SYSTEM AND METHOD FOR CONTROLLING RESONANT WIRELESS POWER SOURCE

Cross Reference To Related Applications

[0001] This application claims priority from provisional application Ser. No. 61/679,298 filed August 3, 2012, which is incorporated herein by reference in its entirety.

Background

[0002] The invention is related to the field of wireless systems, and in particular to a resonant wireless power system that allows power to be transferred throughout with variations in matching networks efficiently.

[0003] In a wireless power system, inductive coils forming a loosely-coupled transformer are used to transfer power in a non-contact fashion. In a resonant wireless power (RWP) system, impedance-matching networks are used to cancel the reactive impedance of those inductive coils, allowing power to be transferred through the system more effectively. In an idealized RWP system with a resistive load, the matching networks can be tuned perfectly to null out all reactive impedance, creating a perfectly matched circuit. In practical RWP systems, however, this idealized model does not hold. Load power varies according to the demands of the device being serviced, resulting in a varying load impedance. The coupling factor between the source and load coils changes with the relative position of the charger and charging device. And, in mass-produced devices, the matching networks suffer from

manufacturing variations, so they are never perfectly in tune.

[0004] There are two complex impedances that can be used to describe most of the important aspects of a RWP system: the open-circuit impedance Zoc and the reflected impedance Zref. Mutual inductance, or coupling, between the source coil and receiver coil can be modeled in a number of different ways. In our preferred method, the coupling is modeled as a current-controlled voltage source in series with the receiver (secondary) coil. The impedance seen by this voltage source, which includes the coil, matching network and load (rectifier, dc/dc, load current), is Zoc. On the source (primary) side, when coupling is present, one can model the effect of the coupling as an impedance in series with the source coil called Zref, the reflected impedance. Both the open-circuit impedance and the reflected impedance are complex quantities— they have real (resistive) and imaginary (reactive) components. For a 1 : 1 RWP system, the reflected impedance is related to the open-circuit impedance by this formula:

_ ) ²

Summary

[0005] According to one aspect of the invention, there is provided a current-mode resonant power source. The current-mode resonant power source includes a source circuit that includes a source coil, an ac driver with a first resistance, representing the equivalent output impedance of the ac driver, and a matching network . A current probe measures the magnitude signal of the instantaneous source coil current. An amplifier receives and compares the magnitude signal and a target signal, and produces an error signal proportional to the difference. A compensation filter receives the error signal and produces the control voltage that determines the ac driver supply voltage.

[0006] According to another aspect of the invention, there is provided a resonant power source with power-factor correction. The resonant power source includes a source circuit that includes a source coil, an ac driver with a first resistance,

representing the equivalent output impedance of the ac driver, and a matching network. A current probe measures the magnitude signal of the instantaneous source coil current. A voltage probe measures the instantaneous ac driver voltage. A phase detector receives and compares the phase of the instantaneous source coil current and the instantaneous ac driver voltage, and produces a first output signal proportional to the phase difference. An amplifier receives and amplifies the first output signal, and produces an amplified signal. A compensation filter receives the amplified signal and produces the control voltage that determines the impedance of a variable element in the source circuit.

[0007] According to another aspect of the invention, there is provided a resonant wireless power system. The resonant wireless system includes a source circuit that includes a source coil, an ac driver with a first resistance, representing the equivalent output impedance of the ac driver, and a matching network. A current probe measures the magnitude signal of the instantaneous source coil current. A voltage probe measures the instantaneous ac driver voltage. A phase detector receives and compares the phase of the instantaneous source coil current and the instantaneous ac driver voltage, and produces a first output signal proportional to the phase difference. An first amplifier receives and compares the magnitude signal and a target signal, and produces an error signal proportional to the difference. A first compensation filter receives the error signal and produces the control voltage that determines the ac driver supply voltage. A second amplifier receives and amplifies the first output signal, and produces an amplified signal. A second compensation filter receives the amplified signal and produces the control voltage that determines the impedance of a variable element in the source circuit.

Brief Description of the Drawings

[0008] FIG 1 is a schematic diagram illustrating a series-parallel matching network used in accordance with the invention;

[0009] FIG 2 is an open-circuit impedance plot of the series-parallel matching network;

[0010] FIG 3 is a schematic diagram illustrating a current-mode resonant wireless power (RWP) source used in accordance with the invention;

[0011] FIG 4 is a plot illustrating a source impedance locus, two-receiver case used in accordance with the invention;

[0012] FIG 5 is schematic diagram illustrating a RWP source with power-factor correction used in accordance with the invention;

[0013] FIG. 6 is a graph illustrating complex-impedance plots for a RWP source with power factor correction;

[0014] FIG 7 is a schematic diagram illustrating a current-mode RWP source with power-factor correction used in accordance with the invention; and

[0015] FIG 8 is a schematic diagram illustrating an adjustable matching network for coil-current control used in accordance with the invention.

Detailed Description

[0016] The invention involves a current-mode resonant wireless power (RWP) source and power-factor correction used by a RWP source for managing variations in operating conditions in a source device. [0017] In practical RWP systems, load resistance is constantly changing based on the instantaneous demands of the load circuit, e.g. battery charger. In the most general case one must consider that load resistance can vary from infinite (no load) to some small value (maximum load). Consider the example of the series-parallel tuning network 1 in FIG. 1. When the load resistance is infinite, the open circuit impedance is simply the series combination of L2, C2a and C2b, which is purely reactive, where L2 represents the receiver coil coupled to the tuning network 1. When the load resistance approaches zero, C2b is shorted out and the open-circuit impedance is just the series combination of L2 and C2a, also purely reactive. For finite resistances, the open-circuit impedance has a positive real part and an imaginary part somewhere between the two extremes.

[0018] It can be shown that, as load resistance decreases from infinity to zero, the open-circuit impedance traces out a counter-clockwise half-circle in the complex impedance plane, as shown in FIG. 2. Generally, only use a portion of this half-circle less than one quadrant. The lower-right quadrant represents high-Q operating points, while the upper-right quadrant represents low-Q operating points. One typically tunes to use high-Q operating points. The location of the center of the half-circle depends on the impedances in the tuning network; one can choose to position it at the origin, or above or below it on the complex axis.

[0019] The nature of a multiple-receiver RWP system is such that reflected impedances always appear in series with the source coil and with each other.

Assuming that the multiple receivers do not directly affect each other's coupling (a reasonable assumption in a typical RWP system), the main mechanism by which they can affect each other is by changing the source coil current. If the source coil were driven by a current source rather than a voltage source, this interaction would be eliminated. Unfortunately, the ac power drivers that are commonly known (e.g.

Class-D, Class-E) resemble voltage sources rather than current sources. Furthermore, if there are parallel elements in the tuning network, as there typically are, the ac current in the amplifier will not match the coil current, and the interaction through reflected impedance can still occur.

[0020] However, if one creates a feedback loop to hold the ac coil current magnitude constant, the RWP driver will behave like a current-source driver and share its advantageous properties. A conceptual block diagram 2 of such a current-mode RWP source is shown in FIG 3. A current probe in series with the source coil detects the magnitude of the coil current II to generate a signal II mag. It is not important whether RMS or peak current is used for the control loop— all give the same system behavior. An amplifier is used to compare II mag to some target value, II target and produce an error signal proportional to the difference. The error signal is passed through a filter for loop compensation to produce the control voltage that determines the ac driver supply voltage Vs. The ac driver modulates the driver supply voltage Vs at a frequency determined by the clock signal CLK to drive ac power into the matching network. The resistor Rs represents the equivalent output resistance of the ac driver. The compensation filter may also provide range limiting and/or slew-rate limiting to control the behavior of the source during transients.

[0021] The magnitude of II mag is linearly proportional to the drive voltage Vs because of the linear nature of the matching network. As a result, a feedback loop is formed that forces Il mag to be equal to Il target in steady-state. This gives the RWP source the properties of a current-driven circuit. Referring to FIG 4, the source current II will be independent of the magnitude of source impedance Zs. In order to maintain the desired II, the source will increase the drive voltage Vs when Zs increases. If new receivers enter the charging field, for example, they will cause an increase in Zs (for example, a reflected impedance Zref2 is introduced due to the mutual inductance between the source coil and a new receiver). This will be compensated by the current-mode loop, resulting in an increase in Vs. Aside from possible transient effects due to finite control loop bandwidth, the other receivers will not be disturbed by the entrance of the new receiver.

[0022] In a voltage-driven RWP system, the source current Is can be measured to estimate the power driven into the source coil, since the voltage Vs is known. In a current-mode system, the situation is reversed. The current Is is a known value set by the control circuits. The drive voltage Vs can be used to give an indication of the power delivered into the source coil.

[0023] In a fixed-tuned RWP system, there is generally a nonzero phase angle between the drive voltage Vd and the drive current Is (refer to the phase angel of the impedance vector Zs in FIG. 4, for example). Equivalently one could state that the power factor is generally less than 1. This results in a lack of available power and a loss of efficiency relative to the ideal case in which the source impedance is tuned to be purely real. Power-factor correction (PFC) is a technique used in ac power

transmission systems to adaptively correct for such unwanted reactive impedances. One can apply this technique to an RWP system to null out the reactive impedance, thus obtaining the optimal power transfer characteristic.

[0024] FIG. 5 shows a block diagram 4 of a RWP source with power factor correction. A current sensor measures the instantaneous current II , while a voltage probe measures the ac drive voltage Vd. The phases of II and Vd are compared by a phase detector (labeled as "Phase diff." in FIG, 5). The output signal indicating a phase difference is amplified and then filtered for loop compensation. The resulting control voltage Vcor is used to control the impedance of a variable impedance element Zcor. Zcor can be implemented by either a variable capacitance or by a variable inductance. The control loop is configured in such a way that Zcor becomes more capacitive when the phase difference is positive. The PFC control loop thus forces the phase difference between Vd and II to zero. This is equivalent to saying that it forces the impedance Zs to be purely real.

[0025] An illustration of how PFC works using complex-impedance plots is shown in FIG 6. This graph depicts a single-receiver scenario. Between conditions 1 and 2, both the coupling factor and the load resistance of the receiver are different. The PFC control loop is able to correct for these differences by adding a negative reactance that just cancels the positive reactance in each case. The resulting source impedance (not shown) is purely resistive. If multiple loads are present, the PFC control loop works exactly the same way. It forces the resultant vector of the reflected impedance plus the correcting impedance Zcor to be purely real. If the source tuning network is more complicated than a pure series network, the PFC loop still works the same way. It may be advantageous to tune the fixed matching network to always provide an inductive impedance, so that Zcor can be a variable capacitor, or vice versa.

[0026] The voltage Vs is fixed in a PFC source, so measuring the current II can be used to estimate power driven into the source coil.

[0027] The two techniques shown so far can correct the magnitude and phase variations that occur when reflected impedance changes. Combining the two techniques yields a source design that suffers from none of the drawbacks previously highlighted for voltage-driven, fixed-tuned sources. A block diagram 6 of such a source is shown in FIG 7. The magnitude and phase of II are both fed back to separate control loops. The magnitude is used for the current-mode loop and the phase is used in the power-factor correction loop.

[0028] In the current-mode source with power-factor correction, the magnitude of the source current Is is controlled by the current-mode loop, so the voltage Vs can be monitored to indicate the source power.

[0029] In the previous examples, the method for adjusting the magnitude of the coil current II was to adjust the voltage Vs. This typically requires a

variable-output- voltage regulator to supply Vs to the power amplifier. Here an alternative technique is proposed for adjusting the magnitude of the coil current II that eliminates the need for this variable-output- voltage regulator. Since this regulator contributes an efficiency loss to the system, this alternative scheme can potentially improve the overall efficiency of the RWP system.

[0030] The general approach is to adjust the source matching network to change the transformation from II to Is, thus changing the transformation of Zref to Zs. By varying Zs, the magnitude of the II current can be adjusted while Vs remains fixed. Although many different matching circuits can be used, one can illustrate the principle using the simple parallel matching circuit 8 shown in FIG 8. The parallel capacitor CI can be adjusted by the power controller. The variable capacitor CI can be implemented using a varactor, a capacitor bank with switches, or any alternate technique. Increasing the impedance of CI to approach the impedance of LI + Zref causes the parallel combination to become more resonant, which causes II to increase relative to Is. Thus power delivered to the coil can be controlled without adjusting VI . Changing CI also changes the imaginary component of Zs, which is undesirable for reasons previously stated. One can correct for that by including a power-factor correction loop, as shown in FIG 8. The PFC loop corrects for the imaginary impedance variations as CI is changed, as well as correcting for the imaginary component of Zref.

[0031] Although the present invention has been shown and described with respect to several preferred embodiments thereof, various changes, omissions and additions to the form and detail thereof, may be made therein, without departing from the spirit and scope of the invention.