JP2013114935A - Mems switch - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem in which, for a MEMS switch, increasing a spring constant of "a spring" for supporting a movable structure is effective means for enhancing its characteristics, especially for enhancing long-term reliability and a switching speed, but the increased spring constant causes a voltage required for switch driving to increase, making it difficult to integrate it with a semiconductor circuit, and it has then been required to reduce the driving voltage while enhancing the MEMS switch characteristics.SOLUTION: A MEMS switch includes a movable structure. Use of mechanical vibration of the movable structure allows a large vibration amplitude to be obtained even with a low drive voltage and facilitates induction of pull-in. Control of a DC voltage and an AC voltage for exciting vibration allow control of the pull-in and release, to achieve make and break operations of the MEMS switch. As a result, even with a spring constant being increased, the drive voltage can be reduced and switch characteristics can be enhanced.

Description

  The present invention relates to a configuration and a driving method for improving characteristics of a MEMS switch. The present invention also relates to a MEMS device that can reduce the driving voltage using this driving method.

  2. Description of the Related Art In recent years, technological advances in semiconductor devices have been great, and they have been used over a wide range such as industrial equipment and consumer equipment. In particular, miniaturization and high integration of semiconductor devices have greatly contributed to downsizing, weight reduction, price reduction, and high functionality of devices and systems equipped with semiconductor devices. However, while miniaturization and high integration are achieved, an increase in the scale and complexity of the manufacturing process and an increase in the scale and cost of the manufacturing apparatus have been induced. Furthermore, the limits of miniaturization are discussed, and development directions other than miniaturization are being sought. A MEMS (microelectromechanical system) device is one example, and is characterized in that new functions can be realized by miniaturizing mechanical elements and adapting them to semiconductor technology. In addition, new technologies and new knowledge obtained through the technological development of MEMS devices are reflected in conventional semiconductor processes and contribute greatly to the advancement of the semiconductor field.

  One such MEMS device is a device called “RF-MEMS”. This is a device that controls a radio frequency (RF) band signal, and is expected to be applied to the communication field such as signal switching (switching), filters, and oscillators. In the case where the RF-MEMS switch is applied to the “signal switching”, the RF-MEMS switch is arranged in a specified region of the high-frequency signal line. As an example of the RF-MEMS switch, a fixed electrode (generally disposed beside the high-frequency signal line) is disposed close to a minute movable structure, and the movable structure and the fixed electrode are disposed. An electric signal (generally a DC voltage) is applied between the movable structure and the movable structure is mechanically deformed, and the distance between the movable structure and the high-frequency signal line is reduced (electrically or mechanically). Including general contact). As a result of such a decrease in the distance, the impedance of the high-frequency signal line is changed, and the characteristics of the high-frequency signal transmitted through the high-frequency signal line are greatly changed, thereby realizing the switch function. Many semiconductor processes can be used for manufacturing such a structure with mechanical deformation, and various RF-MEMS devices have been proposed.

  When constructing a MEMS device, there are roughly two approaches. One is “bulk micromachining” in which fine processing is performed on single crystal silicon, and the other is “surface micromachining” using sacrificial layer etching.

  In bulk micromachining, a single crystal silicon wafer back surface or front surface (or back surface and front surface) is processed using a known means such as etching to produce a three-dimensional structure. Applications to diaphragm type pressure sensors and acceleration sensors have been put into practical use. However, it is also known that there is a limit to realizing a more complicated three-dimensional structure.

  In surface micromachining, a three-dimensional structure is produced by processing means that combines thin film deposition, patterning, and removal of layers between thin films (sacrificial layer etching). In many cases, a thin film of a silicon layer is deposited and the structure is deformed or vibrated, and a metal thin film is sometimes used. Although surface micromachining is easy to divert a general semiconductor production line, it has been pointed out that the mechanical properties of a thin film greatly depend on production conditions.

  FIG. 13 shows a cross-sectional structure of a parallel-connected MEMS switch manufactured by surface micromachining. This figure is described in Non-Patent Document 1 below. In the figure, reference numeral 199 denotes a movable structure made of gold, which is mechanically supported by flexible suspensions 196a and 196b and support bases 195a and 195b. 195 a and 195 b are bonded to the glass substrate 190. A contact 198 and bumps 197a and 197b are formed on one surface of the movable structure 199 (the lower surface in the figure). Further, fixed electrodes 193a and 193b, bump contact portions 194a and 194b, and a coplanar waveguide (which is a high-frequency signal line) are formed on the upper surface of the glass substrate 190. The coplanar waveguide is composed of a signal line 191 and grounds 192a and 192b arranged on both sides thereof.

  FIG. 14 is a cross-sectional structure diagram in the case where a DC voltage is applied between the fixed electrodes 193a and 193b and the movable structure 199 from the outside. The movable structure 199 is drawn in the direction of the fixed electrodes 193a and 193b by the electrostatic force due to the DC voltage. As a result, the contact 198 constituting the movable structure 199 contacts the signal line 191, and the bumps 197a and 197b contact the bump contact portions 194a and 194b, respectively. In this state, the high frequency signal transmitted through the coplanar waveguide is transmitted to the contact 198 side without being transmitted in a direction perpendicular to the paper surface. As a result, the high frequency signal is cut off (the switch is “off”). When the DC voltage applied from the outside is reduced to zero, the movable structure 199 returns to the state shown in FIG. 13 by the restoring force of the “spring” of the suspensions 196a and 196b. In the state of FIG. 13, since the contact 198 is located away from the signal line 191, a high frequency signal transmitted through the coplanar waveguide is transmitted through this switch without being transmitted to the contact side (the switch is “ON”).

MEMS switches are characterized by low insertion loss and high isolation compared to semiconductor switches. In addition, the electrostatic drive type switch has low power consumption and a simple structure. However, the conventional switches shown in FIGS. 13 and 14 have the following problems.
(1) When the “spring” constant of the suspension is reduced, the voltage required for driving the switch (hereinafter also referred to as “drive voltage”) can be reduced. However, since the restoring force of a weak suspension is small, it is impossible to resist the adhesive force generated when the contact point contacts the signal line when the external DC voltage is reduced to zero. For this reason, the operation from the switch “off” to the switch “on” cannot be performed, and the switch does not function as a MEMS switch.
(2) On the other hand, it is also possible to reduce the drive voltage by reducing the distance between the movable structure and the fixed electrode. However, even in this case, since the displacement of the suspension is small, it is difficult to obtain a sufficiently large restoring force of the “spring”, and the same adhesion problem as in (1) occurs.
(3) When the switch is designed with a suspension having a large “spring” constant in order to solve the problems (1) and (2), the drive voltage obviously increases.
(4) When the design of (2) is performed, there is a further problem that the high frequency characteristics of the switch deteriorate. This is because the distance between the movable structure and the signal line is reduced, so that the capacitance between the two increases and when the switch is turned on (191 and 198 in FIG. 13 are separated). This is because the high-frequency signal “runs away” through the contact 198. For this reason, it is difficult to handle a signal having a high frequency with such a switch.
(5) To increase the switching speed, it is necessary to increase the “spring” constant, but this increases the drive voltage of the switch.
(6) When the power of the high-frequency signal transmitted by the switch is large, the drive voltage increases. This is because it is essential to design a large restoring force of the suspension so as to resist an electrostatic force generated by a high-power high-frequency signal. As a result, the drive voltage increases.
(7) When the movable structure and the fixed electrode come into contact with each other, the power supply that supplies a DC voltage from the outside is short-circuited, and the MEMS switch is destroyed. In order to avoid this, it is often used to provide an insulating film on the surface of the fixed electrode to prevent this short circuit. However, when the operation of the MEMS switch is repeated, the insulating film is charged, and the movable structure and the fixed electrode are not separated even when the application of the DC voltage is stopped. The charging phenomenon of the insulating film reduces the durability (long-term reliability) of the MEMS switch, leading to deterioration of characteristics.

  As described in the previous paragraph, although the MEMS switch has advantages, the switch drive voltage is reduced, the switch characteristics are improved (the number of operation repetitions (lifetime) that can be used against adhesion is increased, and the high frequency characteristics are improved. Improvement, higher switching speed, higher power, etc.). In particular, the drive voltage of a MEMS switch that has been released or announced is usually 25 volts or more, which is greatly different from the operating voltage (1 to 2 volts) of a semiconductor circuit, which is a major obstacle to integrating the two. Yes. That is, reducing the driving voltage has been an important solution.

  As one method for solving such a problem, a MEMS switch having a structure and driving method different from those in FIG. 13 has been announced. This switch is described in Non-Patent Document 2 below, and its structural diagram is shown in FIG. FIG. 4A is a diagram showing the structure of a MEMS switch, and FIG. 4B is a diagram showing an example of an application circuit to an amplifier. As shown in FIG. 6A, the circular diaphragm arranged at the center is supported by a suspension named “Support Beam” and floats from the substrate. Around the diaphragm, two sets of fixed electrodes (fixed electrodes) are arranged so as to surround the diaphragm. That is, a structure that is originally used as a MEMS resonator is used. When a high frequency signal is applied to the fixed electrode arranged in the vertical direction (in FIG. 5B, it is switched to the “left-right direction”), the shape of the diaphragm changes from a circular shape to an elliptical shape, and the vertical direction and the horizontal direction Telescopic motion (also called wine glass mode) occurs. Furthermore, when a high frequency signal having the same frequency as the resonance frequency of the diaphragm is applied, the diaphragm and the fixed electrode periodically repeat “contact” and “non-contact” due to a large amplitude at the time of resonance of the diaphragm. Become. With such “contact” and “non-contact” operations, a certain type of switch function is realized. A switch using this resonance has a feature that the switch can be driven with a very low driving voltage even though a large “spring” constant can be obtained. In the circuit example shown in FIG. 5B, the output from the DC power source expressed as “VDD” is switched by the high-frequency signal expressed as “Vi”. The signal from this switched DC voltage (resulting in a high frequency signal) is the output of the amplifier. The illustrated configuration is also referred to as a “Class E” amplifier and is used in some high frequency circuits. However, the application target of the “class E” amplifier is limited, and there is a drawback that it cannot be applied to a more general coplanar waveguide.

  Furthermore, a switch using resonance has a limitation that the switch is limited to operate in a narrow frequency band very close to the resonance frequency. For example, in Non-Patent Document 2 below, a switch having a resonance frequency of 60 MHz is prototyped, but the passband is only 3 kilohertz. In addition, this switch can perform high-speed switching with a period of 17 nanoseconds, but cannot perform switching with other periods, and therefore cannot transmit a signal. This is because this switch is constantly vibrating. For this reason, the application of this switch is narrowly limited.

  As described in the previous paragraph, when the “spring” constant is decreased in order to reduce the drive voltage of the MEMS switch, there is a problem that the switch characteristics are deteriorated. However, reduction of the driving voltage is an essential requirement for integration with a semiconductor circuit. In addition, in order to widen the application range of the switch, there should be no narrow limitation on the drive frequency band of the switch. As described above, in order to meet the great expectation for practical use of MEMS switches, reduction of drive voltage and improvement of switch characteristics have been important issues.

Conference presentation paper Yamamoto et al., “Three-dimensional vibration characteristics of single-crystal silicon MEMS resonators”, Proceedings of the 27th “Sensor Micromachine and Application System” Symposium (CD version), pages 265-269, October 2010 Conference presentation paper Lin et al., “A Resonance Dynamic Approach to Faster, More Reliable Micromechanical Switches”, IEEE Frequency Control Symposium (FCR) Proceedings, 640-645.

  In order to achieve excellent switching characteristics in a MEMS switch, it is essential to increase the “spring” constant to satisfy the requirements for high speed, high power, and long-term reliability. However, in general, there is a drawback that the driving voltage is remarkably increased. For this reason, how to reduce the drive voltage of a switch having a large spring constant is an important problem to be solved. Further, since the deterioration of durability and reliability due to the charging phenomenon deteriorates the characteristics as a switch, it is also a problem to be solved to avoid this phenomenon. Furthermore, since the application of the switch is not limited, it is a problem to realize a MEMS switch for high frequency signals in a wide frequency band.

  In a MEMS switch having a movable structure and a fixed electrode, (1) a signal in which a designated first AC voltage and a designated first DC voltage are superimposed is supplied, and (2) the movable structure is A make operation is induced by mechanical vibration; (3) the first AC voltage is changed to a specified second AC voltage; and the first DC voltage is changed to a specified second AC voltage. The make operation is maintained by a signal changed to a DC voltage. (4) The second AC voltage is changed to a designated third AC voltage, and the second DC voltage is designated. Break operation is induced by the signal changed to the three DC voltages.

The terms used in the present invention will be described. Up to the previous paragraph, the term “switch on / off” has been used. In order to avoid confusion in the interpretation of the term, the following definition will be used.
Make and break: In general, in the “relay”, the state where the built-in contact is closed (contacted) is called “make”, and the state where the built-in contact is opened (not touched) is called “break”. To do. The “make operation” is an operation in which the contact of the relay is closed, and the “break operation” is an operation in which the contact of the relay is not in contact. The “make-up state” is a state in which the node is kept closed after the make-up operation. The “break state” is a state in which a node is kept open after a break operation. That is, “make” and “break” express the physical operation of the switch contacts.
ON and OFF: Expresses the function of the circuit in which the switch is mounted. The state in which this circuit is conducting is “on”, and the non-conducting state is “off”. As described in detail below, in a switch arranged in “parallel” in a high-frequency distributed constant circuit (eg, a coplanar waveguide), the contact of the switch is “ When in the “break state”, the switch is “on”. On the other hand, in order to prevent the high-frequency signal from propagating, the switch is “off” when the contact of the switch is in the “make state”.
Pull-in: When the “movable structure”, which is a component of the MEMS switch that is the subject of the present invention, approaches the “fixed electrode” and falls below a specific distance, it is physically “pulled” to the fixed electrode side. It is. At this time, the MEMS switch performs a “make operation” and shifts to a “make state”.
Hold: The “pull-in” state continues. At this time, the relay is in the “make state”.
Release: Transition from the “pull-in” state to a state in which the distance between the movable structure and the fixed electrode is greatly separated. At this time, the MEMS switch performs a “break operation” and shifts to a “break state”.
Further, the MEMS switch can be regarded as a synonym for “relay” which is widely used. There are many types of relays, but electromagnetic relays are often used. In this relay, an iron piece (called an armature) is attracted by the electromagnetic force (magnetic field) generated by an electromagnet, and the contact provided at one end of the iron piece makes mechanical contact with another fixed contact. The operation principle is that these two contacts are electrically connected. On the other hand, in the MEMS switch, a voltage is applied from the outside between the movable structure and the fixed electrode, and the generated electrostatic force (corresponding to electromagnetic force) attracts the “movable structure” (corresponding to the armature). Contact points are arranged on the “movable structure” and mechanically come into contact with other fixed contacts, and these two contacts are electrically connected (“make”). The present invention is characterized by a method of displacing the movable structure to shorten the distance from the fixed electrode, and a method of returning the displacement to the original state.

  In addition, “inducing a make operation by mechanically vibrating the movable structure” described in the preceding paragraph means that the movable structure mechanically vibrates at a frequency of the AC voltage, and the MEMS switch Indicates that the makeup state is reached. Here, the “mechanical vibration” includes a large-amplitude vibration at the time of resonance. Similarly, “inducing a break operation” indicates that the MEMS switch shifts from a make state to a break state. In addition, in the expression such as “change the first AC voltage to the designated second AC voltage”, “change” includes the following cases. That is, the voltage value of the first AC voltage and the designated second AC voltage are equal and actually “not changed”. Even in such a case, it is expressed as “change” formally from the first AC voltage (for example, 1 volt) to the designated second AC voltage (1 volt). Further, “superimposition” represents adding a DC voltage to an AC voltage (equivalent to applying a DC bias) or adding an AC voltage to a DC voltage.

An example of driving the above-described MEMS switch is shown below.
First AC voltage = 0.6 volts pp (Peak-to-Peak)
First DC voltage = 5.5 volts Second AC voltage = 0 volts pp (Peak-to-Peak)
Second DC voltage = 5.5 volts Third AC voltage = 0 volts pp (Peak-to-Peak)
Third DC voltage = 4.0 volts Under this condition,
A “pull-in” occurs at the first AC voltage and the first DC voltage, and a “make operation” occurs at the same time.
A transition is made to the “hold” state at the second AC voltage and the second DC voltage, and the “make state” continues.
A “release” occurs at the third AC voltage and the third DC voltage, and a “break operation” occurs at the same time.

  In addition, although the said movable structure and the said fixed electrode are formed by the same manufacturing process, it is not restricted to this. Moreover, although the said movable structure and the said fixed electrode are comprised with the same material, it is not restricted to this. When the movable structure and the fixed electrode are not made of the same material, they are manufactured by different manufacturing processes.

  In addition, the constituent material of the said movable structure and the said fixed electrode is not necessarily a metal. For example, a semiconductor such as silicon, a semiconductor containing silicon as a component, a material that imparts conductivity to the surface of an insulator such as resin or resin, a hybrid material that combines various materials, or the like may be used. The driving force for “vibrating the movable structure mechanically” is not limited to electrostatic force, and may be electromagnetic force, piezoelectric strain, thermal strain, or the like. For example, a piezoelectric material is used for the movable structure or a part of the movable structure (for example, the surface of the movable structure), and mechanical distortion is generated by a DC voltage and an AC voltage supplied from the outside, and the movable It is to vibrate the structure. Furthermore, the locus of displacement of the movable structure due to electrostatic force is not necessarily “substantially straight” (in a strict sense, it is an arc when one end of the movable structure is fixed). For example, it may be a straight line or a zigzag.

  Note that the “make state” means that the above-described two contact points are in an electrically conductive state, and is thus independent of the distance between the movable structure and the fixed electrode. Of course, this distance is smaller in the “make state” than in the “break state”, but this distance is never “zero”. Here, “zero” means that the movable structure and the fixed electrode are in direct contact (electrical short circuit). If it is “zero”, both are electrically short-circuited, and therefore it is not preferable because it is necessary to pay special attention to the power supply system that supplies the AC voltage and the DC voltage. In the present invention, even in the “make state”, the movable electrode and the fixed electrode are “in an electrically insulated state”. That is, in the configuration example in which the surface of the fixed electrode or the surface of the movable structure (the surface facing the fixed electrode) is covered with an insulating film, the movable structure is connected to the surface of the fixed electrode and the machine. Even if they are in contact with each other (the mechanical distance is zero), they are not electrically “zero”, and both are electrically insulated. Such a configuration example is more preferable because it does not require special consideration to the power supply system as described above. However, when the surface of the movable structure or the fixed electrode is covered with an insulating film, the insulating film is charged and the characteristics of the MEMS switch are deteriorated while the make and break operations of the MEMS switch are repeated many times. For example, it must be considered to induce a shorter operating life.

  It is possible to eliminate the “covering with an insulating film” exemplified in the previous paragraph. Specifically, a “stopper” is incorporated into the MEMS switch to limit the amount of displacement of the movable structure. For example, there is a configuration in which a taller stopper is disposed in the vicinity of the fixed electrode, and the movable structure is stopped at this position. The position where the stopper is disposed is (1) the lower surface of the movable structure (the surface facing the fixed electrode), the region facing the region where the fixed electrode is disposed, (2) A lower surface of the movable structure (a surface facing the fixed electrode), a region facing a region where the fixed electrode is not disposed, (3) a surface of the fixed electrode, (4) the fixed electrode Is a surface where the fixed electrode is not disposed. In the configurations of (1) and (3), the stopper and the fixed electrode are in mechanical contact with each other in the makeup state. For this reason, it is essential that at least the surfaces of the stopper or the fixed electrode, or the stopper and the fixed electrode be electrically insulating. However, considering that the purpose is to eliminate “covering with an insulating film”, (2) or (4) is a preferable configuration example. In the configuration example of (2) or (4), even if the stopper is made of a conductive material, the stopper, the movable structure of the MEMS switch, and the fixed electrode are separated in a circuit (even if both are mechanically contacted). A configuration that is not electrically conductive, for example, a configuration in which the stopper is in a floating state may be employed.

  Note that the surface of the “stopper” is not necessarily flat. For example, a configuration may be adopted in which a “dent” is provided on the surface, and a peripheral region of the “dent” has a stopper function. Further, a convex portion (bump shape) may be provided in the central region of the surface of the stopper, and the convex portion may have a stopper function.

  As an alternative to incorporating the “stopper” described in the previous paragraph, the “position detection mechanism” of the movable structure is incorporated in the MEMS switch, and the displacement amount of the movable structure is controlled by electrical or mechanical feedback control. It is also possible to regulate the maximum value of. In the above configuration example in which “covering with an insulating film” is excluded, there is an advantage that characteristic deterioration of the MEMS switch (for example, the operating life is shortened) due to charging of the insulating film can be prevented.

  The number of fixed electrodes described above is not limited to one. Two or more fixed electrodes may be arranged.

  Note that the electrostatic force for displacing the movable structure is not limited to the electrostatic force generated between the movable structure and the fixed electrode that are arranged to face each other. Can be adopted. For example, it is possible to use a mechanism of an electrostatic actuator in which one set of comb-shaped electrode groups is combined.

  The first AC voltage that vibrates the movable structure is set to a frequency within a frequency range that does not exceed plus 20 percent and minus 20 percent around the resonance frequency of the movable structure.

  When the movable structure is vibrated at a resonance frequency, the displacement of the movable structure can be increased. However, since it is difficult to control a vibrating body that is resonating, there is a disadvantage that the displacement amount of the movable structure is reduced by driving at a resonance frequency “near”, but easy control is ensured. There are advantages you can do. Further, if the vibration characteristic (Q value) of the movable structure is suppressed (Q value is reduced), even if the movable structure is vibrated at a frequency away from the resonance frequency, the displacement amount is reduced. Can be suppressed. Such suppression of the Q value can be realized by operating the MEMS switch in an atmosphere of “weak vacuum” (for example, 0.1 atm), or configuring the movable structure with a damping material. Furthermore, dimensional errors that occur during the manufacturing process of the MEMS switch cannot be avoided, and therefore, the resonance frequency varies among individual MEMS switches. In order to suppress this influence, using a frequency within a certain range (for example, plus or minus 20 percent) has an advantage that the power supply system of the MEMS switch can be easily designed.

  In the “make state”, the frequency of the AC voltage may be set to a frequency different from the resonance frequency. For example, a frequency near the resonance frequency is used for the “make operation”, and a frequency different from this frequency is used for the “make state” (also “hold” state).

  (1) By supplying the first AC voltage at a designated first time, and (2) superimposing the first DC voltage on the first AC voltage at a designated second time. Inducing a make operation, (3) changing the first AC voltage to the third AC voltage at a designated third time, and changing the first DC voltage to the third DC voltage A break operation is induced by changing.

An example of driving the MEMS switch described in the previous paragraph is shown below.
First time: AC voltage = 0.6 volts pp (Peak-to-Peak)
Second time: AC voltage = 0.6 volts pp, DC voltage = 5.5 volts Third time: AC voltage = 0.6 volts pp, DC voltage = 4.0 volts In this example, the AC voltage is always The MEMS switch is controlled only by controlling the DC voltage value. In the above example, “pull-in” occurs at the second time and “make operation” occurs simultaneously. “Release” occurs at the third time and “break operation” occurs at the same time. Further, the period from the first time to the second time is a standby state (the “release state” and the “break state”). The period from the second time to the third time is the “hold” state. In the example described above, the first AC voltage or the second AC voltage is all 0.6 volts pp. Further, when viewed from the control system of the system on which the MEMS switch is mounted, the second time is the generation time of the “on” signal, and the third time is the generation time of the “off” signal.

  (1) supplying the first DC voltage at a designated first time, and (2) superimposing the first AC voltage on the first DC voltage at a designated second time. Inducing a make operation, (3) changing the first AC voltage to the third AC voltage at a designated third time, and changing the first DC voltage to the third DC voltage A break operation is induced by changing.

An example of driving the MEMS switch described in the previous paragraph is shown below.
First time: DC voltage = 5.5 volts Second time: DC voltage = 5.5 volts, AC voltage = 0.6 volts pp
Third time: DC voltage = 4.0 volts, AC voltage = 0 volts pp
In this example, both AC voltage and DC voltage are controlled. In the above example, “pull-in” occurs at the second time and “make operation” occurs simultaneously. “Release” occurs at the third time and “break operation” occurs at the same time. Further, the period from the first time to the second time is a standby state (the “release state” and the “break state”). The period from the second time to the third time is the “hold” state. Further, when viewed from the control system of the system on which the MEMS switch is mounted, the second time is the generation time of the “on” signal, and the third time is the generation time of the “off” signal.

  There are many variations on the set of voltage values illustrated in the previous paragraph. For example, immediately after the “pull-in” occurs at the second time, the first AC voltage (which is 0.6 volts pp) may be reduced (may be 0 volts) to enter the “hold” state. . Also, immediately after the “release” occurs at the third time, the third DC voltage (4.0 volts) is increased to the first DC voltage (5.5 volts) The same condition as the first time) may wait for a “next ON” signal from the system control system.

  (1) A signal in which a designated first AC voltage and a designated first DC voltage are superimposed is supplied, and (2) a make operation is induced by mechanically vibrating the movable structure. (3) A break operation is induced by increasing or decreasing the frequency of the first AC voltage.

  Supply a signal in which the designated first AC voltage and the designated first DC voltage are superimposed, and increase the frequency of the first AC voltage until it exceeds the resonance frequency of the movable structure. The makeup operation is induced by reducing the frequency of the first AC voltage.

  Note that the method of inducing the make operation described in the previous paragraph uses the non-linear characteristics unique to the MEMS device.

  A plurality of frequency components in a frequency range including the resonance frequency are included in the first AC voltage.

  In the description in the previous paragraph, the “plurality of frequency components” may be included at the same time. That is, when the AC voltage is observed at a specific time, a plurality of frequency components are included. The “plurality of frequency components” may appear in time series. That is, when the AC voltage is observed in a specific time width, a plurality of frequency components appear one after another (the frequency changes on the time axis). In this case, the frequency change may increase sequentially, decrease sequentially, or may be random.

  According to the present invention, a MEMS switch characterized by utilizing the vibration of the movable structure can be realized. According to the present invention, it is possible to significantly reduce the driving voltage without changing the structure that causes the characteristic deterioration of the MEMS switch, and the contribution to the advancement of the MEMS switch is great.

  According to the present invention, the MEMS switch can be operated at a low driving voltage, and a movable structure supported by a suspension having a large “spring” constant as compared with the conventional MEMS switch can be used. became. For this reason, it is possible to prevent the “burn-in” of the switch due to the adhesion effect of the contact, to realize a long life (the limit of the number of repetitions of the switch has been extended), and to provide a MEMS switch with excellent long-term reliability. became.

  A movable structure supported by a suspension having a large “spring” constant can be switched “on” and “off” even when a high power signal is transmitted on the signal line. For this reason, the MEMS switch of the present invention can be used in a system for high power applications.

  Since the movable structure supported by the suspension having a large “spring” constant operates at high speed, the switching time can be shortened and the speed of the switch can be increased.

  By reducing the drive voltage, it is possible to integrate the control circuit system of the MEMS switch and the circuit system of the system on which the MEMS switch is mounted on the same semiconductor substrate.

  In the MEMS switch according to the present invention, reduction of driving voltage, improvement of electrical characteristics and mechanical characteristics can be realized by utilizing mechanical vibration. Therefore, the present invention can be widely applied not only to the MEMS switch but also to the actuator field where there is a great demand for reducing the drive voltage. For example, if it is applied to a capacitance type MEMS sensor (acceleration sensor, gyro sensor, etc.), a narrow gap can be realized, so that a large capacitance change value is output, and high sensitivity of the sensor can also be realized. .

It is a figure explaining the basic composition and the drive method of a MEMS switch. <Example 1> It is a figure which shows the drive principle of a MEMS switch. It is a figure which shows the operation | movement sequence of a MEMS switch. <Example 2> It is a figure which shows the structure of a MEMS switch. <Example 3> It is a figure which shows the preparation methods of a MEMS switch. <Example 4> It is a figure which shows the drive operation | movement range of a MEMS switch. <Example 5> It is a figure which shows the other drive method of a MEMS switch. <Example 6> It is a figure which shows the other structure of a MEMS switch. <Example 7> It is a figure which shows the other structure of a MEMS switch. <Example 8> It is a figure which shows the other structure of a MEMS switch. <Example 9> It is a figure which shows the other structure of a MEMS switch. <Example 10> It is a figure which shows the other structure of a MEMS switch. <Example 11> It is a figure which shows the cross section of the MEMS switch produced by the surface micromachining. <Conventional example 1> It is a figure explaining operation | movement of the switch of <conventional example 1>. It is a structure and circuit example of a MEMS switch using resonance. <Conventional example 2>

  Hereinafter, the MEMS switch will be described in detail with reference to the drawings.

<Basic configuration of MEMS switch using resonance>
FIG. 1 is a diagram illustrating a basic configuration and a driving method of a MEMS switch that is Embodiment 1 of the present invention. The MEMS switch 13 is arranged in parallel between a terminal 11 for inputting a high-frequency signal and a terminal 12 for outputting the high-frequency signal, and the MEMS switch short-circuits or opens the high-frequency transmission lines 17a and 17b. . In this example, an electrical circuit in which an AC voltage source 14 and a DC voltage source 15 are connected in series is used as a drive source for the switch 13, and the make / break operation of the switch 13 is controlled by the drive switch 16. In this example, the switch is provided in parallel with the high-frequency transmission line. However, the present invention is not limited to this configuration. For example, MEMS switches are arranged in series with the transmission line (that is, the switch in the middle of 17a or 17b). It is also possible to insert).

  FIG. 2A is a diagram for explaining the basic configuration of the MEMS switch. In FIG. 2A, one end of the movable structure 21 is fixed by a support 22 and is joined to the surface of the glass substrate 20. A fixed electrode 23 is disposed on the upper surface of the glass substrate 20. Further, a contact 24 is provided on the lower surface of the movable structure 21, while a signal line 25 (drawn like a contact) is provided at a position facing the contact 24 on the surface of the glass substrate 20. ing. In such a configuration, an AC voltage source or a DC voltage source as a switch drive source is connected between the movable structure 21 and the fixed electrode 23. An electric signal composed of an AC voltage and a DC voltage from the driving source generates an electrostatic force that brings the movable structure 21 close to the fixed electrode 23. If the electrostatic force is sufficiently large, the movable structure 21 is greatly displaced and is extremely close to or in contact with the fixed electrode 23. In this state, the contact 24 contacts the signal line 25 mechanically and electrically, and the switch performs a “make” operation. This state corresponds to the “switch-off (signal cutoff)” state described in the conventional example of FIG.

  FIG. 2B shows a driving method of the MEMS switch of the present invention in comparison with a conventional driving method. In the figure, the applied voltage is plotted on the horizontal axis, and the distance (gap length) between the movable structure of the switch and the fixed electrode is plotted on the vertical axis. First, a conventional MEMS switch driving method (applying only DC voltage) will be described. In FIG. 6B, as shown by a curve 27, when the voltage applied to the MEMS switch is increased, the movable structure is attracted to the fixed electrode side by electrostatic attraction, and the distance between the two (gap length). Decrease. When the distance becomes about 2/3 of the initial value of the gap length (the value when the DC voltage is 0 volt), the movable electrode suddenly approaches a distance very close to the fixed electrode ("pull-in"). is there). The applied voltage at this time is called a pull-in voltage. Thereafter, even if the applied voltage is increased, the change in the distance between the two is extremely small. A curve 28 in FIG. 5B shows a change in the distance between the two when the applied voltage is decreased in the “pull-in” state. When the applied voltage is reduced, the change in the distance between the two exceeds the pull-in voltage for a while. However, when the applied voltage becomes smaller than a specific voltage, the movable structure suddenly moves away from the fixed electrode ("release"). The applied voltage at this time is called a release voltage. The difference between the pull-in voltage and the release voltage is usually a large value exceeding 15 volts. For this reason, generally, a large voltage exceeding 30 volts is required for the pull-in voltage.

  Next, a method for driving the MEMS switch according to the present invention will be described with reference to FIG. In the present invention, the vibration phenomenon of the movable structure is used to “pull in” the MEMS switch. For this reason, the drive source shown in FIG. 1 is used. That is, an electrical signal in which an AC voltage is superimposed on a DC voltage is applied to the MEMS switch. When the frequency of the AC voltage is set to “near” the mechanical resonance frequency of the movable structure, the movable structure vibrates with a large amplitude. As shown by a curve 29 in FIG. 5B, when the state oscillates to about ½ of the initial value of the gap, it is drawn into the electrostatic force of the DC voltage that is the component of the applied electric signal. "Pull-in" occurs. It is important that the value of the pull-in voltage in this driving method is lower than the pull-in voltage by the conventional driving method (see curve 27). Further, in the “pull-in” state, the AC voltage is set to zero, and the “pull-in” state is maintained only by applying the DC voltage. Stopping the application of the AC voltage can reduce the power consumption in the drive system of the MEMS switch, and further, unnecessary radiation of the AC voltage to the circuit system arranged around the MEMS switch. There is also an advantage that can be suppressed. In this paragraph, it is described that “the frequency of the AC voltage is set to“ near ”the mechanical resonance frequency of the movable structure”, but this means that the frequency of the AC voltage exactly matches the resonance frequency. It is not limited. For example, a frequency in the range of 20% before and after the resonance frequency may be used as the AC voltage. The range also depends on the vibration damping coefficient of the movable structure. The point is that the vibration amplitude of the movable structure increases with respect to the applied AC voltage component, which can contribute to the reduction of the pull-in voltage.

  Even in the driving method described in the previous paragraph, “release” occurs when the applied DC voltage is decreased in a “pull-in” state and becomes smaller than a specific voltage. The release voltage at this time is the same as the release voltage in the conventional driving method.

  As is clear from the description up to the previous paragraph, in order to keep the “pull-in” state (keep the movable structure close to the fixed electrode or keep the “make-up state”), the release voltage It is necessary to continue to apply a large DC voltage. The voltage that maintains this “pull-in” state is called the hold voltage. The hold voltage is set larger than the release voltage, preferably smaller than the pull-in voltage.

<Operation of MEMS switch using resonance>
In the description up to the previous paragraph, the driving mode of the MEMS switch according to the present invention has been clarified. There are various driving procedures for realizing the operation sequence required for the switch. The “operation sequence” is an order of switch make and break states. Here, the “make state” corresponds to the “pull-in” state, and the “break state” corresponds to the “release” state. In this paragraph, an example of the sequence will be described as Example 2 with reference to FIG. FIG. 3A is a list of the pull-in, hold, and release states of the MEMS switch corresponding to the make state and break state of the switch, and the values of the DC voltage and the AC voltage applied in each state. is there. Here, in the driving method using the vibration of the movable structure, the pull-in voltage is 5.5 volts, the release voltage is 4.0 volts, and the pull-in voltage in the conventional driving method is 30 volts. Assumes. Symbols A to E described in the “operating point” column of FIG. 3A are also displayed as operating points in the drawing of FIG.
First time (initial state): “First time” is the time when power is turned on to the circuit system on which the MEMS switch is mounted. The operating point at this time is A, no electric signal for driving is applied, and the MEMS switch is in the released state.
Second time (“make” operation): Corresponds to a state in which the contacts at the reed relay are connected. When a 6.0 volt DC voltage (denoted as a DC voltage) superimposed with an alternating voltage (denoted as an AC voltage, for example, 0.6 pp as a voltage value) is applied, a “pull-in” is immediately induced. The The operating point is B. In this case, the first AC voltage is 0.6 volts pp, and the first DC voltage is 6.0 volts.
“Make” state: After reaching the “pull-in” state, no AC voltage is applied and only a DC voltage of 6.0 volts is applied. As is clear from the above examples of the voltage values, the MEMS switch shifts to the “hold” state, but the “make state” is maintained. The operating point is C. In this case, the second AC voltage is 0 volts pp and the second DC voltage is 6.0 volts. Note that the “hold” state is maintained even if the DC voltage is not 6.0 volts but is a voltage higher than the release voltage (for example, 4.5 volts). At this time, the second DC voltage is 4.5 volts.
Third time ("break" operation): corresponds to a state in which the contact at the reed relay is separated. When the DC voltage is reduced to 3.0 volts (below the release voltage), the MEMS switch is immediately "released". The operating point is D. In this case, the third AC voltage is 0 volts and the third DC voltage is 3.0 volts.
“Break” state: Upon reaching the “release” state, a DC voltage of 6.0 volts is applied for the next “make” operation. However, the AC voltage is not applied, and since it is below the pull-in voltage (30 volts) in the conventional driving method, the “release” state continues. The operating point is E.
Second time for the second time (“make” operation): Since the AC voltage is newly applied in the previous state, the movable structure starts to vibrate and satisfies the “pull-in” condition. Immediately, “pull-in” is induced and “make-up state” is entered. The operating point at this time is B.
With the above operation, the MEMS switch can repeat a predetermined operation. Note that the sequence, DC voltage, and AC voltage values illustrated in this paragraph are merely examples for explaining the operation, and are not limited thereto.

  In this paragraph, the “AC voltage” is described. The present invention is characterized in that the movable structure is mechanically vibrated with the AC voltage. In order to efficiently generate such mechanical vibrations, it is preferable to supply the resonance frequency (primary resonance frequency is preferable, but not limited) of the movable structure. However, since a processing error is unavoidable in the manufacturing process of the MEMS switch, the size (for example, the length and width) of the movable structure is different for each MEMS switch. As a result, the resonance frequency of the movable structure is different for each MEMS switch. For this reason, the drive system of the MEMS switch is simplified by using a frequency within a specific width (for example, plus 20 percent, minus 20 percent) centered on the resonance frequency. For example, when the resonance frequency of the movable structure is 13 kHz (kHz), the frequency between 10.4 kHz and 15.6 kHz is set as the frequency of the AC voltage.

In this paragraph, another example of operation of the MEMS switch described up to the previous paragraph will be described. Below, each voltage value in each time according to the sequence illustrated to Fig.3 (a) is illustrated.
First time (initial state): “First time” is the time when power is turned on to the circuit system on which the MEMS switch is mounted. Immediately after turning on the power, the first AC voltage (0.5 volts pp) is supplied. However, since there is no DC voltage to be superimposed, no “make operation” is induced. That is, the MEMS switch is in a released state.
Second time ("make" operation): The first DC voltage (6.0 volts) is superimposed on the already supplied first AC voltage. As a result, “pull-in” is immediately induced.
“Make” state: Since the first AC voltage and the first DC voltage are continuously supplied after reaching the “pull-in” state, the MEMS switch shifts to the “hold” state and maintains the “make state”. Has been. In this state, even if the AC voltage is set to 0 volts pp, the DC voltage (6.0 volts) can maintain the “hold” state.
Third time ("break" operation): When the first DC voltage is lowered to the third DC voltage (3.0 volts) while maintaining the first AC voltage (0.5 volts pp) The MEMS switch is immediately "released". In this case, the third AC voltage has the same value as the first AC voltage.
“Break” state: Upon reaching the “release” state, a DC voltage of 6.0 volts is applied for the next “make” operation. The AC voltage remains the third AC voltage, but the “release” state continues because the DC voltage is below the pull-in voltage (30 volts). This state is a standby state in preparation for the subsequent “second time (“ make ”operation”) ”.
In the above sequence, the supplied AC voltage is constant, and the MEMS switch is operated by controlling the value of the DC voltage.

In this paragraph, another example of operation of the MEMS switch will be described. Below, each voltage value in each time according to the sequence illustrated to Fig.3 (a) is illustrated.
First time (initial state): “First time” is the time when power is turned on to the circuit system on which the MEMS switch is mounted. Immediately after power-on, the first DC voltage (6.0 volts) is supplied. However, since there is no superimposed AC voltage, the “make operation” is not induced. That is, the MEMS switch is in a released state.
Second time (“make” operation): The first AC voltage (0.5 volts pp) is superimposed on the already supplied first DC voltage. As a result, “pull-in” is immediately induced.
“Make” state: After reaching the “pull-in” state, the first AC voltage is changed to the second AC voltage (0 volts pp), and the first DC voltage is continuously supplied (second DC voltage) But also) The MEMS switch shifts to the “hold” state, and the “make state” is maintained. In this state, the AC voltage may be kept at 0.5 volts pp.
Third time ("break" operation): The AC voltage remains at 0 volts pp (ie, the third AC voltage is the same as the second AC voltage), and the first DC voltage (or second When the DC voltage of the MEMS switch is reduced to the third DC voltage (3.0 volts), the MEMS switch is immediately "released". In this case, the third AC voltage may be 0.5 volts pp.
“Break” state: Upon reaching the “release” state, a DC voltage of 6.0 volts is applied for the next “make” operation. The AC voltage remains the third AC voltage. Since the DC voltage is equal to or lower than the pull-in voltage (30 volts) in the conventional driving method, the “release” state continues. This state is a standby state in preparation for the subsequent “second time (“ make ”operation”) ”.
In the above sequence, the MEMS switch can perform a desired operation by supplying an alternating voltage of 0.5 volts pp for a short time to induce the “make operation”. As a result, it is possible to suppress power consumption in the drive system of the MEMS switch. Further, since an AC signal is applied only at the moment of the “make operation”, there is an advantage that unnecessary radiation of the electric field to the surrounding area where the MEMS switch is arranged can be suppressed.

<Configuration example of MEMS switch>
FIG. 4 is a diagram illustrating a configuration example of a MEMS switch that is Embodiment 3 of the present invention. In the third embodiment, a MEMS switch using a movable structure made of a conductive material such as metal is illustrated. FIG. 4A is a plan view seen from the upper surface of the MEMS switch, and FIG. 4B is a cross-sectional view along AA ′. Further, on the right side of FIG. 4A, a plan view in which only the central portion of the plan view is enlarged is displayed. In the figure, 45 is a movable structure, 46 is a support, and 48 is a glass substrate. On the upper surface of the glass substrate 48, a coplanar waveguide composed of a signal line 40 through which a high-frequency signal propagates and a ground 41 disposed on both sides thereof is disposed. Further, a fixed electrode 42 for driving is disposed on the upper surface of the glass substrate, and an insulating film 43 for preventing an electrical short circuit is provided on the surface of the fixed electrode 42. The movable structure 45 is coupled to the support 46 via the suspension 34, and the support 46 is joined to the glass substrate 48. In other words, the movable structure 45 is arranged so as to float above the glass substrate 48 and is formed so as to cover the coplanar waveguide provided on the surface of the glass substrate 48. The support 46 is bonded to a ground 41 disposed on the upper surface of the glass substrate 48. Since the movable structure 45 is made of a conductive material, the movable structure 45 is always at a ground potential (generally 0 volts). The fixed electrode 42 is connected to a conductive pattern 47. Further, since the movable structure 45 is displaced in the vertical direction in FIG. 5B, in order to suppress the damping of the air layer existing around the movable structure 45, a large number of small through holes (numbering) Not). Further, a contact 44 is provided on the lower surface of the movable structure 45. The contact 44 is designed to be positioned above the signal line 40 so that when the movable structure 45 is displaced toward the glass substrate 48 and is brought into a pull-in state, the contact 44 is in electrical contact with the signal line 40. It has become.

  In this paragraph, the operation of the third embodiment shown in FIG. 4 will be described. As described above, the potential of the movable structure 45 is the ground potential. The conductive pattern 47 is connected to a drive source (not shown) composed of an AC voltage source and a DC voltage source. For this reason, electrostatic forces attracting each other are generated between the movable structure 45 and the fixed electrode 42, and the distance (gap length) between the movable structure 45 and the fixed electrode 42 is reduced. If the magnitude of the electrical signal (AC voltage and DC voltage) from the drive source satisfies the condition for inducing “pull-in”, the movable structure 45 is in contact with (or extremely close to) the fixed electrode 42. , "Pull-in" occurs, and a "make operation" of the MEMS switch is induced. If mechanical or electrical contact occurs in such a state, an excessive short-circuit current flows between the movable structure 45 and the 42 fixed electrode. To prevent this, An insulating film 43 is disposed on the surface. In the pull-in state, the contact 44 comes into contact with (or is extremely close to) the signal line 40 described above. As a result, the signal line 40 is fixed to the ground potential. That is, the high frequency signal flowing into the signal line 40 propagates to the ground side via the contact 44 and cannot propagate through the coplanar waveguide. With this operation, the function of “off” the switch (the high-frequency signal is cut off) is achieved. On the other hand, when the magnitude of the electric signal from the drive source is a condition for inducing a “release” state, the contact between the contact 44 and the signal line 40 is released, and a “break operation” is induced. As a result, the high frequency signal that also flows into the signal line 40 continues to propagate through the coplanar waveguide. With this operation, the function of “ON” of the switch (a high-frequency signal is connected) is achieved.

  The material of the glass substrate 48 is not limited to glass, but may be any material having a large specific resistance, such as a semiconductor material such as high-resistance silicon or gallium arsenide (GaAs), ceramic, or resin. Furthermore, a material having a small tangent loss (tan δ) is more preferable because attenuation of a high-frequency signal can be suppressed.

  The movable structure 45, the suspension 34, and the support base described above can use materials such as various metals, semiconductors, and conductive plastics. Among metal materials, nickel (Ni) having a high Young's modulus and a low density is particularly preferable. Since nickel can be produced by “plating”, it is suitable for a structure in which the MEMS switch is integrated on the same semiconductor substrate as the semiconductor integrated circuit. Various metals (for example, gold, platinum), semiconductors, conductive plastics, and the like can be used as the contact material. Gold and a gold alloy are particularly preferable because the contact resistance can be reduced, but this is not restrictive.

  In the third embodiment, the movable structure 45 is resonated using the electrostatic force generated by the applied electric signal. However, the resonance may be caused using other physical phenomena. For example, the resonance can be excited by mechanical deformation (strain) generated by depositing a piezoelectric thin film on the upper surface of the movable structure 45 and applying an electric signal to the thin film. As a modified example of such a configuration, the movable structure 45 itself is made of a piezoelectric material, and conductivity is imparted to the surface thereof. Further, the resonance may be induced by utilizing a mechanical force (Lorentz force) induced in the conductive path arranged in the magnetic field. As described above, in the present invention, the “motive force” that causes resonance is not limited to a specific method.

A design example of the MEMS switch having the structure shown in FIG. 4 is shown below.
Transmission line (signal line 40 and ground 41), fixed electrode 42: gold insulating film 43 with a thickness of 1 micron: silicon nitride film with a thickness of 0.1 micron movable structure 45, suspension 34, and support base 46:12 Micron thick nickel Contact 44: 0.15 micron thick gold Initial value of distance between movable structure and fixed electrode: 1 micron

  A MEMS switch designed with the design values illustrated in the previous paragraph causes pull-in by superimposing a 1-volt DC voltage on a 1-volt DC voltage, and the switch is turned on and off repeatedly. It became clear to do. In addition, since the resonance frequency (fundamental wave) of the movable structure is 21.4 kHz, the switching time is about 50 μsec.

<Method for manufacturing MEMS switch>
An example of a manufacturing method of the MEMS switch will be described in detail with reference to FIGS. In these drawings, the same number indicates the same component.
(1) A chromium thin film (thickness 0.02 micron) 51 and a first gold thin film (thickness 1 micron) 52 are formed on a glass substrate 50 having a thickness of 500 microns by vapor deposition. Thereafter, a silicon nitride film (thickness 0.2 μm) 54 is deposited using a second chromium thin film (thickness 0.02 μm) 53 and a CVD apparatus. The chromium thin film 53 is patterned using a known technique (FIG. 5A).
(2) The gold thin film 52 and the chromium thin film 51 are patterned using a known technique (a photolithography technique using a photoresist). Subsequently, an aluminum thin film (thickness: 0.02 microns) 55 is deposited thereon, and then the photoresist is peeled off (so-called “lift-off”). As a result, the aluminum thin film 55 is laminated only in the region where the glass substrate 50 is exposed (FIG. 5B).
(3) After the first copper thin film is deposited on the entire upper surface of the structure shown in FIG. 5B, the surface of the copper thin film is polished using a polishing apparatus, and the upper surface of the silicon nitride film 54 Polishing is performed until the height is the same (that is, the surface of the silicon nitride film 54 is exposed). Reference numeral 56 denotes a first copper thin film at the time when the polishing step is completed (FIG. 5C).
(4) Deposit a second copper thin film 57 by depositing 0.05 micron thick copper on the surface of the copper thin film 56. Subsequently, an opening 58 having a diameter of 20 microns is formed in the region of the copper thin film 57 (FIG. 5D).
(5) A third copper thin film 59 is produced by depositing copper having a thickness of 1 micron on the surface of the second copper thin film 57. Subsequently, after laminating the second gold thin film, a gold contact 60 is formed in the region of the opening 58 by a known technique (FIG. 5E).
(6) A part of the region of the copper thin film 59 (actually composed of the first to third copper thin films) is etched to produce the support opening 61 (FIG. 5F).
(7) A resist having a thickness of 20 microns is laminated on the entire upper surface of the structure shown in FIG. 5 (f) and then patterned (not shown). Nickel plating with a thickness of 12 microns is performed on the opening of the resist pattern to produce a nickel thin film 62. Subsequently, the resist is removed (FIG. 5G).
(8) Finally, the copper thin film 59 (actually composed of the first to third copper thin films) is removed with nitric acid (sacrificial layer etch) (FIG. 5H).
The MEMS switch is manufactured by the above process.

<Drive operation range of MEMS switch>
FIG. 6 shows measurement results when the MEMS switch having the configuration illustrated in FIG. 4 is driven by the driving method according to the present invention described with reference to FIG. 2, and shows an operation range of the manufactured MEMS switch. The horizontal axis of the graph of the figure shows the value of the applied DC voltage, and the vertical axis shows the value of the applied AC voltage (Peak-to-Peak). The asterisk in the figure indicates that the MEMS switch caused a “pull-in”, the movable structure “contacted” (including extreme proximity) to the fixed electrode, and the “contact” was maintained. ing. That is, the conditions for performing the “make-up state” are shown. The “contact” here confirms the occurrence of the “contact” by measuring the electrical resistance between the contact 44 shown in FIG. 4B and the signal line 40. From the graph of FIG. 6, the ♦ mark exists in the area surrounded by the three straight lines 65, 66, and 67. The straight line 65 indicates a case where the applied DC voltage is equal to the hold voltage of the MEMS switch. If the operating condition (which is a combination of the DC voltage value and the AC voltage value) is on the left side of this straight line, the MEMS switch “hold” does not occur (naturally, “pull-in” also occurs. do not do). A straight line 66 indicates a case where the positive peak value (corresponding to the “mountain” of the waveform) of the AC voltage waveform superimposed on the DC voltage is equal to the “pull-in” voltage of the switch. That is, in order for the MEMS switch to “pull in”, it is necessary to set an operating condition above this straight line. Further, a straight line 67 indicates a condition in which the negative peak value (corresponding to the “valley” of the waveform) of the AC voltage waveform superimposed on the DC voltage is equal to or lower than the “release” voltage. For this reason, when the operating condition is set above the straight line, the MEMS switch is “pulled in” (although it is instantaneous), but stable “hold” does not occur, and the switch functions normally. Will not be achieved.

In this paragraph, a driving method of the MEMS switch obtained from the result of FIG. 6 is illustrated. (1) Keep the DC voltage constant and apply the AC voltage to “pull in” the switch. To “hold” the switch, it can be left as it is, or the amplitude of the AC voltage can be reduced. This is determined according to the mutual relationship between the AC voltage value, the DC voltage value, and the hold voltage. On the other hand, a switch in the “hold” state can be “released” by decreasing the DC voltage.
(2) “pull-in” the switch by keeping the AC voltage constant and applying the DC voltage. To “hold” the switch, it can be left as it is, or the DC voltage can be reduced. This is determined according to the relationship between the AC voltage value, the DC voltage value, and the hold voltage. On the other hand, a switch in the “hold” state can be “released” by decreasing the value of the DC voltage.
(3) By making the DC voltage constant and superimposing an AC voltage in the vicinity of the resonance frequency of the movable structure (or, strictly speaking, a frequency in the range of 20% before and after the resonance frequency), the switch becomes “ Pull in ”. To “hold” the switch, it can be left as it is, or the amplitude of the AC voltage can be reduced. This is determined according to the relationship between the AC voltage value, the DC voltage value, and the hold voltage. On the other hand, a switch in the “hold” state is “released” by separating the frequency of the AC voltage from the resonance frequency (to be strictly speaking, setting it outside the range of 20% before and after the resonance frequency). Can be made. In addition to the three representative examples described above, there are many variations in the combination of the DC voltage value, the AC voltage value, the frequency, and the respective application timings for driving the MEMS switch. Is included.

  Note that the frequency of the AC voltage to be applied may not be a single frequency, but may be a signal “including simultaneously” a plurality of frequencies near the resonance frequency of the movable structure. Use of such a method has the advantage that the design of the AC power supply is simplified. Furthermore, as for the frequency of the alternating voltage to be applied, a signal that “sequentially includes a plurality of frequencies in the vicinity of the resonance frequency of the movable structure” (that is, a signal whose frequency changes on the time axis) is used. Is also possible. Here, the frequency may be sequentially increased or decreased.

<Other driving methods for MEMS switches>
It is known that an interesting phenomenon occurs when the movable structure is near the fixed electrode and has not yet been “pulled in”. This is called a nonlinear resonance phenomenon. FIG. 7 is a diagram showing the principle of a switch driving method using the nonlinear resonance phenomenon. This figure shows the results of actual measurement of the prototype MEMS switch. The horizontal axis is the frequency of the AC voltage, and the vertical axis is the amplitude of the mechanical vibration of the movable structure (measured with a laser Doppler device). It is. As shown in the figure, (1) when the frequency of the AC voltage is increased and brought closer to the resonance frequency of the movable structure, the amplitude increases (curve 81). (2) When the frequency becomes higher than the resonance frequency, the amplitude decreases (curve 82). Under such conditions, there is no “pull-in” of the MEMS switch. (3) When the frequency is decreased while the frequency exceeds the resonance frequency, the amplitude increases due to a nonlinear phenomenon (curve 83). (4) When the frequency is further lowered, “pull-in” occurs (curve 84). The driving method using such a nonlinear phenomenon indicates that it is possible to cause the “pull-in” of the MEMS switch even in “driving conditions in which pull-in does not occur” in the driving method not using the nonlinear phenomenon. This is useful for further reducing the driving voltage. It has been confirmed that the prototype MEMS switch causes “pull-in” by a driving method in which the frequency of the alternating voltage is increased and then decreased. On the other hand, it was also confirmed that “pull-in” occurs by an operation method in which the frequency of the AC signal is decreased and the frequency is increased when the frequency is lower than the resonance frequency. A driving method using such a nonlinear phenomenon is also included in the present invention.

<Other configuration of MEMS switch>
FIG. 8 shows a MEMS switch having a configuration different from that of the third embodiment described above. The plan view of the MEMS switch in the present embodiment is the same as that in FIG. 4, but the cross-sectional view along AA ′ is different. FIG. 8 shows the cross-sectional view. 8, the same numbers as those in FIG. 4 indicate the same components. This embodiment is characterized in that the thickness of the fixed electrode 72 is thicker than that of the embodiment (fixed electrode 42) shown in FIG. With this configuration, it is possible to set a small interval (gap length) between the movable structure 45 and the fixed electrode 72. As a result, the magnitude of the electric signal for driving (the above-described DC voltage or AC voltage) can be reduced, which is effective in solving the problem of reducing the driving voltage. Furthermore, since the height of the contact 74 arranged in the movable structure 45 is designed without being uniquely limited by the gap between the movable electrode and the fixed electrode, there is an advantage that the degree of freedom of design is increased.

An example of designing a MEMS switch with the configuration shown in FIG. 8 is shown below.
Transmission line (signal line 40 and ground 41): 1 micron thick gold Fixed electrode 72: 1.44 micron thick gold Insulating film 43: 0.05 micron thick silicon nitride film Movable structure 45, suspension 34, and Support base 46: nickel having a thickness of 4 microns Contact 74: gold having a thickness of 0.5 microns The initial value of the distance between the movable structure 45 and the fixed electrode 72: 0.21 microns

  In the MEMS switch designed with the design values exemplified in the previous paragraph, pull-in is generated by superimposing an AC voltage of 30 millivolts (effective value, rms) on a DC voltage of 0.96 volts, and the switch is repeatedly turned on and off repeatedly. It became clear to do the action. In addition, since the resonance frequency (fundamental wave) of the movable structure is 168 kHz, the switching time is about 10 μsec. Furthermore, even if the high-frequency signal propagating through the signal line has a high power of 2 W, the high-frequency signal can be controlled.

<Other configuration of MEMS switch>
FIG. 9 shows a MEMS switch having a configuration different from that of the third embodiment described above. The plan view of the MEMS switch in the present embodiment is the same as that in FIG. 4, but the cross-sectional view along AA ′ is different. FIG. 9 shows the cross-sectional view. 9, the same numbers as those in FIG. 4 indicate the same components. This embodiment is different from the above-described embodiment 3 (for example, FIG. 4) in the following points. (1) Metal bumps 80a and 80b are provided in place of the insulating film 43 in FIG. (2) The fixed electrode 42 of FIG. 4 is provided as fixed electrodes 81a and 82a and fixed electrodes 81b and 82b. The fixed electrode 81a and the fixed electrode 82a are electrically connected (not shown), and the fixed electrode 81b and the fixed electrode 82b are similarly connected. The height of the metal bump 80a and the metal bump 80b is set larger than the height of the fixed electrodes 81a, 81b, 82a, and 82b. (3) A contact 44 is provided in the central region of the movable structure 45, and the contact 44 comes into contact with the transmission line 40 when the movable structure 45 is pulled in. (4) The metal bumps 80 a and 80 b are set in an electrically floating state or the same potential as the movable structure 45. In such a configuration, when the movable structure 45 is “pulled in”, the displacement (or movement) of the movable structure 45 is limited by the metal bumps 80a and 80b. As described above, due to the difference in height between the metal bumps 80a, 80b and the fixed electrodes 81a, 81b, 82a, 82b, the movable structure 45 is fixed to the fixed electrodes 81a, 81b, 82a in the “pull-in” state. , 82b is avoided. For this reason, in FIG. 4, there is an advantage that the insulating film 43 provided for preventing the short circuit becomes unnecessary.

  This embodiment is characterized in that a short circuit between the movable structure and the fixed electrode is prevented by a metal bump without using an insulating film. For this reason, the movable structure does not adhere to the fixed electrode due to the charging (charge-up) phenomenon of the insulating film, and there is an advantage that the long-term reliability of the MEMS switch can be increased.

<Other configuration of MEMS switch>
FIG. 10 shows a MEMS switch having a configuration different from that of the third embodiment described above. The plan view of the MEMS switch in the present embodiment is the same as that in FIG. 4, but the cross-sectional view along AA ′ is different. FIG. 10 shows the cross-sectional view. 10, the same numbers as those in FIG. 4 indicate the same components. This embodiment is different from the above-described embodiment 3 (for example, FIG. 4) in the following points. (1) There are three contacts arranged on the movable structure 45. These are 95, 96a, and 96b (see FIG. 10). Since the three contacts are integrated with the same movable structure 45 (which is a conductor), they have the same potential. (2) The coplanar waveguide includes a signal line 90 (corresponding to the signal line 40) and grounds 91 (corresponding to the ground 41) arranged on both sides thereof. (3) In a state where the movable structure 45 is “pulled in”, the contact point 95 contacts the signal line 90. Further, the contacts 96a and 96b are in contact with the ground 91. As a result, the signal line 90 constituting the coplanar waveguide is short-circuited with the ground 91. (4) Since the movable structure 45 contacts the coplanar waveguide at “3 points”, the movable structure 45 does not contact the fixed electrode 72. For reference, in the configuration in which there is one contact (the third embodiment in FIG. 4), the movable structure 45 is inclined to one side like a “seesaw” and is in contact with one of the fixed electrodes 72. There is danger. The configuration shown in FIG. 10 has the following advantages. (1) In this embodiment, since there is no possibility of contact with the fixed electrode 72, the insulating film 43 in FIG. 4 becomes unnecessary. As a result, the movable structure 45 does not adhere to the fixed electrode 72 due to the charge-up phenomenon of the insulating film, and there is an advantage that the long-term reliability of the MEMS switch can be increased. (2) Although the coplanar waveguide is short-circuited in the “pull-in” state, the short-circuit current flowing between the signal line 90 and the ground 91 causes a current path from the contact 95 to the contact 96a or the contact 96b. pass. For this reason, there is an advantage that the short-circuit current path (length of the current path) in the conventional example is shortened, and the high-frequency characteristics can be greatly improved.

<Other configuration of MEMS switch>
FIG. 11 shows a MEMS switch having a configuration different from that of the third embodiment described above. The plan view of the MEMS switch in the present embodiment is the same as that in FIG. 4, but the cross-sectional view along AA ′ is different. FIG. 11 shows the cross-sectional view. 11, the same reference numerals as those in FIG. 4 denote the same components. This embodiment is different from the above-described embodiment 3 (for example, FIG. 4) in the following points. (1) An insulating film 101 is disposed on the upper surface of the signal line 100 (corresponding to the signal line 40). The material of the insulating film 101 is a known material such as a nitride film, an oxide film, ceramics, or a resin material. (2) There is no insulating film on the upper surface of the fixed electrode 42. In this configuration, the contact 44 provided on the movable structure 45 is in mechanical contact with the signal line 100 through the insulating film 101 in the “pull-in” state. Such contact does not cause electrical contact. However, when the thickness of the insulating film is 1 μm or less, the capacitance value formed between the contact 44 and the signal line 100 becomes large. Contact (so-called “electrostatic coupling”) will be achieved. In the configuration of the present embodiment, the electrical contact is “indirect”, and is not the contact between the metal (gold of the contact 44) and the metal (gold of the signal line 40) as in the third embodiment. For this reason, there is no fluctuation of the metal-metal contact resistance, and it is possible to suppress the deterioration of the high frequency characteristics. That is, the MEMS switch characteristics can be improved.

<Other configuration of MEMS switch>
FIG. 12 shows a MEMS switch characterized in that a plurality of contacts are arranged in a movable structure and contact with signal lines occurs at a plurality of locations. In the ninth embodiment, a configuration in which three contacts are arranged across the coplanar waveguide is shown, and a plan view thereof is shown in FIG. 4 and a sectional view thereof is shown in FIG. The present embodiment has a configuration obtained by further developing the ninth embodiment. FIG. 12 is an overhead view of the region where the contacts are arranged in the configuration of the present embodiment as viewed obliquely from above. This figure corresponds to an overhead view of the enlarged view shown on the right side of FIG. In the figure, reference numeral 110 denotes a signal line, and a configuration branched into four is illustrated. Reference numeral 111 denotes a movable structure. A contact point is disposed below the region 112. A fixed electrode is disposed below the region 113. In the region 112 of FIG. 12, 4 × 4 contact points are two-dimensionally arranged. The number of the contacts is not limited to 4 × 4, and the arrangement may be one-dimensional. In particular, when the arrangement is one-dimensional and the arrangement direction coincides with the direction of the signal line, it is not necessary to perform “signal line branching” as indicated by 110 in FIG. The present embodiment is characterized in that a plurality of the contact points are simultaneously brought into contact with the signal line. For this reason, even if the area of the region occupied by the contact is small and the allowable current at each location is small, the allowable current value can be increased as a whole. In addition, since the “contact” of the individual contacts is performed in parallel, it is possible to improve the reliability of the contact (even if several contacts cause a “contact” failure, the influence on the whole is small). Furthermore, in this embodiment, since the individual switch dimensions can be reduced, there is an advantage that a large spring constant for “contacting” each contact can be set.

  Note that, in this specification, a both-end fixed beam type MEMS switch is exemplified, but the MEMS switch has many other forms. For example, a disk type (disc shape, etc.), a ring type, a “fishbone” type, and a “connected type switch” in which the movable structure regions of a plurality of adjacent MEMS switches are mechanically coupled. There is. The present invention can be implemented for many of these forms.

  In this specification, a resistance-type MEMS switch having a contact is described as an example. However, the present invention can be easily applied to a capacitance-type MEMS switch in which a capacitance is provided at a contact portion.

  According to the present invention, the driving voltage of the electrostatic actuator can be greatly reduced, and the effect when applied to a MEMS switch is great. In particular, when applied to a MEMS switch, a movable structure having a large “spring” constant can be driven with a low driving voltage, so that switching speed and long-term reliability are improved, and a high-power signal is transmitted. It becomes possible to handle. Further, since the MEMS switch is manufactured using silicon semiconductor technology, when the MEMS switch is manufactured on a silicon substrate, one-chip integration with a drive circuit and a signal processing circuit is possible. It is also effective for upgrading. In this case, the effect of the present invention that the MEMS switch can be driven with a driving voltage of 1 to 2 volts is very effective. Such an improvement in function can greatly contribute to the enhancement of functions of portable devices, particularly portable communication devices. In addition, multi-frequency communication represented by cognitive communication promotes downsizing and low power consumption of devices, and can be used to construct a ubiquitous environment. Furthermore, when a switch matrix is configured using the MEMS switch of the present invention, a small switching board can be manufactured. Since the high frequency characteristics of the MEMS switch are excellent, this switching board can be used to switch the signal of the high frequency circuit. Furthermore, by arranging a number of MEMS switches of the present invention on a high-frequency signal line, it can be widely used for impedance matching circuits, filters, antenna phase shifters, and the like.

DESCRIPTION OF SYMBOLS 11 Input terminal 12 Output terminal 13 MEMS switch 14 AC voltage source 15 DC voltage source 16 Drive switch 17a, 17b Transmission line 20, 48, 50, 190 Glass substrate 21, 45, 111, 199 Movable structure 22, 46, 195a, 195b Support base 23, 42, 72, 81a, 81b, 82a, 82b, 193a, 193b Fixed electrode 24, 44, 74, 95, 96a, 96b, 198 Contact 25, 40, 90, 100, 110, 191 Signal Line 27, 28, 29, 81, 82, 83, 84 Curve 34, 196a, 196b Suspension 41, 91, 192a, 192b Ground 43, 101 Insulating film 47 Conductive pattern 51, 53 Chrome thin film 52 Gold thin film 54 Silicon nitride Film 55 Aluminum thin film 56, 57, 59 Copper thin film 58, 1 opening 60 Gold Contact 62 nickel thin film 65, 66, and 67 linearly 80a for determining operating limits, 80b metal bumps 112 and 113 regions 194a, 194b bump contact portion 197a, 197b bumps

Claims (7)

In a MEMS switch having a movable structure and a fixed electrode,
Supply a signal that superimposes the specified first AC voltage and the specified first DC voltage,
Inducing make motion by mechanically vibrating the movable structure,
By changing the first AC voltage to the designated second AC voltage and changing the first DC voltage to the designated second DC voltage,
Maintaining the makeup action,
By changing the second AC voltage to the designated third AC voltage and changing the second DC voltage to the designated third DC voltage,
A MEMS switch that induces a break operation. The first AC voltage for vibrating the movable structure is:
2. The MEMS switch according to claim 1, wherein the MEMS switch has a frequency within a frequency range of plus 20 percent and minus minus 20 percent around the resonance frequency of the movable structure. Supplying the first alternating voltage at a designated first time;
Inducing a make operation by superimposing the first DC voltage on the first AC voltage at a specified second time,
A break operation is induced by changing the first AC voltage to the third AC voltage at the designated third time and changing the first DC voltage to the third DC voltage. The MEMS switch according to claim 1 or 2, wherein Supplying the first DC voltage at a designated first time;
Inducing a make operation by superimposing the first AC voltage on the first DC voltage at a specified second time,
A break operation is induced by changing the first AC voltage to the third AC voltage at the designated third time and changing the first DC voltage to the third DC voltage. The MEMS switch according to claim 1 or 2, wherein Supply a signal that superimposes the specified first AC voltage and the specified first DC voltage,
Inducing make motion by mechanically vibrating the movable structure,
The MEMS switch according to claim 1, wherein a break operation is induced by increasing or decreasing a frequency of the first AC voltage. Supply a signal that superimposes the specified first AC voltage and the specified first DC voltage,
The makeup operation is induced by increasing the frequency of the first AC voltage until it exceeds a joint frequency of the movable structure and then decreasing the frequency of the first AC voltage. 3. The MEMS switch according to 1 or 2.   3. The MEMS switch according to claim 1, wherein the first AC voltage includes a plurality of frequency components in a frequency range including the resonance frequency. 4.

Images