effect, the initial small gap (d1 ) is considered to be 3 µm.
To minimize the effect of sudden mechanical shock, semicircle-shaped end stoppers are designed at the
end of the masses adjacent to the anchors. Table 2 shows the main design parameters of the multi-DoF
MEMS gyroscope.
Micromachines 2020, 11, x
Figure 1.
Figure Proposed multi-degree
1. Proposed multi-degree of
of freedom
freedom (multi-DoF) microelectromechanical systems
(multi-DoF) microelectromechanical systems (MEMS)
(MEMS)
gyroscope design
gyroscope design with
with 3-DoF
3-DoF drive
drive mode
mode and
and 2-DoF
2-DoF sense
sense mode
mode oscillators.
oscillators.
Table 2. Design parameters of the proposed multi-DoF MEMS gyroscope.
Parameters Values
Overall device size 4.2 mm × 4.2 mm
Structural layer thickness (𝑡) 25 μm
Mass value of the mass (𝑚 ) 2.5239 × 10−7 Kg
Mass value of the decoupling frame (𝑚 ) 3.234 × 10−8 Kg
Length of the parallel sensing plates (𝑙) 500 μm
Table 2. Design parameters of the proposed multi-DoF MEMS gyroscope.
Parameters Values
Overall device size 4.2 mm × 4.2 mm
Structural layer thickness (t) 25 µm
Mass value of the mass (m1 ) 2.5239 × 10−7 Kg
Mass value of the mass (m2 ) 1.1203 × 10−7 Kg
Mass value of the decoupling frame (m f ) 3.234 × 10−8 Kg
Mass value of the mass (m3a ) 9.8125 × 10−8 Kg
Mass value of the mass (m3b ) 5.5343 × 10−8 Kg
Length of the parallel sensing plates (l) 500 µm
Width of the parallel sensing plates (w) 8 µm
Number of parallel sensing plates pair (Ns ) 120
Overlap length between the moving and fixed
450 µm
parallel sensing plates (l0 )
Smaller sense gap size (d1 ) 3 µm
Larger sense gap size (d2 ) 9 µm
Number of drive comb pairs (Nd ) 240
Width of drive combs (wd ) 6 µm
Length of drive combs (ld ) 120 µm
Gap between drive combs (d) 3 µm
3. Microfabrication Process
The MEMS gyroscope is designed following the constraints of the commercially available and
low-cost multiuser SOIMUMPs micromachining process, offered by MEMSCAP Inc. USA [28].
In addition to being a mature fabrication process, SOIMUMPs offers to fabricate high aspect
ratio microstructures with minimum air damping, due to absence of substrate below the moving
microstructure. Both of these attributes are generally desired in the design of high-performance
MEMS inertial sensors. The SOIMUMPs microfabrication process is a single-wafer silicon-on-insulator
(SOI) micromachining process and allows patterning and etching on the SOI wafer using four mask
layers with different thickness. The topmost layer, called Padmetal, is followed by a device layer
of 25 µm thickness. An oxide layer of 2 µm thickness is sandwiched between the device layer and
the 400 µm thick substrate layer. The minimum feature size for the device layer is 2 µm, but to
avoid buckling due to the residual stresses it is suggested, by the foundry, that the beams must have
a thickness of 6 µm or more for a length of more than 100 µm. Figure 2 shows the details of the
microfabrication process step for the proposed MEMS gyroscope design. The main steps involved are;
(a) The process starts with an SOI wafer as shown in Figure 2a. The doping of the top layer of silicon
is done by depositing a phosphosilicate glass (PSG) layer and then annealing is done at 1050 ◦ C
in Argon. It is followed by the removal of the PSG layer using wet etching.
(b) The lift-off process is then used to deposit the first Padmetal layer which consists of 500 nm gold
stack and 20 nm chrome layer, as shown in Figure 2b.
(c) To apply a second mask, a lithography technique is utilized for the patterning of the silicon layer
as shown in Figure 2c.
(d) To apply the TRENCH mask, substrate layer of thickness 400 µm is patterned in which
lithography technique is used with successive etching through Deep Reactive Ion Etching
(DRIE). The final structure is obtained by removing the buried oxide layer in the regions defined
by the TRENCH mask, as shown in Figure 2d.
The electrical isolation between the top structural layer of silicon and substrate is achieved using an
Oxide layer (silicon oxide layer). However, since there is no insulating layer between the top PadMetal
layer and structural silicon layer, so the PadMetal layer patterned on different parts of the structure is
electrically connected to each other due to structural doping of the silicon. To achieve the electrical
isolation between the fixed and moving combs in the electrostatic actuator and parallel displacement
microfabrication process step for the proposed MEMS gyroscope design. The main steps involved
are;
(a) The process starts with an SOI wafer as shown in Figure 2a. The doping of the top layer of silicon
is done by depositing a phosphosilicate glass (PSG) layer and then annealing is done at 1050
Micromachines 2020, 11, 862
°C
6 of 26
in Argon. It is followed by the removal of the PSG layer using wet etching.
(b) The lift-off process is then used to deposit the first Padmetal layer which consists of 500 nm gold
sensing plates,
stack and 205 nm wide and
µm chrome 25 µm
layer, deep isolation
as shown in Figuregrooves,
2b. of different lengths, are made in the
(c) To apply a second mask, a lithography technique is utilized for the1,patterning
structural layer of silicon as shown in Figure 2d. The zoomed sections 2 and 3 show the
of the location
silicon layerof
the various
as shown isolation grooves.
in Figure 2c. The grooves shown in zoomed section 1, disconnect the movable masses
fromTo
(d) theapply
fixed combs and are repeated
the TRENCH on all fourlayer
mask, substrate corners
of ofthickness
the MEMS gyroscope.
400 The zoomed
μm is patterned in section
which
2 shows two isolation grooves which isolate the oscillating masses from the
lithography technique is used with successive etching through Deep Reactive Ion Etchingfixed actuator combs and
sensing parallel
(DRIE). Theplates. The zoomed
final structure sectionby
is obtained 3 shows the isolation
removing the buriedgroove
oxidebetween
layer in the two setsdefined
regions of fixed
sensing parallel
by the TRENCH platesmask,
for the
aselectrical
shown inisolation.
Figure 2d.
Figure 2. Microfabrication process steps for the proposed MEMS gyroscope design.
3.1. Microfabrication Process Limitations

For the accurate release of microstructures and structural integrity, the SOIMUMPs microfabrication
process has certain design rules to be followed. For the complex MEMS devices, like MEMS gyroscopes,
this limits the design options for the MEMS designer. The main limitations that are considered for the
proposed multi-DoF MEMS gyroscope design are discussed below.
3.1.1. Limitation 1
The design cannot have a central anchor as shown in Figure 3. This is because such an
anchor/doughnut feature does not survive during the microfabrication etching step and eventually
falls out. For the proposed multi-DoF MEMS gyroscope, this implies that in comparison to traditional
are considered for the proposed multi-DoF MEMS gyroscope design are discussed below.
3.1.1. Limitation 1
The design cannot have a central anchor as shown in Figure 3. This is because such an
anchor/doughnut feature does not survive during the microfabrication etching step and eventually
falls out. For the proposed multi-DoF MEMS gyroscope, this implies that in comparison to traditional
multi-DoF
multi-DoF MEMSMEMS gyroscope
gyroscope designs
designs reported
reported inin the
the literature,
literature, the
the sense
sense mass
mass cannot
cannot bebe completely
completely
enclosed withinanother
enclosed within anothermass massdue
dueto to
thethe requirement
requirement of anchor
of anchor part part forfixed
for the the fixed parallel
parallel sensingsensing
plates.
plates. Thus, for the proposed MEMS gyroscope, the anchor for the fixed parallel
Thus, for the proposed MEMS gyroscope, the anchor for the fixed parallel sensing plates attached to sensing plates
attached
mass m3b,toismass m3b, isbyprovided
provided by routing
routing the the silicon
silicon fixed beamsfixed
from beams
one sidefrom oneouter
of the sidemass
of themouter mass
1 as shown
m as shown
in1 Figure 1. in Figure 1.
Figure 3.
Figure Schematic of
3. Schematic of invalid
invalid doughnut/anchor
doughnut/anchor feature
feature in
in the
the SOIMUMPs
SOIMUMPs process.
process.
3.1.2. Limitation 2
3.1.2. Limitation 2
As the silicon substrate is completely etched in the SOIMUMPs process, bottom electrodes for
As the silicon substrate is completely etched in the SOIMUMPs process, bottom electrodes for
the out-of-plane actuation and sensing cannot be fabricated in this process. This leaves the MEMS
the out-of-plane actuation and sensing cannot be fabricated in this process. This leaves the MEMS
designer to use only in-plane actuation and sensing techniques.
designer to use only in-plane actuation and sensing techniques.
3.1.3. Limitation 3
3.1.3. Limitation 3
The silicon layer is the structural layer in which the device is patterned, as shown in red in Figure 2.
The silicon layer is the structural layer in which the device is patterned, as shown in red in Figure
As per the design rules of the SOIMUMPs process, the etched part of the silicon layer must be less than
2. As per the design rules of the SOIMUMPs process, the etched part of the silicon layer must be less
33% of the total chip area. This limitation is imposed to make sure that the structural integrity of the
than 33% of the total chip area. This limitation is imposed to make sure that the structural integrity
device remains intact after fabrication. For the proposed MEMS gyroscope, this rule limits the amount
of allowed spacing between the masses and hence on the number of turns of mechanical springs.
3.2. Process Driven Design Modifications

Some design choices had to be made in order to mitigate the limitations imposed by the
microfabrication process. The first limitation states that an anchor in the middle surrounded by empty
space is not permitted. To remedy this, the fixed parallel sensing plates are partially enclosed within
the sense mass m3b . Ideally, the anchor should have been completely enclosed to make the MEMS
gyroscope design symmetric. However, this is not permitted, so all the masses have to be altered so that
support can be provided to the central anchor from the external anchor. This approach allows to alter
the design minimally and keep the design simple and as symmetric as possible. Although the device
does not remain completely symmetric after this modification, the effects of this asymmetry on the
performance of the device are negligible. This is verified through finite element method (FEM)-based
thermal stability analysis, discussed in Sections 5.5 and 5.6. The second limitation is the absence of
a bottom electrode. Thus, for the proposed MEMS gyroscope, we have used in-plane comb drive
actuators and in-plane parallel plate-based displacement sensing mechanism. The third limitation is to
keep the etched area in the structural layer less than 33% of the total device area. This is achieved by
minimizing the air gaps between the masses and by reducing the unused empty spaces in the design.
4. Analytical Modeling of the Multi-DoF MEMS Gyroscope
4.1. Equations of Motion for the Drive and Sense Mode Oscillators
Figure 4a shows the 3-DoF drive mode mass–spring–damper model representation of the proposed
MEMS gyroscope. For an input driving force Fcos(ωt), displacement of mass m1 , m2 and sense mass
(m3a , m3b , m f ) is represented by x1 , x2 , and x3 respectively. The mass m1 and sense mass (m3a , m3b ,
m f ) are anchored to the substrate through mechanical springs k1 and k3 respectively. The mass
m2 , attached to the mass m1 and sense mass (m3a , m3b , m f ) through mechanical springs k12 and k23 ,
respectively, acts as a collider mass and transfers the dynamic energy from the mass m1 to the sense
mass (m3a , m3b , m f ). The mass m2 is also connected to the substrate through mechanical spring k2 .
The k2 allows us to optimize the dynamic displacement amplification between the mass m1 and sense
mass (m3a , m3b , m f ) and the width of the flat operational region between the resonance peaks of drive
mode oscillators [17]. The air damping between the actuator combs attached to the mass m1 and
sensing parallel plates attached to sense mass (m3a , m3b , m f ) are represented by c1x and c2x , respectively.
Figure 4b shows the 2-DoF sense mode mass–spring–damper model of the MEMS gyroscope. The mass
m3b acts as dynamic vibration absorber for the mass m3a and both masses are coupled through the
mechanical spring k y1 . The mass m3a is attached to the decoupling frame m f , which acts as an anchor,
through mechanical spring k y2 and the air damping in the sensing parallel plates attached to the mass
Micromachines 2020, 11, by
m3b is represented x c .
y
Figure 4.
Figure Discrete mass–spring–damper
4. Discrete mass–spring–damper model
model for
for the
the proposed
proposed MEMS
MEMS gyroscope.
gyroscope. (a)
(a) Drive
Drive mode
mode
and (b) sense mode.
and (b) sense mode.
For an angular rate of Ω in the z-axis, the second-order differential equations of motion for the
For an angular rate of Ω in the z-axis, the second-order differential equations of motion for the
proposed MEMS gyroscope written as:
proposed MEMS gyroscope written as:
.. .
𝑚
m1 𝑥 + k𝑘12 ((𝑥
x1 + 𝑐c1x x𝑥1 + k𝑘1 x𝑥1 + x1 −−x𝑥2 ) )==F𝐹D (1)
(1)
𝑚m 𝑥x.. + 𝑘 𝑥 + 𝑘 (𝑥 − 𝑥 ) + 𝑘 (𝑥 − 𝑥 ) = 0 (2)
2 2 + k2 x2 + k12 (x2 − x1 ) + k23 (x2 − x3 ) = 0 (2)
𝑚 + 𝑚 + 𝑚 𝑥.. + 𝑐 𝑥 . + 𝑘 𝑥 + 𝑘 (𝑥 − 𝑥 ) = 0 (3)
m f + m3a + m3b x3 + c2x x3 + k3 x3 + k23 (x3 − x2 ) = 0 (3)
𝑚 𝑦 .. + 𝑘 𝑦 + 𝑘 (𝑦 − 𝑦 ) = 𝐹 ( ) (4)
m3a y1 + k y1 y1 + k y2 ( y1 − y2 ) = Fc( m3a) (4)
𝑚 𝑦 .. + 𝑐 𝑦
. + 𝑘 (𝑦 − 𝑦 ) = 𝐹 ( )
, (5)
m3b y2 + c y y2 + k y2 ( y2 − y1 ) = Fc( m3b) , (5)
where 𝐹 = 𝐹𝑐𝑜𝑠(𝜔𝑡) is the driving actuation force, 𝐹 ( ) = −2𝑚 .𝛺𝑥 and 𝐹 ( ) = −2𝑚 . 𝛺𝑥
where FD = Fcos(ωt) is the driving actuation force, Fc( m3a) = −2m3a Ωx3 and Fc( m3b) = −2m3b Ωx3 are
are the Coriolis forces acting on the 2-DoF sense mode oscillator.
the Coriolis forces acting on the 2-DoF sense mode oscillator.
4.2. Calculation of Mechanical Stiffness
4.2. Calculation of Mechanical Stiffness
Figure 5 shows the configuration of mechanical springs attached to the masses. There is a total
Figure 5 shows the configuration of mechanical springs attached to the masses. There is a total of
of seven sets of mechanical springs in the x and y-axis. Five sets of mechanical springs, namely 𝑘 ,
seven sets of mechanical springs in the x and y-axis. Five sets of mechanical springs, namely k1 , k2 , k3 ,
𝑘 , 𝑘 , 𝑘 and 𝑘 are placed such that they permit motion of the mass 𝑚 , 𝑚 and 𝑚 in the
k12 and k23 are placed such that they permit motion of the mass m1 , m2 and m f in the drive direction,
drive direction, but restrict any motion in the sense direction. Another two sets of mechanical springs
namely 𝑘 and 𝑘 are placed such that they allow movement of the mass 𝑚 and 𝑚 in the
sensing direction only. This configuration of the mechanical springs allows the decoupling frame mf
and the sense masses 𝑚 and 𝑚 to behave as a single mass in the drive direction.
but restrict any motion in the sense direction. Another two sets of mechanical springs namely k y1 and
k y2 are placed such that they allow movement of the mass m3a and m3b in the sensing direction only.
This configuration of the mechanical springs allows the decoupling frame mf and the sense masses m3a
and m3b to behave as a single mass in the drive direction.
Figure 5. Mechanical suspension design for the proposed MEMS gyroscope design.
Figure 5. Mechanical suspension design for the proposed MEMS gyroscope design.
Two types of springs, with different dimensions, are used in the MEMS gyroscope design i.e.,
Two types of springs, with different dimensions, are used in the MEMS gyroscope design i.e.,
double-folded and quadruple-folded serpentine springs. In the drive direction, the double-folded
double-folded and quadruple-folded serpentine springs. In the drive direction, the double-folded
mechanical springs k with length L and width w , connect the mass m1 with the external anchor.
mechanical springs 𝑘1, , with length k𝐿1 and widthk1 𝑤 , connect the mass 𝑚 with the external
Two double-folded springs k2 , with length and width given by Lk2 and wk2 , respectively, are used
anchor. Two double-folded springs 𝑘 , with length and width given by 𝐿 and 𝑤 , respectively,
to connect the mass m2 to the anchors. Two double-folded springs k3 , are used to connect the
are used to connect the mass 𝑚 to the anchors. Two double-folded springs 𝑘 , are used to connect
decoupling frame mf to the anchors and their lengths and widths are given by Lk3 and wk3 respectively.
the decoupling frame mf to the anchors and their lengths and widths are given by 𝐿 and 𝑤
Two mechanical springs, k12 and k23 , are used to interconnect the drive masses, whereas k y1 and k y2 are
respectively. Two mechanical springs, 𝑘 and 𝑘 , are used to interconnect the drive masses,
used to interconnect the sense masses. A total of four mechanical springs k12 with their length and
whereas 𝑘 and 𝑘 are used to interconnect the sense masses. A total of four mechanical springs
width given by Lk12 and wk12 respectively, are used to couple the masses m1 and m2 . Four mechanical
𝑘 with their length and width given by 𝐿 and 𝑤 respectively, are used to couple the masses
springs k23 with length and width of Lk23 and wk23 respectively, are used to connect the mass m2 and
𝑚 and 𝑚 . frame
decoupling Four mechanical springs
m f . In the sense 𝑘 withfour
direction, length and width
springs 𝐿 toand
k y1 areofused 𝑤 the
couple respectively, are
frame m f with
to connect the mass 𝑚 and decoupling frame 𝑚
the sense mass m3a . Their lengths and widths are given by Lky and wk y respectively. Another setare
used . In the sense direction, four springs 𝑘 of
used to couple the frame 𝑚 with the sense mass 𝑚 . Their 1
lengths 1
and widths are
four springs k , is used to interconnect the masses m and m , with length and width given by Lgiven by 𝐿
y2 3a 3b k y2
and
and w 𝑤k . respectively.
The mechanicalAnother setboth
springs, of four
in drive and𝑘sense
springs , is used to interconnect
direction, the masses
can be modeled 𝑚 and
as fixed-guided
y2
𝑚 , with length and width given by 𝐿 and 𝑤 . The mechanical
beams and their equivalent mechanical stiffness can be written as: springs, both in drive and sense
direction, can be modeled as fixed-guided beams and their 3equivalent mechanical stiffness can be
β Etwk j

written as: β  12EI 

k j =  3  =
  , (6)
γ L  γ L3
𝛽 12𝐸𝐼 kj 𝛽 𝐸𝑡𝑤 kj
𝑘 = = , (6)
𝛾 𝐿 𝛾 𝐿
tw3
where E represents the Young’s Modulus of the material, I is the moment of inertia given by I = 12 , t is
the thickness,
where L is the
E represents thelength
Young’s w is the width
andModulus of theof the beams.
material, I is the subscriptofj inertia
Themoment represents theby
given 𝐼=
concerned ,
mechanical spring and can have a value of 1, 2, 3, 12, 23, y and y . The β and γ
t is the thickness, 𝐿 is the length and 𝑤 is the width of the beams. The subscript j represents the
1 2 represent the number
of folds in the
concerned serpentine
mechanical springs
spring andand
canthehave
totalanumber
value of of mechanical
1, 2, 3, 12, springs
23, 𝑦 and respectively.
𝑦 . The 𝛽 and 𝛾
represent the number of folds in the serpentine springs and the total number of mechanical springs
4.3. Calculation of Differential Capacitance Change
respectively.
For the proposed MEMS gyroscope design, an input angular rotation in the z-axis will result in
4.3.
the Calculation of Differential
Coriolis force in the y-axisCapacitance
which willChange
lead to the displacement in the 2-DoF sense mode oscillator.
To detect theproposed
For the displacement
MEMSingyroscope
the sense design,
mass, the
an sensing parallel
input angular plates in
rotation arethe
attached to the
z-axis will sense
result in
the Coriolis force in the y-axis which will lead to the displacement in the 2-DoF sense mode oscillator.
To detect the displacement in the sense mass, the sensing parallel plates are attached to the sense
mass 𝑚 . These sensing plates are arranged in differential gap–antigap configuration and the
overall capacitance change, corresponding to the sense mass displacement, can be calculated as [29]:
mass m3b . These sensing plates are arranged in differential gap–antigap configuration and the overall
capacitance change, corresponding to the sense mass displacement, can be calculated as [29]:

2yεo lo tN d22 − d21
∆C = , (7)
d22 − y2 d21 − y2
where N is the number of parallel plates on each side, εo is vacuum permittivity, lo is the overlap length
of the fixed and moving parallel plates, y is the sense mass displacement and d1 and d2 are the small
and large air gaps respectively.
4.4. Air Damping Analysis

For the proposed MEMS gyroscope, both the slide and squeezed film air damping contribute
towards the energy dissipation. The slide film air damping is the only energy dissipation mechanism
in the drive direction. The moving drive combs and moving sense parallel plates both move laterally
in the drive direction and slide over the fixed parallel plates, hence contributing to the slide film air
damping. The movement of masses in the sense direction results in the squeezed film air damping
phenomenon. Due to the movement of sense mass, the moving parallel plates displace towards the
fixed parallel plates and the thin air film between moving and fixed parallel plates is squeezed in.
The slide film (bsl ) and squeezed film (bsq ) damping coefficients can be calculated as [30,31]:
!
1 1
bsl = Nµe f f A + (8)
d1 d2
 
3
1 1 
bsq = Ns µe f f lo t  3 + 3 ,
 (9)
d1 d2
where N is the number of moving combs on each side of the mass, Ns is the number of sensing parallel
plates, A is the overlap area of the parallel plates, lo is the overlap length, µe f f is the effective viscosity
of air and d1 and d2 are the smaller and larger gap sizes respectively. For drive combs, both d1 and
d2 are of the same size i.e., d but for parallel sensing plates the values of d1 and d2 are 3 µm and
9 µm respectively.
The squeezed film air damping depends on the viscous, elastic and inertial effects of the thin air
film present between the moving and fixed parallel plates. There are several types of flow regimes for
the thin air film and are generally classified depending on a dimensionless value called the Knudsen
number, Kn , which is defined as the ratio of the mean free path of air to the thickness of the air gap,
and is given as:
λ
Kn = , (10)
d1
where λ is the mean free path of air at a given operating temperature and pressure and d1 is the nominal
air gap thickness. The mean free path of the air can be calculated as [32]:
µ
r
πKB T
λ= , (11)
P 2mair
where µ is the air viscosity at atmospheric pressure, mair is the mass of air, P is the air pressure, T is
the air temperature and KB is the Boltzman constant. The classification of flow regimes based on
the Kn values are discussed in [33]. For the proposed MEMS gyroscope, the smaller air gap between
the sensing parallel plates is 3 µm which results in a Kn value 0.0225 at atmospheric pressure and
room temperature. This shows that, since this value is within 0.01 < Kn < 0.1, the flow regime for the
proposed MEMS gyroscope is slip flow. In the proposed MEMS gyroscope operating temperature
range of −40 ◦ C to 100 ◦ C, the dominant flow regime remains slip flow with Kn values of 0.0175 and
0.0275 at −40 ◦ C and 100 ◦ C, respectively.
The air damping force acting on the MEMS gyroscope can be either viscous, elastic or inertial and
is dependent on the oscillation frequency of the mass. The relative effect of the air damping force is
generally classified based on the dimensionless squeeze number σ, which is given as:
12µe f f t2𝑡ω⍵
12µ
σ𝜎== ,, (12)
(12)
𝑃𝑑2
Pd 1
where 𝑡 is the characteristic length and for the proposed MEMS gyroscope represents the thickness
where t is the characteristic length and for the proposed MEMS gyroscope represents the thickness
of the parallel sensing plates, µ is the effective viscosity, P is the air pressure, 𝑑 is the smaller air
of the parallel sensing plates, µe f f is the effective viscosity, P is the air pressure, d1 is the smaller air
gap and ⍵ is operating frequency of the MEMS gyroscope. At low values of the squeeze number,
gap and ω is operating frequency of the MEMS gyroscope. At low values of the squeeze number,
the viscous damping force dominates the elastic force since the air between the fixed and moving
the viscous damping force dominates the elastic force since the air between the fixed and moving
parallel sense plates is not compressed completely. However, at high values of the squeeze number,
parallel sense plates is not compressed completely. However, at high values of the squeeze number,
the elastic damping force dominates, since the thin air film fails to escape from the gap between the
the elastic damping force dominates, since the thin air film fails to escape from the gap between the
fixed and moving parallel sensing plates. The term µ in Equation (12) is the effective viscosity of
fixed and moving parallel sensing plates. The term µe f f in Equation (12) is the effective viscosity
air at a given temperature and pressure, and is related to the air viscosity at ambient conditions, given
of air at a given temperature and pressure, and is related to the air viscosity at ambient conditions,
as [34]:
given as [34]:
µµ
µµe f f == .. (13)
(13)
11+ 9.638𝐾n1.1.
+9.638K
The above equation
The above equationshows
shows thatthat µ µis
the the is dependent
e f fdependent on theon 𝐾 the
andKhence
n andonhence on the
the operating
operating temperature. Figure 6 shows the effect of temperature variations on the
temperature. Figure 6 shows the effect of temperature variations on the squeeze number in the squeeze number
in the operating temperature range −40
of to ◦ C to 100 ◦ C. The results show that the squeeze number
operating temperature range of −40 °C 100 °C. The results show that the squeeze number decreases
decreases for the MEMS
for the proposed proposed MEMS as
gyroscope gyroscope as the temperature
the temperature increases. Theincreases. The of
inertial effect inertial
dampingeffect of
force
damping force
is generally is generally
explained explained
by using by using
Reynold’s Reynold’s
number given number
as [35]: given as [35]:
𝑑2 𝜌⍵
𝑅 =d1 ρω . (14)
Re = µ . (14)
µe f f
where 𝑑 is the sense gap size, 𝜌 is the density of air and ⍵ is the operating frequency of the MEMS
where d1 is the
gyroscope. For sense gap size,
low values ρ is the density
of Reynolds number of i.e.,
air and𝑅 <<ω is1,the
theoperating frequency
inertial effects of the
are very MEMS
small and
gyroscope. For low values of Reynolds number i.e., R << 1, the inertial effects are
hence can be ignored. For the proposed MEMS gyroscope, the 𝑅 values are 0.00985 and 0.0105 at
e very small and hence
can Re damping ◦ C to
−40MEMS
−40 be
°Cignored.
to 100 °C.For theshows
This proposed
thatMEMS gyroscope,
the inertial effect ofthethe values areforce
0.00985 and 0.0105
is negligible foratthe
100 ◦ C. This shows that the inertial effect of the damping force is negligible for the MEMS gyroscope.
gyroscope.
Squeeze Number
Temperature (oC)
Figure 6. Effect
Figure 6. Effect temperature
temperature variations
variations on
on the
the squeeze
squeeze number
number for
for the
the proposed
proposedMEMS
MEMSgyroscope.
gyroscope.
5. FEM Analysis for the MEMS Gyroscope
5. FEM Analysis for the MEMS Gyroscope
5.1. Modal Analysis
5.1. Modal Analysis
To determine the resonant frequency values and corresponding mode shapes for the MEMS
To determine
gyroscope, the resonant
a FEM-based frequency
analysis was values
carried out in theand corresponding
CoventorWare modemodule.
Analyzer shapes Figure
for the7 MEMS
shows
gyroscope, a FEM-based analysis was carried out in the CoventorWare Analyzer module. Figure 7
shows the desired three drive mode shapes and two sense mode shapes for the MEMS gyroscope.
The first, sixth and tenth modes were in the drive direction with a frequency of 1.608 kHz, 3.329 kHz
and 4.822 kHz respectively. The second and seventh modes are in the sense direction with a frequency
of 1.707 kHz and 3.709 kHz respectively. These results show that the two sense mode resonance
frequencies are close to the first two drive mode frequencies, thus making the operational region of
modes separate from the flat region by at least 200 Hz. Table 3 compares the FEM-based modal
analysis results with the analytical model, which show a close correspondence.
the desired three drive mode shapes and two sense mode shapes for the MEMS gyroscope. The first,
sixth and Table
tenth 3.modes were of
Comparison in the
theanalytical
drive direction
and FEMwith a frequency
analysis-based of frequency
modal 1.608 kHz, 3.329 kHz and
values.
4.822 kHz respectively. The second and seventh modes are in the sense direction with a frequency
of 1.707 kHz and 3.709 ModekHz Shape Analytical
respectively. TheseModel (kHz)
results FEMthe
show that Model
two (kHz)
sense mode resonance
1st (Drive
frequencies are close mode)
to the first two drive mode1.738frequencies, thus making1.608
the operational region of the
MEMS gyroscope2nd fall(Sense mode)
in the flat region between 1.800
the resonance peaks. The1.707
third, fourth and fifth modes
6th (Drive mode) 3.507
are out of the plane and twisting at 1.79 kHz, 1.9 kHz and 2 kHz respectively. 3.329 These undesired modes
separate from the7th flat(Sense
region mode) 3.8673 compares the FEM-based
by at least 200 Hz. Table 3.709 modal analysis results
10th (Drive mode) 4.960
with the analytical model, which show a close correspondence. 4.822
Figure
Figure 7.7. Modal analysis
analysisresults
resultsand
andthethe corresponding
corresponding modemode shapes
shapes for MEMS
for MEMS gyroscope
gyroscope (a) 1st(a) 1st
mode
mode
(1.607 (1.607 kHz)
kHz) (b) 2nd(b) 2nd (1.707
mode mode kHz)
(1.707and
kHz)
(c) and (c) 6th(3.329
6th mode modekHz)
(3.329
andkHz) andmode
(d) 7th (d) 7th mode
(3.709 (3.709
kHz).
kHz).
Table 3. Comparison of the analytical and FEM analysis-based modal frequency values.
5.2. Pull-in Voltage Analysis

Mode Shape Analytical Model (kHz) FEM Model (kHz)
The accurate operation of the parallel plate-based
1st (Drive mode) 1.738 rotation sensing 1.608
mechanism in the proposed
2nd (Sense mode) 1.800
gyroscope requires initial biasing of the sense plates. Due to the bias voltage,1.707 the moving sensing
6th (Drive mode) 3.507 3.329
plate moves towards the fixed plate under the influence of an electrostatic force. The maximum value
7th (Sense mode) 3.867 3.709
of the bias voltage 10this(Drive
limited by the pull-in effect.
mode) 4.960 The voltage at which 4.822 the electrostatic force
overcomes the mechanical force of the suspension beams is called the pull-in voltage and
corresponding displacement is called the pull-in displacement. The pull-in voltage for the differential
5.2. Pull-in Voltage Analysis
parallel sensing plates can be estimated as [36]:
The accurate operation of the parallel plate-based rotation sensing mechanism in the proposed
8𝑘 𝑑to the bias voltage, the moving sensing plate
gyroscope requires initial biasing of the sense plates. Due
𝑉 = , (15)
moves towards the fixed plate under the influence of27𝑙 𝑡𝑁 𝜀
an electrostatic force. The maximum value of the
bias voltage is limited by the pull-in effect. The voltage at which the electrostatic force overcomes the
mechanical force of the suspension beams is called the pull-in voltage and corresponding displacement
is called the pull-in displacement. The pull-in voltage for the differential parallel sensing plates can be
estimated as [36]: s
8k y d31
Micromachines 2020, 11, x Vpull−in = , (15)
27l0 tNs ε
wherek y𝑘is the
where spring
is the constant,
spring constant, 𝑑 initial
d1 is the is the air gap air
initial between the moving
gap between theand fixed parallel
moving and fixed plates, ε is
parallel
the
plates, 𝜀 is theof
permittivity free space and
permittivity Ns isspace
of free and 𝑁 ofismoving
the number and fixed
the number sensing and
of moving platefixed
pairs.sensing
To estimate
plate
the maximum
pairs. bias the
To estimate voltage that can
maximum bevoltage
bias appliedthat
to the
cansensing parallel
be applied plates
to the in the
sensing proposed
parallel MEMS
plates in the
gyroscope, a pull-in
proposed MEMS voltage analysis
gyroscope, a pull-inwas carried
voltage out in was
analysis the CoventorWare MEMS+
carried out in the module. Figure
CoventorWare MEMS+ 8
shows
module.theFigure
applied8 bias
showsvoltage versus sense
the applied mass displacement.
bias voltage versus sense The
massresults show thatThe
displacement. pull-in occurs
results at
show
athat
biaspull-in
voltageoccurs
of 3.4 Vatwith a pull-in
a bias voltagedisplacement
of 3.4 V withof 0.5 µm. Thus,
a pull-in for the proposed
displacement of 0.5 MEMS
μm. Thus, gyroscope,
for the
the maximum
proposed MEMS biasgyroscope,
voltage mustthebe below 3.4bias
maximum V. voltage must be below 3.4 V.
1
Sense mass displacement (µm)
0.5
-0.5
-1
-1.5
-2
-2.5
-3
0 0.5 1 1.5 2 2.5 3 3.5 4
Bias Voltage (V)
Figure 8. Sense mass displacement vs. applied bias voltage graph for the MEMS gyroscope.
Figure 8. Sense mass displacement vs. applied bias voltage graph for the MEMS gyroscope.
5.3. Frequency Response Analysis
5.3. Frequency Response Analysis
To analyze the frequency response of the 3-DoF drive mode and 2-DoF sense mode oscillator,
To analyze
an FEM-based the frequency
harmonic analysis response
is carriedof out
the 3-DoF drive mode and
in the CoventorWare 2-DoF sense
analyzer module.mode oscillator,
A harmonic
force was applied to the mass m1 in the drive axis by applying an actuation voltage of 80 Vharmonic
an FEM-based harmonic analysis is carried out in the CoventorWare analyzer module. A DC and
5force
V ACwas applied
to the the mass 𝑚electrostatic
combtodrive-based in the drive axis by applying
actuators. an actuation
The analysis voltage by
was performed of 80 V DC and
considering
5 Vair
the ACdamping
to the comb drive-based
effect. electrostatic
Figure 9 shows actuators.response
the frequency The analysis
of thewas performed
3-DoF by considering
drive mode oscillator.
the air damping effect. Figure 9 shows the frequency response of the 3-DoF
The results show that at the first resonant frequency, the displacement in the mass m3 is dynamicallydrive mode oscillator.
The results show that at the first resonant frequency, the displacement in the mass
amplified with an amplification ratio of three with respect to the mass m1 . The displacement in the 𝑚 is dynamically
amplified
m1 , mwith an amplification ratio of three with respect to the mass 𝑚 . The displacement in the
mass 2 and m3 is 0.25 µm, 0.6 µm and 0.92 µm respectively. At the second and third resonant
mass 𝑚 , 𝑚 and 𝑚 is 0.25
frequency, the displacement inμm,
the 0.6
senseμmmass
and 0.92
m3 isμm respectively.
minimum At the second
and reduced and third
with respect resonant
to mass m1 .
frequency, the displacement in the sense mass 𝑚 is minimum and reduced
At the first resonant frequency, the three masses were in phase while at second and the third resonant with respect to mass
𝑚 . At thethey
frequency firstwere
resonant
out offrequency,
phase. In the theflat
three masses were
operational inbetween
region phase while at second
the first and resonant
and second the third
resonant frequency they were out of phase. In the flat operational region between
frequency, the displacement in the mass m3 was dynamically amplified with an average amplification the first and second
resonant
ratio frequency,
of nearly two. theThedisplacement the mass 𝑚
displacementinamplitude of thewasmass
dynamically amplified
m3 was nearly 0.22with
µm an average
in the flat
amplification
operational region.ratio of nearly two. The displacement amplitude of the mass 𝑚 was nearly 0.22 μm
in the flat operational region.
1.4
Mass m1
1.4 Mass m2
1.2 Mass m1
Mass m3
Mass m2
(µm)(µm)
1.2
Mass m3
1
displacement
1
displacement 0.8
0.8
0.6
massmass
0.6
0.4
Drive
0.4
Drive
0.2
0.2
0
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000
0
1000 1500 2000 2500
Frequency
3000 3500
(Hz)
4000 4500 5000 5500 6000
Frequency (Hz)
Figure 9. The frequency response of the 3-DoF drive mode oscillator.
Figure 9. The
Figure 9. The frequency
frequency response
response of
of the
the 3-DoF
3-DoF drive
drivemode
modeoscillator.
oscillator.
Figure 10 shows the frequency response of the 2-DoF sense mode oscillator. At the first resonant
Figure
frequency, 10 shows the frequency response
𝑚 of thedynamically
2-DoF sense mode oscillator. with anAt the first resonant
Figurethe
10 displacement in the mass
shows the frequency response was
of the 2-DoF senseamplified
mode oscillator. amplification
At ratio
the first resonant
frequency,
of 1.85. the
This is displacement
because both in
the the mass
mass 𝑚 m was
and 𝑚dynamically
were in amplified
phase at thewith
firstan amplification
sense mode ratio
resonant
frequency, the displacement in the mass 𝑚 was dynamically amplified with an amplification ratio
3b
of 1.85. This is because
frequency. both the mass m and m3b were in phase at were
the first
outsense modethus,resonant
of 1.85. ThisAtisthe second
because resonant
both the massfrequency,
𝑚 3a andsince
𝑚 both werethe masses
in phase at the first of phase,
sense there
mode resonant
frequency.
was At
nearly At the
zero second resonant frequency, since both the masses were out of phase, thus, there was
frequency. thedynamic amplification.
second resonant However,
frequency, since in thethe
both flatmasses
operational
were region between
out of phase, thethere
thus, two
nearly
resonant zero dynamic
peaks, thereamplification.
was an However,
amplified in the
displacement flat operational
in the mass region
𝑚 between
with the
respect two
to resonant
mass 𝑚
was nearly zero dynamic amplification. However, in the flat operational region between the two
peaks,
with anthere was an amplified displacement
average in2.6.
the The m3b with response
massfrequency respect to mass mshown3a with anFigures
average9
resonant peaks, amplification
there was an ratio of nearly
amplified displacement in the mass 𝑚 results with respect toinmass 𝑚
amplification
and ratio of nearly 2.6. The frequency response results shown in Figures 9 and 10 show that
with 10 show that
an average the most ratio
amplification suitable operational
of nearly 2.6. Thefrequency
frequency region
response was between
results shownthe resonant
in Figures 9
the most suitable
frequencies of operational
1.707 kHz andfrequency
3.329 region
kHz with was
a between theof
bandwidth resonant
1.622 frequencies
kHz for the of 1.707 kHz
propose MEMSand
and 10 show that the most suitable operational frequency region was between the resonant
3.329 kHz with a bandwidth of 1.622 kHz for the propose MEMS gyroscope design.
gyroscope
frequenciesdesign.
of 1.707 kHz and 3.329 kHz with a bandwidth of 1.622 kHz for the propose MEMS
gyroscope design.
0.045
Mass m3a
0.045
0.04 Mass m3b
Mass m3a
(µm)(µm)
0.04
0.035 Mass m3b
displacement
0.035
0.03
displacement
0.03
0.025
0.025
0.02
massmass
0.02
0.015
Sense
0.015
0.01
Sense
0.01
0.005
0.005
0
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000
0
1000 1500 2000 2500
Frequency
3000 3500
(Hz)
4000 4500 5000 5500 6000
Figure
Figure 10. Frequency
10. The frequency response (Hz) sense mode oscillator.
of the 2-DoF oscillator.
Figure 10. The frequency
5.4. Oscillation Frequency-Dependent response
Air Damping of the 2-DoF sense mode oscillator.
Analysis
5.4. Oscillation Frequency-Dependent Air Damping Analysis
The air damping
5.4. Oscillation force acting onAir
Frequency-Dependent the gyroscope can either be viscous, elastic or inertial depending
The air damping force acting on theDamping
gyroscopeAnalysis
can either be viscous, elastic or inertial depending
on the oscillation frequency. To analyze the dominant air damping mechanism for the proposed
on the oscillation
The frequency.
air damping To analyze
force acting the dominant
on the gyroscope air damping
can either mechanism
be viscous, for thedepending
elastic or inertial proposed
MEMS gyroscope, a detailed FEM-based analysis was carried out in the DampingMM module of
MEMS gyroscope, a detailed FEM-based analysis was carried out in the DampingMM
on the oscillation frequency. To analyze the dominant air damping mechanism for the proposed module of
CoventorWare. Figure 11a,b shows the effect of oscillation frequency on the slide film and squeeze film
CoventorWare.
MEMS gyroscope, Figure 11a,b shows
a detailed the effect
FEM-based of oscillation
analysis frequency
was carried out inon
thethe slide film and
DampingMM squeeze
module of
air damping respectively at room temperature and pressure conditions and the relative contribution of
film air damping respectively at room temperature and pressure conditions and
CoventorWare. Figure 11a,b shows the effect of oscillation frequency on the slide film and squeeze the relative
film air damping respectively at room temperature and pressure conditions and the relative
contribution
both the viscousof both the viscous and
and elastic/spring dampingelastic/spring
forces. The damping forces.
results show The
that forresults show that
the oscillation for the
frequency
oscillation frequency in the range of 6 kHz, the viscous damping force
in the range of 6 kHz, the viscous damping force was the main energy dissipation mechanism and was the main energy
dissipation
effect of themechanism andforce
elastic/spring effectofofair
thedamping
elastic/spring force of airThese
is negligible. damping is negligible.
results were alsoThese results
verified by
were also verified by the Equation (12), since the squeeze number value
the Equation (12), since the squeeze number value for the MEMS gyroscope, at 6 kHz and room for the MEMS gyroscope, at
6 kHz and room temperature and pressure was only 0.005. Figure 11c shows
temperature and pressure was only 0.005. Figure 11c shows that with the higher values of oscillation that with the higher
values of oscillation
frequency, frequency,
the elastic/spring air the
forceselastic/spring
became dominant air forces became
and dominant
increased linearlyand increased
with linearly
the oscillation
with the oscillation frequency, while the viscous damping forces decreased.
frequency, while the viscous damping forces decreased. The FEM results presented in this section The FEM results
thus
presented in this section thus show that for the proposed MEMS gyroscope
show that for the proposed MEMS gyroscope the main damping force will be viscous air damping. the main damping force
will values
The be viscous air damping.
of squeezed filmThe
andvalues
slide of squeezed
film damping filmcoefficients
and slide film damping
obtained coefficients
through FEM obtained
analysis
through FEM analysis
−4 were 10.9 10
−5 Ns/m and 1.1 10 Ns/m,
were 10.9 × 10 Ns/m and 1.1 × 10 Ns/m, respectively. The analytically-calculated values respectively. The analytically-
of both
calculated values of both squeezed and slide film damping
−4 coefficients were
squeezed and slide film damping coefficients were 6.2 × 10 Ns/m and 1.29 × 10 Ns/m, respectively. 6.2
−5 10 Ns/m and
1.29simulated
The 10 Ns/m, value respectively.
of the squeezedThe simulated
film damping value of the squeezed
coefficient filmthan
was higher damping coefficient
the analytical was
value.
higher than the analytical value. This mismatch can be explained by the
This mismatch can be explained by the assumption that, in the analytical formulation, ideal sensingassumption that, in the
analytical formulation, ideal sensing plate edge conditions were assumed while the simulation results
plate edge conditions were assumed while the simulation results from the DampingMM module of
from the DampingMM module of CoventorWare also accounted for the non-ideality of the sensing
CoventorWare also accounted for the non-ideality of the sensing plate edges which leads to a more
plate edges which leads to a more accurate and relatively higher value of damping coefficient [37].
accurate and relatively higher value of damping coefficient [37].
Figure 11. Effect of oscillation frequency on (a) squeezed film air damping for frequency up to 6 kHz
Figure 11. Effect of oscillation frequency on (a) squeezed film air damping for frequency up to 6 kHz
(b) slide film air damping for frequency up to 6 kHz (c) squeezed film air damping for frequency up to
(b) slide film air damping for frequency up to 6 kHz (c) squeezed film air damping for frequency up
10 MHz.
to 10 MHz.
5.5. Effect of Operating Temperature and Pressure Variations on MEMS Gyroscope

As discussed in Section 4.4, both the squeezed and slide film air damping are dependent on the
effective air viscosity which is strongly influenced by the changes in the operating temperature and
pressure conditions. The effect of temperature and pressure variations on the MEMS gyroscope was
analyzed through FEM simulations
Micromachines 2020, 11, x in the DampingMM module of CoventorWare in the form of the
energy loss function (1/Q air ).of Operating

5.5. Effect Figure 12 shows
Temperature and Pressure that
Variationseffect of change in temperature in the range of
on MEMS Gyroscope
◦ ◦
−40 C to 100 C on energy
As discussed in Section 4.4, both the squeezed and slide film air damping are dependent on the
loss is negligible
effective air viscosity which is stronglyfor thebyfixed
influenced values
the changes of atmospheric
in the operating temperature and pressure at 1 kPa and
pressure conditions. The effect of temperature and pressure variations on the MEMS gyroscope was
101 kPa. However, the energy loss function values were strongly dependent
analyzed through FEM simulations in the DampingMM module of CoventorWare in the form of the on the variation in the
energy loss function (1⁄𝑄 ). Figure 12 shows that effect of change in temperature in the range of
device operating pressure−40conditions. The results show that for a fixed temperature
°C to 100 °C on energy loss is negligible for the fixed values of atmospheric pressure at 1 kPa and the value of energy
101 kPa. However, the energy loss function values were strongly dependent on the variation in the
loss function at 1 kPa wasdevice
almostoperatingtwo pressureorders of results
conditions. The magnitude
show that for aless than the
fixed temperature value
the value of at 101 kPa.
energy loss function at 1 kPa was almost two orders of magnitude less than the value at 101 kPa.
0.152 0.00144
101 kPa
0.15 0.00142
1 kPa
Loss factor at 101 kPa
Loss factor at 1 kPa

0.148
0.0014
0.146
0.00138
0.144
0.00136
0.142
0.00134
0.14
0.138 0.00132
0.136 0.0013
-40 -20 0 20 40 60 80 100
Temperature (oC)
Figure 12. Energy loss factor for MEMS gyroscope with varying operating temperature and pressure
Figure 12. Energy loss factor for MEMS gyroscope with varying operating temperature and
values.
pressure values. In addition to energy loss factor, the effect of varying temperature and pressure values on the
frequency response of the absorber mass (𝑚 ) in the drive mode and sense mass 𝑚 in the sense
mode was analyzed through FEM simulations. Figure 13a,b shows that the effect of temperature
In addition to energy loss factor, the effect of varying temperature and pressure values on the
variations, in the range of −40 °C to 100 °C, was negligible on the response amplitude of the absorber
mass both in the drive and sense direction. Figure 14a,b shows the effect of change in the operating
frequency response of the absorber mass (m ) in the drive mode and sense mass m3b in the sense mode
air pressure, for 1 kPa and 101 kPa, on the frequency response of the absorber mass in the drive
direction and sense mass in the sense3 direction. The results show that the displacement amplitude at
resonance peaks was significantly increased by decreasing the operating air pressure. However, in
was analyzed through FEM simulations. Figure 13a,b shows that the effect of temperature variations,
the flat operational region between the resonant peaks, both in the drive and sense mode, there is a
in the range of −40 ◦ C to 100 C, was negligible on the response amplitude of the absorber mass both
◦ change in the displacement amplitude of the masses. The results in Figures 13 and 14 show
negligible
that the proposed MEMS gyroscope is robust against both the operating temperature and air pressure
variations in the flat operating region between the resonant peaks in the drive and sense mode.
in the drive and sense direction. Figure 14a,b shows the effect of change in the operating air pressure,
for 1 kPa and 101 kPa, on the frequency response of the absorber mass in the drive direction and sense
mass in the sense direction. The results show that the displacement amplitude at resonance peaks
was significantly increased by decreasing the operating air pressure. However, in the flat operational
region between the resonant peaks, both in the drive and sense mode, there is a negligible change in
the displacement amplitude of the masses. The results in Figures 13 and 14 show that the proposed
MEMS gyroscope is robust against both the operating temperature and air pressure variations in the
flat operating region between the resonant peaks in the drive and sense mode.
1
100 ℃
Drive mass displacement (µm)
0.9
20 ℃
0.8 -40 ℃
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000
Frequency (Hz)
(a)
0.045
100 ℃
Sense mass displacement (µm)
0.04
20 ℃
0.035 -40 ℃
0.03
0.025
0.02
0.015
0.01
0.005
0
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000
Frequency (Hz)
(b)
Figure 13. Effect of operating temperature variations on the frequency response of (a) drive mass 𝑚
Figure 13. Effect of operating temperature
and (b) sense mass 𝑚 . variations on the frequency response of (a) drive mass m3
and (b) sense mass m3b . 1 kPa
10
9
101 kPa
8
1
0.02
Sense mass
0.015
0.01
0.005
0
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000
Frequency (Hz)
(b)
Figure 13. Effect of operating temperature variations on the frequency response of (a) drive mass 𝑚
and (b) sense mass 𝑚 .
10
1 kPa

9
101 kPa
8
0
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000
Frequency (Hz)
(a)
0.35
Sense mass displacement (µm) 1 kPa
0.3 101 kPa
0.25
0.2
0.15
0.1
0.05
0
1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000
Frequency (Hz)
(b)
Figure 14. Effect of operating air pressure variations on the frequency response of (a) drive mass 𝑚
Figure 14. Effect of operating air pressure variations on the frequency response of (a) drive mass m3
and (b) sense mass 𝑚 .
and (b) sense mass m3b .
5.6. Thermal Stability Analysis of MEMS Gyroscope

The proposed MEMS gyroscope was designed considering the SOIMUMPs microfabrication
process with silicon as a structural material. The material properties of silicon thin-film,
including elasticity, are affected by the temperature. The change in Young’s modulus of silicon
with temperature can result in variation in the resonant frequency of the MEMS gyroscope.
The semi-empirical expression for the effect of temperature on Young’s modulus of the silicon
has been reported in [38] and is given as:
T0

E(T ) = E0 − BT exp − , (16)
T
where E0 is Young’s modulus at 0 K, T is the temperature, B and T0 are constants which are dependent
on the Grueneisen parameter, Debye temperature, Anderson–Grueneisen parameter and material
volume at 0 K. For silicon, the values of B and T0 are estimated as 15.8 MPa and 317 K respectively [39].
Figure 15 shows Young’s modulus value for the silicon for the operating temperature range of −40 ◦ C to
100 ◦ C for the proposed MEMS gyroscope. The results show that, for the desired operating temperature
range for the MEMS gyroscope, the effect of temperature on stiffness variations is negligible.
In addition to material properties changing, the device operating temperature may result in the
thermal deformation in the microstructures. For the MEMS gyroscope, thermal deformation may
result in the change in the gap and planarity between the electrostatic comb drive-based actuator
and parallel plate-based capacitive sensor. To analyze the effect of the operating temperature in the
range of −40 ◦ C to 100 ◦ C on the thermal deformation in the MEMS gyroscope, an FEM-based thermal
analysis was carried out. The thermal deformation results in Figure 16a show that at 100 ◦ C there was
structural expansion in the MEMS gyroscope with a maximum value of 4.66 µm in the mechanical
suspension beams. Figure 16b shows that at −40 ◦ C, the mechanical structure contracted and maximum
deformation in the mechanical springs was 3.21 µm. Although the deformation in the mechanical
springs was high, the maximum expansion and contraction in the drive and sense mass was very low,
with a maximum value of 0.4 µm at 100 ◦ C and 0.27 µm at −40 ◦ C. In the electrostatic comb drive
actuator, attached to a mass m1 , the net effect of thermal deformation in the comb drives and hence on
the electrostatic force is negligible due to symmetric deformation of the mass m1 .
dependent on the Grueneisen parameter, Debye temperature, Anderson–Grueneisen parameter and
material volume at 0 K. For silicon, the values of B and 𝑇 are estimated as 15.8 MPa and 317 K
respectively [39]. Figure 15 shows Young’s modulus value for the silicon for the operating
temperature range of −40 °C to 100 °C for the proposed MEMS gyroscope. The results show that, for
the desired operating temperature range for the MEMS gyroscope, the effect of temperature on
stiffness variations is negligible.
169.04
Young's Modulous (GPa)

169.02
169
168.98
168.96
168.94
168.92
-40 -20 0 20 40 60 80 100
Temperature (°C)
Figure 15. Young’s modulus of silicon in the temperature range of 40 ◦ C to 100 ◦ C.
Figure
Micromachines 2020, 11, x 15. Young’s modulus of silicon in the temperature range of 40 °C to 100 °C.
In addition to material properties changing, the device operating temperature may result in the
thermal deformation in the microstructures. For the MEMS gyroscope, thermal deformation may
result in the change in the gap and planarity between the electrostatic comb drive-based actuator and
parallel plate-based capacitive sensor. To analyze the effect of the operating temperature in the range
of −40 °C to 100 °C on the thermal deformation in the MEMS gyroscope, an FEM-based thermal
analysis was carried out. The thermal deformation results in Figure 16a show that at 100 °C there was
structural expansion in the MEMS gyroscope with a maximum value of 4.66 μm in the mechanical
suspension beams. Figure 16b shows that at −40 °C, the mechanical structure contracted and
maximum deformation in the mechanical springs was 3.21 μm. Although the deformation in the
mechanical springs was high, the maximum expansion and contraction in the drive and sense mass
was very low, with a maximum value of 0.4 μm at 100 °C and 0.27 μm at −40 °C. In the electrostatic
comb drive actuator, attached to a mass 𝑚 , the net effect of thermal deformation in the comb drives
and hence on the electrostatic force is negligible due to symmetric deformation of the mass 𝑚 .
Figure 16. Structural

Figure16. Structural thermal
thermal deformation
deformation in
in MEMS
MEMSgyroscope
gyroscope(a)
(a)at 100◦°C,
at100 C, (b)
(b)at −40 ◦°C.
at−40 C.
To
To get
get more
more detailed
detailed information
information on on the
the thermal
thermal deformation
deformation and and itsits effect
effect on
on the
the capacitive
capacitive
sensing plates, aa deformation
sensing plates, deformationpath pathisisadded
addedininthe theFEM
FEM analysis
analysis along
along with
with thethe sense
sense mass
mass 𝑚 m, 3b
as,
as shown in Figure 17a. The results of the expansion and contraction along this path at −40 ◦C
shown in Figure 17a. The results of the expansion and contraction along this path at −40 °C and 100
and 100 ◦ C are shown in Figure 17b,c, respectively. The results show that the deformation due to
°C are shown in Figure 17b,c, respectively. The results show that the deformation due to temperature
temperature
did not remain didconstant
not remain constantthe
throughout throughout
length ofthethelength of theand
sense mass sense massminimized
it was and it wasatminimized
the center
at the center and increased on either side. The thermal deformation in the
and increased on either side. The thermal deformation in the x, y and z-axis may lead to change x, y and z-axis may lead in
to
the initial overlap length between sensing parallel plates, initial gap size and overlap thicknessoverlap
change in the initial overlap length between sensing parallel plates, initial gap size and values
thickness
respectively.values
Therespectively. The maximum
maximum deformation in thedeformation in the
x-axis was only 0.3x-axis
μm whichwas only
was 0.3
much which
µmless thanwas
the
much less than the initial overlap length of 450 µm between parallel sensing
initial overlap length of 450 μm between parallel sensing plates. In the z-axis, thermal deformation plates. In the z-axis,
thermal
was onlydeformation
0.002 μm and waswasonlynegligible
0.002 µm to andthewas negligible
initial overlapto thickness
the initial value
overlap of thickness value of 25
25 μm. However, in
µm. However, in the sense y-axis, the deformation the mass m was 0.165 µm and 0.239 µm at −40 ◦C
the sense y-axis, the deformation the mass 𝑚 was 0.165 μm and 0.239 μm at −40 °C and 100 °C,
3b
and 100 ◦ C, respectively.
respectively. This results in This
an results in initial
effective an effective
air gapinitial air gap
of nearly 2.84ofμmnearly
and 2.84 μm and
2.76 µm 2.76 µm
between the
between
parallel the parallel sensing
capacitive capacitive sensing
plates. plates.
This This decrease
decrease in thegap
in the initial initial gap results
results in theindecrease
the decrease in
in the
the maximum pull-in voltage value and maximum input angular
maximum pull-in voltage value and maximum input angular rate measurement. rate measurement.
respectively. The maximum deformation in the x-axis was only 0.3 μm which was much less than the
initial overlap length of 450 μm between parallel sensing plates. In the z-axis, thermal deformation
was only 0.002 μm and was negligible to the initial overlap thickness value of 25 μm. However, in
the sense y-axis, the deformation the mass 𝑚 was 0.165 μm and 0.239 μm at −40 °C and 100 °C,
respectively. This results in an effective initial air gap of nearly 2.84 μm and 2.76 μm between the
Micromachines 2020, 11,
parallel 862
capacitive sensing plates. This decrease in the initial gap results in the decrease in the 19 of 26
maximum pull-in voltage value and maximum input angular rate measurement.
(a) Thermal
Figure 17. Figure 17. (a) deformation analysis
Thermal deformation path along
analysis with with
path along the sense
the sense m3b 𝑚
massmass (b) thermal deformation
(b) thermal
at −40 ◦ C. deformation
(c) thermalatdeformation
−40 °C. (c) thermal deformation
at 100 ◦ C. at 100 °C.
5.7. Analysis of Fabrication Process Tolerances on Frequency Response

In addition to the
Micromachines environmental
2020, 11, x variations, the microfabrication process tolerances may also
contribute towards uncertainty in the MEMS gyroscope performance. The SOIMUMPs is a relatively
5.7. Analysis of Fabrication Process Tolerances on Frequency Response
mature microfabrication process with tolerances of ±1 µm for the device layer and ±5 µm for the
In addition to the environmental variations, the microfabrication process tolerances may also
substrate layer. Thetowards
contribute main uncertainty
effect of indevice
the MEMS layer thickness
gyroscope is on The
performance. theSOIMUMPs
resonantis frequency
a relatively and sense
capacitance change due to stiffness variation of mechanical springs and overlap for
mature microfabrication process with tolerances of ±1 μm for the device layer and ±5 μm area
the between the
substrate layer. The main effect of device layer thickness is on the resonant frequency and sense
sensing parallel plates. The effect of change in structural layer thickness with 25 µm ± 1 µm was studied
capacitance change due to stiffness variation of mechanical springs and overlap area between the
and its effect sensing
on the parallel
resonant frequency
plates. The effectchange
of changeand capacitance
in structural change was
layer thickness analyzed.
with 25 μm ± 1 μmFigure
was 18a shows
that the effect of change in resonant frequency due to thickness tolerance of ±1 µm resulted in nearly
studied and its effect on the resonant frequency change and capacitance change was analyzed. Figure
2% change in18a shows that the effect of change in resonant frequency due to thickness tolerance of ±1 μm resulted
resonant frequency from the nominal values. Figure 18b shows that thickness tolerances
in nearly 2 % change in resonant frequency from the nominal values. Figure 18b shows that thickness
result in a capacitance
tolerances resultchange variation
in a capacitance of nearly
change 3.1%
variation from
of nearly 3.1the nominal
% from values.
the nominal These
values. Theseresults show
that the proposed
results MEMS
show thatgyroscope
the proposedisMEMS
robust againstisthe
gyroscope microfabrication
robust process tolerances.
against the microfabrication process
tolerances.
2
Mode 1
Resonant frequency change (%)
1.5 Mode 2
Mode 6
1
Mode 7
0.5
0
-0.5
-1
-1.5
-2
24 24.25 24.5 24.75 25 25.25 25.5 25.75 26
(a)
4
3
Capacitacne change (%)
2
1
0
-1
-2
-3
-4
24 24.25 24.5 24.75 25 25.25 25.5 25.75 26
Structural layer thickness (µm)
(b)
Figure 18. Effect of fabrication process tolerances on (a) resonant frequencies (b) capacitance change.
Figure 18. Effect of fabrication process tolerances on (a) resonant frequencies (b) capacitance change.
6. Sensitivity Analysis of MEMS Gyroscope
6.1. Mechanical Sensitivity Analysis

As discussed in Section 2, the working of the proposed 3-DoF drive and 2-DoF sense mode
MEMS gyroscope was based on the Coriolis effect. An applied actuation voltage to the comb drive-
based electrostatic actuator results in the oscillation of the mass 𝑚 in the x-axis (𝑥 ). For an input
rotation in the z-axis Ω, a Coriolis force 𝐹 = 2𝑚 Ω𝑥 was generated in the y-axis. This force results
in a dynamically-amplified displacement (𝑦 ) in the sense mass 𝑚 . The proposed MEMS
gyroscope was designed to be operated in the flat operating region between the first two resonant
6. Sensitivity Analysis of MEMS Gyroscope
6.1. Mechanical Sensitivity Analysis

As discussed in Section 2, the working of the proposed 3-DoF drive and 2-DoF sense mode MEMS
gyroscope was based on the Coriolis effect. An applied actuation voltage to the comb drive-based
electrostatic actuator results in the oscillation of the mass m3 in the x-axis (x3 ). For an input rotation
.
in the z-axis Ω, a Coriolis force Fc = 2m3a Ωx3 was generated in the y-axis. This force results in a
dynamically-amplified displacement ( y2 ) in the sense mass m3b . The proposed MEMS gyroscope was
designed to be operated in the flat operating region between the first two resonant peaks of the drive
and sense mode. To obtain the mechanical sensitivity for the MEMS gyroscope, transient analysis
was carried out in the CoventorWare MEMS+ module. An actuation voltage of 80 V DC and 5 V AC
was applied to the electrostatic actuator with a frequency of 2.5 kHz, which results in an oscillation
amplitude of 0.22 µm in the mass m3 . For this oscillation amplitude and with a rotation of 50 rad/s in
the z-axis, a Coriolis force of 1.943 µN was generated in the y-axis Figure 19a,b shows the input rotation
and drive mass displacement, respectively. Figure 19c shows the y-axis displacement in the mass
m3b corresponding
Micromachines to an
2020, 11,induced
x Coriolis force with an amplitude of 0.114 µm. Assuming the linear
relation between the input rotation and induced Coriolis force and hence sense mass displacement,
displacement in the mass 𝑚 corresponding to an induced Coriolis force with an amplitude of 0.114
the mechanical sensitivity
μm. Assuming for the
the linear proposed
relation betweenMEMS
the inputgyroscope
rotation andwas 0.00228
induced µm/rad/s.
Coriolis The analytical
force and hence
value of the oscillation
sense amplitude
mass displacement, of mass msensitivity
the mechanical 3 in the drive
for the direction
proposed was
MEMS 0.26 µm.
gyroscope The
was corresponding
0.00228
magnitudeμm/rad/s. The analytical
of the Coriolis force,value
in theof y-axis
the oscillation
for an amplitude of massof𝑚50in
input rotation the drive
rad/s was direction was The y-axis
2.2963 µN.
0.26 μm. The corresponding magnitude of the Coriolis force, in the y-axis for an input rotation of 50
displacement in mass m μN.due
rad/s was 2.2963 3b
to this Coriolis force was 0.1347 µm and the mechanical sensitivity was
The y-axis displacement in mass 𝑚 due to this Coriolis force was 0.1347 μm
0.002694 µm/rad/s.
and the mechanical sensitivity was 0.002694 μm/rad/s.
60
Angular rate (rad/s)
50
40
30
20
10
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
(a)
0.4
0.3
displacement (µm)
0.2
Drive mass
0.1
0
-0.1
-0.2
-0.3
-0.4
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
(b)
0.14
0.12
displacement (µm)
0.1
Sense mass
0.08
0.06
0.04
0.02
0
-0.02
-0.04
-0.06
-0.08
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Time (s)
(c)
(a) Input
Figure 19.Figure angular
19. (a) Input velocity
angular velocityof
of 50 rad/s.
50 rad/s. (b) (b) Oscillation
Oscillation responseresponse the 𝑚
ofmass
of the drive . (c) mass m3 .
drive
Oscillation
(c) Oscillation response
response of the
of the sense mass 𝑚
sensemass m3b. .
6.2. Integration of MEMS Gyroscope with Readout Electronics and Voltage Sensitivity Analysis
The voltage sensitivity of the MEMS gyroscope is not only dependent on the mechanical design,
but also on the readout electronics used for the conversion of capacitance change, corresponding to
an input rotation, to an equivalent output voltage. For the proposed MEMS gyroscope, a
commercially available capacitance to voltage conversion Universal Readout ICTM MS3110 is
considered [40]. This IC is vastly used for signal conditioning of MEMS devices and provides an
output voltage proportional to the change in differential capacitance. Figure 20 shows the integration
of the proposed MEMS gyroscope with the internal circuitry of the MS 3110 IC in Matlab Simulink
environment, which mentions the input and output ports of the MEMS gyroscope model, a
6.2. Integration of MEMS Gyroscope with Readout Electronics and Voltage Sensitivity Analysis
The voltage sensitivity of the MEMS gyroscope is not only dependent on the mechanical design,
but also on the readout electronics used for the conversion of capacitance change, corresponding to an
input rotation, to an equivalent output voltage. For the proposed MEMS gyroscope, a commercially
available capacitance to voltage conversion Universal Readout ICTM MS3110 is considered [40]. This IC
is vastly used for signal conditioning of MEMS devices and provides an output voltage proportional to
the change in differential capacitance. Figure 20 shows the integration of the proposed MEMS gyroscope
with the internal circuitry of the MS 3110 IC in Matlab Simulink environment, which mentions the
input and output ports of the MEMS gyroscope model, a combination of actuation voltages in the drive
mode and a DC bias voltage applied to the sensing parallel plates. The differential output capacitances
from the sensing parallel plates are shown as Cap1 and Cap2 and are connected as inputs (CS1IN and
CS2IN ) to the MS 3110. The internal circuitry of the MS 3110 IC is shown as a simplified block diagram.
The circuit contains a charge amplifier followed by a sample and hold circuit. A low pass filter is used
forMicromachines
the signal2020, 11, x
conditioning and finally, the signal is passed through a buffer amplifier, which has offset
trim functionality along with the output level selection. The MS 3110 IC has multiple programmable
followed by a sample and hold circuit. A low pass filter is used for the signal conditioning and finally,
attributes, including the gain of the charge amplifier, the cut-off frequency of the filter and the buffer
the signal is passed through a buffer amplifier, which has offset trim functionality along with the
amplifier gain. The MSThe
output level selection. 3110
MScan3110operate in two modes
IC has multiple with either
programmable a single-ended
attributes, including thesense capacitor
gain of
input or a differential
the charge input.
amplifier, the Infrequency
cut-off differential mode,
of the filterthe
andoutput capacitance
the buffer fromThe
amplifier gain. theMS
MEMS gyro Cap1
3110 can
and Cap2 may
operate in twohave
modes mismatched values
with either and may sense
a single-ended lead to a DC offset
capacitor inputinorthe output values.
a differential input.InInthis case,
thedifferential mode, theCS1
internal capacitors output capacitance
and CS2 fromare
of MS3110 the MEMS to
adjusted gyro 𝐶𝑎𝑝1theand
reduce DC𝐶𝑎𝑝2offsetmay have
by balancing the
mismatched values and may lead to a DC offset in the output values. In this
common mode capacitance. The output voltage of MS3110 IC can range between 0.5 V to 4 V and is case, the internal
capacitors
given as: 𝐶𝑆1 and 𝐶𝑆2 of MS3110 are adjusted to reduce the DC offset by balancing the common
mode capacitance. The output voltage of MS3110 IC 2∆C can range
+ (CS2between
− CS10.5) V to 4 V and is given as:
Vo = 1.14 × G × V2P25 × + Vre f , (17)
2Δ𝐶 + (𝐶𝑆2C−F 𝐶𝑆1)
𝑉 = 1.14 𝐺 𝑉2𝑃25 +𝑉 , (17)
𝐶
where G is the gain with a value of 2, V2P25 and Vre f are set to 2.25 V, ∆C is the change in capacitance
and CF is𝐺the
where is the gain with
feedback capacitor of 2,a𝑉2𝑃25
a valuewith nominal 𝑉
andvalue are set pF.
of 2.8 to 2.25 Δ𝐶 isofthe
TheV,value CFchange
can beinchanged
capacitance and 𝐶 is the feedback capacitor with a nominal value of 2.8 𝑝𝐹. The
according to the sense capacitance range to make the output voltage reside within the desired value of 𝐶 can be range
of 0.5 V to 4 V. Figure 21 shows the effect of variation in the CF on the output voltage oftheMS 3110
changed according to the sense capacitance range to make the output voltage reside within
desired range of 0.5 V to 4 V. Figure 21 shows the effect of variation in the 𝐶 on the output voltage
IC for the proposed MEMS gyroscope. The graph shows that for the MEMS gyroscope, for smaller
of MS 3110 IC for the proposed MEMS gyroscope. The graph shows that for the MEMS gyroscope,
values of CF , the output voltage is higher than the upper limit of 4 V for the same input capacitance
for smaller values of 𝐶 , the output voltage is higher than the upper limit of 4 V for the same input
change. Moreover,
capacitance change.ifMoreover,
the capacitance change is
if the capacitance too large
change is toothen
largethe
thenoutput voltage
the output may
voltage mayalso
alsoincrease
beyond thebeyond
increase maximum value of 4value
the maximum V, soofhigher values
4 V, so CF must
higherofvalues of 𝐶be must
selected to keeptothe
be selected voltage
keep the in the
acceptable
voltage inrange.
the acceptable range.
Figure
Figure 20. Integrationofofthe
20. Integration the MEMS
MEMS gyroscope
gyroscope model
model with with the readout
the readout electronics.
electronics.
5.5
tput Voltage (V)
4.5
3.5
Figure 20. Integration of the MEMS gyroscope model with the readout electronics.
5.5
Output Voltage (V)

5
4.5
3.5
2.5
2
0 1 2 3 4 5 6 7 8 9
Value of feedback capacitor (pF)

Micromachines
Figure2020,
21.11, x
Effect of variation in the feedback
feedback capacitor
capacitor on
on output
output voltage
voltage for
for MEMS
MEMS gyroscope.
gyroscope.
Figure 22
Figure 22 shows
shows the
the output
output capacitance
capacitance response
response ofof the
the proposed
proposed MEMSMEMS gyroscope
gyroscope for for an
an input
input
rotation of
rotation of 50 rad/s. The output capacitance 𝐶 − in Figure 22a shows the decrease
output capacitance C− in Figure 22a shows the decrease in the capacitance in the capacitance
on one
on one side
side of
of the
the differential
differential parallel
parallel sensing plates of MEMS gyroscope with a value of 3.64 3.64 pF.
Figure 22b shows
Figure 22b shows the increase in capacitance 𝐶 + from rest value for on the other side of the
C+ from rest value for on the other side of the parallel parallel
sensingplates,
sensing plates,with
withaavalue
valueofof3.795
3.795pF.pF. These
These results
results show
show that
that thethe
netnet capacitance
capacitance (∆C(Δ𝐶)
change
change for
) for the
the proposed
proposed MEMS MEMS gyroscope
gyroscope is 155isfF 155
forfFanforinput
an input rotation
rotation of 50 of 50 rad/s
rad/s whichwhich is in close
is in close agreement
agreement with
with
an an analytically
analytically calculated
calculated value of value
136 of
fF 136
using fFEquation
using Equation (7). Assuming
(7). Assuming a lineararelation
linear relation
betweenbetween
the net
the net capacitance
capacitance change,
change, the the capacitive
capacitive (∆C/Ω) of(Δ𝐶/Ω)
sensitivitysensitivity of the MEMS
the proposed proposed MEMS isgyroscope
gyroscope 3.1 fF/rad/sis
3.1 fF/rad/s and the corresponding output voltage of the
and the corresponding output voltage of the MS3110 IC is 11.36 mV/rad/s.MS3110 IC is 11.36 mV/rad/s.
3.77
Output capacitance C-
3.75
3.73
3.71
(pF)
3.69
3.67
3.65
3.63
3.61
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
(a)
3.83
Output capacitance C+
3.81
3.79
3.77
3.75
(pF)
3.73
3.71
3.69
3.67
3.65
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Time (s)
(b)
Figure 22.
Figure (a)Output
22. (a) Outputsense
sensecapacitance C−.
capacitance 𝐶 −. (b)
(b) Output
Output sense capacitance C𝐶++.
sense capacitance .
7. Discussion
One of the most important performance parameters for MEMS devices is their reliability. The
MEMS gyroscopes are supposed to work in harsh environments and under a wide range of
temperature variations, usually in the range of −40 °C to 100 °C, while under constant oscillations. In
addition to the structural design, the selection of material for the fabrication process of MEMS
gyroscopes is very important, since reliability issues like fatigue limit under cyclic loading and
7. Discussion
One of the most important performance parameters for MEMS devices is their reliability.
The MEMS gyroscopes are supposed to work in harsh environments and under a wide range
of temperature variations, usually in the range of −40 ◦ C to 100 ◦ C, while under constant oscillations.
In addition to the structural design, the selection of material for the fabrication process of MEMS
gyroscopes is very important, since reliability issues like fatigue limit under cyclic loading and stiffness
degradation under the combined effect of loading and temperature (creep effect) are strongly dependent
on the thin film material and its thickness. Most of the vibratory multi-DoF MEMS gyroscopes presented
in the literature (given in Table 1) are fabricated either using silicon or Nickel thin films as a structural
material. In comparison to Nickel, the silicon has much better mechanical properties and thermal
stability. Thus, the selection of silicon as the structural material for our MEMS gyroscope gives us an
added advantage of mechanical reliability and thermal stability. The complex structural design of
multi-DoF vibratory MEMS gyroscopes, with multiple masses nested together, requires high accuracy
during the microfabrication. Most of the silicon-based multi-DoF MEMS gyroscope designs presented
in the literature are fabricated using in-house microfabrication processes. The costs involved in the
development of such microfabrication facilities are very high. We believe that this is the reason that
the research work on the development of non-resonant multi-DoF MEMS gyroscopes presented in
literature is very limited.
The commercial foundry-based microfabrication of multi-DoF MEMS gyroscopes, based on
the multiuser process concept, is a relatively low-cost solution for MEMS researchers. However,
since the design rules offered by multiuser foundry services are fixed, the design must meet the
fabrication process constraints and at the same time achieve the desired performance parameters.
The multi-project wafer-based, low-cost and commercially available SOIMUMPs microfabrication
process, offered by MEMSCAP Inc. USA, allows the use of silicon as a structural material with 25 µm
thickness, which meets the basic requirements for MEMS gyroscope of high mass, mechanical reliability
and thermal stability. The design rules of SOIMUMPs process for the design of multi-mass-based
multi-DoF MEMS gyroscopes offer additional challenges, as discussed in Section 3. In this paper,
we have focused on the design of multi-DoF MEMS gyroscope that not only meets the design rules of
the fabrication process but also achieves better performance.
The performance characteristics of capacitive MEMS gyroscopes including voltage sensitivity,
dynamic range and noise floor are dependent√ on the sense electronics parameters like resolution.
The MS3110 has a resolution of 4 aF/ Hz and the minimum measurable angular rate for a MEMS
gyroscope is dependent on this resolution. The resolution for the proposed MEMS gyroscope is
estimated to be 0.0519 rad/s using the capacitive sensitivity and sense electronics resolution. The total
noise equivalent in MEMS gyroscope is also dependent on the resolution of the sensing electronics.
The total noise equivalent rate for the MEMS gyroscope at room√temperature and atmospheric air
pressure and with MS3110 readout electronics IC is 0.00328 rad/s/
√ Hz with electrical√noise equivalent
rate and mechanical noise equivalent rate of 0.00129 rad/s/ Hz and 0.00302 rad/s/ Hz respectively.
The dynamic range of the proposed MEMS gyroscope is estimated to be 69.2 dB. These values of noise
equivalent and dynamic range for the proposed MEMS gyroscope are obtained by using the analytical
expressions discussed in [41,42]. Table 4 shows the comparison of the proposed 3-DoF drive and 2-DoF
sense mode gyroscope with the multi-DoF MEMS gyroscopes presented in the literature. In addition to
the inherent advantages of being designed using a commercially available process, robustness against
environmental variations and thermal stability the proposed MEMS gyroscope also exhibits better
sensitivity and increased operational bandwidth in comparison to previous designs.
Table 4. Comparison of the proposed multi-DoF MEMS gyroscope design with the literature.
Structural
Layer
References Configuration Device Size Sensitivity Bandwidth
Material Thickness (µm)
Schofield et al. 1 Dof drive,
Silicon 75 - 2.3 µV/(◦ /s) 600 Hz
[14] 2 DoF sense
1 Dof drive,
Trusov et al. [15] Silicon 50 3 × 3 mm2 56 µV/(◦ /s) 250 Hz
2 DoF sense
1 Dof drive,
Acar et al. [21] Polysilicon 100 4 × 4 mm2 0.0303 mV/(◦ /s) 50 Hz
2 DoF sense
3 Dof drive,
Saqib et al. [18] Nickle 20 3.2 × 3 mm2 0.565 fF/rad/s 1.2 kHz
1 DoF sense
3.1 fF/rad/s, 198.9
3 Dof drive,
This work Silicon 25 4.2 × 4.2mm2 µV/(◦ /s), 0.04 × 10−3 1.62 kHz
2 DoF sense
µm/(◦ /s)
8. Conclusions
In this paper, we have presented a design of a 3-DoF drive mode and 2-DoF sense mode vibratory
MEMS gyroscope considering the design rules of commercially available multi-user foundry process
SOIMUMPs. The operational bandwidth of the proposed MEMS gyroscope is 1.62 kHz with a dynamic
amplification ratio of 2 and 2.6 in the drive and sense direction respectively. The results showed that
the effect of both squeezed and slide film air damping under varying operating temperature and
pressure conditions is negligible in the operational bandwidth of gyroscope. The effect of temperature
in the range of −40 ◦ C to 100 ◦ C on the stiffness variation of mechanical springs and hence on the
dynamic response is also negligible for the MEMS gyroscope. The FEM-based structural thermal
deformation analysis showed that the net effect of thermal deformation in the comb drives and hence
on the electrostatic force is negligible due to symmetric deformation of the mass. In the sensing
parallel plates, the thermal deformation is also symmetric and results in nearly 5% change in the
initial air gap. The system modeling-based results showed that the proposed MEMS gyroscope has
capacitive and voltage sensitivity of 3.1 fF/rad/s and 198.9 µV/(◦ /s) respectively when integrated
with the commercially available capacitance to voltage conversion circuit. The dynamic range of the
proposed MEMS gyroscope is estimated to be 69.2 dB with minimum and maximum angular rate
of 0.0519 rad/s and 150 rad/s respectively. The total noise√equivalent rate for the MEMS gyroscope
considering sense electronics resolution is 0.00328 rad/s/ Hz.
Author Contributions: Conceptualization, M.M.S.; methodology, S.A.R.B., M.M.S.; software, U.S.K., A.H.;
validation, S.A.R.B., R.I.S.; resources, J.I.; writing—original draft preparation, S.A.R.B., M.M.S., U.S.K.; supervision,
M.M.S.; funding acquisition, M.M.S., R.I.S., A.H. All authors have read and agreed to the published version of
the manuscript.
Funding: This research was funded by the Higher Education Commission (HEC), Pakistan, grant number
TDF-02-065 and The APC was funded by the National University of Sciences and Technology(NUST), Pakistan.
Acknowledgments: The authors would like to acknowledge the National Centre of Robotics and Automation
(NCRA) for their support.
Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the
study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to
publish the results.
References
1. Wu, G.Q.; Chua, G.L.; Gu, Y.D. A dual-mass fully decoupled MEMS gyroscope with wide bandwidth and
high linearity. Sens. Actuators A Phys. 2017, 259, 50–56. [CrossRef]
2. Shakoor, R.I.; Marc, B.; Iqbal, S. Experimental evaluation of resonant frequencies with associated mode shapes
and power analysis of thermally actuated vibratory microgyroscope. Microsyst. Technol. 2018, 24, 3601–3613.
[CrossRef]
3. Yang, J.S.; Fang, H.Y. A new ceramic tube piezoelectric gyroscope. Sens. Actuators A Phys. 2003, 107, 42–49.
[CrossRef]
4. Dellea, S.; Giacci, F.; Longoni, A.F.; Langfelder, G. In-plane and out-of-plane MEMS gyroscopes based on
piezoresistive NEMS detection. J. Microelectromech. Syst. 2015, 24, 1817–1826. [CrossRef]
5. Bochobza-Degani, O.; Seter, D.J.; Socher, E.; Nemorivsky, Y. A novel micromachined vibrating rate-gyroscope
with optical sensing and electrostatic actuation. Sens. Actuators A Phys. 2000, 83, 54–60. [CrossRef]
6. Chuang, W.C.; Lee, H.L.; Chang, P.Z.; Hu, Y.C. Review on the modeling of electrostatic MEMS. Sensors
2010, 10, 6149–6171. [CrossRef]
7. Nguyen, M.N.; Ha, N.S.; Nguyen, L.Q.; Chu, H.M.; Vu, H.N. Z-axis micromachined tuning fork gyroscope
with low air damping. Micromachines 2017, 8, 42. [CrossRef]
8. Xu, Q.; Hou, Z.; Kuang, Y.; Miao, T.; Ou, F.; Zhuo, M.; Xiao, D.; Wu, X. A Tuning Fork Gyroscope with a
Polygon-Shaped Vibration Beam. Micromachines 2019, 10, 813. [CrossRef]
9. He, C.; Zhao, Q.; Huang, Q.; Liu, D.; Yang, Z.; Yan, G. A MEMS vibratory gyroscope with real-time
mode-matching and robust control for the sense mode. IEEE Sens. J. 2014, 15, 2069–2077. [CrossRef]
10. Balachandran, G.K.; Petkov, V.P.; Mayer, T.; Balslink, T. A 3-axis gyroscope for electronic stability control
with continuous self-test. IEEE J. Solid-State Circ. 2015, 51, 177–186.
11. Jia, J.; Ding, X.; Gao, Y.; Li, H. Automatic Frequency tuning technology for dual-mass MEMS gyroscope
based on a quadrature modulation signal. Micromachines 2018, 9, 511. [CrossRef] [PubMed]
12. Dyck, C.W.; Allen, J.J.; Huber, R.J. Parallel-plate electrostatic dual-mass resonator. In Micromachined Devices
and Components V; International Society for Optics and Photonics: Santa Clara, CA, USA, 1999; Volume 3876.
13. Acar, C.; Shkel, A.M. Nonresonant micromachined gyroscopes with structural mode-decoupling. IEEE Sens. J.
2003, 3, 497–506. [CrossRef]
14. Schofield, A.R.; Trusov, A.A.; Shkel, A.M. Effects of operational frequency scaling in multi-degree of freedom
MEMS gyroscopes. IEEE Sens. J. 2008, 8, 1672–1680. [CrossRef]
15. Trusov, A.A.; Schofield, A.R.; Shkel, A.M. Performance characterization of a new temperature-robust
gain-bandwidth improved MEMS gyroscope operated in air. Sens. Actuators A Phys. 2009, 155, 16–22.
[CrossRef]
16. Sahin, K.; Sahin, E.; Alper, S.E.; Akin, T. A wide-bandwidth and high-sensitivity robust microgyroscope.
J. Micromech. Microeng. 2009, 19, 074004. [CrossRef]
17. Erismis, M. Design and modeling of a new robust multi-mass coupled-resonator family with dynamic motion
amplification. Microsyst. Technol. 2013, 19, 1105–1110. [CrossRef]
18. Saqib, M.; Mubasher Saleem, M.; Mazhar, N.; Awan, S.U.; Shahbaz Khan, U. Design and analysis of a
high-gain and robust multi-DOF electro-thermally actuated MEMS gyroscope. Micromachines 2018, 9, 577.
[CrossRef]
19. Saleem, M.M.; Bazaz, S.A. Design and robustness analysis of structurally decoupled 3-DoF MEMS gyroscope
in the presence of worst-case process tolerances. Microsyst. Technol. 2011, 17, 1381–1391. [CrossRef]
20. Verma, P.; Khan, K.Z.; Khonina, S.N.; Kazanskiy, N.L.; Gopal, R. Ultraviolet-LIGA-based fabrication
and characterization of a nonresonant drive-mode vibratory gyro/accelerometer. J. Micro-Nanolithogr.
MEMS MOEMS 2016, 15, 035001. [CrossRef]
21. Acar, C.; Shkel, A.M. Inherently robust micromachined gyroscopes with 2-DOF sense-mode oscillator.
J. Microelectromech. Syst. 2006, 15, 380–387. [CrossRef]
22. Riaz, K.; Bazaz, S.A.; Saleem, M.M.; Shakoor, R.I. Design, damping estimation and experimental
characterization of decoupled 3-DoF robust MEMS gyroscope. Sens. Actuators A Phys. 2011, 172, 523–532.
[CrossRef]
23. Verma, P.; Shekhar, C.; Arya, S.K.; Gopal, R. New design architecture of a 3-DOF vibratory gyroscope with
robust drive operation mode and implementation. Microsyst. Technol. 2015, 21, 2175–2185. [CrossRef]
24. Zhu, X.H.; Chu, H.J.; Shi, Q.; Qiu, A.P.; Su, Y. Experimental study of compensation for the effect of temperature
on a silicon micromachined gyroscope. Proc. Inst. Mech. Eng. Part N J. Nanoeng. Nanosyst. 2008, 222, 49–55.
[CrossRef]
25. Guo, Z.; Fu, P.; Liu, D.; Huang, M. Design and FEM simulation for a novel resonant silicon MEMS gyroscope
with temperature compensation function. Microsyst. Technol. 2018, 24, 1453–1459. [CrossRef]
26. Cui, M.; Huang, Y.; Wang, W.; Cao, H. MEMS Gyroscope Temperature Compensation Based on Drive Mode
Vibration Characteristic Control. Micromachines 2019, 10, 248. [CrossRef] [PubMed]
27. Dwivedi, A.; Khanna, G. Numerical simulation and modelling of a novel MEMS capacitive accelerometer
based microphone for fully implantable hearing aid. Microsyst. Technol. 2019, 25, 399–411. [CrossRef]
28. Cowen, A.; Hames, G.; Monk, D.; Wilcenski, S.; Hardy, B. SOIMUMPs Design Handbook (Revision 8.0.).
Available online: http://www.memscap.com (accessed on 27 June 2020).
29. Edalatfar, F.; Yaghootkar, B.; Qureshi, A.Q.A.; Azimi, S.; Bahreyni, B. Design, fabrication and characterization
of a high performance MEMS accelerometer. IEEE Sens. J. 2016, 1–3. [CrossRef]
30. Bao, M.; Yang, H. Squeeze film air damping in MEMS. Sens. Actuators A Phys. 2007, 136, 3–27. [CrossRef]
31. Abdolvand, R.; Amini, B.V.; Ayazi, F. Sub-micro-gravity in-plane accelerometers with reduced capacitive
gaps and extra seismic mass. J. Microelectromech. Syst. 2007, 16, 1036–1043. [CrossRef]
32. Mol, L.; Rocha, L.A.; Cretu, E.; Wolffenbuttel, R.F. Squeezed film damping measurements on a parallel-plate
MEMS in the free molecule regime. J. Micromech. Microeng. 2009, 19, 074021. [CrossRef]
33. Syed, W.U.; Brimmo, A.; Waheed, O.; Bojesomo, A.; Ali, M.H.; Ocak, I.; Elfadel, I.A.M. Numerical modeling
and validation of squeezed-film damping in vacuum-packaged industrial MEMS. J. Micromech. Microeng.
2017, 27, 075016. [CrossRef]
34. Veijola, T.; Raback, P. Methods for solving gas damping problems in perforated microstructures using a 2D
finite-element solver. Sensors 2007, 7, 1069–1090. [CrossRef]
35. Morris, C.J.; Forster, F.K. Oscillatory flow in microchannels. Exp. Fluids 2004, 36, 928–937. [CrossRef]
36. Somà, A.; Saleem, M.M.; De Pasquale, G. Effect of creep in RF MEMS static and dynamic behavior.
Microsyst. Technol. 2016, 22, 1067–1078. [CrossRef]
37. CoventorWare Analyzer Field Solver Reference; Coventor Inc.: Cary, NC, USA, 2018.
38. Wachtman, J.B., Jr.; Tefft, W.E.; Lam, D.G., Jr.; Apstein, C.S. Exponential temperature dependence of Young’s
modulus for several oxides. Phys. Rev. 1961, 122, 1754. [CrossRef]
39. Gysin, U.; Rast, S.; Ruff, P.; Meyer, E.; Lee, D.W.; Vettiger, P.; Gerber, C. Temperature dependence of the force
sensitivity of silicon cantilevers. Phys. Rev. B 2004, 69, 045403. [CrossRef]
40. Datasheet MUCRI. MS3110 Universal Capacitive ReadoutTM IC; MicroSensors, Inc.: Costa Mesa, CA, USA, 2004.
41. Bao, M.H. Micro Mechanical Transducers: Pressure Sensors, Accelerometers and Gyroscopes; Elsevier:
San Diego, CA, USA, 2000.
42. Apostolyuk, V. Theory and Design of Micromechanical Vibratory Gyroscopes in MEMS/NEMS; Springer:
Boston, MA, USA, 2006; pp. 173–195.
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Micromachines 11 00862

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Micromachines 11 00862

Uploaded by

Copyright:

Available Formats

micromachines

Abstract: This paper presents microfabrication process-driven design of a multi-degree of

Keywords: MEMS gyroscope; multi-degree of freedom (multi-DoF); non-resonant; microfabrication;

Micromachines 2020, 11, 862; doi:10.3390/mi11090862 www.mdpi.com/journal/micromachines

measurement of displacement of the proof mass corresponding to an input rotation. In MEMS

Table 1. Multi-degree of freedom (multi-DoF) non-resonant microelectromechanical systems (MEMS)

2. Structural Design of Multi-Degree of Freedom (Multi-DoF) Microelectromechanical Systems

Table 2. Design parameters of the proposed multi-DoF MEMS gyroscope.

Table 2. Design parameters of the proposed multi-DoF MEMS gyroscope.

3.1. Microfabrication Process Limitations

3.2. Process Driven Design Modifications

4. Analytical Modeling of the Multi-DoF MEMS Gyroscope

4.4. Air Damping Analysis

5.2. Pull-in Voltage Analysis

5.5. Effect of Operating Temperature and Pressure Variations on MEMS Gyroscope

energy loss function (1/Q air ).of Operating

Loss factor at 1 kPa

Drive mass displacement (µm)

5.6. Thermal Stability Analysis of MEMS Gyroscope

Young's Modulous (GPa)

Figure 16. Structural

5.7. Analysis of Fabrication Process Tolerances on Frequency Response

6.1. Mechanical Sensitivity Analysis

6. Sensitivity Analysis of MEMS Gyroscope

6.1. Mechanical Sensitivity Analysis

Output Voltage (V)

Value of feedback capacitor (pF)

You might also like