MEMS piston-tube based capacitive accelerometer

Abstract

A novel high resolution, low noise MEMS capacitive accelerometer is disclosed. The accelerometer utilizes a piston-tube electrode configuration that enables the use of a wide area for the electrodes. Therefore, a high capacitive sensitivity is achieved. The accelerometer consists of two structures: upper and lower. The lower structure contains a plurality of fixed electrodes that are attached to the base and have a piston-style shape (teeth). Those pistons form the sensing electrodes of the accelerometer. The upper structure contains a plurality of moving electrodes that have a tube-style shape (through holes), and they are attached to a substrate via restoring mechanical springs. The proof mass of the accelerometer is distributed around these tubes to reduce squeeze thin film damping in the system. The accelerometer is able to sense linear acceleration along the z-axis and/or the angular acceleration about the in-plane axes (x and y).

Claims

1. A MEMS capacitive accelerometer, comprising: a. a first base plate having a top surface and a thickness; b. an array of spaced apart teeth electrodes constructed in the thickness of said base plate; c. each said tooth electrode having a shape, a length, a height, and a thickness; d. said array of teeth being grouped in one or multiple sub-arrays of teeth each comprising one or more teeth, wherein the teeth in each sub-array of teeth being electrically connected to each other and electrically isolated from other sub-arrays of teeth; e. each sub-array of teeth being electrically addressable, whereby each sub-array of teeth forms a sensing electrode of said accelerometer; f. a second plate attached to said base plate and having a structure sized to fit on said base plate, said second plate having a fixed peripheral structure and a moving central structure, whereby the fixed structure being attached to the base plate and the moving central structure forms a proof mass of said accelerometer; g. an array of spaced apart openings constructed in the thickness of said moving structure said second plate; h. said array of openings sized and designed to receive and interdigitate with said array of teeth, and said array of openings with each said opening having a length, a width and a height, whereby said array of openings forms a moving electrode of said accelerometer; i. a plurality of spring means supports and holds said moving structure of said second plate wherein said openings are aligned with respect to said array of teeth and to return the said moving structure to its initial position in the absence of external acceleration, each said spring having a length, a thickness, and a height, whereby the teeth of the base plate penetrate into the openings of the top plate in response to linear acceleration along an out-of-plane axis (z-axis), or/and rotational acceleration about an in-plane axes (x and y) being measured by means of capacitance change detection that is formed between the teeth and the openings.

2. The MEMS capacitive accelerometer of claim 1, wherein said array of teeth electrodes being arc shaped and being radially co-centric with a predefined radial spacing.

3. The MEMS capacitive accelerometer of claim 1, wherein said array of teeth electrodes being substantially rectangular, circular, triangular, trapezoidal, pentagonal, or hexagonal shaped, and each said sub-array of teeth being aligned linearly along their length and forming several linear rows in each said sub-array.

4. The MEMS capacitive accelerometer of claim 1, wherein said array of teeth electrodes being substantially rectangular, circular, triangular, trapezoidal, pentagonal, or hexagonal shaped, and each said sub-array of teeth being aligned radially along their length and forming several curved rows in each said sub-array.

5. The MEMS capacitive accelerometer of claims 3 and 4, wherein said sensing tooth electrodes and moving electrode (openings) form four capacitors.

6. The MEMS capacitive accelerometer of claim 1, wherein the proof mass is partially solid (lumped) at the center of the said moving structure of said second plate, and surrounded by an array of openings.

7. The MEMS capacitive accelerometer of claim 1, wherein said springs being long springs having high ratio of a longitudinal stiffness to a transverse stiffness to allow for a large travel range of the proof mass.

8. The MEMS capacitive accelerometer of claim 1, wherein said springs being extended radially inwards.

9. The MEMS capacitive accelerometer of claim 1, wherein said springs having the same thickness as the height of the opening, whereby eliminating the fabrication step needed for etching the second plate to soften the springs, thereby simplifying the fabrication process.

10. The MEMS capacitive accelerometer of claim 1, wherein a plurality of said pistons are extended upward such that their top surfaces are at the same plane with top surface of said second plate.

11. The MEMS capacitive accelerometer of claim 10, wherein said accelerometer measures the linear acceleration along two directions of the out-of-plane axis.

Description

BRIEF DESCRIPTION OF DRAWINGS

(1) Embodiments herein will hereinafter be described in conjunction with the appended drawings provided to illustrate and not to limit the scope of the claims, wherein like designations denote like elements, and in which:

(2) FIG. 1A illustrates the lower structure (pistons) of the piston-tube capacitive accelerometer.

(3) FIG. 1B is a top view illustrating the lower structure (pistons) of the piston-tube capacitive accelerometer.

(4) FIG. 2A is a top view illustrating the upper structure (tubes) of the piston-tube capacitive accelerometer.

(5) FIG. 2B is a perspective view illustrating the upper structure (tubes) of the piston-tube capacitive accelerometer.

(6) FIG. 3 is an exploded view of the piston-tube capacitive accelerometer illustrating the lower structure containing pistons (fixed electrodes) and the upper structure containing the tubes (moving electrodes).

(7) FIG. 4 is a perspective view illustrating the final bonded structure of the piston-tube capacitive accelerometer.

(8) FIG. 5A shows top, cross-sectional, and detailed views illustrating the piston-tube capacitive accelerometer.

(9) FIG. 5B shows cross-sectional and detailed views illustrating the piston-tube capacitive accelerometer during the application of linear and angular acceleration (only faces along the B-B line are shown for the sake of clarity).

(10) FIG. 6 shows top and isometric views of an embodiment of the piston-tube capacitive accelerometer in which the proof mass is partially solid (lumped).

(11) FIG. 7A illustrates the upper structure layer being bonded to the lower structure before etching and patterning of the tubes and extended pistons of the bi-directional MEMS accelerometer.

(12) FIG. 7B illustrates a cross-sectional view of the structure of the accelerometer after DRIE etching of the upper structure.

(13) FIG. 7C shows a detailed view of the cross-section of the final structure of the accelerometer after etching the tubes and extended pistons.

(14) FIG. 7D is a top view of the final structure of the bi-directional accelerometer illustrating the top sides of extended pistons (black) and the lower pistons (grey).

(15) FIG. 8A is a detailed cross-sectional view of the bi-directional accelerometer illustrating the capacitances between the extended pistons and lower pistons in conjunction with the corresponding tubes of the bi-directional accelerometer.

(16) FIG. 8B is a detailed cross-sectional view of the bi-directional accelerometer during the upward movement (i.e. away from the substrate) of the proof mass (tubes) illustrating the change in the capacitances between the extended pistons and lower pistons with the corresponding tubes.

(17) FIG. 8C is a detailed cross-sectional view of the bi-directional accelerometer during the downward movement (i.e. towards the substrate) of the moving structure (tubes) illustrating the change in the capacitances between the extended pistons and lower pistons with the corresponding tubes.

(18) FIG. 8D shows the change of capacitance between the extended pistons and lower pistons with the corresponding tubes in response to a displacement of the proof mass (tubes).

(19) FIG. 9 is a snapshot of a fabricated prototype of an embodiment of the piston-tube based accelerometer with partially solid (lumped) proof mass.

DETAILED DESCRIPTION OF THE INVENTION

(20) An embodiment of the MEMS capacitive accelerometer is illustrated in FIG. 1-5. It consists of two structures. The lower structure 100 in the embodiment is shown in FIG. 1A-B, and it forms a plurality of rectangular-shaped pistons (teeth) 110 111 112 113, and the upper structure 200 is shown in FIG. 2A-B, and it forms a plurality of rectangular-shaped tubes (openings) 210, that correspond to the pistons in the lower structure.

(21) The pistons could form one or multiple sensing electrodes (preferably four), and they protrude vertically towards the upper structure and extend horizontally along the two in-plane axes (x and y). Each sensing electrode is electrically isolated form other sensing electrodes by utilizing the Buried Oxide (BOX) layer 130 of an SOI wafer or any other dielectric layer. The pistons within each single group (sensing electrode) are electrically connected to each other via a small thickness (height) layer 140 forming a single capacitor with the moving (common) electrode in the upper structure. It will be readily clear to the one with ordinary skills in the art that this layer 140, shown in FIG. 1B, is connected to the outer fixed periphery which is divided into different electrical pads by means of DRIE etching. Each pad 990, refer to FIG. 9, is electrically addressable. The goal of dividing the pistons into multiple electrodes is to enable the detection of the angular acceleration about the in-plane axes.

(22) The tubes 210 form the moving electrode, which is a common electrode that in conjunction with the lower sensing electrodes 110 111 112 113 make up the capacitors. The mass of the tubes is considered the proof mass of the accelerometer. The tubes are rectangular through holes, and they are vertically aligned with the pistons (sensing electrodes). The cross-section of the tubes is larger than that of the pistons so that the pistons penetrate along the tubes during the acceleration. In other words, a horizontal gap 224 between the outer sides of the pistons and inner sides of the tubes exist, refer to the cross sectional view in FIG. 5, where only faces along the B-B line are shown in the figure for the sake of clarity. The moving structure containing the tubes is attached to the substrate by multiple supporting springs 220 221 222 223.

(23) When an external linear acceleration 510, please refer to FIG. 5B, is applied towards the upper structure (i.e. the positive direction of the z-axis), the relative engagement of the moving structure (tubes) and the lower structure (pistons) increases. The change of capacitance between the moving electrode (tubes) and the sensing electrodes (pistons) is detected by the read-out circuit. Hence the acceleration can be determined.

(24) When an external angular acceleration 520, refer to FIG. 5B, is applied about an in-plane axis (i.e. y-axis), the relative motion of the moving electrode (tubes) with respect to the fixed electrode (pistons) is a rotation about the same in-plane axis (i.e. y-axis) in a direction opposite to the applied angular acceleration. The resulting change in the engagement of the tubes to the corresponding pistons is in one half of the accelerometer structure (capacitors 500 that lie on one side of the axis of rotation where the engagement has increased) leads to a change in the capacitance which can be detected by the read-out circuit. The rotary motion of the tubes creates a slight non-linear change of capacitance. The non-linearity is caused by changes in the horizontal gaps between the long sides of the pistons and the tubes. Therefore, in the design of the accelerometer for applications that require angular acceleration measurements, a consideration has to be taken into account for specifying the amount of the horizontal gap such that during the rotation, the change in the gap should not exceed one third of the original horizontal gap 224 to avoid pull-in instability.

(25) The configuration of the springs holding the moving structure could be designed in many ways as it will be clearly apparent to the one who is skilled in the art. One may use suspension springs that extend radially to support the tubes. Another preferred configuration of springs is that the springs extend along the peripheral of the tubes' plate. This configuration of the springs has a number of advantages. First, it allows springs to be largely long (leads to having a high ratio consisting of the longitudinal stiffness to the transverse stiffness as the longitudinal stiffness of the beams is inversely proportional to the length and the transverse stiffness is inversely proportional to the length cube). This high value of the ratio pushes the lateral instability limit further; hence the pull-in voltage value is larger. This is a very important advantage when the accelerometer is operating in the closed loop mode because a large electrostatic force can be generated to keep the mass in the rest position by applying voltages below the pull-in voltage; hence, a wide dynamic range of the accelerometer could be achieved. The second advantage is that it provides a wide room for the electrodes to be multiplied as opposed to the spring configuration in which the springs are extending radially leading to shrinkage of the area of the tubes. Finally, the springs could have the same thickness as that of the tubes, meaning that the fabrication step needed for etching the upper structure to soften the spring could be eliminated, which simplifies the fabrication process.

(26) In another embodiment of the current invention, one may design the proof mass of the accelerometer such that it is partially lumped (solid) by removing some of the tubes at the center of the accelerometer. This lumped part 600, shown in FIG. 6, of the proof mass provides a control of the quality factor of the accelerometer by adding a squeeze thin film damping to the system as desired based on the application requirements. This step might be useful when the dynamic range of the accelerometer is needed to be wide as the squeeze thin film damping would reduce the ringing effect near the resonant frequency.

(27) The preferred embodiment of the accelerometer illustrated in FIG. 1-5 is intended to measure the linear vibration/acceleration in a single direction along the out-of-plane axis as the sensing electrodes are positioned on the lower direction of the tubes (i.e. direction towards the substrate). However, to measure the linear acceleration in two directions (positive and negative direction along the out-of-plane axis), two unidirectional accelerometers have to be mounted opposing each other in the measurement unit, so that acceleration in two directions along the out-of-plane axis can be measured. But that requires an additional area of the device.

(28) To overcome the limitations of the unidirectional accelerometer, another embodiment of the accelerometer that provides a bidirectional measurement of the vibration/acceleration is described. The bidirectional accelerometer, shown in FIGS. 7-8, has fixed sensing electrodes that measure the linear displacement of the tubes in two directions, positive and negative directions along the z-axis. The accelerometer has a structure similar to the structure of the unidirectional embodiment of the accelerometer. However, about half of the pistons in the lower structure are extended to the level of the tubes in the upper structure. This extension is achieved by adding the features of the extended pistons (rectangles) to the mask containing the tube features. For illustration, FIG. 7A shows the structure of the accelerometer after bonding of the upper structure 701 to the lower one 702 during the fabrication process (e.g. the MicraGEM process). The features of the tubes and the extended pistons, contained in the mask, are photo-lithographically patterned and DRIE etched. FIG. 7B shows the final structure of the accelerometer. The high alignment accuracy between the features in the lower and upper structures in fabrication processes such as the MicraGEM process enables the fabrication of such extended pistons. FIG. 7C shows a detailed cross sectional view of a segment the accelerometer. There are two types of capacitances in the bidirectional accelerometer. The first one is the capacitance that is formed between the lower pistons 740 and the inner sides of the corresponding tubes 710, referred to as C.sub.Lp A890, and the second one is formed between the extended pistons 730 and the inner sides of the corresponding tubes 720, referred to as C.sub.Ep A891. The lower pistons and the extended pistons are electrically separated by a dielectric layer 750 (i.e. BOX layer when using an SOI wafer).

(29) FIG. 8A-C show the operation modes of the bidirectional accelerometer. In the absence of acceleration, refer to FIG. 8A, the proof mass (tubes) is at the rest position, and the capacitance C.sub.Ep is at its maximum value as the overlapping area between the extended pistons 820 and the corresponding tubes 840 is at its maximum height (maximum height of the capacitors is bounded by the height of the tubes 870). The capacitance C.sub.Lp is at its minimum value (ideally zero) as the overlapping area between the lower pistons 810 and the corresponding tubes 830 is ideally zero. Fringe field capacitance exists in the structure. However, it can be neglected as the change rate in the overlapping area in the presence of external acceleration is very large as opposed to the fringe fields.

(30) These values of the capacitances C.sub.Ep (maximum) and C.sub.Lp (minimum) at the rest position of the proof mass are considered as parasitic or rest capacitances. Hence, reference capacitors are connected in parallel with these capacitances, C.sub.Ep and C.sub.Lp, in the read-out circuit of the accelerometer to eliminate their effect.

(31) When an external acceleration 930, refer to FIG. 8B, is applied towards the lower structure (negative direction of the z-axis), the positioning of the proof mass (tubes) changes with respect to the lower structure such that C.sub.Ep B891 reduces linearly, and hence this linear change of capacitance is detected by the read-out circuit. On the other hand, C.sub.Lp B890 ideally doesn't change as there is no electrode engagement between the lower pistons and the corresponding tubes (the change in the overlapping area of C.sub.Lp is still zero similar to that of the rest position of the proof mass).

(32) When an external acceleration 960, refer to FIG. 8C, is applied towards the upper structure (positive z-axis direction), the positioning of the proof mass (tubes) changes with respect to the lower structure such that C.sub.Lp C890 increases linearly while C.sub.Ep C891 ideally doesn't change as the net change in the overlapping area is still zero as the overlapping area is similar to the rest position of the proof mass. This linear change of capacitance is detected by the read-out circuit.

(33) Another configuration for capacitance measurement by the read-out circuit is achieved by electrically connecting the extended pistons (C.sub.Ep) to the lower pistons (C.sub.Lp) to form a single capacitance, i.e C.sub.total=C.sub.Ep+C.sub.Lp. A reference capacitor has the value of the parasitic capacitance of C.sub.Ep and C.sub.Lp (C.sub.total at rest position) is connected in parallel with the accelerometer total capacitance C.sub.total to an op amp circuit. When an acceleration 960 is applied upwards along the positive direction of the out-of-plane axis (z-axis), the distance between the lower structure and moving structure of the accelerometer reduces, hence the overall accelerometer capacitance (C.sub.total) increases almost linearly (the increase is only in C.sub.Lp while the change of C.sub.Ep is zero), and the output voltage of the op amp is positive as C.sub.total is greater than C.sub.ref. When the acceleration is applied downward 930, the moving structure of the accelerometer moves away from the lower structure, hence the capacitance of the accelerometer decreases almost linearly (the decrease is only in C.sub.Ep while the change of C.sub.Lp is zero); the output of the op amp is negative as the accelerometer capacitance (C.sub.total) becomes less than C.sub.ref. FIG. 9 shows such an accelerometer capacitance change. This estimated capacitance change is for an embodiment of area of 44 mm incorporating thousands of pistons and tubes in response to a relative displacement of the moving structure of the accelerometer along the z-axis. One with ordinary skills in the art will readily understand that for the bi-axial angular acceleration measurement, the accelerometer will have four sensing electrodes (i.e. four arrays of pistons), each sensing electrode of the four comprises extended pistons and lower pistons forming a single capacitor, i.e. C.sub.total, with the moving electrode.

(34) Such bidirectional accelerometer has the ability to measure the linear acceleration in two directions along the z-axis; however, it has less capacitive sensitivity when compared to the unidirectional accelerometer (the sensitivity is reduced by half as half of the pistons are used to measure the acceleration on each side).

(35) It will be also readily apparent to one who is skilled in the art that the shape of the pistons (teeth) and the corresponding tubes in the present invention could be designed in different geometrical shapes. They could be rectangular as shown in the embodiments illustrated in the present invention, or square-shaped, circular, pentagonal, hexagonal or trapezoidal. The configuration of the springs will then be arranged accordingly such that the longitudinal direction of the springs (the direction that corresponds to the largest stiffness of the springs) is normal to the parallel plate capacitance between the inner sides of the tubes and the outer sides of the pistons (to avoid the lateral instability). And the transverse (sensitive) direction is parallel to the axis of the acceleration needed to be measured.

REFERENCES

(36) [1] B. V. Amini, R. Abdolvand, and F. Ayazi, Sub-micro-gravity capacitive SOI microaccelerometers, Dig. Tech. Pap. Int. Conf. Solid State Sensors Actuators Microsystems, TRANSDUCERS '05, vol. 1, pp. 515-518, 2005. [2] D. Zhao, A Low-Noise CMOS Interface for Capacitive Microaccelerometers, Georgia Institute of Technology, 2009. [3] B. V. Amini and F. Ayazi, Micro-gravity capacitive silicon-on-insulator accelerometers, J. Micromechanics Microengineering, vol. 15, pp. 2113-2120, 2005. [4] M. A. ERM, MEMS Accelerometers and Gyroscopes for Inertial Measurement Units, MIDDLE EAST TECHNICAL UNIVERSITY, 2004. [5] N. Yazdi, F. Ayazi, K. Najafi, and S. Member, Micromachined Inertial Sensors, vol. 86, no. 8, 1998. [6] L. San and V. Foronda, Automotive Events Detection using MEMS Accelerometers, Universitat Politcnica de Catalunya. [7] J. Bernstein, R. Miller, W. Kelley, and P. Ward, Low-noise MEMS vibration sensor for geophysical applications, J. Microelectromechanical Syst., vol. 8, no. 4, pp. 433-438, 1999. [8] H. Kulah, J. C. J. Chae, and K. Najafi, Noise analysis and characterization of a sigma-delta capacitive silicon microaccelerometer, TRANSDUCERS '03. 12th Int. Conf. Solid-State Sensors, Actuators Microsystems. Dig. Tech. Pap. (Cat. No. 03TH8664), vol. 1, no. 2, pp. 352-361, 2003. [9] M. F. Pantano, L. Pagnotta, and S. Nigro, A numerical study of squeeze-film damping in MEMS-based structures including rarefaction effects, Ital. Res. smart Mater. MEMS, vol. 23, pp. 103-113, 2013. [10] C. Acar and A. Shkel, MEMS Vibratory Gyroscopes: Structural Approaches to Improve Robustness. 2009, p. 104.