JPH06160419A - Semiconductor acceleration sensor and its detecting device - Google Patents

Abstract

PURPOSE:To provide a semiconductor acceleration sensor the characteristics of which do not vary much and the output of which has excellent linearity against acceleration changes. CONSTITUTION:After forming an n-layer (diffusion layer) 10a at a prescribed middle position of a silicon board 10 formed basically of P-type silicon, the P-type silicon is removed from the prescribed part of the board 10 by performing electrochemical etching from the bottom side until the etching reaches the layer 10a and a beam section 13 is constituted of the remaining layer 10a. Since the thickness of the layer 10a can be accurately controlled, the thickness of the section 13 can be accurately controlled and the sensitivity variation between each sensor can be suppressed. In addition, when the areas S1 and S2 of electrodes and distances d1 and d2 between the electrodes and this sensor are set so that the relation between them can be expressed by d2=d1(S2/S1)<1/3>, the output of the sensor in its normal condition coincides with the changing point of the output characteristics of the sensor.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a capacitance type semiconductor acceleration sensor and its detecting device.

[0002]

2. Description of the Related Art A conventional electrostatic capacitance type semiconductor acceleration sensor is, for example, as shown in FIG. In the illustrated example, a so-called differential type is used, in which a weight portion 3 is integrally formed at the center of a silicon plate 1 which is a semiconductor substrate via a beam portion 2, and the upper and lower surfaces of the silicon plate 1 are both formed. The glass plate 4 is placed in the. At this time, a predetermined gap is formed between the upper and lower surfaces of the weight portion 3 and the facing surface of the glass plate 4 to allow the weight portion 3 to swing. The upper and lower surfaces of the weight portion 3 are used as the first and second movable electrodes 5 and 6, and the first and second movable electrodes 5 and 6 are fixed at predetermined positions on the inner surface of the glass plate 4 by aluminum vapor deposition or the like. The electrodes 7 and 8 are formed.

By the way, the first movable electrode 5 and the first movable electrode 5
A static capacitance C1 is generated between the second movable electrode 6 and the second fixed electrode 8, and a static capacitance C2 is generated between the second movable electrode 6 and the second fixed electrode 8. Then, when acceleration is applied to this sensor, the beam portion 2 bends, the distance between the opposing electrodes 5 and 7, 6 and 8 changes, and the capacitances C1 and C2 also change due to the change in this distance. . At this time, the acceleration is obtained by detecting the capacitance difference ΔC (ΔC = C1−C2) in both sexes.

By adopting such a configuration (differential type),
Since the temperature characteristics and the parasitic capacitances of the first and second electrostatic capacitances C1 and C2 are subtracted from each other and cancelled, the characteristics are improved. Further, the linearity of the output (capacitance difference ΔC) with respect to the change in acceleration is improved for the following reason. That is, the electrode area on the upper side (the area where the first movable electrode and the fixed electrode are overlapped, which are usually equal to each other or equal to the size of the first movable electrode 5 in order to enlarge the fixed electrode side) is S1. , The lower electrode area (equal to the size of the second movable electrode 6 for the same reason) is S2, and the distance between the opposing electrodes is d1,
Let d2. In this state, acceleration is applied and the weight part 3 becomes x.
The capacitances C1 and C2 when displaced by (moving downward for convenience for convenience) are C1 = εS1 / (d1 + x) (1) C2 = εS2 / (d2-x) (2) The difference in capacitance ΔC is

[0005]

ΔC = εS1 / (d1 + x) −εS2 / (d2-x) (3) In terms of characteristics, it is preferable to form the shape in a vertically symmetrical manner. Therefore, when S1 = S2 and d1 = d2, the capacitance and the difference between the capacitance and the acceleration are shown in FIG.
Shown in 1. The displacement x is assumed to be proportional to the magnitude of acceleration. Then, as is clear from the figure, in the vicinity of the acceleration of 0 G, the linearity is better in ΔC than in the case of C1 and C2 alone.

[0006]

However, the above-mentioned conventional differential type acceleration sensor has the following problems. That is, the beam portion 2 and the weight portion 3 are formed by anisotropically etching a flat plate-shaped silicon substrate and removing a predetermined portion of the substrate. The removal amount, that is, the thickness of the beam portion 2 is Controlled by etching time,
Since the silicon wafers to be processed have different thicknesses, the thickness of the manufactured beam portion 2 also varies. Then, the rigidity (elastic coefficient) of the beam portion 2 also varies, and the sensitivity of the sensor also varies.

When detecting the acceleration in the vertical direction (the same direction as the direction of gravitational acceleration), in the steady state,
Since the gravitational acceleration (1 G) is constantly applied to the weight portion 3, the weight portion 3 is actually located below the state shown in FIG. 10, that is, displaced by a predetermined distance x. Therefore, the lower distance d2 becomes shorter than the upper distance d1, and the capacitance changes around the point Q in the characteristic diagram shown in FIG. As a result, the linearity decreases.

Further, all of the conventional acceleration sensors have been designed to detect a change in capacitance as a change in frequency. For example, such an acceleration sensor controls the operation of an active suspension of a vehicle or an airbag system. When it is used as a detection unit for this purpose, a sufficient response cannot be obtained by acceleration detection using such a frequency.

The present invention has been made in view of the above background, and an object of the present invention is to provide a semiconductor acceleration sensor having a small variation in characteristics and excellent linearity with respect to changes in acceleration. Another object of the present invention is to provide a detection device used therefor.

[0010]

In order to achieve the above object, in a semiconductor acceleration sensor according to the present invention, a weight portion which is integrally connected to a frame body through a beam portion and which is displaced in accordance with acceleration. A semiconductor plate having movable electrodes formed on both surfaces thereof, and a substrate such as a glass plate provided with fixed electrodes facing the movable electrodes with a predetermined gap, respectively, and sandwiching the semiconductor plate. A semiconductor acceleration sensor that detects acceleration from a difference in electrostatic capacitance between the movable electrode and the fixed electrode that form a pair with the displacement, and the semiconductor plate is arranged at a predetermined position. A boundary surface between the two layers is located on the surface of the beam portion that supports the weight portion so as to be displaceable and the extension portion thereof.

Further, preferably, the distance between the movable electrode formed on one surface of the weight portion and the fixed electrode facing the movable electrode and the electrode area thereof, and the distance formed on the other surface of the weight portion. The distance between the movable electrode and the fixed electrode facing the movable electrode and the electrode area thereof are set so that both capacitances generated between the paired electrodes are substantially equal in a reference state where the acceleration of the detection target is not measured. It is to be.

Further, as a detection device of a differential type semiconductor acceleration sensor, which is provided with at least two sets of movable electrodes and fixed electrodes, and which detects the generated acceleration from the difference in each capacitance generated between both electrodes, An oscillation circuit for inputting a predetermined oscillation signal to each of the electrically connected movable electrodes and an oscillation circuit connected to each of the fixed electrodes and obtained by differentiating the given oscillation signal by each of the capacitances. And means for obtaining an output according to the acceleration by subtracting each PWM signal.

Further, as another detecting device, there is provided a means which is connected to each of the fixed electrodes and which alternately applies a predetermined voltage between the pair of electrodes, and the movable electrode which is electrically connected. And a means for simultaneously inputting charges accumulated in each capacitance by means for applying the voltage to the differential amplifier circuit, wherein the differential amplifier circuit calculates the difference between the capacitances as a voltage. The output voltage corresponding to the acceleration may be obtained by converting into

[0014]

Since the p layer and the n layer are provided at predetermined positions of the semiconductor substrate on which the movable electrode is formed, when the electrochemical etching is performed from one side (for example, the p layer), the etching portion of the p layer is removed until the n layer is reached. To be done. Therefore, the thickness of the beam portion can be accurately controlled by forming the beam portion with such an n-layer portion, for example.

Then, in the reference state (when gravitational acceleration is applied and the weight portion (movable electrode) is displaced by a predetermined amount,
If the capacitances are set to be approximately equal after the displacement (nothing other than gravitational acceleration), the linearity of the capacitance difference (sensor output) is good. And becomes a highly sensitive sensor.

Further, if the detection device of the present invention is used, the difference in each capacitance can be taken out as a DC component,
Acceleration with high responsiveness is detected.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of a semiconductor acceleration sensor and its detecting device according to the present invention will be described in detail below with reference to the accompanying drawings. FIG. 1 shows a first embodiment of an acceleration sensor according to the present invention. As shown, a glass plate 11 is arranged on both upper and lower sides of a silicon plate 10 which is a semiconductor plate, and the silicon plate 10 has a plurality of beams with respect to a square frame 12. Weight part 14 through part 13
It is composed of a cantilevered support connected. Then, the upper and lower surfaces of the weight portion 14 are attached to the first and second movable electrodes 15, 1
6, and the first and second fixed electrodes 17, 1 are formed on the surface of the glass plate 11 facing both electrodes 15, 16 by aluminum vapor deposition or the like.
8 is the same as the conventional one.

In the present invention, the n layer (diffusion layer) 10a is formed by, for example, doping at a predetermined position in the middle of the silicon plate 10 formed based on P type silicon. And this n layer constitutes the upper side of the beam part 13 and the weight part 14 continuous to it.

To manufacture the sensor (silicon plate) having such a structure, a plate-shaped silicon plate on which the n layer 10a is formed is electrochemically etched from a predetermined portion from below. Then, the P-type silicon is removed until it reaches the n-layer 10a. Therefore, the thickness of the beam portion 13 is n layer 1
0a, and the thickness can be controlled easily and precisely by the diffusion depth of the n-layer 10a, so that variations in performance (sensitivity) among the manufactured sensors can be suppressed as much as possible. .

In this example, since the electrochemical etching is performed from the lower side, the area S1 of the first movable electrode 15 becomes larger than the area S2 of the second movable electrode 16,
Along with this, the areas of the corresponding fixed electrodes are also made different (substantially the same as the areas of the pair of movable electrodes), but as shown in FIG. 10 shown as a conventional example, it is of course possible to apply the shapes to the upper and lower sides. . Further, instead of providing the n diffusion layer at a predetermined portion of P-type silicon as described above, a plate-like (having substantially the same plane shape) P-type silicon and N-type silicon are laminated and arranged in layers. (In such a case, the N-type silicon and the P-type silicon are also located over the entire circumference of the frame body 12).

Further, in this example, since the areas of the upper and lower electrodes are different as described above, if the distances d1 and d2 between the electrodes forming each pair are made equal, the linearity of the output with respect to the change in acceleration is lowered. Further, as shown in FIG. 2, when the beam portion 12 and the weight portion 13 are arranged in a horizontal state in order to detect the acceleration in the vertical direction, gravitational acceleration is constantly applied to the weight portion 13, and the acceleration to be detected is detected. Even when there is no load, the weight portion moves downward as shown in the figure, and even if d1 = d2 is set in the state shown in FIG. 1, d1> d2 in the actual usage situation.

In this case, in order to improve the linearity, the relationship between the areas S1 and S2 and the distances d1 and d2 is set as shown below. That is, from the reference state (distance between paired electrodes is d1 and d2, respectively) where no acceleration (excluding gravitational acceleration that is always applied) is applied, a predetermined acceleration is applied and the weight portion is applied. The output ΔC when 13 is displaced downward by x is as described above.

[0023]

## EQU00002 ## .DELTA.C = .epsilon.S1 / (d1 + x)-. Epsilon.S2 / (d2-x) (3).

Therefore, the condition that the linearity of ΔC with respect to the acceleration at this time is the best is that in the reference state, that is, ΔC when x = 0 becomes the transition point P shown in FIG. That is, specifically, the second derivative of the above equation (3) becomes zero. Therefore, the second derivative Δ
C ″ is

[0025]

ΔC ″ = k (S1 / (d1 + x) ³⁻ S2 / (d2-x) ³ ) (4) Here, k is an arbitrary constant.

The condition for ΔC ″ = 0 to hold when x = 0 is d2 = d1 (S2 / S1) ^1/3 (5) by substituting each value into the equation (4). .

Therefore, each value is set so that the above conditional expression (5) is satisfied. That is, since the areas S1 and S2 are first determined in the manufacturing process, the area is substituted into the equation (5) to obtain the ratio of d1 and d2. The semiconductor plate 1 shown in FIG.
By removing a predetermined amount of the upper surface side 10b of 0 by etching or the like, it is possible to obtain a sensor structure having the best linearity. The adjustment for obtaining the optimum structure is not limited to the above, and various methods such as removing an arbitrary portion can be adopted. In addition, the above-mentioned electrode areas S1 and S2 are actually effective areas, that is, the area where the pair of electrodes overlap (when the area of the fixed electrode is small, the area of the fixed electrode).

Furthermore, the linearity is best satisfied if the above conditional expression (5) is satisfied, but it is not necessary to satisfy this condition exactly, and according to the specifications (target allowable error range) and the like, It may be set with a certain range.

As shown in FIG. 2, the amount of deflection due to gravitational acceleration is calculated from the weight of the weight portion 14 and the elastic coefficient of the beam portion 13 to find out how much it is displaced from the horizontal position, and is actually manufactured. When doing, consider the amount of displacement,
It is necessary to make adjustments such as addition to d1 and d2 obtained by the above conditional expressions.

FIG. 3 shows a first embodiment of the detection device according to the present invention. As shown in the figure, the capacitance C generated between the two sets of electrodes of the acceleration sensor in the above embodiment
1, C2 and two resistors R1 and R2 form a bridge circuit. The connection side of the two electrostatic capacitances C1 and C2 is configured by connecting the wirings connected to the first and second movable electrodes 15 and 16 formed on the silicon plate 10, for example, to the both movable sides. Oscillation circuit 20 on electrodes 15 and 16
The output of is applied.

As a connection structure of both electrodes 15 and 16, for example, as shown in FIG. 4, the second movable electrode 16 is formed by applying aluminum vapor deposition 19 or the like on the lower surface side of the weight portion 13. , The aluminum vapor-deposited end 19a is an n-layer 1
By connecting to 0a, both movable electrodes 15 and 16 have the same potential, and wiring can be easily drawn out.

Further, the other ends of the electrostatic capacitances C1 and C2 (wirings of the fixed electrodes 17 and 18 in this example) are connected to the differential amplifier circuit 23 via the Schmitt circuits 21a and 21b and the rectifier circuits 22a and 22b. .

Next, the operating principle of the above-described embodiment will be described with reference to the timing chart shown in FIG.
First, the oscillating circuit 20 outputs a rectangular wave having a period T ((A) in the same figure), and differentiates it by the electrostatic capacitances C1 and C2 ((B) in the same figure). Then, by inputting the differentiated signals to the Schmitt circuits 21a and 21b, a rectangular wave is created ((C) in the figure). Then, the pulse width t of this rectangular wave
Is proportional to the time constants C1R1 and C2R2 (PWM
signal).

Then, the rectangular wave is rectified by the rectifier circuits 22a and 2a.
2b, and a DC signal corresponding to the pulse width t, that is, the electrostatic capacitances C1 and C2 is obtained there ((D) in the same figure). Then, these both DC signals are fed to the differential amplifier circuit 2
By inputting the signal to the circuit 3, a DC signal proportional to the above ΔC is output from the circuit 23. In the illustrated example, each waveform has the same shape because C1 = C2. However, if acceleration changes and the capacitance changes,
The duty ratio of each waveform changes appropriately.

With this configuration, the difference (ΔC) between the two capacitances can be seen as the magnitude (change) of the DC signal with the magnitude (change) of the acceleration, and the responsiveness is good. When mounted in an airbag system, an operation signal is issued when the output of the differential amplifier circuit 23 exceeds a certain threshold value (for example, by simply inputting the output of the circuit and the threshold value of the previous period to the comparison, Can be configured) and so on.

An example of a concrete circuit configuration of this embodiment is shown in FIG. In this example, the output is 2.5V in the standard state.
Has been adjusted to be. Further, in the present embodiment, the DC signal is output. However, for example, when the DC signal is not required in the subsequent device, the Schmitt circuit 21
By taking out the PWM signals from a and 21b, digital processing becomes possible.

The detection device according to the present invention can be applied not only to the output detection of the above-mentioned acceleration sensor according to the present invention but also to the conventional differential type acceleration sensor other than that shown in FIG. May be applied to (the same applies below).

FIG. 7 shows a second embodiment of the detecting device according to the present invention. As shown in the figure, first, a direct current voltage (Vin) is applied to the first and second fixed electrodes 17 and 18 of the semiconductor acceleration sensor 15 via the first and second analog switches 30a and 30b. I have to. Further, the first and second movable electrodes 15 and 16 which are electrically connected to each other and serve as a common electrode are connected to an inverting input terminal of an operational amplifier 32 which is a differential amplifier via a third analog switch 31. I am trying.

Then, the first and second analog switches 3
The operation timings φ1 and φ2 of 0a and 30b are as shown in FIG. 8, and the operation timing of the third analog switch 31 is the same as that of the second analog switch 30b. With such a configuration, a switched capacitor circuit is formed, and its capacitance C
It can be detected as a differential output (voltage) corresponding to the difference (ΔC) between 1 and C2.

FIG. 9 shows a third embodiment of the detection device according to the present invention. The structure thereof is substantially the same as that of the second embodiment described above, but the first and second fixed electrodes 17 are provided. , 18
The voltages V1 and V2 applied to the are different. Since the other configurations and operations are similar to those of the second embodiment described above, the same reference numerals are given and the description thereof is omitted.

Predetermined charges Q1 and Q2 are stored in the electrostatic capacitances C1 and C2 in the acceleration sensor to which the voltages V1 and V2 are applied, and the charges Q1 and Q2 are stored.
Is output as an output voltage Vout as shown in the following expression via the operational amplifier 32 (note that if V1 = V2 in the following expression (6), it becomes the output voltage in the body 2 embodiment).

[0042]

## EQU00004 ## Vout = .alpha. (Q1-Q2) =. Alpha..epsilon. (S1 * V1 / (d1 + x) + S2 * V2 / (d2-x)) (6) where .alpha. Is a constant determined by the circuit constant The condition that the linearity of Vout with respect to the acceleration is the best is that it is located at the inflection point P shown in FIG. 11 in the reference state (x = 0) as described above.
When the second derivative (Vout ″) of the above equation (6) when x = 0 is solved as 0,

[0043]

From Vout ″ = k (S1 * V1 / (d1 + x) ³ −S2 * V2 / (d2-x) ³ V2 = V1 * (S1 / S2) * (d2 / d1) (7) .

Therefore, by setting the voltages V1 and V2 that satisfy the condition of the above expression (7), the linearity with respect to the acceleration can be optimized. That is, even if there is a difference or variation in the electrode area of the differential type acceleration sensor or the distance between the electrodes, the linearity is good only by applying a voltage satisfying the above formula (7) or a condition close thereto. Acceleration sensor (detection device) can be configured.

[0045]

As described above, in the semiconductor acceleration sensor according to the present invention, the beam portion is made of a material (n layer or p layer) different from that of other weight portions, frame bodies, etc. By removing the unnecessary portion by etching or the like, the beam portion can be configured only with the relevant n layer (p layer). That is, the thickness of the beam portion becomes the thickness of the n layer (p layer), and can be accurately formed to a predetermined thickness. Therefore,
Variations in characteristics (sensitivity) are reduced. When the area of each electrode and the distance between the electrodes are set so that the capacitances are substantially equal in the reference state, output characteristics with good linearity with respect to changes in acceleration can be obtained. Further, in the detection device according to the present invention, it is possible to output as a DC signal corresponding to the difference between the electrostatic capacities, so that the detection is simple and the response is good.

[Brief description of drawings]

FIG. 1 is a sectional view showing a preferred embodiment of a semiconductor acceleration sensor according to the present invention.

FIG. 2 is a diagram showing an example of a usage state.

FIG. 3 is a diagram showing a first embodiment of a detection device according to the present invention.

FIG. 4 is a block diagram showing an example of connection of first and second movable electrodes.

5 is a timing chart of the circuit shown in FIG.

6 is a circuit diagram showing a specific configuration of the circuit shown in FIG.

FIG. 7 is a diagram showing a second embodiment of the detection device according to the present invention.

FIG. 8 is a diagram showing operation timings of analog switches.

FIG. 9 is a diagram showing a third embodiment of the detection device according to the present invention.

FIG. 10 is a sectional view showing a preferred embodiment of a conventional semiconductor acceleration sensor.

FIG. 11 is an endless example of characteristics of electrostatic capacitance with respect to acceleration and the difference thereof.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 10 Silicon plate (semiconductor plate) 10a n layer 11 Glass plate (substrate) 12 Frame body 13 Beam part 14 Weight part 15 1st movable electrode 16 2nd movable electrode 17 1st fixed electrode 18 2nd fixed electrode 20 Oscillation circuit 21a, 21b Schmitt circuit (means for obtaining output according to acceleration) 22a, 22b Rectifier circuit (means for obtaining output according to acceleration) 23 Differential amplifier circuit (means for obtaining output according to acceleration) 30a, 30b First and second analog switches (means for alternately applying a predetermined voltage) 31 Third analog switches (means for inputting to a differential amplifier circuit) 32 Operational amplifier (differential amplifier circuit)

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Yoshiyuki Morita, 10 Odoroncho, Hanazono Tadodo-cho, Ukyo-ku, Kyoto City, Kyoto Prefecture

Claims (4)

[Claims] 1. A semiconductor plate in which movable electrodes are formed on both surfaces of a weight portion that is integrally connected to a frame body through a beam portion and that is displaced in accordance with acceleration; and a predetermined gap between the movable electrodes. And a substrate such as a glass plate arranged so as to sandwich the semiconductor plate, between the movable electrode and the fixed electrode that are paired with the displacement. Is a semiconductor acceleration sensor that detects acceleration from the difference in electrostatic capacitance that occurs in the semiconductor plate, wherein the semiconductor plate includes an n-layer and a p-layer disposed at predetermined positions, and the beam portion that displaceably supports the weight portion. A semiconductor acceleration sensor in which a boundary surface between the two layers is located at an extension portion of the surface of the. 2. The distance between the movable electrode formed on one surface of the weight portion and the fixed electrode facing the movable electrode and its electrode area, and the movable electrode formed on the other surface of the weight portion. The distance between the fixed electrode and the fixed electrode and the area of the fixed electrode are adjusted so that both capacitances generated between the paired electrodes are substantially equal in a reference state where no acceleration is applied. 1. The semiconductor acceleration sensor according to 1. 3. A differential type semiconductor acceleration sensor detecting device, comprising at least two sets of movable electrodes and fixed electrodes, for detecting acceleration from a difference between electrostatic capacitances generated between the electrodes. An oscillation circuit for inputting a predetermined oscillation signal to each of the movable electrodes that are electrically connected, and an oscillation circuit that is connected to each of the fixed electrodes and is provided by differentiating the given oscillation signal by each of the capacitances. A detection device for a semiconductor acceleration sensor, comprising: means for obtaining an output according to acceleration by subtracting each PWM signal. 4. A detection device of a differential type semiconductor acceleration sensor, comprising at least two sets of movable electrodes and fixed electrodes, and detecting acceleration from a difference in electrostatic capacitances generated between the electrodes, Means connected to each fixed electrode and alternately applying a predetermined voltage between each pair of electrodes, and means electrically connected to the movable electrode to apply each voltage to each electrostatic electrode. A means for simultaneously inputting the charges accumulated in the capacitance to the differential amplifier circuit is provided, and the differential amplifier circuit converts the difference between the capacitances into a voltage to obtain an output voltage according to the acceleration. Detection device for semiconductor acceleration sensor.