JPH01152369A - Capacity type acceleration sensor - Google Patents

Abstract

PURPOSE:To obtain a highly sensitive and highly accurate sensor excellent in impact resistance with a simple construction, by etching an Si plate on both sides thereof to form a mobile electrode playing a role of a weight at the tip of a cantilever beam and a glass plate having a solid electrode is bonded on both the sides of the Si plate. CONSTITUTION:An Si plate 1 with a diameter of 3-4 inch, and a thickness of 0.2mm is etched on both sides thereof to form a mobile electrode 3 playing a role of a weight at the tip of a cantilever 2 with a thickness of several ten m. The mobile electrode 3 is etched so as to be 2-3mum in the depth from the surface of the Si plate. Glass plates 6 and 7 having solid electrodes 4 and 5 on the surface thereof are bonded on both sides of the Si plate 1. A distance changes between the solid electrodes 4 and 5 and the mobile electrode 3 depending on a Y-wise acceleration thereby detecting acceleration as change in the capacity between the electrodes. The cantilever 2 is so thick as to be excellent in the impact resistance. A resonance point gives about 1kHz to achieve a high accuracy at 0-10Hz as measuring range without damping. A small gap between the electrodes assures a high sensitivity.

Description

Detailed Description of the Invention [Field of Industrial Application] The present invention relates to an acceleration sensor, and particularly to an acceleration sensor suitable for an automobile body control system. A device that can detect the range of ~1 OHz with high accuracy is required. Here, I
G=9.8m/S2.

[Conventional technology]

Acceleration sensors include piezoelectric types that use the piezoelectric effect of piezoelectric materials, strain gauge types that use piezoresistive effects, and servo types that have a force feedback mechanism.

A number of methods are known, including a magnetic method using a differential transformer, a front-end method using a photointerrupter, and a capacitive method using silicon microfabrication technology.

Regarding these acceleration sensors, for example, silicon
Diaphragm Capacitive Sensor for Pressure Flow Acceleration and Attitude Measurement (Transducers'
87, The 4th International
lConference on 5solid-5tat
e 5 sensors and actuators, 1
The technique described in June 1987) is publicly known.

[Problem that the invention seeks to solve]

Among the various acceleration sensors mentioned above, the piezoelectric type is easy to manufacture, but cannot detect acceleration of low frequency components.

The strain gauge type is small and highly accurate, but it is resistant to W1! ! inferior to sex. The servo type has high precision, but has a complicated structure and is expensive.

The magnetic type is large and has a complicated structure, and the front type has low measurement resolution. Furthermore, the capacitive type using silicon microfabrication technology has low sensitivity and resolution, and is affected by acceleration components in other axis directions. As described above, the acceleration sensors of the prior art have the above-mentioned drawbacks when considering application to a vehicle body control system.

An object of the present invention is to provide an acceleration sensor that has a simple structure (low cost), excellent wIqJ resistance, high sensitivity, and high precision.

[Means for solving problems]

The above object is achieved by devising the structure of the detection section in a capacitive acceleration sensor that utilizes silicon microfabrication technology. That is, after etching a silicon plate from both sides to form a movable electrode that acts as a weight at the tip of the cantilever, glass plates with fixed electrodes formed on the surface are adhered to both sides of the silicon plate, and the gap between the movable electrode and the fixed electrode is By adopting a structure in which acceleration is detected from a change in capacitance, the drawbacks of the prior art can be overcome.

If two of the above cantilevers are provided and arranged concentrically, a double-supported beam will be formed. Even with this configuration, similar actions and effects can be obtained.

By forming a movable electrode at the tip of a silicon cantilever and mounting it so that the gap between the movable electrode and the fixed electrode is several μm, accelerations of about 0 to ±IG can be measured with high sensitivity with small displacements of the movable electrode. be able to. At this time, even if the silicon cantilever is made as thick as several tens of micrometers, the capacitance change between the electrodes with respect to acceleration can be large, so high sensitivity can be achieved without impairing impact resistance. When the center of gravity of the movable electrode is located on the central axis of the cantilever, the cantilever does not receive a bending moment with respect to acceleration components other than the direction of displacement of the movable electrode. Therefore, only the acceleration component perpendicular to the movable electrode can be detected.

When etching a silicon plate from both sides to form a cantilever and a movable electrode, there is no need for special passivation during two-sided etching because there is no diffused resistor or electrode like in a semiconductor strain gauge type acceleration sensor. The yield is improved and costs can be reduced. Since the silicon cantilever can be made as thick as several tens of micrometers, the resonance point of the displaced portion consisting of the cantilever and the movable electrode will be on the order of 1 kHz.

This value is sufficiently large compared to the measurement range of 0 to 10 Hz, and it is possible to improve impact resistance and detection accuracy even without damping the displacement part by filling it with silicone oil or the like. can. This simplifies the structure and manufacturing method of the detection section, resulting in lower costs. A silicon plate with a cantilever and a movable electrode is bonded between it and two glass plates having approximately the same coefficient of thermal expansion. Therefore, the thermal distortion of the contact area is small, and the temperature change during use (-40 to +10
0° C.), the displacement of the movable electrode with respect to temperature is extremely small. Therefore, the temperature characteristics of the acceleration sensor are good,
A highly accurate acceleration sensor can be realized. Note that the detection section has a symmetrical stacked structure with respect to the silicon cantilever. Therefore, the thermal strain of the adhesive portion is equal in the upper glass plate portion and the lower glass plate portion, and the factors that cause the movable electrode to move up and down in response to temperature changes can be canceled out and the displacement can be further reduced. In this way, the symmetrical stacked structure of the detection section also greatly contributes to improving the temperature characteristics.

An embodiment of the present invention will be described below with reference to FIG. The detection part of the acceleration sensor according to this embodiment is constructed by anisotropically etching a silicon plate 1 from both sides to form a cantilever 2 and a movable electrode 3 serving as a weight at its tip. 5 and a glass plate 6 formed by vapor deposition or other method.
and 7 are bonded to both sides of a silicon plate 1. The movable electrode 3 is made of silicon film by a groove-like through groove 8.
Since it is separated from the main body, it can be displaced up and down according to the acceleration component perpendicular to it (in the direction of the arrow Y in the figure). Note that 9 is an extraction electrode of the movable electrode 3, and is connected to an external electronic circuit via this.

When the movable electrode 3 is displaced up and down according to the acceleration G, the capacitance between the movable electrode 3 and the fixed electrodes 4 and 5 changes. The sensor according to the present invention detects acceleration from this capacitance change. As an acceleration sensor for vehicle body control, it is necessary to detect a specific acceleration component (in the Y direction in the figure) with high precision, and it is necessary to detect acceleration components in other axis directions (in the X direction perpendicular to the Y axis in the figure). must not. The silicon plate 1 is etched from both sides so that the center of gravity of the movable electrode 3 having a weight function is located on the central axis of the silicon cantilever 2 in the longitudinal direction. With this configuration, when the cantilever 2 receives acceleration in the X-axis direction, a bending moment that displaces the movable electrode 3 up and down (in the Y-axis direction) does not act on the cantilever 2 . As a result, the acceleration sensor according to the present invention is insensitive to the acceleration component in the X-axis direction,
Only the acceleration component in the Y-axis direction can be accurately detected.

If the gap between the movable electrode 3 and the fixed electrodes 4 and 5 changes due to factors other than acceleration (for example, changes in ambient temperature), a detection error will occur. Since the acceleration sensor is used within a wide temperature range of -40 to +100° C. after being attached to the vehicle body, it is required to be less affected by temperature.

The silicon plate 1 and the glass plate 6,
Since the temperature effect of thermal strain on the adhesive portion with 7 becomes a major problem, it is necessary to minimize the thermal strain on this portion. The glass plates 6 and 7 on which the fixed electrodes 4 and 5 are formed by vapor deposition or other methods are made of borosilicate glass whose thermal expansion coefficient is equal to that of the silicon plate 1. By bonding the silicon plate 1 and the glass plate 4.5 using the well-known anodic bonding method (a method of bonding by applying a high voltage at a high temperature of 300°C or higher), thermal strain at the bonded portion is minimized. Can be made smaller. As is clear from the figure, silicon plate 1. Cantilever 2. The structure of the detection section made of the movable electrode 3 is a vertically symmetrical stacked body.

The predisposition for vertically displacing the movable electrode 3 due to thermal strain at the bonded portion between the glass plates 6 and 7 is canceled out by each other due to the symmetrical laminated structure. In this way, in this embodiment, the structure of the detection section itself also improves the temperature characteristics of the acceleration sensor.

Figure 2 shows the basic principle of a capacitive acceleration sensor.

The capacitance C1 between the movable electrode 3 and the fixed electrode 4 and the capacitance C2 between the movable electrode 3 and the fixed electrode 5 are connected to an electronic circuit (not shown) via electrode lead leads 10, 11, 12. )
connected to and detected. The initial gap dO between the movable electrode 3 and the fixed electrode 4 is made to be several μm. Now, when the movable electrode 3 receives a force due to acceleration and is displaced by W toward the fixed electrode 5, the capacitance C1 decreases and the capacitance C2 increases.

The acceleration G acting on the vehicle body is detected from this capacitance difference value ΔC (Δc=cz−Ct).

Figure 3 shows an exploded view of the detection unit. The detection part consists of an upper glass plate part, a middle silicon plate part, and a lower glass plate part. Holes 15 and 16 are made in the glass plate 6 by sandblasting or other methods, and the fixed electrode 4 is formed on one side and the lead-out M1 pole 13 is formed on the opposite side by a method such as vapor deposition or sputtering.

The fixed electrode 4 is connected to the extraction electrode 13 and a plating connection lead portion 14 formed by plating on the inner surface of the hole 15, and is connected to an external capacitance detection circuit (not shown) from the lead extraction electrode 18 by wire bonding. .

After the cantilever 2 and the movable electrode 3 are etched on the silicon plate 1, a lead extraction electrode 17 is formed on the silicon plate 1.
is formed by vapor deposition or other methods.

A fixed electrode 5 is formed on the glass plate 7 by vapor deposition, sputtering or other methods, and is connected to an external capacitance detection circuit (not shown) through a lead extraction electrode 19.

Note that after the upper glass plate part, the middle silicon plate part, and the lower glass plate part are bonded in a wafer state, they are diced into each detection chip. Chips and cleaning water during dicing are removed from the movable electrode 3.
Before dicing, a soft adhesive (for example, silicone rubber, etc.) is injected into the holes 15 and 16 to prevent it from entering between the electrode 4.5 and the fixed electrode 4.5.

FIG. 4 shows details of the silicon plate 1. Silicon plate 1 is (
100) consists of a wafer processed from both sides by well-known anisotropic etching in an alkaline solution. A slot-like through groove 8 is formed around the cantilever 2 and the movable electrode 3, and the movable electrode 3 can be displaced up and down in response to acceleration using the fixed end of the cantilever 2 as a fulcrum.

Note that 21 indicates a diagonally etched portion that occurs when the through groove 8 is processed.

Cutting allowance mark 20. Lead extraction passage 22゜23 (2
3 formed on the back surface of the silicon plate 1) is etched to a depth of several μm. Lead extraction passage 22.23
The electrode extraction parts of the fixed electrodes 4 and 5 pass through, and the fixed electrode 4
, 5 should not come into contact with the silicon plate 1. Through hole 2
Reference numeral 4 is provided for injecting a soft adhesive into the lead extraction passage 23 through a hole 16 (see FIG. 3) formed in the glass plate 6. Reference numeral 25 denotes a lead extraction space for wire bonding which is etched to connect the fixed electrode 5 to an external capacitance detection circuit from the lead extraction electrode 19 by wire bonding.

A v-■ cross section is shown in FIG. As shown in this figure, a cantilever 2 and a movable electrode 3 are formed by etching a silicon plate 1 from both sides, and the center of gravity of the movable electrode 3, which also functions as a weight, is aligned on the central axis of the cantilever 2 in the longitudinal direction. The silicon wafer 1 is processed so as to be symmetrical with respect to the thickness direction thereof. In this example, since a diffusion resistor as in a strain gauge type acceleration sensor is not provided, no special passivation measures are required when etching the cantilever, and the cantilever can be manufactured at a high yield.

An example of the manufacturing process of the acceleration sensor according to the present invention is shown in the sixth example.
The six illustrated upper glass wafers, middle silicon wafers, and lower glass wafers in this example have a diameter of 3 to 4 inches and a thickness of 0, 3 + m, 0,
A detection section was prototyped using 2 na, 0, and 3 m. After forming holes 15 and 16 in the upper glass wafer by side blasting, a conductive metal was formed on the inner surface of the holes 15 by electroless plating.

In order to electrically connect with this plating connection lead 14,
Extracting electrode 13 on one side. Lead extraction electrode 18. A fixed electrode 4 was formed on the other surface by vapor deposition.

The middle silicon wafer has a thermal oxide film used as a protective mask during alkali etching, and a slit portion that becomes a through groove 8, a movable electrode 3. Processing was performed in the order of cantilever 2. In addition,
The lead extraction passages 22 and 23 were machined at the same time as the movable electrode 3, and the through hole 24 and the lead extraction space 25 were machined at the same time as the through groove 8 of the slit portion. Movable electrode 3. The lead extraction passages 22 and 23 are etched from the surface of the silicon plate 1 to a depth of several micrometers (preferably 2 to 3 micrometers).

Fixed electrode 5 and lead extraction electrode 19 were simultaneously formed on the lower glass wafer by vapor deposition. Note that a metal such as AQ or Au was used as the electrode material.

After stacking the upper glass wafer, middle silicon wafer, and lower glass wafer, the three wafers are aligned using an appropriate alignment jig. After heating these laminated wafers to a high temperature of 300° C. or higher, a high voltage was applied to bond them by a so-called anodic bonding method. At this time.

The movable electrode 3 is bonded to the fixed electrodes 4 and 5 so as to be located in the center thereof. When a borosilicate glass material is used for the glass plate, the coefficients of thermal expansion of the glass plate and the silicon plate are approximately equal, and the thermal distortion of the bonded surface between the two becomes extremely small. Therefore, even if the ambient temperature of the detection part changes, the vertical displacement of the movable electrode 3 about the fixed end of the cantilever 2 as a fulcrum is minute due to thermal distortion of the adhesive part. As a result, the change in capacitance between the movable electrode 3 and the fixed electrodes 4 and 5 due to changes in ambient temperature is very small, making it possible to obtain a highly accurate acceleration sensor with little temperature influence. In addition, since the etching process is performed so that the center of gravity of the movable electrode 3, which also functions as a weight, is on the central axis of the silicon cantilever 2, the detection part has a vertically symmetrical structure with respect to the displacement direction of the movable electrode 3. ing. Therefore, there is an adhesive strain at the interface between the upper glass plate and the silicon plate, and an adhesive strain at the interface between the lower glass plate and the silicon plate.

The factors that cause the movable electrode 3 to move up and down cancel each other out, and the movable electrode 3 and fixed electrodes 4, 5 in response to changes in ambient temperature
This further reduces the change in capacitance between the two and contributes to higher performance acceleration sensors.

Next, a lead extraction electrode 26 (see FIG. 7, which will be described later) for wire bonding of the movable electrode 3 was formed on the silicon plate 1 by vapor deposition.

A soft adhesive was injected into the holes 15 and 16 to prevent chips, water, etc. from entering the vicinity of the movable electrode 3 during dicing. Next, check the bonding status of the bonding interface by visual observation, measure the capacitance between the movable electrode 3 and the fixed electrodes 4 and 5 using a capacitance tester, extract and mark defective areas, and sort after dicing. , # allowed to leave.

Next, the stacked wafers were simultaneously diced to produce several hundred detection chips at a time. In this way, by performing adhesion 2 sorting and dicing of the detection portions in the wafer state, manufacturing costs can be reduced.

As shown in FIG. 7, a good detection chip is coated with an organic adhesive 34.
After bonding to the stem 31, wire bonding and cab bonding work were performed, and comprehensive characteristics evaluation as an acceleration sensor was performed.

FIG. 7 shows the subassembly structure of the detection section of the acceleration sensor according to the present invention. Glass plate 6. A detection chip in which a silicon plate 1 and a glass plate 7 are laminated is fixed onto a stem 31 using a soft organic adhesive 34 (for example, silicone rubber) having a small longitudinal elastic modulus. A hole 27 is provided in the metal stem 31, and a lead 29 is mounted in a honeycomb manner using a glass material 28. Since the thermal expansion coefficients of the stem 31, the bola, and the filter plate 7 are not equal, a soft adhesive such as silicone rubber is used for the adhesive 34 to prevent the adhesive strain in this part from being transmitted to the upper detection chip. there was.

Next, a conducting wire 30 is connected to the lead extraction electrode 26 and the lead 29.
The cap 33 is joined to the stem 31 in sNz electrically connected by wire bonding or in a dry atmosphere,
Nz or dry atmosphere was sealed in the room 32.

The view from ■-■ in FIG. 7, that is, the plan view is shown in FIG.
Leads 29-a, 29-b hermetically sealed within 27-b, 27-c.

29-C is electrically connected to an external capacitance detection circuit by a conductor. .

The relationship between displacement of the movable electrode and capacitance change will be explained with reference to FIG. 9. The movable electrode 3 moves up and down in the Y-axis direction using the fixed end 36 of the cantilever 2 as a fulcrum in response to the acceleration component in the Y-axis direction. Let dO be the initial gap between the movable electrode 3 and the fixed electrodes 4 and 5. do is several μm from the point of view of high sensitivity.
is desirable, and a prototype with do=3 μm was manufactured and evaluated. Assume that the center of gravity of the movable electrode 3 is 35 and that the center of gravity is displaced by W due to acceleration as shown by the dotted line.
As shown in the figure, as the displacement W of the center of gravity 35 of the movable electrode 3 increases, the deviation in parallelism between the movable electrode 3 and the fixed electrodes 4 and 5 increases.

Considering this discrepancy in parallelism, the movable electrode 3 and the fixed electrode 4
, 5 was analyzed. The results are shown in FIG. 10, where the movable electrode 3 is
The relationship between the capacitance CI, C2+, and the capacitance difference value ΔC (ΔC=C2Cs) between the fixed electrodes 4 and 5 is shown. The surface electrode portion of the movable electrode 3 has a length of 1 and 26
mm, width 1, 76 mn+.

The analysis results are shown when the length of the cantilever 2 is 0.8 mm. As the displacement W of the movable electrode 3 increases, the nonlinearity of the capacitances CI and C2 and their difference value ΔC with respect to the displacement W increases.

FIG. 11 shows an analysis result of non-linear error with respect to acceleration of the acceleration sensor according to the present invention. With respect to the acceleration of ±IG, the displacement W of the cantilever 2 is ±0.6 μm.
The width of cantilever 2 is 0.2 mm.

The analysis results when the thickness was etched to 0.02 mm are shown. It has a nonlinear error of approximately ±3% for the measurement range of 0 to ±IG. Therefore, linearity was improved by using a nonlinear error compensation circuit as described below. In addition, the capacitance CI between the movable electrode 3 and the fixed electrodes 4 and 5
, C2 changes ΔCt/C1 and ΔC2/Cz are approximately 0.2
0, the value is about 10 times the sensitivity ΔR/R (R is the resistance value of the diffused resistor) of the strain gauge type acceleration sensor, and high sensitivity can be achieved. Moreover, when an acceleration of ±IG is applied, the maximum stress generated at the fixed end 36 of the cantilever 2 is as small as about IMPa, which can be suppressed to about several tenths of the stress in the case of a strain gauge type acceleration sensor. , it is possible to improve impact resistance while achieving high sensitivity. Therefore, even if the cantilever 2 is dropped from a height of approximately 1 m onto concrete (at this time, an impact acceleration of approximately 100 G is instantaneously applied to the movable electrode 3), the cantilever 2 will not be damaged and is highly reliable. showed that. Also, cantilever 2
The thickness of the sensor can be increased to 20 μm (approximately 10 μm in the case of a strain gauge type acceleration sensor), so the natural frequency of the detection section consisting of the cantilever 2 and the movable electrode 3 is approximately 1.5 kHz. Therefore, the natural frequency is sufficiently large compared to the frequency component of 0 to 10 Hz that should be detected by the acceleration sensor for vehicle body control, and the displacement part is filled with liquid such as silicone oil.
In particular, even without damping, it is possible to achieve both impact resistance and high accuracy during detection. This simplifies the partial assembly structure and assembly work of the detection unit.

Suitable for lowering prices.

A first embodiment of the acceleration sensor according to the present invention is different from the above.
Shown in Figure 2.

The basic structure of the detection section is the same as the embodiment shown in FIG. 1 except for the following points.

That is, the movable electrode 3 is supported by a pair of silicon beams 40, 41 placed on the left and right sides of the figure.

Compared to the cantilever structure shown in FIG. 1, the one shown in FIG. 12 has a double-end structure, which reduces the sensitivity to about 1/1o, but can greatly influence G. In other respects, the same effects as in the previous embodiment can be obtained.

FIG. 13 shows the configuration of the detection circuit of the acceleration sensor of this embodiment. The detection unit 53 includes a movable electrode 3 and fixed electrodes 4 and 5 formed at the tip of the cantilever 2 supported by the fixed end 36. Movable electrode 3 that changes in response to acceleration
The capacitance between the fixed electrodes 4 and 5 was measured by a capacitance detection circuit 50, and after amplification by an amplifier circuit 51, a nonlinear compensation circuit 52 improved the linearity of the output voltage Vo with respect to acceleration.

An example of the output characteristics of the prototype sensor is shown in FIG.

As shown in the figure, it was confirmed that acceleration of 0 to ±IG can be detected with high accuracy, and that it is suitable as an acceleration sensor for vehicle body control.

〔Effect of the invention〕

The capacitive acceleration sensor according to the present invention has a simple structure. Since the structure is simple, the price is low and it can be configured to be small and lightweight. Highly reliable. Moreover, it has excellent impact resistance and high sensitivity.

[Brief explanation of the drawing]

Fig. 1 is a schematic diagram showing the basic structure of the detection section of an acceleration sensor according to the present invention, Fig. 2 is a basic structure diagram of a capacitive acceleration sensor, Fig. 3 is a developed view of the detection section, and Fig. 4 is a silicon plate of the detection section. 5 is a sectional view thereof, FIG. 6 is an explanatory diagram of the manufacturing process of the detection section, FIG. 7 is a sectional view showing a partially assembled structure of the detection section, FIG. 8 is a plan view thereof, and FIG. 9 The figure is an explanatory diagram of the relationship between the displacement of the movable electrode and the capacitance change, Figure 10 is a diagram showing the analysis results, Figure 11 is a diagram showing the analysis results of non-linear error, and Figure 12 is the acceleration sensor detection according to the present invention. FIG. 13 is an explanatory diagram of the configuration of the detection circuit, and FIG. 14 is a chart showing the output characteristics of a prototype sensor. 1... Silicon plate, 2... Cantilever, 3...
Movable electrode, 4, 5... Fixed electrode, 6, 7... Glass plate, 40° 41... Beam.

Claims (1)

[Scope of Claims] 1. (a) A movable electrode formed by etching a silicon plate from both sides and a cantilever formed by etching a silicon plate from both sides are integrally connected, and the movable electrode is formed by etching a silicon plate from both sides. a movable electrode is elastically supported; (b) the silicon plate on which the movable electrode and the cantilever are formed is sandwiched between a pair of glass plates; and (c) each of the pair of glass plates is connected to the movable electric plate. (d) A gap is created between the fixed electrode fixed to the glass plate and the movable electrode elastically supported by the cantilever. A capacitive acceleration sensor characterized by the following: 2. Claim 1, characterized in that the center of gravity of the movable electrode is located on the center line of the cantilever.
The capacitive acceleration sensor described in section. 3. (a) A movable electrode formed by etching a silicon plate from both sides, and a double-sided beam formed by etching the silicon plate from both sides and integrally connected to the movable electrode, (b) A silicon plate forming the movable electrode and the double-supported beam is sandwiched between a pair of glass plates, and (c) A portion where each of the pair of glass plates faces the movable electric plate. A plate-shaped fixed electrode is fixed to the glass plate, and (d) a gap is formed between the fixed electrode fixed to the glass plate and the movable electrode elastically supported by the cantilever. Capacitive acceleration sensor.