Patent Application
20050162244
This invention relates to a switch, for use on an electronic circuit or the like, adapted for switching over a propagation path for an external signal by attracting or repelling the movable member to or from the electrode.
The conventional RF-MEMS switch is a mechanical switch having movable members in a membrane or rod form supported at both ends or in a cantilever, so that by placing them into or out of contact with the electrodes, signal propagation path can be switched over. Although the power sources for driving the membrane or movable member, in many cases, use those based on electrostatic force, there are released ones using magnetic force.
As a micro-fabricated switch in a size around 100 μm, there is known one described in IEEE Microwave and Wireless Components letters, Vol. 11 No 8, August 2001 p 334. This switch forms a signal line for radio-signal transmission over a membrane, to provide a control electrode immediately beneath the signal line. In case a direct current potential is applied to the control electrode, the membrane is pulled and deformed toward the control electrode by an electrostatic attractive force. By a contact with a ground electrode formed on the substrate, the signal line formed on the membrane becomes a shorted state. Due to this, the signal flowing through the signal line is attenuated down or blocked off.
Unless a direct current potential is applied to the control electrode, there is no deformation in the membrane. The signal flowing through the signal line on the membrane is allowed to pass through the switch without encountering the loss through the ground electrode.
Meanwhile, as a conventional method for controlling the positioning of the movable member, there is known an art shown in JP-A-2-7014. This structure is arranged to open and close an optical path by a micro-switch, thereby turning the signal on/off. When to pass light, a voltage is applied to between a vibration plate and a flat plate, to lift the element through an electrostatic force. When to block light, voltage is rendered zero to cancel the electrostatic force whereby it is returned to the former position by a spring force of the vibration plate. Due to this, the element blocks the light.
At this time, in case the voltage is abruptly applied or reduced to zero, a phenomenon called chattering takes place, resulting in vibration of the element. It takes a time in reaching a stability. Consequently, it is a practice to apply a voltage called a preparatory voltage pulse before applying a control voltage, thereby preventing chattering. The condition for stabilization is determined by a preparatory pulse voltage V1 and a pulse width τ1, and a spacing τ2 between the preparatory pulse voltage V2 and the major control voltage. In case V1=V2 and τ1=τ2 is assumed, then τ1 has a boundary condition of one-sixth of the eigenfrequency.
The research and development of RF-MEMS switch in the IEEE Microwave and Wireless Components letters originates aiming at those for the military and aerospace applications, wherein the research and development is focused on by what means signal propagation characteristic is to be improved. However, in the case of the home-use application including personal digital assistants, there is a desire for an RF-MEMS switch meeting simultaneously various conditions of durability, high-speed response, low consumption power, low driving voltage, size reduction and the like, besides improved signal propagation characteristic as a natural matter.
However, the direct current voltage of as high as approximately 30 V or more is required to attract the membrane toward the control electrode. It is not preferred to build such a switch as needing a high voltage within a radio transceiver apparatus.
Meanwhile, in order to achieve high electrical isolation on a switch, it is required to provide a comparatively wide gap between a movable member and an electrode. In such a case, it is critical by what means the movable member is to be driven with a great displacement and high speed on a low drive voltage.
Also, on the RF-MEMS switch for example, when the movable member is attracted on the electrode, in case the drive voltage is turned off into a state not to give an attractive force to the movable member, the movable member is returned by its own spring force to a predetermined position distant from the electrode. For attracting the movable member at high speed to the electrode by a low drive voltage, the spring force of the movable member must be weakened. This, however, poses a problem of low response speed for the movable member to return to a predetermined position.
Also, on a mechanical switch, in returning the movable member contacted with the electrode to a position where isolation is high not to cause a capacitance coupling of movable member and electrode, there is a problem of overshoot, i.e. the movable member is to displace beyond the predetermined position. Where the overshoot of movable member is great, capacitance coupling possibly takes place on the electrode and movable member, as a signal propagation path, resulting in forming an incorrect signal path.
On the other hand, the switch of JP-A-2-7014 requires a sufficient connection area in order to secure a capacitance during switch-on. In the case the beam assumably has a width of several μm, the beam has a length on the order of several hundred μm. Accordingly, it is difficult to fix a beam having a length of several hundred μm only at one end. Higher stability is available rather by a both-ends-supported beam fixed at both ends.
However, where fixed at both ends, the substrate and beam materials, if different, cause a change of internal stress due to a difference in the thermal expansion coefficient between the materials, thereby changing the spring constant. The eigenfrequency of a structural body is determined by a mass and spring constant of the beam, as shown in Equation 1. Accordingly, temperature change causes eigenfrequency change correspondingly.
f=1/π{square root}{square root over ( )}k/m Eq. 1
Even in case a preparatory pulse voltage is applied to avoid chattering, a switch temperature change causes a change of eigenfrequency, hence changing the optimal preparatory pulse voltage. For example, when the preparatory pulse voltage is optimized at room temperature, a rise in switch temperature causes an eigenfrequency increase. Based on a preparatory pulse voltage same as that at room temperature, it is impossible to prevent chattering.
From these problems and requests, there is a desire for a switch realized with switch high-speed response on low driving voltage and a widened gap at between the movable member and the electrode, enabling to increase the response speed for the movable member attracted on the electrode to return to a predetermined position distant from the electrode and to control the magnitude of an overshoot of the movable member.
It is an object of the present invention to provide a high-performance switch realized with signal propagation characteristic improvement, high speed response, low consumption power and low driving voltage, and an electronic appliance using the same.
A switch of the present invention is a switch for switching over an external signal propagation path by attracting or repelling a movable member to or from an electrode, the switch comprising: an input port for inputting an external signal; and a movable member connected to the input port; a first electrode for propagating the external signal; a first control power supply connected to the first electrode and for generating a control signal; a second electrode for blocking the external signal; and a second control power supply connected to the second electrode and for generating a control signal; whereby the first control power supply provides a control signal to the first electrode, the movable member being displaced by a driving force generated based on a potential difference between the movable member and first electrode and a potential difference between the movable member and second electrode, thereby being attracted to the first or second electrode. This makes it possible to realize a switch for signal propagation characteristic improvement, high-speed response, low consumption power and low driving voltage.
Exemplary embodiments of the present invention are demonstrated hereinafter with reference to the accompanying drawings.
In a drive scheme, a control signal 21 of alternating current voltage as shown in
Incidentally, even in a mode other than the self-resonant mode of the movable member 2, in the case a vibration speed and low drive voltage is obtainable to satisfy a sufficient vibratory displacement and desired response speed for switching during vibration of the movable member 2, vibration and switching is feasible at a frequency other than the self-resonant frequency of the movable member 2.
Meanwhile, besides the alternating current voltage control signal, it is possible to use a control signal in another waveform such as a rectangular waveform.
Also, although embodiment 1 showed the vibration driving scheme to the movable member by an electrostatic force, it is possible to realize a switch on a vibration driving scheme using another kind of driving force such as magnetic force.
According to the embodiment 1, the movable member 2 can be driven with a great displacement at high speed on a low drive voltage, making it possible to provide a comparatively broad gap at between the movable member 2 and the electrode 3, 4. This enables high electrical isolation on the switch, to realize a high-performance switch having a high signal on/off ratio.
Meanwhile, by designing and fabricating the movable member 2 to have a self-resonant frequency corresponding to a vibration speed higher than a desired response speed, a higher response speed can be realized for the movable member 2.
Incidentally, by attracting the movable member to the electrode with the movable member always vibrated at higher speed than a desired response speed, it is possible to realize a high response speed corresponding to a vibration frequency.
Also, by vibrating the movable member at a higher speed than a desired response speed from a predetermined position of the movable member distant from the electrode, a high response speed can be realized.
Also, by vibrating the movable member at a higher speed than a desired response speed with a state the movable member attracted on the electrode, a high response speed can be realized. In this case, the frequency for vibrating the movable member may be at a self-resonant frequency of the movable member in a form that the movable member is attracted on the electrode.
Also, by vibrating the movable member with a state the movable member attracted on the electrode, the movable member can be released from the electrode and returned, with high electrical isolation, to a predetermined position at high speed without causing a capacitance coupling between the movable member and the electrode.
In the case of signal reception, an RF signal inputted at an antenna end 507 is inputted to the receiving section 502 through the transmission/reception switching section 501 where the signal is amplified and frequency-converted and thereafter outputted, at an IF terminal 507, to the IF section 505. In the case of sending a signal, operation is reverse to the above, i.e. the signal outputted from the IF section 505 is inputted to the transmitting section 504 through the IF end 508 where it is frequency-converted and amplified, thereafter being passed through the transmission/reception switching section 501 and outputted at the antenna end 507.
The transmission/reception switching section 501, because of requiring a low-loss device, uses the switch of embodiment 1.
In order to prevent chattering similarly to embodiment 1, there is a need for a control signal that is not a simple control signal. Using
The switch 507 is structured by two movable electrodes 531, 532 fixed at both ends. In case a direct current potential is applied between the movable electrodes 531, 532, the movable electrodes 531, 532 are pulled and contacted with each other. The movable electrodes 531, 532 are arranged in such a spacing that an sufficient isolation is secured during off but driving is possible on low voltage during on. For example, in the case that each movable electrode 531, 532 has a width 2 μm, a thickness 2 μm and a length 500 μm, then the spacing between the movable electrodes 531, 532 is sufficiently 0.6 μm. Incidentally, the movable electrodes 532, 531 must not be both movable electrodes, i.e. it is satisfactory that either one is movable.
In switching from on state to off state, the control voltage is rendered zero, to open the movable electrodes 531, 532 into off state. At this time, chattering takes place whereby the movable electrodes 531, 532 returns to the initial state while vibrating at an eigenfrequency.
Where the movable electrodes 531, 532 fixed at both ends are applied to a switch as in this embodiment, in case the materials forming the substrate and beams are different, internal stress is changed by a difference in thermal expansion coefficient. This relationship is shown in Equation 2. E represents the Young's modulus, Δa the difference in thermal expansion coefficient, and Δt the temperature change. If assuming the beam material A1 and the substrate material Si, then E=77 GPa and Δσ=21×10−6 t l/k results. In the case the temperature is changed from −20° C. to +80° C., internal stress changes 160 MPa and the eigenfrequency changes, as shown in
Δσ=EΔaΔt Eq. 2
It is a general practice, in the control method not using a feedback system, to compute a parameter of control signal on the basis of an eigenfrequency of the beam. In case the control signal optimized at room temperature is used at every temperature, it is impossible to obtain a sufficient chattering preventive effect, i.e. chattering may be increased in a certain case.
Consequently, this embodiment 2 provides a temperature measuring section 510 nearby or within the transmission/reception switching section 501, in order to give an optimal control signal meeting a switch temperature. The temperature measuring section 510 can be configured by a well-known temperature compensation circuit, e.g. a simple temperature compensation circuit utilizing transistor temperature characteristics, as shown in
According to an output signal from the temperature measuring section 510, the control section 506 outputs a control signal matched to a switch temperature. In this case, it is satisfactory to previously store a table having optimal control signals based on temperature so that the control section 506 can output an optimal signal depending upon an operating temperature. Otherwise, an analog circuit may be provided to output an optimal signal.
The optimal control signal is to be computed as follows. Because the movable electrode is applied by a spring force, an electrostatic force and further a damping force, it is possible to compute a position Z of the movable member at time t from the equation of motion as shown in Equation 3. Z represents the position at time t, b the damping coefficient, k the spring constant, Fe the electrostatic force shown in Equation 4. dd shows the electrode-to-electrode distance, S the electrode area and g the electrode-to-electrode distance. Meanwhile, the initial condition of the equation of motion is taken as a speed 0 at time 0 and a position as a latch position.
Md2z(t)/dt+b{1.2−z(t)/g}−3/2dz(t)/dt+kz(t)−Fe=0 Eq. 3
Fe=(½)(εS/dd2)V2, and ZZ′(0)=z(0)=−g Eq. 4
This equation of motion must be determined by a numerical solution instead of a general solution, because it is a nonlinear equation of motion.
Consequently, the present embodiment does not simply render the control signal 0, i.e., after the control signal is rendered 0, the control signal is again applied for a certain time thereby stabilizing the dynamic characteristic of the movable electrode.
It is well known that, generally, in the case to drive the electrode on an electrostatic force, the linear control range of a movable electrode is one-third of a gap. For example, when the gap is 0.6 μm, the linear control range is 0.2 μm. For this reason, the control signal is applied at a time that the spacing between the electrodes becomes 0.2 μm. In
Next, the application voltage is computed. In case applying the potential of a spring in a manner to cancel it all by an applied electrostatic force, an application voltage can be computed from a balance of potential as shown in Equation 5. The potential of the spring is shown in the left-handed term, which is shown by a spring constant k and a displacement amount, i.e. electrode-to-electrode initial gap g. Meanwhile, the potential based on an electrostatic force is shown in the right-handed term, wherein ε represents the dielectric constant, V the application voltage, d the electrode-to-electrode distance, S the electrode area and x the movable range. Because electrostatic force is applied only within a linear range, if g is assumed 0.6 μm, then d is from 0.4 μm to 0.6 μm while x is 0.2 μm. In the case of the above electrode and at room temperature, the application voltage V is 10 V.
(½)kg2(V/dd)2Sx Eq. 5
Next, explanation is made on the movable electrode dynamic characteristic in the case internal stress is changed by a temperature change.
For this reason, similarly to the case at room temperature, the optimal voltage at an elevated temperature is computed by Equation 5. This voltage is applied to the movable electrode.
In the case the switch temperature is lowered, pull-in voltage decreases because of lowered internal stress. Consequently, in case a control voltage same as that at room temperature is applied, the movable electrode, before returning to the initial position, is pulled toward the fixed electrode by the control voltage. For this reason, the optimal voltage for a lowered temperature is computed by Equation 5, which voltage is applied to the movable electrode.
In this manner, it is emphasized to apply an optimal control signal suited for the temperature. This embodiment makes it possible to apply an optimal control voltage for a temperature change.
Incidentally, although the above explanation measured the temperature to compute a change of resonant frequency, the physical amount to be measured may be anything, besides temperature, provided that a change of resonant frequency can be computed. For example, various methods are applicable, including a method to directly read out a change in resonant frequency, a method to compute a resonant frequency from a change in pull-in voltage, a method to compute a change in internal stress from an electrode-to-electrode capacitance change, and a method to directly measure an electrode position.
In using the switch, where the movable member is vibrated at all times, there is a problem that a signal is propagated to the output port with a period of a self-resonance of the movable member. As a switch this problem is solved, shown is a method that two switches are connected in series, for use as one switch.
In order to cut off the signal outputted at a self-resonant frequency of the movable member 2a from the switch 1a, the switch 1b is driven in reverse phase to the switch 1a. Namely, when the signal outputted at an on side of switch 1a reaches the switch 1b, the switch 1b is off. Consequently, the signal outputted from the switch 1a propagates to the ground of the off-side electrode 4b of the switch 4b. In order to drive the switches 1a and 1b reverse in phase, it is satisfactory to make the control signal reverse in phase between the on-side control power supply 5a and off-side control power supply 6a of the switch 1a, and the on-side control power supply 5b and off-side control power supply 6b of the switch 1b.
In the switch of this embodiment, when the switch 1a is on, the switch 1b must be on in order to propagate the signal. When the switch 1a is off, the switch 1b is advantageously placed in an off state in order to enhance isolation.
Incidentally, there is a problem that the control signal of the on-side control power supplies 5a, 5b go on the transmission line, and the control signal further propagates to the output port 8. However, the control signals of the on-side control power supplies 5a, 5b are reverse in phase. In case the switch 1a and the switch 1b are arranged at a sufficient near distance, the both signals offset with each other, causing no problem. Also, as shown in
Meanwhile, there is a problem that direct current flows from the on-side control power supply 5a to the ground for the movable member 2b of the switch 1b. However, this can be solved by connecting a capacitor 14 between the switch 1a and the switch 1b, as shown in
The present drive scheme can be used as a hybrid drive scheme combined with another drive scheme, such as an electrostatic drive scheme, a magnetic force drive scheme, an electromagnetic drive scheme or a piezoelectric drive scheme, enabling to realize a switch higher in performance. For example, it is possible to apply a hybrid drive scheme combining the electrostatic and Lorentz force drive schemes that the movable member 2 and the electrode 9 are attracted to each other by an electrostatic force wherein, only when returning the movable member 2 to a predetermined position, a drive force based on a repellent Lorentz force is provided.
Incidentally, the signal propagation path can be switched over by using a drive force using an attractive and repellent Lorentz force caused by flowing drive currents through the movable member 2 and electrode 9. The two drive currents, if opposite in direction, causes an attractive force upon the movable member 2 and electrode 9, whereby the electrode 9 is attracted to the electrode 9. Meanwhile, in case the drive currents are in the same direction, a repellent force acts between the movable element 2 and electrode 9, whereby the moving member 2 is returned to the predetermined position distant from the electrode 9. The currents are under control of the control power supply 10.
Meanwhile, a high resistive material may be used in either one of the movable member 2 or the electrode 9, to utilize a polarity inversion speed due to a comparatively low carrier mobility of the high resistive material. Due to this, with the movable member 2 and the electrode 9 in contact with by an attractive force, the polarity of the movable member 2 or electrode 9 is inverted in which instance the movable member 2 and the electrode 9 turn into the same polarity to cause a repellent force. This force can be used as a drive force for returning the movable member 2 to a predetermined position.
Otherwise, a high dielectric insulation material comparatively low in polarity inversion speed may be used in an insulation layer to be formed on an electrode between the movable member 2 and the electrode 9. Due to this, with the movable member 2 and the electrode 9 in contact with by an attractive force, the movable member 2 is inverted in polarity in which instance the movable member 2 and the insulation layer surface turn into the same polarity to cause a repellent force. This repellent force can be used as a drive force for returning the movable member 2 to a predetermined position.
These methods enables to increase the response speed for the movable member to return to the predetermined position.
In the mechanical switch, in the case the movable member contacted with the electrode is returned to a predetermined position distant from the electrode where isolation is high not to cause capacitance coupling between the movable member and the electrode, there is a problem of overshoot, i.e. the movable member displaces beyond a predetermined position. This is because, when the movable member is greatly overshot, capacitive coupling takes place on the electrode and movable member as signal propagation paths, forming an incorrect signal path. In order to solve such problems, embodiment 5 is to control the magnitude of an overshoot of the movable member.
Referring to
Next explained as embodiment 6 is another method for controlling the magnitude of movable member overshoot on a switch shown in
To the movable member 2 is applied, as a control signal 151, a pulse signal opposite in direction to and corresponding in magnitude to an overshoot. As the overshoot of movable member 2 becomes greater, the greater control signal 151 is provided so that the movable member 2 can be returned through a stronger force to a predetermined position distant from the electrode 9a. In this case, the direction the force is applied is changed depending upon a vibration direction of movable member 2 due to overshoot. Comparing between the curves 152 and 153, the following is to be understood. Namely, it can be seen that, as compared to a position (curve 152) of the movable member prior to control where the movable member 2 is to return to a predetermined position distant from the electrode 9a by only a spring force of the movable member 2 without a control signal 151, the movable member after being controlled with a control signal 151 is in a position (curve 153) decreased in the vibration amplitude due to overshoot of the movable member 2.
It can be seen that, because the movable member 2 before control has a great spring constant, vibration is caused on the movable member 2 by an overshoot in returning to a predetermined position, as on the curve 154. Consequently, a control signal 151 is applied in order to always apply an asymmetric force of 10:1 to the movable member 2, alternately at the electrode 9a and the electrode 9b thereof. By doing so, the displacement of movable member 2 can be controlled to reduce the magnitude of overshoot and increase the response speed for the movable member 2 to return to a predetermined position. Meanwhile, by asymmetrically applying a force to the movable member 2 depending upon a direction of overshoot, the movable member 2 can be pulled back to a predetermined position by a strong force, reducing the magnitude of overshoot.
Next explained is an embodiment on a method for controlling to relieve the magnitude of an overshoot in one direction of the movable member in a switch shown in
The control signal of embodiment 5-7 makes it possible to control the magnitude of an overshoot of the movable member 2, thus preventing an incorrect signal path from being formed by a capacitance coupling between the movable member 2 and the electrode 9a, 9b. Also, the response speed can be increased for the movable member 2 to return to a predetermined position.
Incidentally, although embodiment 5-7 explained the vibration driving scheme on a movable member by an electrostatic force, the vibration driving scheme may use another driving force, such as a magnetic force.
Meanwhile, the driving scheme may be a hybrid driving scheme combining a plurality of driving schemes discretely or including other driving schemes.
Also, the switch of embodiment 5-7 can be utilized for a switch to drive a movable member in a desired direction, e.g. vertical driving type or horizontal driving type.
Also, the switch of embodiment 5-7 can be utilized for a switch of a multi-output port type, switch as SPDT or SPNT.
Also, the switch of embodiment 5-7 can be mounted on an electronic apparatus in various kinds.
Next, as shown in
Then, as shown in
Next, as shown in
Furthermore, as shown in
Finally, as shown in
Incidentally, although this embodiment used the high resistive silicon substrate 9 as a substrate 201, it may use a usual silicon substrate, a compound semiconductor substrate or an insulation material substrate.
Also, although the silicon oxide film 202, the silicon nitride film 203 and the silicon oxide film 204 were formed as insulation films on the high resistive silicon substrate 201, these insulation films may be omittedly formed where the silicon substrate has a sufficiently high resistance. Meanwhile, on the silicon substrate 201 was formed the three-layer structured insulation film having the silicon oxide film 202, silicon nitride film 203 and silicon oxide film 204. However, in case the silicon nitride film 203 has a film thickness sufficiently greater as compared to the silicon nitride film deposited on the beam, i.e. a film thickness not to vanish even through so-called an etch back process, the silicon oxide film 204 forming process can be omitted.
Incidentally, in this embodiment, as the material forming the beam 208 Al is used. Alternatively, another metal material may be used, such as Mo, Ti, Au, Cu or the like, a semiconductor material introduced with an impurity in a high concentration, e.g. amorphous silicon, or a polymer material having conductivity. Furthermore, although sputtering was used as a film-forming method, forming may be by a CVD process, a plating process or the like.
Incidentally, in the case of attracting the movable member of a mechanical switch by an electrostatic force, the movable member and the electrode may have a contact interface in a wave form, rectangular form or the like. When forming a movable member and an electrode by a plating process, there is a need to form, through the use of a sacrificial layer 205, a gap vertically high in aspect ratio between the movable member and the electrode or an extremely narrow gap between the movable member and the electrode. By making the sacrificial layer 205 in a waveform or rectangular form, the sacrificial layer 205 is made ready to stand, enabling to form a contact interface or gap between the movable member and electrode with higher accuracy. Meanwhile, conventionally, there is a problem that, in a contact interface between the rectangular movable member and electrode, the corner of a convex part is cut into a round or the corner deep in a concave is not accurately cut leaving a sacrificial layer. However, by the structure waveform-rounded in the contact interface between the movable member and the electrode, it is possible to realize an accurate contact interface/gap of movable member and electrode uniformly cut in an etching process on a sacrificial layer 205.
The switch of this embodiment has an increased contact area of the movable member and the electrode, thereby increasing the electrostatic force acting between the movable member and electrode. The switch is high in energy efficiency to generate a greater electrostatic force on the same control voltage, realizing to increase the response speed.
As described above, the present invention can realize switch high-speed response and low driving voltage, and also a relatively wide gap between the movable member and the electrode.
Also, realized is an increase in the response speed for the movable member attracted on the electrode to return to a predetermined position distant from the electrode. Furthermore, it is possible to control the magnitude of overshoot of a movable member.
Meanwhile, it is possible to realize a high-performance switch realizing signal propagation characteristic improvement, high-speed response, low consumption power and low drive power directed toward establishing a great-capacity, high-speed radio communication technology and an electronic apparatus using the same.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2002-217871 | Jul 2002 | JP | national |
| 2003-192682 | Jul 2003 | JP | national |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10625154 | Jul 2003 | US |
| Child | 11063282 | Feb 2005 | US |
