Magnetic resonance coupling arrangement

Abstract

A wireless power transfer system is disclosed. The system includes a first resonator having a first resonant frequency ωo1, a half power bandwidth Δω1, and an unloaded quality factor Qo1=ωo1/Δω1 coupled through a first coupling circuit to a power source, a second resonator having a second resonant frequency ωo2, a half power bandwidth Δω2, and an unloaded quality factor Qo2=ωo2/Δω2 coupled through a second coupling circuit to a load, the first resonator disposed a distance away from the second resonator, wherein the distance is smaller than the first and second resonant wavelengths, the first and second coupling circuits are configured so that up to a maximum achievable power transfer efficiency between the first and second resonators can be achieved, wherein Qo1 and Qo2 can be less than 100.

Description

TECHNICAL FIELD

The present disclosure relates to the field of magnetic and electrical coupling and in particular to resonant coupling between resonators for wireless energy transfer.

BACKGROUND

This section introduces aspects that may help facilitate a better understanding of the disclosure. Accordingly, these statements are to be read in this light and are not to be understood as admissions about what is or is not prior art.

Wireless power transfer (WPT) technologies are gaining increased popularity as they address key limitations associated with “tethered” and battery operated consumer and medical electronics including miniaturization, cost, and safety. Traditionally, WPT techniques are based on near-field magnetic induction or far-field radiofrequency radiation. Far-field techniques offer the capability to wirelessly power over a large coverage area but have low end-to-end efficiency due to high material electromagnetic (EM) interaction and 1/r^2 power density dependence. In contrast, near-field magnetic induction can operate at high efficiency but only within centimeter ranges. Recently, the development of magnetic resonance coupling (MRC) has gained momentum as a promising WPT method due to its capability of operating at much longer distances than traditional inductive methods while maintaining higher end-to-end power transfer efficiency than far-field radiative techniques. In addition, MRC systems demonstrate capabilities of power “relaying” and preferentially designed power distribution to multiple loads, features of which do not exist in traditional induction and far-field techniques. However, practical implementation of MRC has been challenging. Conventional MRC requires a minimum of four inductively coupled coils to generate the resonant structure of which the optimization parameters are difficult to control and are difficult to miniaturize. Furthermore, the coupled-mode theory and equivalent circuit model derived design equations are overly complex and become impractical as a design guide especially when incorporating “relay” coils and power distribution to multiple loads. Other methods utilizing direct coupled methods and impedance matching networks have been presented. However, among the current technology, there does not exist a unified and simple approach for achieving the optimal IM solution in MRC systems. This is especially apparent as systems become overly complex due to the addition of relay resonators.

Therefore, there is an unmet need for a novel arrangement and methodology in WPT utilizing MRC that allows for simple IM optimization and flexible design that can be reconfigured for when relay coils are being added.

SUMMARY

A wireless power transfer system is disclosed. The system includes a first resonator having a first resonant frequency ω_o1, a half power bandwidth Δω₁, and an unloaded quality factor Qo₁=ω_o1/Δω₁ coupled through a first coupling circuit to a power source. The system further includes a second resonator having a second resonant frequency ω_o2, a half power bandwidth Δω₂, and an unloaded quality factor Qo₂=ω_o2/Δω₂ coupled through a second coupling circuit to a load, the first resonator disposed a distance away from the second resonator, wherein the distance is smaller than the first and second resonant wavelengths. The first and second coupling circuits are configured so that up to a maximum achievable power transfer efficiency between the first and second resonators can be achieved, wherein Qo₁ and Qo₂ can be less than 100.

A method of providing a maximum wireless power transfer efficiency between a first resonator and a second resonator is also disclosed. The method includes providing a first resonator having a first resonant frequency ω_o1, a half power bandwidth Δω₁, and an unloaded quality factor Qo₁=ω_o1/Δω₁ coupled through a first coupling circuit to a power source. The method also includes providing a second resonator having a second resonant frequency ω_o2, a half power bandwidth Δω₂, and an unloaded quality factor Qo₂=ω_o2/Δω₂ coupled through a second coupling circuit to a load, the first resonator disposed a distance away from the second resonator, wherein the distance is smaller than the first and second resonant wavelengths. The method further includes configuring the first and second coupling circuits so that up to a maximum power transfer efficiency between the first and second resonators can be achieved, wherein Qo₁ and Qo₂ can be less than 100.

BRIEF DESCRIPTION OF THE DRAWINGS

The above and other objects, features, and advantages of the present invention will become more apparent when taken in conjunction with the following description and drawings wherein identical reference numerals have been used, where possible, to designate identical features that are common to the figures, and wherein:

FIG. 1 is a schematic representation of a magnetic resonance coupling (MRC) circuit according to the present disclosure.

FIG. 2a is another schematic representation of a MRC circuit according to the present disclosure.

FIG. 2b is a schematic of a lumped element capacitive K-inverter network exhibiting a characteristic impedance K, input impedance Z_C and real output impedance, R_out.

FIG. 3 is yet another schematic representation of a MRC circuit according to the present disclosure.

FIG. 4 is a graph of power transfer efficiency (PTE) response as a function of inter-resonator coupling (which represents coil separation distance/misalignment) for a single value of Q1=Q2=475.

FIG. 5 is a graph of an inter-resonator coupling coefficient vs. coil separation (in mm).

FIG. 6 is a graph of a measure of PTE vs. inter-resonator coupling coefficient.

FIG. 7 is a graph of a simulation of PTE vs. frequency for different values of a capacitive network are increased and decreased from the optimum calculated value to show peak PTE at the optimal value for a particular value of k_12tgt of 0.01.

FIG. 8 is a flow chart of a method according to the present disclosure.

FIGS. 9 and 10 represent schematic drawings of magnetic resonance coupling (MRC) circuit with repeaters according to the present disclosure.

The attached drawings are for purposes of illustration and are not necessarily to scale.

DETAILED DESCRIPTION

For the purposes of promoting an understanding of the principles of the present disclosure, reference will now be made to the embodiments illustrated in the drawings, and specific language will be used to describe the same. It will nevertheless be understood that no limitation of the scope of this disclosure is thereby intended.

In the present disclosure, an alternative method to achieving wireless power transfer (WPT) via magnetic resonance coupling MRC, based on a bandpass filter (BPF) synthesis methodology is disclosed to address the aforementioned challenges of conventional MRC while also providing added functionality in the form of arbitrary load impedance matching. In contrast with conventional four coil MRC, BPF synthesized MRC shrinks to only a two-coil resonant structure. This shrinking is accomplished via impedance transformation using passive lumped component immitance inverters (or otherwise identified as K-inverter), thus, resulting in a more compact and controllable system. The filter design equations as applied to MRC are concise and can be solved mathematically to predict and control the frequency and coupling response phenomena unique to resonantly coupled circuits. In addition, the parameter optimization procedures in BPF synthesized MRC shows practical advantages over conventional MRC systems. For example, maximizing power transfer dynamically over coil misalignment and separation can be achieved via adaptive tuning of the characteristic impedance of the immitance inverters e.g. changing capacitance values. From this point a designer can utilize any lumped element network that is a K-inverter network and achieve optimal or maximum achievable PTE by manifesting the optimally derived characteristic impedance of the system. This is significantly more practical than adaptive tuning in conventional MRC which typically requires system level tuning such as optimization of coil coupling coefficients which is both difficult to accomplish physically, difficult to measure, and is highly variable.

The designs of the K-inverters are based on developing a general source and load n±2 coupling matrix. Relevant design equations are then derived from this coupling matrix and used to optimize and maximize wireless power transfer given a pre-determined coupling coefficient between the transmit and receive coils. Addition of relay coils and/or multiple received devices require simply updating the n±2 coupling matrix and re-deriving the design equations or doing so numerically and implementing the optimization procedure used for a system of only 2 resonators. In addition, arbitrary load impedance matching can also implemented and also only requires updating the n±2 coupling matrix and re-deriving the relevant design equations from this coupling matrix.

In the present disclosure, resonators may be shown as RLC circuits, however, other types of resonators are possible, including but not limited to cavity, mechanical, optical, fluid, etc. In each case, the resonant frequency ω_r of the resonator is a physical characteristic associated with the resonator. A half-power bandwidth Δω, represents bandwidth over which the power of vibration is greater than half the power at the resonant frequency. Quality factor of the resonator, defined in alternative ways in this document, is defined in one sense as ω_r/Δω.

Referring to FIG. 1, a MRC arrangement 100 according to the present disclosure is provided. The arrangement 100 includes a source 102, a first resonator 104, a second resonator 106, and a load 108. The source 102 is electrically coupled to the first resonator 104 via an electrical coupling link 116. There is also a normalized coupling coefficient identified by M_S1 110 between the source 102 and the first resonator 104. The load 108 is electrically coupled to the second resonator 106 via an electrical coupling link 120. There is also a normalized coupling coefficient identified by M_2L 114 between the load 108 and the second resonator 106. There is also a normalized coupling coefficient identified by M₁₂ 112 between the first resonator 104 and the second resonator 106. Finally there is an inter-resonator coupling (later identified as k₁₂) identified as 118 between the first resonator 104 and the second resonator 106.

Referring to FIG. 2a, another schematic of a detailed model of an equivalent BPF modeled MRC system is provided. On the left hand side there is a source V_s with a source impedance Z_s interfacing with a first k-inverter K_s1 that is then coupled to a first resonator Res. 1 having lumped parameter characteristics C₁ and L₁. Input impedance looking into K_S1 equals Z_S1. Inter-resonator magnetic coupling between the first resonator Res.1 and a second resonator Res.2 is identified as K₁₂. On the right hand side there is a load with a load impedance Z_L interfacing with a second k-inverter K_2L that is then coupled to the second resonator Res. 2 having lumped parameter characteristics C₂ and L₂.

Referring to FIG. 2b, a lumped element capacitive K-inverter network exhibiting a characteristic impedance K, input impedance Z_C and real output impedance, R_out is provided.

Referring to FIG. 3, another schematic of a detailed model of an equivalent BPF modeled MRC system is provided. On the left hand side there is a source V_s with a source impedance Z_s interfacing with a first k-inverter K_s1opt that is then coupled to a first resonator Res. 1 having lumped parameter characteristics C₁, L₁, and parasitic resistance R_p1. The first k-inverter includes a capacitive network of C_p1 and −C_s1, where −C_s1 and C₁ combine to provide a capacitive equivalent C_1s1. Inter-resonator magnetic coupling between the first resonator Res.1 and a second resonator Res.2 is identified as K₁₂. On the right hand side there is a load with a load impedance Z_L interfacing with a second k-inverter K_2Lopt that is then coupled to the second resonator Res. 2 having lumped parameter characteristics C₂, L₂, and patristic resistance R_p2. The second k-inverter includes a capacitive network of C_p2 and −C_s2, where −C_s2 and C₂ combine to provide a capacitive equivalent C_2s2. The capacitive networks C_p1, −C_s1, C_p2, and −C_s2 are configured such that the system will exhibit K_S1opt and K_2Lopt characteristic impedances, (which are uniquely derived from the methodology described herein). This will enable maximum achievable PTE for the given resonator parameters and desired inter-resonator coupling coefficient.

In typical WPT applications, inductors L₁ and L₂ are fixed according to the coil design and physical dimensions at a pre-specified ω₀. Thus, the resonant frequency of the system can be set by choosing series capacitors C₁ and C₂ accordingly, based on:

${\omega }_0=2\pi f_0=\frac{1}{\sqrt{L_1{C}_1}}=\frac{1}{\sqrt{L_2{C}_2}}$

The generic matched condition for a lossless (Q_0n=∞) 2-stage BPF occurs when input impedance Z_S1 is the complex conjugate of Z_S. In the present disclosure, the source and load impedances, Z_S and Z_L respectively, are fixed to 50Ω to enable convenient measurement with standard 50Ω testing equipment. The relationships between impedances Z_s1 Z₁₂, Z_2L (See FIG. 2a) can be determined by applying

$Z_{\mathrm{in}}=\frac{K^2}{Z_L},$ at each K-inverter, wherein Z_in is the input impedance looking into the K-inverter network, K is the real valued characteristic impedance of the inverter, and Z_L is the load impedance. The resulting matched condition for a lossless system (i.e., infinite Q) can be shown to beK₁₂Z_S=K_S1K_2L

The inter-resonator K-inverter characteristic impedance, K₁₂, is directly related to the inter-resonator coupling coefficient, k₁₂. The generic BPF IM method, for lossless MRC systems, is realized by manifesting arbitrarily chosen values of K_S1 and K_2L which result in the equivalency described by K₁₂Z_S=K_S1K_2L. However, the ambiguity in choosing values K_S1 and K_2L does not lead to the optimal IM conditions, which are unique based on the resonator parameters including resonator Q. The source and load included coupling matrix is a convenient and powerful tool commonly used in BPF design and analysis.

The BPF modeled MRC system shown in FIG. 2a can be generalized as a maximally flat 2-pole (n=2) Butterworth BPF with source and load included external couplings. This system can be conveniently described by its coupling topology which is illustrated in FIG. 1. Using FIG. 1, the coupling matrix for the 2-stage BPF modeled MRC system can be easily determined as

$\left[M\right]=\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}\\ S\end{array}\\ 1\end{array}\\ 2\end{array}\\ L\end{array}\left[\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}S\\ 0\end{array}\\ M_{S1}\end{array}\\ 0\end{array}\\ 0\end{array}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}1\\ M_{S1}\end{array}\\ -{\mathrm{jM}}_{11}\end{array}\\ M_{12}\end{array}\\ 0\end{array}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}2\\ 0\end{array}\\ M_{L2}\end{array}\\ -{\mathrm{jM}}_{22}\end{array}\\ M_{2L}\end{array}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}L\\ 0\end{array}\\ 0\end{array}\\ M_{2L}\end{array}\\ 0\end{array}\right]$ where e j=√{square root over (−1)}, subscripts S and L represent the source and load respectively, and subscripts 1 and 2 represent the first and second resonator, respectively. This coupling matrix can be extended to include N>2 resonators for the purposes of power relay capabilitiy. The M values of the above matrix represent the normalized coupling coefficients of the system. In particular, M_S1 and M_2L are the normalized external couplings between the source to resonator 1 and resonator 2 to load respectively. Normalized coupling coefficients M₁₁ and M₂₂ represent the normalized self-coupling terms of resonator 1 and 2 respectively and is related to Q_0n of each resonator by

$M_{11}=\frac{1}{{\mathrm{FBWQ}}_{01}}$ $M_{22}=\frac{1}{{\mathrm{FBWQ}}_{02}}$ where FBW is the fractional bandwidth of the filter and is defined as

$\mathrm{FBW}=\frac{{\omega }_2-{\omega }_1}{{\omega }_0}$ where ω₂−ω₁ is the 3 dB passband-edge angular frequency of the BPF prototype MRC model. Typically, FBW is defined by the desired requirements of the BPF.

The unloaded Q factor of each series LC resonator can be determined as

$Q_{0n}=\frac{{\omega }_0{L}_n}{R_{\mathrm{Pn}}}$ where R_Pn is the frequency dependent equivalent series resistance (ESR) of the coil inductors. The values of M₁₁ and M₂₂ can be determined through measurement of resonator Q_0n.

M-parameter, M₁₂, represents the normalized inter-resonator coupling coefficient which is given by

$M_{12}=\frac{k_{12}}{\mathrm{FBW}}$

From BPF theory, it can be shown that M_S1, M₁₂, and M_2L are directly related to the characteristic impedance of each K-inverter for the 2-stage BPF modeled MRC system shown in FIG. 2a. Assuming a fixed 50Ω source and load impedance, these relationships are given byK_S1=M_S1√{square root over (50L₁ω₀FBW)}K₁₂=FBW ω₀M₁₂√{square root over (L₁L₂)}K_2L=M_2L√{square root over (50L₂ω₀FBW)}

The transfer response for the 2-Stage BPF Modeled MRC can be computed directly from the M-matrix in terms of the scattering parameters byS₂₁=−2j[A]_n+2,1⁻¹ S₁₁=1+2j[A]_1,1⁻¹ where matrix [A] is given by[A]=[M]+Ω[U]−j[q]where

$\Omega =\frac{1}{\mathrm{FBW}}\left(\frac{\omega }{{\omega }_0}-\frac{{\omega }_0}{\omega }\right)$ $\mathrm{and}\left[U\right]=\left[\begin{array}{cccc}0& 0& 0& 0\\ 0& 1& 0& 0\\ 0& 0& 1& 0\\ 0& 0& 0& 0\end{array}\right].\left[q\right]=\left[\begin{array}{cccc}1& 0& 0& 0\\ 0& 0& 0& 0\\ 0& 0& 0& 0\\ 0& 0& 0& 1\end{array}\right]$

The S₁₁ and S₂₁ parameters are the power reflection and transmission ratios, respectively. It is important to note that in the present disclosure, power transfer efficiency (PTE) is defined as the wireless transmission efficiency which does not include the power loss due to the impedance of a power source. Specifically, this PTE is related to S₂₁ by

$\mathrm{PTE}=|S_{21}{|}^2\times 100$ $S_{21}=\frac{2{\mathrm{jM}}_{S1}{M}_{12}{M}_{2L}}{M_{12}^2+\left(M_{S1}^2+M_{11}+j\Omega \right)\left(M_{22}+M_{2L}^2+j\Omega \right)}$ taking the magnitude at ω₀ (Ω=0) simplifies the above equation to:

$|S_{21}{|}_{\omega ={\omega }_0}=\frac{2M_{12}{M}_{2L}{M}_{S1}}{M_{12}^2+\left(M_{22}+M_{2L}^2\right)\left(M_{11}+M_{S1}^2\right)}$ This equation fully describes the power transmission topology of the 2-stage BPF modeled in the present disclosure. In order to find the maximum power transfer efficiency, a partial derivative of the above equation is obtained with respect to M_S1 and M_2L. By inspection, M₁₁, M₁₂, and M₂₂ are known constants determined by the resonator parameters (M₁₁ and M₂₂) and the specified inter-resonator coupling coefficient, k₁₂ (M₁₂) at which optimal IM is desired to occur. This targeted inter-resonator coupling value is redefined as k_12tgt. Likewise, M₁₂ is redefined as M_12tgt.

To determine the optimal M_S1 and M_2L functions, the partial derivative of (26) is taken with respect to (w.r.t) M_S1 and M_2L. Both partial derivative functions are set to equal zero (indicating global maximum). This yields two equations with two unknowns, the optimal M_S1 and M_2L values, redefined as M_S1opt and M_2Lopt. By solving the systems of two equations, the M_S1opt and M_2Lopt functions are derived to be

$M_{S1\mathrm{opt}}=\frac{{M_{11}^{1\text{/}4}\left(M_{12\mathrm{tgt}}^2+M_{11}{M}_{22}\right)}^{1\text{/}4}}{M_{22}^{1\text{/}4}}$ $M_{2\mathrm{Lopt}}=\frac{{M_{22}^{1\text{/}4}\left(M_{12\mathrm{tgt}}^2+M_{11}{M}_{22}\right)}^{1\text{/}4}}{M_{11}^{1\text{/}4}}$

Using

$M_{11}=\frac{1}{{\mathrm{FBWQ}}_{01}},M_{22}=\frac{1}{{\mathrm{FBWQ}}_{02}},\mathrm{and}{M}_{12}=\frac{k_{12}}{\mathrm{FBW}},$ S₁₁=1+2j[A]_1,1⁻¹ can be simplified to

$M_{S1\mathrm{opt}}={\left(\frac{1+k_{12\mathrm{tgt}}^2{Q}_{01}{Q}_{02}}{{\mathrm{FBW}}^2{Q}_{01}^2}\right)}^{1\text{/}4}$ $M_{2\mathrm{Lopt}}={\left(\frac{1+k_{12\mathrm{tgt}}^2{Q}_{01}{Q}_{02}}{{\mathrm{FBW}}^2{Q}_{02}^2}\right)}^{1\text{/}4}$ which represent the global optimum normalized external coupling M-matrix values. As shown, M_S1opt and M_2Lopt can be determined analytically as functions of the unloaded Q_0n of the resonators and the targeted location, k_12tgt, at which the optimal IM is desired to occur.

The above equations can be substituted in to K_S1=M_S1√{square root over (50L₁ω₀FBW)} and K_2L=M_2L√{square root over (50L₂ω₀FBW)} resulting in the determination of global optimum external coupling characteristic impedance functions no redefined as K_S1opt and K_2Lopt as follows:

$K_{S1\mathrm{opt}}={\left(1+k_{12\mathrm{tgt}}^2{Q}_{01}{Q}_{02}\right)}^{1\text{/}4}\sqrt{\frac{50L_1{\omega }_0}{Q_{01}}}$ $K_{2\mathrm{Lopt}}={\left(1+k_{12\mathrm{tgt}}^2{Q}_{01}{Q}_{02}\right)}^{1\text{/}4}\sqrt{\frac{50L_2{\omega }_0}{Q_{02}}}$ Therefore, for a known C₁, L₁, R_p1 and C₂, L₂, and R_p2,

${\omega }_0=2\pi f_0=\frac{1}{\sqrt{L_1{C}_1}}=\frac{1}{\sqrt{L_2{C}_2}}$ $Q_{0n}=\frac{{\omega }_0{L}_n}{R_{\mathrm{Pn}}}$ k_12tgt is measured, thus k_s1opt and k_2Lopt can be found from the above equations. With k_s1opt and k_2Lopt parameters in hand, referring back to FIG. 2a, the input impedance of a K-inverter is expressed as

$Z_{\mathrm{in}}=\frac{K^2}{Z_L},$ where Z_in is the input impedance looking into the K-inverter network, K is the real valued characteristic impedance of the inverter, and Z_L is the load impedance.

However, before we can determine −C_sn and C_pn, we need to revisit FIG. 2a. A cursory analysis of the ABCD matrix for a quarter-wave transmission line leads to the ABCD matrix describing K-inverters, which is given by

$\left[\begin{array}{cc}A& B\\ C& D\end{array}\right]=\left[\begin{array}{cc}0& \pm \mathrm{jK}\\ \stackrel{\_}{+}\frac{j}{K}& 0\end{array}\right]$

The input impedance, Z_C, of the K-inverter circuit in FIG. 2b is given by

$Z_C=\frac{R_{\mathrm{out}}}{1+C_{\mathrm{pn}}^2\hspace{0.17em}{R}_{\mathrm{out}}^2{\omega }_0^2}+j\left(\frac{1}{C_{\mathrm{sn}}{\omega }_0}-\frac{C_{\mathrm{pn}}{\omega }_0}{\frac{1}{R_{\mathrm{out}}^2}+C_{\mathrm{pn}}^2{\omega }_0^2}\right)$ where ω₀ is the operating angular frequency. Using the above equation, the relationships of −C_sn and C_pn to the K-inverter characteristic impedance, K, can be determined. Specifically, this is done by equating (3) with the inversion relationship of a K-inverter described by

$\left[\begin{array}{cc}A& B\\ C& D\end{array}\right]=\left[\begin{array}{cc}0& \pm \mathrm{jK}\\ \stackrel{\_}{+}\frac{j}{K}& 0\end{array}\right].$ Recall that the characteristic impedance of a K-inverter is real valued. Thus, by using the relationships given by

$\mathrm{Re}\left\{Z_C\right\}=\frac{K^2}{R_{\mathrm{out}}}$ $\mathrm{Im}\left\{Z_C\right\}=0$ Thus the capacitances −C_sn and C_pn can be solved as follows:

$-C_{\mathrm{sn}}=\frac{R_{\mathrm{out}}}{K{\omega }_0\sqrt{-K^2+R_{\mathrm{out}}^2}}$ $C_{\mathrm{pn}}=\frac{\sqrt{-K^2+R_{\mathrm{out}}^2}}{R_{\mathrm{out}}K{\omega }_0}$ By placing the K_S1opt and K_2Lopt into the −C_sn and C_pn equations above, −C_sn and C_pn can be determined to provide the optimum impedance matching and hence the optimum PTE.

To determine the effect tuning k_12tgt has on the optimal PTE response at ω=ω₀ as a function of resonator separation distance/angular misalignment; represented through resonator coupling coefficient, k₁₂, the following analysis is provided. The determination of optimal external characteristic impedances of the 2-stage BPF modeled MRC system requires the resonator parameters and a specification on the desired resonator coupling point, k_12tgt, at which optimal IM is to occur. The effect k_12tgt has on the optimal PTE response as a function of k₁₂ can be determined by modifying

$|S_{21}{|}_{\omega ={\omega }_0}=\frac{2M_{12}{M}_{2L}{M}_{S1}}{M_{12}^2+\left(M_{22}+M_{2L}^2\right)\left(M_{11}+M_{S1}^2\right)}.$ Specifically, M_S1opt and M_2Lopt are substituted in place of M_S1 and M_2L. Variables, M₁₁, M₂₂, and M₁₂ are left unchanged arriving at:

$|S_{21}{|}_{{\mathrm{opt}}_{—}k}=\frac{2Q_{01}{Q}_{02}{k}_{12}\sqrt{\frac{1+k_{12\mathrm{tgt}}^2{Q}_{01}{Q}_{02}}{Q_{01}{Q}_{02}}}}{2+k_{12\mathrm{tgt}}^2{Q}_{01}{Q}_{02}+k_{12}^2{Q}_{01}{Q}_{02}+2\sqrt{1+k_{12\mathrm{tgt}}^2{Q}_{01}{Q}_{02}}}$

FIG. 4 shows the plotted results of the above equation as a function of a k₁₂ range and four example values of k_12tgt values. FIG. 4 is a representation of the PTE response as a function of inter-resonator coupling (which represents coil separation distance/misalignment) for a single value of Q1=Q2=475. Each curve represents a different k_12tgt point in the optimal IM method. For example: if it is known that the resonators are going to be 4 cm apart and it is known that 4 cm apart corresponds to a k₁₂ value of 0.02, then the IM network can be optimized by making k_12tgt=0.02. This will result in the corresponding PTE vs. k₁₂ response shown in FIG. 4, for k_12tgt=0.02. The unloaded Q-factors, Q₀₁ and Q₀₂ are fixed to a value of 475. As shown, decreasing the value of k_12tgt causes a corresponding shift in PTE response and the location of maximum achievable PTE. This point is known as the critical coupling point, k_12crit. The k₁₂ regions lower and higher than k_12crit are characterized as the under-coupled and over-coupled regions, respectively. This tuning result indicates that minimizing k_12tgt has the effect of maximizing range at the expense of maximum achievable PTE. For our BPF modeled MRC system, this tuning optimization and controllability can be accomplished in a practical manner as tuning the k_12tgt value can be manifested through the analytically determined capacitance values for −C_sn and C_pn provided above. Referring to FIG. 4, theoretical PTE versus inter-resonator coupling coefficient, k₁₂, for various k_12tgt tuning values at ω=ω₀ are provided.

A unique system behavior can be observed upon examination of FIG. 8. Specifically, there is an observable discrepancy between the PTE at k₁₂=k_12tgt (diamonds in FIG. 8) and k₁₂=k_12crit point (dots in FIG. 4). Furthermore, the magnitude of discrepancy changes in relation to each k_12tgt tuned PTE response. This phenomena indicates that the desired location of the critical coupling, ideally set by k_12tgt, deviates from the actual location, k_12crit.

Referring to FIG. 5, shows the measured range of k₁₂ between the fabricated coils as a function of resonator separation distance. The range of achievable k₁₂ was measured by obtaining the S-parameters of the coils through two port Vector Network Analyzer (VNA) measurements. The design parameters summarized in Table I and the specified k_12tgt points in FIG. 5 are used to determine K_S1opt and K_2Lopt by equations for K_S1opt and K_2Lopt, respectively. The corresponding optimal capacitance values are determined by substituting the specific K_S1opt and K_2Lopt values into equations for −C_sn and C_pn provided above.

TABLE I
1. Experimental resonator parameters
		3. Resonator	4. Resonator
	2. Parameters	1 (Tx)	2 (Rx)
	5. f0	6. 13.56 MHz	7. 13.56 MHz
	8. Ln	9. 1410 nH	10. 1460 nH
	11. Rpn	12. 0.2529 Ω	13. 0.2619 Ω
	14. Q0n	15. 475	16. 475
	17. Cn*	18. 97.702 pF	19. 94.356 pF

Referring to FIG. 6, theoretical and measured resonant frequency PTE response plotted as a function of k₁₂. Each curve indicates a particular optimal IM design for the chosen k_12tgt tuning point. The dot indicates the theoretically predicted critical coupling point.

FIG. 6 also shows the measured PTE responses at ω=ω₀ (13.56 MHz) as a function of k₁₂ (separation distance of 30 mm to 330 mm) for the 5 different optimal IM conditions based on the chosen k_12tgt optimization points. Importantly, the measured PTE response and critical coupling point converge with the predicted results both in magnitude and location. A comparison of the S₁₁ response between the measured and circuit simulation was conducted to investigate the measurement and theory discrepancy shown in FIG. 6. This result indicates that slight deviation from the optimal capacitance values of the optimal IM networks are likely to blame for the discrepancy in PTE response shown in FIG. 6. This observation points to the importance of proper capacitance value tuning and minimization of PCB parasitic for achieving the optimal IM condition.

Referring to FIG. 7, a graph of |S_2l| vs. frequency is depicted using the techniques of the present application to generate the solid line (optimal condition). The dashed lines indicate K_S1 (See FIG. 2a) increased and decreased by 10% from the optimal value. The change in the values indicate that there is a marked drop from the optimal condition by simply varying K_s1 by 10% in opposing directions.

Referring to FIG. 8, a flowchart for determining the optimal k-inverter values to arrive at the maximum PTE is provided. At step 500, various physical parameters of the resonators (two of them) are determined. Next, at step 502, inter-resonator coupling value k_12tgt is determined. Next, at step 504, K values for S1opt and 2Lopt are calculated. Finally, at step 506, optimal k inverter IM network is calculated by determining −C_sn and C_pn.

The method and system described herein can also be expanded to account for repeaters. While 2 resonators have been shown, the number of resonators can be N+2 where +2 represents the source and load resonators. Referring to FIGS. 9 and 10 two schematic drawings are presented that depict repeaters in an MRC system. In this case the coupling matrix (provided below in Table II) will be updated to account for the number of resonators greater than 2. The transfer function of this updated N+2 set of resonator can then be solved (with N>2 resonators; numerical methods are more practical). An optimization algorithm can then be implemented to determine the external coupling M variables such that the transfer function is optimized (which represents optimal PTE). The optimal external coupling M variables are used to determine the optimal external coupling K-inverters. The external coupling K-inverters are manifested by a lumped element K-inverter network and thus optimal IM conditions can be realized.

TABLE II
Coupling Matrix for the case of repeaters
	S	1	2	n − 1	n	L
S	0	MS1	MS2	MS, n−1	MSn	MSL
1	M1S	M11	M12	M1, n−1	M1n	M1L
2	M2S	M21	M22	M2, n−1	M2n	M2L
n − 1	Mn−1, S	Mn−1, 1	Mn−1, 2	Mn−1, n−1	Mn−1, n	Mn−1, L
n	MnS	Mn1	Mn2	Mn, n−1	Mnn	MnL
L	MLS	ML1	ML2	ML, n−1	MLn	0

The invention has been described in detail with particular reference to certain preferred aspects thereof, but it will be understood that variations, combinations, and modifications can be effected by a person of ordinary skill in the art within the spirit and scope of the invention.

Claims

1. A wireless power transfer system, comprising: a first resonator having a first resonant frequency ωo1, a half power bandwidth Δω1, and an unloaded quality factor Qo1=ωo1/Δω1 coupled through a first coupling circuit to a power source; anda second resonator having a second resonant frequency ωo2, a half power bandwidth Δω2, and an unloaded quality factor Qo2=ωo2/Δω2 coupled through a second coupling circuit to a load, the first resonator disposed a distance away from the second resonator;wherein the configuring of the first and second coupling circuits based on expressing transmission power from the first resonator to the second resonator as a function of external coupling with the source and external coupling with the load and determining partial derivatives of the transmission power with respect to the external couplings in order to find optimal external couplings and in order to find optimal characteristic impedance of the first and second coupling circuits,

2. The wireless power transfer system of claim 1, optimal characteristic impedance of the first and second coupling circuits are used to find lumped parameter values for components of the first and second coupling circuits.

3. The wireless power transfer system of claim 2, wherein the lumped parameter values for components of the first and second coupling circuits are governed by:

4. The wireless power transfer system of claim 1, the first and second coupling circuits are re-configured to continue to provide a maximum power transfer efficiency between the first and second resonators as the distance between the first and second resonators change.

5. A method of providing a maximum wireless power transfer efficiency between a first resonator and a second resonator, comprising: providing a first resonator having a first resonant frequency ωo1, a half power bandwidth Δω1, and an unloaded quality factor Qo1=ωo1/Δω1 coupled through a first coupling circuit to a power source;providing a second resonator having a second resonant frequency ωo2, a half power bandwidth Δω2, and an unloaded quality factor Qo2=ωo2/Δω2 coupled through a second coupling circuit to a load, the first resonator disposed a distance away from the second resonator;wherein the configuring of the first and second coupling circuits based on expressing transmission power from the first resonator to the second resonator as a function of external coupling with the source and external coupling with the load and determining partial derivatives of the transmission power with respect to the external couplings in order to find optimal external couplings and in order to find optimal characteristic impedance of the first and second coupling circuits,

6. The wireless power transfer system of claim 5, wherein finding optimal characteristic impedance of the first and second coupling circuits are used to find lumped parameter values for components of the first and second coupling circuits.

7. The method of claim 6, wherein the lumped parameter values for components of the first and second coupling circuits are governed by:

8. The method of claim 7, wherein Qo1 and Qo2 is less than 100.

9. The system of claim 1, wherein Qo1 and Qo2 is less than 100.