micromachines Article
Z-Axis Micromachined Tuning Fork Gyroscope with Low Air Damping Minh Ngoc Nguyen 1,2 , Nhat Sinh Ha 1 , Long Quang Nguyen 1 , Hoang Manh Chu 1, * and Hung Ngoc Vu 1, * 1
2
*
International Training Institute for Materials Science, Hanoi University of Science and Technology, Hanoi 100000, Vietnam;
[email protected] (M.N.N.);
[email protected] (N.S.H.);
[email protected] (L.Q.N.) Faculty of Electronic and Electrical Engineering, Hung Yen University of Technology and Education, Hung Yen 160000, Vietnam Correspondence:
[email protected] (H.M.C.);
[email protected] (H.N.V.); Tel.: +84-4-3868-0787 (H.M.C. & H.N.V.)
Academic Editor: Hiroshi Toshiyoshi Received: 14 December 2016; Accepted: 20 January 2017; Published: 1 February 2017
Abstract: This paper reports on the design and fabrication of a z-axis tuning fork gyroscope which has a freestanding architecture. In order to improve the performance of the tuning fork gyroscope by eliminating the influence of the squeeze-film air damping, the driving and sensing parts of the gyroscope were designed to oscillate in-plane. Furthermore, by removing the substrate underneath the device, the slide-film air damping in the gap between the proof masses and the substrate was eliminated. The proposed architecture was analyzed by the finite element method using ANSYS software. The simulated frequencies of the driving and sensing modes were 9.788 and 9.761 kHz, respectively. The gyroscope was fabricated using bulk micromachining technology. The quality factor and sensitivity of the gyroscope operating in atmospheric conditions were measured to be 111.2 and 11.56 mV/◦ /s, respectively. Keywords: Bulk micromachining; Tuning fork gyroscope; Finite element analysis
1. Introduction Micro-gyroscopes have the advantages of small size, low cost, and suitability for mass production. The micro-gyroscopes can be used in the automotive industry to make anti-rollover systems, antiskid controls, and electronic stability controls [1]. Various applications for consumer equipment range from camera stabilization, cell phone stabilization, virtual reality, and inertial mouse to navigation for portable electronics [2]. The micro-gyroscopes have further applications in robotics and military equipment including inertial navigation for aeronautics and astronautics, and platform stabilization [3]. Mostly, the micro-gyroscopes have vibrating structures suspended above a silicon substrate. Among them, the tuning fork gyroscope (TFG) with an electrostatic drive and capacitive-type sensing is mostly preferred thanks to its reported potential capabilities and advantages such as common-mode rejection and low power consumption [4–9]. Generally, for vibratory gyroscopes, the mechanical quality factor (Q) of the operating modes is a key performance parameter. In order to achieve higher sensitivity and better rate resolution, the micro-gyroscope needs to be designed with a large sense-mode Q. A large drive-mode Q is also necessary for improving the bias stability as well as ensuring large drive amplitudes using small drive voltages [10]. The overall mechanical Q for an operating mode relates to energy-dissipation mechanisms in the TFG such as air damping (QAir ), dissipation through the substrate (QSupport ), thermo-elastic damping (QTED ), and surface loss (QSurface ) [11], and it is expressed by 1/Q = 1/QAir + 1/QSupport +1/QTED + Micromachines 2017, 8, 42; doi:10.3390/mi8020042
www.mdpi.com/journal/micromachines
Micromachines 2017, 8, 42 Micromachines 2017, 8, 42
2 of 10 2 of 10
Among them, the effect of air damping is considered as the primary loss mechanism for the device, 1/Q Surface . Among them, the effect of air damping is considered as the primary loss mechanism for the which operates in atmospheric conditions. device, which operates in atmospheric conditions. In a typical capacitive detection gyroscope, In a typical capacitive detection gyroscope, where where aa comb-type comb-type detection detection structure structure is is commonly commonly used, the squeeze-film air damping and the slide-film air damping between the proof masses andand the used, the squeeze-film air damping and the slide-film air damping between the proof masses substrate usually resultresult in a low [12,13].[12,13]. This requires a costly avacuum packaging solution the substrate usually in aQ-factor low Q-factor This requires costly vacuum packaging to achieve better sensitivity [14]. Developing high-performance micro-gyroscopes operating at solution to achieve better sensitivity [14]. Developing high-performance micro-gyroscopes operating atmospheric pressure is an effective means to reduce the cost. at atmospheric pressure is an effective means to reduce the cost. In recent years, years, different different structure structure designs designs have have been been pursued pursued to to improve improve the the performance performance of of In recent micro-gyroscope operating in atmospheric conditions. Geen et al. in Analog Devices developed micro-gyroscope operating in atmospheric conditions. Geen et al. in Analog Devices developed a a single-chip single-chip surface surface micro-machined micro-machined integrated integrated gyroscope gyroscope with with an an atmospheric atmospheric hermetic hermetic package package [15]. [15]. ◦ /h with This device device attained attained aa root root Allan Allan variance varianceof of50 50°/h with aa full full scale scalerange rangeof of±150°/s. This 150◦ /s. Alper Alper et et al. al. reported a silicon-on-insulator microelectromechanical systems (MEMS) gyroscope which operates reported a silicon-on-insulator microelectromechanical systems (MEMS) gyroscope which operates at ofof 1.5°/s [16]. CheChe et al.etreported on the at atmospheric atmospheric pressure pressurewith withaashort-term short-termbias biasstability stability 1.5◦ /s [16]. al. reported onhigh the Q-value for an electrostatic-driven tuning fork micro-machined gyroscope with a bar structure that high Q-value for an electrostatic-driven tuning fork micro-machined gyroscope with a bar structure operates at atmospheric pressure [17]. A A single-crystal that operates at atmospheric pressure [17]. single-crystalsilicon-based silicon-basedlateral lateralaxis axis tuning tuning fork fork gyroscope (TFG) with a vertically torsional sensing comb was developed by Guo et al. [12]. However, gyroscope (TFG) with a vertically torsional sensing comb was developed by Guo et al. [12]. However, the qualityfactor factorofof sensing mode in study this study was rather low duesqueeze-film to the squeeze-film air the quality thethe sensing mode in this was rather low due to the air damping damping between the proof mass and the glass substrate. Hu et al. reported on the sensitivity between the proof mass and the glass substrate. Hu et al. reported on the sensitivity improvement of improvement a slot-structureworking micro-gyroscope working in atmospheric through a a slot-structureofmicro-gyroscope in atmospheric conditions through aconditions tunable electrostatic tunable electrostatic spring constant attained by the triangular-shaped fixed electrodes [18]. spring constant attained by the triangular-shaped fixed electrodes [18]. In this study, study, we we aim aim at at improving improving the the sensitivity sensitivity of of the the angular angular rate rate sensor sensor by by increasing increasing the the In this quality factor of the driving and sensing modes. We propose a capacitive-type z-axis tuning fork quality factor of the driving and sensing modes. We propose a capacitive-type z-axis tuning fork gyroscope, a freestanding freestanding architecture. architecture. In sensitivity of the tuning tuning gyroscope, which which has has a In order order to to enhance enhance the the sensitivity of the fork masses and and the the substrate substrate is fork gyroscope, gyroscope, the the slide-film slide-film air air damping damping in in the the gap gap between between the the proof proof masses is eliminated by removing the substrate part underneath the device. The optimal design of the tuning tuning eliminated by removing the substrate part underneath the device. The optimal design of the fork gyroscope is carried out out by by simulating simulating its its mechanical mechanical behavior behavior with the finite finite element element method method fork gyroscope is carried with the (FEM) using using ANSYS ANSYS software. software. The (FEM) The proposed proposed z-axis z-axis tuning tuning fork fork micro-gyroscope micro-gyroscope has has been been fabricated fabricated and characterized. and characterized. 2. Design and Simulation The proposed TFG architecture is is shown shown in in Figure Figure 1. 1.
Figure 1. Schematic tuning fork gyroscope: (1) (1) outer frame, (2) Schematicdrawing drawingofofthe theproposed proposedfreestanding freestanding tuning fork gyroscope: outer frame, inner frame, (3) (3) driving comb electrodes, (4)(4) parallel (2) inner frame, driving comb electrodes, parallelplate platesense senseelectrode, electrode,(5) (5)double doublefolded folded beams, beams, self-rotation ring. ring. (6) anchors, (7) linear beams and (8) self-rotation
The TFG comprises comprises two proof masses. masses. Each for driving driving and and The TFG two proof Each proof proof mass mass has has an an outer outer frame frame (1) (1) for an inner one (2) for sensing. The set of driving comb electrodes (3) is attached to the outer frame. This an inner one (2) for sensing. The set of driving comb electrodes (3) is attached to the outer frame. design is such that the masses can oscillate in opposite directions along the x-axis under applied electrostatic force in driving mode. The motion of this set obeys the slide-film mechanism, offering
Micromachines 2017, 8, 42
3 of 10
This design is such that the masses can oscillate in opposite directions along the x-axis under applied electrostatic force in driving mode. The motion of this set obeys the slide-film mechanism, offering highly stable actuation. A pair of parallel plate sense electrode sets (4) is placed symmetrically within each inner frame. Upon the rotation, the Coriolis force excites this frame along an in-plane motion, such that the differential capacitance can be detected. In the design, the double folded beams (5), which have their fixed ends connected to the substrate by anchors (6), are employed for the suspension of the frames. The suspension of the frames is designed to allow the structure to oscillate in two orthogonal modes. In the design, the single folded beams are utilized to join the inner frame with the outer one such that the motion in the orthogonal direction can be easily sensed. The TFG is operated in the anti-phase drive mode, which excites the masses oscillated in opposite directions. To achieve an anti-phase oscillation in the driving mode, the two masses are coupled with a diamond-shaped coupling spring formed by four linear beams (7). The anti-phase vibration for the sensing mode is guaranteed by using a self-rotation ring (8), which is positioned inside the diamond-shaped coupling spring. This design can be used to stiffen the in-phase resonance of the structure, while leaving the anti-phase compliant. Based on this design the noise is decreased, resulting in the improvement of the sensitivity of the TFG [4]. The TFG design model was verified by finite element analysis (FEA) using ANSYS software. The dimensional parameters of the proof masses and the suspension beams were investigated to reach the optimally designed mechanical structure of the TFG, since the vibration modes are strongly dependent on these parameters. The modal simulation results for the drive and sense vibration modes of the micro-gyroscope are shown in Figure 2. The resonant frequencies of the driving and sensing modes were determined to be 9.788 and 9.761 kHz, respectively. Thus, the sensor bandwidth was evaluated to be 27 Hz. This bandwidth value satisfies the requirement of both the sensitivity and response time related to the gyroscope operation [19]. Other analysis results of the vibrating modes (not shown in this paper) showed that the higher unwanted modes have frequencies far greater than 15.4 kHz. Thus, there is a difference of about 57.5% between the two sensing and driving modes and the higher parasitic ones. It means that the crosstalk effect can be suppressed. Thus, the designed architecture using the mechanical coupling between driving and sensing proof masses exhibits the desired operation features. As mentioned above, the structure design of the TFG utilizes a set of interdigitated comb capacitors based on the variable area for large-stroke lateral actuation. For a parallel-plate actuator structure with N plates on each side, with thickness t, overlap length L and plate gap g, the total electrostatic drive force in a balanced actuation scheme is given by [19]: Fdrive = 2Nε0
tL VDC VAC g2
(1)
where V DC is the constant bias voltage applied to the moving fingers and V AC is the time-varying voltage applied to the fixed ones. We calculated the electrostatis force with V AC = 10 V, V DC = 11 V and used them as the input parameters for the simulation. As shown in the design model, the principle of differential detection is employed by symmetrically placing capacitive electrodes on two opposing sides of the inner mass frames, so that the capacitance change in the electrodes is in opposite directions. It means that a fully differential capacitive bridge is formed. The change in capacitance for an electrode set with N fingers on each side can be calculated as: ∆C = 2Nε0
tL Y g2
(2)
Where Y is the displacement of the sensing electrode in the motion direction. In fact, for the design of the sensing capacitive structure, there are two gaps, one of which is a smaller one corresponding to g, and the other is larger on the opposite side. Here, g = 2.5 µm, which is compliant with the resolution of photolithography. In order to optimize the large gap for
Micromachines 2017, 8, 42
4 of 10
the arrangement of the sensing fingers, several calculations were performed. Fixing a displacement of 1 µm in 2017, the sensing motion, a maximum large gap of 7 µm was determined by relying on Micromachines 8, 42 4 ofthe 10 calculation of the change in capacitance ∆C using Equation (2). Subsequently, the total gap between the two two adjacent fingers on one the interdigitated capacitor set is 12.5 The maximum change in the adjacent fingers onside oneofside of the interdigitated capacitor setµm. is 12.5 µm. The maximum capacitance within the above limited condition, calculated by applying Equation (2), is in the change in capacitance within the above limited condition, calculated by applying Equation (2), is range in the between 0.1 to 0.1 0.4to pF.0.4 pF. range between The structure parameters and calculated characteristics of the TFG TFG are are shown shown in in Table Table1.1. Table 1. Design parameters parameters of of the the micro-gyroscope. micro-gyroscope. Table 1. Design
Parameters Parameters Length of driving comb finger (ldcl) Length of driving comb finger (ldcl ) Driving comb overlap length (Ldc) Driving comb overlap length (Ldc ) Length of sensing comb finger (lscl) Length of sensing comb finger (lscl ) comb overlap (Lsc) Sensing Sensing comb overlap length (Llength sc ) Gap between comb fingers (g c) Gap between comb fingers (gc ) Gap between proof mass and substrate (h) Gap between proof mass and substrate (h) Device thickness (t) Device thickness (t) NumberNumber of driving (Ncomb of comb driving (Ndc) dc ) NumberNumber of sensing comb (Ncomb sc ) of sensing (Nsc) Sensing Sensing area (Asens ) (Asens) area Driving Driving area (Adriv ) (Adriv) area Driving mass (Mdriv ) Driving mass (Mdriv) Sensing mass (Msens ) Sensing mass (Msens) Driving mode stiffness (kdriv ) Driving mode stiffness (kdriv) Sensing mode stiffness (ksens ) Sensing Fdrivmode stiffness (ksens) Fdriv
Dimensions/Performance Dimensions/Performance 30 µm 30 µm 10 µm 10 µm 100 µm 100 µm 80 µm µm 80 2.5 µm 2.5 µm µm 4 µm µm 30 µm 1584 480 480 −7 m2 8.844 2 8.844××10 10−7 m −6 m2 1.685 × 10 1.685 × 10−−67 m2 1.17 kg 1.17×× 10 10−7 kg 6.18 × 10−−88 kg 6.18 × 10 kg 468.471 N/m 468.471 N/m 240.054 N/m 240.054 N/m 134.64 µN 134.64 N
The to the the rotation. rotation. A The operation operation principle principle of of the the TFG TFG is is based based on on aa tuning tuning fork’s fork’s response response to A z-axis z-axis input rotation signal signal perpendicular perpendicularto tothe thedevice deviceplane planecauses causesa aCoriolis-induced Coriolis-induced transfer energy input rotation transfer of of energy to to the sense vibration mode. The resulting in-plane displacement along the y-axis is sensed the sense vibration mode. The resulting in-plane displacement along the y-axis is sensed capacitively capacitively at sense electrodes attached to the of inner frame mass. of the proof mass. at sense electrodes attached to the inner frame the proof The designed micro-gyroscope is aimed at operating non-vacuum-packaged conditions. In The designed micro-gyroscope is aimed at operating in in non-vacuum-packaged conditions. In the the micro-gyroscope’s performance, air damping is considered to be the dominant damping micro-gyroscope’s performance, air damping is considered to be the dominant damping mechanism mechanism which includes slide-film damping and squeeze-film this the study, both and the which includes slide-film damping and squeeze-film damping. Indamping. this study,Inboth driving driving and sensing masses were designed to move in-plane, which results in limiting the squeezesensing masses were designed to move in-plane, which results in limiting the squeeze-film damping. film damping.
Figure 2. Fundamental Fundamental operation operation modes modes of of tuning tuning fork fork gyroscope gyroscope obtained obtained by by finite finite element element analysis: analysis: (a) sensing mode and (b) driving mode.
To eliminate the limitation of the slide-film air damping in this design, the part of the silicon substrate underneath the movable structures of the TFG was removed, which resulted in a freestanding architecture. This enabled high Q-factors for both the driving and sensing modes at atmospheric pressure. In the mentioned gyroscope performance, air damping is considered to be the dominant damping mechanism which includes slide-film damping and squeeze-film damping. Slide-film
Micromachines 2017, 8, 42
5 of 10
To eliminate the limitation of the slide-film air damping in this design, the part of the silicon substrate underneath the movable structures of the TFG was removed, which resulted in a freestanding architecture. This enabled high Q-factors for both the driving and sensing modes at atmospheric pressure. In the mentioned gyroscope performance, air damping is considered to be the dominant damping mechanism which includes slide-film damping and squeeze-film damping. Slide-film damping, or lateral damping, occurs when two plates of an area A, separated by a distance d, slide parallel to each other. The lateral damping coefficient can be expressed as: Cslide = µeff
A d
(3)
The effective viscosity µeff is approximated as [20]: µeff =
µ 1 + 2Kn + 0.2Kn0.788 e−Kn /10
(4)
The Knudsen number Kn is the measure of the gas rarefaction effect, which is a function of the gas mean free path λ and the gap d: λ (5) Kn = d Squeeze-film damping occurs when two parallel plates move toward each other and squeeze the air film in between them. The effective viscosity µeff in squeeze-film damping is given as [20]: µeff =
µ 1 + 9.638Kn1.159
(6)
The squeeze-film damping coefficient is calculated as: Csqueeze =
πµeff Lw3 N 3µ LwtpN + eff 3 4h3 g
(7)
Based on model parameters as shown in Table 1 and the desired working pressure (atmospheric pressure), we calculated the damping coefficients to estimate the quality factors of the TFG. To obtain a large Coriolis force and a detectable signal in the sense mode, large quality factors in the drive and sense modes are required. In order to get a high quality factor with the micro-gyroscope still operated in air, the damping effect must be reduced. In our micro-gyroscope, both the driving mass and sensing mass are designed to move parallel to the silicon substrate. In the air gaps between the movable part and fixed part of driving and sensing, the dominant viscous damping is slide-film damping between the proof mass and the silicon substrate. By removing a part of the silicon substrate under the movable structure, the slide-film damping was removed, which enabled high Q-factors for both the driving and sensing modes at atmospheric pressure. As shown in Figure 1, all the substrate underneath the moving part of the micro-gyroscope was removed. The quality factor (Q) in each operation mode is a critical design parameter of the TFG. A high Q in the drive mode is necessary to get a large drive mode vibration amplitude using a small drive voltage, and a high Q in the sensing mode contributes to improved rate resolution. Based on the above damping model, we calculated the damping coefficients with the Couette-type model. In the drive mode, the damping coefficient Cdrive is the total of the slide damping caused by the combs’ structure. Meanwhile, in the sensing mode, the damping coefficient Csense is related to only the squeeze damping of the sensing combs as variable-gap capacitive detection in the sense mode is applied. Under the atmospheric air condition, Qdrive and Qsense are calculated to be:
√ Qdrive =
kdrive Mdrive Cdrive
(8)
Micromachines 2017, 8, 42
6 of 10
√ Qsense =
ksense Msense Csense
(9)
where kdrive , Mdrive and ksense , Msense are the stiffness and weight of the driving mode and the sensing mode, respectively. Micromachines 2017, 8, 42 All these parameters were calculated and are shown in Table 1. 6 of 10 Table 2 shows the calculated damping coefficients and quality factors of the drive and sensing where drivethe , Mnon-freestanding drive and ksense, Msense arefreestanding the stiffnessTFG. and weight of the driving mode and the sensing modeskfor and mode, respectively. All these parameters were calculated and are shown in Table 1. Damping factor, and sensitivity of the drive andofsensing mode of sensing the Table 2. Table 2 shows thecoefficient, calculatedquality damping coefficients and quality factors the drive and non-freestanding and freestanding TFG for comparison. modes for the non-freestanding and freestanding TFG. type Cdrive quality (kg/s) factor, Csense Qsense S (pF/rad/s) TableTFG 2. Damping coefficient, and(kg/s) sensitivity ofQthe sensing mode of the drivedrive and non-freestanding Non-freestandingand freestanding−5TFG for comparison. 250 84.7 0.021 4.5 × 10 4.59 × 10−5 TFG (kg/s) sense Q drive Qsense −drive 5 −5(kg/s) Freestanding TFG TFG type 2.95 × 10C 381 91.8 S (pF/rad/s) 0.034 4.24 ×C10 Non-freestanding TFG 4.5 × 10−5 4.59 × 10−5 250 84.7 0.021 Freestanding TFG 2.95 × 10−5 4.24 × 10−5 381 91.8 0.034
In order evaluate the sensitivity of the micro-gyroscope, we simulated sensing displacement In ordertoto evaluate the sensitivity of the micro-gyroscope, we the simulated the sensing depending on the input angular rate. From this calculated displacement, we derived thewe sensitivity displacement depending on the input angular rate. From this calculated displacement, derived of the sensor by calculating the change of the capacitance using Equation (2). Here, we input the the sensitivity of the sensor by calculating the change of the capacitance using Equation (2). Here, we calculated damping coefficients as shown in Table 2 to investigate the effect of damping on the sensing input the calculated damping coefficients as shown in Table 2 to investigate the effect of damping on displacement for the two applying an input angular raterate of of ΩzΩ=z =10 the sensing displacement fordesign the twocases. design When cases. When applying an input angular 10 rad/s, rad/s, the sensing displacement and driving displacement of the TFG are 0.104 and 6.03453 µm, respectively. the sensing displacement and driving displacement of the TFG are 0.104 and 6.03453 µm, Compared with the non-freestanding TFG, the displacements in both operated modes at the resonant respectively. Compared with the non-freestanding TFG, the displacements in both operated modes frequency are higher. The are differences the drive mode mode aresensing 6% andmode 48%, respectively. at the resonant frequency higher. in The differences inand the sensing drive mode and are 6% and Applying a range ofApplying input angular ratesofoninput the structure, received dependence the sensing 48%, respectively. a range angular we rates on thethestructure, we of received the displacement on the input angular rate for the freestanding micro-gyroscope structure, as shown in dependence of the sensing displacement on the input angular rate for the freestanding Figure 3a. Figurestructure, 3b showsasthe dependence change in the output capacitance on change the input micro-gyroscope shown in Figureof3a.the Figure 3b shows the dependence of the in angular rate of the micro-gyroscope. The change in capacitance was calculated using Equation (2). the output capacitance on the input angular rate of the micro-gyroscope. The change in capacitance The calculated sensitivityusing of theEquation two micro-gyroscope structures is shown in Table 2. structures The results show the was (2). The sensitivity of the two micro-gyroscope is shown in sensitivity of the freestanding TFG was evaluated to be 0.034 pF/rad/s, which is 1.6 times higher than Table 2. The results show the sensitivity of the freestanding TFG was evaluated to be 0.034 pF/rad/s, that ofisthe TFG which 1.6non-freestanding times higher than thatdesign. of the non-freestanding TFG design.
Figure 3. Dependence of the sensing displacement (a) and the change in output capacitance C (b) Figure 3. Dependence of the sensing displacement (a) and the change in output capacitance ∆C (b) are are investigated as a function of the input angular rate z. investigated as a function of the input angular rate Ωz .
3. Fabrication 3. Fabrication The proposed tuning fork micro-gyroscope was fabricated by bulk micro-machining technology. The proposed tuning fork micro-gyroscope was fabricated by bulk micro-machining technology. The main fabrication steps are shown in Figure 4. A 2 cm × 2 cm silicon-on-insulator (SOI) wafer with The main fabrication steps are shown in Figure 4. A 2 cm × 2 cm silicon-on-insulator (SOI) wafer a 30-µm-thick device layer and a 2 µm buried silicon dioxide layer was used for fabricating the with a 30-µm-thick device layer and a 2 µm buried silicon dioxide layer was used for fabricating micro-gyroscope. The SOI wafer was first cleaned by a standard cleaning process and thermally the micro-gyroscope. The SOI wafer was first cleaned by a standard cleaning process and thermally oxidized at 1100 °C for 1 h to form a protecting mask for the dry etching process (Figure 4a). Next, the oxidized at 1100 ◦ C for 1 h to form a protecting mask for the dry etching process (Figure 4a). Next, gyroscope patterns were transferred to the SOI wafer surface by photolithography and developing processes (Figure 4b,c). The DRIE process was then performed to a depth of 30 µm to reach the buried dioxide layer of the SOI wafer (Figure 4d). After the etching process, Au electrodes were defined using the lift-off process (Figure 4e). Removing the substrate underneath the device layer was carried out by back-side photolithography and an additional back-side etching step. Finally, a vapor HF etching process was used to etch the SiO2 layer underneath the device layer and to release the
Micromachines 2017, 8, 42
7 of 10
the gyroscope patterns were transferred to the SOI wafer surface by photolithography and developing processes (Figure 4b,c). The DRIE process was then performed to a depth of 30 µm to reach the buried dioxide layer of the SOI wafer (Figure 4d). After the etching process, Au electrodes were defined using the lift-off process (Figure 4e). Removing the substrate underneath the device layer was carried out by back-side photolithography and an additional back-side etching step. Finally, a vapor HF etching process was used to etch the SiO2 layer underneath the device layer and to release the movable electrodes, beams, and proof masses. The microstructures of the fabricated micro-gyroscope were inspected by2017, scanning Micromachines 8, 42 electron microscopy (SEM). 7 of 10 Micromachines 2017, 8, 42
7 of 10
Figure 4. 4. Fabrication (a) (a) Starting SOISOI wafer; (b) Figure Fabricationprocess processofofthe theproposed proposedtuning tuningfork forkgyroscope. gyroscope. Starting wafer; photolithography; (c) pattern transferring; (d) deep silicon etching; (e) sputter metallization; (f) Figure 4. Fabrication process of the proposed tuning fork gyroscope. wafer; (b) (b) photolithography; (c) pattern transferring; (d) deep silicon etching;(a)(e)Starting sputterSOI metallization; backside etching; (g) buried oxide etching. photolithography; transferring; (f) backside etching;(c) (g)pattern buried oxide etching.(d) deep silicon etching; (e) sputter metallization; (f) backside etching; (g) buried oxide etching.
Figure 5a–c show the whole top view of the TFG, a close-up of the self-rotation ring, and the Figure 5a–c show the whole top view of the TFG, a close-up of the self-rotation ring, and the sensing comb electrodes, The components the micro-gyroscope are defined wellthe by Figure 5a–c show therespectively. whole top view of the TFG, aofclose-up of the self-rotation ring, and sensing comb electrodes, respectively. The components of the micro-gyroscope are defined well by the the proposed fabrication process. sensing comb electrodes, respectively. The components of the micro-gyroscope are defined well by proposed fabrication process. the proposed fabrication process.
Figure 5. SEM images of a fabricated tuning fork gyroscope: (a) TFG chip, (b) close-up of the self-rotation ring, and (c)of sensing comb electrodes. Figure 5. SEM images of fabricated tuning fork fork gyroscope: gyroscope: (a) (a) TFG TFG chip, chip, (b) (b) close-up close-up of of the Figure 5. SEM images aa fabricated tuning the self-rotation ring, and (c) sensing comb electrodes. self-rotation ring, and (c) sensing comb electrodes.
4. Device Characterization and Discussion
4. Device and Discussion The Characterization frequency and angular rate response of the fabricated tuning fork gyroscope were investigated in atmospheric conditions. Figure shows measured frequency response ofwere the angular rate rate response of the6 fabricated tuning fork gyroscope investigated The frequency frequencyand and angular response of the the fabricated tuning fork were gyroscope sensing modeinofconditions. the TFG. The resonant frequency about 11.125 kHz. From this, the qualityof factor in atmospheric Figure 6 shows the measured frequency response of the sensing mode investigated atmospheric conditions. Figure 6was shows the measured frequency response the of the micro-gyroscope sensing mode was evaluated to be 111.2. The resonant frequency and quality of the TFG. resonant frequency about was 11.125 kHz. From this, thethis, quality factor of the sensing modeThe of the TFG. The resonantwas frequency about 11.125 kHz. From the quality factor factor of the micro-gyroscope mode were 11.250 respectively. measured micro-gyroscope sensing mode drive was evaluated to be 111.2. The and resonant frequency andThe quality factor of the micro-gyroscope sensing mode was evaluated to bekHz 111.2. The349, resonant frequency and quality results the micro-gyroscope resonant drive frequencies were different from the simulated which The is results themeasured result of of the micro-gyroscope mode were 11.250 kHz11.250 and 349, respectively. The measured of the factor ofofthe drive mode were kHz and 349,results, respectively. the change of the spring beam width caused by the DRIE process and the holes designed on the resonant differentwere fromdifferent the simulated results, whichresults, is the result results offrequencies the resonantwere frequencies from the simulated whichofisthe thechange result of micro-gyroscope structure, which caused the mass ofDRIE sensor’s proof mass decrease. In our spring caused by the DRIE process and the holes designed on the micro-gyroscope the changebeam of thewidth spring beam width caused by the process and theto holes designed oncase, the the drive and sensing mode frequencies had a mismatch of about 125 Hz, corresponding to the sensor structure, which caused the mass of sensor’s proof mass to decrease. In our case, the drive and sensing micro-gyroscope structure, which caused the mass of sensor’s proof mass to decrease. In our case, bandwidth. This result the of requirement to optimize sensitivity and bandwidth [4]. mode frequencies had mode asatisfies mismatch about 125 Hz, corresponding to the sensor bandwidth. result the drive and sensing frequencies had a mismatch ofthe about 125 Hz, corresponding toThis the sensor satisfies the This requirement to optimize the sensitivity and bandwidth [4]. and bandwidth [4]. bandwidth. result satisfies the requirement to optimize the sensitivity
factor of the micro-gyroscope drive mode were 11.250 kHz and 349, respectively. The measured results of the resonant frequencies were different from the simulated results, which is the result of the change of the spring beam width caused by the DRIE process and the holes designed on the micro-gyroscope structure, which caused the mass of sensor’s proof mass to decrease. In our case, the drive and sensing Micromachines 2017, 8, 42 mode frequencies had a mismatch of about 125 Hz, corresponding to the sensor 8 of 10 bandwidth. This result satisfies the requirement to optimize the sensitivity and bandwidth [4]. Micromachines 2017, 8, 42 8 of 10 Figure 7 shows the block diagram and the experimental setup of the measurement system for the TFG angular rate characterization. The atmosphere packaged gyroscope was investigated by coupling the angular rate on a rotation table with a controllable speed by programming. To excite the micro-gyroscope, an alternating current (AC) voltage of 10 V, with the driving frequency fixed to the resonance frequency of the driving mode at 11.25 kHz, was applied to the movable comb electrode. A direct current (DC) bias of 5 V was applied to the stationary electrode. In order to supply the voltage for the driving comb electrode and detect the capacitance change of the sensing comb structure, the micro-gyroscope was bonded on a printed circuit board (PCB). Wiring from the comb electrodes to the PCB was carried out by using an ultrasonic wedge bonding machine. The capacitance change was transformed to the MS3110 capacitance readout circuit. MS3110 is a commercial product which is used for differential capacitance sensing. This is a preferred C-V readout Figure6.6.Frequency Frequencyresponse responseof of TFGsensing sensingmode. mode. Figure product due to its stability and detective ability in the TFG fF range. The output of MS3110 is voltage which is proportional to the change in the capacitance of the sensing comb electrodes. The signal was processed a low-pass filter amplifier and intosetup a personal (PC) system via the data Figure 7by shows the block diagram and thetransformed experimental of the computer measurement for the acquisition (DAQ) module USB6009. module packaged was used gyroscope to interfacewas the investigated output signalby from the TFG angular rate characterization. The This atmosphere coupling 3110 circuit board with the PC. with The signal processing was performed by using LabVIEW software. theMS angular rate on a rotation table a controllable speed by programming.
Figure 7. The measurement system is setup for characterizing TFG: (a) block diagram and (b) Figure 7. The measurement system is setup for characterizing TFG: (a) block diagram and experimental setup; the measurement system includes (1) rate table controller; (2) rate table; (3) micro(b) experimental setup; the measurement system includes (1) rate table controller; (2) rate table; gyroscope; (4) C/V transformation; (5) DAQ; (6) computer for signal processing and display; (7) DC (3) micro-gyroscope; (4) C/V transformation; (5) DAQ; (6) computer for signal processing and display; power; (8) AC power; (9) computer controller. (7) DC power; (8) AC power; (9) computer controller.
Figure 8 shows the measured output signal versus the input angular rate. The output voltage −1 to To linearly excite the micro-gyroscope, an alternating current of 10 V, with the−1driving was proportional to the input angular rate in the(AC) rangevoltage from −200°·s 200°·s . The frequency fixed to gyroscope the resonance frequency to of be the11.56 driving mode at 11.25 kHz, was applied to the sensitivity of the was determined mV/°/s. movable comb electrode. A direct current (DC) bias of 5 V was applied to the stationary electrode. In order to supply the voltage for the driving comb electrode and detect the capacitance change of the sensing comb structure, the micro-gyroscope was bonded on a printed circuit board (PCB). Wiring
Micromachines 2017, 8, 42
9 of 10
from the comb electrodes to the PCB was carried out by using an ultrasonic wedge bonding machine. The capacitance change was transformed to the MS3110 capacitance readout circuit. MS3110 is a commercial product which is used for differential capacitance sensing. This is a preferred C-V readout product due to its stability and detective ability in the fF range. The output of MS3110 is voltage which is proportional to the change in the capacitance of the sensing comb electrodes. The signal was processed by a low-pass filter amplifier and transformed into a personal computer (PC) via the data acquisition (DAQ) module USB6009. This module was used to interface the output signal from the MS 3110 circuit board with the PC. The signal processing was performed by using LabVIEW software. Figure 8 shows the measured output signal versus the input angular rate. The output voltage was linearly proportional to the input angular rate in the range from −200◦ ·s−1 to 200◦ ·s−1 . The sensitivity ◦ of the gyroscope Micromachines 2017,was 8, 42determined to be 11.56 mV/ /s. 9 of 10
Figure 8. The output voltage Vout of the TFG is investigated as a function of the input angular rate z. Figure 8. The output voltage V out of the TFG is investigated as a function of the input angular rate Ωz .
5. Conclusion 5. Conclusions A z-axis tuning fork micro-gyroscope which has a freestanding and anti-phase controlled A z-axis tuning fork micro-gyroscope which has a freestanding andtuning anti-phase controlled architecture was designed and fabricated. The performance of the designed fork gyroscope architecture was designed and fabricated. The performance of the designed tuning fork gyroscope was was improved by limiting the influence of the squeeze-film air damping and slide-film air damping improved by limiting influence of the squeeze-film air damping slide-film due to due to the in-plane the operation of the driving and sensing modes of and the sensor andair thedamping freestanding thearchitecture in-plane operation of the driving and sensing modes of the sensor and the freestanding architecture as well. The influence of the crosstalk effect caused by higher parasitic modes on the as well. The modes influence the crosstalk effect caused by higher parasitic modes themicro-machining operating modes operating wasof eliminated. The designed micro-gyroscope was fabricated byon bulk was eliminated.The Thequality designed micro-gyroscope was fabricated by bulk micro-machining technology. technology. factors of the micro-gyroscope sensing and driving modes operated in condition were 111.2 and 349, respectively. The bandwidth wasoperated measuredinto atmospheric be 125 Hz. Theatmospheric quality factors of the micro-gyroscope sensing and driving modes The sensitivity of the tuning fork micro-gyroscope was measured to be 11.56 mV/°/s. condition were 111.2 and 349, respectively. The bandwidth was measured to be 125 Hz. The sensitivity ◦ Although tuning fork micro-gyroscope most mV/ challenging of the tuning forkthe micro-gyroscope was measured istothe be 11.56 /s. type of transducer in the micro-world, its tuning advantage high sensitivity ismakes it have great potential a varietyinofthe Although the forkofmicro-gyroscope the most challenging type of for transducer applications in everyday life. For our proposed device, one robust structure for improving the micro-world, its advantage of high sensitivity makes it have great potential for a variety of applications sensitivity the micro-gyroscope operated in atmospheric conditions was introduced and verified. in everyday life. For our proposed device, one robust structure for improving the sensitivity the However, impact factors such as long-termconditions operation and temperature the operation micro-gyroscope operated in atmospheric was environment introduced and verified. on However, impact of the sensor were also investigated. In addition, integration with accelerometers in an inertial factors such as long-term operation and environment temperature on the operation of the sensor were measurement unit is also of interest for further development also investigated. In addition, integration with accelerometers in an inertial measurement unit is also Acknowledgments: work is supported by the Ministry of Science and Technology (MOST), Vietnam, as the of interest for furtherThis development. project of NAFOSTED coded 103.99-2014.34. Acknowledgments: This work is supported by the Ministry of Science and Technology (MOST), Vietnam, Author Contributions: M.N.N, N.S.H, L.Q.N. conceived and designed the device; H.M.C fabricated the devices, as the project of NAFOSTED coded 103.99-2014.34. M.N.N, N.S.H, L.Q.N. and H.N.V. set up the experiments and characterized the device; all authors took part in Author Contributions: M.N.N, N.S.H, L.Q.N. conceived and designed the device; H.M.C fabricated the devices, writing the paper. M.N.N, N.S.H, L.Q.N. and H.N.V. set up the experiments and characterized the device; all authors took part in Conflicts of Interest: The authors declare no conflict of interest. writing the paper.
References 1. 2.
Classen, J.; Frey, J.; Kuhlmann, B.; Ernst, P.E.; Bosch, R. MEMS gyroscopes for automotive applications. In Advanced Microsystems for Automotive Applications; Springer: Berlin, Germany, 2007; pp. 291306. Shkel, A.M. Micromachined Gyroscopes: Challenges, Design Solutions, and Opportunities, In Smart Structures and Materials 2001: Smart Electronics and MEMS, Proceedings of the SPIE, Newport Beach, CA,
Micromachines 2017, 8, 42
10 of 10
Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2.
3. 4. 5.
6. 7. 8. 9. 10. 11. 12.
13. 14.
15.
16. 17. 18. 19. 20.
Classen, J.; Frey, J.; Kuhlmann, B.; Ernst, P.E.; Bosch, R. MEMS gyroscopes for automotive applications. In Advanced Microsystems for Automotive Applications; Springer: Berlin, Germany, 2007; pp. 291–306. Shkel, A.M. Micromachined Gyroscopes: Challenges, Design Solutions, and Opportunities. In Smart Structures and Materials 2001: Smart Electronics and MEMS, Proceedings of the SPIE, Newport Beach, CA, USA, 4 March 2001; pp. 74–85. Grewal, M.S.; Weill, L.R.; Andrews, A. Global Positioning Systems, Inertial Navigation and Integration; John Wiley & Sons: Hoboken, NJ, USA, 2001. Weinberg, M.; Kourepenis, A. Error sources in in-plane silicon tuning-fork MEMS gyroscopes. J. Microelectromech. Syst. 2006, 15, 479–491. [CrossRef] Gomez, U.M.; Kuhlmann, B.; Classen, J.; Bauer, W.; Lang, C.; Veith, M.; Esch, E.; Frey, J.; Grabmaier, F.; Offterdinger, K.; Raab, T.; Faisst, H.J.; Willig, R.; Neul, R. New surface micromachined angular rate sensor for vehicle stabilizing systems in automotive applications. In Proceeding of 13th International Conference on TRANSDUCERS, 5–9 June 2005; pp. 184–187. Zotov, S.A.; Simon, B.R.; Prikhodko, I.P.; Trusov, A.A.; Shkel, A.M. Quality factor maximization through dynamic balancing of tuning fork resonator. IEEE Sens. J. 2014, 14, 2706–2714. [CrossRef] Azgin, K.; Temiz, Y.; Akin, T. An SOI-MEMS Tuning Fork Gyroscope with Linearly Coupled Drive Mechanism. In Proceeding of the 20th International Conference on MEMS, Kobe, Japan, 21–25 January 2007; pp. 607–610. Zaman, M.F.; Sharma, A.; Hao, Z.; Ayazi, F. A mode-matched silicon-yaw tuning-fork gyroscope with subdegree-per-hour allan deviation bias instability. J. Microelectromech. Syst. 2008, 17, 1526–1536. [CrossRef] Wang, R.; Cheng, P.; Xie, F.; Young, D.; Hao, Z. A multiple-beam tuning-fork gyroscope with high quality factors. Sen. Actuators A. 2011, 166, 22–33. [CrossRef] Yazdi, N.; Ayazi, F.; Najafi, K. Micromachined inertial sensors. Proc. IEEE 1998, 86, 1640–1659. [CrossRef] Trusov, A.A.; Schofield, A.R.; Shkel, A.M. A substrate energy dissipation mechanism in in-phase and anti-phase micromachined z-axis vibratory gyroscopes. J. Micromech. Microeng. 2008, 18, 095016. [CrossRef] Guo, Z.Y.; Yang, Z.C.; Zhao, Q.C.; Lin, L.T.; Ding, H.T.; Liu, X.S.; Cui, J.; Xie, H.; Yan, G.Z. A lateral-axis micromachined tuning fork gyroscope with torsional Z-sensing and electrostatic force-balanced driving. J. Micromech. Microeng. 2010, 20, 025007. [CrossRef] Trusov, A.A.; Schofield, A.R.; Shkel, A.M. Micromachined rate gyroscope architecture with ultra-high quality factor and improved mode ordering. Sens. Actuators A. 2011, 165, 26–34. [CrossRef] Tekin, T.; Ngo, H.D.; Wittler, O.; Bouhlal, B.; Lang, K.D. Packaging of MEMS/MOEMS and Nanodevices: Reliability, Testing, and Characterization Aspects. In Proceeding of the SPIE MOEMS-MEMS, San Francisco, CA, USA, 10 February 2011; pp. 24–25. Geen, J.A.; Sherman, S.J.; Chang, J.F.; Lewis, S.R. Single-chip surface-micromachined integrated gyroscope with 50/h root Allan variance. In Proceedings of the 2002 IEEE International SolidState Circuits Conference, San Francisco, CA, USA, 3–7 February 2002; pp. 426–427. Alper, S.E.; Azgin, K.; Akin, T. A high-performance silicon-on-insulator MEMS gyroscope operating at atmospheric pressure. Sens. Actuators A. 2007, 135, 34–42. [CrossRef] Che, F.L.; Xiong, B.; Li, Y.F.; Wang, Y.L. A novel electrostatic-driven tuning fork micromachined gyroscope with a bar structure operating at atmospheric pressure. J. Micromech. Microeng. 2010, 20, 015025. [CrossRef] Hu, S.C.; Jin, Z.H.; Zhu, H.J.; Wang, H.; Ma, M.J. A slot-structure MEMS gyroscope working at atmosphere with tunable electrostatic spring constant. J. Microelectromech. Syst. 2013, 22, 909–918. [CrossRef] Acar, C.; Shkel, A.M. MEMS Vibratory Gyroscopes: Structural Approaches to Improve Robustness; Springer Science & Business Media: Berlin, Germany, 2009. Veijola, T.; Turowski, M. Compact damping models for laterally moving microstructures with gas-rarefaction effects. J. Microelectromech. Syst. 2001, 10, 263–273. [CrossRef] © 2017 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/).
