JP4983107B2 - Inertial sensor and method of manufacturing inertial sensor - Google Patents

Description

  The present invention relates to an inertial sensor and a method for manufacturing the inertial sensor.

  As a conventional technique, a vibration gyroscope that is processed using a semiconductor processing technique using a material such as silicon (Si) is known. This type of gyroscope oscillates an inertial mass in a certain direction, and detects the magnitude of angular velocity by displacement due to Coriolis force generated when the angular velocity is input. This angular velocity sensor can be applied to an input interface, a camera shake correction of a video camera or a still camera, and the like.

  In general, such an angular velocity sensor drives a vibrator manufactured by a semiconductor process by an electric signal, and the vibrator is perpendicular to the driving direction by the generated Coriolis force due to the speed of this driving and the angular velocity applied to the system. By measuring the amount of displacement, the angular velocity can be measured.

  It is known that the Coriolis force is very small compared to the driving force, and the displacement is also small, so that it is difficult to detect. As a known technique, there is a method in which a coupling is generated and a gain is obtained by bringing a driving frequency and a detection frequency close to each other. For example, when the Q value is sufficiently large, a gain of about 10 times can be obtained by setting the ratio between the drive frequency and the detection frequency (this ratio is hereinafter referred to as detuning degree) to 1.05 or 0.95. .

  As a method for adjusting the degree of detuning, a reference electrode, a piezoelectric body, a drive electrode, a detection electrode, etc. of a vibrator formed from a single crystal silicon substrate are formed by reactive ion etching, isotropic ion etching, or crystal anisotropic etching. It is disclosed that a desired degree of detuning is obtained by grinding other than the surface (for example, see Patent Document 1). Further, although it is premised on driving in the atmosphere, a technique is disclosed in which the relationship between the degree of detuning α and the Q value of the detection frequency is numerically limited (see, for example, Patent Document 2).

  In addition, a sensor that realizes multiple axes by itself has been developed in recent years (see, for example, Patent Document 3). Using such a composite sensor is more advantageous in terms of both cost and mounting area than mounting a single unit with a plurality of axes.

  In order to obtain a sufficient gain for Coriolis displacement using a single-axis sensor that is advantageous in terms of both cost and mounting area, it is necessary to bring the drive frequency and the detection frequency close to each other.

  However, as shown in FIG. 14, the single-structure multi-axis sensor has a plurality of detection frequencies with respect to the drive frequency, and thus the detection frequencies approach each other. For this reason, coupling occurs between detections of a plurality of axes, and even a slight imbalance due to the process causes a deterioration in the sensitivity of other axes. The other-axis sensitivity referred to here is, for example, a ratio of magnitude that misidentifies the angular velocity generated around the first axis as the angular velocity of the second axis that intersects the first axis. Specifically, for example, when an angular velocity of 100 deg / sec is generated around the X axis, the other axis is detected if 1 deg / sec is detected around the Y axis even though no angular velocity is generated around the Y axis. The sensitivity is 1%.

  In general, a single-axis angular velocity sensor is mass-produced with a sensitivity of about 1% for other axes. However, the multi-axis composite sensor can achieve only about several percent of other-axis sensitivity due to the coupling between the detection frequencies.

  When the sensitivity of other axes deteriorates, for example, the angular velocity of the other axes becomes an error factor in the control of robots, vehicles, etc., increasing drift and remarkably worsening the angular error at the time of integration. Normal control is not possible.

  For example, when a gain of 10 times is desired, the degree of detuning may be designed to be 1.05 or 0.95. At this time, the difference between the drive frequency and the detection frequency is 5% when the drive frequency is considered as a reference. However, if the device is manufactured without any ingenuity, the detection frequency may be low or high on the drive frequency side. Will be biased to. At this time, if each axis requires a gain for a single drive frequency, the axes are separated from each other by 5%. The frequency ratio between the detection axes is 1% or less. For example, if the detection frequencies of 0.5% are separated from each other, when the detection Q value is sufficiently high, there is a gain of about 100 times. Even if an angular velocity is applied only around the first axis due to process variations, etc., the transducer does not move completely parallel to the direction of the Coriolis force corresponding to the first axis, for example, corresponds to the second axis. It is also displaced in the direction of Coriolis force. The ratio of vibration leakage at this time has a gain of 100 times, and the other-axis sensitivity is greatly deteriorated.

JP-A-2005-241382 JP 2002-310661 A JP 2005-31096 A

  The problem to be solved is that the single-structure multi-axis sensor has a plurality of detection frequencies with respect to the drive frequency, and therefore the detection frequency is biased toward the low frequency side or the high frequency side of the drive frequency. In other words, the coupling between the detections of a plurality of axes occurs, and even a slight imbalance due to the process causes the sensitivity of other axes to deteriorate.

  The present invention makes it possible to improve the sensitivity of other axes by changing the spring constant of the elastic support that supports the vibrator so that the drive mode frequency has different detection mode frequencies on the low frequency side and the high frequency side, respectively. To.

  The inertial sensor of the present invention is an inertial sensor including a vibrator supported so as to be displaceable in a floating state with respect to a substrate, and a displacement detection unit that detects the displacement of the vibrator. The driving mode frequency has different detection mode frequencies on the low frequency side and the high frequency side, respectively.

  The inertial sensor of the present invention has separate detection mode frequencies on the low frequency side and the high frequency side of the drive mode frequency, so that the two detection mode frequencies can be separated. As a result, the gain of the coupling between driving and detection is obtained, and the coupling between the detections of the other axes is minimized, and the sensitivity of the other axes is suppressed.

  The inertial sensor manufacturing method according to the present invention includes a vibrator supported so as to be displaceable in a floating state with respect to a substrate, a first elastic support body and a second elastic support body that support the vibrator, and the vibrator. The inertial sensor manufacturing method includes a displacement detection unit that detects a displacement of the first elastic unit so as to have different detection mode frequencies on the low frequency side and the high frequency side of the drive mode frequency of the vibrator. The support and the second elastic support are formed so as to have different spring constants.

  In the inertial sensor manufacturing method of the present invention, the first elastic support and the second elastic support are made to have different spring constants so that the drive mode frequency has different detection mode frequencies on the low frequency side and the high frequency side, respectively. Therefore, the vibration frequency of the first elastic support and the vibration frequency of the second elastic support are different. Therefore, since either one has a detection mode frequency on the low frequency side and the other side has a detection mode frequency on the high frequency side higher than the low frequency side, the first elastic support body and the second elastic support body. Have separate detection mode frequencies. As a result, the gain of the coupling between driving and detection is obtained, and the coupling between the detections of the other axes is minimized, and the sensitivity of the other axes is suppressed.

  According to the inertial sensor of the present invention, since the drive mode frequency has different detection mode frequencies on the low frequency side and the high frequency side, the coupling between the detections of the other axes is minimized and the sensitivity of the other axes is suppressed. Therefore, in the control of the robot, the vehicle, etc., there is an advantage that the detection error of the other axis is reduced and the detection can be performed with high accuracy. This eliminates the need for frequent correction by an inclination sensor or the like as in the prior art, and has an advantage that normal control can be performed.

  According to the method for manufacturing an inertial sensor of the present invention, the spring constant is set between the first elastic support and the second elastic support so that the drive mode frequency has different detection mode frequencies on the low frequency side and the high frequency side, respectively. Since they are formed differently, it is possible to manufacture an inertial sensor capable of suppressing the sensitivity of the other axis while minimizing the coupling between the detections of the other axes. As a result, in the control of the robot, the vehicle, etc., there is an advantage that it is possible to provide an inertial sensor in which detection errors of other axes are reduced and highly accurate detection is possible.

  An embodiment (first example) relating to the inertial sensor of the present invention will be described with reference to FIGS. 1 and 2 show, as an example, an inertial sensor including a multi-axis composite sensor including an angular velocity sensor. (1) of FIG. 1 is a schematic cross-sectional view showing an inertial sensor, and (2) is a perspective view showing an outline of a configuration example of a vibrator and electrodes arranged on the upper part. (1) of FIG. 2 is a front view showing the positional relationship between the vibrator, the elastic support and the electrode, and (2) is a plan view.

  As shown in FIGS. 1 and 2, a support portion 12 is formed on the first substrate 100, and one end side of the elastic supports 13-1 to 13-4 is supported by the support portion 12. The vibrator 11 is supported on the other end side of each of the elastic supports 13-1 to 13-4 in a state of being separated from the first substrate 100 and the second substrate 200 described later. In addition, a displacement detection unit 14 that detects the displacement of the vibrator 11 and outputs a signal is provided, for example, on the side of the second substrate 200 facing the vibrator 11. Here, the displacement of the vibrator 11 is detected by the capacitance change between the vibrators 11 and the electrodes 211-1 to 211-4 provided above the vibrator 11. Further, the driving electrode 212 provided above the vibrator 11 drives the vibrator 11 in, for example, a third axis (for example, Z axis) direction in a three-dimensional coordinate system. Further, for example, an electrode 111 for monitoring the driving of the vibrator 11 is formed on the first substrate 100 below the vibrator 11.

  The elastic supports 13-1 to 13-4 are, for example, elastic supports in the X-axis direction so that the vibrator 11 has different detection mode frequencies on the low frequency side and high frequency side of the drive mode frequency. (First elastic support members) 13-1 and 13-3 and Y-axis elastic support members (second elastic support members) 13-2 and 13-4 are formed to have different spring constants. . That is, the elastic support bodies at positions facing each other with the vibrator interposed therebetween have the same spring constant. This is because the vibration mass of each mode is common and the vibration frequency can be controlled by adjusting the spring constant of the elastic support 13. As shown in FIG. 3, for example, when it is desired to separate the detection frequencies by 5% on both sides of the drive frequency, ω = 2πf = √ (k / m), where ω is the vibration frequency and f is the vibration of the vibrator. The frequency, k is the spring constant of the elastic support 13), and the spring constant may be separated by 21%.

  For example, when controlling by the length of the spring, since the resonance frequency is inversely proportional to the square of the length of the spring, the length of the spring supporting the vibrator is set in the second axial direction with respect to the spring in the first axial direction, for example. By adjusting the length of the spring to √ (1 / 1.1) = 0.95 times, the detection frequencies of the first axis and the second axis can be separated by 10% across the drive frequency.

  Since the spring width does not ideally affect the resonance frequency, it is practically difficult to adjust only by the spring width. It is appropriate to consider to the extent that fine adjustment is performed when an unnecessary mode appears in the aspect ratio with the length.

  When controlling by the thickness of the spring, since the primary resonance frequency of the beam bending is proportional to the thickness of the spring, the second axial spring is 1.1. By adjusting the thickness twice, the detection frequencies of the first axis and the second axis can be separated by 10%.

  When adjusting with the spring material, it is proportional to the square root √ (E / ρ) of the ratio of Young's modulus and density of the material. The detection frequencies of the two axes can be separated by 10%.

  There is also a method of providing a difference in detection frequency by inserting a slit, groove, hole or the like in the spring to weaken the spring.

  The inertial sensor 1 according to the present invention detects the acceleration in the direction of the first axis (for example, the X axis) and the angular velocity around the second axis (for example, the Y axis) in the three-dimensional coordinate system, and the vibration 11 Excitation means (drive electrode 212) that vibrates the child 11 in the direction of the third axis (for example, the Z axis) in the three-dimensional coordinate system at a sufficiently high frequency with respect to the required response of acceleration and angular velocity, and a displacement for detecting this displacement Signal separation means (not shown) for separating the low frequency component and the component around the excitation frequency in the detection unit 14 (acceleration and angular velocity detection electrodes 211-1 to 211-4) and the signal obtained by the displacement detection unit 14. 2) and a calculation means (not shown) for obtaining acceleration in the first axis (for example, the X axis) direction and an angular velocity calculation means (not shown) for obtaining the angular velocity around the second axis (for example, the Y axis) in the low frequency component. Z)) It can detect both velocity and acceleration.

  Hereinafter, the operation principle of the inertial sensor 1 will be described.

  First, an angular velocity detection method will be described.

  An AC voltage that drives the vibrator 11 at its resonance frequency is applied between the vibrator 11 and the drive electrode 212 in the direction of the third axis (for example, the Z axis in the three-dimensional coordinate system). An electrostatic force is generated between the drive electrodes 212 to drive the vibrator 11 periodically.

Here, when an angular velocity is applied around the first axis (for example, the X axis in the three-dimensional coordinate system), the Coriolis force F _coriolis is generated in the direction of the second axis (for example, the Y axis in the three-dimensional coordinate system). This Coriolis force F _coriolis is expressed by the following equation.

F _coriolis = 2mvΩ

  Here, m is the mass of the vibrator 11, v is the vibration velocity in the driving direction, and Ω is the angular velocity applied from the outside.

  When the Coriolis force is generated in the direction of the second axis (for example, the Y axis in the three-dimensional coordinate system), the force is applied to the vibrator 11 and is displaced in the second axis direction. As shown in FIG. 4, since the vibrator 11 has a different center of gravity position and the height of the support positions of the elastic supports 13-1 to 4, a moment is generated by the Coriolis force and vibrates in the twisting direction. The driving direction by the driving electrode 212 is the Z direction. This displacement in the twisting direction is detected by the capacitance change of the four electrodes 211-1 to 211-4.

  For example, among the four electrodes 211-1 to 211-4, the two electrodes 211-1 and 211-4 on the side where the gap is inclined and the gap is widened, the capacitances C <b> 1 and C <b> 4 are reduced, and the gap is narrowed and the gap is narrowed The two electrodes 211-2 and 211-3 have increased capacitances C2 and C3. By taking the sum of the capacitances C1 + C4 on the widened sides and taking the sum of the capacitances C2 + C3 on the narrowed sides, the difference (C1 + C4) − (C2 + C3) between the sums of the capacitances of the respective electrodes is taken to efficiently twist. It is possible to detect the displacement due to, that is, the angular velocity.

  Further, when an angular velocity is applied around the second axis (for example, the Y axis in the three-dimensional coordinate system), a Coriolis force is generated in the direction of the first axis (for example, the X axis in the three-dimensional coordinate system). Similarly, since the angular velocity generated around the second axis due to the capacitance change of the four electrodes 211-1 to 211-4 can be detected, the angular velocity for two axes can be detected.

  At this time, by setting the first axis direction spring and the second axis direction spring to appropriate lengths, the detection frequency of the first axis on the low frequency side and the high frequency side of the drive resonance frequency, and the second axis The detection frequency can be moved, and the degree of detuning can be set individually. Here, as shown in FIG. 5, when the Q value of the detection mode is sufficiently higher than the expected gain, in order to obtain a gain of 2 or more with a clear multiplication effect, the detuning degree is set to 0.71 to 1.22. It may be set between. Similarly, for a gain of 5 times or more, the detuning degree may be set between 0.90 and 1.09. In the case of 10 times or more, it may be set to 0.95 to 1.04. However, if a high gain is to be gained, it becomes weaker against process variations, and an adjustment mechanism becomes essential, leading to an increase in cost.

  Next, an acceleration detection method will be described.

  When the mass of the vibrator 11 is m and an acceleration α in a predetermined direction acts on the vibrator 11, a force F = mα acts in the same direction as the acceleration α. Further, since the displacement x of the elastic support 13 when a force is applied is represented by F = kx, x∝α, and the acceleration can be known by detecting the displacement.

  When acceleration occurs around the first axis (for example, the X axis in the three-dimensional coordinate system), a moment due to inertial force is generated as in the case of angular velocity, and displacement occurs in the twisting direction. This displacement in the twisting direction is detected as a change in capacitance of the electrodes 211-1 to 211-4. Although the acceleration around the first axis and the direction of displacement due to the angular velocity around the second axis are the same, generally it is sufficient to detect the acceleration up to 200 Hz. (Several kHz to several tens of kHz), it can be easily separated by a filter or the like.

  The same applies to the acceleration around the second axis (for example, the Y axis in the three-dimensional coordinate system).

  Here, FIG. 6A shows an example of a driving mode in the inertial sensor 1 that can simultaneously detect multiaxial angular velocities and accelerations. FIGS. 6B and 6C show examples of detection modes in the inertial sensor 1. FIG. FIG. 6 (2) shows a detection mode for detecting rotation around the X axis, and FIG. 6 (2) shows a detection mode for detecting rotation around the Y axis.

  After the inertial sensor 1 is formed by a semiconductor process, it is sealed under reduced pressure in an atmosphere lower than atmospheric pressure using a ceramic package or the like. When using resonance of several kHz to several tens of kHz as in the inertial sensor 1, it is known that the attenuation due to the atmosphere is much larger than the structural attenuation such as internal loss. For this reason, it is possible to improve the Q value of driving and detection by sealing under reduced pressure. At this time, for example, if a high vacuum of 1 Pa or less is used, it is necessary to perform degassing processing and ensure the robustness of the package, which increases the load of the manufacturing process and increases the manufacturing cost. On the other hand, when the degree of vacuum is about 100 Pa or more, the influence of degassing due to heat treatment or the like can be almost ignored, and the package can be simplified.

  Next, an embodiment (a first example of the manufacturing method) relating to the manufacturing method of the inertial sensor of the present invention will be described with reference to FIGS. 7 to 11 show, as an example, a manufacturing process of the inertial sensor 1 including a multi-axis composite sensor including the angular velocity sensor described in the first embodiment.

  As shown in FIG. 7A, a substrate 30 having a three-layer structure in which a first layer 31, a second layer 32, and a third layer 33 are sequentially stacked is used. An example of such a substrate 30 is an SOI substrate. Here, a silicon layer formed on the lower first layer 31, an insulating layer formed on the second layer 32, and a silicon layer formed on the upper third layer 33 were used. For the insulating layer, an insulator such as silicon oxide or silicon nitride can be used. Here, the first and second layers 31 and 32 on both sides are made conductive. For example, conductivity is provided by doping an n-type impurity or a p-type impurity. The upper third layer 33 is formed thinner than the lower first layer 31. This is because an elastic support is formed by an upper silicon layer, and is formed thin in order to provide flexibility when a predetermined thickness is reached. The lower first layer 31 is formed thick to form a mass part (vibrator).

  A multi-axis sensor is manufactured by processing the substrate (SOI substrate) 30. First, as shown in FIG. 7B, the lower surface of the substrate 30 is divided into predetermined blocks by removing the first layer 31 using reactive ion etching or the like. In this etching step, since there is a sufficient etching selectivity between the silicon oxide layer and the silicon layer, the second layer (silicon oxide layer) 32 can be used as an etching stopper.

  Next, as shown in FIG. 7C, the second layer (silicon oxide layer) 32 is etched to remove the silicon oxide layer. At this time, the upper third layer (silicon layer) 33 functions as an etching stopper.

  Next, as shown in FIG. 8A, a first substrate 100 that serves as a lower substrate is prepared. A groove 110 is formed in the first substrate 100, and an electrode 111 is formed in the groove 110.

  Next, as shown in FIG. 8B, the first substrate 100 is bonded to the lower surface of the first layer 31 of the substrate 30. For example, anodic bonding is used for this bonding. Techniques such as silicon / silicon oxide film bonding, silicon / silicon bonding, and metal / metal bonding may be used.

  Subsequently, the third layer 33 is selectively etched from the upper surface of the upper third layer 33 using a mask that covers a part of the vibrator serving as the elastic support and mass part. As a result, as shown in the plan view of FIG. 9 (1) and the schematic configuration sectional view of (2), the vibrator 11, the elastic support 13 that supports the vibrator 11, and the support that supports the elastic support 13 A structure in which the portion 12 is formed on the first substrate 100 is obtained. At this time, since the vibration mass of each mode is common, the vibration frequency can be controlled by adjusting the spring constant of the elastic support 13. As shown in FIG. 3, for example, when it is desired to separate the detection frequencies by 5% on both sides of the drive frequency, ω = 2πf = √ (k / m), where ω is the frequency and f is the vibration frequency. From the relationship of the vibration frequency, k is the spring constant of the elastic support 13, the spring constant may be separated by 21%.

  For example, when controlling by the length of the spring, since the resonance frequency is inversely proportional to the square of the length of the spring, the length of the spring supporting the vibrator is set in the second axial direction with respect to the spring in the first axial direction, for example. By adjusting the length of the spring to √ (1 / 1.1) = 0.95 times, the detection frequencies of the first axis and the second axis can be separated by 10% across the drive frequency.

  Since the spring width does not ideally affect the resonance frequency, it is practically difficult to adjust only by the spring width. It is appropriate to consider to the extent that fine adjustment is performed when an unnecessary mode appears in the aspect ratio with the length.

  When controlling by the thickness of the spring, since the primary resonance frequency of the beam bending is proportional to the thickness of the spring, the second axial spring is 1.1. By adjusting the thickness twice, the detection frequencies of the first axis and the second axis can be separated by 10%.

  When adjusting with the spring material, it is proportional to the square root √ (E / ρ) of the ratio of Young's modulus and density of the material. The detection frequencies of the two axes can be separated by 10%.

  There is also a method of providing a difference in detection frequency by inserting a slit, groove, hole or the like in the spring to weaken the spring.

  Considering the convenience in the actual semiconductor process, it is difficult to make a difference in thickness and material. Therefore, it is realistic to adjust the length and to make a spring such as a slit weak.

  Next, a second substrate 200 as shown in FIG. 10 is prepared, and a wiring groove 210 is processed on the lower surface side thereof. A normal method such as silicon etching can be used for this processing. Further, an electrode (detection electrode) 211 and an electrode (drive electrode) 212 are formed in the groove 210.

  Next, as shown in FIG. 11, the second substrate 200 is bonded to the substrate 30 including the vibrator 11, the elastic support 13, and the like using a bonding method such as anodic bonding. Here, a plurality of weight-like through holes (not shown) are formed in the second substrate 200, and the lower silicon conductive layer can be observed. Here, by depositing a metal such as gold on the upper surface of the second substrate 200, each wiring terminal is formed by depositing a metal layer on the wall surface of the weight-shaped through hole, and unnecessary metal films are removed by etching or the like. Then, the inertial sensor 1 as illustrated can be obtained. Thereafter, the inertial sensor 1 is mounted on a package (not shown).

  According to the manufacturing method of the inertial sensor 1, the Q value can be increased even when the degree of vacuum is close to the atmospheric pressure, so that gas leaks from the outside to the inside of the package or occurs inside the package. Since the influence of gas or the like can be prevented, there is an advantage that the hermetic structure of the package can be simplified and the package cost can be reduced. Further, since the increase in the Q value on the detection side can be suppressed as compared with increasing the Q value by increasing the degree of vacuum, the stabilization time of the damped vibration when the angular velocity is applied can be shortened. There is an advantage that the SN ratio and responsiveness can be improved.

  Next, an embodiment (second example) relating to the inertial sensor of the present invention will be described with reference to FIGS. 12 and 13 show an inertial sensor having an electrode configuration different from that of the first embodiment as an example. Therefore, the inertial sensor 2 of the second embodiment is different from the inertial sensor 1 of the first embodiment only in the configuration of the vibrator, the elastic support, and the electrodes above the vibrator. This is the same as the inertial sensor 1 of the embodiment. 12A is a perspective view showing an outline of a configuration example of the vibrator and the electrodes disposed on the upper portion, and FIG. 12B is a schematic sectional view showing an inertial sensor. (1) of FIG. 13 is a front view showing the positional relationship between the vibrator, the elastic support and the electrodes, and (2) is a plan view.

  As shown in FIGS. 12 and 13, a support portion 12 is formed on the first substrate 100, and one end side of the elastic supports 13-1 to 13-4 is supported by the support portion 12. The vibrator 11 is supported on the other end side of each of the elastic supports 13-1 to 13-4 in a state of being separated from the first substrate 100 and the second substrate 200 described later. In addition, a displacement detection unit 14 that detects the displacement of the vibrator 11 and outputs a signal is provided, for example, on the side of the second substrate 200 facing the vibrator 11. Here, the displacement of the vibrator 11 is detected by the capacitance change between the vibrators 11 and the electrodes 211-1 to 211-4 provided above the vibrator 11. Further, the driving electrodes 212-1 and 212-2 provided above the vibrator 11 are alternately applied with a voltage so that the vibrator 11 is rotated around the first axis (for example, the X axis) like a pendulum. Drive. Further, for example, an electrode 111 for monitoring the driving of the vibrator 11 is formed on the first substrate 100 below the vibrator 11.

  Each of the elastic supports 13-1 to 13-4 has, for example, a diagonal direction of the vibrator 11 so that the vibrator 11 has different detection mode frequencies on the low frequency side and the high frequency side of the drive mode frequency. The elastic supports (first elastic supports) 13-1 and 13-3 are different from other diagonal elastic supports (second elastic supports) 13-2 and 13-4 of the vibrator 11. The spring constant is formed. In other words, the elastic support at the position facing the diagonal direction of the vibrator has the same spring constant.

  Next, an operation method of the inertial sensor 2 will be described. As shown in FIG. 12, a voltage is alternately applied to the electrodes 212-1 and 212-2, and the vibrator 11 is driven around the first axis like a pendulum. When an angular velocity is applied around the first axis with respect to this driving direction, a Coriolis force acts in the direction of the third axis (for example, the Z axis), and the vibrator moves in the third axis direction according to the magnitude of the angular velocity. This is detected by a change in capacitance of the electrodes 211-1 to 211-4. When an angular velocity is applied around the third axis, Coriolis force acts in the direction of the second axis (for example, the Y axis). At this time, the vibrator 11 performs a pendulum motion because the support point and the center of gravity are displaced. When this movement is performed, the capacitance between the electrodes 211-1 to 211-4 and the vibrator 11 changes, and for example, the detection capacitance increases between the electrodes 211-1 and 211-4 and the vibrator 11. The detection capacitance decreases between the electrodes 211-2 and 211-3 and the vibrator 11. By taking this difference, the angular velocity around the first axis can be separated.

  Further, as in the first embodiment, the vibration mass of each mode is common, and by adjusting the spring constant of the elastic supports 13-1 to 13-4 related to the mode shape, it is shown in FIG. Thus, two detection frequencies can be arranged on both sides across the drive frequency.

It is drawing which showed one Embodiment (1st Example) which concerns on the inertial sensor of this invention. It is drawing which showed one Embodiment (1st Example) which concerns on the inertial sensor of this invention. It is a figure explaining the relationship between the vibration mode and detection mode of the inertial sensor of 1st Example. It is a perspective schematic diagram explaining the vibration mode of the inertial sensor of 1st Example. It is a relationship diagram between a gain and a degree of detuning. It is a perspective schematic diagram explaining the detection mode of the inertial sensor of 1st Example. It is a manufacturing-process figure which showed one Embodiment (1st Example) which concerns on the manufacturing method of the inertial sensor of this invention. It is a manufacturing-process figure which showed one Embodiment (1st Example) which concerns on the manufacturing method of the inertial sensor of this invention. It is a manufacturing-process figure which showed one Embodiment (1st Example) which concerns on the manufacturing method of the inertial sensor of this invention. It is a manufacturing-process figure which showed one Embodiment (1st Example) which concerns on the manufacturing method of the inertial sensor of this invention. It is a manufacturing-process figure which showed one Embodiment (1st Example) which concerns on the manufacturing method of the inertial sensor of this invention. It is drawing which showed one Embodiment (2nd Example) which concerns on the inertial sensor of this invention. It is drawing which showed one Embodiment (2nd Example) which concerns on the inertial sensor of this invention. It is a figure explaining the relationship between the vibration mode and detection mode of the conventional inertial sensor.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Inertial sensor, 11 ... Vibrator, 14 ... Displacement detection part, 100 ... Board | substrate

Claims (5)

A vibrator supported to be displaceable in a floating state with respect to the substrate;
And a displacement detector for detecting a displacement of the vibrator,
Said vibrator,
Supported by the first elastic support and the second elastic support having different spring constants so as to have different resonance frequencies on the low frequency side and the high frequency side of the drive mode frequency, respectively .
The first elastic support body is formed to extend in the first axial direction so as to have a resonance frequency on the low frequency side, and the first elastic support body has the resonance frequency on the high frequency side. Formed extending in a second axial direction perpendicular to the axial direction,
Driven by the drive mode frequency in the third axis direction orthogonal to the first axis direction and the second axis direction by the excitation means;
The displacement detector is
Detecting a displacement in the twisting direction of the first axis and detecting a displacement in the twisting direction of the second axis;
Inertial sensor. The length of the spring of the first elastic support is made different from the length of the second elastic support spring, and the resonance frequency of the first elastic support is made different from the resonance frequency of the second elastic support. The inertial sensor according to 1. The excitation means includes
Formed with a drive electrode fixed to the substrate,
An alternating voltage having a driving mode frequency is applied between the vibrator suspended by the first elastic support and the second elastic support relative to the substrate, and the driving electrode and the vibrator are The vibrator is periodically driven by generating an electrostatic force between them
The inertial sensor according to claim 1. The vibrator is sealed in a decompressed atmosphere.
The inertial sensor according to claim 1. And displaceably supported vibrator in suspension state with respect to the substrate, in the manufacturing method of an inertial sensor comprising a displacement detector, the detecting the displacement of the vibrator,
The vibrator is supported by the first elastic support and the second elastic support having different spring constants so as to have different resonance frequencies on the low frequency side and the high frequency side of the drive mode frequency,
The first elastic support is formed to extend in the first axial direction so as to have a resonance frequency on the low frequency side, and the second elastic support is configured to have the resonance frequency on the high frequency side. Extending in the second axial direction perpendicular to the axial direction,
Forming electrodes as excitation means for driving the vibrator at the drive mode frequency in a third axis direction orthogonal to the first axis direction and the second axis direction;
As the displacement detector,
Detecting a displacement in the twisting direction of the first axis, and forming a detector for detecting a displacement in the twisting direction of the second axis;
Inertial sensor manufacturing method.

Images