JP2010232793A - Resonance circuit, method of manufacturing the same, and electronic device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a resonance circuit that suppresses variation in resonance frequency due to variance in structural size of a vibrator, and to provide a method of manufacturing the resonance circuit. <P>SOLUTION: The resonance circuit 30 includes a substrate, a MEMS resonator 10 which includes a fixed electrode 12 formed on the substrate and a movable electrode 13 having a movable portion opposed to at least part of the fixed electrode, and a voltage application means 20 of applying a bias voltage to the MEMS resonator, and the voltage application means includes a voltage dividing circuit which includes a resistance R11 for compensation which is composed of the same layer with the movable portion and varies in resistance value with the thickness of the layer, and a reference resistance R12 which is connected to the resistance for compensation and composed of a layer different from the movable portion, and outputs the potential at the connection point between the resistance for compensation and reference resistance to at least one terminal 10b between a first terminal and a second terminal of the MEMS resonator. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a resonant circuit, a manufacturing method thereof, and an electronic device, and more particularly, to a configuration and a manufacturing method of a resonant circuit using a MEMS resonator.

  In the recent electronic device market, products using MEMS (Micro Electro-Mechanical System) technology are accelerating. A product using MEMS technology refers to an electromechanical device (MEMS device) having a minute structure (MEMS structure) formed on a substrate using semiconductor manufacturing technology. Specific examples of devices using MEMS technology include various types of microsensors such as an acceleration sensor, an angular velocity sensor, an inertial sensor, and a pressure sensor, but the market has been rapidly expanding in recent years. In this way, MEMS technology has the potential to create new devices, and in the future, devices that take advantage of the features of microstructures such as micromechanical relays (switches) and variable capacitance elements, and combinations thereof, will be used. It is expected to be put to practical use one by one.

  As a new application example using the MEMS technology, there is a resonance circuit including a MEMS resonator. As an operation principle of a resonance circuit provided with this type of MEMS resonator, an electrostatic drive / detection method and a piezoelectric drive / detection method are representative, but the former is particularly compatible with a semiconductor process such as a CMOS process. Therefore, it can be said to be an advantageous technique for downsizing and cost reduction. In this electrostatic drive / detection method, a MEMS resonator having a movable electrode and a fixed electrode is formed, and vibration of the movable electrode caused by electrostatic force is detected by a change in capacitance between the movable electrode and the fixed electrode. is there. Examples of typical transmission and phase characteristics of the resonant circuit are shown in FIGS. 3 (a) and 3 (b). As documents disclosing such a resonance circuit, there are the following patent documents 1 to 3.

US Pat. No. 6,249,073 US Pat. No. 6,424,074 JP 2004-58228 A

  By the way, in the above-described resonance circuit, the resonance frequency varies due to variations in the structural dimensions of the MEMS resonator. Therefore, in order to ensure the accuracy of the resonance frequency, it is necessary to reduce the variation in the structural dimensions at the time of manufacturing. If the variation cannot be reduced, it is necessary to correct the frequency by providing a trimming step after the manufacturing. The manufacturing process becomes complicated and the manufacturing cost increases.

  In particular, among the structural dimensions of MEMS resonators, the thickness of the movable part of the movable electrode has a large effect directly on the resonance frequency and the thickness of the movable part is as thin as several um or less. Will change.

  Accordingly, the present invention is to solve the above-described problems and to realize a resonance circuit and a method for manufacturing the same that can suppress fluctuations in the resonance frequency caused by variations in the structural dimensions of the vibrator.

  In view of such circumstances, the resonant circuit of the present invention includes a substrate, a fixed electrode formed on the substrate, and a MEMS resonator having a movable electrode provided with a movable portion facing at least a part of the fixed electrode. Voltage applying means for applying a bias voltage to the MEMS resonator, and the voltage applying means includes a compensation resistor that is formed of the same layer as the movable portion and has a resistance value that varies depending on the thickness of the layer. A reference resistor connected to a compensation resistor and composed of a layer different from that of the movable portion, and a potential at a connection point between the compensation resistor and the reference resistor is selected from a first terminal and a second terminal of the MEMS resonator. A voltage dividing circuit for outputting to at least one of the terminals.

  According to the present invention, the voltage application means includes a compensation resistor composed of the same layer as the movable portion, and when the thickness of the movable portion layer changes, the resistance value of the compensation resistor changes due to the change in the thickness of the same layer. To do. This change in the resistance value of the compensation resistor changes the output potential of the voltage dividing circuit, thereby changing the bias voltage. Therefore, if the change in the resonance frequency due to the change in the thickness of the movable part is compensated by the change in the bias voltage due to the change in the thickness of the compensation resistor, the change in the resonance frequency is suppressed autonomously. The frequency accuracy can be improved and stabilized, and the trimming step can be omitted if the resonance frequency accuracy is within an allowable range.

  In one aspect of the present invention, the voltage applying means outputs the connection point potential to one of the first terminal and the second terminal of the MEMS resonator, and Another potential is output to the other terminal. According to this, the voltage applying means outputs the connection point potential to one terminal of the MEMS resonator and outputs the other potential to the other terminal, so that the potentials of both terminals of the MEMS resonator are both controlled by the voltage applying means. As a result, an arbitrary bias voltage can be reliably applied to the MEMS resonator.

  In this case, in the voltage dividing circuit, the compensation resistor and the reference resistor are connected in series between a high potential and a low potential, and when the junction potential is higher than the other potential, the compensation resistor The high potential is connected to the resistance side and the low potential is connected to the reference resistance side. When the connection point potential is lower than the other potential, the low potential is connected to the compensation resistance side. At the same time, the low potential is connected to the reference resistance side. According to this, when the potential at the connection point is higher than the other potential, a potential higher than that at the reference resistance side is supplied to the compensation resistance side in the voltage dividing circuit. When the connection point potential has a negative correlation (that is, when the resistance value increases as the resistance layer becomes thinner), if the connection point potential decreases, the bias voltage decreases, so that fluctuations in the resonance frequency are suppressed. . Contrary to the above, when the potential at the connection point is lower than the other potential, a potential lower than that on the reference resistance side is supplied to the compensation resistance side, thereby connecting to the thickness of the compensation resistance layer. When the point potential has a negative correlation (that is, when the resistance value increases as the resistance layer becomes thinner), if the connection point potential increases, the bias voltage decreases, so that fluctuations in the resonance frequency are suppressed. Therefore, in any case, the compensation resistor can be constituted by a resistance layer having a resistance value having a negative correlation with the thickness of the layer, so that a normal resistance layer, that is, a resistance layer having a pair of connection terminals at the edge portion. This makes it possible to configure a compensation resistor.

  In another aspect of the present invention, an input-side coupling capacitor connected between the first terminal and the input terminal of the MEMS resonator, and an output-side coupling capacitor connected between the second terminal and the output terminal. And further comprising. According to this, the input side coupling capacitor is connected between the first terminal of the MEMS resonator and the input terminal of the resonance circuit, and the output side coupling is connected between the second terminal of the MEMS resonator and the output terminal of the resonance circuit. By connecting the capacitor, the MEMS resonator and the external circuit are separated in a direct current, so that a desired bias voltage can be applied to the MEMS resonator regardless of the potential of the external circuit. The effect of suppressing the fluctuation of the resonance frequency can also be enhanced.

  Another resonant circuit of the present invention includes a substrate, a fixed electrode formed on the substrate, and a MEMS resonator having a movable electrode having a movable portion facing at least a part of the fixed electrode, A first circuit for supplying a first potential to a first terminal of the MEMS resonator; and a second circuit for supplying a second potential to a second terminal. The MEMS resonator includes the first circuit. Voltage applying means for applying a bias voltage corresponding to the difference between the potential and the second potential, and at least one of the first circuit and the second circuit is the same layer as the movable part. A compensation resistor having a resistance value that varies depending on the thickness of the layer; and a reference resistor that is connected to the compensation resistor and is composed of a layer different from the movable portion. The connection between the compensation resistor and the reference resistor The point potential is the first potential or the first potential. Characterized in that it is a dividing circuit for outputting a potential.

  Further, in a different aspect of the present invention, in each of the above-described inventions, the movable portion and the compensation resistor are configured by a conductor layer formed of the same material, and the reference resistance is the movable portion and the compensation resistor. Consists of conductor layers made of different materials. Here, it is preferable that the movable part and the compensation resistor are formed of a polysilicon layer.

  Furthermore, in a further different aspect of the present invention, in each of the above-mentioned inventions, the movable part and the compensation resistor are constituted by adjacent conductor layers. In this case, it is preferable that the movable part and the compensation resistor are accommodated in a common accommodation space. Also in this case, it is desirable that the movable part and the compensation resistor are formed of a polysilicon layer. In the present specification, the adjacent conductor layers mean that other conductors (other conductor layers, wiring layers, etc.) are arranged on the substrate in a manner that they are not interposed therebetween.

  Next, a method for manufacturing a resonant circuit according to the present invention includes a MEMS resonator having a fixed electrode and a movable electrode having a movable portion facing at least a part of the fixed electrode, and voltage application for applying a bias voltage to the MEMS resonator. The voltage applying means includes a compensation resistor and a reference resistor, and a connection point potential between the compensation resistor and the reference resistor is applied to at least one terminal of the MEMS resonator. A voltage dividing circuit for outputting, forming the movable part, forming the compensation resistor in the same layer as the movable part, and forming the reference resistance in a layer different from the movable part; It is characterized by comprising. Here, the step of forming the movable part and the compensation resistor and the step of forming the reference resistance may be performed first.

  In one aspect of the present invention, the movable part and the compensation resistor are made of a conductive layer made of the same material and in the same method, and the reference resistance is a material different from the movable part and the compensation resistor. Or it is comprised with the conductor layer formed by the different method. Here, it is preferable that the movable part and the compensation resistor are formed of a polysilicon layer.

  In another aspect of the invention, the movable part and the compensation resistor are formed of adjacent conductor layers. According to this, since the movable portion and the compensation resistor are adjacent to each other, it is possible to increase the relevance with respect to the film thickness value depending on the formation conditions (for example, film formation conditions), so that the resonance frequency compensation effect Can be increased.

  In another aspect of the present invention, it is preferable that the method further includes a step of forming a common accommodation space in which the movable portion and the compensation resistor are accommodated together. According to this, since the movable part and the compensation resistor are arranged together in the accommodating space, it is possible to prevent the compensation resistor from being increased in temperature by surrounding elements. In addition, since the compensation resistance is subjected to the etching action under the same conditions at the same time during the release process for the movable part, the contents of the post-processing applied after the formation of the movable part and the compensation resistance are the same. The influence (change in structural dimensions, etc.) due to the difference can be reduced, and as a result, the resonance frequency compensation action by the compensation resistor can be enhanced. Also in this case, it is desirable that the movable part and the compensation resistor are formed of a polysilicon layer.

  In another aspect of the present invention, the movable portion and the compensation resistor are formed on the same layer. According to this, since the movable part and the compensation resistor are formed on the same base, it is possible to make the formation (film formation) conditions of the movable part and the compensation resistor more equivalent. The relevance of the film thickness can be further increased, and as a result, the resonance frequency compensation effect by the compensation resistor can be further improved.

  Next, an electronic device according to the present invention is characterized in that any one of the resonance circuits described above and a signal circuit unit connected to the resonance circuit are formed on the same substrate. According to this, since the resonance circuit including the MEMS resonator and the voltage applying unit and the signal circuit unit can be integrally formed on the same substrate, it can be configured compactly, and in particular, a plurality of components such as the MEMS resonator and the signal circuit unit. By forming them in parallel, it becomes possible to form a plurality of components in the same process, so that it is possible to improve manufacturing efficiency and reduce manufacturing costs. Here, the electronic device is typically configured as a semiconductor device (semiconductor chip) in which the above configuration is integrally provided on a single semiconductor substrate.

  As an example of such an electronic device, there is a case where an oscillation circuit is configured using an amplifier circuit having a negative resistance as the signal circuit portion. Examples of the substrate include a semiconductor substrate such as a silicon substrate, a glass substrate, and a ceramic substrate. When a MEMS resonator is configured as a vibrator, a semiconductor device using a semiconductor substrate such as a silicon substrate is configured. It is preferable to do.

  According to the present invention, it is possible to achieve an excellent effect that the fluctuation of the resonance frequency of the vibrator due to the variation in the structural dimensions of the movable part can be suppressed and the accuracy of the frequency characteristics of the resonance circuit can be improved.

The schematic perspective view which shows the structural example of the vibrator | oscillator of the resonance circuit which concerns on this invention. The longitudinal cross-sectional view (a) and (b) of the vibrator | oscillator. Graphs (a) and (b) showing the pass characteristics of the vibrator and the frequency dependence of the phase displacement. The graph which shows the bias voltage dependence of the passage characteristic of a vibrator. The circuit diagram which shows the structure of 1st Embodiment of a resonance circuit. The circuit diagram which shows the structure of 2nd Embodiment of a resonance circuit. The block diagram which shows typically the relationship of the structural dimension of the vibrator | oscillator of a resonance circuit, and the resistance element for compensation. The schematic perspective view (a) and (b) which shows the more concrete example of arrangement | positioning of the component of a resonance circuit. The schematic plan view which shows the more specific other example of arrangement | positioning of the component of a resonance circuit. FIG. 6 is a circuit diagram illustrating a configuration example of an electronic device including a resonance circuit. The fragmentary top view (a) and fragmentary sectional view (b) which show the example of composition of a standard resistance element. Circuit diagrams (a) to (c) showing other configuration examples of the reference resistance element. Schematic process sectional drawing (a) thru | or (h) which show the manufacturing method of a resonant circuit. Schematic process sectional drawing (a) thru | or (h) which show the other example of the manufacturing method of a resonance circuit.

  Next, embodiments of the present invention will be described in detail with reference to the accompanying drawings. FIG. 1 is a schematic perspective view showing a configuration example of a MEMS resonator 10 that can be used in the resonance circuit of the present embodiment, and FIG. 2 is the same vibrator showing cross sections taken along lines AA and BB in FIG. 10 is a cross-sectional view (a) and (b) of FIG.

The MEMS resonator 10 is configured on a substrate 1 made of a semiconductor such as a silicon single crystal. An insulating film 2 such as silicon oxide (SiO ₂ ) is formed on the substrate 1, and a base layer 11 of the vibrator 10 such as silicon nitride is further formed on the insulating film 2. On the base layer 11, a fixed electrode 12 made of polycrystalline silicon or the like imparted with conductivity by doping and a movable electrode 13 made of the same material are formed. Here, a part of the fixed electrode 12 constituting the vibrator electrode of the vibrator 10 and the movable portion 13a of the movable electrode 13 are arranged on the substrate 1 so as to face each other in the vertical direction at an appropriate interval (several um or less). . The movable portion 13a is configured to bend in the vertical direction according to the electrostatic attractive force generated between the movable electrode 13a and the movable electrode 13a so as to be movable in the direction in which the distance between the movable electrode 13a and the fixed electrode 12 is changed.

  Although the illustrated MEMS resonator 10 has a cantilever structure in which the movable electrode 13 is formed in a cantilever shape to form a movable portion 13a, this is merely an example of the vibrator of the present invention. In the invention, it is possible to use a MEMS resonator adopting various shapes and structures, such as a movable part of the MEMS resonator formed in a doubly-supported beam shape or a disk-shaped movable part.

FIG. 3 is a schematic characteristic diagram showing the electrical characteristics of the MEMS resonator 10 of the resonance circuit according to the present invention. Here, FIG. 3A is a graph showing pass characteristics when an AC signal is input to the MEMS resonator 10, and FIG. 3B is a graph showing phase displacement. As described above, the MEMS resonator 10 has a resonance characteristic having an appropriate resonance frequency. The resonance frequency f _n is _expressed by the following equation (1).
_{f n = (1 / 2π)} · (k m / m eff) 1/2 ··· (1)
Here, _{k m} is the spring constant determined from the structure of the MEMS resonator _{10, m eff} is the effective mass of the structure.

On the other hand, electrostatic attraction between the application appropriate bias voltage (DC bias) both electrodes between the fixed electrode 12 and movable electrode 13 is generated in the MEMS resonator 10, it acts as to reduce the spring constant k _m To do. For this reason, the resonance frequency f _n is reduced by the above equation (1). In general, the resonance frequency f _n decreases as the bias voltage increases.

FIG. 4 is a graph showing the resonance characteristics of the AC signal when the bias voltage Vb applied to the MEMS resonator 10 is changed. Thus, the resonance frequency f _n decreases as the bias voltage Vb increases.

  FIG. 5 is a circuit diagram showing a configuration of the resonance circuit 30 including the MEMS resonator 10 used in the first embodiment. In this example, the input terminal 30 a and the output terminal 30 b of the resonance circuit 30 are the first terminal 10 a and the second terminal 10 b of the MEMS resonator 10 (one of which is connected to the fixed electrode 12, whichever The other terminal is connected to the movable electrode 13 side terminal), the first terminal (input side terminal) 10a of the MEMS resonator 10 and the input terminal 30a, and the second terminal A bias circuit 20 serving as a voltage application unit is connected between the (output side terminal) 10b and the output terminal 30b. An input side coupling capacitor Ca is connected between the input terminal 30a and the first terminal 10a, and an output side coupling capacitor Cb is connected between the output terminal 30b and the second terminal 10b.

  The bias circuit 20 is a component for applying a DC bias voltage to both terminals 10 a and 10 b of the MEMS resonator 10. The bias circuit 20 includes a voltage dividing circuit 21 in which a compensation resistor R11 and a reference resistor R12 are connected in series, and a voltage is applied to both ends of the series circuit of both resistors. In the case of the illustrated example, in the voltage dividing circuit 21, the DC power supply potential Vdc that is a high potential is supplied to the compensation resistor R11 side, and the reference resistor R12 side is connected to a ground potential that is a low potential. The connection point potential Vp between the compensation resistor R11 and the reference resistor R12 is applied to the second terminal 10b on the output side of the MEMS resonator 10. This connection point potential Vp becomes the second potential Vp = Dv · Vdc corresponding to the potential difference pressure Vdc at both ends of the voltage dividing circuit 21 and the voltage dividing ratio Dv = R12 / (R11 + R12).

  On the other hand, the bias circuit 20 is provided with a potential supply unit 22 connected to the first terminal 10a on the input side of the MEMS resonator 10, and the potential supply unit 22 is connected to a constant potential (grounded in the illustrated example) via a resistor R13. The first potential Vq is applied to the input terminal 10a of the MEMS resonator 10. Here, in the present embodiment, the connection potential Vp which is the second potential is configured to be always higher than the first potential Vq. In this manner, the MEMS resonator 10 has a bias voltage corresponding to the difference between the second potential Vp supplied by the voltage dividing circuit 21 of the bias circuit 20 and the first potential Vq supplied by the potential supply unit 22. Vb = Vp−Vq is applied. The resonance circuit 30 is used with an input terminal 30a and an output terminal 30b connected to an external circuit. However, the external circuit and the MEMS resonator 10 are DC-isolated by an input side coupling capacitor Ca and an output side coupling capacitor Cb. Therefore, only the bias voltage Vb from the bias circuit 20 is applied to the MEMS resonator 10 as a DC bias.

  In the first embodiment, the second potential Vp from the voltage dividing circuit 21 is supplied to the second terminal 10b on the output side of the MEMS resonator 10, and the other first voltage is supplied to the first terminal 10a on the input side. The potential Vq is supplied by the potential supply unit 22, but conversely, the potential Vp from the voltage dividing circuit 21 is supplied to the first terminal 10a on the input side as the first potential, and the second potential on the output side is supplied. Another potential Vq may be supplied to the terminal 10b as the second potential. In addition, the bias circuit 20 may be configured by only the voltage dividing circuit 21 that supplies the potential Vp to any one terminal of the MEMS resonator 10. Furthermore, although the case where the potential Vq is lower than the potential Vp supplied by the voltage dividing circuit 21 has been described in the above example, for example, the voltage dividing circuit can be configured by connecting the potential Vq to a high potential via the resistor R13. 21 may be higher than the potential Vp supplied by the power supply 21.

  FIG. 6 is a circuit diagram showing a configuration of a resonance circuit 30 ′ of the second embodiment. This example is different from the first embodiment in that a bias circuit 20 ′ for supplying potentials Vp and Vq by a voltage dividing circuit is provided at both ends of the MEMS resonator 10, but all other configurations are the same. Is omitted.

  Here, a voltage dividing circuit 21 having a compensation resistor R11 and a reference resistor R12 similar to those in the first embodiment is connected to the second terminal 10b on the output side of the MEMS resonator 10. Here, in the voltage dividing circuit 21, the compensation resistor R11 side is connected to the high potential Vdc1, and the reference resistor R12 side is connected to the low potential ground potential.

  A voltage dividing circuit 22 ′ having a reference resistor R 14 and a compensating resistor R 15 is connected to the first terminal 10 a on the input side of the MEMS resonator 10. In the voltage dividing circuit 22 ', the reference resistor R14 side is connected to the high potential Vdc2, and the compensation resistor R15 side is connected to the low potential ground potential.

  In this case, the connection point potential Vq of the voltage dividing circuit 22 ′ having the reference resistor R14 and the compensation resistor R15 is supplied to the first terminal 10a on the input side of the MEMS resonator 10 as the first potential, and the output side Since the connection point potential Vp of the voltage dividing circuit 21 including the compensation resistor R11 and the reference resistor R12 is supplied to the second terminal 10b as the second potential, the bias voltage of the MEMS resonator 10 is Vb as described above. = Vp-Vq. In this embodiment, the connection point potential Vp is always higher than the connection point potential Vq.

  In each of the embodiments described above, the compensation resistor R11 is formed of the same layer as the movable portion 13a of the vibrator 10. For example, the movable portion 13a and the compensation resistor R11 are configured simultaneously by forming a common conductive layer and patterning the conductive layer in a lump.

FIG. 7 is a diagram schematically showing the configuration of the MEMS resonator 10 and the compensation resistor R11 (that is, in a state in which the MEMS resonator 10 and the compensating resistor R11 are adjacently arranged irrespective of the actual arrangement). The MEMS resonator 10 has the fixed electrode 12 and the movable electrode 13 as described above, but the movable electrode 13 has a movable portion 13a facing the fixed electrode 12 of the fixed electrode 12, and a thickness H of the movable portion 13a, The resonance frequency fn of the MEMS resonator 10 is expressed by the following equation depending on the length (beam length) Lo (length extending in the extending direction in the range overlapping the fixed electrode 12) in the extending direction (left-right direction in the drawing) of the movable portion 13a. It is known that it can be calculated as (1 ').
fn = (1 / 2π) · (35E / 33ρ) ^1/2 · H / Lo ² (1 ′)
Here, E is the Young's modulus of the material constituting the movable portion 13a, and ρ is the density of the material. Therefore, according to the above equation (1 ′), when the thickness H of the movable portion 13a is increased, the resonance frequency fn is increased, and when the movable portion 13a is thinned, the resonance frequency fn is decreased. It can be seen that there is a positive correlation.

On the other hand, the resistance value of the compensation resistor R11 is orthogonal to the length Lr between both ends when considering the vicinity of both ends in the plane direction (near the left and right ends in the figure) as a connection point. Since the width Wr and the thickness (not shown) are the same as those of the movable portion 13a, they are given by the following expression (2).
R11 = ρr · Lr / (Wr · H) (2)
Here, ρr is the resistivity of the material. Therefore, according to this equation (2), the resistance value decreases as the resistance layer thickness H of the compensation resistor R11 increases, and the resistance value increases as the thickness H decreases, that is, the thickness H and the compensation resistor R11. It can be seen that the resistance value has a negative correlation.

  On the other hand, the reference resistor R12 has a structure different from that of the movable portion 13a, for example, a different layer formed in a separate process, or an external resistor. Accordingly, the compensation resistor R11 is common to the movable portion 13a with respect to the layer thickness H, and thereby the thickness H of the compensation resistor R11 varies in conjunction with the variation in the thickness H of the movable portion 13a. Does not necessarily relate to variations in the thickness H of the movable portion 13a. In the above description, the thickness of the movable portion 13a and the thickness of the compensation resistor R11 are the same. However, if both are formed of the same layer, the thickness increase / decrease due to the thickness variation at the time of manufacture is linked. Therefore, the thicknesses of the two do not necessarily match.

  The bias circuit sections 20 and 20 'are both configured to increase the supply potential Vp because the resistance value of R11 decreases when the layer thickness H increases by a predetermined amount. Specifically, when the compensation resistor R11 is lowered, unless the reference resistor R12 is similarly lowered, the voltage dividing ratio Dv is increased, so that the connection point potential Vp moves to the high potential side. As described above, in this embodiment, since the layer thickness H and the bias voltage Vb = Vp−Vq have a positive correlation, the thickness H of the movable portion 13a is determined by the above-described equation (1 ′) and the characteristics shown in FIG. When the resonance frequency fn is about to change due to a change (for example, an increase of a predetermined amount), a change in the resistance value of the compensation resistor R11 caused by the same change in the thickness H causes a change in the bias voltage and suppresses this. As a whole, fluctuations in the resonance frequency due to variations in the thickness of the movable portion 13a are reduced.

  The above items are common to both the first embodiment and the second embodiment. However, in the second embodiment, the compensation resistor R15 is formed of the same layer as the movable portion 13a in the same manner as the compensation resistor R11, so that the resistance value of the compensation resistor R15 decreases as the thickness H increases. Since the connection point potential Vq of the voltage circuit 22 'decreases, the bias voltage Vb = Vp-Vq (Vp> Vq) is increased in the same manner as described above. Therefore, the voltage dividing circuit 22 'also functions to suppress fluctuations in the resonance frequency fn caused by the change in the thickness H of the movable portion 13a.

  Therefore, in the second embodiment, the resistance R15 is a compensation resistance configured by the same layer as the movable portion 13a in the same manner as the resistance R11. As a result, both the second potential Vp and the first potential Vq change. It is possible to increase the rate of change of the bias voltage Vp due to the change of H. However, in the illustrated example, only the resistor R11 may be a compensation resistor, or only the resistor R15 may be a compensation resistor. If the resistor R11 is not a compensation resistor and the resistor R15 is not a compensation resistor, the first potential Vq is given by being connected to another constant potential as in the first embodiment.

  Note that the above description is made on the assumption that the potential Vp is higher than the potential Vq, but conversely, the case where the potential Vp is lower than the potential Vq is also conceivable. For example, in the configuration shown in FIG. 5, the potential Vq is connected to the high potential via the resistor R13 as described above, so that the configuration can be higher than the potential Vp supplied by the voltage dividing circuit 21. . As described above, when the potential Vp is lower than the potential Vq, the description about the level of the potential in the voltage dividing circuit that outputs the potential Vp and the increase / decrease of the bias voltage and the increase / decrease of the potential are reversed. In other words, it is appropriate.

  Similarly to the above, the relationship in which the second potential Vp of this embodiment is higher than the first potential Vq is changed so that the second potential Vp is lower than the first potential Vq. In this case, the connection position between the compensation resistor R11 and the reference resistor R12 in the dividing circuit 21 may be replaced. In the second embodiment, the connection position between the reference resistor R14 and the compensation resistor R15 in the voltage dividing circuit 22 '. Should be replaced.

  In each of the above embodiments, the compensation resistor R11 included in the voltage dividing circuit 21 is configured by a resistance layer having a resistance value having a negative correlation with the thickness of the layer. For example, electrodes are provided on both the front and back surfaces of the resistance layer. You may comprise by the resistance layer provided with the resistance value which connects and has a positive correlation with the thickness of a layer. In this case, it is only necessary to replace the connection position of the compensation resistor R11 and the reference resistor R12 in the dividing circuit 21 and connect the compensation resistor R11 to the low potential side. Similarly, in the voltage dividing circuit 22 ′ of the second embodiment, when the compensation resistor R15 is a resistance layer having a resistance value having a positive correlation with the layer thickness as described above, the compensation resistor R15 is The reference resistor R14 may be connected to the low potential side on the high potential side.

  FIG. 8A is a schematic perspective view schematically showing the configuration of the first embodiment. In FIG. 8, the connection relationship between circuit elements is schematically drawn with lines. In this figure, the substrate and the insulating film, base layer, interlayer insulating film and the like formed thereon are omitted. In the illustrated example, the fixed electrode 12 is formed by the first conductive layer formed on the substrate, and then an insulating film (not shown) (not shown) is formed on the fixed electrode 12, and then the second conductive layer is used. The movable electrode 13 and the compensation resistor R11 are formed. Here, the movable portion 13a of the movable electrode 13 and the resistance layer of the compensation resistor R11 are collectively formed by forming the second conductive layer and patterning it.

  Further, the resistance layer constituting the reference resistor R12 is formed in a process different from the process of forming the movable portion 13a. For example, it can be formed together with the fixed electrode 12 by forming the first conductive layer and patterning it. However, although the forming process of the compensation resistor 11 and the forming process of the reference resistor R12 are different processes, if they are common in that they are processes using the same kind of material and film forming method, they are formed by both processes. Since there is a possibility that the structure dimension is related, it is preferable that the formation process of the reference resistor R12 is different from the formation process of the compensation resistor R11 in at least one of the formation material and the formation method. For example, as a forming material, a metal layer or an impurity region in the substrate is used for conductive polycrystalline silicon, or as a forming method, sputtering method, vapor deposition method, ion implantation method is used for CVD method. And the like.

  An interlayer insulating film (not shown) is formed on the MEMS resonator 10, the compensation resistor R11, and the reference resistor R12 formed as described above, and a metal film is further formed on the interlayer insulating film. To the input side first terminal 10a, the output side second terminal 10b, the compensation resistance terminals 20a and 20b, which are conductively connected to the lower layer structure through through holes formed in the interlayer insulating film. Reference resistance terminals 20c and 20d and wiring 10c are formed. The first terminal 10 a on the input side is conductively connected to the movable electrode 13, and the second terminal 10 b on the output side is conductively connected to the fixed electrode 12. The output-side second terminal 10b is conductively connected to the compensation resistor terminal 20b and the reference resistor terminal 20c through the wiring portion 10c. The compensation resistor terminal 20a and the reference resistor terminal 20b are conductively connected to both ends in the longitudinal direction of the compensation resistor R11 extending in the plane direction, and the reference resistor terminals 20c and 20d are elongated in the plane direction of the reference resistor R12. Conductive connections are made to both ends in the direction.

  In the above configuration, the fixed electrode 12 and the movable portion 13a constituting the MEMS resonator 10 are arranged in a housing space 10A (indicated by a two-dot chain line in the drawing) where an interlayer insulating film or the like is not formed. The accommodation space 10A is formed by partially etching the interlayer insulating film in a release process by a method described later after forming each of the above structures on the substrate. In the accommodation space 10 </ b> A, the fixed electrode 12 and the movable portion 13 a are directly opposed to each other, and the movable portion 13 a is configured to be movable in a direction in which the facing interval is changed above the fixed electrode 12.

  In this illustrated example, the compensation resistor R11 and the reference resistor R12 are arranged such that the extending direction of the resistance layer is arranged in the same straight line on the substrate. Further, the extending direction of the compensation resistor R11 and the reference resistor R12 is the extending direction of the fixed electrode 12 and the movable electrode 13 (in the illustrated example, the direction coinciding with the direction in which the movable electrode 13 protrudes on the fixed electrode 12). Are formed so as to be parallel to each other.

  FIG. 8B is a schematic partial perspective view showing another configuration example different from the configuration shown in FIG. 8A of the first embodiment. Here, since the connection relation between each terminal part and wiring is the same as that of the above-described configuration, the description thereof is omitted. This configuration example is different from the example shown in FIG. 8A in that the compensation resistor R11 is disposed in the accommodation space 10A ′ formed on the substrate. By disposing the compensation resistor R11 in the accommodating space 10A ′ in which the MEMS resonator 10 is disposed in this way, the movable portion 13a and the compensation resistor R11 can be formed under the same manufacturing conditions.

  Also in this illustrated example, the compensation resistor R11 and the reference resistor R12 are arranged such that the extending direction of the resistance layer is arranged in the same straight line on the substrate. Further, the extending direction of the compensation resistor R11 and the reference resistor R12 is the extending direction of the fixed electrode 12 and the movable electrode 13 (in the illustrated example, the direction coinciding with the direction in which the movable electrode 13 protrudes on the fixed electrode 12). Are formed so as to be parallel to each other. However, in this example, the MEMS resonator 10 including the movable portion 13a and the compensation resistor R11 are disposed adjacent to each other in a direction intersecting (orthogonal) with the extension direction. That is, no conductor such as another conductive layer or wiring layer is interposed between the MEMS resonator 10 and the compensation resistor R11 on the substrate. Since the movable portion and the compensation resistor are adjacent to each other in this manner, the relevance can be increased with respect to the film thickness value depending on the film formation condition at the time of formation, so that the resonance frequency compensation effect can be enhanced.

  In the present embodiment, the accommodation space 10A ′ extends in a direction intersecting (orthogonal) with the extension direction and includes both the movable portion 13a and the compensation resistor R11 adjacent to each other. Thus, the movable part 13a and the compensation resistor R11 are arranged together in the common accommodation space 10A, so that the influence of etching when forming the accommodation space 10A ′ is equal to the movable part 13a and the compensation resistor R11. Therefore, it is possible to prevent a decrease in the relationship between the thicknesses due to the presence or absence of the etching process.

  FIG. 9 is a schematic plan view showing an example of a planar structure different from the above. In FIG. 9 as well, a part of the connection relation between circuit elements is schematically drawn with lines. In this example, the movable electrode 13 is connected to the first terminal 10a, and the fixed electrode 12 is connected to the second terminal 10b. However, the first terminal 10a is connected to the connection point between the compensation resistor R11 and the reference resistor R12, and the second terminal 10b is connected to the grounding resistor R13. The other end of the grounding resistor R13 is connected to the ground potential. The first terminal 10a is connected to an input side coupling capacitor (not shown), and the second terminal 10b is connected to an output side coupling capacitor (not shown). Even in this configuration, the compensation resistor R11 is disposed adjacent to the MEMS resonator 10 including the movable portion 13a. The MEMS resonator 10 including the movable portion 13a is accommodated in the accommodating space 10A ″. In this case, the accommodating space 10A ″ may be expanded to accommodate the adjacent compensation resistor R11.

  8 and 9 may be modified so that the fixed electrode 12 is connected to the first terminal 10a and the movable electrode 13 is connected to the second terminal 10b. Moreover, it is preferable that each said electrode and resistance are comprised by a polysilicon layer altogether. Generally, the sheet resistance of the polysilicon layer is easily obtained as a value of about 1.0 to 2.0 kΩ / □ (ohm / square).

  FIG. 10 is a schematic circuit configuration diagram showing an embodiment of an electronic device having a resonance circuit 30 according to the present invention and a signal circuit unit 40 connected to the resonance circuit 30. Here, in the illustrated example, a case where an oscillation circuit using the resonance circuit 30 of the first embodiment is configured is shown. In this example, the resonance circuit 30 'of the second embodiment may be used instead of the first embodiment.

  In this embodiment, the input terminal 30 a and the output terminal 30 b of the resonance circuit 30 are connected to the signal circuit unit 40, and the signal circuit unit 40 includes an amplifier circuit INV 1 having a negative resistance connected in parallel with the resonance circuit 30. And a feedback resistor Rfb connected in parallel with the resonance circuit 30 is provided. Further, both ends (input terminal 30a and output terminal 30b) of the amplifier circuit INV, the feedback resistor Rfb, and the resonance circuit 30 are connected to a constant potential (ground potential) via capacitances Cg and Cd. These parallel circuits are connected to the amplifier circuit INV2, and the output potential Vout is output from the amplifier circuit INV2.

  FIG. 11 is a partial plan view (a) and a partial cross-sectional view (b) showing a configuration example of reference resistors such as the resistors R12 and R14 of the bias voltage supply unit 20. In this configuration example, the MEMS resonator 10 and the bias circuit 20 are formed on the substrate 1 made of a common semiconductor. An insulating film 2 made of silicon oxide or the like is formed on a substrate 1 made of a semiconductor such as a silicon substrate, and wiring portions 3B and 3C made of aluminum or the like are formed on the insulating film 2. Impurity regions 1A are formed in the surface layer portion of the substrate 1, and contact regions 1B and 1C are formed on the surface portion in the impurity regions 1A. The contact regions 1B and 1C have a higher impurity concentration than the impurity regions 1A and have improved conductivity. Here, for example, if substrate 1 is a p-type substrate, impurity region 1A is an n-type impurity region, and contact regions 1B and 1C are high-concentration n-type impurity regions. The wiring portions 3B and 3C are conductively connected to the contact regions 1B and 1C through the through holes of the insulating film 2, respectively.

  Since the reference resistor R12 is constituted by the impurity region 1A formed in the surface layer portion of the substrate 1 as described above, the resistance constituent material and the forming method are different from those of the compensation resistor R11, so that the movable resistor formed on the substrate is formed. It can be set as the resistance element which does not interlock | cooperate with the structural dimension of the part 13a. In this case, the resistance value can be appropriately set according to the impurity concentration of the impurity region 1A, the region size, the planar distance between the contact portions 1B and 1C, and the like. In addition, since the main part of the resistance element is formed on the surface layer portion of the substrate 1, there is an advantage that the structure provided on the substrate 1 is hardly affected and the degree of freedom in designing the structure can be increased.

  FIG. 12 is a circuit diagram schematically showing another configuration example of the reference resistor. In the example shown in FIG. 12A, the channel resistance when a predetermined gate voltage is applied to the gate of a field effect transistor (FET) formed on the substrate is used as a resistance element. In the example shown in FIG. A constant current source having a gate connected to a source is configured and used as a resistance element. Further, the example shown in FIG. 12C shows an example in which a voltage dividing circuit connected in series by connecting gates and sources of a p-channel MOSFET and an n-channel MOSFET is formed. The bias circuit can be configured by replacing one of the reference resistors as the reference resistor and the other with the compensation resistor.

  Note that the reference resistance may be adjusted by trimming a resistor such as a polysilicon resistor formed in the substrate to a desired value. The reference resistor may be arranged as a separate element outside the substrate on which the MEMS resonator is formed. In short, the reference resistor may be formed of a structure or element whose resistance value does not change in conjunction with the compensation resistor R11 or R15, or both.

  Finally, with reference to FIG. 13, a method for manufacturing a resonant circuit according to the present invention will be described. In this example, a semiconductor substrate (for example, a p-type semiconductor substrate made of single crystal silicon or the like) 1 shown in FIG. 13A is prepared, and then oxidized on the substrate 1 as shown in FIG. 13B. An insulating film 2 made of silicon or the like is formed. Next, as shown in FIG. 13C, a base layer 11 made of an insulating film such as silicon nitride formed under the MEMS resonator 10 is formed. The underlayer 11 functions as a release etching stop layer in a release process described later.

  Thereafter, as shown in FIG. 13D, a fixed conductive layer 12 is formed by forming a first conductive layer such as polycrystalline silicon having conductivity on the base layer 11 and patterning it. This step can be performed, for example, by implanting impurities after film formation of polycrystalline silicon to impart conductivity, and then performing a patterning process such as resist formation or etching. Then, as shown in FIG. 13E, a sacrificial layer 3 made of an insulating film such as PSG (phosphorus doped glass) is formed on the fixed electrode 12. The sacrificial layer 3 can also be formed of a thermal oxide film generated by subjecting the fixed electrode 12 to thermal oxidation.

  Further, as shown in FIG. 13 (f), a second conductive layer made of conductive polycrystalline silicon or the like similar to the above is formed on the sacrificial layer 3, and patterned to form the movable electrode 13 and the resistance layer. R11 is formed simultaneously. Here, the movable electrode 13 includes the movable portion 13a facing the fixed electrode 12 as described above. The resistance layer R11 constitutes the compensation resistor described above, and is made of the same material as the movable electrode 13 (the movable portion 13a thereof) and in the same method (that is, the same film forming process and patterning process). It is formed. Therefore, since the movable portion 13a and the resistance layer R11 are basically formed with the same film formation amount and patterning accuracy, there is a tendency that variations in structural dimensions, for example, thickness, planar length and width, are linked. That is, generally, if a certain structural dimension of the movable portion 13a is larger than the design value, the same structural dimension of the resistance layer R11 is also larger than the design value. In particular, since the thickness of the movable portion 13a and the resistance layer R11 is originally a small dimension, the ratio of the variation in the thickness is much larger than the other structural dimensions, and accordingly, the influence on the resonance frequency is large as described above.

  13D, 13E, and 13F, the base layer 11 is left in the formation region of the compensation resistor R11, and the compensation resistor R11 is formed on the base layer 11. In general, as described above, the base layer 11 functions as an etching stop layer for protecting the lower layer structure from release etching when the storage space 10A is formed. It may be formed only in the region to be formed, and may be removed from the forming region of the compensation resistor R11 which is another region. In this case, the compensation resistor R11 is formed on the insulating film 2.

  Thereafter, as shown in FIG. 13G, interlayer insulating films 4, 6, 8 made of silicon oxide or the like and wiring layers 5, 7 made of aluminum or the like are alternately laminated. Further, although not shown, a structure portion including another resistance layer R11 which is a reference resistance, another resistance layer R13, and a wiring portion connecting them is also formed on the substrate 1, whereby the bias circuit 20 is formed on the substrate 1. It is constructed integrally. In this case, the MEMS resonator 10 and the bias circuit 20 are connected via the wiring layers 5 and 7. Furthermore, it is preferable that the above-described electronic device is integrally formed by also integrally forming the structural portion constituting the signal circuit unit 40 on the substrate 1.

  A protective film 9 made of a resin such as a resist or silicon nitride is formed on the interlayer insulating film 8, and an opening 9 a is formed above the MEMS resonator 10 in the protective film 9. Then, the sacrificial layer 3 and the interlayer insulating films 4, 6, and 8 are removed through the opening 9 a with an etching solution such as buffered hydrofluoric acid or an etching gas, and the accommodation space 10 </ b> A shown in FIG. Is formed (release process). As a result, the movable portion 13a is released to be movable, and the MEMS resonator 10 is operable.

  In addition, when forming the accommodating space 10A ′ in which the resistance layer R11 is disposed as shown in FIG. 8B, the manufacturing method is not different from the above, and the range to be removed in the release step is not limited. What is necessary is just to set so that the formation area of resistance layer R11 may be covered.

  FIG. 14 is a process sectional view showing an example of another manufacturing method different from the above. In this manufacturing method, as shown in FIGS. 14A to 14D, the steps until the insulating film 2, the base layer 11 and the fixed electrode 12 are formed on the substrate 1 are the same as those shown in FIG. .

  Thereafter, as shown in FIG. 14E, an insulating film 3 'such as silicon oxide is formed on the entire surface by CVD or thermal oxidation, and this insulating film 3' is patterned to obtain the structure shown in FIG. As shown, the region where the support of the movable electrode 13 is formed is removed. At this time, the insulating film 3 'remains in the region where the compensation resistor R11 is formed.

  Next, as shown in FIG. 14G, the second conductive layer similar to the above is formed and patterned to form the movable electrode 13 and the resistance layer R11 at the same time. At this time, since the movable portion 13a and the compensation resistor R11 are formed on the same insulating film 3 'as a base layer, respectively, the deposition conditions of the movable portion 13a and the compensation resistor R11 are further equivalent. can do. Therefore, the variation aspect of the film thicknesses of the movable portion 13a and the compensation resistor R11 can be approximated, so that the above-described resonance frequency compensation effect by the compensation resistor R11 can be further enhanced. However, the support portion other than the movable portion 13a of the movable electrode 13 is provided on the base layer 11 in a region where the insulating film 3 'is not formed. Thereafter, similarly to the case of FIG. 13, the upper space structure is formed as shown in FIG. 14 (h), and then the accommodation space 10A is formed by a release process.

  Each embodiment described above is only an example for carrying out the present invention. Therefore, the present invention is not limited to the illustrated example described above. For example, the feature points of the respective embodiments are mutually described. Needless to say, various modifications can be made without departing from the gist of the present invention, for example, as long as they do not contradict each other.

DESCRIPTION OF SYMBOLS 1 ... Board | substrate, 2 ... Insulating film, 10 ... MEMS resonator, 10a ... 1st terminal, 10b ... 2nd terminal, 11 ... Underlayer, 12 ... Fixed electrode, 13 ... Movable electrode, 13a ... Movable part, 20 ... Bias circuit, R11, R15: Compensation resistor, R12, R14: Reference resistor, 30, 30 '... Resonant circuit, 30a ... Input terminal, 30b ... Output terminal, 40 ... Signal circuit unit, Vq ... First potential, Vp ... Second potential, INV1, ... Amplifier circuit having negative resistance

Claims (15)

A MEMS resonator having a substrate, a fixed electrode formed on the substrate, and a movable electrode having a movable portion facing at least a part of the fixed electrode, and a voltage applying means for applying a bias voltage to the MEMS resonator And comprising
The voltage applying means is composed of the same layer as the movable part and has a compensation resistor whose resistance value varies with the thickness of the layer, and a reference resistor connected to the compensation resistor and composed of a layer different from the movable part. And a voltage dividing circuit that outputs a potential at a connection point between the compensation resistor and the reference resistor to at least one of the first terminal and the second terminal of the MEMS resonator. circuit.   The voltage applying means outputs the connection point potential to one of the first terminal and the second terminal of the MEMS resonator and outputs another potential to the other terminal of the MEMS resonator. The resonance circuit according to claim 1, wherein:   In the voltage dividing circuit, the compensation resistor and the reference resistor are connected in series between a high potential and a low potential, and when the connection point potential is higher than the other potential, the compensation resistor and the reference resistor are connected to the compensation resistor side. When the high potential is connected, the low potential is connected to the reference resistance side, and when the connection point potential is lower than the other potential, the low potential is connected to the compensation resistance side and the reference resistance The resonance circuit according to claim 2, wherein the low potential is connected to the side.   The MEMS resonator further includes an input side coupling capacitor connected between the first terminal and the input terminal, and an output side coupling capacitor connected between the second terminal and the output terminal. The resonance circuit according to any one of claims 1 to 3. A MEMS resonator having a substrate, a fixed electrode formed on the substrate, and a movable electrode having a movable portion facing at least a part of the fixed electrode, and a first terminal of the MEMS resonator A first circuit for supplying a potential; and a second circuit for supplying a second potential to a second terminal; corresponding to a difference between the first potential and the second potential in the MEMS resonator Voltage applying means for applying a bias voltage to
At least one of the first circuit and the second circuit is formed of the same layer as the movable part, and a compensation resistor whose resistance value varies with the thickness of the layer, and the movable resistor connected to the compensation resistor. A voltage dividing circuit including a reference resistor formed of a layer different from the unit, and outputting a potential at a connection point between the compensation resistor and the reference resistor as the first potential or the second potential. Resonant circuit.   The movable part and the compensation resistor are made of a conductor layer made of the same material, and the reference resistance is made of a conductor layer made of a material different from the movable part and the compensation resistor. The resonance circuit according to any one of claims 1 to 5.   The resonance circuit according to claim 1, wherein the movable portion and the compensation resistor are formed of adjacent conductor layers.   The resonance circuit according to claim 1, wherein the movable portion and the compensation resistor are accommodated in a common accommodation space. In a method of manufacturing a resonance circuit, comprising: a MEMS resonator having a fixed electrode and a movable electrode having a movable portion facing at least a part of the fixed electrode; and a voltage applying means for applying a bias voltage to the MEMS resonator.
The voltage application unit includes a compensation resistor and a reference resistor, and has a voltage dividing circuit that outputs a potential at a connection point between the compensation resistor and the reference resistor to at least one terminal of the MEMS resonator,
Forming the movable part and forming the compensation resistor in the same layer as the movable part;
Forming the reference resistance in a layer different from the movable part;
A method of manufacturing a resonant circuit comprising:   The movable part and the compensation resistor are composed of a conductor layer formed of the same material and in the same method, and the reference resistance is a conductor layer formed of a material different from or different from the movable part and the compensation resistor. The method for manufacturing a resonance circuit according to claim 9, comprising:   The method of manufacturing a resonance circuit according to claim 9, wherein the movable portion and the compensation resistor are formed of adjacent conductor layers.   The method for manufacturing a resonance circuit according to claim 9, further comprising a step of forming a common accommodation space in which the movable part and the compensation resistor are accommodated.   9. An electronic device comprising: the resonance circuit according to claim 1; and a signal circuit unit connected to the resonance circuit formed on the same substrate.   The electronic device according to claim 13, wherein the movable portion and the compensation resistor are disposed together in a housing space formed on the substrate.   15. The electronic device according to claim 13, wherein the signal circuit unit includes an amplifier circuit having a negative resistance and constitutes an oscillation circuit.

Images