Atty, Docket. Mo.: 016904 /PCT/02/ETCH/TAHOE/MDD

THREE-COIL INDUCTIVELY COUPLED PLASMA SOURCE

WITH INDIVIDUALLY CONTROLLED COIL CURRENTS FROM A SINGLE RF POWER GENERATOR

CROSS-REFERENCE O' RELATED APPLICATIONS

[0011 This application claims priority of U.S.

Application Serial No, 13/863,614 filed April 16, 2013 entitled THREE-COIL INDUCTIVELY COUPLED PLASMA SOURCE WITH INDIVIDUALLY CONTROLLED COIL CURRENTS FROM A SINGLE RE PO¾ER GENERATOR, by Leonid Dorf, et al., which is a continuation-in-part of U.S. Application Serial No,

13/836,890 filed March 15, 2013 entitled THREE-COIL

INDUCTIVELY COUPLED PLASMA SOURCE WITH INDIVIDUALLY CONTROLLED COIL CURRENTS FROM A SINGLE RF POWER GENERATOR, by Leonid Dorf, et al., which claims the benefit of U.S. Provisional Application Serial Ho.

61/637,468 filed April 24, 2012 entitled THREE-COIL

INDUCTIVELY COUPLED PLASMA SOURCE WITH INDIVIDUALLY CONTROLLED COIL CURRENTS FROM A SINGLE RF POWER GENERATOR, by Leonid Dorf, et al.

BACKGROUND

[0021 The recent growth in size semiconductor wafers in integrated circuit fabrication is making it more difficult to obtain the needed degree of uniformity of plasma process rate across the treated surface of the wafer. The process rate ma be an etch rate or a

deposition rate, for example. Plasma processing tools for processing 300 mm diameter wafers employ a pair of concentric coils {coii antennas) , namely an inner coil and an outer coil, to control uniformity of radial distribution of process rate. The RF power levels applied to the different coils are selected to counteract non-uniformities . For 450 mm diameter wafers, we feel that three separately powered RF coils are needed.

Separate adjustment of the different RF power levels delivered to the three different coils would provide finer control of uniformity. Impedance matching is required for all three coils. The problem is that operation of the impedance match and control of the different power levels delivered to the different coils is unpredictable and complicated by internal resonances and mutual inductances among the three coils and between the coils and the plasma. A related problem is that varying the three different power levels in an effort to improve uniformity may cause the system to enter resonant modes, in which plasma instabilities can arise.

Resonances also cause very high RF voltages to appear on the coils and their circuit elements,, which can cause arcing .

SUMMARY

[0Q3J A plasma reactor for processing a workplace includes an RF power generator, an impedance match coupled to the RF power generator, and. three coil antennas having respective driven ends and return ends, the return ends being connected to a common potential . Three current divider branches are coupled, between the impedance match and the driven end of respective ones of the coil antennas, a pair of the three current divider branches comprising respective variable impedance

elements. A user interface is provided, A current apportionmen controller controls impedances of the variable impedance elements of the pair of current divider branches in response to a desired apportionment of currents to the three coil antennas defined at the user interface.

[0043 In one embodiment, the current apportionment controller comprises a look-up table storing a sequence of pairs of impedanc values of the variable impedance elements . The sequence of pairs of values defines a linear programming sequence. In one embodiment, the user interface is adapted to record a user-selected point along the path and. transmit the user-selected point to the current apportionment controller.

[005] In one embodiment, the sequence of pairs of values defines a path in a 2~dimensional space whose diiTsensions are the impedance values of the variable impedance elements. In a related embodiment, the path includes: (a) a reference point at which currents in the three coil antennas are at least approximately equal, {b} a first point at which current in a first one of the coil antennas, relative to currents a second and third one of the coil antennas, is near a maximum, (c) a second point at which current in the second coil antenna, relative to currents in the first and third coil antennas, is near a maximum, and (d) a third point at which current in the third coil antenna, relative to currents in the first and second coil antennas, is near a maximum.

[0063 In one embodiment, the apportionment controller comprises: (a) a memory containing dat representing ^' respective currents in the three coil antennas relative to total current as functions of values of the variable impedance elements, and (fo) a processor for determining from the data the impedance values of the variable impedance elements at which the respective currents in the three coil antennas at least nearly correspond to the desired apportionment.

[0073 In accordance with another aspect, a method is provided for individually controlling currents in three coil antennas of a plasma reactor supplied from an RF generator through an RF impedance match. The method includes: (a) providing respective current divider branches between the impedance match and respective ones of the coil antennas,, each of at least two of the current divider branches comprising a respective variable

impedance element; (b) providing a model comprising a first set of equations defining voltage drops in the current divider branches in accordance with mutual inductances of the coil antennas and corresponding portions of a plasma in the reactor and a second set of equations defining voltage drops in the corresponding portions of the plasma in accordance with the mutual inductances (c) solving the first and second sets of equations for currents in the coil antennas relative to total current as 3-dimensionai functions of impedance values of the variable impedance elements in the at least two current divider branches; and (d) correlating

information corresponding to the functions to a user- defined apportionment of currents in the coil antennas.

[008] In one embodiment of the method, the

correlating comprises: (a) searching the functions for respective points in a space representing pairs of values of the variable impedance elements at which coil currents are maximum for respective ones of the coil antennas; (b) defining a linear programming path that includes the respective points; and {c) enabling a user to select a location on the linear programming path to control

impedance values of the variable impedance elements.

[0093 In a related embodiment, the correlating further comprises searching the functions for a set of impedance values of the variable impedance elements corresponding to an appo tionment of currents among the coil antennae that at least nearly approximates a user- defined apportionment.

[0010] in accordance with a yet further aspect, a method is provided for individually controlling currents in three coil antennas of a plasma reactor supplied from an RF generator through an RF impedance match . The method includes; (a) providing respective current divider branches between the impedance match and respective ones of the coii antennas, at least two of the current divider branches comprising respective variable impedance

elements; and {fa) correlating information, corresponding to coil antenna currents as 3-dimens ona.l functions of impedance values of the variable impedance elements, to a user-defined apportionment of currents in the coil antennas .

[0011] In one embodiment, the correlating comprises: (a) defining respective points in a space representing pairs of values of the variable impedance elements at which coil antenna currents are maximum for respective ones of the coil antennas; (fo) defining a linear

programming path that includes the respective points; and (c) enabling a user to select a location on the linear programming path to control impedance values of the variable impedance elements .

BRIEF DESCRIPTION OF THE DRAWINGS

[0012] So that the xaanner in which the exemplary embodiments of the present invention are attained can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to the embodiments thereof which are

illustrated in the appended drawings. It is to be appreciated that certain well known processes are not discussed herein in order to not obscure the invention,

[0013] FIGS. 1A, IB and IC depict a plasma reactor in accordance with one embodiment,

[001 J FIG, 2 is a schematic block diagram depicting an embodiment of a 3~port controller of FIG, I A,

[0015] FIG, 3A is a schematic diagram depicting internal structure in one embodiment of an impedance xaatch of FIG. 1A .

[0016] FIG. 3B is a schematic diagram depicting internal structure in one embodiment of the three port controller of FIG. 2.

[0017] FIG. 4 is a schematic block diagram depicting architecture of an embodiment of a processor of the three port controller of FIG. 2,

[0018] FIG. 5 depicts a matrix of mutual inductances in a model employed in the processor of FIG. 4. C0019J FIG. 6 depicts a definition of three plasma sections aligned with the three RF coil antennas for modeling plasma elements in the matrix of FIG. 5,

[0020] FIG. 7 illustrates an example of the outer coil current s a 3-dimensional function of the values of two variable capacitors in a current divider of FIGS. 3B, relative to total current in all three coils.

[0021] FIG. 8 illustrates an example of the middle coil current as a 3-dimensional function of the values of two variable capacitors in the current divider of FIGS , 3B, relative to total current in all three coils.

[0022] FIG. 9 illustrates an example of inner outer coil current as a 3-dimensionai function of the values of two variable capacitors in a current divider of FIGS, 33, relative to total current in all three coils.

[0023] FIG, 10 depicts a method for programming and operating a programmable path processor,

[0024] FIG. 11 is a graph depicting a linear

programming path defining a scala p ocframming parameter employed in one embodiment of the processor of FIG, 4♦

[0025] FIG, 12A is a graph depicting the current in the outer coil as a function of position along the programming path of FIG. 10, relative to total current in all three coils.

[0026] FIG, 12B is a graph depicting the magnitude of the current in the middle coil as a function of position along the programming path of FIG. 10, relative to total curren in all three coils.

C0027J FIG, 12C is a graph depicting the current in the inner coil as a function of position along the programming path of FIG. 10, relative to total current i all three coils.

[0028] FIG. 13 is a schematic block diagram of an alternative embodiment of the processor of FIG. 4 that employs a search operation instead of a linear

programming path.

[0029] FIG. 14 depicts a modification of the

embodiment of FIG. 1A in which the outermost coil antenna is a side coil antenna.

[00303 FIG . 15 is a simplified block diagram depicting a general form of the 3-port controller of FIG. 33, configurable in numerous operable species, of which FIG. 3E corresponds to one of the species.

[0031] FIG. 16 is a block diagram depicting another one of the operable species of FIG. 15.

[00323 FIG. 17 is a block diagram depicting variations of the embodiment of FIG. 15 in which one or more

optional series impedance elements are introduced.

[0033J To facilitate understanding, identical

reference numerals have been used, where possible, to designate identical elements that are common to the figures. It is contemplated that elements and features of one embodiment may be beneficially incorporated in other embodiments without further recitation. It is to be noted, however, that the appended drawings illustrate only exemplary embodiments of this invention and are therefore not to be considered limiting of its scope, for the invention may admit to other equall effective embodiments .

DETAILED DESCRIPTION

[0034] Referring to FIGS, 1Ά and IB, a plasma reactor includes a vacuum chamber enclosure 100 including

cylindrical side wall 102, a disk shaped ceiling 104 and a chamber floor 106. A support pedestal 108 extends through the floor and holds a workplace support 110 having a workplace support surface 1.12 and an internal workplace support electrode 114 beneath and insulated from the support surface 112. A planar workplace 116, such as a sendconduc or wafer, a photolithographic mask, a precursor for a plasraa display or a solar cell array, may be held on the support surface 112 for plasraa

processing. Three (or more) RF coil antennas 120, 122, 124, are disposed over the ceiling 104 in alignment with respective concentric circular (annular) zones 126, 128, 130 (FIG. IC) coaxial with an axis of symmetry of the chamber enclosure 100.

{00353 In one embodiment, an RF bias power generator 140 is coupled, through a RF bias impedance match 142 to the workpiece support electrode 114.

[0036J An RF source power generator 150 has an output terminal coupled, through an RF impedance match 152 to a three port controller 160 and return terminal connected to RF ground. The impedance match 152 consists of a match tuning circuit 154 and a match controller 156 governing the match tuning circuit 154 in accordance with the output of a standing wave ratio (S R) detector 158 disposed at or in series with the output of the RF source power generator 150, The SWR detector 158 measures the ratio between forward and reflected power or voltage, and may be a conventional RF device such as a dual

directional coupler, for example, The three port

controller 160 has three outputs 160a, 160b and 160c coupled to the three RF coils 120, 122 and 124

respectively. The currents provided at the outputs 160a, 160b and 160c are separately adjustable relative to total current, as will be described below.

[0037] Referring to FIG. 2, the three port controller 160 includes a variable reactance current divider 162 coupled between the output of the impedance match 152 and the outputs 160a, 160b and 160c, As will be described in detail below, the variable reactance current divider 162 includes plural variable reactive elements (not shown in PIG. 2} such as variable capacitors or variable

inductors, whose adjustment controls the apportionment of currents at the three outputs 160a, 160b and 160c. This adjustment is performed under control of a processor 164, in response to user preferences for apportionment of currents among the three RF coil antennas 120, 122 and 124, As will be described in detail below, the processor 164 may translate a user preference for apportionment of the three currents to specific values of the variable reactances of the variable reactance current divider 162. A variable reactance controller 166 may convert the specific variable reactance values furnished by the processor 164 to corresponding commands that produce the desired reactance values in the variable reactance current divider 162, For example, the variable reactance elements may be capacitors with stepper motors governed by the commands . In the most practical embodiment , the user may control the variable reactances by following a pre-established linear programming path that covers nearly all possibilities,, as will be described later in this specification.

[0038] An embodiment of the match tuning circuit 154 is depicted in FIG. 3&, and an embodiment of the variable reactance current divider 162 is depicted in FIG. 33. The match tuning circuit 154 of FIG. 3A includes a variable load capacitor CI connected in series with a series inductor Llser between the RF source power

generator 150 and a power input. 161 of the three port, controller 160. A parallel inductor LIpar is connected across the series combination of the load capacitor CI and the inductor Llser. The match tuning circuit ISA further includes a variable tune capacitor C2 coupled between the output terrains! of the RF generator 150 and RF ground through a series inductor L2.

[0039J The variable reactance current divider 162 has three parallel branches 162-1, 162-2 and 162-3 coupled between the power input 161 and the three outputs 160a, 160b and 160c to the respective RF coils 120, 122 and 124, The first branch 162-1 has no discrete impedance elements and is simply a conductor providing in effect a D.C. short. The second branch 162-2 consists of a series reactive element connected in series with the middle coil 122 and a parallel variable reactive element connected, in parallel wi th the middle coil 122. In the illustrated embodiment of the second branch 162-2, the series

reactive element is a capacitor C21 while the parallel variable reactive element is a variable capacitor C22. Similarly, the third branch 1 2-3 consists of a series reactive element connected in series with the inner coil 124 and a parallel variable reactive element connected in parallel with the inner coil 124. In the illustrated embodiment of the th rd branch 162-3, the series reactive element is a capacitor C31 while the parallel variable reactive element is a variable capacitor €32. In the embodiment of FIG. ^" :■K , the variable reactance controller 166 includes a first controller 166-1 that translates a first capacitance value received from the processor 164 to a command for the variable capacitor 022, and a second controller 166-2 that translates another capacitance value received from the processor 164 to command for the variable capacitor C32.

[0040] In the embodiment of FIG, 4, the processor 164 may optionally include a mathematical model 170 of the system including the variable reactance current divider 162, and an equation solver 172 adapted to employ conventional numerical methods to solve the simultaneous equations constituting the mathematical model 170. The results are stored in look-up tables, as will be

described below. The look-up tables may be used to obtain the best values for the variable capacitors C22 and C32 to realize a desired apportionment of currents among the different coils. As plasma conditions

(particularly pressure) change, the values predicted using the results stored in the look-up tables become less accurate. But modest inaccuracies can be

compensated by the user employing methods described below, without redoing the calculations involving the model. However, if plasma conditions change

significantly or radically, then a new set of results must be obtained from the model (using conventional numerical methods to solve the simultaneous equations constituting the mathematical model 170) , based upon the latest values of plasma conditions {e.g.. , chamber pressure). The look-up tables are then updated.

[0041] The mathematical model 170 is derived in accordance with the topology of th current divider 162 and the plasma, and is as follows:

Zpip ⁺ i* Mpiiro

V _c is the voltage at the power input 161.

[0042] The foregoing model is based upon current functions including the complex term et ^xot , so that all current functions, ϊχ and. I _p are actually the

corresponding complex amplitudes. For any current function, the actual current that corresponds to it is given by Rm{I± e ^xe,t ) . The useful result of solving the model 170 is the magnitudes of the three coil currents for i — I, 2 _f 3, which are denoted simply as l _lf X _Xl X ₃ , in bold font. [0043] The matrix M is depicted in FIG, 5, and is symmetrical across its diagonal, and only the lower diagonal half is depicted in FIG. 5 for the sake of simplicity. The off-diagonal elements in the matrix in the foregoing raodel are the mutual inductances among the three RF coils 120, 122 and 124 and the three sections of the plasma in the reactor chamber 100 underlying the three circular /annular zones 126,, 128 and 130 under the respective coils. Elements on the matrix diagonal are self-inductances. The division of the plasma into three corresponding solid sections 126' , 128' ana 130' is depicted in PIG. 6. The subscripted indices of the matrix elements in M denote the elements mutually coupled as follows : the three coils 120, 122 and 124 correspond to indices 1, 2 and 3 respectively; the three plasma sections 126'',. 128' ana 130' correspond to indices , 5 and 6, respectively. Each of the matrix elements is computed based upon the geometry of the reactor. In turn, the plasma impedance, , depends upon the pressure in the chamber containing the plasma.

[004 J Equation (1) expresses the voltage drops attributable to current in the first coil 120 and the mutual inductances involving the first coil 120.

Equation (2) expresses the voltage drops attributable to current in the second coil 122 and the mutual inductances involving the second coil 122. Equation (3) expresses the voltage drops attributable to current in the third coil 124 and the mutual inductances involving the third coil 124. Equation (4) expresses, in a compressed manner, the voltage drops att ibutable to individual currents through the plasma sections 126', 128' and 130' and the mutual inductances involving the respective plasma sections. [0045] Equation (4) is actually three equations , one for each value of the subscripted index p, ranging from 4 to 6, corresponding to the three plasma sections 126' , 123' and 130'. in equation (4) , the Z _p is the complex plasma impedance, and is computed f om the plasma

resistance R _p as follows: where j- (-l) ' , ω is the angular frequency of the RF source power generator 150, and v is the mean electron collision frequency in the plasma . The plasma resistance R _p of each plasma section and the collision frequency v is determined from plasma conditions in accordance with conventional principles. In particular, the collision frequency v is determined from the pressure of chamber containing the plasma. Thus, the complex plasma

impedance is a function of pressure. The model provides accurate results if an accurate value of chamber pressure was used to determine the plasma resistance R _p and the model may continue to be used until a significant change in chamber pressure is sensed by the processor 164. When that occurs, the equations of the model must be updated according to the latest values of plasma conditions

(e.g., pressure),, and solved. The processor 164 may monitor chamber pressure using a pressure sensor (not illustrated) coupled to or inside of the chamber, to determine whether the model needs to be updated.

[0046] Returning now to FIG. 4, in one embodiment, the equation solver 172 is adapted to solve the simultaneous equations of the model 170 for the three coil currents as three two-dimensional functions I _2f Is and I3 of the values of the two variable capacitors C22 and C$ ₂ . The three coil current functions are normalized to the total current through all three coils, and are denoted Χχ' _t ϊ ₂ ' , la' , respectively. Thus, for each value of the index Each two dimensional function !±' , X _s ' , I3' is stored in a respective memory 174-1, 174-2 and 174-3. A working example of the two dimensional functions χ' , Χ ₂ ' and Χ ₃ ' stored in the memories 174-1, 174-2 and 174-3 is depicted in the graphs of FIGS, 7, 8 and 9 respectively. In each one of FIGS. 7, 8 and , a plane is defined by two axes labeled (¾;> (the vertical axis) and C¾ _s (the horizontal axis) , and may be referred to as the {<¾, C½) -plane or two-dimensional (O * Css) -space. The magnitude of the coil current in each graph of FIGS. 7-9 is depicted on an axis orthogonal to the {C;> , C _¾ ) -plane .

[0047] FIGS. 7, 8 and 9 demonstrate a remarkable aspect of our discovery, namely that only two variable capacitors, {¾- and €3; are needed to realize nearly ail possible apportionments of the three coil currents 1χ ' , Iz' and. I ₃ '. The ability to cover such large range of apportionments is due in part to the relatively gradual slopes characteri ing the peaks in each of the graphs of FIGS. 7-9.

[0048J One rofoleia is ho to easily find the pair of capacitor values for C ₂₂ and C32 {out of all possible combinations) that provides the exact (or nearly exact} apportionment of Χχ ' , X _s ' and 1 ₃ ' specified by the user. This problem is solved in a first embodiment by a parameter path processor 176. Referring to FIG. 10, the parameter path processor 176 can be implemented by software or by a human operator as follows: Find three respective points PI, P2 and P3 in two-dimensional

(C->2,Cy>) -space at which a respective one of Χχ' , ¾>' and X ₃ ^e is raaximum (block 190 of FIG. 10} . Find a point P0 in two-dimensional (C;??, Ca _: >) -space at which, at least

approximately, ι'=1 ₃ '=Χ ₃ ' (block 192 of FIG . 10). Next, establish, a linear programming path in two-dimensional (C-22 , ¾) --space that includes all four points P0, PI, P2 and P3 (block 194 of FIG. 10} . Preferably, the path segments between pairs of the points P0, PI, P2 and P3 are straight lines. Ά user may then vary CV; and C½ along the pre-established linear programming path by varying only one parameter, until a desired apportionment of Xj ' , I ₂ ' and X ₃ ' is found (e.g., by trial and error}. The user may need to depart from the programming path to obtain optimum results, particularly if the plasma operating conditions (e.g., chamber pressure) used to implement the model have changed after the look-up tables were

generated from the model.

[0049] In another embodiment, the parameter path processor 176 uses the two-dimensional functions of J ₂ ' , Xz' and I3' to calculate corresponding discrete values of la ' , Xz' and X ₃ ^f at successive points along the linear programming path and stores these values in a look-up table 180 as a function of position along the linear programming path processor (block 196 of FIG. 10) . In one embodiment, during operation, a desired value of C ₂₂ and C32 may be found quickl by addressing the look-up table with a discrete set of desired values of Χ _% ' , and X ₃ ^r provided from a user interface 182 (block 198 of FIG. 10)■ [0050] An example of a linear programming path is depicted in FIG . 11 lying in the (C?;, C3?) -plane, and corresponds to the locations of peaks In each of the graphs of FIGS. 7-9, The same linear programming path is superimposed on the (<¾, C _¾; >) -plane in each of FIGS. 7-9. FIGS. 12A, 12E and 12C are graphical representations of the contents of the Iook- ^' up table 180 of FIG. 4 for 2 ₂ ' , I* ' and I?', respectively. For each position along the linear programming path of FIG. 11 (horizontal axis of FIGS. 12A-12C) , a corresponding discrete magnitude of a respective one of Xi ' , I2' and I3' (vertical axis) is gi en .

[0051] In one embodiment, once a representation of the linear programming path of FIG. 11 is stored in the look ^¬ up table 180, the other components of the processor 164 may be de-activated or removed, and only the look-up table ISO is required for complete control, unless or until operating conditions or reactor geometries are significantly altered. Thus,, in a highly economical implementation , the Iook- ^' up table 180 is programmed in the factory in the manner previously described with reference to blocks 190-196. Then, the components of the processor 164 are removed with the exception of the look up table 180,, and the reactor is shipped. Variations in processing conditions may be accommodated by furnishing different look-up tables that have been programmed in accordance with the different processing conditions.

[0052] FIG . 13 depicts a related embodiment, in which a linear programming path is not employed. Instead, in FIG, 13 the processor 164 includes a search processor 200 in place of the programmable path processor 176 of FIG. 4. The search processor 200 of PIG. 13 is adapted to search through the three 2-dimensional functions stored in the memories 174-1, 174-2 and 174-3 to find a pair of values of Cn and C _¾ 2 that provides a set of discrete values of X _t ' _f I ₃ ^f and X ₃ ' most closely resembling the desired values specified by the user interface 182. The user interface 182 may facilitate the selection of apportionment of the discrete values of X% ' , X ₂ ' and X ₃ ' by referring to process rate distribution information. Such information may reveal a pattern of non- uniformity in process rate distribution, enabling a user {or

intelligent machine) to counteract a perceived non- uniformit pattern by selecting a countervailing pattern of the three coil currents.

[0053] In one embodiment, once the data in the

memories 174-1, 174-2 and 174-3 are populated, the components 170 and 172 may be removed or deactivated for a more economical implementation. Such reactor may foe delivered with a set of tables stored in the memories 174-1, 174-2 and 174-3 to facilitate a range of chamber operating conditions and/or chamber geometries and/or source geometries .

[0054] Any one of the variable or fixed capacitors Q . _r 2 _f C3 _. 3 _. and/or C ₃ ? of FIG. 3 may be combined with a additional impedance element (not illustrated) , such as an inductor or a resistor, for example, in carrying out the embodiments described above with reference to FIGS, 1Ά-13,

[0055] FIG. 14 depicts a version of the embodiment of FIG, 1A, in which the outer coil 120 is wrapped around the side wall 102. in FIG. 14, the outer coil 120 may be below the plane of the ceiling 104. The side wall 102 may include a cylindrical dielectric window 102a and the ceiling 104 may include a disk-shaped dielectric window 104a, for example,

[0056] FIG. 15 is a simplified schematic block diagram depicting a general form of the current divider 162 of FIG, 38. The three branches 162-1, 162-2 and 162-3 of the current divider 162 ar referred to in this

specification as current divider branches. In FIG. 15, each of the three current divider branches 162-1, 162-2 and 162-3 is depicted as having a generic series

impedance element and a parallel impedance element .

Thus, the first current divider branch 162-1 includes of a series impedance element Z _u and a parallel impedance element ·2, t e second current divider branch 162-2 includes of a series impedance element ¾ and a parallel impedance element ¾, and the third current divider branch 162-3 includes of a series impedance element Z _t and a parallel impedance element Z^. Each impedance element Zn, Z , Z ■. , 2 : _: and ¾; includes a capacitor, and, optionally, may include an additional element such as an inductor or a resistor.

[0057] A set of rules defining the operable species of FIG, 15 is as follows:

Rule 1: At least two of the three current divider branches 162-1, 162-1 and 162-3 must each have at least one variable impedance element and a series impedance element of a finite non-zero impedance . In some cases, the variable impedance element and the series impedance element may foe the same element. ule 2: The two variable impedance elements of Rule 1 must both be either parallel impedance elements (i.e., they must be any two of Z%z, Zn and Z33) , or series impedance elements {i.e., they must be any two of Ζ \ , Z21 , and Z -y. ) .

Rule 3: The parallel impedance elements {i.e., Ζχζ, %2z and Z yy ) may not be short circ its.

Rule 41: If the number of current divider branches having variable impedance elements is exactly two, and if the variable impedance elements are series impedance elements (i.e., if they are any two of Ζπ, and Z y _x ) , then the remaining branch mast have a finite non-zero series impedance element.

[0058] An alternative or modified set of rules

defining the operable species of FIG. 15 is as follows:

Rule Ά: At least two of the three current divider branches 162-1, 162-1 and 162-3 must each have at least one variable impedance element and a series impedance element of a finite non-zero impedance, and are referred to herein as ^tunable current divider branches". In some cases, the variable impedance element and the series impedance element may be the same element .

Rule B: It is not required that the two variable

impedance elements of Rule A are both parallel impedance elements {e.g., any two of Z i z _f 1 and Z y , or both series impedance elements (e.g., any two of Ζ , Ζζι, and 31 ) _t in that, in some cases, one may be a parallel impedance element while the other a series impedance element . ule C: The parallel impedance elements (i.e., Ζη, Z2 and Zi2) may not foe short circuits,

Rule D: If the nuraher of tunable current divider

branches as defined in Rule A is exactly two, and if at least one of the variable impedance elements is a series impedance element (i.e., any of Ζχχ, Z21, or ¾) , then the remaining branch must have a finite non-zero series impedance element.

[0059] The current divider 162 of F G. 3E is an operable species within the foregoing rules defining operable species of FIG. 15, in which Ζχ·. - 0 (is a short circuit} , Ziz - ⁰⁰ {is open) and Zzz and are the variable impedance elements,

[0060] Another operable species is depicted in FIG. 16, in which the series impedance elements Z _n , Ζ ₂ · _* and ¾ are each non-zero finite impedances and two of them are variable, and in which the parallel impedance

elements {Ζχζ, Zn and z _3i ) are of any impedance values

(including infinity) other than zero. In the specific example of FIG. 16, the two variable impedance elements are and .

[00613 As used herein, the term "current divider branch" or "current branch" refers to each one of a set of three current divider branches conforming to the rules stated above.

[0062 ] In each of the operable species defined above, at least two of the current divider branches have a variable impedance element, which is controlled by a respective one of the controllers 166a, 166b of FIG. 15. [0063 J In each of the operable species defined above , each current divider branch 162-1, 162-2, 162-3 is coupled between the output of the match tuning circuit 152 and a respective one of the coil antennas. In the embodiments of FIG. 15, this coupling is a direct connection from the match tuning circuit 152 to each current divider branch 162-1, 162-2, 162-3, FIG. 17 depicts an equivalent embodiment, in which coupling of the match tuning circuit 152 to each current divider branch 162-1, 162-2, 162-3 is indirect and is made through intervening series impedances 300 , 301 and/or 302, and in which coupling of each current divider branch 162-1, 1 62-2 , 162-3 to a respective coil antenna is indirect and is made through intervening series

impedances 303, 304 and/or 305. The term "between" is defined in this specification such that each current divider branch 162-1, 162-2, 162-3 is coupled "between" the match tuning circuit 152 and the respective coil antenna whether or not any of the series impedances 300, 301, 302, 303, 304 or 305 are present. Each one of the series impedances 300 , 301, 302, 303 , 304 or 305 may include a capacitor, an inductor or a. resistor.

[0064] While the foregoing is directed to embodiments of the present invention, other and further embodiments of the invention may be devised without departing from th basic scope thereof, and the scope thereof is determined by the claims that follow.