JP2013007743A - Mems sensor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a MEMS sensor which is small, and by which high sensitivity is obtained in comparison with a conventional case.SOLUTION: A movable electrode part has: a plurality of movable support parts 50 arranged at intervals in the X1-X2 direction; and a plurality of movable armatures 60 which extend from the sides of the respective movable support parts 50, and are arranged to the respective movable support part 50 at intervals in the Y1-Y2 direction. A fixed electrode part has: a plurality of fixed support parts 51 arranged at intervals in the X1-X2 direction; and a plurality of fixed armatures 62 which extend from the sides of the respective fixed support parts 51, and are arranged to the respective fixed support parts 51 at intervals in the Y1-Y2 direction. The plurality of movable support parts and the plurality of fixed support parts are alternately arranged in the X1-X2 direction, the movable support parts and the fixed support parts adjacent to each other are made into sets 66, and the movable armatures 60 and the fixed armatures 62 are alternately arranged between the movable support parts and the fixed support parts of the respective sets.

Description

  The present invention relates to a capacitance type MEMS sensor formed by processing a silicon layer, and more particularly to an electrode structure.

  An electrostatic capacitance type MEMS sensor has a weight part, a movable electrode part, and a fixed electrode part. For example, when acceleration acts on the MEMS sensor, the weight part and the movable electrode part move, and a capacitance change occurs between the movable electrode part and the fixed electrode part. It is possible to detect acceleration based on this change in capacitance.

  FIG. 31 is an explanatory view (plan view) schematically showing a movable electrode portion and a fixed electrode portion of a conventional MEMS sensor. This MEMS sensor detects acceleration in the in-plane direction, and can detect acceleration acting in the X1-X2 direction, for example.

  FIG. 31A shows the stationary state of the MEMS sensor. As shown in FIG. 31 (a), the MEMS sensor includes a first detection unit 1 and a second detection unit 2, and each of the first detection unit 1 and the second detection unit 2 has a plurality of comb teeth. The fixed electrodes 1a and 2a and the plurality of comb-like movable electrodes 1b and 2b are alternately arranged at intervals in the X1-X2 direction.

  As shown in FIG. 31 (a), in the first detector 1, a first gap a0 having a predetermined distance between the movable electrode 1b and the fixed electrode 1a located on the X1 side with respect to the movable electrode 1b. Is provided. Further, in the second detector 2, a second gap b0 having a predetermined distance is provided between the movable electrode 2b and the fixed electrode 2a located on the X2 side with respect to the movable electrode 2b.

  As shown in FIG. 31B, when the movable electrode 1b in the first detection unit 1 and the movable electrode 2b in the second detection unit 2 move in the X1 direction by the action of acceleration, respectively, FIG. In FIG. 31A, the first gap a1 is smaller than the gap a0 in the stationary state, while the second gap b1 is larger than the gap b0 in the stationary state in FIG.

  As shown in FIG. 31C, when the movable electrode 1b in the first detection unit 1 and the movable electrode 2b in the second detection unit 2 are moved in the X2 direction by the action of acceleration, respectively, ) Is larger than the gap a0 in the stationary state of FIG. 31A, while the second gap b2 is smaller than the gap b0 in the stationary state of FIG. .

  As shown in FIG. 31 (b) and FIG. 31 (c), when the acceleration acts in the X1-X2 direction, the gaps a, b between the movable electrodes 1b, 2b and the fixed electrodes 1a, 2a change. This changes the capacitance. At this time, the capacitance change in the first detection unit 1 and the capacitance change in the second detection unit 2 are reversed.

  Patent Document 1 describes an electrode structure in which a capacitance change is caused by a change in a gap between a movable electrode and a fixed electrode, as in FIG. 31 described above. In addition, Patent Document 1 also describes an electrode structure that can detect a change in capacitance due to a change in the facing area between the movable electrode and the fixed electrode. Patent Documents 2 and 3 also describe electrode structures that cause a change in capacitance due to the variation in the facing area.

  Patent Document 4 discloses an acceleration sensor having comb-like electrodes. Patent Document 4 discloses an electrode structure in which support parts extending in parallel are provided in both the movable electrode part and the fixed electrode part, and comb-like electrode elements are alternately arranged from the inner surface side of each support part. ing.

JP 2007-139505 A Japanese Patent Laid-Open No. 2002-22446 JP-A-10-313123 JP 2010-513888 A

  However, with the conventional electrode structure, it has been difficult to sufficiently improve the sensitivity as well as downsizing the MEMS sensor.

  In addition, each patent document does not disclose an electrode structure for improving the linearity of sensitivity.

  Therefore, the present invention solves the above-described conventional problems, and an object of the present invention is to provide a MEMS sensor that is smaller than the conventional one and that can obtain high sensitivity and good linearity.

The present invention provides a MEMS sensor having a movable electrode portion and a fixed electrode portion.
When the two directions orthogonal to each other in the horizontal plane are defined as the first direction and the second direction, the first direction is the moving direction of the movable electrode part,
The movable electrode portion includes a plurality of movable support portions that are arranged with an interval in the first direction and extend in the second direction, and each movable support portion toward the first direction. A plurality of movable electrode elements extending from the side portions of the movable support portions and spaced from each other in the second direction at each movable support portion,
The fixed electrode portion is arranged with a gap in the first direction, and a plurality of fixed support portions whose extending direction from the proximal end side to the distal end side is opposite to the movable support portion; A plurality of fixed electrode elements extending in a direction opposite to the extending direction of the movable electrode element from a side part of the support part, and arranged at intervals in the second direction at each fixed support part; Have
A plurality of the movable support portions and a plurality of the fixed support portions are alternately arranged at intervals in the first direction, and the adjacent movable support portions and the fixed support portions are paired. A plurality of movable electrode elements and a plurality of fixed electrode elements are alternately arranged in the second direction with an interval between each set of the movable support part and the fixed support part. And
By moving the movable electrode portion in the first direction, it is possible to detect a change in capacitance based on a change in the facing area between each movable electrode and each fixed electrode.
In a stationary state, the distance in the first direction from the side of the movable support portion to the tip of the fixed electrode is L1, and the width of the movable electrode and the fixed electrode in the second direction When the dimension is T1, and the maximum movable distance when the movable electrode moves in the direction approaching the fixed support portion is L2, (L1-L2) ² / T1 is 2.2 (μm) or more. It is characterized by being.

In the present invention, in the stationary state, an interval in the first direction between the movable support portion and the fixed support portion of the adjacent sets is L3, and the movable electrode and the fixed electrode are When the width dimension in the direction of 2 is T1, and the maximum movable distance when the movable electrode moves in the direction approaching the fixed support portion of the adjacent group is L4, (L3-L4) ² × T1 is preferably 4 (μm ³ ) or more.

Alternatively, the present invention provides a MEMS sensor having a movable electrode portion and a fixed electrode portion.
When the two directions orthogonal to each other in the horizontal plane are defined as the first direction and the second direction, the first direction is the moving direction of the movable electrode part,
The movable electrode portion includes a plurality of movable support portions that are arranged with an interval in the first direction and extend in the second direction, and each movable support portion toward the first direction. A plurality of movable electrode elements extending from the side portions of the movable support portions and spaced from each other in the second direction at each movable support portion,
The fixed electrode portion is arranged with a gap in the first direction, and a plurality of fixed support portions whose extending direction from the proximal end side to the distal end side is opposite to the movable support portion; A plurality of fixed electrode elements extending in a direction opposite to the extending direction of the movable electrode element from a side part of the support part, and arranged at intervals in the second direction at each fixed support part; Have
A plurality of the movable support portions and a plurality of the fixed support portions are alternately arranged at intervals in the first direction, and the adjacent movable support portions and the fixed support portions are paired. A plurality of movable electrode elements and a plurality of fixed electrode elements are alternately arranged in the second direction with an interval between each set of the movable support part and the fixed support part. And
By moving the movable electrode portion in the first direction, it is possible to detect a change in capacitance based on a change in the facing area between each movable electrode and each fixed electrode.
In a stationary state, an interval in the first direction between the movable support portion and the fixed support portion of the adjacent sets is L3, and the movable electrode element and the fixed electrode element are in the second direction. (L3−L4) ² × T1 is 4 where L1 is the maximum movable distance when the movable electrode element moves in a direction approaching the fixed support portion of the adjacent group. (Μm ³ ) or more.

  In the present invention, a plurality of movable support portions constituting the movable electrode portion and a plurality of fixed support portions constituting the fixed electrode portion are alternately arranged in the first direction in the same detection portion. Then, adjacent movable support portions and fixed support portions are made into the same set, and a plurality of movable electrode elements and fixed electrode elements are arranged in the second direction between the movable support portions and the fixed support portions in each set. Alternatingly arranged. As a result, a large number of movable electrodes and fixed electrodes can be arranged efficiently in the detection unit, and a small size and high sensitivity can be obtained as compared with the conventional electrode structure.

  In the electrode structure in which the gap (distance) between the movable electrode and the fixed electrode is changed as shown in FIG. 31, the capacitance is inversely proportional to the gap (distance). As shown as an example, when the sensitivity is increased, the capacitance change is abrupt with respect to the variation in distance, and the effective detection range (dynamic range) of the physical quantity by the MEMS sensor is narrowed. Further, there is a problem that the linearity of sensitivity within the effective detection range is deteriorated.

  On the other hand, in the electrode structure that changes the capacitance by changing the facing area between the movable electrode electrode and the fixed electrode electrode as in the present invention, the capacitance is proportional to the facing area. Compared to the method of changing the capacitance by changing the distance between the two, the effective detection range (dynamic range) of the physical quantity can be widened and the linearity of the sensitivity within the effective detection range can be effectively improved. Become.

Moreover, in the present invention, (L1-L2) ² / T1 is set to 2.2 (μm) or more, and (L3-L4) ² × T1 is set to 4 (μm ³ ) or more. This makes it possible to improve the sensitivity linearity more effectively.

In the present invention, it has a first detection unit and a second detection unit, each of the detection unit includes the movable electrode unit and the fixed electrode unit,
The extending direction of the movable electrode element and the fixed electrode element in the first direction is reversed between the first detection unit and the second detection unit, and is obtained by the first detection unit. It is preferable that a differential output can be obtained by the change in capacitance and the change in capacitance obtained by the second detection unit.

  Further, in the above configuration, each of the first detection unit and the second detection unit is divided into a plurality of detection regions, and each of the detection regions includes the movable electrode unit and the fixed electrode unit. Is preferred. Thereby, while maintaining the intensity | strength of each electrode part, a movable electrode part and a fixed electrode part can be arrange | positioned efficiently, while being able to implement | achieve size reduction of a MEMS sensor, a sensitivity can be raised more effectively.

According to the present invention, the movable body including the movable electrode portion and an anchor portion connected via a spring portion are provided, and the first portion of the movable body from a stationary state is interposed between the movable body and the anchor portion. A gap is formed that regulates the movable dimension in the direction of
The spring portion is formed to extend in the second direction and be folded between the anchor portion and the movable body so as to suppress vibration of the movable body in the second direction. preferable. Thereby, the stability of sensitivity can be improved.

  According to the MEMS sensor of the present invention, as compared with the conventional electrode structure, it is small, and high sensitivity and good linearity can be obtained.

FIG. 1 is a plan view showing a functional layer of a MEMS sensor according to an embodiment of the present invention. 2A is a partially enlarged plan view of a stationary state showing a part of the movable electrode portion and the fixed electrode portion shown in FIG. 1, and FIG. 2B is subjected to acceleration from the stationary state of FIG. It is a partial enlarged plan view which shows the state which the movable electrode part moved to X2 direction. 3A is a partially enlarged plan view of a stationary state showing a part of the movable electrode portion and the fixed electrode portion shown in FIG. 1, and FIG. 3B is a view showing the movable electrode portion in the X1 direction from the stationary state shown in FIG. FIG. 3C is a partially enlarged plan view showing a state in which the movable electrode portion has moved in the X2 direction from the stationary state shown in FIG. FIG. 4 is a partial vertical cross-sectional view of the MEMS sensor shown in FIG. 1 taken along the line AA and viewed from the direction of the arrow. FIG. 5A is a simulation result showing the relationship between the acceleration and the capacitance in the comparative example. FIG. 5B is the result of FIG. 5A without changing the size of the detection part of the MEMS sensor. FIG. 6 is a simulation result showing a relationship between acceleration and capacitance when sensitivity is increased. FIG. 6 shows a differential output curve between the capacitance change in the graph (1) in FIG. 5A and the capacitance change in the graph (2). FIG. 7 is a simulation result showing the relationship between acceleration and sensitivity. FIG. 8 is a simulation result showing a relationship between acceleration and capacitance in the example. FIG. 9 shows a differential output curve between the capacitance change shown in the graph (3) in FIG. 8 and the capacitance change shown in the graph (4). FIG. 10 is a simulation result showing the relationship between acceleration and sensitivity in the example. It is a simulation result which shows the relationship between the acceleration in the Example which shows the relationship between the magnitude | size of L1 (however, L2 was 2 micrometers, L3 was 5 micrometers), and the maximum shake width of a sensitivity. It is a simulation result which shows the relationship between L1-L2 (L2 = 2micrometer) and the maximum fluctuation width of a sensitivity. It is a simulation result which shows the relationship between (L1-L2) ² / T1 (L2 = 2micrometer) and the maximum fluctuation width of a sensitivity. It is the simulation result which shows the relationship between the acceleration in the Example which shows the relationship between the magnitude | size of L1 (however, L2 was 3 micrometers and L3 was 6 micrometers), and the maximum shake width of a sensitivity. It is a simulation result which shows the relationship between L1-L2 (L2 = 3micrometer) and the maximum fluctuation width of a sensitivity. It is a simulation result which shows the relationship between (L1-L2) ² / T1 (L2 = 3 micrometer) and the maximum fluctuation width of a sensitivity. It is a simulation result which shows the relationship between the acceleration in the Example which shows the relationship between the magnitude | size of L1 (however, L2 was 4 micrometers and L3 was 7 micrometers), and the maximum fluctuation width of a sensitivity. It is a simulation result which shows the relationship between L1-L2 (L2 = 4micrometer) and the maximum fluctuation width of a sensitivity. It is a simulation result which shows the relationship between (L1-L2) ² / T1 (L2 = 4 μm) and the maximum amplitude of sensitivity. It is a simulation result which shows the relationship between the acceleration in the Example which shows the relationship between the magnitude | size of L3 (however, L1 was 5 micrometers and L4 was 2 micrometers), and the maximum fluctuation width of a sensitivity. It is a simulation result which shows the relationship between L3-L4 (L4 = 2micrometer) and the maximum fluctuation width of a sensitivity. It is a simulation result which shows the relationship between (L3-L4) ² (L4 = 2 μm) × T1 and the maximum sensitivity fluctuation range. It is a simulation result which shows the relationship between the acceleration in the Example which shows the relationship between the magnitude | size of L3 (however, L1 was 6 micrometers and L4 was 3 micrometers), and the maximum shake width of a sensitivity. It is a simulation result which shows the relationship between L3-L4 (L4 = 3micrometer) and the maximum fluctuation width of a sensitivity. It is a simulation result which shows the relationship between (L3-L4) ² (L4 = 3 μm) × T1 and the maximum sensitivity amplitude. It is a simulation result which shows the relationship between the acceleration in the Example which shows the relationship between the magnitude | size of L3 (however, L1 was 7 micrometers and L4 was 4 micrometers), and the maximum fluctuation width of a sensitivity. It is a simulation result which shows the relationship between L3-L4 (L4 = 4micrometer) and the maximum fluctuation width of a sensitivity. It is a simulation result which shows the relationship between (L3-L4) ² (L4 = 4 μm) × T1 and the maximum sensitivity amplitude. It is explanatory drawing regarding the technical meaning of (L1-L2) ² / T1. (L3-L4) It is explanatory drawing regarding the technical meaning of ² * T1. It is explanatory drawing (plan view) for demonstrating the capacitance change by the movement to the X1-X2 direction of a movable electrode while showing the conventional electrode structure typically.

  FIG. 1 is a plan view showing a functional layer of a MEMS sensor according to an embodiment of the present invention. FIG. 2A is a partially enlarged plan view in a stationary state showing a part of a movable electrode portion and a fixed electrode portion shown in FIG. 2 (b) is a partially enlarged plan view showing a state in which the movable electrode portion is moved in the X2 direction by receiving acceleration from the stationary state of FIG. 2 (a), and FIG. 3 (a) is a movable electrode shown in FIG. FIG. 3B is a partially enlarged plan view showing a state in which the movable electrode portion has moved in the X2 direction from the stationary state shown in FIG. 1, and FIG. FIG. 4C is a partially enlarged plan view showing a state in which the movable electrode portion has moved in the X1 direction from the stationary state shown in FIG. 1, and FIG. 4 is a partial longitudinal sectional view of the MEMS sensor in this embodiment.

  As shown in FIG. 1, the MEMS sensor S has, for example, a rectangular shape in which the X1-X2 direction (first direction) is a long side and the Y1-Y2 direction (second direction) is a short side. The MEMS sensor S shown in FIG. 1 constitutes an uniaxial detection acceleration sensor for detecting acceleration acting in the X1-X2 direction.

  As shown in FIG. 4, the MEMS sensor S includes an SOI substrate 13 in which a support substrate 10 and a functional layer 11 are stacked with an insulating layer 12 interposed therebetween, and an SOI substrate 13 that faces the SOI substrate 13 in the height direction via a metal connection portion 14. And the wiring board 15 bonded to each other. Here, FIG. 1 shows only the functional layer 11 in the MEMS sensor S, and further shows the weight portion 22 and the electrode portion located inside the frame body 16 (see FIG. 4) in the functional layer 11.

For example, the support substrate 10 and the functional layer 11 are made of silicon, and the insulating layer 12 is made of SiO ₂ .

  As shown in FIGS. 1 and 4, the functional layer 11 is formed by separating the fixed electrode portions 20a to 20d, the movable electrode portions 21a to 21d, the weight portion 22 and the frame body 16 from the silicon substrate. Among these, the movable electrode portions 21a to 21d and the weight portion 22 are integrally formed, and constitute a movable body 25 supported so as to be movable in the X1-X2 direction, which is the first direction, by application of acceleration.

  As shown in FIG. 1, the planar shape of the functional layer 11 is 180 degrees rotationally symmetric with respect to the center (centroid) O in the X1-X2 direction and the Y1-Y2 direction, and passes through the center O in the X direction. It is symmetrical in the vertical direction (Y1-Y2 direction) with respect to the extending line.

  As shown in FIG. 1, the first detection unit 23 is provided on the X1 side of the center O, and the second detection unit 24 is provided on the X2 side. Further, the first detection unit 23 is divided into a first detection region 23a on the Y1 side and a first detection region 23b on the Y2 side. The second detection unit 24 is divided into a second detection region 24a on the Y1 side and a second detection region 24b on the Y2 side.

  As shown in FIG. 1, the first detection region 23a on the Y1 side includes a fixed electrode portion 20a and a movable electrode portion 21a. The first detection region 23b on the Y2 side includes a fixed electrode portion 20b and a movable electrode portion 21b. The second detection region 24a on the Y1 side includes a fixed electrode portion 20c and a movable electrode portion 21c. The second detection area 24b on the Y2 side is composed of a fixed electrode portion 20d and a movable electrode portion 21d. The insulating layer 12 is not formed between each movable electrode part 21a-21d and the support substrate 10 (refer FIG. 4). On the other hand, the fixed electrode portions 20 a to 20 d are fixed to the support substrate 10 via the insulating layer 12.

  A portion excluding the first detection unit 23 and the second detection unit 24 in the movable region is the weight unit 22. In FIG. 1, the weight portion 22 is located around the first detection portion 23 and the second detection portion 24.

  As shown in FIG. 1, the weight portion 22 includes an X1 side region 22a, a Y1 side region 22b, an X2 side region 22c, and a Y2 side region 22d.

  As shown in FIG. 1, a first anchor portion 26 formed separately from the weight portion 22 is provided on the Y1 side of the weight portion 22 from the Y1 side region 22b. The first anchor portion 26 is formed in a long and narrow shape in the X1-X2 direction. The first anchor portion 26 is fixed via the support substrate 10 and the insulating layer 12 shown in FIG.

  As shown in FIG. 1, a second anchor portion 27 formed separately from the weight portion 22 is provided on the Y2 side of the weight portion 22 from the Y2 side region 22d. The second anchor portion 27 is formed in a long and narrow shape in the X1-X2 direction. The second anchor portion 27 is fixed via the support substrate 10 and the insulating layer 12 shown in FIG.

  As shown in FIG. 1, the X1 side region 22 a of the weight portion 22 extends further to the X1 side than the first anchor portion 26 and the second anchor portion 27. A gap (interval) 28 having a predetermined width is formed between the Y1 side end portion 22a1 of the X1 side region 22a and the first anchor portion 26. Further, a gap 29 having a predetermined width is formed between the Y2 side end portion 22a2 of the X1 side region 22a and the second anchor portion 27. The widths of the gaps 28 and 29 in the X1-X2 direction are the same, and the gaps 28 and 29 regulate the dimension by which the weight portion 22 can move in the X1 direction from the stationary state of FIG.

  Further, as shown in FIG. 1, the X2 side region 22 c of the weight portion 22 extends further to the X2 side than the first anchor portion 26 and the second anchor portion 27. A gap 30 having a predetermined width is formed between the Y1 side end portion 22c1 of the X2 side region 22c and the first anchor portion 26. Further, a gap 31 having a predetermined width is formed between the Y2 side end portion 22c2 of the X2 side region 22c and the second anchor portion 27. The widths of the gaps 30 and 31 in the X1-X2 direction are the same, and the gaps 30 and 31 regulate the dimension by which the weight portion 22 can move in the X2 direction from the stationary state of FIG.

  Further, as shown in FIG. 1, in the X1 side region 22 a of the weight portion 22, a space portion 35 that extends inward in the Y1-Y2 direction from the gaps 28 and 29 and has a slightly wider width than the gaps 28 and 29. , 36 are formed. In addition, in the X2 side region 22c of the weight portion 22, space portions 37 and 38 that extend inward in the Y1-Y2 direction from the gaps 30 and 31 and have a slightly wider width than the gaps 30 and 31 are formed. Yes.

  And in each space part 35-38, the spring parts 40-43 which connect each anchor part 26 and 27 and the weight part 22 are formed. Each spring part 40-43 is the location which cut out the silicon substrate integrally with each anchor part 26 and 27 and the weight part 22, and gave the elasticity to a X1-X2 direction. The insulating layer 12 is not formed between the spring portions 40 to 43 and the weight portion 22 and the support substrate 10 (see FIG. 4). Therefore, when the acceleration is received, the weight portion 22 can move in the X1-X2 direction by the elastic deformation of the spring portions 40-43.

  As shown in FIG. 1, the spring portions 40 to 43 are formed long in the Y1-Y2 direction, and are formed by being folded back between the anchor portions 26 and 27 and the weight portion 22. Thereby, each spring part 40-43 has rigidity in the Y1-Y2 direction, and is suppressing the vibration to the Y1-Y2 direction of the weight part 22. FIG.

  As shown in FIG. 1, the movable electrode portions 21 a to 21 d are formed integrally with the weight portion 22. Each of the movable electrode portions 21a to 21d is provided with a plurality of movable support portions 50 extending in the Y1-Y2 direction at intervals in the X1-X2 direction. Each movable support portion 50 extends inward from the Y1 side region 22b and the Y2 side region 22d of the weight portion 22. In FIG. 1, only one movable support portion 50 is assigned a symbol to each movable electrode portion 21 a to 21 d. Further, as shown in FIG. 1, each of the fixed electrode portions 20a to 20d is arranged to be spaced from the fixed base portion 52 in the X1-X2 direction and extend from the fixed base portion 52 in the Y1-Y2 direction. A plurality of fixed support portions 51 are provided. In FIG. 1, only one fixed support portion 51 is assigned a symbol to each of the fixed electrode portions 20 a to 20 d. In each detection region 23a, 23b, 24a, 24b, the extending direction of the movable support portion 50 and the fixed support portion 51 is opposite. In each detection region 23a, 23b, 24a, 24b, a plurality of movable support portions 50 and a plurality of fixed support portions 51 are alternately arranged at intervals in the X1-X2 direction.

  As shown in FIG. 4, each fixed base 52 is fixed via the support substrate 10 and the insulating layer 12. The insulating layer 12 may be interposed between each fixed support portion 51 and the support substrate 10, but since the fixed support portion 51 is thin, the insulating layer 12 between the fixed support portion 51 and the support substrate 10 is removed by etching. Accordingly, the fixed support portion 51 is in a state of being lifted from the support substrate 10 as in the case of the movable support portion 50 as shown in FIG. However, since each fixed support portion 51 is connected to the fixed base portion 52, it does not move in the X1-X2 direction even if it receives acceleration.

  As shown in FIG. 1, the X1-X2 direction side portions of the movable support portions 50 and the X1-X2 direction side portions of the fixed support portions 51 are comb-like movable with an interval in the Y1-Y2 direction. Electrodes and fixed electrodes are formed. The movable electrode and the fixed electrode will be described with reference to FIGS.

  FIG. 2A shows a first detection region 23a and a second detection region 24a around II surrounded by a circle shown in FIG. As shown in FIG. 2 (a), in the first detection region 23a, an elongated movable support portion 50 is formed in a straight line shape from the inner portion 22b1 of the Y1 side region 22b of the weight portion 22 toward the Y2 direction. In the first detection region 23a, a plurality of movable support portions 50 are formed at intervals in the X1-X2 direction. Similarly, in the second detection region 24a, an elongated movable support portion 50 is formed in a straight line shape from the inner side portion 22b1 of the Y1 side region 22b of the weight portion 22 toward the Y2 direction. In the second detection region 24a, a plurality of movable support portions 50 are formed at intervals in the X1-X2 direction.

  As shown in FIG. 2A, the movable electrode 60 extending in the X2 direction from the X2-side end 50a of the movable support 50 provided in the first detection region 23a is spaced in the Y1-Y2 direction. A plurality are formed. In FIG. 2A, only one movable electrode 60 is assigned a reference to the movable support portion 50.

  As shown in FIG. 2A, the movable electrode 61 extending in the X1 direction from the X1-side end 50b of the movable support 50 provided in the second detection region 24a is spaced in the Y1-Y2 direction. Are formed. In FIG. 2A, only one movable electrode 61 is assigned a reference to the movable support portion 50. As shown in FIG. 2A, the length dimension of the movable electrode elements 60 and 61 in the X1-X2 direction is sufficiently shorter than the length dimension of the movable support portion 50 in the Y1-Y2 direction. Yes.

  Further, as shown in FIG. 2A, in each of the detection regions 23a and 24a, an elongated fixed support portion 51 is formed in a straight line shape from the fixed base portion 52 toward the Y1 direction. As shown in FIGS. 1 and 3, a plurality of fixed support portions 51 are formed in the detection regions 23 a and 24 a at intervals in the X1-X2 direction. Each fixed support portion 51 is a movable support portion 50. And are arranged alternately.

  As shown in FIG. 2A, fixed electrode elements 62 extending in the X1 direction from the X1 side end portion 51b of the fixed support portion 51 provided in the first detection region 23a are spaced apart in the Y1-Y2 direction. A plurality are formed. As shown in FIG. 2A, the fixed electrode elements 62 are alternately arranged in the Y1-Y2 direction with a gap from the movable electrode element 60. In FIG. 2A, only one fixed electrode 62 is attached to the fixed support portion 51.

  Further, as shown in FIG. 2A, the fixed electrode element 63 extending in the X2 direction from the X2 side end portion 51a of the fixed support portion 51 provided in the second detection region 24a is spaced in the Y1-Y2 direction. Are formed. As shown in FIG. 2A, the fixed electrode elements 63 are alternately arranged in the Y1-Y2 direction with a gap from the movable electrode element 61. In FIG. 2A, only one fixed electrode 63 is attached to the fixed support portion 51. As shown in FIG. 2A, the length dimension of the fixed electrode elements 62 and 63 in the X1-X2 direction is sufficiently shorter than the length dimension of the fixed support portion 51 in the Y1-Y2 direction. Yes.

  In FIG. 2A, the electrode structure of the first detection region 23a on the Y1 side and the second detection region 24a on the Y1 side illustrated in FIG. 1 has been described, but the first detection region 23b on the Y2 side and the second detection region on the Y2 side are illustrated. Each electrode structure in the detection region 24b is in a line-symmetric relationship with the electrode structure in FIG. 2A with respect to a line passing through the center O and extending in the X1-X2 direction.

FIG. 3A shows an electrode structure around III surrounded by a circle shown in FIG.
As shown in FIG. 3A, in the same detection region 23a, each movable support portion 50 and each fixed support portion 51 that are alternately arranged in the X1-X2 direction are grouped 66 one by one. A plurality of movable electrode elements 60 and a plurality of fixed electrode elements 62 are alternately arranged at intervals in the Y1-Y2 direction between the movable support part 50 and the fixed support part 51 of each set 66. Yes.

  In the stationary state shown in FIG. 3A (the state where no physical quantity is applied), the distance in the X1-X2 direction from the X2 side end 50a of the movable support 50 to the tip 62a of the fixed electrode 62 is L1. is there. The width dimension of the movable electrode 60 and the fixed electrode 62 in the Y1-Y2 direction is T1. In FIG. 3A, the width dimension of the movable electrode 60 is typically indicated by T1.

  Further, in the stationary state shown in FIG. 3A, an interval in the X1-X2 direction between the movable support 50 and the fixed support 51 of the adjacent set 66 is indicated by L3. In the stationary state shown in FIG. 3A, the overlapping length of the movable electrode 60 and the fixed electrode 62 is L5.

  Moreover, each movable electrode element 60 is located at the approximate center between the fixed electrode elements 62 and 62 located on both sides in the Y1-Y2 direction.

  As shown in FIG. 3B, when the MEMS sensor S receives acceleration and the weight portion moves in the X2 direction, each movable support portion 50 also moves in the X2 direction. The facing area between each fixed electrode 62 (area facing in the Y1-Y2 direction) is larger than that in the stationary state of FIG. If the maximum movable distance when the movable support portion 50 moves in the X2 direction is L2, and the movable support portion 50 moves in L2 in FIG. 3B, the stationary movement shown in FIG. The interval that was L1 in the state becomes L1-L2. Further, the interval L3 in the stationary state of FIG. 3A becomes L3 + L2.

  As shown in FIG. 3C, when the MEMS sensor S receives acceleration and the weight portion 22 moves in the X1 direction, each movable support portion 50 also moves in the X1 direction, and each movable electrode in each set 66 The facing area between the child 60 and each fixed electrode 62 is smaller than that in the stationary state of FIG. At this time, if the maximum movable movable distance when the movable support portion 50 moves in the X1 direction is L4, and the movable support portion 50 moves L4 in FIG. 3B, the stationary movement of FIG. The interval that was L3 in the state becomes L3-L4. Further, the interval L1 in the stationary state of FIG. 3A becomes L1 + L4.

  The distance variation relationships shown in FIGS. 3A, 3B, and 3C are the same in the first detection region 23b on the Y2 side shown in FIG.

  On the other hand, as shown in FIG. 3B, when the MEMS sensor S receives acceleration and the weight portion moves in the X2 direction, the first detection regions 23a and 23b have movable electrodes as described in FIG. 3B. Although the opposing area between the element 60 and the fixed electrode element 62 increases, the opposing area between the movable electrode element 61 and the fixed electrode element 63 decreases in the second detection regions 24a and 24b. As shown in FIGS. 1 and 2, in the first detection areas 23a and 23b and the second detection areas 24a and 24b, the X1-X2 directions of the movable electrode elements 60 and 61 and the fixed electrode elements 62 and 63 are used. This is because the extending direction to is reversed.

  Thereby, when the opposing area between the movable electrode 60 and the fixed electrode 62 increases in the first detection regions 23a and 23b and the capacitance increases, the second detection regions 24a and 24b The facing area between the fixed electrode elements 63 is reduced, and the capacitance is reduced. On the other hand, when the opposing area between the movable electrode 60 and the fixed electrode 62 decreases in the first detection regions 23a and 23b and the capacitance decreases, the movable electrode 61 and the fixed electrode 61 are fixed in the second detection regions 24a and 24b. The opposing area between the electrode elements 63 increases and the capacitance increases.

  Note that the maximum movable distances L2 and L4 of the weight portion 22 and the movable electrode portions 21a to 21d are restricted by the size of the gaps 28 to 31 shown in FIG. 1 in the X1-X2 direction.

  A frame body 16 (not shown in FIG. 1) is separated from the weight portion 22 shown in FIG. 1 and surrounds the periphery of the weight portion 22. As shown in FIG. 12 is fixed.

As shown in FIG. 4, the frame 16 and the fixed bases 52 constituting the functional layer 11 and the wiring board 15 are connected by the metal connection part 14. In FIG. 4, the wiring substrate 15 is shown in a single layer structure, but in reality, an insulating layer is formed on the surface of the silicon substrate (on the side facing the functional layer 11), and the wiring layer 70 is formed inside the insulating layer. It is a formed structure. The wiring layer 70 is electrically connected to the fixed base 52 via the metal connection portion 14, and the wiring layer 70 is connected to the pad portion 71 outside the frame body 16.
Although not shown in FIG. 4, a ground pad or the like is also formed on the outside of the frame body 16.

  In the present embodiment, it is possible to obtain a differential output by the capacitance change obtained from the first detection unit 23 and the capacitance change obtained from the second detection unit 24. Based on the differential output, the direction of the acceleration can be known from the magnitude of the acceleration and the sign (plus value or minus value) of the differential output.

(Experiment 1; Relation between acceleration and capacitance and relation between acceleration and sensitivity in comparative example)
The movable electrode elements 60 and 61 and the fixed electrode elements 62 and 63 shown in FIGS. 1 to 3 are not formed, and the plurality of movable support parts 50 and the plurality of fixed support parts 51 are comb-like electrodes. The experiment was conducted using the electrode structure as a comparative example. In the comparative example, when the weight portion 22 is moved in the X1-X2 direction in response to acceleration, a change in capacitance can be obtained by changing the gap (distance) between the comb-like electrodes (see FIG. 31).

  In the comparative example shown in the following experiment, the configuration other than the electrode structure is the same as that of the example. Further, in the example and the comparative example, the detection unit shown in FIG. 1 has the same size. However, in the comparative example, there are no movable electrode elements 60 and 61 and fixed electrode elements 62 and 63 as in the example. The distance between the movable electrode (the portion corresponding to the movable support portion 50 of the embodiment) and the fixed electrode (the portion corresponding to the fixed support portion 51 of the embodiment) was reduced to increase the number of electrodes. The principle of capacitance change in the comparative example is as described with reference to FIG.

  FIG. 5A is a simulation result showing the relationship between acceleration and capacitance in the comparative example. In the experiment of FIG. 5A, the sizes of the gaps a and b shown in FIG. 31 were set to 1.7 μm. Here, in the graph of (1) shown in FIG. 5A, when a positive acceleration is applied, the gap between the movable electrode and the fixed electrode is increased, the capacitance is decreased, and a negative acceleration is applied. The change in the capacitance at the detection unit where the gap between the movable electrode and the fixed electrode becomes smaller and the capacitance increases is shown. On the other hand, in the graph of (2) shown in FIG. 5A, when a negative acceleration is applied, the gap between the movable electrode and the fixed electrode is increased, the capacitance is decreased, and a positive acceleration is applied. The change in the capacitance at the detection unit where the gap between the movable electrode and the fixed electrode becomes smaller and the capacitance increases is shown. Here, “plus value” and “minus value” refer to a relationship in which, for example, if the acceleration of a plus value is in the X1 direction, the minus value is in the opposite X2 direction.

  FIG. 5B shows the acceleration and electrostatic capacity when the spring constants of the spring portions 40 to 43 are made smaller than in FIG. 5A and the sensitivity is increased without changing the size of the detection part of the MEMS sensor. It is a simulation result which shows the relationship with a capacity | capacitance. In the experiment of FIG. 5B, the sizes of the gaps a and b shown in FIG. 31 were set to 1.7 μm. As shown in FIG. 5 (b), the capacitance change becomes sharper than that in FIG. 5 (a), and the effective acceleration detection range (dynamic range) r2 in FIG. 5 (b) is shown in FIG. 5 (a). It was found that the dynamic range becomes narrower than the dynamic range r1.

  FIG. 6 shows a differential output curve based on the capacitance change in the graph (1) in FIG. 5A and the capacitance change in the graph (2). As shown in FIG. 6, it was found that the greater the acceleration (absolute value), the greater the change in capacitance (differential output).

  FIG. 7 is a simulation result showing the relationship between acceleration and sensitivity. The sensitivity is indicated by the slope of the differential output curve shown in FIG. As shown in FIG. 7, it was found that the sensitivity curve does not become a flat straight line with respect to the acceleration (absolute value), and changes greatly as the acceleration (absolute value) increases.

(Experiment 2; Relationship between acceleration and capacitance, and relationship between acceleration and sensitivity in Examples)
FIG. 8 is a simulation result showing a relationship between acceleration and capacitance in the example. In the experiment of FIG. 8, L1 and L3 shown in FIG. 3 (a) were set in the range of 4 to 6 μm, and the width dimension T1 of the electrode was set to 1.2 μm. Here, in the graph of (3) shown in FIG. 8, when a positive acceleration is applied, the facing area between the movable electrode and the fixed electrode is reduced and the capacitance is reduced (the state of FIG. 3 (c)). ) When a negative acceleration is applied, the facing area between the movable electrode and the fixed electrode is increased and the capacitance is increased (the state shown in FIG. 3B). ing. On the other hand, in the graph of (4) shown in FIG. 8, when a negative acceleration is applied, the facing area between the movable electrode and the fixed electrode is reduced, the capacitance is reduced, and a positive acceleration is applied. The change in capacitance at the detection unit is shown in which the facing area between the movable electrode and the fixed electrode increases and the capacitance increases. Here, “plus value” and “minus value” refer to a relationship in which, for example, if the acceleration of a plus value is in the X1 direction, the minus value is in the opposite X2 direction.

  FIG. 9 shows a differential output curve based on the capacitance change shown in the graph (3) in FIG. 8 and the capacitance change shown in the graph (4). FIG. 10 is a simulation result showing the relationship between acceleration and sensitivity in the example. The sensitivity is indicated by the slope of the differential output curve shown in FIG.

  In the embodiment, as shown in FIG. 9, the differential output curve is linearly inclined (primary curve) with respect to the acceleration (absolute value), and the sensitivity is substantially reduced with respect to the acceleration (absolute value) as shown in FIG. I found it flat.

  When comparing the sensitivity in the comparative example of FIG. 7 with the sensitivity in the example of FIG. 10, it was found that the sensitivity of FIG. 10 is higher than the sensitivity at the bottom of the sensitivity curve shown in FIG.

  Therefore, in the comparative example, if the sensitivity equivalent to that of the embodiment is to be obtained in the portion of the region where acceleration is small, for example, the number of comb-shaped electrodes must be increased to increase the facing area between the electrodes. It turns out that it will enlarge. In the embodiment, while maintaining the high sensitivity of the MEMS sensor as compared with the comparative example, the effective detection range (dynamic range) of the acceleration can be expanded and the linearity of the sensitivity within the effective detection range can be improved. I knew it was possible.

(Experiment 3; Experiment on optimization of L1 shown in FIG. 3A)
Next, the width T1 of the electrode shown in FIG. 3A was set to 1.2 μm, 0.8 μm, or 1.6 μm, and L3 in FIG. 3A was fixed to 5 μm.

  Further, as shown in FIG. 3B, the maximum movable distance L2 of the movable support portion 50 when each movable electrode 60 approaches the fixed support portion 51 forming a set is set to 2 μm. 2 μm corresponds to the case where an acceleration (absolute value) of 100 G is applied.

  As shown in FIG. 11, L1 is set to about 3 μm or more, and the sensitivity in a stationary state (at rest) in FIG. 3A is measured, and further, the movable support 50 is 2 μm in the direction of the fixed support 51. Sensitivity (100 G (absolute value)) when moving was measured. Then, [sensitivity (100 G (absolute value)) / sensitivity (at rest)] × 100 (%) was defined as the maximum fluctuation width of the vertical axis in FIG. Note that the spring constant was adjusted so as to move 2 μm when 100 G was applied.

  As shown in FIG. 11, it was found that the maximum amplitude of sensitivity increases as L1 decreases. The smaller the maximum sensitivity fluctuation, the better. For practical use, it is preferable to set the maximum amplitude (absolute value) to 10% or less. It was found that it is preferable to set L1 to about 3.6 μm or more in order to set the maximum sensitivity fluctuation width to 10% or less. Thus, the maximum fluctuation width of the sensitivity is increased by reducing L1 in addition to the fluctuation of the facing area between the movable electrode and the fixed electrode, as well as the fluctuation of the portion other than the facing area. This is because the capacitance is likely to change.

  FIG. 12 shows the relationship between L1-L2 (L2 is 2 μm) shown in FIG. 3B and the maximum sensitivity fluctuation range. As shown in FIG. 12, it was found that the maximum amplitude (absolute value) of sensitivity can be suppressed to 10% or less by setting L1-L2 to about 1.6 μm or more.

Subsequently, (L1-L2) ² / T1 is calculated, and FIG. 13 shows the relationship between (L1-L2) ² / T1 and the maximum sensitivity amplitude.

As shown in FIG. 13, it was found that (L1-L2) ² / T1 (where L2 is 2 μm) is substantially the same curve regardless of the width T1 of the electrode (normalization). Then, as shown in FIG. 13, it was found that the maximum fluctuation width (absolute value) of sensitivity can be suppressed to within 10% by setting (L1-L2) ² / T1 to about 2 (μm) or more.

  Subsequently, the width T1 of the electrode shown in FIG. 3A was set to 1.2 μm, 0.8 μm, or 1.6 μm, and L3 in FIG. 3A was fixed to 6 μm.

  In addition, when 100 G acts, the maximum movable distance L2 (see FIG. 3B) of the movable support portion 50 when each movable electrode 60 approaches the fixed support portion 51 that forms a set is 3 μm. The spring constant was adjusted.

  As shown in FIG. 14, L1 is set to about 4 μm or more, and the sensitivity (at rest) in FIG. 3A is measured, and further, the movable support portion 50 is 3 μm in the direction of the fixed support portion 51. Sensitivity (100 G (absolute value)) when moving was measured. Then, [sensitivity (100 G (absolute value)) / sensitivity (at rest)] × 100 (%) was defined as the maximum fluctuation width of the vertical axis in FIG.

  As shown in FIG. 14, it was found that L1 is preferably about 4.6 μm or more in order to set the maximum sensitivity fluctuation width to 10% or less.

  FIG. 15 shows the relationship between L1-L2 (L2 is 3 μm) shown in FIG. 3B and the maximum sensitivity fluctuation range. As shown in FIG. 15, it was found that the maximum amplitude (absolute value) of sensitivity can be suppressed to 10% or less by setting L1−L2 to about 1.6 μm or more.

Subsequently, (L1-L2) ² / T1 was calculated, and FIG. 16 shows the relationship between (L1-L2) ² / T1 and the maximum sensitivity amplitude.

As shown in FIG. 16, it was found that (L1−L2) ² / T1 (where L2 is 3 μm) is substantially the same curve regardless of the width T1 of the electrode (normalization). . Then, as shown in FIG. 16, it was found that the maximum amplitude (absolute value) of sensitivity can be suppressed to within 10% by setting (L1-L2) ² / T1 to about 2.2 (μm) or more. .

  Subsequently, the width T1 of the electrode shown in FIG. 3A was set to 1.2 μm, 0.8 μm, or 1.6 μm, and L3 in FIG. 3A was fixed to 7 μm.

  Further, when 100 G acceleration is applied, the maximum movable distance L2 (see FIG. 3B) of the movable support portion 50 when each movable electrode 60 approaches the fixed support portion 51 that forms a pair is 4 μm. The spring constant was adjusted so that

  As shown in FIG. 17, L1 is set to about 5 μm or more, and the sensitivity (at rest) in FIG. 3A is measured. Further, the movable support portion 50 is 4 μm in the direction of the fixed support portion 51. Sensitivity (100 G (absolute value)) when moving was measured. Then, [sensitivity (100 G (absolute value)) / sensitivity (at rest)] × 100 (%) was defined as the maximum fluctuation width of the vertical axis in FIG.

  As shown in FIG. 17, it was found that L1 is preferably about 5.6 μm or more in order to set the maximum fluctuation width of sensitivity to 10% or less.

  FIG. 18 shows the relationship between L1-L2 (L2 is 4 μm) shown in FIG. 3B and the maximum sensitivity fluctuation range. As shown in FIG. 18, it was found that the maximum amplitude (absolute value) of sensitivity can be suppressed to 10% or less by setting L1-L2 to about 1.6 μm or more.

Subsequently, (L1−L2) ² / T1 is calculated, and FIG. 19 shows the relationship between (L1−L2) ² / T1 and the maximum amplitude of sensitivity.

As shown in FIG. 19, it was found that (L1−L2) ² / T1 (where L2 is 4 μm) has substantially the same curve regardless of the width T1 of the electrode (standardization). . Then, as shown in FIG. 16, it was found that the maximum amplitude (absolute value) of sensitivity can be suppressed to within 10% by setting (L1-L2) ² / T1 to about 2.2 (μm) or more. .

According to the experiments shown in FIGS. 11 to 19, by setting (L1-L2) ² / T1 to 2.2 or more, the maximum sensitivity fluctuation range (absolute value) can be set to 10% or less. In particular, it has been found that the sensitivity linearity can be improved.

The technical meaning of (L1-L2) ² / T1 will be described with reference to FIG.
The movable electrode 60 and the fixed electrode 62 shown in FIG. 29A are comb-shaped electrodes having no width, and the overlap area (the range of the dotted line) is as shown between the left and right diagrams in FIG. ), It is possible to obtain a sensitivity characteristic with good linearity.

  However, as shown in FIG. 29B, actually, support portions 50 and 51 for supporting each electrode 60 and 62 are required, and each electrode 60 and 62 has a width (width dimension T1). (See also FIG. 3).

  Thus, since each electrode element 60 and 61 has the width dimension T1, it is between each electrode element 60 and 62 and the support parts 50 and 51 of the left figure of FIG.29 (b) (part shown by the arrow). ) Occurs, that is, the change in capacitance cannot be captured by the simple change in the overlap area shown in FIG. The right diagram in FIG. 29B is an extraction of the electrostatic capacitance component of the arrow portion in the left diagram in FIG. 29B. The parallel plate type as shown in the right diagram in FIG. A capacitive component is also added.

  The capacity component of the parallel plate type increases in inverse proportion to the distance, and increases in proportion to the area (electrode width dimension T1).

  Therefore, the linearity varies as the width dimension T1 of each electrode element increases and in inverse proportion to the interval (gap) L1 between each electrode element and the support portion.

  Therefore, with respect to L1-L2 in FIGS. 12, 15, and 18, normalization is performed with the width dimension T1 of each electrode element, and normalization is performed with 1 / (L1-L2). Since it is generated and becomes L1-L2, in order to normalize by 1 / (L1-L2), as shown in the following formula 1, L1-L2 is changed to T1 and 1 / (L1-L2) Divided by.

  (L1-L2) / [T1 · [1 / (L1-L2)]] (Equation 1)

  From the above formula 1, the following formula 2 can be obtained.

(L1-L2) ² / T1 (Formula 2)

As described above, (L1-L2) ² / T1 is a mathematical expression having a technical meaning that the maximum fluctuation width (absolute value) of sensitivity can be set to 10% or less.

(Experiment 4: Experiment for optimization of L3 shown in FIG. 3A)
Next, the width T1 of the electrode shown in FIG. 3A was set to 1.2 μm, and L1 in FIG. 3A (at rest) was fixed to 5 μm. Then, as shown in FIG. 3C, the maximum movable distance L4 of the movable support portion 50 when each movable electrode 60 approaches the adjacent fixed support portion 51 is set to 2 μm. 2 μm corresponds to the case where acceleration of 100 G (absolute value) is applied. As shown in FIG. 20, L3 is set to about 3 μm or more, and the sensitivity in the stationary state (at rest) in FIG. 3A is measured. As described above, the movable support 50 is 2 μm, and the adjacent group The sensitivity (100 G) when moving in the direction of the fixed support portion 51 was measured. Then, [sensitivity (100 G (absolute value)) / sensitivity (at rest)] × 100 (%) was defined as the maximum fluctuation width on the vertical axis of FIG. The experimental result is a graph of (5) shown in FIG.

  Further, an experiment was performed in which the width T1 of the electrode shown in FIG. 3A was 1.2 μm, and L1 in FIG. 3A (at rest) was linked to L3. That is, if L3 is 4 μm, L1 is also set to 4 μm. Similarly to the experiment in the graph of (5) above, the maximum movable distance L4 of the movable support 50 when each movable electrode 60 approaches the adjacent fixed support 51 is set to 2 μm (100 G ( [At the time of acceleration action of absolute value)] [sensitivity (100 G (absolute value)) / sensitivity (at rest)] × 100 (%) was measured. The experimental result is a graph of (6) shown in FIG.

  As shown in FIG. 20, it was found that it is preferable to set L3 to about 3.8 μm or more in order to set the maximum sensitivity fluctuation (absolute value) to 10% or less.

  FIG. 21 shows the relationship between L3-L4 (L4 is 2 μm) shown in FIG. 3C and the maximum sensitivity fluctuation range. The graph (5) shown in FIG. 21 is based on the graph (5) in FIG. 20, and the graph (6) shown in FIG. 21 is based on the graph (6) in FIG. . In FIG. 21, L1 and L3 are interlocked, and the experiment in which the electrode width T1 is 0.8 μm (graph (7)), and L1 and L3 are interlocked and the electrode width is adjusted. An experiment (graph (8)) in which the dimension T1 was 1.6 μm was also performed.

  As shown in FIG. 21, it was found that the maximum sensitivity fluctuation (absolute value) can be suppressed to 10% or less by setting L3-L4 (L4 = 2 μm) to about 1.8 μm or more.

Subsequently, (L1-L2) ² × T1 is calculated, and FIG. 22 shows the relationship between (L3-L4) ² × T1 and the maximum sensitivity amplitude. The three curves shown in FIG. 22 are based on the graphs (6), (7), and (8) shown in FIG.

As shown in FIG. 22, it was found that (L3−L4) ² × T1 (where L4 is 2 μm) has substantially the same curve regardless of the width T1 of the electrode (standardization). . Then, as shown in FIG. 22, it was found that by setting (L3-L4) ² × T1 to 4 (μm ³ ) or more, the maximum sensitivity fluctuation range (absolute value) can be suppressed to within 10%.

  Subsequently, the width T1 of the electrode shown in FIG. 3A was set to 1.2 μm, and L1 in FIG. 3A (at rest) was fixed to 6 μm. When the acceleration of 100 G is applied, the maximum movable distance L4 of the movable support 50 when each movable electrode 60 approaches the adjacent fixed support 51 is 3 μm (see FIG. 3C). The spring constant was adjusted so that

  As shown in FIG. 23, L3 is set to about 5 μm or more, and the sensitivity in the stationary state (at rest) in FIG. 3A is measured. As described above, the movable support 50 is 3 μm, and the adjacent group The sensitivity (100 G) when moving in the direction of the fixed support portion 51 was measured. Then, [sensitivity (100 G (absolute value)) / sensitivity (during rest)] × 100 (%) was defined as the maximum fluctuation width of the vertical axis in FIG. The experimental result is the graph of (9) shown in FIG.

  Further, an experiment was performed in which the width T1 of the electrode shown in FIG. 3A was 1.2 μm, and L1 in FIG. 3A (at rest) was linked to L3. That is, if L3 is 6 μm, L1 is also set to 6 μm. Similarly to the experiment in the graph of (9), the maximum movable distance L4 of the movable support 50 when each movable electrode 60 approaches the adjacent fixed support 51 is set to 3 μm (100 G ( [At the time of acceleration action of absolute value)] [sensitivity (100 G (absolute value)) / sensitivity (at rest)] × 100 (%) was measured. The experimental result is the graph of (10) shown in FIG.

  As shown in FIG. 23, it was found that L3 is preferably about 5 μm or more in order to set the maximum sensitivity fluctuation (absolute value) to 10% or less.

  FIG. 24 shows the relationship between L3-L4 (L4 is 3 μm) shown in FIG. 3C and the maximum sensitivity fluctuation range. The graph (9) shown in FIG. 24 is based on the graph (9) in FIG. 23, and the graph (10) shown in FIG. 24 is based on the graph (10) in FIG. . In FIG. 24, L1 and L3 are interlocked, the electrode width T1 is 0.8 μm (graph (11)), and L1 and L3 are interlocked and the electrode width An experiment (graph (12)) in which the dimension T1 was 1.6 μm was also performed.

  As shown in FIG. 24, it was found that the maximum amplitude (absolute value) of sensitivity can be suppressed to 10% or less by setting L3−L4 (L4 = 3 μm) to about 1.8 μm or more.

Subsequently, (L3−L4) ² × T1 is calculated, and FIG. 25 shows the relationship between (L3−L4) ² × T1 and the maximum amplitude of sensitivity. Note that the three curves shown in FIG. 25 are based on the graphs (10), (11), and (12) shown in FIG.

As shown in FIG. 25, it was found that (L3−L4) ² × T1 (where L4 is 3 μm) is substantially the same curve regardless of the width T1 of the electrode (normalization). Then, as shown in FIG. 25, it was found that by setting (L3−L4) ² × T1 to 4 (μm ³ ) or more, the maximum sensitivity fluctuation (absolute value) can be suppressed within 10%.

  Subsequently, the width T1 of the electrode shown in FIG. 3A was set to 1.2 μm, and L1 in FIG. 3A (at rest) was fixed to 7 μm. When the acceleration of 100 G is applied, the maximum movable distance L4 of the movable support 50 when each movable electrode 60 approaches the adjacent fixed support 51 is 4 μm (see FIG. 3C). ) To adjust the spring constant.

  As shown in FIG. 26, L3 is set to about 6 μm or more, and the sensitivity in a stationary state (at rest) in FIG. 3A is measured. As described above, the movable support 50 is 4 μm, and the adjacent group The sensitivity (100 G) when moving in the direction of the fixed support portion 51 was measured. Then, [sensitivity (100 G (absolute value)) / sensitivity (at rest)] × 100 (%) was defined as the maximum fluctuation width of the vertical axis in FIG. The experimental result is a graph of (13) shown in FIG.

  Further, an experiment was performed in which the width T1 of the electrode shown in FIG. 3A was 1.2 μm, and L1 in FIG. 3A (at rest) was linked to L3. That is, if L3 is 7 μm, L1 is also set to 7 μm. Similarly to the experiment in the graph of (13), the maximum movable distance L4 of the movable support 50 when each movable electrode 60 approaches the adjacent fixed support 51 is 4 μm (100 G ( [At the time of acceleration action of absolute value)] [sensitivity (100 G (absolute value)) / sensitivity (at rest)] × 100 (%) was measured. The experimental result is a graph of (14) shown in FIG.

  As shown in FIG. 26, it was found that L3 is preferably about 6 μm or more in order to set the maximum sensitivity fluctuation (absolute value) to 10% or less.

  FIG. 27 shows the relationship between L3-L4 (L4 is 4 μm) shown in FIG. 3C and the maximum sensitivity fluctuation range. The graph of (13) shown in FIG. 27 is based on the graph of (13) in FIG. 26, and the graph of (14) shown in FIG. 27 is based on the graph of (14) in FIG. . In FIG. 27, L1 and L3 are interlocked, the electrode width T1 is 0.8 μm (graph (15)), and L1 and L3 are interlocked and the electrode width An experiment (graph (16)) in which the dimension T1 was 1.6 μm was also conducted.

  As shown in FIG. 27, it was found that by setting L3-L4 (L4 = 4 μm) to about 1.8 μm or more, the maximum sensitivity fluctuation range (absolute value) can be suppressed to 10% or less.

Next, (L3-L4) ² × T1 is calculated, and FIG. 28 shows the relationship between (L3-L4) ² × T1 and the maximum sensitivity amplitude. Note that the three curves shown in FIG. 28 are based on the graphs (14), (15), and (16) shown in FIG.

As shown in FIG. 28, it was found that (L3-L4) ² × T1 (where L4 is 4 μm) is substantially the same curve regardless of the width T1 of the electrode (normalization). As shown in FIG. 28, it was found that the maximum amplitude (absolute value) of sensitivity can be suppressed to 10% or less by setting (L3−L4) ² × T1 to 4 (μm ³ ) or more.

The technical meaning of (L3-L4) ² × T1 will be described with reference to FIG.
In FIG. 30 (a), the capacitance component generated in the gap (gap) L1 between each electrode and the support portion is the width of the movable electrode 60 (shown thickly as 60b) and the width of the movable electrode 60. And the width of the fixed electrode element 62 (shown thick as reference numeral 51b) and the width of the fixed electrode element 62 (shown thick as reference numeral 62b) and the movable support element opposite the width of the fixed electrode element 62 This is the total that occurs between 50 widths (shown thick as 50b).

  Further, as shown in FIG. 30 (a), the capacitance component generated in the gap (gap) L3 between the adjacent pair of movable support portions and fixed support portions is the width 50c of the movable support portion 50 (shown thick). And the width 51c (shown thick) of the fixed support portion 51.

  In FIG. 30A, the movable electrode 60 and the fixed electrode 62 are configured to have such a width that they are in contact with each other. For this reason, if L1 and L3 are equal, the capacitance component generated in the gap (gap) L1 between each electrode and the support part and the gap (gap) between the adjacent movable support part and the fixed support part. ) Capacitance component generated in L3 is equal, and therefore, by forming a differential circuit, the capacitance component generated in the gap (gap) L1 between each electrode and the support portion and the movable support portion of the adjacent set is fixed. It is possible to cancel the capacitance component generated at the gap (gap) L3 between the support portion and the support portion.

  However, as shown in FIG. 30 (b), in actuality, the electrodes 60, 62 have a width (width dimension T1) and are separated from each other. The right diagram in FIG. 30B is obtained by extracting the capacitance component (parallel plate capacitance component) indicated by the arrow in the left diagram in FIG. In this way, in FIG. 30 (b), the width dimension of each electrode element 60, 62 is reduced as compared with FIG. 30 (a). As a result, even if L1 and L3 are equal, The capacity component generated in the interval (gap) L1 between the two and the capacity component generated in the interval (gap) L3 between the adjacent pair of movable support portions and fixed support portions is not equal. Thus, a deviation from the ideal state occurs.

  That is, the linearity varies in inverse proportion to the width T1 of each electrode 60, 61. Thus, the smaller the width T1 of each electrode element 60, 61, the larger the variation in linearity. Similarly, the variation in linearity increases in inverse proportion to (L3-L4).

  Therefore, in order to normalize with L1 / L4 of FIGS. 21, 14, and 17 by 1 / T1 and to normalize with 1 / (L3-L4), as shown in Equation 3 below, L3-L4 was divided by 1 / T1 and 1 / (L3-L4).

  (L3-L4) / [(1 / T1) × [1 / (L3-L4)]] (Formula 3)

  From Equation 3, the following Equation 4 can be obtained.

(L3-L4) ² × T1 (Formula 4)

As described above, (L3−L4) ² × T1 is a mathematical expression having a technical meaning that can set the maximum fluctuation width (absolute value) of the sensitivity to 10% or less.

  When L1 and L3 are not linked, it is preferable to satisfy L3> L1 at the time of stationary in FIG. In the experiment shown in FIG. 11, L1 is about 3.6 μm when the maximum sensitivity fluctuation is 10 (%). At this time, L3 is 5 μm. On the other hand, for example, as shown in the graph of (5) in FIG. 20, L3 is about 4.4 μm when the maximum sensitivity fluctuation width is 10 (%). At this time, L1 is 5 μm. Thus, when the dimensional relationship between L1 and L3 at rest is viewed, the smaller dimension (L1) can be set smaller in the experiment in FIG. 11 where L3> L1. Therefore, the downsizing of the MEMS sensor can be promoted while maintaining high sensitivity and good linearity.

  In FIG. 3, L5 is equal to or greater than L2 and L4. As a result, even if the movable electrode moves to the maximum in the X1-X2 direction as shown in FIGS. 3B and 3C, the movable electrode does not move outside the fixed electrode, and the linearity of sensitivity is reduced. Can be suppressed.

  The MEMS sensor S in the present embodiment includes a plurality of movable support portions 50 constituting the movable electrode portions 21a to 21d and a plurality of fixed support portions 51 constituting the fixed electrode portions 20a to 20d in the same detection portion. They are alternately arranged in the X1-X2 direction (first direction). Adjacent movable support portions 50 and fixed support portions 51 are made into the same set 66, and a plurality of movable electrode elements 60 and 61 are fixed between the movable support portion 50 and the fixed support portion 51 in each set 66. Electrodes 62 and 63 were alternately arranged in the Y1-Y2 direction (second direction). As a result, a large number of movable electrode elements 60 and 61 and fixed electrode elements 62 and 63 can be efficiently arranged in the detection unit, which is smaller and has higher sensitivity than the conventional electrode structure.

  By adopting the electrode structure of the present embodiment, it is possible to effectively increase the variation area of the opposing area between the movable electrode elements 60 and 61 and the fixed electrode elements 62 and 63 in the detection area of the same size, and the sensitivity. Can be increased.

  Further, in the method of causing a change in capacitance by changing the gap (distance) between the movable electrode and the fixed electrode, the capacitance is inversely proportional to the gap (distance), and as shown in FIGS. When the sensitivity is increased, the capacitance change is abrupt with respect to the variation in distance (see FIG. 5B), and the effective detection range (dynamic range) of the physical quantity by the MEMS sensor is narrowed. Furthermore, there is a problem that the linearity of sensitivity within the effective detection range cannot be improved (see FIG. 7).

  On the other hand, in the method of changing the capacitance by changing the facing area between the movable electrode 60, 61 and the fixed electrode 62, 63 as in the present embodiment, the capacitance is proportional to the facing area. Compared with the comparative example method in which the capacitance is changed by the above-described distance variation, the effective detection range (dynamic range) of the physical quantity can be widened and the linearity of the sensitivity within the effective detection range can be improved. Is possible.

In addition to the above configuration, in the present embodiment, as shown in the above experiment, (L1-L2) ² / T1 is set to 2.2 (μm) or more, and (L3-L4) ² × T1 is set to 4 (μm). ³ ) By setting above, it is possible to effectively improve the linearity of sensitivity.

  Further, as shown in FIG. 1, the first detection unit 23 is divided into a plurality of detection regions 23a and 23b, the second detection unit 24 is divided into a plurality of detection regions 24a and 24b, and each detection region 23a, 23b, 24a and 24b is divided. By providing the movable electrode portions 21a to 21d and the fixed electrode portions 20a to 20d, the movable support portion 50 and the fixed support portion 51 of this embodiment need not be formed extremely long, and the strength of each electrode portion The sensitivity can be improved while maintaining a sufficient value.

  FIG. 1 shows the MEMS sensor for detecting the acceleration acting in the X1-X2 direction. However, if the MEMS sensor is rotated 90 degrees from the state of FIG. 1, the MEMS sensor for detecting the acceleration acting in the Y1-Y2 direction is used. it can. Further, the MEMS sensor that detects the acceleration in the X-axis direction shown in FIG. 1 and the MEMS sensor that detects the acceleration in the Y-axis direction obtained by rotating the MEMS sensor of FIG. 1 by 90 degrees are combined into a two-axis acceleration detection sensor. It is also possible to do.

  This embodiment is applicable not only to acceleration sensors but also to general physical quantity sensors such as angular velocity sensors and impact sensors.

S MEMS sensor 10 Support substrate 11 Functional layer 12 Insulating layer 13 SOI substrate 14 Metal connection portion 15 Wiring substrate 16 Frame bodies 20a to 20d Fixed electrode portions 21a to 21d Movable electrode portion 22 Weight portion 23 First detection portions 23a and 23b First Detection area 24 2nd detection part 24a, 24b 2nd detection area 26, 27 Anchor part 28-31 Gap 40-43 Spring part 50 Movable support part 51 Fixed support part 52 Fixed base part 60, 61 Movable electrode element 62, 63 Fixed electrode Child 66 set 70 Wiring layer

Claims (6)

In a MEMS sensor having a movable electrode portion and a fixed electrode portion,
When the two directions orthogonal to each other in the horizontal plane are defined as the first direction and the second direction, the first direction is the moving direction of the movable electrode part,
The movable electrode portion includes a plurality of movable support portions that are arranged with an interval in the first direction and extend in the second direction, and each movable support portion toward the first direction. A plurality of movable electrode elements extending from the side portions of the movable support portions and spaced from each other in the second direction at each movable support portion,
The fixed electrode portion is arranged with a gap in the first direction, and a plurality of fixed support portions whose extending direction from the proximal end side to the distal end side is opposite to the movable support portion; A plurality of fixed electrode elements extending in a direction opposite to the extending direction of the movable electrode element from a side part of the support part, and arranged at intervals in the second direction at each fixed support part; Have
A plurality of the movable support portions and a plurality of the fixed support portions are alternately arranged at intervals in the first direction, and the adjacent movable support portions and the fixed support portions are paired. A plurality of movable electrode elements and a plurality of fixed electrode elements are alternately arranged in the second direction with an interval between each set of the movable support part and the fixed support part. And
By moving the movable electrode portion in the first direction, it is possible to detect a change in capacitance based on a change in the facing area between each movable electrode and each fixed electrode.
In a stationary state, the distance in the first direction from the side of the movable support portion to the tip of the fixed electrode is L1, and the width of the movable electrode and the fixed electrode in the second direction When the dimension is T1, and the maximum movable distance when the movable electrode moves in the direction approaching the fixed support portion is L2, (L1-L2) ² / T1 is 2.2 (μm) or more. A MEMS sensor characterized by being. In a stationary state, an interval in the first direction between the movable support portion and the fixed support portion of the adjacent sets is L3, and the movable electrode element and the fixed electrode element are in the second direction. (L3−L4) ² × T1 is 4 where L1 is the maximum movable distance when the movable electrode element moves in a direction approaching the fixed support portion of the adjacent group. The MEMS sensor according to claim 1, which is (μm ³ ) or more. In a MEMS sensor having a movable electrode portion and a fixed electrode portion,
When the two directions orthogonal to each other in the horizontal plane are defined as the first direction and the second direction, the first direction is the moving direction of the movable electrode part,
The movable electrode portion includes a plurality of movable support portions that are arranged with an interval in the first direction and extend in the second direction, and each movable support portion toward the first direction. A plurality of movable electrode elements extending from the side portions of the movable support portions and spaced from each other in the second direction at each movable support portion,
The fixed electrode portion is arranged with a gap in the first direction, and a plurality of fixed support portions whose extending direction from the proximal end side to the distal end side is opposite to the movable support portion; A plurality of fixed electrode elements extending in a direction opposite to the extending direction of the movable electrode element from a side part of the support part, and arranged at intervals in the second direction at each fixed support part; Have
A plurality of the movable support portions and a plurality of the fixed support portions are alternately arranged at intervals in the first direction, and the adjacent movable support portions and the fixed support portions are paired. A plurality of movable electrode elements and a plurality of fixed electrode elements are alternately arranged in the second direction with an interval between each set of the movable support part and the fixed support part. And
By moving the movable electrode portion in the first direction, it is possible to detect a change in capacitance based on a change in the facing area between each movable electrode and each fixed electrode.
In a stationary state, an interval in the first direction between the movable support portion and the fixed support portion of the adjacent sets is L3, and the movable electrode element and the fixed electrode element are in the second direction. (L3−L4) ² × T1 is 4 where L1 is the maximum movable distance when the movable electrode element moves in a direction approaching the fixed support portion of the adjacent group. A MEMS sensor characterized by being (μm ³ ) or more. A first detection unit and a second detection unit, and each of the detection units includes the movable electrode unit and the fixed electrode unit;
The extending direction of the movable electrode element and the fixed electrode element in the first direction is reversed between the first detection unit and the second detection unit, and is obtained by the first detection unit. The MEMS sensor according to any one of claims 1 to 3, wherein a differential output can be obtained by a change in capacitance and a change in capacitance obtained by the second detection unit.   The MEMS sensor according to claim 4, wherein each of the first detection unit and the second detection unit is divided into a plurality of detection regions, and each of the detection regions includes the movable electrode unit and the fixed electrode unit. . The movable body having the movable electrode portion and an anchor portion connected via a spring portion are provided. Between the movable body and the anchor portion, the movable body moves in the first direction from a stationary state. A gap is formed that regulates the possible dimensions,
The spring portion is formed to extend in the second direction and be folded between the anchor portion and the movable body so as to suppress vibration of the movable body in the second direction. 6. The MEMS sensor according to any one of 1 to 5.

Images