JP5147857B2 - Switchable capacitor array - Google Patents

Description

  The present invention relates to a switchable capacitor array, and more particularly to a switchable capacitor array using a micro electromechanical system (MEMS).

  International Publication WO 2006/054246 describes an apparatus that includes a control matching stage for matching multiple stages such that, for example, in mobile phone applications, the power amplification stage is matched to the antenna stage. The control matching stage includes a switchable MEMS element. The control matching stage may include a capacitor that is switched by the MEMS element. The MEMS element needs to be driven at a relatively high voltage, but driving at the high voltage can be a problem particularly from the viewpoint of the lifetime of the element.

According to the present invention, a switchable capacitor array according to claim 1 is provided.
In a circuit as disclosed in WO 2006/054246, in which a MEMS capacitor is used to switch a radio frequency (RF) signal, a large signal is generated particularly in a high output power and low capacity state. The inventors have found that it appears in MEMS capacitors. In these states, the MEMS element can be self-actuated.

The inventor has discovered that two different types of self-operation are possible.
First, “auto pull-in” occurs. In this case, even if there is no control signal, the MEMS element is pulled in and the element is switched. This is particularly problematic when the RMS (root mean square) voltage of the signal being switched is greater than the pull-in voltage. Second, non-release occurs. In this case, the MEMS element cannot be opened. This occurs when the RMS voltage is greater than the device pullout voltage. Therefore, in any case, it becomes a problem when the signal is high output.

  In order to prevent these problems, it is necessary to use a MEMS element having a relatively high pull-out voltage and pull-in voltage, that is, a relatively high self-operation voltage.

  However, another problem with MEMS devices is the lifetime of the devices. The lifetime of a MEMS device is limited by dielectric charging. This can be minimized by using a relatively low operating voltage.

  Therefore, it is difficult to adjust the switching reliability with respect to the demand for longer life.

  According to the present invention, at least some of the capacitor cells have a plurality of MEMS elements connected in series in order to reduce the RF voltage drop across each element and to reduce the effect of the RF signal passing through the element. Including.

In a more preferred configuration, a drive circuit for driving the switchable capacitor array is further provided. The drive circuit is configured to switch the capacitance of the circuit in order to reduce the possibility of self-actuation in such situations when a predetermined predetermined high power condition occurs .

  For better understanding of the present invention, embodiments will be described with reference to the accompanying drawings.

It is a figure which shows the capacitor array in the 1st Embodiment of this invention. It is a figure which shows the capacitor array in the 2nd Embodiment of this invention. It is a figure which shows the capacitor array in the 3rd Embodiment of this invention. FIG. 4 is a circuit diagram using the capacitor array shown in any of FIGS. It is a figure which shows the driver which drives a capacitor array. It is a timing chart showing a transition period. In embodiment of this invention, it is a figure which shows the drive voltage used in order to drive a MEMS element. In embodiment of this invention, it is a figure which shows the drive voltage used in order to drive a MEMS element. 6 is a timing chart of signals used in the embodiment for controlling the drive circuit. It is a figure which shows the package of embodiment of this invention. It is a figure which shows the capacity | capacitance compensation circuit in embodiment of this invention. It is a figure which shows the capacity | capacitance compensation circuit in other embodiment of this invention. It is a figure which shows the additional circuit in other embodiment of this invention. It is a figure which shows other embodiment of this invention.

  The figures are schematic and not to scale. In the different drawings, the same reference numerals are given to the same or similar components.

  A switchable capacitor array according to a first embodiment of the invention is shown in FIG. This is intended for use when switching RF signals in, for example, mobile phones and similar devices. This array can also be used in other applications.

  The embodiment for the capacitor array 22 uses a two-terminal RF-MEMS switch 38 as the capacitor type. When the RF-MEMS switch is activated, a predetermined capacitance is obtained. When the switch is inactive, the capacitance is much smaller in the inactive state.

  The array includes a plurality of capacitor cells 34 provided in parallel between input node 30 and output node 32. Each cell 34 is configured to provide a given capacitance when switched. In the present embodiment, there are five cells to which capacitances of 0.5 pF, 1 pF, 2 pF, 4 pF, and 8 pF are given, and the 8 pF cell is configured by two branches in parallel. Thus, each cell represents a capacitance corresponding to consecutive bits of the digital capacitor array 22 that represents a minimum of 0.5 pF to 15.5 pF. For example, “10000” represents 8 pF, and “00011” represents 1.5 pF. One control input 42 is provided for each cell, and is assigned codes b0 to b4 representing the least significant bit to the most significant bit, respectively. The control inputs 42 are each connected to the MEMS switch 38 via the bias resistor 43.

  This array is composed of a plurality of RF-MEMS switches 38, each being connected in series with a DC blocking capacitor 40.

  In this array, for a 4 pF cell, a single RF-MEMS switch 38 is used with a DC blocking capacitor 40 connected in series. The 8 pF cell is simply a parallel connection of two 4 pF cells driven in common. However, in the lower bits (2 pF, 1 pF, and 0.5 pF), each cell 34 includes two RF-MEMS switches 38 connected in series, each switch connected in series with a DC blocking capacitor.

  By using two switches connected in series, in use, ½ of the RF voltage between the input and output nodes 30 and 32 is applied to each of the lower-order bits of the RF-MEMS switch 38. This is a non-opening problem, that is, when the control voltage of each control input 42 is lowered to a voltage that turns off the RF-MEMS element, the RF voltage applied to the capacitor turns on the RF-MEMS element. Significantly reduce the risk of leaving.

  The controller 24 (see FIG. 4) is used to determine the voltage to be driven. This controls the driver 44 (see FIG. 5). However, for the purposes of the present invention, it is noted that the controller has additional functions. Specifically, the minimum capacitance value is used in a state where self-operation is very likely to occur. In the present embodiment, it is realized by turning on the cell of the least significant bit despite the mismatch.

  The state where self-actuation is likely to occur specifically includes the highest output state where the maximum RF power is used, and the power represented here depends on the mode and frequency band. The control device 24 detects the presence of these states and turns on the least significant bit.

  In this way, the risk that the least significant bit capacitor is turned on unintentionally is reduced.

  Note that in FIG. 1, the circuit further includes a common bias input 68. This bias input 68 is used to drive the RF-MEMS switch 38 opposite the input 42 with an appropriate DC voltage. Details will be described below.

As an alternative to turning on the least significant bit cell at high power, an additional capacitor is provided that is turned on at high power to ensure minimum capacitance at high power. In this method, since the least significant bit can be used, there is an advantage that the fine adjustment of the capacitance is possible even at high output. This additional capacitor can be thought of as an additional cell. The additional capacitor has a capacitance, but this capacitance need not be the same as the least significant bit capacitance. Therefore, the relationship of 1: 2: 4:... ²ⁿ does not have to be established in all cells.

  Another embodiment will be described with reference to FIG. Instead of the two-terminal capacitive RF-MEMS switch 38 in FIG. 1, a three-terminal capacitive RF-MEMS switch 38 is used. Such a switch is disclosed in WO 2006/117709 and no separate DC blocking capacitor 40 is required as described in this patent document.

  In the configuration shown in FIG. 2, two 3-terminal RF-MEMS switches 38 connected in series are used in 0.5 pF and 1 pF cells, and a single 3-terminal RF-MEMS switch 38 is 2 pF, Used for 4pF and 8pF cells.

By using a plurality of switches 38 connected in series with the least significant bit cell, as in the configuration of FIG. 1, the voltage across the element is reduced, thereby causing a non-opening problem. The risk of voltage of the RF power signal passing between 32 is reduced.

  Furthermore, a driving method in which the least significant bit is turned on at a high output determined as a function of the mode and the frequency band is also applied.

  The configuration of FIG. 2 has further advantages. In the configuration of FIG. 2, the capacitors used for the 1pF, 2pF, 4pF and 8pF cells have the same design as the 2pF cell. A 4 pF cell is obtained by connecting two 2 pF capacitors in parallel. An 8 pF cell can be obtained by connecting four 2 pF cells in parallel. A 1 pF cell uses two capacitors connected in series. The use of this single design increases the accuracy of the capacitance values being 1: 2: 4: 8: 16, especially the accuracy of these capacitance values using the same design.

  The three-terminal capacitive RF-MEMS element has many more advantages. Since the three-terminal MEMS element is effectively configured so that two capacitors are connected in series, the effective RF voltage applied to the MEMS plates is about a given power level and capacitance value. Decrease by a multiple of 2. This further reduces the need for automatic pull-in and non-opening. This means that MEMS switches can be driven at much lower voltages and can be designed to reduce dielectric charging and improve lifetime.

  FIG. 3 is a diagram illustrating another approach. In this approach, the lower bit cell 34 includes two MEMS elements 38 and a DC blocking capacitor 40. Thus, this configuration has half the capacitance of each of the two capacitors connected in series, similar to the cell of the configuration of FIG. This configuration saves chip area and eliminates one of the two bias resistors 43 of FIG. This can reduce circuit loss, particularly when only the least significant bit is used.

  The capacitor array 22 shown in FIG. 1, FIG. 2, or FIG. 3 can be used for adaptive matching, as will be described below with reference to FIGS.

With reference to FIG. 4, a high level functional block diagram of a portion of a portable device is shown.
The portable device includes a plurality of output stages 10, each output stage 10 configured to be able to transmit and receive according to different mobile phone standards. Specifically, the first output stage 12 is configured to transmit and receive according to the UMTS standard, and the second output stage 14 is configured to transmit and receive according to the GSM standard.

  The antenna switching unit 16 includes a plurality of FET switches 18 for switching between output stages. The FET switch 18 is used to switch between transmission and reception when it is necessary to switch between transmission and reception, for example, in the GSM output stage.

  The antenna switching unit 16 is connected to the antenna 20 via the controlled impedance stage 21. The controlled impedance stage 21 provides a switchable variable impedance, as will be described in detail below.

  Accordingly, various output stages 10 are connected to the antenna 20 via the controlled impedance stage 21 and the antenna switching unit 16.

The control device 24 controls the controlled impedance stage 21.
The controlled impedance stage 21 includes a switched capacitor array 22 that includes a plurality of RF-MEMS switch cells according to any of the configurations of FIGS.

  FIG. 5 is a diagram showing details of the driver 44 in the controlled impedance stage 21. The driver is realized on a silicon substrate and designed with many considerations. In particular, it is desirable to reduce the silicon area required for the design.

  In particular, as will be described below, the present embodiment takes into account the need to be able to turn on / off the MEMS device at a high speed and in a manner that reduces spurious emission. A driver for driving a plurality of MEMS elements with both polarities and with one or more specific drive signals is provided.

  The driver is used to drive the control input 42 of the switched capacitor array 22. The driver 44 includes signal inputs 46 and 48. The first signal input 46 receives the transition period signal TP and the second signal input 48 receives the activation / hold signal. The transition period signal TP indicates all events in the transition period, and the operation / holding signal is used to select a high voltage output (60 V) or a low voltage output (30 V).

  The driver also includes additional signal inputs, a voltage selection input 49 and a bridge input 47. Their importance will be explained below.

  Driver 44 includes a charge pump 50 that cooperates with capacitor 52 to provide a 60V output.

  Control switch 54 is used to selectably connect this capacitor and node 56. The nodes 56 are provided in parallel and are connected in sequence to the High side of a plurality of HV drive circuits each including a push-pull circuit 58. The low side of each push-pull circuit is grounded.

  The plurality of HV drive circuits 70 are provided in parallel to drive the plurality of MEMS switches. Each HV drive circuit 70 includes a switch control unit 60 for driving the push-pull circuit 58. Each circuit is connected in parallel between node 56 and ground. A first HV drive circuit 70 (as shown) is connected to a first control input 42 (b0) for driving the capacitor of the first bit. The other four HV drive circuits 70 including the switch control unit 60 and the push-pull circuit 58 are provided to drive other control inputs (b1 to b4). A further HV drive circuit 70 is used to drive a common bias input 68 for the full bridge, as described below. Another further HV driving circuit 70 is used to drive a band switching input 69 (see FIG. 10) described later. Therefore, the single drive circuit 44 is used to drive all of the switchable MEMS.

  The voltage control is performed by connection between the node 56 and the high voltage detector 62 connected to the first signal input 46. The first signal input 46 receives a TP signal for switching between 30V mode and 60V mode. The Low output and High output of the high voltage detection unit 62 are connected to a low voltage control unit 64 whose output is connected to the node 56, and an oscillator 66 that drives the pump 50 via an AND gate circuit 67. The AND gate circuit 67 is connected to the first signal input 46 for driving the pump only when the TP signal is High. The low voltage control unit 64 is realized by a window comparator that drives a capacitor with a pair of charge pumps. One of the charge pumps charges the capacitor positive when the voltage falls below a low window value (eg, 29V), and the other charge pump charges the capacitor when the voltage rises above a high window value (eg, 31V). Configured to discharge. A flip-flop circuit may be used to control the voltage.

  The charge pump 50, the capacitor 52, the high voltage detection unit 62, and the transmitter 66 form a high voltage (60V) control loop for generating 60V. The high voltage detection unit 62 and the 30V control unit 64 form a low voltage (30V) control loop for generating 30V.

  The capacitor 52 is provided in the high voltage loop but not in the low voltage loop. This is because the high voltage is used to actuate (switch on) the MEMS capacitor switch and the low voltage is used to maintain the on state of the MEMS capacitor switch. Therefore, the capacitor 52 needs to be appropriately charged in order to switch the switch at high speed without using the extremely large charge pump 50.

  FIG. 6 is a diagram showing a W-CDMA signal indicating a 50 μs transition period between levels in the UMTS stage 12. The transition period signal TP is High during the 50 μs transition period. It should be noted that when the GSM / EDGE signal is used, the transition period signal TP is High during idle slots.

Several drive patterns performed by the driver 44 are shown in FIG.
The drive pattern shown in FIG. 7a is configured such that the MEMS switch is switched only during the transition period. In the first predetermined period, the switch is driven with a higher voltage (60 V in the example) at the start of the transition period. A longer period may be arbitrarily selected as the transition period. Thereafter, the drive voltage drops to 30 V, which is a voltage that holds the MEMS switch without switching. This predetermined period can be set to, for example, 50 to 500 μs.

  After the predetermined period, the voltage may be lowered to a lower voltage over a second predetermined period. The second predetermined period can be set to 150 to 500 μs, for example. The first and second predetermined periods need not be the same. Actually, the second predetermined period does not need to be controlled, but may be simply a time for the low voltage control unit 64 to reduce the voltage from the high voltage to the low voltage. To avoid transient effects, the transition from high voltage to low voltage should be smooth. For this reason, it is preferable to discharge the current source.

  A further feature of this configuration is that the settling time of the MEMS element can be close to 50 μs. Since this is the length of the transition period, the MEMS element should be turned on at the beginning of the transition period, preferably in the first 5 μs.

  In order to implement this without using the buffer capacitor 52, the area of the charge pump required on the semiconductor substrate becomes very large. Instead, by using a 1-2 nF buffer capacitor 52, the on-state can be reasonably and stably achieved until the end of the transition period, so that a 200 pF MEMS element can be charged in 5 μs.

  Note that, as shown in FIG. 5, in contrast to the high voltage loop, a similar buffer capacitor for holding node 56 at a low voltage of 30V is not provided. This is because rapid changes to this voltage are not necessary. This reduces the area of silicon occupied by the drive circuit 44. Instead, the low voltage control loop maintains a near correct voltage. This is accomplished by a window comparator that turns on the current source when the voltage is too low and turns on the current sink when the voltage is too high. Since the leakage current is only about 5 nA (typically less than 10 nA), the current source need not be able to generate a large current.

  The amount of current discharged through the current sink sets the time until the high voltage is lowered to the low voltage. When this is very short, individual selective discharge can be performed.

  Note that the time during which the voltage is held at a high voltage and the time for changing to a low voltage are both much longer than the transition period. For example, the MEMS element is stable after 50 μs, but it is not particularly necessary for the switch AH to supply a low voltage after 50 μs, and switching is performed, for example, after 200 μs. The start timing of the high voltage is important, and the end timing is not important.

  By using this holding voltage of 30 V, the lifetime of the MEMS switch is increased as compared with the case where this pattern is not used.

  It should be noted that a MEMS switch of the type described here can be turned off by applying the same voltage on both sides of the device. That is, in the configuration of FIG. 2, the same voltage (ground voltage or high / low voltage) is supplied to the common bias terminal 68 and the respective control inputs 42. Similarly, to turn on these MEMS switches, a high voltage is applied to one common bias terminal and ground is applied to the other common bias terminal and the respective control inputs 42.

FIG. 8 is a diagram showing signals given to the driver shown in FIG. 5 in order to obtain the result of FIG. The TP signal applied to input 46 indicates a transition period. The AH signal applied to the input 48 effectively indicates that the driver 44 generates a high voltage when AH is High and generates a low voltage when AH is Low.

  First, consider how the AH, TP and VS signals provide high or low voltage.

  The VS signal applied to input 49 simply determines a high voltage value (eg, 60V) and a low voltage value, which in this embodiment is half the high voltage.

  The first TP pulse in FIG. 8 turns on the selected push-pull circuit 58 in combination with the AH signal, and selects the selectable capacitor array 22 as shown in FIGS. 7a and 7b. The control input 42 and the common bias input 68 are driven.

  Note that after the TP signal becomes Low (after the end of the transition period) and the AH is still High, the low voltage control loop functions to hold the high voltage of 60V. Only when AH becomes Low, in the example, after 200 μs from the start of the transition period, the HV detection circuit 62 operates to detect a low voltage, and the low voltage control unit 64 controls the voltage of the node 56 to a low voltage. To be switched.

  This is used to recharge capacitor 52 during periods of no switching. For example, TP becomes High every 0.005 seconds (that is, at a frequency of 200 Hz), the capacitor 52 is recharged, and the state of being charged at a high voltage is maintained. AH only goes High every 0.1 seconds to switch MEMS elements.

  When TP becomes Low again, the switch 71 is closed, and the low voltage control loop of the HV detection circuit 62 and the low voltage controller 64 drives the node 56 again at a low voltage. Note that by preventing the capacitor from holding 30 V in this loop, an additional switch for disconnecting the capacitor during the charging operation of the capacitor 52 becomes unnecessary.

Second, consider the effect of the bridge signal BR. In order to realize the drive pattern of FIG. 7a, a bridge signal BR is applied to the bridge input 47 and at the start of the transition period, ie
Switching is performed at substantially the same timing as the AH and TP signals. As shown in FIG. 8, the bridge signal BR is switched from Low to High, or from High to Low.

  A state in which the control input 42 of the MEMS element that is turned on is held in the ground state and the common bias 68 is the drive voltage generated by the driver 44 is a starting point. The drive voltage is a low voltage of 30V at this point in the cycle. The MEMS element that is off is at the same voltage for both the control input and the common bias input, and is at a low voltage of 30 V at this point in the cycle.

  After the switching of the BR signal, the common bias 68 is maintained in the ground state, and the control input of the element that is turned on is connected to the drive voltage generated by the driver 44. This driving voltage becomes a high voltage of 60 V due to the simultaneous change of the AH signal. The high voltage starts to decrease after 200 μs, and in the illustrated embodiment, reaches the low voltage of 30 V after another 200 μs.

  The MEMS element that is off is at the same voltage for both the control input and the bias input, and at this point in the cycle is 0V of ground voltage.

  At the next switching, i.e. the start of the next transition period, AH, BR and TP are again switched simultaneously. The common bias 68 is held at a high voltage of 60 V, and the control input of the turned-on element is in the ground state.

  The MEMS element that is off is at the same voltage for both the control input and the bias input, and is at a high voltage of 60V at this point in the cycle.

  The high voltage starts to decrease after 200 μs, and in the illustrated embodiment, reaches the low voltage of 30 V after another 200 μs. This returns to the starting point.

  Therefore, by using the circuit as described above, the polarity of the voltage is reversed every cycle, thereby increasing the lifetime of the device.

  Further, in FIG. 8, the control input can also be switched at the same timing as the switches AH, BR, and TP that change the MEMS element held on at the start of each transition period.

  The identification of the control input 42 and the common bias input 68 is determined by the type of the MEMS element. In the case of a two-terminal element, these two inputs correspond to the two terminals as they are. In the case of a three-terminal element, the central input is the control input, and the remaining both terminals can be used as the common bias input 68.

  In the above description, the case where the MEMS switch is turned on has been described. The MEMS switch is switched off by being driven with the appropriate voltage at the same point in the cycle, ie at the beginning of the transition period (eg during the first 5 μs of this period). Since dielectric charging is not performed in the off state, the voltage applied to the element is switched to 0 V, and the voltage may be held in that state.

  By switching the switched capacitor array off and on only during the transition period, the effects of high voltage transients are greatly reduced. Otherwise, transients that occur when the MEMS switch is switched can easily cause spurious emissions.

  Another effect that causes spurious emissions is the effect of the high voltage clock output of the oscillator 66 that generates the ripple. This ripple can be mixed with the RF carrier and cause spurious emissions.

  In the present embodiment, the oscillator 66 that drives the charge pump 50 is switched only during the transition period (ie, from on to off or vice versa). By activating the charge pump only during the transition period, spurious emissions from the charge pump can be minimized. Specifically, the transmitter 66 that drives the pump is activated only when the TP signal is High. For the remainder of the transition period, the capacitor 52 remains constant at 60V and the clock signal is not activated except during the transition period. Note that when the clock signal is activated for a time shorter than the full time of, for example, 200 μs, the capacitor 52 is set to 60 V for 200 μs in order to prevent a sudden change of voltage that causes a transient phenomenon by itself. In the next 200 μs that is applied and subsequently, the voltage is reduced from 60V to 30V.

  A further advantageous effect of this embodiment is that if this embodiment is not used, a change in the transfer function during adaptation due to switching of the switchable capacitor array 22 can cause distortion of the output signal. .

  Therefore, in this embodiment, to reduce spurious emissions from the clock, the clock is only operated during the transition period, and the ripple of the operating bias voltage is reduced, thereby reducing the ripple that is mixed into the RF carrier. In order to reduce the spurious emission, a combination of driving the MEMS element with a high voltage for a longer period is performed.

  Furthermore, the lifetime of the MEMS element is increased by supplying a low driving voltage (30 V) for maintaining the state in which the MEMS element is operated.

  In the above embodiment, the high voltage is 60V and the low voltage is 30V. However, this voltage selection is based on the use of a specific MEMS element, and this selection may be changed.

  The MEMS element generally has some hysteresis, and a pull-in voltage that is a voltage necessary for activating the switch and a pull-out voltage that is a low voltage necessary for deactivating the switch are provided. is there. The driver needs to generate a voltage sufficiently higher than the pull-in voltage and a low voltage between the pull-in voltage and the pull-out voltage.

  The drive circuit 44 has a voltage selection input 49 for supplying a voltage input that specifies a high voltage. Conveniently, this voltage is a scaled voltage, for example a voltage of 2.8V is supplied to the voltage selection input 49 for 60V output and 1.4V is input for 30V output. . In this way, the drive circuit can generate different voltages simply by adjusting this voltage. However, this input is arbitrary and the voltage selection input 49 may be omitted in the production device.

  The RF-MEMS switch 38 of the switchable capacitor array 22 is sealed. Typically, this is realized by covering the common substrate 84 with a cap 82 and sealing with a solder ring that functions as a sealing ring, as shown in FIG.

  The bias resistor 43 and the MEMS switch 38 (see FIG. 2) are integrated on the common substrate 84. Further, the detection inductor may be similarly integrated.

All of this reduces the number of wires crossing the sealing solder ring 86 and minimizes the influence of the solder ring 86 on device performance.

  Input 30 and output 32 are placed in close proximity to minimize self-inductance. In addition, the largest capacitor in the upper bits is also placed close to the input 30 and output 32 to minimize self-inductance.

  Optionally, the circuit may include additional elements to cope with the large capacitance to ground constituted by the solder ring 86. Since this is on the order of 0.3 to 0.5 pF, it is comparable to the least significant bit. In general, the circuit may need to operate at very different frequencies, and any matching to compensate for the capacitance to the solder ring ground must work in any frequency band. There is. To that end, the circuit may use elements that are switched based on a band signal input to terminal 69 of switchable capacitor array 22 (see FIG. 3).

One approach is to use an inductor in series with the switching capacitor, but in this case a large RF-MEMS capacitor such as a switching capacitor having a very large area of about 1.5 mm ² , for example. A switch is needed. Furthermore, large RF-MEMS capacitor switches have a self-resonant frequency that is inappropriately close to the high band of operation.

  Therefore, the preferred embodiment uses a switchable parallel LC circuit to remove unwanted parasitics by resonance. FIG. 10 is a diagram showing a configuration example including two switchable LC compensation circuits 90, one being arranged between the output node 32 and the ground, and the other being arranged between the input signal path and the ground. Is done.

  In order to expand the impedance compensation range in a plurality of frequency bands, a switchable parallel LC array 92 using a multi-switch for obtaining a switchable impedance is provided.

  FIG. 11 shows another configuration example in which the switchable parallel LC array 92 of FIG. 8 is replaced with a non-switchable parallel LC array 94 in which an inductor and a capacitor are connected in parallel to the input signal path. Non-switchable parallel LC array 94 is configured to self-resonate between the low and high bands so that it operates as an inductor in the low band and as a capacitor in the high band. This allows both low and high band operation without adding a plurality of switches used in the LC array 92 shown in FIG.

  A schematic diagram of still another embodiment is shown in FIG. For clarity, some components of all elements are omitted. The omitted constituent elements are the same as those in the above-described embodiment.

  In this embodiment, an additional filter function is provided to reduce the effects of switching transients.

  The first additional filter is constituted by a resistor 71 between the capacitor 52 and the push-pull circuit 58. This filter cooperates with the capacitor 52 to reduce the slope of the bias pulse when the push-pull circuit is turned on.

  Second, a series low-pass filter 72, as shown as an RC filter in the embodiment, is provided between the push-pull circuit 58 and the switchable capacitor array 22.

  Third, the push-pull array transistors are reduced in size to increase the transistor resistance. This also has the effect of a low pass filter.

  Fourth, grounded shunt inductors 74 and 76 are provided along the RF path to reduce transients. This is also an indispensable path for impedance matching in order to match the impedance of the antenna 20 and the drive circuit. Further, the shunt inductor 74 between the antenna 20 and the switchable capacitor array 22 reduces the transmission of electrostatic discharge phenomena returning from the antenna 20 to the switchable capacitor array 22 and the driver circuit. This relaxes the requirements of these circuit elements in this respect.

  Fifth, a high-pass filter is formed between the antenna switch 16 and the switchable capacitor array 22 by the shunt resistor 78 grounded in combination with the DC blocking capacitor 80.

  By using some or all of these additional filter techniques, the impact of high voltage switching that causes transients to propagate through the network can be reduced.

  This is a specific problem when the antenna switch 16 that is simply realized by a p-type high electron mobility transistor (HEMT) that can be switched at a voltage of about 2.5 V is used. The 60V switching signal for switching the MEMS capacitors of the capacitor array 22 can easily interfere with these HEMTs and cause spurious effects.

  In the embodiment of FIG. 12, the slope of the bias pulse is lowered by the low-pass filter by the resistor 71, the low-pass filter 72, and the high-resistance transistor of the push-pull circuit 58. Further improvements can be made by providing a high pass filter in the RF signal line.

  It is not necessary to use fixed high and low voltages such as 30V and 60V to drive the MEMS element, but as an alternative, a specifically determined voltage can be used, and the same voltage is applied to each element. May be used in parallel, or different voltages may be used for each element.

  In the above-described embodiment, the high voltage (60 V) is an input at the voltage selection input 49, and the voltage selection input 49 selects the correct voltage.

  In yet another embodiment, the driver uses a special calibration mode that gradually increases the voltage applied to the MEMS switch while measuring the capacitance of the switch as shown in FIG. Have. For this, a voltage selection input 49 driven by the control device 24 (see FIG. 2) and a capacitance measuring unit 80 are used. The switch is switched and the voltage at which the capacitance increases is measured. Thereafter, the voltage is gradually decreased until the switch is turned off. The voltage at which the switch is switched is used as the high voltage, and a voltage slightly higher than the voltage at which the switch is turned off is used as the low voltage.

  A special calibration mode is performed at start-up, as an alternative or in addition to it at specific time intervals, or after a certain amount of usage.

The above-described embodiments are not exhaustive, and those skilled in the art who are familiar with the technology in this field can use other necessary elements and configurations. Even if it is recommended that a specific element be used for a specific function in the description of this embodiment mode or circuit diagram, a person skilled in the art who is familiar with the technology in this field You will notice that alternative elements can be used.

  Although the switchable capacitor array of FIGS. 1-3 has been described above in a particular application for mobile phones, the switchable capacitor array described above is applicable to other load-lines. It may be used in applications and other configurable RF networks.

Claims (9)

A switchable capacitor array (22) comprising:
Comprising a plurality of capacitor cells (34) connected in parallel between an input (30) and an output (32);
Each said capacitor cell is
At least one MEMS switch (38) for switching the capacitor cell;
At least one control input for receiving a control signal for switching said or each MEMS switch;
Each cell (34) has a respective capacitance value that is the capacitance value of the cell when the cell is activated;
The capacitance value increases from the lowest value to the highest value and changes such that at least one cell has a low capacitance value and at least one cell has a high capacitance value;
Each cell with a low capacitance value has a capacitance value lower than a predetermined value,
Each cell with a high capacitance value has a capacitance value higher than the predetermined value,
Each capacitor cell with a capacitance value up to the predetermined value uses two MEMS switches (38) connected in series,
Each capacitor cell having a capacitance value greater than the predetermined value uses a single MEMS switch (38) , a switchable capacitor array. The ratio of the minimum capacitance value to a maximum capacitance value is substantially 1: 2: ...: containing 2m-1 to become the m capacitor cells (34), can be switched according to 請 Motomeko 1 Capacitor array. Each said MEMS switch (38) is 3 a terminal capacitive RF-MEMS switches, switchable capacitor array according to any one of claims 1 or 2. The switchable capacitor array according to claim 1 or 2 , wherein each MEMS switch (38) is a two-terminal capacitive RF-MEMS switch. Switchable capacitor array (22) according to any one of claims 1 to 4 ,
A variable impedance circuit comprising a driver (44) connected to the control input of the capacitor cell (38) and configured to control the capacitor cell to select a variable capacitance. The driver (44) determines a predetermined high output state of a signal passing between the input (30) and the output (32) of the switchable capacitor array (22), and the predetermined The variable impedance circuit according to claim 5 , wherein a predetermined capacitor between input and output is switched when it is determined that the high output state is determined. The predetermined high output state is an RF output above a certain level,
The variable impedance circuit of claim 6 , wherein the level is a function of the mode and / or frequency band used. The variable impedance circuit according to claim 6 or 7 , wherein the predetermined capacitor is a cell (34) having a lowest capacitance value that is switched on when the predetermined high power state is determined. A method of operating the switched capacitor array (22), comprising:
The switched capacitor array is:
A plurality of capacitor cells (34) of different capacitance values increasing in parallel from the lowest value to the highest value , connected in parallel between the input (30) and the output (32);
Each said capacitor cell is
Each capacitance value,
At least one MEMS switch for switching the capacitor cell;
At least one of the low capacitance capacitor cells uses at least two MEMS switches connected in series;
Each high capacitance capacitor cell uses a single MEMS switch,
The operating method is:
Determining a predetermined high power state of a signal passing between the input (30) and the output (32) of the switchable capacitor array (22);
Switching the predetermined capacitor (34) between input and output when the predetermined high output state is determined.

Images