JP3189259B2 - Vibration transducer and method of manufacturing the same - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vibrating transducer for detecting a physical quantity such as an applied strain by detecting a frequency that changes in response to a strain or the like. The present invention relates to a vibrating transducer improved so as to be able to measure at a high speed.

[0002]

2. Description of the Related Art FIG. 30 is a perspective view of a structure using a conventional vibration transducer as a pressure sensor, and FIG.
FIG. 32 is a sectional view showing an AA 'section in FIG. 31, and FIG. 33 is an electrical equivalent circuit showing the structure shown in FIG. 31. FIG. These conventional vibrating transducers are disclosed, for example, in U.S. Pat. No. 4,926,143.

As shown in FIG. 30, reference numeral 10 denotes a p-type silicon single crystal substrate having an upper surface having a crystal plane (100), for example, having an impurity concentration of 10 ¹⁵ atoms / cm ³ or less and a conduction type of p-type. A diaphragm 11 is provided on one surface of the substrate 10.
Are dug up from the back surface by etching to form a thin wall.

The thick portion 12 around the diaphragm 11
Is joined to the base 14 with the pressure guide hole 13 in the center, the pressure P _M to be measured in the impulse line 15 is introduced. An n ⁺ diffusion layer (omitted in the drawing) having an impurity concentration of about 10 ¹⁷ is partially formed on the surface of the diaphragm 11 which is not etched as indicated by the symbol A, and a vibrator 16 is formed on a part of the n ⁺ diffusion layer. Are formed in the crystal axis <001> direction (FIG. 31). The vibrator 16 processes, for example, an n ⁺ diffusion layer and a p layer formed in the diaphragm 11 using techniques of photolithography and under-etching.

Reference numeral 17 denotes a vibrator 16 at the upper center of the vibrator 16.
And a magnet provided in a non-contact state orthogonal to
Is S _i O ₂ film as an insulating film (see FIG. 32). 19
a, 19b is a metal electrode such as aluminum, the contact holes 2 and one end of the metal electrode 19a is that through the S _i O ₂ layer provided on the n ⁺ layer that extends from the transducer 16
0a, and the other end is connected to one end of a comparison resistor _R0 substantially equal to the resistance value of the vibrator 16 and the input end of the amplifier 21 via a lead wire.

An output signal is taken out from an output terminal of the amplifier 21 and connected to one end of a primary coil L ₁ of a transformer 22. The other end of the primary coil L ₁ is connected to a common line. The other end of the comparison resistor R ₀ is connected to the secondary end of the coil L ₂ of the transformer 22 the midpoint of which is connected to the common line, the other end of the secondary coil L ₂ to the other end of the vibrator 16 A metal electrode 19b similarly formed,
It is joined to the n ⁺ diffusion layer via the contact hole 20b.

In the above structure, the p-type layer (substrate 10)
When a reverse bias voltage is applied between the and the n ⁺ diffusion layer (oscillator 16) to insulate the oscillator 16 and an alternating current i flows through the oscillator 16, the impedance of the oscillator 16 in the resonance state of the oscillator 16 is represented by R. An equivalent circuit as shown in FIG.

In this way, a bridge is formed by the secondary coil L ₂ having the middle point C ₀ connected to the common line, the comparison resistor R ₀ , and the impedance R. When the output is detected and the output thereof is positively fed back to the primary coil L ₁ via the feedback line 23, the system causes self-sustained pulsation at the natural frequency of the oscillator 16.

Here, when the pressure P is applied to the diaphragm 11 through the pressure introducing hole 13, the tension acting on the vibrator 16 changes, and the resonance frequency changes. For this reason, the natural frequency of the vibrator 16 changes, and the magnitude of the pressure P can be known from the change in the natural frequency.

[0010]

However, the above-mentioned vibration type sensor has a disadvantage that the magnet 17 having a large size is required, and the fixing structure is complicated.
Therefore, as a configuration that does not use this magnet, for example, an electrostatic force is used as an excitation means, a piezo resistor is used as a detection element, and these are incorporated in an oscillation loop to cause self-excited oscillation. Although it is conceivable to detect distortion by the change of the temperature, according to this configuration, the current flows to the vibrator, so the temperature rises due to heat generation and an error factor is created, and it is necessary to build a piezo gauge on the vibrator. This has the disadvantage of complicating the manufacturing process. In addition, a configuration in which a piezoelectric drive is used to detect with a pressure-sensitive element is also conceivable. However, according to this configuration, the piezoelectric material is limited. In particular, since silicon has no piezoelectricity, PZT or the like must be formed on silicon. However, there is a problem that fabrication is difficult.

[0011]

According to the present invention, a first main structure for solving the above problems is a semiconductor substrate having a first conductivity type, and a semiconductor substrate formed on a surface of the substrate and having a conductive type formed thereon. A channel sandwiched between a drain and a source having a second conductivity type opposite to the type, and a drain and a source which are fixed at both ends while maintaining a gap so as to be displaceable from the surfaces of the drain and the source. And a plate-shaped vibrating gate that is disposed so as to cover at least one of them and the previous channel and that is displaced by an electrostatic force generated between the drain and the self-excited oscillation. The second main configuration includes: A semiconductor substrate having a first conductivity type, a channel formed on a surface of the substrate and sandwiched between a drain and a source having a second conductivity type opposite to the previous conductivity type, formed on a surface of the previous substrate; Done The oxide film, the insulating film that covers the gate oxide film and has high resistance to hydrofluoric acid, and the insulating film that is mainly made of polysilicon and faces the previous channel while maintaining a gap. And a conductive vibrating gate having both ends fixed to the previous substrate, and a shell covering the vibrating gate and having the inside kept in vacuum.
The main structure of the vibrating method is to form a shell having a vibrating gate fixed at both ends with a predetermined spacing on the upper surface of a silicon substrate having a thin bottom and forming a shell covering the periphery with a spacing therebetween. In the method for manufacturing a transducer, (A)
A gate oxide film is formed on the previous substrate, and after that, a drain and a source are formed by ion-implanting impurities into the previous substrate only at a portion corresponding to the channel, and then the gate oxide is formed. The film is covered with an insulating film having high resistance to hydrofluoric acid, and (B) a first sacrificial oxide film is formed on the insulating film, and then polysilicon serving as a vibration gate is deposited thereon. This is doped with an impurity to impart conductivity to a predetermined shape, on which a second sacrificial oxide film is formed to form a first shape.
Etching the second sacrificial oxide film with hydrofluoric acid to form a gap corresponding portion having a predetermined shape, and (C) forming an oxide film corresponding to the gap covering the gap corresponding portion and the insulating film, Polysilicon as a shell is deposited thereon, and then the gap-corresponding portion and the oxide film corresponding to the gap are removed by etching with hydrofluoric acid to form a vibrating gate, a shell, and a void surrounded by these. And (D) thereafter, the former shell and the former insulating film are covered with polysilicon in a vacuum, and the above-mentioned gap is kept in a vacuum to manufacture the semiconductor device.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a configuration diagram showing the configuration of one embodiment of the present invention.

The silicon substrate 24 has, for example, a conduction type of n
The electrode 25 is fixed here, and the electrode 25
Are connected to a common potential point COM. On the upper surface of the silicon substrate 24, a p-type impurity is diffused and the source S
Is formed, and an electrode 26 for taking out the potential of the source S is formed here. The silicon substrate 24
Although not shown, a pressure P _M to be measured or the like is applied to the lower surface.

Also, a p-type impurity is diffused into the upper surface of the silicon substrate 24 at a predetermined distance W from the source S to form a drain D, and an electrode for taking out the potential of the drain D is formed here. 27 are formed.

[0015] Above the portion of the predetermined distance W of the silicon substrate 24, convex portions 28 and 29 are formed apart x _1, impurity functions as has been plate-shaped vibrator conductivity is imparted conductive diffusion Both ends of the vibration gate 30 (for convenience, the symbol G is used) are fixed to these convex portions 28 and 29. In other words, the vibration gate 30 and the silicon substrate 24 are arranged apart from each other by x except for both ends, and although not shown, a channel CNN is provided between the drain D and the source S on the silicon substrate 24 corresponding to the vibration gate 30.
Is formed.

Between the electrode 27 and the common potential point COM,
The resistor R1 and the DC power supply E1 are connected in series, and the potential of the drain D is kept at a negative potential with respect to the common potential point COM. Further, a direct current power supply E2 is connected to the oscillation gate 30 so as to have a negative potential with respect to the common potential point COM.

FIG. 2 is an explanatory diagram for explaining the operation of the embodiment shown in FIG. The configuration includes a cross section of the silicon substrate 24 as viewed from the longitudinal direction of the vibration gate 30. Since a negative potential is applied from the DC power supply E2 to the vibrating gate 30 functioning as a gate, electrons are transferred from the lower surface of the vibrator 30 to the inside of the silicon substrate 24 (in FIG. ), And conversely, the holes become attracted to the surface.

The channel CNN1 which is a thin P-type conductive layer is formed on the surface by the attracted holes (P-type), and connects the source S (P-type) and the drain D (P-type) with the P-type. Therefore, a current _id1 flows between the source S and the drain D.

The drain D generated by the current _id1
Is subjected to a phase shift by a drain resistance R _D and a capacitance C _D formed between the drain and the silicon substrate 24, and the potential change resulting from the phase shift causes the oscillation gate 30 and the drain D to be connected to each other. changing the electrostatic attraction between changing the distance x _1.

The change in the interval x ₁ causes the channel CNN to change.
1, thereby changing the current _id1 ;
This causes a change in the potential of the drain. Although oscillates by repeating this, the oscillation drain resistance R _D and the drain D and the capacitance C _D and the product of the oscillation angular velocity ω of the oscillation between the silicon substrate 24 (.omega.R _D C _D) is compared to the 1 Continue by choosing to be extremely large.

When the pressure PM is applied to the silicon substrate 24 as shown in the figure while the self-sustained pulsation is maintained as described above, the pressure P _M is applied to the silicon substrate 24 via the projections 28 and 29 for fixing the vibration gate 30. applied to the pressure P _M distortion vibrating gate 30 by which the corresponding natural frequency changes. Therefore, by taking out a change in the natural frequency, it is possible to detect the value of pressure P _M.

Next, the above points will be described in detail using mathematical expressions. FIG. 3 illustrates the device configuration illustrated in FIG. 1 as an equivalent circuit. FIG. 4 shows the arrangement of the signs of the respective parts of the device configuration shown in FIG. 1 necessary for performing the following calculation. FIG. 5 shows the transfer function obtained by decomposing the respective parts of the equivalent circuit shown in FIG. FIG.

In FIG. 3, R _D is the drain resistance of the drain D, C _D is the capacitance between the drain D and the silicon substrate 24, and the vibration gate 30 is arranged facing the substrate 24 in an insulated manner. The vibration gate 30 vibrates while approaching or leaving the substrate 24 at the oscillation angular velocity ω of oscillation.
A DC power supply E2 is applied to the vibration gate 30. Incidentally, R ₁ for simplicity will be described as zero.

For such an arrangement, as shown in FIG. 4, the width of the vibration gate 30 is b, the thickness is h, and
It is assumed that apart x ₁ with respect to 4. The vibration gate 30 is disposed so as to cover only the drain D formed on the substrate 24 and the width a. By covering in this manner, the electrostatic attraction between the drain D and the vibration gate 30 can be effectively used for oscillation.

The distance between the source S and the drain D on the substrate 24, that is, the channel width is set to the length w.
FIG. 5 shows a process in which oscillation is caused in the configuration shown in FIGS. Using this, analysis is performed sequentially.

Here, K ₁ is a transfer function which is a ratio of a change ΔF in electrostatic attraction between the drain D and the vibrating gate 30 to a change Δe _{d1 in the} drain voltage generated in the drain D, and K ₂ is a static function. electrostatic transfer function which is the ratio of the displacement [Delta] x ₁ vibration gate 30 for sucking force [Delta] F, K ₃ is a transfer function which is the ratio of the drain current .DELTA.i _d1 of the drain D to the displacement [Delta] x ₁ vibration gate 30, K ₄ denotes a drain Transfer functions each representing a ratio of a change Δd _d2 of the drain voltage to a current Δid ₁ are shown.

[0027] Here, the conditions which such systems can cause oscillations, and _{G L = K 1 K 2 K} 3 K 4 (1), if the ∠G _L between the phase angle of G _L, G _L> 1 _{(2) ∠G L = 0 (} 3) becomes the oscillation conditions.

First, the electrostatic attraction force F generated per unit length of the vibrating gate 30 is as follows: F = ε ₀ a (V / x) ² (4) Here, V = E ₁ −E ₂ and ε ₀ is a dielectric constant. The electrostatic attraction force F that changes when V changes by Δe _d1
The [Delta] F of the change, when expressing the δ as partial derivative symbol, ΔF = (δF / δV) Δe d1 = 2ε 0 aV · Δe d1 / x 1 2 (5) ∴ K 1 = ΔF / Δe d1 = 2ε 0 aV / x ₁ ² (6)

Second, the displacement Δx of the beam fixed at both ends due to the distributed load ΔF is as follows: Δx ₁ = ΔF · L ⁴ / 384EI (1 + jγω−τ ₀ ² ω ² ) (7) Here, L is the length of the vibration gate 30, I is the moment of inertia of the vibration gate 30, γ is the damping rate, and τ ₀ is the natural period. Note that E is E = E ₀ (1 + Sη) S = 0.24 (L / h) ² , E ₀ is the Young's modulus of the vibrating gate 30, and η is the strain. Therefore, K ₂ is given by K ₂ = Δx ₁ / ΔF = L ⁴ / 384E ₀ I (1 + jγω−τ ₀ ² ω ² ) (8)

[0030] Thirdly, the drain current i _d1, in the pinch-off region, i _d1 = become _{(1/2) μC 0 L (E} 2 -V T) 2 / w (9). Here, mu is the electron mobility, C ₀ is the capacitance per unit area between the channels, the V _T is the threshold voltage. C ₀ is C ₀ = ε ₀ / x ₁ (10)

The change in the drain current i _d1 when vibration gate 30 approaches the silicon substrate 24 side by a [Delta] x ₁ .DELTA.i _d1
Is as follows: Δid ₁ = (δid ₁ / δx ₁ ) (− Δx ₁ ) = (− ／) με ₀ L [(E ₂ −V _T ) / x ₁ ] ² (−Δx ₁ ) / w Therefore, _{K 3 = Δi d / Δx 1} = (1/2) με 0 L [(E 2 -V T) / x 1] 2 / w (11).

Fourth, the drain current i_d1Change Δi_d1To
Drain voltage e_d1Change Δe _d1Ask for. This place
In this case, description will be given with reference to FIG. Constant in pinch-off area
Since this can be regarded as the current source CC1, FIG.
As shown in FIG. Fig. 6
An equivalent circuit shown in FIG. The equivalent times of FIG.
From the road, K_Four= Δe_d2/ Δi_d1 = -R / (1 + jωC_DR_D)  (12)

[0033] _{Thus, G L = K 1 K 2} K 3 K 4 = 2ε 0 a (E 1 -E 2) / x 1 2 · L 4 / 384EI (1 + jγω-τ 0 2 ω 2) · (1/2 ) Με ₀ L [(E ₂ −V _T ) / x ₁ ] ² / w · [−R / (1 + jωC _D R _D )]

[0034] In this case, and _{ω = ω 0 = 1 / τ} 0, ω 0 C D R
Assuming that _D is extremely large as compared with 1, _GL = −με ₀ ² L ⁵ a (E ₁ −E ₂ ) (E ₂ −V _T ) ² · (1 / jγω ₀ ) (1 / jω ₀ C _D ) become ₍₁ / 384EIx 1 ^4).

Here, assuming that the sharpness of resonance is Q, Q = 1
As demonstrated by _{/ γω 0, G L = -με} 0 2 L 5 a (E 1 -E 2) (E 2 -V T) 2 · (Q / j) (1 / jω 0 C D) · (1 / 384EIx ₁ ⁴⁾ obtaining (12).

Further, when I = bh / 12 E = E ₀ [1 + 0.24 (L / h) ² η] is substituted into the equation (12), G _L = Qμε ₀ ² L ⁵ a (E ₁ −E ₂ ) and a ^{_{(E 2 -V T) 2 /}} 32E 0 [1 + 0.24 (L / h) 2 η] bh 3 x 1 4 wω 0 C D (13).

[0037] ^{_{Here, E 0 = 1.6X10 11 N /}} m 2 ε 0 = 8.85X10 ^{over 12 / m μ = 0.1m 2 /} Vs ( impurity concentration: 1X10 ¹⁶ / c
The use of m ^3), G _L = 2.43X10 ^{over _{37 QL 5 a (E 1 -E}} 2) (E 2 -V
_{^{T) 2 / E 0 [1}} + 0.24 (L / h) 2 η] bh 3 x 1 4 w
the ω ₀ C _D.

[0038] As an example, for example, L = 150 × 10 ^-6
m, a = w = 5X10 ^over 6 m, h = 1X10 ^over 6 m, b = 1
0X10 ^over 6 m, x ₁ = 0.5X10 ^over 6 m, η = 0, f 0 =
500X10 ³ Hz, C _D = 10X10 ^{over 12} F, E ₁ -E ₂
= 1V, if _{E 2 -V T = 1V, G} L = 61X10 ^{- 4}
Q. Therefore, if Q = 10 ⁴ , then G _L = 61.

That is, if a vibrating transducer is constructed with a length of 150 μm, a width of 10 μm, and a gap of 0.5 μm, G _L = 61, and the phase of G _L is also the same, which sufficiently satisfies the oscillation condition. In equation (13), G _L is (E ₂
−V _T ) ² , so by adjusting E ₂ ,
The amplitude of _GL can be controlled.

The products K ₁ , K ₂ and K ₃ correspond to the vibration gate 3
The relationship between the change in drain voltage Δe _d1, which is the potential difference between 0 and the drain D, and the change in drain current Δi _d1 is shown. In a resonance state, ω ₀ C _D R _D at ω = ω ₀ = 1 / τ _0. Is very large compared to 1, so that considering Q = 1 / γω ₀ , there is no phase difference between K ₁ and K ₃ (see equations (6) and (11)), but 90 ° in K ₂ It can be seen that the following phase difference (Equation (8)) occurs. Accordingly, it can be seen that the drain current has a phase difference of 90 ° with respect to the potential difference between the oscillation gate 30 and the drain D.

FIG. 7 shows a specific circuit configuration. Reference numeral 31 denotes a vibration gauge disclosed in FIG. Vibrating gate 30
The drain current is 9 with respect to the potential difference between
0 While ° phase advances, the resistance R _d and with the capacitor C _d is the phase 90 ° delayed are in phase is satisfied oscillation conditions, input the voltage through the capacitor C ₂ is the resistor R ₂ It is output to the connected amplifier 32, and outputs the voltage V ₀ at its output.

This voltage V ₀ is output to the automatic gain control circuit 33 so as to correspond to this value.
_The DC voltage is applied to the oscillation gate 30 so as to be compared with the reference voltage _Vref and correspond to the reference voltage _Vref .

The automatic gain control circuit 33 includes a rectifier circuit 34
, The voltage V ₀ is changed to a corresponding DC voltage, and the reference voltage V _ref is applied to one end of the input terminal. The DC voltage is applied to the other end of the input of the comparator 35, and is applied to the oscillation gate 30 here. Automatically adjusts the magnitude of DC voltage. Since the loop constituted by these conditions satisfies the oscillation condition, self-excited oscillation occurs, and the oscillation amplitude can be controlled by adjusting the value of the reference voltage _Vref .

FIG. 8 is a perspective view showing the structure of another embodiment of the present invention, and FIG. 9 is a cross-sectional view near the center. However,
The shell part and the diaphragm part covering the vibration gate are omitted. The embodiment shown in FIG. 8 has the following improvements over the embodiment shown in FIG.

First, a bias voltage is applied to the oscillation gate (FIG. 1).
To simplify the circuit configuration as configuration without applying a corresponding) to E ₂ of the influence of the bias voltage variations is removed to improve the frequency stability with. Second, by eliminating the bias voltage, the vibrating gate is prevented from adhering to the substrate due to the electrostatic attraction force, thereby preventing inoperability. Third,
The silicon substrate facing the oscillation gate is protected with a protective film having high resistance to hydrofluoric acid to ensure stability and facilitate fine processing. Fourth, it is possible to make the vibration gate conductive by doping impurities while using polysilicon as the vibration gate and enabling fine processing.

Hereinafter, a description will be given with reference to FIGS. The silicon substrate 35 has, for example, an n-type conduction type, and a p-type impurity is diffused on the upper surface of the silicon substrate 35 to form a source S. Here, aluminum is used to extract the potential of the source S. electrode 36 is formed via the wiring part W _S shown by a dotted line. Further, although not shown in the lower surface of the silicon substrate 35 such as a pressure P _M to be measured here it is formed on the diaphragm recessed is applied.

Also, at a predetermined distance from the source S, a p-type impurity is diffused into the upper surface of the silicon substrate 35 to form a drain D, where the drain D is formed.
Aluminum electrodes 37 for taking out the potential of the are formed through the wiring portion W _D shown by a dotted line.

[0048] Above the silicon substrate 35, the fixed end 38 and 39 are spaced apart by a gap x _2, both ends of the polysilicon plate-like vibration gate 40 which impurities are diffused conductivity is imparted, These fixed ends 38 and 39 are integrally fixed. The length of the beam of the vibration gate 40 is L.
The vibrating gate 40 is connected to the electrode 4 made of aluminum.
It is connected via a wiring portion W _G indicated by 1 and the dotted line.

That is, the vibration gate 40 and the silicon substrate 3
5 and are spaced apart by a gap x ₂ except the ends, channel CNN2 is formed between the drain D and the source S of the silicon substrate 35 corresponding to the vibration gate 40.

On the drain D, the channel CNN2 and the source S formed on the upper surface of the silicon substrate 35, a protective film 4 having high corrosion resistance to hydrofluoric acid (HF) is formed.
2, for example, _{S i3 N 4, S i C} x N y, S i C, and the like AL ₂ O _3, is a two-layer structure film 44 made of an oxide film 43 is formed.

A gap is provided between the two-layer structure film 44 and the vibration gate 40 so that the vibration gate 40 can vibrate up and down with the fixed ends 38 and 39 as nodes. Thus, the vibration gauge 45 is configured.

FIG. 10 is an explanatory diagram for explaining the open loop characteristics of the vibration gauge 45 shown in FIG. v _io is an AC voltage applied to obtain characteristics, and V _g is a DC bias voltage applied to the oscillation gate 40. The characteristics of the vibration gauge 45 shown in FIGS. 11 to 13 will be described with reference to FIG.

FIG. 11 is a characteristic diagram showing characteristics of DC bias voltage versus AC output. The vertical axis indicates the degree of amplification using the effective value of the AC voltage _vOO . In this case indicates whether the AC voltage v _OO output to the drain D while applying a constant AC voltage v _io how changes by the DC bias voltage V _g.

FIG. 11 shows that the largest output voltage is obtained when no bias voltage _Vg is applied to the oscillation gate 40, that is, when _Vg = 0. This indicates that the efficiency is highest when E ₂ = 0 in FIG.

Next, there is a region where there is no signal near V _g = −2 volts. It exists some potential difference (approximately -2 volts) between the drain D and the source S, the suction force to offset this by electrostatic and the electrostatic force by the V _g is against vibration gate 40 It is because it disappears. Also,
When V _g exceeds -7 volts, the vibration gate 40 adheres to the silicon substrate side by electrostatic force and no signal is output.

FIG. 12 shows the relationship between the DC bias voltage and the frequency output.
It is a characteristic view showing a characteristic. The vertical axis is the intersection that appears in the drain D.
Current voltage v_OOResonance frequency f_OOIndicated by And
In this case, the constant AC voltage v_ioDray while applying
AC voltage v output to _OOResonance frequency f_OOIs DC
Bias voltage V_gShow how it changes
ing.

It is shown that the change in f _OO with respect to V _g is smallest when no bias voltage V _g is applied to the oscillation gate 40, that is, when V _g = 0. In other words, the state of V _g = 0 is a state in which the dependence of the resonance frequency f _{OO on} the bias voltage V _g is small, and the state is stable.

From the characteristic curves shown in FIGS. 11 and 12, it can be seen that the configuration shown in FIG. 8 in which no bias DC voltage is applied to the oscillation gate 40 has good efficiency and frequency stability.

FIG. 13 shows the drain-source voltage V
FIG. 9 is a characteristic diagram showing characteristics of _ds versus AC output. The vertical axis indicates the amplification using the effective value of the AC voltage _{vOO at} the drain.

Then, in this case, a constant AC voltage v _io is applied, and the AC voltage v _OO output to the drain D in a state where V _g = 0 is changed by the change of the DC voltage E ₁ , that is,
It shows how the voltage varies with the source-to-source voltage V _ds . The length L of the beam of the vibration gate 40 is changed to L ₁ and L ₂ (L
₁ <L ₂ ). This means that since the source voltage is zero, the drain voltage of the AC voltage _vOO is dependent on the DC voltage.

FIG. 14 shows a vibrating transducer in which a self-excited oscillation circuit is formed using the vibration gauge 45 having the above-described characteristics. The drain D has one end connected to a bias constant current source CC2 connected to the DC power supply -E1.

The source S is connected to the common potential point COM. Vibrating gate 40 which functions as a gate G opposed to these drain D and the source S, the other end of the resistor R ₃ is connected to one end of the resistor R ₃ is connected to the common potential point COM. Furthermore, the moving electrode 40 is connected to the output terminal of the variable amplifier 46 via a capacitor C _3.

[0063] The non-inverting input terminal (+) to the inverting input of the operational amplifier 47 to which a bias voltage E _R1 is applied (-) is its output end is connected to the drain D of the vibration gauge 45 by a resistor R ₄ It is connected.

The output terminal of the operational amplifier 47 is connected to the input terminal of the variable amplifier circuit 46 via the 90 ° phase shift circuit 48 and to the input terminal of the rectifier / smoothing circuit 49.
The output terminal of the rectifier / smoothing circuit 49 is connected to the inverting input terminal (-) of the error amplifier 50 to which the set voltage E _R2 is applied to the non-inverting input terminal (+). This error amplifier 50
Controlling the gain of the variable amplifier 46 by the error voltage appearing at the output terminal V _e.

Next, the operation of the vibrating transducer 51 configured as described above will be described. The AC drain current _id2 flowing to the drain D at the resonance point flows with a 90 ° phase advance with respect to the voltage between the drain D and the oscillation gate 40. The operational amplifier 47 converts the drain current _id2 into a voltage signal _Vd2 and outputs it to the 90-degree phase shift circuit 48.
90 ° phase shift circuit 48 is applied to the variable amplifier 46 by a voltage signal V _d2 Okurashi a 90 ° phase, it is applied to the vibrating gate 40 via a capacitor C ₃ from the output end.

As described above, a self-excited oscillation circuit is formed by the loop connecting the vibration gauge 45, the operational amplifier 47, the 90 ° phase circuit 48, the variable amplifier circuit 46, and the vibration gauge 45, and the oscillation is continued. Oscillation frequency is detected at the terminal T _0.
The constant current source CC2 sets a DC bias current so that the operational amplifier 47 does not saturate.

In this case, the AC voltage signal V _d2 is rectified / smoothed by the rectification / smoothing circuit 49 to be a DC voltage.
The error amplifier 50 compares the DC voltage with the set voltage E _R2 and outputs an error voltage V _e that becomes equal to the set voltage E _R2 to the variable amplifier circuit 46, controls the amplification degree, and the oscillation amplitude as a whole is reduced. The amplitude is controlled so as to be constant.

FIG. 15 shows a vibrating transducer in which a self-excited oscillation circuit is formed using an inductor. FIG.
(A) is a vibration gauge 52 configured as a P-channel type,
FIG. 15B shows an example of an N-channel vibration gauge 45.

In FIG. 15A, a vibration gate G of the vibration gauge 52 is connected to a source S via a resistor R ₃ , a drain D is connected to one end of an inductor L _n1 , and the inductor L
_n1 and the other end of which is connected to the negative electrode of the DC power source E _3. Further, the positive electrode of the DC power source E ₃ is connected to the common potential point COM and the source S. Here, the resistance R ₃ is for immobilizing a potential vibration gate G.

[0070] Further, in FIG. 15 (b), the vibration gate G of the vibration gauge 45, the common potential point C through a resistor R ₃
The OM and the source S, the drain D to one end of the inductor L _n2, the other end of the inductor L _n2 are respectively connected to the positive electrode of the DC power source E _3. And the DC power supply E ₃
Are connected to a common potential point COM.

In the above configuration, no DC bias voltage is applied to the oscillation gate G, but if necessary,
In place of the resistor R _3, a negative potential with respect to the source S to the vibration gate G in the circuit of FIG. 15 (a), a positive potential with respect to the source S in the circuit in the vibration gate G shown in FIG. 15 (b) The current amplification factor may be controlled by changing the mutual conductance of the vibration gauges 52 and 45 by applying the voltage.

Next, the operation of the vibration transducer shown in FIG. 15A will be described with reference to the waveform chart shown in FIG. With respect to the AC voltage V _gd (FIG. 16A) applied between the drain D and the oscillation gate G, the AC drain current _id3 flowing through the drain D is 90 ° as shown in FIG. 16B. The phase advances.

Accordingly, the drain D is
The drain voltage V _d3 of the alternating current generated at this time is as follows: V _d3 = jωL _n1 · _{id 3} (14), and the phase is advanced by 90 ° with respect to _{id 3} . Where ω = 2
πf, f are resonance frequencies.

Therefore, when the drain voltage V _d3 is viewed between the oscillating gate G and the drain D, the phase is inverted by 180 ° to become V _d3 ′ (FIG. 16 (d)), The voltage V _gd (FIG. 16A) has the same phase as the voltage V _gd and satisfies the oscillation condition.

Therefore, by detecting the frequency of the drain voltage V _d3 , it is possible to detect distortion caused by pressure or the like. The above points are the same in the case of FIG. 15B, and self-excited oscillation occurs in the same manner.

FIG. 17 shows another example of a vibrating transducer constituted by using a narrow-band operational amplifier. The vibration gauge 52 is constituted by a P channel. The oscillation gate G is connected to the common potential point COM together with the source S via the resistor R3.

The inverting input terminal (-) of the operational amplifier 53 is
Connected to the rain D and a resistor R _FiveIts output via
Connected to the end. The non-inverting input terminal (+)
Setting power supply E_FourNegative potential with respect to the common potential point COM
Is set to

In this case, the operational amplifier 53 uses a narrow-band operational amplifier, and the potential of the inverting input terminal (-) has a frequency characteristic as shown in FIG. The horizontal axis indicates the frequency f, A indicates the phase characteristic, and B indicates the gain characteristic. Inverting input terminal of the operational amplifier 53 (-) characteristic of the range of frequencies f _A and f _B, the phase is constant offset 90 °, the gain varies linearly. Resonance frequency of the vibration gauge uses an operational amplifier 53 such that between the frequencies f _A and f _B.

[0079] The operational amplifier 53 is an inverting input terminal (-) and the potential of the non-inverting input terminal (+) is operated to be the same, the inverting input terminal (-) voltage is kept the same as setting the power E ₄ The DC power is supplied to the vibration gauge 52.

The characteristic of the inverting input terminal (-) of the operational amplifier 53 utilizes the characteristic that the phase is shifted by 90 degrees as shown in FIG. The voltage at the inverting input terminal (-) has a 90 degree phase advance with respect to the drain current _id4 . Therefore, the same operation as in FIG. 15 is performed. However, in this case, since the drain voltage _Vd4 is extracted from the output terminal of the operational amplifier 53, there is a merit that the output can be performed with low impedance.

FIG. 19 shows another example of a vibrating transducer using a capacitor to form a self-excited oscillation circuit. In this case, the vibration gauge 52 is configured by a P-channel, and the phase shift element is configured to use a capacitor instead of an inductor. Vibration gate G of vibration gauge 52
, A bias voltage E ₆ is applied as a negative potential with respect to the common potential point COM, and its source S is also
It is connected to the.

The drain D is connected to the negative electrode of the DC power supply E ₅ via a constant current source CC 3 for supplying the DC drain current _ID, and the positive electrode of the DC power supply E ₅ is connected to the common potential point COM. A phase shift capacitor _CP is connected to both ends of the constant current source CC3.

In this case, since the voltage value is set so as to satisfy the relationship of E ₆ > E ₅ , V _gd ′ in FIG.
As shown in FIG. 7, the phase does not reverse with respect to V _gd . Thus, by using the capacitor C _P, AC drain current i _d5
In contrast, the phase of the drain voltage V _d5 is delayed by 90 ° to satisfy the oscillation condition as the same phase.

FIG. 20 shows another vibrating transducer in which a self-excited oscillation circuit is formed using an inductor. FIG.
Is an example configured as a P-channel type. Vibration gauge 5
2 is a resistor R ₆ for detecting a drain current _{id 6.}
To the common potential point COM, the source S is a DC power source E ₆
It is connected to the.

The voltage across the resistor R ₆ is applied to the base of the transistor Q ₁ via a coupling capacitor C ₄ .
A divided voltage for bias obtained by dividing the DC power supply E ₆ by the resistors R ₇ and R ₈ is applied.

[0086] Further, between the DC power supply E ₆ and the common potential point COM, an inductor L _n3, the collector and emitter of the transistor Q _1, resistor R ₉ to determine the idling current of the transistor Q ₁ is connected in series ing.

The collector of the transistor Q ₁ is connected to the vibration gauge 52 via a coupling capacitor C _5.
The oscillation gate G has a resistor R for bias.
Each is connected to the source S via ₁₀ . These vibrations gauge 52, the resistance R _6, R _10, 90 ° slow circuit 54 at a portion surrounded by a dotted line, except for the capacitor C ₅ is constructed.

Next, the operation of the vibrating transducer configured as described above will be described with reference to the waveform diagram shown in FIG. At the resonance point, as shown in FIG. 21, the phase of the drain current _id6 advances by 90 ° with respect to the voltage V _GD between the oscillation gate G and the drain D. The drain current i _d6 is converted to the drain voltage by the resistor R _6, it is amplified by the transistor Q ₁ through a high-pass filter constituted by the parallel resistance of the capacitor C ₄ and the resistor R ₇ and R _8, the collector current i
It flows in the inductor L _n3 as _d6 '. In this case, the phases of the collector current _id6 'and the drain current _id6 are in phase.

Therefore, the voltage between both ends of the inductor L _n3 , that is, the voltage V _SG between the vibrating gate and the source of the vibrating gauge 52 advances by 90 ° with respect to the drain current _{id 6} , so that the voltage V _GD between the vibrating gate and the drain becomes Drain current i
_The phase is delayed by 90 ° with respect to _d6 , the oscillation condition is satisfied, and the self-excited oscillation is continued.

FIG. 22 shows another vibrating transducer in which a self-excited oscillation circuit is formed using a capacitor. FIG.
Is an example configured as a P-channel type. Vibration gauge 5
2 is a resistor R ₁₁ for detecting a drain current _id7.
, The source S is connected to the DC power source E ₇
Connected to each other. Field effect transistor Q ₂
The voltage across the resistor R ₁₁ is applied to the gate G of the capacitor C _6.
It is applied through a high-pass filter composed of a resistor R ₁₂ and.

[0091] The drain D and source S of the field effect transistor Q ₂ is the power supply through respective resistors R ₁₃ and R _{1 4} E ₇
And the common potential point COM. A capacitor C ₇ and a resistor R are provided between the drain D and the source S.
₁₅ and are connected in series, which like the connection point is connected to the gate G of the field effect transistor Q _3.

The drain D of the field effect transistor Q ₃ is connected to a power supply E ₇ , and the source S is connected to a common potential point C via a resistor R _16.
OM. Further, the source S is connected to the vibrating gate G of the vibration gauge 52 via a high-pass filter composed of a capacitor C ₅ and the resistor R _10.

The vibration gauge 52, the resistance R ₁₁ ,
90 in the portion surrounded by the dotted line except for R ₁₀ and capacitor C ₅
° A delay circuit 55 is configured. Specifically, the resistance R
_13, R _14, R _15, a field effect transistor Q ₂ and a capacitor C ₇ is 90 ° lagging phase, the field effect transistor Q ₃ are functions as Sosufuorowa. The operation is the same as that of the vibration transducer shown in FIG.

In the above description, the description has been given of the point that the vibration gauge and the electronic circuit as shown in FIG. 8 are combined to form the vibration transducer. Next, a vibration gauge 45 as a component of such a vibration type transducer is described.
Will be described with reference to manufacturing process diagrams shown in FIGS. 23 and 24. Although the manufacturing process shown in FIG. 24 is continuous with the manufacturing process shown in FIG. 23, it is divided into two for convenience of explanation.

In the structure shown in FIG. 8, the cross-sectional structure generated in the course of the manufacturing process is different between the portion of the vibration gate 40 and the portions of the fixed ends 38 and 39 for fixing the same at both ends. These are separated and shown separately. The figure on the left shows the cross-sectional structure of the central part of the vibration gate 40.
The drawing on the right shows the cross-sectional structure of the fixed end 38. In addition,
The fixed end 39 has the same structure as the fixed end 38 and will not be described.

Step 1 shows a gate oxide film forming step. A gate oxide film 61 is formed on an n-type silicon single crystal substrate 60 to a thickness of, for example, about 500 angstroms. In this step, the cross-sectional structures at the center and the fixed end of the vibration gate are formed to be the same. Thereafter, the process proceeds to step 2. Thereafter, each step proceeds according to the step number.

Step 2 shows an ion implantation process. Here, boron is ion-implanted into a predetermined region as a p-type impurity. Thus, corresponding to W _D of (W corresponds to _S in FIG. 8) and the drain part 63 (FIG. 8 source 62 of p-type at a predetermined interval W _N plan to form a channel CCN in the central portion of the vibrating gate ), And a p-type gate lead portion 64 (corresponding to _WG in FIG. 8) is formed at the fixed end.

Step 3 shows a channel forming step.
Here, the channel CCN is located at the center of the oscillation gate.
Are formed at predetermined intervals W _N of the channel portions 64 (CC
N2) is ion-implanted with boron at a shallow depth. Thereby, the resistance between the source and the drain can be controlled to a predetermined value. In this case, there is no change at the fixed end.

Step 4 shows a nitride film forming step. In this step, in order to protect the gate oxide film 61 against hydrofluoric acid (HF) to be used in the subsequent step, the as an insulating film 65 strong resistance against hydrofluoric acid e.g. S _i C _x N _y film A film is formed on the gate oxide film 61 to a thickness of about 1000 angstroms.

Step 5 shows a first sacrifice layer oxide film forming step. In this step, first, CVD is performed as a lower sacrificial layer for finally forming a space around the oscillation gate 40.
An oxide film 66 is formed on the insulating film 65 to a thickness of about 5000 angstroms by a (Chemical Vapor Deposition) method.
To form Next, the gate oxide film 61, the insulating film 65, and the oxide film 66 in the portion where the fixed end 38 is to be formed by photolithography are opened as openings 67 with respect to the fixed end.

Step 6 shows a polysilicon film forming step. This step is a pre-step for finally forming the vibration gate 40 and the fixed end 38. First, a polysilicon 68 is formed on the oxide film 66 and the opening 67 to a thickness of, for example, about 1 μm. Thereafter, boron is doped to impart conductivity. Next, a portion corresponding to the vibration gate 40 and a portion corresponding to the opening 67 are masked by photolithography, and then RIE (Reactive Ion Etchi
In step (ng), the polysilicon 68 is etched into a predetermined shape to form a plate-like beam 69 that finally becomes a vibration gate and a Y-shaped column 70 in the opening 67.

Step 7 shows a step of forming a second sacrificial layer oxide film. In this step, C is finally formed as a sacrificial layer except for the lower side for forming a gap around the oscillation gate 40.
An oxide film 71 is formed on the oxide film 66, the beam 69, and the support 70 to a thickness of about 5000 angstroms by the VD method.

Step 8 shows an oxide film etching step. First, the vicinity of the beam 69 is masked at the center of the vibrating gate by photolithography, and the vicinity of the column 70 and the portion of the column 70 except for the Y-shaped central portion 72 are masked at the fixed end. The films 66 and 71 are etched with hydrofluoric acid to form gap corresponding portions 73 and 74.

Step 9 shows a film formation process corresponding to a gap. In this step, an oxide film 75 as a sacrificial layer for introducing an etching solution used in a later step is formed by approximately 500.
The insulating film 65 and the gap corresponding portions 73 and 74 are formed on the entire surface by a CVD method so as to have a thickness of about Å. Thereafter, the oxide film 75 on the Y-shaped central portion 72 is removed by etching using photolithography technology.

Step 10 shows a shell corresponding portion forming step. 1 is formed on the oxide film 75 formed in the step 9.
The polysilicon 76 is formed to have a thickness of about μm. Thereafter, the stress remaining in the polysilicon of the shell and the vibration gate is removed by heat treatment for a short time by RTA (Rapid Thermal Aneal) to prevent them from being deformed. Thereafter, masking is performed using a photolithography technique, and the polysilicon 76 is etched by RIE to form a shell corresponding portion 77 in a range covering the vibration gate.

Step 11 shows an etching gap forming step. In this step, in order to form an oscillation gate and a shell, the oxide film 75 is etched using hydrofluoric acid and removed to form an introduction hole 78, and then a gap is formed through the introduction hole 78. The corresponding parts 73 and 74 are also removed. In this way, the vibration gate 40, the fixed end 38,
And a shell 79 are formed.

Step 12 shows a vacuum sealing step. In this step, the shell 79, the introduction hole 78, the insulating film 6
5 is formed with polysilicon 80 to a thickness of about 5000 angstroms, and the inside of the shell 79 is kept at a vacuum.

Step 13 shows a contact hole forming step. A gate oxide film 61, an insulating film 65, and a polysilicon 80 on the source portion 62 and the drain portion 63;
Are partially opened using photolithography technology and RIE to form contact holes 81 and 82. Similarly, a contact hole 84 can be formed in the gate portion.

Step 14 shows a diaphragm forming step. Using a potassium hydroxide (KOH) solution, the bottom of the silicon single crystal substrate 60 is etched so that the central portion is thin and the peripheral portion is thick, and the diaphragm 8 is formed.
Form 3

Step 15 shows a bonding step.
The electrodes 37, 36 made of aluminum are formed in the contact holes 81, 82, 84. The above is the manufacturing method of forming the diaphragm by covering the vibration gauge of the vibration type transducer with the shell.

In the etching gap forming step of Step 11, the oxide film 75 and the gap corresponding portions 73 and 74 are formed.
Although it has been described that hydrofluoric acid is used as an etching agent for removing the hydrogen fluoride, etching can be more effectively performed by using a predetermined gas phase etching instead of the hydrofluoric acid. Next, this will be described.

In the liquid phase etching using hydrofluoric acid, there is a problem that the vibration gate 40 easily adheres to the polysilicon shell 79 side or the insulating film 65 side due to the surface tension of the etching solution after the etching.

In such a case, for example, nitrogen gas 9
Vapor phase etching is performed using an etching gas mixed at a ratio of 5%, hydrogen fluoride gas 4.99%, and water vapor 0.01%. When a small amount of water vapor is mixed in this way, the reaction itself becomes a liquid-phase reaction by hydrogen fluoride gas due to the presence of the small amount of water vapor, but the surface is dried by the small amount of water vapor and the surface tension is greatly reduced. In addition, it is effective in preventing the attachment of the vibration gate.

[0114] oxide film as an etching gas and the sacrificial layer in this case reaction with such (S _i O ₂₎ 75 is approximately _{(S i O 2 + HF +} H 2 O) → (S i F 4 ↑ + H 2 O ↑) Becomes Of H ₂ O is water vapor left reaction, right H ₂
O ↑ indicates vaporized water vapor. S _i F ₄ ↑ indicates that fluoride silicon is volatilized (gasified).

[0115] Note that when the vapor-phase etching, a large proportion of the nitrogen gas (N ₂₎ is used, which is a a fluoride silicon gas (S _i F ₄ ↑) and water vapor (H ₂ O ↑) It is to remove enough.

Next, different configurations of the vibration gate will be sequentially described. FIG. 25 is a partial perspective view showing another configuration of the vibration gate corresponding to FIG. 9 (FIG. 8). Since no direct current bias is applied to the vibrating gate 40 shown in FIG. 8, there is no adhesion due to electrostatic force generated between the vibrating gate and the drain, but these are extremely close to each other. Since the surface has a flat surface with good smoothness, the vibration gate may be protected by the intermolecular force or the like in some cases.
May remain on the surface. In such a situation, the oscillation stops.

Therefore, the contact area between the vibration gate 85 and the protective film 42 is reduced to reduce the "adhesive force <restoring force".
A minute projection 86 is formed on the bottom of the vibration gate 85 along the longitudinal center axis of the vibration gate 85 so as to satisfy the following relationship. The projection 86 is formed, for example, as one or more continuous lines having a width of 1 μm or less, or a projection having a height of 500 Å or more in a broken line.

The projection 86 is provided with the vibration gate 8 in FIG.
5 is provided only at the bottom, but in addition to this,
A projection may be provided on the inner surface side of the shell covering the vibration gate 85 or on the upper part of the vibration gate 85, and these may be used together.

FIG. 26 shows a projection 8 on the bottom of the vibration gate 85.
6 shows a manufacturing method in the case where 6 is provided. This step is
This is added as a part of the steps shown in FIGS. Step A1 is inserted in the step next to Step 5 shown in FIG. In step A1, a resist 87 is applied on the oxide film 66 formed by the CVD method as a sacrificial layer in step 5, and a portion where the vibration gate 85 is to be finally formed is exposed by EB (Electron Beam) exposure. Or D
Patterning is performed by eep UV exposure or the like, and a groove 88 having a small projection width is formed in the oxide film 66 by RIE. After that, the resist 87 is removed.

Step A2 shows a polysilicon film forming step. In this step, the projecting beam 89 corresponding to the plate-like beam 69 which finally becomes the vibration gate in step 6 of FIG.
It is a process of forming as. Subsequent steps are the same as steps 7 to 10 in FIG. 23. In step 11, the vibration gate 85 having the protrusion 86 is formed.

If a predetermined tension is not applied to the vibration gate, buckling occurs when a compressive stress is applied. Therefore, it is necessary to apply a tensile strain to the vibrating gate, the center plane of the order to the tension film having a residual tensile strain so that the vibration gate shrinks than the substrate, for example such as S _i C _X N _y vibrating gate, or It is formed so as to be symmetrical on the upper and lower surfaces.

FIG. 27 is a partial perspective view showing a first configuration for applying tension to the vibration gate shown in FIG. 9 (FIG. 8). A tension film 91 is symmetrically formed on a plate-shaped central surface of the vibration gate 90 by sandwiching the tension film 91 between upper and lower vibration beams 92 and 93 made of polysilicon in a sandwich manner. The Membrane 91 is constituted by a S _i C _X N _y, the tension can be calculated thickness of the tension layer 91 of the vibrating beam 92 and 93, the Young's modulus, and the like residual strain.

Now, the total cross-sectional area of the vibrating beams 92 and 93 is A _S , the Young's modulus is E _S , the linear expansion coefficient is a _S , its tension is zero, the cross-sectional area of the tension film 91 is A _f , and the Young's modulus is To E _f ,
If the linear expansion coefficient a _f, the tension T, the temperature difference between (t-t _0), the tension X vibration gate _{90, X = (A S · E} S) T / (A S · E S + A _f · E _f ) (15).

The tension change ΔX _T due to the temperature change is given by ΔX _T = (A _S E _S A _F E _f ) (a _S -a _f ) (t-t ₀ ) / (A _S E _S + A _f · E _f ) (16) Here, a _S = 2.5 × 10 ⁻⁶ (1 / °)
C), a _f = 0 (1 / ° C.), E _f = 2E _S , A _S = 10A
_Assuming that _f , the tension change ΔX _T due to this temperature change is ΔX _T / (A _S · E _S ) = 0.418 (με / ° C).

The value of the strain obtained from the result and the data of the frequency change due to the temperature change is about 40 με / 100 ° C.
Becomes This temperature change is mainly caused by a difference in linear expansion coefficient between the polysilicon of the vibrating beams 92 and 93 and the nitride film of the tension film 91.

FIG. 28 is a process chart showing a method of manufacturing the vibration gate 90 as described above. Here, a description will be given as a form in which the manufacturing process shown in FIG. 28 is inserted into the process diagrams shown in FIGS.

Step B1 is inserted in the step following Step 5 shown in FIG. Step B1 is a first polysilicon film formation step, in which a polysilicon 94 is formed as a sacrificial layer in Step 5 on the oxide film 66 formed by the CVD method, for example, to a thickness of about 5000 angstroms. , Doping with boron to impart conductivity.

Next, the process proceeds to a step B2. Step B2 shows a nitride film forming step. In this step, the carbonized nitride layer _{(S i C x N y)} 95 over the polysilicon 94 is deposited by, for example, about 1000 Angstroms thick.

Step B3 shows a second polysilicon film formation step. This step is similar to the step of forming the first polysilicon film in step B1.
A film is formed with a thickness of about 00 angstroms, and thereafter, conductivity is imparted by doping with polon.

Step B4 shows a photolithography process. The polysilicon 94, the nitride film 95, and the polysilicon 96 are removed by masking using photolithography and RIE to form a vibrating gate corresponding portion 97. After this, Figure 2
The process is the same as Steps 8 to 10 of Step 3, and reaches Step B5 (corresponding to Step 11 in FIG. 23).

FIG. 29 is a partial perspective view showing a second configuration for applying tension to the vibration gate shown in FIG. 9 (FIG. 8). Polysilicon 99 is arranged on the plate-shaped central surface of the vibration gate 98, and is symmetrically sandwiched between upper and lower tension films 100 and 101 in a sandwich manner. These tension films 100 and 101 is constituted by a S _i C _X N _y,
What is necessary is just to perform the manufacturing method according to the case of the vibration gate 90. Further, the tension of the vibration gate 98 can be calculated in the same manner as in the case of the vibration gate 90.

[0132]

As described above in detail with the embodiments, according to the first aspect of the present invention, a drain and a source are separated from a semiconductor substrate at a predetermined interval and a channel is formed between them. Then, the vibrating gate is arranged at an interval facing the channel and the drain, and self-oscillation is performed by utilizing the relationship between the electrostatic attraction force of the vibrating gate and the drain and the phase of the drain current. It does not require magnets of large dimensions as in the past, so there is no need to precisely arrange the gap between the vibration gate (vibrator) and the magnet,
There is an advantage that the size can be reduced and the manufacturing process can be simplified.

In addition to the above advantages, firstly, since the bias voltage (corresponding to E2 in FIG. 1) is not applied to the oscillation gate, the circuit configuration is simplified to reduce the influence of the fluctuation of the bias voltage. By removing it, the frequency stability can be improved. Secondly, by eliminating the bias voltage, it is possible to prevent the vibrating gate from adhering to the substrate due to the electrostatic attraction force and to prevent the operation gate from becoming inoperable. Third, since polysilicon is used for the vibration gate, it is possible to impart conductivity to the vibration gate by doping impurities while enabling fine processing.

Further, according to the manufacturing method of the eleventh aspect , the step of using the two-layer structure in which the silicon substrate facing the vibrating gate is covered with an insulating film having high resistance to hydrofluoric acid is used. The film does not appear on the surface of the substrate, so that the film is resistant to contamination by alkali metal ions and the like, and the stability is improved, and the vibrating gate and shell can be finely processed with high precision.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing a configuration of one embodiment of the present invention.

FIG. 2 is an explanatory diagram for explaining the operation of the embodiment shown in FIG. 1;

FIG. 3 is an equivalent circuit expressing the device configuration shown in FIG. 1 as an equivalent circuit.

FIG. 4 is an explanatory diagram showing a rule of a sign of each part of the device configuration shown in FIG. 1 necessary for calculation.

FIG. 5 is an explanatory diagram for decomposing each part of the equivalent circuit shown in FIG. 3 to obtain a transfer function.

6 is an equivalent circuit diagram obtained by further simplifying the equivalent circuit shown in FIG.

FIG. 7 is an overall configuration diagram in which the present invention is configured to function as a self-excited oscillation circuit.

FIG. 8 is a perspective view showing the configuration of another embodiment of the present invention.

FIG. 9 is a cross-sectional view showing a cross section near the center of FIG. 8;

FIG. 10 is an explanatory diagram for explaining open-loop characteristics of the vibration gauge shown in FIG. 9;

11 is a characteristic diagram showing characteristics of a DC bias voltage versus an AC output of the vibration gauge shown in FIG.

12 is a characteristic diagram showing a DC bias voltage versus frequency output characteristic of the vibration gauge shown in FIG.

FIG. 13 is a characteristic diagram showing a characteristic of a voltage V _ds between drain / source and AC output of the vibration gauge shown in FIG. 9;

FIG. 14 shows a vibration type transducer in which a self-excited oscillation circuit is configured using the vibration gauge shown in FIG.

FIG. 15 shows a vibrating transducer of the present invention in which a self-excited oscillation circuit is formed using an inductor.

FIG. 16 is a waveform chart illustrating the operation of the vibration transducer shown in FIG. 15 (a).

FIG. 17 shows another example of a vibrating transducer of the present invention configured using a narrow band operational amplifier.

18 is a characteristic diagram showing characteristics of the operational amplifier shown in FIG.

FIG. 19 shows still another example of the vibration type transducer of the present invention in which a self-excited oscillation circuit is formed using a capacitor.

FIG. 20 shows another vibrating transducer of the present invention in which a self-excited oscillation circuit is formed using an inductor.

FIG. 21 is a waveform chart illustrating the operation of the vibration transducer shown in FIG.

FIG. 22 shows another vibrating transducer of the present invention in which a self-excited oscillation circuit is formed using a capacitor.

FIG. 23 is a manufacturing process diagram mainly explaining the manufacturing process for manufacturing the vibration gauge shown in FIG. 8;

FIG. 24 is a manufacturing process diagram illustrating the manufacturing process following FIG. 23;

FIG. 25 is a partial perspective view showing another configuration of the vibration gate corresponding to FIG. 9 (FIG. 8).

FIG. 26 is a process chart showing a manufacturing method for manufacturing the vibration gate shown in FIG. 25.

FIG. 27 is a partial perspective view showing a first configuration for applying tension to the vibration gate shown in FIG. 9 (FIG. 8).

FIG. 28 is a process chart showing a method of manufacturing the vibration gate shown in FIG. 27.

FIG. 29 is a partial perspective view showing a second configuration for applying tension to the vibration gate shown in FIG. 9 (FIG. 8).

FIG. 30 is a perspective view of a configuration using a conventional vibration sensor as a pressure sensor.

31 is a configuration diagram in which a portion A in FIG. 29 is enlarged and a vibration detection circuit is connected thereto.

FIG. 32 is a cross-sectional view showing an AA ′ cross section in FIG. 31;

FIG. 33 is an explanatory diagram showing the configuration shown in FIG. 31 as an electrical equivalent circuit.

[Explanation of symbols]

 10, 24, 35 Silicon substrate 11, 83 Diaphragm 12 Thick part 13 Pressure guiding hole 15 Pressure guiding tube 16, 31 Vibrator 17 Magnet 21 Amplifier 22 Transformer 23 Feedback line 30 Vibration gate 33 Automatic gain control circuit 38, 39 Fixed end 40 , 85, 98 Vibration gate 45, 52 Vibration gauge 46 Variable amplifier circuit 48 90 ° phase shift circuit 50 Error amplifier 60 Substrate 61 Gate oxide film 62 Source part 63 Drain part 65 Insulating film 73, 74 Gap corresponding part 78 Inlet hole 79 Shell S source D drain E1, E2, E3, E5, E6 DC power supply CNN1, CNN2 Channel CC1-CC3 Constant current source

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Masaaki Shinkoku 2-9-132 Nakamachi, Musashino City, Tokyo Inside Yokogawa Electric Corporation (72) Inventor Shigeru Goto 2-9-132 Nakamachi, Musashino City, Tokyo Inside Yokogawa Electric Corporation (72) Inventor Shunichi Miyazaki 2-9-132 Nakamachi, Musashino-shi, Tokyo Examiner in Yokogawa Electric Corporation Masahide Kawaguchi (56) References JP-A-61-100627 (JP, A) JP-A-61-222178 (JP, A) JP-A-2-196472 (JP, A) JP-A-59-95420 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H01L 29/84 G01L 1/10 G01L 9/00

Claims (14)

(57) [Claims] 1. A semiconductor substrate having a first conductivity type, a channel formed on a surface of the substrate and sandwiched between a drain and a source having a second conductivity type opposite to the conductivity type, and a surface of the substrate. The gate oxide film formed on
High resistance to hydrofluoric acid over the gate oxide film
An insulating film and the insulating film mainly made of polysilicon.
And maintain the gap so as to face the channel
A conductive vibrating gate fixed at both ends to the substrate;
A vibrating transducer having a shell covering the moving gate and having a vacuum maintained inside . 2. At least one of upper and lower surfaces of the vibrating gate.
Either one or one with a width of 1 μm or less on the inner surface of the shell
Above 500 angstroms solid line or broken line
Claims characterized by having an upper projection
2. The vibration transducer according to claim 1 . 3. The method according to claim 1, wherein said oscillation gate is formed of polysilicon and polysilicon.
The tension film with tensile residual strain
The vibration type transducer according to claim 1, wherein the vibration type transducer has a structure in which the vibration type transducer is laminated . 4. A phase of a drain current flowing through the drain.
Is shifted by 90 ° by 90 ° phase shifting means.
2. The vibrating transducer according to claim 1, wherein the vibration type transducer returns to the step (c) to continue self-excited oscillation . 5. An inductor is used as said 90 ° phase shift means.
Resonant transducer of paragraph 4, wherein the range of the claims, characterized in that there. 6. A vibratory transducer according to claim 4, wherein a capacitor is used as said 90 ° phase means. 7. The vibratory transducer according to claim 4, wherein a narrow band amplifier is used as said 90 ° phase shift means. 8. The phase of a drain current flowing through the drain
90 ° phase shift to output 90 ° phase voltage
Means and the drain voltage corresponding to this drain current
Error output compared to the constant set voltage
Width means, the 90 ° phase voltage being input and the error voltage
Controls the amplification degree so as to correspond to the set voltage.
To the oscillation gate as a feedback signal of constant amplitude
The vibratory transducer according to claim 1 , further comprising a variable amplifying means . 9. The drain resistance RD of the drain and the drain.
The capacitance CD between the rain and the substrate and the oscillation
The product (ωRDCD) with the oscillation angular velocity ω is much more than 1
The vibrating transducer according to claim 1 , wherein these values are selected so as to increase . 10. A oscillating gate is generated by applying a DC voltage.
The vibration type transducer according to claim 1 , which controls the amplitude of vibration. 11. A silicon substrate having a thin bottom portion.
A vibrating gage with both ends fixed with a predetermined spacing on the upper surface of
Form a shell that covers and is spaced around it
The method of manufacturing a vibration transducer that was formed in the gate oxide film over (A) the substrate, after this,
Only the part corresponding to the channel is spaced from the substrate
Forming drain and source by ion implantation of pure material
Have a high resistance to hydrofluoric acid on the gate oxide film.
There is covered with an insulating film, after forming the first sacrificial oxide film on the (B) the insulating film
Deposit polysilicon on this to be the oscillation gate
Doping impurities to impart conductivity and form a predetermined shape,
A second sacrificial oxide film is formed thereon to form the first and second sacrificial oxide films.
Etch the oxide film with hydrofluoric acid to form a gap
Forming a応部, (D) gap pairs to cover the insulating film and the gap corresponding portion
A suitable oxide film, and form a shell on top of this.
After depositing the gas, the gap corresponding portion and the acid corresponding to the gap are
The oxide film is removed by etching with hydrofluoric acid and
Forming the door and shell and the gap portion surrounded by these, (E) Thereafter, and said insulating film and said shell in vacuum port
Cover the gap with vacuum by vacuum
Of manufacturing a vibrating transducer. 12. After forming the first sacrificial oxide film,
Groove with electron beam in the upper part to be the vibration gate
And then depositing the polysilicon
Claim 11 characterized by shifting to a process.
13. A method for manufacturing a vibration-type transducer described in the section . 13. The polysilicon is provided with conductivity.
Then, a film is formed on this with a tension film having a residual tensile strain, and
Is further formed of the same polysilicon as the polysilicon.
To provide conductivity, and thereafter, the second sacrificial oxide film
Claims characterized by shifting to a film forming process
Item 12. The method for manufacturing a vibration transducer according to item 11 . 14. The method according to claim 1, wherein the liquid phase etching is performed using hydrofluoric acid.
Instead, a small proportion of water vapor and a large proportion of nitrogen gas
The first and second etching gases are mixed with hydrogen fluoride.
12. The method according to claim 11, wherein the sacrificial oxide film is subjected to gas phase etching .