WO2012085119A2

WO2012085119A2 - A wireless power receiver for receiving a power signal over an inductive coupling, and an improvement in the method for operating a wireless power receiver

Info

Publication number: WO2012085119A2
Application number: PCT/EP2011/073643
Authority: WO
Inventors: Harri Elo
Original assignee: Harri Elo
Priority date: 2010-12-21
Filing date: 2011-12-21
Publication date: 2012-06-28
Also published as: FI20100427L; WO2012085119A3; FI20100427A0

Abstract

A wireless power receiver (6) for receiving a power signal over an inductive coupling (300) from a power transmitting circuit (100) comprising a voltage generator (101) and for passing it to a load (800), the wireless power receiver (6) comprises: - a resonator (601, 602, 603) for receiving a power signal over the inductive coupling (300), the resonator (601, 602, 603) being suitable for converting a changing magnetic flux at its input port to an AC voltage at its output port; - energy storage means (601, 602) comprising at least one inductive component (601) and at least one capacitive component (602) for storing energy comprised in the power signal received by the resonator (601, 602, 603); and - a rectifying and regulating circuitry (610) for rectifying and regulating the received power signal before passing it to a load (800). The improvement is characterized in that the rectifying and regulating circuitry (610) comprises a buck regulator (611, 612, 613, 614, 615, 616) that is configured to carry out impedance matching between the load (800) and the resonator (601, 602, 603). The patent application comprises also an independent patent claim for an improvement in the method for operating a wireless power receiver.

Description

A wireless power receiver for receiving a power signal over an inductive coupling, and an improvement in the method for operating a wireless power receiver

Field of the invention

The present invention relates to wireless power transfer, in particular to rectification and regulation of regenerated alternating voltage by a wireless power receiver.

Background art

Wireless power transfer is a new, strongly emerging industry segment. The idea dates more than a Centillium back, from onset of electrical industry. There is an application of wireless power transfer, which has recently gained plenty of publicity: recharging of batteries of mobile devices. Examples of these devices are cell phones, MP3 players, cameras, electrical books and laptops. It should be pointed out wireless power transfer is not the only and not even the most efficient way of recharging. That is, there is always the usual way of plugging a wire in the device or the battery for supply of power.

It is rather challenging to introduce wireless recharging of batteries as a new feasible technology, as its performance will always be compared to that of conventional, wired technology. Therefore, efficiency below 70 % is not considered sufficient in case of wireless recharging.

Both in transmitter and receiver of the most commonly employed wireless power transfer embodiments there is an inductive antenna, which is also the inductive portion of an LC resonator. Maximum efficiency of power transfer depends on coupling coefficient between the antennae, which in turn depends on geometry of system of the two antennae. Aiming at 70 %

efficiency, size of the antennae needs in practice to be close to equal and distance between them smaller than dimensions of the antennae. These are of course quite confining constrains, as opposed e.g. to wireless data networks. There are many applications of wireless power transfer, in case of which it is inconvenient or practically impossible to provide any kind of a wire for power supply: e.g. sensor networks embedded inside walls an medical devices within human body. In this case i.e. in absence of a wired alternative, moderate efficiency in order of 10 - 70 % is acceptable. This provides more degrees of freedom for form factor and spatial placement of the antennae. The peer antenna does not necessarily need to be in the immediate vicinity, as moderate effiency may be achieved, although distance between the antennae was larger than their size .

One example of an application within the moderate efficiency domain is Near Field Communication or NFC. In more detail, peers of an typical NFC transaction are an initiator, which generates a magnetic field, and a target, which usually does not generate a field. The initiator is able to supply the target via the magnetic field, in which case one speaks about a passive target. Thereby, the initiator is in effect a wireless power

transmitter, and the target is in effect a wireless power receiver .

In case of basically any electronic application, the prefered embodiment minimises component count, cost and size of the device, while maximising the performance. Wireless power transfer is no exception: there is a need for technology, which fulfills the said design constrains, covering also domain of moderate efficiency.

Summary of the invention

An objective of the invention is to improve the efficiency of wireless power transfer, or to enable improving of the same, especially in a situation where the distance between the wireless power receiver from a circuit that transmits power to the wireless power receiver is in the order of (as large as) the antenna size, or larger. This objective can be achieved with a wireless power receiver according to claim 1, and with the improved method according to claim 8.

The dependent claims describe various advantageous aspects of the wireless power receiver.

Advantages of the invention

A wireless power receiver that is suitable i) for receiving a power signal over an inductive coupling from a power

transmitting circuit that comprises a voltage generator and ii) for passing it to a load comprises:

- a resonator for receiving a power signal over the inductive coupling, the resonator being suitable for converting a changing magnetic flux at its input port to an AC voltage at its output port; - energy storage means comprising at least one inductive component and at least one capacitive component for storing energy comprised in the power signal received by the resonator; and

- a rectifying and regulating circuitry for rectifying and regulating the received power signal before passing it to a load .

The improvement in the wireless power receiver is characterized in that the rectifying and regulating circuitry comprises a buck regulator that is configured to carry out impedance matching between the load and the resonator.

Similarly, according to a further aspect of the invention, the method for operating a wireless power receiver to receive a power signal over an inductive coupling from a power

transmitting circuit comprising a voltage generator and to pass it to a load, the power receiver comprising:

- a resonator for receiving a power signal over the inductive coupling, the resonator being suitable for converting a changing magnetic flux at its input port to an AC voltage at its output port;

- energy storage means comprising at least one inductive component and at least one capacitive component for storing energy comprised in the power signal received by the resonator; and

- a rectifying and regulating circuitry for rectifying and regulating the received power signal before passing it to a load; and is improved by using in the rectifying and regulating circuitry a buck regulator that is configured to carry out impedance matching between the load and the resonator.

The inventor has found out that surprisingly good power transfer efficiency for wireless power transfer can be achieved by using a buck converter in the wireless power receiver for impedance matching .

If the buck regulator is implemented with a control integrated circuit operating two changeover switches, the buck regulator can be controlled via the integrated circuit. This increases the control possibilities of the wireless power receiver and so also its versatility.

If one changeover switch is implemented by means of PMOS transistors while the other changeover switch is implemented by means of NMOS transistors, the efficiency of the wireless power receiver may be increased further.

If the buck regulator is configured to perform resistive and/or reactive load modulation, semiconductor switches, resistive components and/or reactive components maay be omitted which are conventionally used for implementation of a modulator. By means of the buck regulator, efficiency or output power of the wireless power transfer system may be increased, compensating for resistive or reactive impedance mismatch, respectively. If, in beginning of a power transfer session, the microcontroller of the rectifying and regulating circuitry is configured i) to switch off the load, ii) to make the buck regulator to operate on an internal load which is constant for a period of time for determining the optimal pulse position, and iii) to switch on the load and disable the internal load, disturbances caused by switching noise that might disturb the wireless power receiver can be better avoided.

If the microcontroller of the rectifying and regulating

circuitry is configured, in steady state, to use the buck regulator to measure pulse positions in regular intervals, in particular by measuring weighed pulse widths of adjacent bins, and whenever it seems beneficial to decrement or increment the bin index employed for power transmission, the operation efficiency of the wireless power receiver in steady state can be maintained over time.

If the microcontroller is configured to request more or less power based on temporal variations in loading conditions, in particular the requesting being done over a data transfer channel from the wireless power receiver to the power

transmitting circuit, the amount of power that is received by the wireless power receiver may be controlled by the wireless power receiver itself. Preferably, abrupt changes in loading conditions are handled by means of impedance matching and long term steady state is maintained by means of data transfer between the wireless power receiver and the power transmitting circuit .

Brief description of the drawings

In the following, the invention is described in more detail with reference to the examples shown in FIG 2 to 22 of the

accompanying drawings, of which:

FIG 1 is a simplifed schematic of a state of the art

wireless power transfer system; is a simplifed schematic of a wireless power transfer system, which comprises the inventive circuit ; llustrates a combined rectifier and regulator; show different ways to implement a MOS switch; shows a controller inherent in circuitry of FIG 3; is a diagram of steady state waveforms associated with combined rectification and regulation; is a small signal equivalent of a wireless power transfer system; is a flow chart model of the wireless power receiver; illustrates implementation of an exhaustive search algorithm; is a small signal equivalent of a wireless power transfer system, which was used for comparing performance of a state of the art embodiment and tl invented embodiment of a wireless power receiver; provides a comparison between theoretical

efficiency, that of a state of the art system and < the inventive embodiment; illustrates dependencies between efficiency, impedance level and resonant frequency of a state < the art system; is a diagram, which illustrates transients, which work against natural oscillation and are inherent . state of the art embodiments; illustrates achievable maximum efficiency in case < optimal load; FIG 15 shows some examples of receiver circuit topologies known from prior art compared with the wireless power receiver according to the present invention;

FIG 16 illustrates simulated performance of different

wireless power receiver configurations;

FIG 17 illustrates the equivalent circuit for wireless

power transfer system;

FIG 18 two rectifiers known from prior art and a buck

regulator which according to the present invention is used in the wireless power receiver;

FIG 19 efficiency vs. load for two values of QK;

FIG 20 a resonator which is loaded by a periodic current source;

FIG 21 an adapted configuration of the wireless power

receiver; and

FIG 22 a transistor level implementation of the wireless power receiver.

Detailed description

Pickup of power carried by magnetic flux, regeneration of AC voltage as well as rectification and regulation of the

regenerated signal are the essential functionalities of a wireless power receiver. It is the general object of this invention to introduce an embodiment with capability to

concurrent rectification and regulation. Such embodiment is motivated in three ways: it minimises component count and hence physical size of the device, it maximises practically achievable efficiency of power transfer and it provides a means or

compensate for manufacturing variations, in particular finite tolerance of component values .

Conventionally, switch mode regulators are implemented by means of two semiconductor switches, one of which connects reactive components of the regulator to the input voltage, whereas the other switch provides a current path, while the input is detached. Likewise conventionally, the said switches are controlled by an integrated circuit, which monitors the output and adjust duty cycle of the control based on measured output voltage. By arranging these switches in a new inventive manner, AC voltage may be both rectified and regulated by means of the same circuitry. Compared to state of the art wireless power receivers, which ordinarily comprise one piece of circuitry for rectification and an other one for regulation, this is an improvement.

Key advantage of the invention is maximisation of efficiency of power transfer. The most commonly employed embodiment of a wireless power transfer arrangement is based on LC resonators, which are tuned exactly or almost at the same frequency.

Furthermore, there is an inductive coupling between the said resonators . For any quality factor of resonators and coupling coefficient between them, there is a theoretical maximum of efficiency. Basically, the invention advises a way of practical implementation, which provides the closest match between the theoretical and the realisable efficiency: due to intelligent way of power acquisition, the receiving resonator may be loaded by the impedance which maximises the efficiency.

The other major advantage of the invention is likewise based on inventiveness of concurrent rectification and regulation. In case of mass production, reactances of individual components are random variables. Consequently, manufactured receivers are never identical to the nominal design. The invented circuit provides a new degree of freedom: besides of width of current pulses, their temporal position is relevant. As the matter of fact, by controlling timing of the pulses in an appropriate manner, one is able to compensate for the said tolerances and thereby maximise both effiency and output power. In other words, a means of active power factor correction is introduced.

Yet an other advantage of the invention is related to data transmission functionality, which is usually a desired property of wireless power transfer systems. In case of an embodiment, which is built, as advised in description of this invention, one may implement a data modulator with no need to install physical components for the purpose.

FIG 1 shows a simplifed schematic of a state of the art wireless power transfer system. A transmitter 100 comprises an AC voltage generator 101, a capacitor 103 and an inductor 104. Together, capacitor 103 and inductor 104 compose an LC resonator. Resistor 102 represents inevitable losses of a practical inductor and capacitor. A receiver 200 comprises likewise a lossy LC

resonator formed by a capacitor 202, an inductor 201 and a resistor 203. Exact magnitude of the said losses is

characterised by means of quality factor Q, which in this simplified presentation is always the same in the transmitter and the receiver. Furthermore, the receiver 200 comprises a rectifier 210 and a regulator 220. The receiver is coupled to the transmitter 100 in magnetic fashion via a mutual inductance 300. Strength of this coupling is characterised by means of coupling coefficient K. The receiver 200 is loaded by a resistor 400, when ever enabled by a load enable switch 401. The rectifier 210 is a diode bridge rectifier, which is composed of four diodes and an output capacitor 211. The regulator 220 is a buck regulator, which comprises a changeover switch 221, an integrated control circuit 222, an inductor 223 and a capacitor 224. The switch 221 may be implemented by means of a MOS transistor between the rectifier 210 and the inductor 223 and a diode, anode of which is grounded, while the cathode is

connected to the inductor 223. Alternatively, a second MOS transistor may be used instead of the said diode. Furthermore, circuit elements of the switch 221 do not necessarily need to be discrete semiconductor devices, as they may be integrated in the control circuit 222.

FIG 2 shows a simplifed schematic of a wireless power transfer system, which comprises an inventive circuit implementing functionalities of both the rectifier and the regulator. The transmitter 500, the mutual inductance 700 and the load 800 are identical to their counterparts in FIG 1. Also, resonator comprising an inductor 601, a capacitor 602 and a resistor 603 is the same as the one comprising the inductor 201, the

capacitor 202 and a resistor 203. However, as parallel instead of series equivalent of the said resonator is shown in FIG 2, resistance 603 is Q2 times larger than resistance 203. It is to be emphasised this is a difference between the presentations, but not between the physical circuits.

Instead of the rectifier 210 and the regulator 220, FIG 2 shows an inventive circuit 610, which combines the said

functionalities in a single block of circuitry. Instead of the single changeover switch 221, there are now two of them: 612 and 613. By means of these switches, the resonator is connected to a MOS transistor 614, an inductor 615 and a capacitor 616.

Together with the switches, the said components compose a buck regulator. Finally, an integrated control circuit 611 of FIG 2 is the counterpart of the circuit 222 of FIG 1: it also measures the output voltage and controls switches at input of the regulator. In addition, the circuit 611 measures voltage difference between output terminals of the resonator and determines the appropriate control signal based on knowledge of both the input and the output voltage of the combined rectifier and regulator 610.

In more detail, an inductor 104 and capacitor 103 of the transmitter side LC resonator are connected in series with an AC voltage generator 101 and a resistor 102, which represents losses of the LC resonator. There are several known ways to implement the voltage generator 101. For instance, input terminals of the LC resonator may be periodically switched between some DC voltage and short circuit. As this switching is applied at resonant frequency of the LC resonator, the input signal is in effect a square wave, fundamental frequency of which is band pass filtered by the resonators and coupled to the receiver .

Due to mutual inductance 700 between the transmitter and the receiver, a proportion of the supplied power is delivered to the receiver. In particular, it is interesting to evaluate amount of power, which is delivered to a load 800. This proportion of power may be maximised by means of the combined rectifier and regulator 610.

The capacitor 602 and the inductor 601 are connected in parallel with the resistor 603, which represents losses of the receiver side LC resonator. Terminals of the LC resonator are connected to the changeover switches 612 and 613, which thereby connect the resonator to the buck converter composed of the transistor 614, the inductor 615 and the capacitor 616. The control circuit 611 is connected to terminals of the LC resonator, to control electrodes of the switches 612 and 613, and to the output node, which connects the inductor 615 to the capacitor 616.

FIG 3 shows a more detailed view of the circuit 610, which conducts both rectification and regulation. There are various kind of transistors, all of which may be used as switches:

junction and insulated gate bipolar transistors as well as junction and MOS field effect transistors. The latter are probably best suited for the invented application. In FIG 3, the changeover switch 612 is implemented by means of PMOS

transistors. The changeover switch 613 is implemented by means of NMOS transistors. The load enable switch 801 is implemented by means of a PMOS transistor, which is integrated in the control circuit 611 in case of FIG 3.

Now operation of the circuit 610 will be described in more detail. An NMOS transistor conducts, as the gate voltage exceeds the source voltage by V_T,NM0S, which is specific for the

particular transistor, often in order of 1 V. A PMOS transistor conducts, as the gate voltage subcedes the source voltage by T,PMosi which is specific for the particular transistor, often in order of -1 V. In case of the circuit 610, either of the NMOS transistors conducts almost always, as each of them conducts on every other half cycle. Therefore, either of terminals of the resonator is most of the time grounded via the switch 613.

Consequently, the switch 613 behaves in a passive manner like diodes of a diode bridge. However, PMOS transistors of the switch 612 are controlled by the control circuit 611. Therefore, inductor and capacitor of the buck converter load the resonator only by request of the control circuit, which monitors the output voltage and instantanous voltage over the resonator and determines sufficient position and duration of pulses applied on gates of the PMOS transistors. When ever either of the PMOS transistors conducts, the loop is closed and the buck converter is connected in parallel to the resonator. Consequently, a fragment of energy is transfered to the buck and finally to the load.

Regarding switches 612 and 613, not all of the details are shown in FIG 3: neither of the switches may be implemented simply by means of two MOS transistors. As MOS transistors are used as switches, a practical phenomen needs to be taken into account: in any enhancement type MOS transistor, there is always a diode embedded in the device. As signals with bipolar swing are controlled by means of MOS switches, these diodes are forward biased. This happens, as source voltage of an NMOS transistor exceeds its drain voltage or as drain voltage of an PMOS transistor exceeds its source voltage.

FIG 4a shows an approach in which two transistors 41 are connected in series in such a way that either of the bulk diodes (also known as body diodes) which is formed by the PN junction and is always reverse biased. FIG 4c shows an approach in which two transistors 43 are connected in series in such a way that either of the bulk diodes is always reverse biased.

Fig 4b shows an approach in which a resistor 45 is connected between source and bulk electrodes of the transistor 46 in series in such a way that potential leakage current through the bulk diode is significantly reduced. Fig 4d shows an approach in which a resistor 47 is connected between source and bulk electrodes of the transistor 48 in series in such a way that powetential leakage current through the of bulk diode is significantly reduced. One of the approaches shown in FIG 4a to 4d needs to be employed in order to prevent undesired leakage of the switch.

FIG 5 shows an exemplary view of the control circuit 611. It comprises a microcontroller core 5005 with the following peripherals: an analog to digital conversion (ADC) unit 5001, a differential amplifier 5011, some data and program memory 5009, a NOR gate 5013, five discrete MOS transistors 5015, 5017, 5019, 5021, 5023, a digital to analog conversion (DAC) unit 5003 and a phase locked loop (PLL) 5007. The ADC unit converts the measured analog quantities, voltage between terminals of the resonator and over the load, into digital presentation format, which is readable for the microcontroller 611. The ADC unit may comprise two separate converters or the inputs may be time division multiplexed for a single converter. Voltages of individual terminals of the resonator are accessed via input signals AIN and BIN, difference of which is determined by means of the differential amplifier and supplied to the ADC unit. The other input of the ADC unit is SUPPLY, which is also supply voltage of the control circuit. In case of a prior art buck regulator, this would be the only monitored signal.

Previously, it was stated that the switch 613 behaves in a passive manner like diodes of a diode bridge, whereas PMOS transistors of the switch 612 are controlled by the control circuit 611. For the purpose, the control circuit comprises two NMOS transistors, which control the switch 612 by driving signals AOUT and BOUT. Unless supply voltage of the

microcontroller core 5005 exceeds an implementation dependent treshold value, the microcontroller 611 is not able to operate and drive the NMOS transistors. In this case, outputs of the microcontroller 611 are in high impedance state, the NMOS transistors controlling the switch 612 are driven by AIN and BIN and the whole "diode bridge" operates in a passive manner. That is, the control circuit implements two modes of operation: it is powered up in passive mode, but operates in active mode in the steady state. In the steady state, AIN and BIN are most of the time driven low by the microcontroller 611. Therefore, fragmented sine wave instead of a continuous one is delivered to the ADC unit. This is no problem, as time dependent behaviour of a sinusoidal wave is known a priori: the fragments comprise sufficient information for reconstruction of the original signal. Basically when ever, a transistor may be a discrete device or integrated in a circuit. In the exemplary case shown in FIG 5, transistors 614 and 801 are integrated in the control circuit. The former drives signal C providing continuous current path for the output inductor 615. As neither AIN nor BIN is driven high by the microcontroller 611, the transistor 614 is driven in active mode by the NOR gate. The transistor 801 implements the load enable switch . The PLL synchronises internal timing of the microcontroller 611 to oscillation of the resonator. Based on analog to digital converted fragments of the sine wave, the microcontroller 611 calculates the phase error twice per period and delivers the result to the PLL, which is locked, as the phase error vanishes. The memory block comprises non-volatile memory for the embedded program as well as volatile memory for intermediate calculation results and temporary variables. Also, portions or all of the executable code may be copied to the volatile memory on power up for shorter access times. Purpose of the DAC unit and the transistor connected to SUPPLY will be explained below.

FIG 6 shows an example of steady state waveforms associated with combined rectification and regulation. The signals shown in FIG 6 are voltage over the resonator comprising components 601, 602 and 603, current flowing through the capacitor 602 and current flowing through the capacitor 615. The most interesting

instances of time are Δχ and Δ₂ shown in FIG 6. At Δχ , the microcontroller 611 drives BIN high, thereby driving BOUT low. In this case, the PMOS transistor connected to BOUT conducts. At Δ₂ , the microcontroller 611 drives AIN high, thereby driving AOUT low. In this case, the PMOS transistor connected to AOUT conducts. The output voltage is not shown in FIG 6. In case of a time-invariant load, the output signal is a sum of a DC voltage and switching frequency ripple, which in practice vanishes with sufficient value of output capacitance 616.

Both in case of Δχ and Δ₂, a large voltage is momentarily applied over the inductor 615 and the capacitor 616. Both Δχ and Δ₂ occur once per period. Between these incidents, the transistor 614 conducts and voltage over the output inductor and capacitor is reversed and a lot lower. Hence, current through the inductor 615 decreases smoothly, as neither half of the switch 612 conducts. At Δχ and Δ₂, the inductor 615 "steals" its current from the resonator. As current of the inductor 601 may not change abruptly, current of the capacitor 602 increases

reflecting current of the inductor 615.

FIG 7 illustrates a problem related to practical implementation of wireless power transfer systems. The circuit shown is a small circuit equivalent of coupled resonators of FIG 1 and 2. Making use of equivalent T-circuit of the transformer, the system of two resonators is reduced into simplified representation on right side of FIG 7. The simplified system comprises a sine wave generator, transmitter and receiver side resistances and an impedance Z, which may be capacitive or inductive. In case resonant frequencies of both of the resonators are equal to the signal frequency fO, Z vanishes (gets shorted) maximising the output power. Inductances and capacitances of practical mass produced

components are however never quite accurate: actual value is typically several percents higher or lower than the nominal value. 10 % tolerance is a quite typical for these components. Moreover, actual value of inductance and particularly that of capacitance is in general case temperature dependent. In challenging conditions, for instance as Q = 100 and K = 0.01, these tolerances may have significant impact on wireless power transfer: instead of vanishing, magnitude of Z could be an order of magnitude higher than RT + RR. In case of state of the art technology, one tackles the

introduced problem merely by increasing the transmitted signal voltage in the same proportion, which is lost due to impedance mismatch: e.g. by an order of magnitude, in case current is decreased that much due to the series reactance. Once again, this leads to implementation with larger and more expensive components. This could be avoided, in case there was a way of automatic calibration of the resonator, in other words a means of power factor correction. At the transmitter side, the signal frequency may be calibrated during the manufacturing in a way that it equals the resonant frequency of the resonator. However, as the receiver has no control over the signal frequency, a more inventive solution is needed.

In the steady state, the receiving resonator has no means to determine whether it is loaded by an active source or a passive impedance. Therefore, the combined rectifier and regulator appears in general case as a complex load impedance, magnitude and phase of which depend on amplitude and position of the current pulses drawn from the resonator. Based on notations of FIG 6, in case the pulses are applied within the first and the third quardant, the said impedance is a parallel connection of a resistor and a capacitor, but in case the pulses are applied within the second and the fourth quardant, the effective load impedance is a parallel connection of a resistor and an

inductor. Consequently, the inventive circuit provides a way to adjust resonant frequency of the resonator and compensate for the finite tolerances.

In case of mass production, capacitances and inductances of an individual receiver are random variables instead of

deterministic values. Therefore, the actual resonant frequency is always higher or lower than the nominal resonant frequency. Too high resonant frequency is compensated by means of parallel capacitance, whereas too low resonant frequency is compensated by means of parallel inductance. That is, the appropriate quadrant for the load pulses is selected based on sign of difference between the nominal and the actual resonant frequency. After that, the resonant frequency may be fine tuned by adjusting position of the pulses within the selected

quadrant. Besides of position of the pulses, the effective load impedance depends on their width and amplitude. The width, in turn, depends on voltage swing over the resonator and the amplitude depends on the load resistance 800. Finally, amplitude of the resonating signal depends not only on effective load on the receiver, but also on signal amplitude and output impedance of the transmitter. Unless the pulse position is optimised and the effective resonant frequency thereby fine tuned, the loop impedance may be too high, as advised in FIG 7, and the receiver may fail to provide sufficient power to the output. This problem is solved by means of the DAC unit and the transistor driven by it in FIG 5. In beginning of a power transfer session, the microcontroller 611 disables the load by means of the transistor switch 801. Instead, a well behaved internal load is provided by means of the DAC and the transistor. As the load is known to remain constant for a certain period of time, the control circuit may concentrate in determination of the optimal pulse position. In this phase, the load current should be fairly large, close or equal to the supported maximum load current, in order to emulate the most challenging conditions. After the pulse position has been optimised, the internal load is disabled and the external one enabled.

The software executed by the microcontroller 611 divides each half cycle of the resonating signal in N bins. Bin 0 is the first bin of the first and the third quadrant, whereas bin N - 1 is the last bin of the second and the fourth quadrant. Current pulses, which are applied twice per period, are aligned to one of these N bins. Choice of the bin determines phase of the effective load: bin 0 maximises capacitance, whereas bin N - 1 maximises inductance. Centers of the bins are located at temporal positions tn. In addition or instead of the described way of operation of the microcontroller 611, the microcontroller 611 may be configured to maintain impedance matching between the resonator and output power of the wireless power receiver, thereby confining

effective load impedance displayed to the resonator in vicinity of the optimal value, maximizing efficiency of wireless power transfer.

FIG 8 shows an exemplary flow chart modeling operation of the inventive wireless power receiver. The flow chart is implemented by means of an algorithm executed by the microcontroller 611. As the receiver is brought in magnetic field of wireless power transmitter, charge and thereby also voltage of capacitor 616 starts to increase (step A801), as electrical charge is

recovered from the received magnetic field. Control circuit 611 is supplied by capacitor 616 and is therefore powered up and able to operate only after voltage of capacitor 616 reaches a certain level (e.g. 3 volt) . As discussed previously, the receiver operates in passive manner (step A803) until the supply voltage is sufficient for operation of the microcontroller 611. As the microcontroller 611 wakes up, it enables the internal load and disables the external load in a manner described above. As soon as voltage of the output capacitor 616 stabilises, the software executed by microcontroller 5005 starts (step A805) an exhaustive search for the optimal bin for the current pulses. Comparison of two consequtive bins is based on width of current pulses in case of constant internal load, which is considered a stationary condition. Within the first and the third quadrant, width of the pulses is expected to decrease with increasing bin index, whereas the opposite is expected within the second and the fourth quadrant .

FIG 9 illustrates an exemplary implementation of the exhaustive search algorithm in more detail. On power up, the default bin is N/2. After the output voltage settles, width of the current pulses, W_N/2, is stored in memory. Next, adjacent bins N/2-1 and N/2+1 are tested in the same manner. In case of FIG 9, best impedance matching that far is achieved at bin N/2+1. Therefore, it does not make sense to test any other bins preceeding N/2. Instead, the search is extended further in the second quadrant. At each step, weighed pulse width

is compared to that of the previous bin. Here, unit of W_n is radians. In case, for instance, weighed pulse width of bin N/2+5 is larger than that of N/2+4, the latter bin provides the best matching between the operating frequency and effective resonant frequency of the receiving resonator, and the search may be terminated . Given n was considered the optimal bin after exhaustive search, weighed pulse width would be W_n and weighed pulse widths associated with the neighbour bins, W_n_i and W_n+i are expected to be larger than W_n. In the steady state this is confirmed in regular intervals, by decrementing or incrementing bin index, waiting until the output voltage settles and determining whether the weighed pulse width really increases. Given the neighbour bin yields a smaller weighed pulse width, the said neighbour bin should be used thereafter, until a new comparison justifies a new change of bin index. After the exhaustive search, the microcontroller 611 disables the internal load and enables the external load and waits until the output voltage settles e.g. within 5 % (step A807 in FIG 8) . From now on, the receiver operates in steady state. However, as parameters like operating frequency may vary over time, the receiver needs to keep on monitoring impact of pulse position in order to ensure the timing remains optimal. Therefore, weighed pulse widths of the adjacent bins are measured in regular intervals in a manner shown in FIG 8: in step A809 bins for t_n and t_n_i are measured. In step A811 it is determined whether it is beneficial to decrement the bin index employed for power transmission from n to n-1. If yes, in step A809 the bin index is further decremented by one. If not, in step A813 the bin index is incremented by one.

In step A815 it is determined whether it is beneficial to increment the bin index employed for power transmission from n to n+1. If yes, in step A813 the bin index is further

incremented by one. If not, the software waits until a timer expires (e.g. after one second) and returns to step A809.

When ever it seems beneficial to decrement or increment the bin index employed for power transmission, the receiver does so. A timer interrupt may be made use of in implementing the said regular interval between individual updates of the optimal bin.

In order to minimise the loop impedance at the system level, besides of the receiver, also the transmitter needs to match its resonant frequency to the operating frequency. In case of the transmitter, this is fairly straightforward, as it controls the operating frequency. In case of FIG 8, we are assuming that the transmitter has determined and frozen the operating frequency by the time microcontroller 611 of the receiver wakes up and starts searching for the optimal pulse position. In the steady state, the transmitter keeps on monitoring the loop impedance and potentially adjusting the operating frequency. This is no problem from perspective of the receiver: it maintains

synchronisation by means of the PLL and reacts to altered conditions by adjusting temporal position of the current pulses.

In case of wireless power transfer systems, it is useful to provide a data transfer channel, at least a unidirectional one from the power receiver to the power transmitter. Thereby, the receiver is able to request for more or less power from the transmitter based on temporal variations in loading conditions. In other words, the data transfer channel enables remote regulation of power transfer.

State of the art implementations of data communication are usually based on load modulation. Ordinarily, one or more circuit elements like resistors or capacitors are connected to the receiver via switches. As these switches are controlled, impedance displayed to the transmitter varies with time yielding a signal, which may be demodulated by the transmitter.

In case of the invented receiver, it is not necessary to install physical components merely for the purpose of load modulation.

That is because the very same effect may be achieved by means of an appropriate SW implementation. For instance, in case bins of quadrants 1 and 3 are used in steady state, momentary use of reverse phase bins in quadrants 2 and 4 has the same effect as reactive load modulation. Furthermore, by making use of the DAC and the related NMOS transistor of FIG 5, one may apply more sophisticated modulation schemes than the commonly used binary shift keying.

FIG 10 shows a simplified model of a wireless power transfer system, which was used for comparing performance of a state of the art diode bridge rectifier 1011 and the invented combined rectifier and regulator. In other words, performance of a passive rectifier is compared to that of an active one. The simulation was conducted with four different values of coupling coefficient K (0,003; 0,01; 0,03; and 0,1) to load 1013 (for which R_L was 40 Ω, 7 Ω, 4 Ω and 3 Ω, respectively. For each value of K, amplitude of the input signal and the load

resistance RL were chosen such way that the output voltage was ca. 5 V in case of the active rectifier. Magnitude of impedances of the reactances was 10 Ω at the resonance frequency, which was equal to the signal frequency and the same for both of the resonators. Q value of the resonators was 100.

Maximum efficiency of wireless power transfer depends on product of quality factor and coupling coefficient, QK. This dependence is represented by a solid curve in FIG 11. Performance of the active rectifier is depicted by circles and that of the passive rectifier by squares. In case of each QK, amplitude of the input signal and all of the component values are same, the rectifier circuit being the only difference. Lossy models of transistors and diodes were used in the simulations in order to yield realistic results. Based on the results, performance of the active rectifier is superior to that of the passive one: by means of the active circuit, one may get quite close to the theoretical maximum.

There is a way to boost performance of a passive rectifier by increasing impedance level i.e. increasing inductance and decreasing capacitance of the receiving resonator. This is done in a way product of the inductance and the capacitance, LC, remains the same in order to preserve the resonance frequency, but ratio L/C is increased. Square root of this ratio is the loop impedance i.e. ratio of voltage and currents swings of the resonating signal:

Z_R= (L/C) .

FIG 12 shows performance of the active rectifier by means of thick lines and loop impedance dependent performance of the passive rectifier by means of square, circle and triangle symbols. Evidently, efficiency of power transfer may be improved a lot by choosing the appropriate loop impedance, in case of which there is not much difference between the passive and the active rectifier implementation. Also, FIG 12 shows that there is no categorically optimal impedance level of a passive receiver. Instead, Z_0PT(QK) increases with decreasing QK .

Besides of the loop impedance, impact of resonance frequency of the receiving resonator is studied in FIG 12. In case capacitive component of the resonator is even 5 % below or above the nominal value, performance is severely impared. In principle, there are several ways to control resonant frequency of a resonator, and most of them are applicable also to case of a passive rectifier. However, any such means would probably call for a complex implementation and increased component count. In case of the invented active rectifier, control of the resonant frequency is based on approriate timing of the current pulses convoyed by the rectifier. Hence, no single electrical component needs to be provided for the purpose. Finally, it is to be noted that the shown results are a bit optimistic in case of the passive rectifier, as contribution of the voltage regulator is neglected. In practice, a regulator between the rectifier and the load is inevitable. The regulator has always some impact on the efficiency. To summarise, one may demonstrate pretty good performance of a passive rectifier in laboratory conditions, but there is no practical way to mass produce passive receivers, which would challenge a receiver employing the invented active circuitry in any conditions.

It is interesting to figure out howcome the passive rectifier behaves, as it does. In case of each practical diode, there is a finite forward voltage drop over the device. Instantanous power consumed by the diode equals to product of this voltage and current though the diode. Therefore, up to a certain limit, it is beneficial to increase the loop impedance, as this reduces power consumption in the diodes .

Based on results shown in FIG 12, efficiency of the passive rectifier starts to decrease, as the loop impedance exceeds ZOPT. It is true that power consumed in the diodes keeps on decreasing ad infinitum with increasing loop impedance. However, there is no way to get rid of transients shown in FIG 13 which illustrates time behaviour of voltages over inductor 201, capacitor 202 and resistor 203. Twice per period, direction of the loop current is changed, in which case roles of the diodes in the rectifier 210 are reversed. Each time this happens, oscillation is in effect stopped momentarily. There is no way to get rid of this phenomen by increasing the loop impedance. The invented active rectifier does not suffer from these transients.

Within prior art wireless power transfer, there is an embodiment called synchronous rectifier, in case of which each of the four diodes of the rectifier 210 in FIG 1 is replaced by a MOS transistor, but otherwise the circuit topology is remained the same. As was implied previously, this is the way MOS switches of FIG 3 work, before the microcontroller 611 is powered up. A synchronous rectifier is by definition an active rectifier, given the transistors are controlled by a microcontroller 611 or a respective circuit. However, qualitative behaviour of

efficiency, η(Ζ_κ), is similar to that of a diode bridge

rectifier. Now, each of the four transistors reverses its role twice per period, from cut off to conductive state and back, yielding transients similar to those in FIG 13.

FIG 14 illustrates the maximum efficiency of wireless power transfer. Almost all practical implementations are based on LC resonators, which are coupled in an inductive manner. Therefore, as can be seen in FIG 14, the maximum efficiency is strongly dependent on product of quality factor Q and coupling

coefficient K. The coupling coefficient, in turn, depends on geometry of the system i.e. form factor of the inductive antennae and distance between them.

In most applications of wireless power transfer, besides of picking up magnetic flux generated by the transmitter and turning it into voltage, the receiver needs to rectify the recovered power in order to supply DC voltage to the load. The said rectification process always consumes some power, but in most cases also introduces an effective load, which yields lower efficiency than the value shown in FIG 14.

FIG 15 shows some examples of receiver circuit topologies known from prior art compared with the wireless power receiver according to the present invention (in FIG 15 the rightmost circuit). Combination of a conventional diode bridge rectifier and a regulator (in FIG 15 the leftmost circuit) is still the most common way to implement a wireless power receiver.

Recently, some semiconductor vendors have come up with

synchronised rectifiers (in FIG 13 in the middle), in case of which some or all of the diodes have been replaced by field effect transistors (FETs).

FIG 16 illustrates simulated performance of the diode bridge rectifier, the combined rectifier/regulator and the optimal - so far never manifested - performance of the FET bridge. In this case, it is assumed that dead-time is exactly zero, which is however never a practical case. A finite dead-time needs to be applied twice per period in order to prevent short circuiting the resonator via two (/four) FETs unintentionally conducting simultanously. In case of the invented circuitry, the dead-time is not an issue. In case of diode and FET bridges, a regulator is unavoidable. In case of curves of figure 3, efficiency of this regulator is 95 %, which may be an optimistic value. Based on these assumptions, theoretical efficiency of the FET bridge comes pretty close to the theoretical optimum, but only in case of one particular value of load resistance. In comparison, the novel circuitry synthesises a close to optimal load resistance, in effect hiding the actual load.

By curves of FIG 16, it is also shown that efficiency of a FET bridge is optimised, as the load impedance is one or two decades lower than impedance of the resonator. This is bad: ordinarily, order of magnitude of the load impedance is the same or larger than that of impedance of the resonator. Examples:

• wireless recharging of batteries of mobile devices: typical resonator impedance = typical load impedance = in order of 10 Ω · Near Field Communication: typical load impedance is a lot larger than resonator impedance, which is in order of 100 Ω

In case of the invented circuitry, efficiency is optimised in case of most typical load impedances. One might argue that there is a simple means of increasing impedance of the receiver resonator in such way that ratio of load/resonator impedances is optimised for the FET bridge topology. This is actually quite true. However, impedance of the resonator may not be increased without consequences :

• With increasing impedance, voltage swing over the resonator increases as well. The more voltage components of the receiver need to withstand, the larger and more expensive they are.

• The impedance is increased by increasing inductance (and decreasing capacitance). Volume of an inductor increases with its inductance yielding a form factor, which is less suitable for devices like cell phones.

On might wonder whether the invented circuitry is very complex. However, it is actually surprisingly simple. As functions of a rectifier and a regulator and integrated, component count decreases due to redundancy. Minimum implementation of a receiver comprises nothing more than an LC resonator, one integrated circuit and output inductor + capacitor. Also, it may be shown that compared to a FET bridge, less silicon area is consumed - even less than in case of a passive diode bridge.

The wireless power receiver can be used in particular in the following applications:

• Currently, installation of furniture integrated wire- free base stations make it necessary to mill the tables in which the base stations are installed, as at most a few mm thick layer of surface material may be left on the transmitter coil. By aid of the invention, one could simply attach the coil to bottom surface of a table without sacrificing efficiency. · As was anticipated already years ago, models of

several mobile phone manufacturers are requipped with Near Field Communication (NFC) interface. An NFC enabled mobile phone may be used as a contactless credit card, for example. In this case it would of course be practical to be able to use NFC functionality even with an empty battery. With the present invention, the NFC interface of a mobile phone may be powered by magnetic field of a payment terminal . FIG 17 illustrates an equivalent circuit for basically any wireless power transfer system. It comprises two lossy LC resonators which are coupled in an inductive manner. Assuming identical resonators i.e. such a situation that - the resistance RT of the transmitter and the resistance RR of the receiver is so chosen that R = RT = RR;

- the inductance LT of the transmitter and the inductance LR of the receiver is so chosen that L = LT = LR; and - the capacitance CT of the transmitter and the capacitance

CR of the receiver is so chosen that C = CT = CR, it may be shown that the loop impedance becomes real valued, in more detail R[1+(QK)²] . Here, Q is quality factor of the resonators and K coupling coefficient between them. In other words, the transmitter is supplying receiver impedance R(QK)² via its output resistance R.

Obviously, something is missing from FIG 17 : the shown system does not actually transfer power to any meaningful load but it rather burns the supplied power in the resonators. The load may however not be coupled directly to the resonator. Instead, the resonator is ordinarily loaded by combination of a rectifier, regulator and the actual load resistance. Regulation and rectification are ideally lossless operations - in best

practical case "close to lossless". Furthermore, as the load is seen through the two said circuit blocks, it appears as R_L,EFF, which is usually not equal to R_L.

Although there are numerous ways to implement regulation and rectification, two rectifiers and the buck regulator -according to the present invention- illustrated in FIG 18 are actually a pretty representative set of the most feasible embodiments. The rectifier may be connected either in series or in parallel to the resonator; both half and full wave modes are commonly employed. By means of the synchronous buck implementation of FIG 18, efficiency of regulation may be way above 90 %. As may be predicted based on intuition, maximum achievable efficiency increases with increasing QK: QK = 1 yields at most 17 % efficiency, but 80 % may be exceeded, as QK is increased to 10. Related algebra yields also a less intuitive result: the maximum efficiency is strongly dependent on the effective load resistance, see FIG 19 for the two values of QK. This fact poses a rarely mentioned challenge for wireless power transfer: as neither QK nor load resistance are known a priori, how does one maximise efficiency? We will discuss first what should be done and right afterwards how this may be done. Obviously, we need a circuit element, which performs the said functions i.e. rectification and regulation, and does this in such manner that the optimal load is introduced to the receiver resonator - regardless on the actual load. Although not yet known to state of the art, there is actually a way to come pretty close to the said goal, hence potentially improving efficiency of future implementations.

Let us consider the receiver of FIG 20: the resonator is loaded by a periodic current source, period of which equals that of the transmitted signal convoying power. Based on simple arithmetics, i.e. Fourier series expansion of the current wave

• Fundamental frequency in-phase component of the series expansion represents the real power transferred to the load. · Fundamental frequency quadrature phase component of the series expansion represents the reactive power i.e. in effect a parallel capacitance or inductance depending on phase shift between resonator voltage and the current pulses . · Due to orthogonality, contribution of harmonic components vanishes.

In the steady state , the resonator has no way to determine whether it is loaded by an active current source or a passive impedance. That is, the periodic current source is perceived as a lumped resistor connected in parallel to an optional capacitor or inductor. Magnitude and phase of the load impedance are dictated by amplitude, duration and phase of the current pulses. There is a fundamental difference between the introduced approach and the conventional bridge rectifiers: in the latter case, the rectifier is transparent exposing the load as such to the resonator, whereas an optimal effective load may be

synthetised by means of the novel circuit. Hence, load

variations may in effect be hidden from the resonator, which yields close to optimal efficiency over wide range of actual load values. Consequently, depending on properties of resonators and the load impedance, the novel circuit may e.g. double or triple the output power with the same total power consumption of the system.

How about the imaginary part of the synthetised impedance, is this just a worthless side effect? No, there are many ways to make use of synthetised capacitance/inductance as well, for instance:

• In case of most battery recharging systems, one employs remote regulation, which is based on a closed control loop: the power transmitter receives data packets from the power receiver and adjusts its output as advised by contents of the said packets. As this kind of regulation mechanism is fairly slow, it is convenient to complement it with local regulation - based on way of adjusting the effective load reactance. That is the way of regulating the output without sacrificing efficiency.

• In case of state of the art wireless power transfer systems, data communication is based on resistive or reactive load modulation. Practical implementation calls for installation of physical switches and passive components. In case of the synthetised load, however, there is no need for physical overhead; resistance, reactance or the both may be modulated merely by adjusting amplitude and/or phase of the current pulses .

The "steady state" means that amplitude of the resonating signal remains roughly constant long enough to be considered

stationary, e.g. for some hundreds of periods. Ordinarily, this condition is fulfilled in case of wireless power transfer. We recall the buck of FIG 18 and adapt it to AC input, cf . FIG 21. Because two changeover switches are used instead of a single one, the regulator is able to handle bipolar input - and rectifies/regulates concurrently, by means of the very same circuitry. The control circuit is able to synchronise switching frequency to oscillation of the resonator and synthetise optimal load for the particular resonator impedance and quality factor.

In FIG 21, C(tco) and D(tco) are functions of a continuous angular variable and to is the instantanous phase of oscillation of the resonator. By means of the changeover switches, the output inductor and capacitor are connected periodically in parallel to the resonator. As the control signal is synchronised to the oscillation, in effect constant amplitude and polarity is introduced to the output components - in the same manner, as in case of more conventional DC to DC conversion. Most of the time, neither of the changeover switches conducts, and current path is provided via the right hand side on-off switch. See also the transistor level implementation of FIG 22.

As in case of any technology aimed for mass production, silicon area consumption and amount of external components are the things that determine feasibility. The circuitry shown on left side of FIG 22 represents probably the most compact and

efficient solution known. To improve efficiency, all diodes of the bridge rectifier have been replaced by FETs. The buck added up, we wind up with total of six power transistors and three large external reactances. Estimating silicon area consumption, impact of the control circuit in effect vanishes; only the transistors convoying "the large current" are relevant.

Components with essential contribution to size and cost of the invented circuit are shown on right side of FIG 22. We

immediately notice that we get rid of one power transistor and one external capacitor - and in addition obtain improved performance. As pulsed instead of continuous current is

delivered to the rectifier/regulator, four leftmost transistors may conduct e.g. 10 % of time instead unavoidable 50 % of the bridge rectifier case. Power reduction by a factor of 5 reduces size of the transistors by a factor 5.

The most potential target domains for the new technology are definitely - but not exclusively - LF and HF band inductive power transfer applications. Both wire-free recharging of mobile devices and NFC belong to these domains. However, the related algebra is valid more universally, not only in case of inductive power transfer. In more detail, the invention is potentially beneficial, when ever · AC signal convoying power is to be rectified prior to use.

• Source impedance of the transmitter is not negligible in respect to the impedance introduced by the load.

The conditions above are valid for quite wide range of

applications, like wide band RF energy harvesting and charge recovery from kinetic energy.

Claims

Claims :

1. A wireless power receiver (6) for receiving a power signal over an inductive coupling (300) from a power transmitting circuit (100) comprising a voltage generator (101) and for passing it to a load (800), the wireless power receiver (6) comprising :

- a resonator (601, 602, 603) for receiving a power signal over the inductive coupling (300), the resonator (601, 602, 603) being suitable for converting a changing magnetic flux at its input port to an AC voltage at its output port;

- energy storage means (601, 602) comprising at least one inductive component (601) and at least one capacitive component (602) for storing energy comprised in the power signal received by the resonator (601, 602,

603 ) ; and

- a rectifying and regulating circuitry (610) for

rectifying and regulating the received power signal before passing it to a load (800); and

wherein the improvement is characterized in that: the

rectifying and regulating circuitry (610) comprises a buck regulator (611, 612, 613, 614, 615, 616) that is configured to carry out impedance matching between the load (800) and the resonator (601, 602, 603).

2. A wireless power receiver (6) according to claim 1, wherein: the buck regulator (611, 612, 613, 614, 615, 616) is implemented with a control integrated circuit (611) operating two changeover switches (612, 613).

3. A wireless power receiver (6) according to claim 2, wherein: one changeover switch (612) is implemented by means of PMOS transistors while the other changeover switch (613) is

implemented by means of NMOS transistors.

4. A wireless power receiver (6) according to any one of the preceding claims, wherein the buck regulator (611, 612, 613, 614, 615, 616) is configured to perform resistive and/or reactive load modulation.

5. A wireless power receiver (6) according to any one of the preceding claims, wherein: in beginning of a power transfer session, a microcontroller (611) is configured i) to switch off the load (800), ii) to make the buck regulator (611, 612, 613, 614, 615, 616) to operate on an internal load which is constant for a period of time for determining the optimal pulse position, and iii) to switch on the load (800) and disable the internal load .

6. A wireless power receiver (6) according to any one of the preceding claims, wherein: the microcontroller (611) is

configured, in steady state, to use the buck regulator to measure pulse positions in regular intervals, in particular by measuring weighed pulse widths of adjacent bins, and whenever it seems beneficial to decrement or increment the bin index employed for power transmission.

7. A wireless power receiver (6) according to any one of the preceding claims, wherein: the microcontroller (611) is

configured to request more or less power based on temporal variations in loading conditions, in particular the requesting being done over a data transfer channel from the wireless power receiver (6) to the power transmitting circuit (100).

8. An improvement in the method for operating a wireless power receiver (6) to receive a power signal over an inductive coupling (300) from a power transmitting circuit (100) comprising a voltage generator (101) and to pass it to a load (800), the power receiver (6) comprising:

- energy storage means (601, 602) comprising at least one inductive component (601) and at least one capacitive component (602) for storing energy comprised in the power signal received by the resonator (601, 602, 603) ; and

- a rectifying and regulating circuitry (610) for

wherein the improvement is characterized by: using in the rectifying and regulating circuitry (610) a buck regulator (611, 612, 613, 614, 615, 616) that is configured to carry out impedance matching between the load (800) and the resonator (601, 602, 603).