GB2596903A - A hybrid metal dielectric resonator - Google Patents

Abstract

A resonator that permits the efficient transfer of electromagnetic energy at very high frequencies and high powers while reducing or eliminating the difficulties (e.g. skin effect) that afflict conventional copper coils is disclosed. An embodiment of the resonator comprises at least one conductive portion 40, 50 and at least one dielectric portion 60, 70 arranged in a continuous path, for example a ring or an oval. the resonator has a major dimension (overall dimension of the resonator substantially perpendicular to the B field generated by it) to minor dimension (dimension of the conductor cross section) ratio of between 15:1 and 3:1. The resonator operates on the basis of a combination of a displacement current in the dielectric portions and a conduction current in the conductive portions rather than a pure conduction current, which reduces or eliminates many of the disadvantages of conventional conducting coils at very high frequencies.

Description

A HYBRID METAL DIELECTRIC RESONATOR Field of the Invention Transmitting energy from an electromagnetic coil driven by alternating current (AC) is known. Applications include wireless charging (via a suitably located receiver coil), induction heating (for cooking hobs and industrial uses) and chemical processing among many others. Wireless charging (also known as inductive power transfer) is widely used for low power applications such as mobile phones, but its application is more problematic for higher power applications due to limitations imposed by physical laws.

Using wireless charging for electric vehicle applications is attractive for several reasons beyond convenience. More readily-available charging would permit the size of the onboard vehicle battery to be reduced, and there would be no issues around damaged charging contacts on vehicles used in hostile environments such as at sea. Wireless charging also permits opportunity charging when a vehicle is queueing or loading/unloading, so there is no additional downtime caused by the charging process. It also allows the possibility of dynamic charging (i.e. charging while a vehicle is on the move).

Some wireless vehicle-charging arrangements have been demonstrated, but the charge rate is disappointing and precise alignment between the charging station transmitter coil(s) and receiver coil(s) on the vehicle is critical. Heavy duty vehicles (whether road-going or category G off-road vehicles used in agriculture, construction or mining, for example) cannot be charged quickly enough for wireless charging to be a viable option. Such vehicles typically require much greater ground clearance than cars which provides a further challenge due to a large distance between ground-mounted and vehicle-mounted coil(s). Providing heavy duty vehicles with wireless charging capability is particularly important as net-zero carbon incentives and policies proliferate around the world.

The efficiency of the energy transfer in these known systems is also poor (at least when compared to wired charging) and attempts to improve charging rates will also increase losses.

The present invention seeks to ameliorate these drawbacks. Summary of the Invention According to a first aspect of the present invention there is provided a resonator for transferring electromagnetic energy in free space, comprising one or more conductive portions and one or more dielectric portions arranged alternately with the conductive portions in a continuous path, the resonator having a major dimension to minor dimension ratio of between 15:1 and 3:1, such that an alternating displacement current can be induced into the dielectric portions of the resonator and an alternating conduction current can be induced into the conductive portions of the resonator.

The only way to address the difficulties outlined above is to increase the power density in the magnetic field. However, existing systems based on copper coils are subject to power limits imposed by laws of physics and efficiency levels.

Dielectric ring resonators (DRR) may address at least some of the power transfer and efficiency limitations of the known conductive coils but have other difficulties, particularly regarding cost, weight and manufacturing, and the resonant frequency being dictated predominantly by the dielectric material and resonator geometry. This first aspect of the present invention provides a metal-dielectric hybrid ring resonator (HRR) with many of the benefits of DRRs but which is easier to manufacture and has lower cost and weight, and flexibility in setting the resonant frequency independently of dielectric material and resonator geometry.

The ratio between the major dimension (overall dimension of the resonator substantially perpendicular to the B field generated by it) and minor dimension (dimension of conductor cross-section) is important. The resonator requires a relatively shallow form factor to radiate the B field effectively. A more preferred range for this ratio is between 12:1 and 4:1.

While the resonator may comprise a single conductive portion and a single dielectric portion, it is preferred that the resonator comprises at least two conductive portions and at least two dielectric portions in order to limit electric field strength between adjacent conductive portions.

A resonator having three conductive portions and three dielectric portions is a good compromise. The preferred layout for E-field mitigation is obtained by evenly arranging the dielectric sections.

The conductive portions of the resonator preferably comprise hollow conductive tubes to reduce weight and manufacturing cost. The conductive portions preferably have a conductivity of at least 3.5 x 10 S/m. Such tubes may be made of any metal, but copper is preferred for reasons of cost and low ohmic resistance. Silver would also be suitable. The resistance may be further reduced by coating the conductive portions with graphene. Tubes have the additional advantage of providing a channel for liquid cooling, in applications where this is necessary.

The resonator comprises at least one part where a conductive portion is interrupted by a dielectric portion. The conductive portions on each side of the dielectric portion are preferably substantially aligned (i.e. overlap completely) for reasons of efficiency.

A number of dielectric materials (such as ceramics or other dielectrics commonly used in capacitors particularly high voltage capacitors are suitable) but sapphire is preferred due to its low loss and high breakdown voltage. Ideally the dielectric will have a relative permittivity of at least 5, preferably at least 7.5.

The shape of the resonator is dictated at least to some extent by the application, but is generally toroidal. A circular plan view is the most efficient for projecting the B field to a distance, if not the easiest shape to make. The cross section is generally ovaloid, although circular is easier to make, but some advantage is often obtained by having a different cross sectional shape, particularly an oval. The resonant frequency of the resonator is determined by a number of factors including the material and thickness of the dielectric portions and the degree to which the dielectric portions overlap the adjacent conducting portions. By moving the conducting portions closer to/further from each other or by moving the dielectric substantially perpendicularly to the adjacent conductive portion, frequency control can be achieved. This process may be automated in use.

Another factor that affects the resonant frequency is the ratio of conductive portion or portions to dielectric portion or portions. This ratio can vary over a wide range and can also be used to adjust the quality factor or Q of the resonator. A ratio of less than 5% dielectric portion is preferable The invention is advantageously used in conjunction with frequencies of at least about 20 MHz. It will generally be used at frequencies less than about 500 MHz but higher frequencies may be used. Dimensions work well in the range of about 27-300 MHz. It is particularly useful for high power transmission in the range about 27-90 MHz.

A preferred embodiment of the resonator is designed to operate at substantially 40 MHz in the industrial, scientific and medical (ISM) band.

Embodiments are particularly useful for high power transmission, particularly with power measured in kW, preferred embodiments operating above about 10 kW, preferably above 50 kW.

In order to improve efficiency of the resonator, it is preferably located in a conductive cavity. For some applications, the cavity will need to be open on one side providing a cavity which at least partially surrounds the resonator. VVhere the resonator is circular, a radial cavity is preferred. The cavity preferably has a diameter of less than or equal to double the diameter of the resonator, preferably less than 150% of the diameter of the resonator and still further preferably only slightly greater, for example 10% greater, than that of the resonator. The cavity is optional, however.

In order to excite the resonator (or derive an output therefrom), an electrical coupler is provided for exchanging electromagnetic energy with the resonator. The electrical coupler is arranged in proximity to the resonator, for example in a co-planar or co-axial location. The electrical coupler may comprise a conducting coil or (at higher frequencies) a waveguide. The electrical coupler is preferably substantially smaller than the resonator, typically having a diameter of less than or equal to half that of the resonator, more preferably having a diameter of less than or equal to a fifth of that of the resonator. In one embodiment it is preferred that the electrical coupler is arranged off-axis from the resonator and alternatively or in addition, the electrical coupler may be arranged co-planar with the resonator.

Where the resonator is arranged within a cavity (open or closed) the electrical coupler is preferably arranged in proximity to the cavity wall (closer to the cavity wall than the resonator) to excite a displacement current in the resonator.

According to a second aspect of the present invention there is provided a system for wireless transfer of power comprising a first resonator structure and an electrical coupler according to the first aspect of the invention, and a second resonator structure, the system further comprising means for introducing energy into the electrical coupler of the first resonator structure, and means for retrieving energy from the second resonator structure.

Energy may be extracted directly by a conventional coil (i.e. the second resonator structure is a metal coil), particularly for low-power applications such as charging a bicycle or small car.

However, for high-power applications it is preferred that both the transmit side and receive side comprise a H RR.

According to a third aspect of the present invention there is provided a wireless power transmitter comprising a resonator structure and an electrical coupler according to the first aspect of the invention, further comprising a source of alternating current coupled to the electrical coupler.

This aspect of the present invention may comprise the energy transmitting portion of a system according to the second aspect of the present invention or it may be arranged to transfer energy to various objects. This may have application in industrial processing such as steel or chemicals. Alternatively, it may be used to provide an induction hob or a means of eliminating pests as described in UK patent application GB2562765A and international patent publication The third aspect of the invention may also be applied to detection of conducting objects by measuring the effect of external objects on the Q factor of the excitation coil. This could be applied to metal detecting or identifying the presence of (conducting) pests on (nonconducting) crops.

According to a fourth aspect of the present invention there is provided a wireless power receiver comprising a resonator structure and an electrical coupler according to the first aspect of the invention, further comprising rectifying means coupled to the electrical coupler to provide a DC power output.

This aspect corresponds to the receiver-side of the system according to the second aspect of the present invention, when the receiver-side of the system comprises a HRR (i.e. suitable for high power transfer).

Brief description of the drawinqs

The present invention will now be described by way of example, with reference to the accompanying drawings, in which: Figure 1 shows a first view of a first embodiment of a hybrid ring resonator according to the present invention, Figure 2 shows another view of the first embodiment with dimensions selected to provide a preferred resonant frequency, Figure 3 shows the resonance frequency with respect to the angular proportion of dielectric, Figure 4 shows a HRR of the first embodiment together with an excitation coil and an optional resonator cavity, Figure 5 shows the delivered power with respect to a ratio of major diameter to minor diameter of the resonator, Figure 6 shows a cross-sectional view of a dielectric portion of a hybrid resonator ring, Figures 7 to 9 are graphs that illustrate how the resonant frequency of the HRR may be selected and adjusted, Figures 10 and 11 are graphs that illustrate the quality factor improvement of embodiments of the present invention, Figure 12 shows a perspective view of a pair of hybrid ring resonators arranged for charging a vehicle, and Figure 13 shows a perspective view of a resonator and an electrical coupler arranged in a conductive cavity.

Detailed description

The embodiments of the present invention are based upon the principle of inductive power transfer. Inductive power transfer systems may be thought of as a transformer (having primary and secondary coils) without a conventional core. The quantity of energy transferred or power density is proportional to the square of the operating frequency multiplied by the square of the magnetic flux density (i.e. proportional to f2B2).

Existing wireless charging technologies are slow because, even though they rely upon large copper coils around ferrite cores, the power density (and hence the energy transferred) is low. In addition, transfer distances are poor and imprecise alignment (axially or angularly) between the transmitter and receiver coils has a significant effect on efficiency, so vehicle batteries regularly do not receive full charge.

The reason is that existing technologies are subject to physical laws that limit the operating frequency and the magnetic flux density. Since power density is proportional to the squares of these two factors, charging rates are consequently limited.

A high alternating magnetic flux density is hard to generate since the correspondingly large alternating current required entails high ohmic losses in the electrical conductors involved, which typically may be copper, due to the well-known skin effect. The skin resistance of the electrical conductor results in a low unloaded quality (Q) factor, and has a profound effect on efficiency as is discussed further below.

Considering the operating frequency, state of the art vehicle charging systems operate at kHz and are highly optimised, given the technology upon which they are based. The skin effect is already significant at this frequency (increasing the ohmic losses) and the limitations imposed by ferrite cores also increase with increasing frequency. Consequently, increasing the frequency will lead to unacceptable power losses.

By contrast, the embodiments described below generate an alternating magnetic field using a combination of displacement currents in a dielectric part of a ring and conduction currents in a conductive part of the ring rather than continuous conduction currents in metal coils. Since displacement current is a volumafic effect, there is no flow of electrons and no ohmic heating. There is thus no frequency limit nor current limit imposed by the skin effect in the dielectric.

A pure dielectric ring resonator (DRR) of suitable geometry for inductive power transfer applications made, for example, of ceramics is expensive and heavy with general geometry and dielectric material dictated by the required resonant frequency making them unsuitable for most vehicle applications, especially in a weight-critical application like a drone. A pure DRR operates solely using displacement currents.

The following embodiments of the present invention comprise at least one hybrid conductor/dielectric resonator constructed from at least one conductive portion and at least one dielectric (insulating) portion. Such arrangements provide much of the benefit of a pure DRR while being significantly lighter and cheaper to manufacture, with flexibility to choose the general geometry and dielectric material independently of the required resonant frequency.

The resonator form, geometric characteristics and composition minimise conductor skin losses, permitting operation at high frequencies while maintaining a high quality factor (Q), thereby permitting efficient transmission of more power over greater distances.

The advantage in efficient power transmission over that for copper coil systems stems from being able to operate at a higher frequency with low losses, by virtue of a resonator design that minimises conduction losses at high frequencies.

In addition to the above, the resonator detailed construction confers tunability and the ability to operate with high levels of stored energy for transfer of high levels of power over relatively large distances. A further key point of the HRR is the flexibility it affords to set any desired resonant frequency within a certain range applicable for a given geometric scale, rather than be constrained to what is made possible by selection of different dielectric materials.

The present embodiments have been designed to operate at 40.68 MHz, which has been found to be particular effective for electric vehicle charging. The near field at this frequency is up to approximately 4 metres and so it also enables a useable range for vehicle charging.

Another attraction is that 40.68 MHz is designated by the International Telecommunications Union (ITU) as an industrial, scientific and medical (ISM) band, which has more relaxed limits on radiated power and is available for use around the world.

However, there is no reason why the principles disclosed could not be applied to systems having operating frequencies between 1 MHz and 1 GHz with appropriate adjustment of various parameters as discussed further below. Embodiments of the present invention are particularly effective in the Very High Frequency (VHF) band between approximately 27 MHz and 300 MHz.

If a is the ratio of frequencies being compared for a system with the same overall dimension, then: Ratio of power input required for unloaded system to generate the same f.B = a312.

For 85 kHz/40 MHz, a = 2.1 x 10-3. So the ratio of power input required for unloaded 85 kHz system with respect to the 40 MHz system is approximately (2.1 x 10-3)-312 = 104.

This means that 10,000 times more power input is required from an 85 kHz WPT system to achieve the same power transfer capability as a 40 MHz system, on a like-for-like basis. Conversely, for a given input power dissipation, the output power capability will be 10,000 less for an 85 kHz system than for a 40 MHz system. This is regardless of the number of turns in the transmitter coils.

Some limited compensation can be made for 85 kHz system by use of ferrite cores. However, ferrite cores are massive and expensive.

Another way to compensate for this is to increase the number of turns in 85 kHz system pickup loop (i.e. the receiver coils). The power output is proportional to f2.B2 at the location of the receiver, where f is frequency and B is magnetic flux density, and is 10,000 times less for an kHz system than for a 40 MHz system with a single turn and a given transmitter input power. The induced EMF in the pickup coil is proportional to f.B, which is A/10,000 = 100 times less for the 85 kHz system. For the same input power for 85 kHz transmitter, we therefore need approximately 100 more "ideal" turns on the receiver than for a 40 MHz system. We now need to take into account that the turns are not "ideal": having 100 more turns increases the resistive losses of the 85 kHz system receiver by a factor of 100; additionally, having more turns means that each turn is less efficient due to the magnetic shielding of tightly-pack coils. Hence, 100 extra turns does not fully compensate for the 10,000 times less power transfer. Increasing the number of turns still further will contribute increasingly less per turn, and therefore starts to become a limiting factor after around 100 turns. This also means having times more copper mass, which is expensive. It also adds weight and bulk to the receiving structure on the vehicle.

Additionally, the value of B drops rapidly axially from the centre of a WPT transmitter coil. At 1 coil radius above coil, B will have reduced by a factor of 2.9, and so B2 will have reduced by a factor of 8.4. The drop off is not linear. However, it can be shown that for an equivalent input power and output power, the transfer distance of a 40 MHz WPT system works out to be 6.4 coil radii compared to an 85 kHz WPT system at 1 coil radius. This means that for a coil radius of 250 mm, if an 85 kHz system can project 250 mm with sufficient performance, then a 40 MHz system can project 1600 mm.

Figure 1 shows a first embodiment of the hybrid ring resonator (H RR). The H RR comprises a first U-shaped conducting portion 40 and a second U-shaped conducting portion 50 arranged in an oval. Between each end of the conducting portions are arranged two insulating portions shown as dielectric plate assemblies 60 and 70. The figure further shows a larger scale view of one of the dielectric plate assemblies 70. The end of copper tube 40 is attached to a flange 41, which may be welded to the tube or part of an extruded or machined tube. The flange may be chamfered or sharp-edged. The remaining flange 51 and those at the other ends of the tubes 40, 50 are similar. Structural or support features can be fitted to the tubes 40, 50 as necessary.

The conducting portions may be constructed from any material that conducts electricity but, to maximise efficiency, the resistance needs to be low at the operating frequency. Copper is a good choice and is cheap but a range of other materials such as silver, silvered copper or graphene-coated copper could be used. When using coated materials, the depth of the coating needs to be of a suitable depth for the skin depth of the material at the relevant frequency. The thickness of the conducting tubes is not important, provided that they meet this minimum thickness, and may be dictated by the structural strength required.

It is worth noting that the cross-section of the conducting portions need not be circular, the same, or even be consistent within a single conducting portion. The cross section could be elliptical, rectangular and so on and may change at any point around the path of the HRR. However, a rounded profile is preferred for projecting magnetic energy.

The ends of the conducting tubes can be the same cross-sectional shape as the tubes or could be larger with a fitted end-plate or smaller in the case of a pointed tube. The size and shape of these ends has a strong effect on the resonance frequency and will preferably be modelled using available simulation packages.

The dielectric plate assembly 70 comprises a pair of dielectric plates 71, 72 with a central join 80. While two plates are shown, a single plate may be used and is preferred for some applications. However, more plates can be used as desired. The dielectric plates may comprise barium strontium titanate (BST) mixtures or aluminium oxide (alumina) in either its amorphous or crystalline forms (i.e. sapphire). Suitable liquid dielectrics include silicone oil, pentadecane, paraffin oil and so on as will be apparent to the skilled reader. One preferred material is sapphire with the optical "C" axis of the sapphire crystal oriented normal (i.e. perpendicular) to the plate's faces. This arrangement provides the lowest loss, has a dielectric constant of 11.5 and a loss tangent of 8.6E-5.

At high powers, the peak magnetic energy of the source (transmitting) resonator must be high. The current circulating in the resonator is therefore high and the peak voltages developed between the ends of the conducting portions will be high. Consequently, the electric field across the dielectric portions will be high, making it essential that a dielectric with high breakdown strength is used. Air is not a sufficient dielectric at high powers. The dielectric must also be relatively low loss.

For the same reason, air must also be eliminated at the interfaces between conductive portions and dielectric portions. Air present in voids at the interfaces would be exposed to significant displacement current density and be prone to breakdown with consequent damage to dielectric and conductors, and additional power dissipation. One means of achieving this is to fuse the ends of the conductive portions to the dielectric. Care must be taken to ensure that the dielectric is not contaminated by the conductor(s) which could cause significant local degradation of the dielectric material.

Another solution is to use a fluid dielectric in addition to a solid dielectric to ensure that there are no air gaps between the dielectric plates and the flanges on the conducting tubes. A thin layer of liquid dielectric having sufficient dielectric constant and breakdown strength to sustain a high displacement current density is used. It must also exhibit low loss at the operating frequency to avoid excessive dissipation of power and heat. Using a thin layer also mitigates the dielectric losses in the liquid. Suitable liquid dielectrics include pentadecane (or other long-chain alkane) that is liquid over the required temperature range and preferably has as high a boiling point as possible. Another option is silicone oil (polydimethylsiloxane) which is less volatile than pentadecane but has better materials compatibility and may also be cheaper. Figure 6, discussed below, shows construction details.

It is also important to eliminate air breakdown paths around the dielectric portions by taking interface precautions. These might include wrapping the dielectric around the ends of the conducting portion or extending the dielectric portions outwards to lengthen any air path between adjacent conducting portions.

The dielectric plates join the end of the tube in a disc-shape in this example but if the ends of the tube were rectangular then the join would be rectangular, assuming that the dimensions of the dielectric plate exceed those of the end of the tube. A perfect 100% overlap between the dielectric plate and the tube is not necessary. Reducing the overlap will affect the resonant frequency and using this phenomenon to control the frequency (e.g. to compensate for the effects of a load) is discussed in more detail below.

The dielectric plate assemblies (where there are two or more) need not be identical and variations as discussed can be applied to either or both assemblies 60, 70.

The alignment of the conducting portions on either side of a dielectric portion is also important.

In the figure, they are shown as perfectly aligned (i.e. 100%) but, while they do need to overlap, they do not need to be perfectly aligned. They could be offset by, say, 20% or even 50% but this will negatively impact power transfer efficiency. 100% alignment is the most efficient.

Figures 2a and 2b show an engineering drawing of the components of an embodiment of the hybrid ring resonator(HRR) including dimensions in millimetres. The thickness of the tube walls is 3.25 mm. These dimensions provide a resonant frequency around 40 MHz. Exact tuning of the resonant frequency will be discussed further below. Note that the figures show a long side and a short side (68 mm shorter at each end) of the HRR. However, it is important to realise that the conducting tubes 40, 50 do not need to be identical, nor do they need to be U-shaped. The size and shape of the HRR strongly affects the resonant frequency and the shape of the magnetic field. Readily-available simulation tools permit the design and modelling of resonant frequency and B-field parameters.

Equally, the HRR could comprise only one conductive portion and only one dielectric portion or could comprise more than two of each. The number of portions provides a trade-off between complexity and electric field strength across the dielectric as discussed further below. The preferred number of conducting portions is three.

The resonant frequency may be determined by the length of the dielectric portion(s) relative to the conductive portion(s). Figure 3 shows the resonance frequency with respect to the angular proportion of dielectric. This assumes a circular resonator in which the dielectric portions sum to a total angular proportion of D/Do. D = Do represents a resonator completely made from dielectric (i.e. a DRR). D = D0/2 represents a resonator that is exactly half conductive and half dielectric. The graph shows that lower proportions of dielectric reduce the resonance frequency, giving significant design flexibility over the size of the resonator for a given operating frequency. For example, a 25% proportion of dielectric results in the resonance frequency being halved. This also allows cheaper, lower permittivity material such as alumina or sapphire to be used rather than expensive higher permittivity materials. One drawback of low proportions of dielectric is that the quality factor Q will deteriorate due to greater losses in the conductive portions. Q and its consequences are discussed further below.

Embodiments of the present invention may be constructed with 10% or less and preferably 5% or less of the path of the resonator comprising dielectric. The example shown in Figure 1 has a dielectric portion of around 1%.

Different dielectric materials may also be used to tailor the resonance frequency as the resonance frequency is inversely proportional to the square root of the permittivity of the dielectric.

Figure 4 shows a hybrid ring resonator 30 of Figures 1 and 2 located in a metallic walled cavity 10 with an open lid 11 and a metallic base 12. The HRR is located such that the mid-point in the xy-plane is co-planar with the top of the cavity walls, but it can be located anywhere within the cavity (if present). The HRR is driven by an excitation coil 20. The cavity is shown as a rectangular box 210 mm high with the other walls being 765 mm by 925 mm. However, the cavity may be circular or other suitable shape for the application. It may be constructed of a conducting material such as copper or silver, or a material coated in graphene or other material to a thickness determined by the skin depth of the material at the relevant frequency. The cavity is not essential but increases efficiency, as discussed below, by increasing the -unloaded of the resonator.

HRRs may be manufactured up to metre scales in any shape needed to obtain the correct frequencies to project B fields across metre scales in the near field. While the HRRs described here are single turn, providing multi-turn HRRs is equally possible.

The excitation coil 20 is required to couple energy into the HRR and may be a conventional copper coil or (at higher frequencies such as 915 MHz) a waveguide. At 40 MHz a waveguide is likely to be too large so a coil is preferred. The placing and parameters of the coil will be dependent upon operating frequency, power and so on. In this example it is located below the HRR in a central position in the cavity and comprises a single turn with major radius 57.6 mm and minor radius 6 mm. The position of the coil can be adjusted manually or automatically (i.e. electronically) to optimise the efficiency of excitation.

While the excitation coil is shown located beneath the HRR, this need not be the case. The excitation coil may be located to the side (i.e. in the same plane as the HRR) or elsewhere dependent upon the application.

In order to drive the excitation coil at high powers, a high-power, high-frequency RE generator is required. One such generator is the RFG-30K-AC-40 available from Coaxial Power Systems of Eastbourne, United Kingdom www.coaxialpowe,r.com. This solid state generator has a power rating of 30 kW at 40 MHz. Higher powers may be achieved by parallelizafion of such generators or by use of the same company's tube/triode model RFG100k with an output power of 100 kW. A preferred embodiment operates at 400 kW, which may be achieved by using four such generators arranged in parallel.

To extract energy from the arrangement, a further coil or HRR structure is required. In order to derive a modest amount of energy, all that is required is a conventional metallic coil located, say, 400 mm above the HRR. This would be sufficient to charge a small electric vehicle such as a bicycle or a small car. A larger coil will be required to charge larger vehicles in a reasonable time and Figure 12 shows an arrangement that uses a H RR on both the transmitter and receiver sides.

While this embodiment shows the hybrid ring resonator (HRR) 30 mounted in a cavity 10, this is not a requirement. However, the cavity does increase the unloaded quality factor Q and therefore power transfer efficiency as discussed further below. The cavity need not be a complete cavity and an open top (e.g. 5 sided) structure will improve the unloaded quality factor and still permit a second HRR, such as an HRR mounted in a vehicle, to receive electromagnetic energy from the HRR 30.

The present embodiment comprises hollow copper pipes but other materials and structures are possible. The cross-sectional shape of the conductors need not be circular -it could be square, rectangular or other shape to suit the application -although a rounded profile is preferred. The cross-section need not be the same throughout the length of the conductor.

Furthermore, since the electrical current is confined to the outer portion of the conductor due to the skin effect, an insulator coated in graphene would serve as a suitable conductor and indeed a metal conductor could be coated with graphene to improve efficiency.

As discussed further below, the efficiency of the arrangement is such that very little power is dissipated by the HRR. In many applications it may be designed to run warm and dissipate a small amount of heat directly to the air (or used forced air cooling). Alternatively, liquid cooling is easy to provide in a hollow tube using known liquid coolants such as water or SF6. Where liquid cooling is used, it may be beneficial to make the walls of the tube as thin as possible, commensurate with the skin depth at the relevant frequency.

At first sight the hybrid resonator according to some aspects of the present invention might appear similar to a loop gap resonator (LGR). However, LGRs are unsuitable for the present applications. LGRs are typically arranged in a cylindrical shape and have an air gap dielectric. The purpose of such resonators is to produce a uniform B-field in the centre of the resonator as required by Electron Paramagnetic Resonance spectroscopy (EPR) or Nuclear Magnetic Resonance (NMR) applications. The length of the resonator is consequently much greater than the diameter and this largely confines the B-field within the resonator. A typical example has a radius of 32.5 mm and a length of 100 mm. By contrast, the purpose of the present hybrid resonator is to deliberately project the B-field which requires a somewhat different form factor (diameter of the conductive portions relative to the overall diameter of the resonator).

Another point to note is that LGRs can only accommodate a relatively small gap since asymmetry in the B field will result from larger gaps.

In the example described above, the major diameter of the resonator conductive portion is 400 mm and the minor diameter is 70 mm giving a form factor ratio of approximately 6:1. In practice this can vary between 15:1 at one end of the spectrum and 3:1 at the other end.

Preferably, the form factor ratio is between 12:1 and 4:1. More generally 0.e. given that the resonator and conductive portions thereof need not be circular) a major dimension comprises the largest dimension of the resonator substantially perpendicular to the B-field within the resonator and a minor dimension comprises the largest cross-section of the conductive portions (excluding any flange for dielectric interface etc.) If we assume that the form factor ratio (major dimension:minor dimension) and the relative thickness and material of the dielectric portions stay the same, we can examine the effects of making the resonator bigger and smaller as follows: Since inductance is proportional to the linear scale factor, a larger resonator will have a larger inductance. Since the resistance of copper at a given frequency is the same, the unloaded Q factor will increase linearly with the size of the resonator. Thus a larger resonator will be more efficient, but the effects may not be critical as will be explained with reference to the efficiency equation below.

The maximum peak current that can be handled (for maximum tolerable displacement current density) is proportional to the square of the linear scale factor and so the maximum peak possible magnetic energy is proportional the fifth power of the linear scale factor. As shown in Figure 5, a larger resonator (i.e. with length scale T > baseline length scale To) can transfer considerably more energy or, conversely, a small increase in resonator size can yield large improvement in energy transfer. For example, a 10% increase in size (T/To) will yield a 61% increase in power (P/Po), a 50% increase in size (T/To) will yield 659% more power (P/Po), and a 151% increase in size (T/To) will yield 100 times more power (P/Po). The range of the resonator will also improve considerably.

The relationship of fifth power between linear scale and projected energy comes about from the following equation: U = (LP)12 L is proportional to size and 1 is proportional to the square of the size. I squared is thus proportional to the fourth power of the size. The magnetic potential energy U is therefore proportional to the fifth power of the size, and the transferred power is proportional to U when the loaded Q is kept constant.

While a resonator having an oval shape has been described, other shapes are possible. While a circular shape is the most efficient for projecting the B-field to a distance, it may pose manufacturing challenges. It is often easier to locate the dielectric portions on short, straight sections of the conductive portion as described.

Although the present description focusses on HRRs used to charge electric vehicles, there are a large number of other applications to which the HRR is well suited. Indeed, the HRR is suited to any application that uses near-field inductive power transfer such as flash heating steels, shrink wrapping and chemical processing. The higher frequencies permitted by the use of a HRR will be of benefit in a number of applications. For example, in surface hardening of metals, the skin depth will be lower, meaning that less material is heated and efficiency is increased. In chemical processing, smaller conductive particles (having higher resonant frequencies) can be loaded into a polymer. Typical state of the art devices are limited to around 3 M Hz.

Another application of the HRR is agricultural invertebrate pest control using the differential electrical conductivity between plants and animals to kill or incapacitate pests without damaging the crops and without recourse to chemical pesticides. This also works for vertebrate pests such as rodents or moles. This is described further in UK patent application GB2562765A and international patent application W02018215975A.

A related application, also described in the above patent publications, is humane stunning of vertebrates such as sheep and broilers in a slaughterhouse which is Halal compliant. The HRR is used to warm the brain of the animal by a few degrees to induce a coma prior to dispatch using known methods. Alternatively, the HRR could be used to dispatch the animal directly.

Other uses of near fields, that require minimal power transfer, include detecting conductive materials such as metal detection or detecting pests such as slugs on a crop since the slug presents a high load to the HRR compared to the plant. The slug causes a change in Q of the system which can be detected and plotted as a function of location. This is far superior to the traditional method which requires a farmer to examine the crop in person.

Figure 6 shows a detail of another dielectric plate assembly. Conducting tubes 40 and 50 are as before but now a single piece sapphire dielectric 73 is surrounded by dielectric fluid 100 On this case pentadecane). Gaps 91 and 92, containing the pentadecane, are engineered between the sapphire plate and the ends of the conducting tubes to a dimension of 0.3 mm. This mitigates a problem that can arise when high displacement current densities are required through the dielectric, such as for operation and transfer at high power. In such a case, an E-field of high peak rate of change, and therefore also of high peak value, is required to drive the displacement current, particularly through regions of low dielectric constant.

However, a potential problem arises given that, even when in contact, due to the microscopic irregularities of their surfaces, the interface between the conductor and the solid dielectric will have many voids. If not otherwise filled, these voids will be filled by air (or other atmosphere present), which will have a lower electrical breakdown strength and lower relative dielectric constant (at little more than 1) than the solid dielectric. Thus electrical breakdown across the voids will be a limiting factor in the density of the displacement current which can pass across the interface. Such electrical breakdown would result in significant and concentrated power dissipation at the interface and risk serious degradation of the dielectric material.

Filling the interfacial voids instead with a liquid dielectric such as for example pentadecane or polydimethylsiloxane increases the maximum displacement current that may cross the interface by virtue of both the higher dielectric strength of the dielectric liquid and its higher dielectric constant. It is then desirable to keep this liquid layer as thin as possible in order to minimise the power losses due to that layer as a liquid dielectric typically has a higher dielectric loss tangent than a low-loss solid dielectric. A thicker layer occupying an engineered space between the solid entities can be beneficial in order to adequately even out non-uniformities in the density of the displacement current that may otherwise arise with localised regions of higher current density in the neighbourhood of points of contact or points of very close proximity between high spots on the respective surfaces.

In principle, any number of dielectric portions can be included, but there is a trade-off between manufacturing cost and the strength and extent of the electric fields produced which peak at moments of current reversal in the high frequency oscillation. These electric fields result from the opposing electric potentials produced between adjacent interfaces of conductor and dielectric sections, as a result of current flow. Pairs of such opposing potentials constitute electric dipoles oscillating at the resonant frequency of the loop. These can be a source of radiative power loss and their fields can cause non-magnetic (i.e. electrostatic) coupling effects which may not be desirable. However, if there is a multiplicity of such N dipoles, preferably evenly arranged around the ring, then each would have a magnitude generally of 1/(N2) of that which would exist if only one dielectric section and one conductor section existed.

Further, the N dipoles would have different locations and different directions, creating a multipole, such that the resultant aggregate electric field would be much weaker and the aggregate radiated power also would be much weaker. Hence with two short dielectric portions, instead of a single short dielectric portion (assuming the same dielectric material, ring dimensions and resonant frequency), each dielectric portion would be half the length and the voltage developed between its ends would be half as much. Thus the dipole moment of each would be one quarter that which would exist in the case where there was only a single dielectric section. Further the two smaller dipoles are placed diametrically opposite each other and of opposite polarities, resulting in considerable mutual cancellation of their fields creating both a much weaker more localised aggregate electric field and weaker radiation. For the current embodiments, three dielectric portions (and consequently three conductive portions) is considered a good compromise between manufacturing complexity and the risk posed by E-fields across the dielectric portions. Readily-available simulation tools will allow the skilled reader to determine the optimum number of dielectric portions for a particular application.

There is a large range of dielectric materials that are suitable for the dielectric portions of the HRR and these can control the resonant frequency and size of the HRR. Solid and/or liquid dielectrics may be used. Where the HRR comprises more than one dielectric portion, the portions may differ in structure and composition.

Figure 7 shows how varying the thickness of a single-piece sapphire dielectric alters the frequency of one embodiment of the HRR with two dielectric portions. Engineering gaps (see 91 and 92 in Figure 6) of 0.3 mm are filled with pentadecane. A frequency range from below 32 MHz to above 42 MHz can be obtained by varying the thickness of the sapphire between 4 mm and 10 mm. This provides quite fine control of frequency and a thickness of 8.23 mm results in the desired ISM frequency of 40.68 MHz.

Figure 8 shows how varying the thickness of the single-piece sapphire dielectric alters the frequency of another embodiment of the HRR with two dielectric portions. This also has a fixed engineering gap of 0.3 mm filled with pentadecane or air. The frequency control is quite fine in this example at 2 kHz/pm, but a wide frequency range is again evident.

Figure 9 shows the effect of dynamic frequency control for the same embodiment as Figure 8, by sliding both of the dielectric plate assemblies (60, 70, Figure 1) from between the ends of the conducting tubes (40, 50, Figure 1). A displacement or more than 70 mm is shown as the dielectric plate assembly needs 88 mm in this example to clear the flange on the end of the tube. Dynamic frequency control may be important in certain applications where loading of the HRR by a receiver coil or HRR affects the resonance frequency. A feedback circuit may be arranged to control the location of the dielectric plate assemblies relative to the ends of the conducting tubes.

We have stated above that conventional inductive charging systems have deteriorating efficiency as frequency increases. The reason for this is the quality factor (or Q) of the existing conductive coils (particularly given that the coils are typically multi-turn and composed of litz wire, both of which measures become less effective as frequency increases), in conjunction with their tuning capacitors and necessary drive coupling arrangement, and increased lossiness in ferrite cores. Due to the ohmic resistance of electrical conductors such as copper, the unloaded Q of existing copper coils is significantly less than 103, at best generally around 200. Power transfer efficiency equals one minus loaded Q divided by unloaded Q: Efficiency = 1-o -loaded-unloaded Inspection of this equation shows that low values of 0 -unloaded result in poor efficiency.

A pure DRR constructed from the right materials may have values of -unloaded as high as 2 x 106 or higher but will be at least 2 x 106. The hybrid ring resonator described herein typically has a value of 0 -unloaded of greater than 2.5 x 103. Although there is some dielectric heating of the HRR due to high E-fields generated inside the dielectric, ohmic losses due to the skin effect of the conductor and some radiative losses for open cavities, the HRR is still far more efficient than copper coils for generating powerful B fields. Figure 10 shows a graph of efficiency against 0 -unloaded for Qloaded = 10. It can be seen that improving the n -unloaded factor results in a law of diminishing returns and there is little benefit to improving 0 -unloaded beyond about 10'.

However, the efficiency improvement over copper coils (Q unloaded around 200) provided by the HRR (Qunieaded of several thousand) is clear. Known techniques for improving (i.e. reducing) loaded Q include improving the coupling between the transmitter and receiver by moving them closer together, expanding the size of the receiver, and placing the HRR in a cavity.

Figure 11 is a graph which illustrates that setting the resonator deep within the cavity gives the best values of 0 -unloaded but sacrifices power transfer range. Figure 11, and related calculations, provide a basis for making design decisions in which these two properties are traded against one another.

Figure 12 shows a pair of hybrid ring resonators arranged to transfer electromagnetic energy from one to the other. A first HRR 130 is arranged as a transmitter and a second HRR 140 is arranged as a receiver, mounted on the underside 160 of a vehicle.

The first HRR 130 is located within an open-topped metallic cavity together with an excitation coil 120 and buried in the ground. When a vehicle carrying the second HRR 140 is located above the cavity, power is transferred across a distance of up to 4 metres into the second HRR 140. This induces a current in a pick-up coil or receiver coil 170 from which electricity is recovered. The location and design of the receiver coil relative to the HRR is subject to similar considerations as for the excitation coil and will be application-dependent.

In order to derive DC current for charging batteries, the output of the pick-up coil must be rectified. Conventional high speed diodes may not be readily available at the power ratings required. In this case it is possible to parallelise a number of such devices. Alternatively, an active rectifying circuit may be driven in synchronisation with the detected signal. Very fast switching semiconductor devices such as LDMOS or GaN FET arranged in known rectifier structures can be driven by high speed electronics in synchronisation with the signal in the pick-up coil.

Counterintuitively, at resonance, the currents in the excitation coil 120 and receiver coil 170 are negligible. The overall power loss of a HRR is less than 0.7%.

The arrangement of Figure 12 shows the transmitter buried in the ground. However, it could (with appropriate positioning of the receiver on the roof of the vehicle) be placed on an overhead gantry or even be mounted vertically and transfer power horizontally.

Figure 13 shows a system for transferring electrical power 1300 comprising a conductive cavity 1310 containing a resonator 1320 and an electrical coupler 1330 arranged in the same plane as the resonator. The electrical coupler, in this example, is a coil of conductive material into which electrical energy can be introduced. The electrical coupler is preferably made from copper for reasons of electrical conductivity and cost, although other materials such as silver will be apparent to the skilled reader.

The electrical coupler 1330 is preferably smaller than the resonator, having a diameter of 50% or less than the diameter of the resonator and more preferably less than 20% of the diameter of the resonator in order to limit losses in the electrical coupler. The coupler is preferably offset from the axis of the resonator. The coupler is preferably located in proximity to the cavity wall, closer to the cavity wall than the resonator 1320.

The cavity preferably has a diameter of no more than 200% of the diameter of the resonator and preferably no more than 150% of the diameter of the resonator. Indeed, good performance is achieved when the diameter of the cavity is only slightly greater than the diameter of the resonator, say 10% greater.

While a dielectric other than air has been disclosed, in a less-preferred embodiment (such as for a lower-power application) one or more of the dielectric portions may comprise air provided that they are short compared with the minor diameter of the HRR, otherwise the E-field and displacement current will not follow the curve of the HRR envelop.

Described embodiments of wireless power transfer very high frequency systems can project well in excess of 400 kW per module to a distance of around 800 mm because they use high frequencies with high unloaded Q factors. This compares to state-of-the-art copper coil systems where 50 kW and 300 mm are considered excellent. The efficiency of the HRR is also very high (greater that 99.3% for the wireless part). The high power transfer may equally be applied to cold ironing ships (i.e. shore-to-ship power) or recharging electric ferries whilst docked.

Another benefit over the prior art is that the disclosed system is more tolerant of mis-alignment of the transmitter and receiver resonators. Researchers at Warwick University have identified a requirement for lateral alignment within 500 mm which the disclosed system can tolerate, even without a large receiver-side resonator: "A potential solution would be to offer lateral alignment advice at distances greater than 50 cm." (How driver behaviour and parking alignment affects inductive charging systems for electric vehicles, Transportation Research Part C 58 (2015) 721-731).

Finally, it can be important to reduce interference generated by the apparatus disclosed herein, even when operating in the ISM bands. One measure that may need to be taken is to include a second ring arranged around the HRR which is driven out of phase. The second ring may be a conducting coil, a DRR or another HRR. This exploits the bipolar far field destructive interference effect and reduces radiative emissions to around 1% or less.

Claims (29)

1. CLAIMS1. A resonator for transferring electromagnetic energy in free space, comprising one or more conductive portions and one or more dielectric portions arranged alternately with the conductive portions in a continuous path, the resonator having a major dimension to minor dimension ratio of between 15:1 and 3:1, such that an alternating displacement current can be induced into the dielectric portions of the resonator and an alternating conduction current can be induced into the conductive portions of the resonator.
2. 2. A resonator as claimed in claim 1, wherein the major dimension to minor dimension ratio is between 12:1 and 4:1.
3. 3. A resonator as claimed in claim 1 or claim 2, the resonator comprising at least two conductive portions and at least two dielectric portions.
4. 4. A resonator as claimed in claim 1, claim 2 or claim 3, wherein the dielectric portions are generally in the form of plates positioned between the conductive portions.
5. 5. A resonator as claimed in any one of the preceding claims, wherein at least one of the one or more conductive portions comprise conductive tubes.
6. 6. A resonator as claimed in claim 5, wherein at least one conductive portion comprises a material having a conductivity of at least 3.5 x 107 Sim, preferably comprising copper tube.
7. 7. A resonator as claimed in claim 5 or claim 6, further comprising means for liquid cooling of the conductive tubes.
8. 8. A resonator as claimed in any one of the preceding claims, wherein at least one of the conductive portions is coated with graphene.
9. 9. A resonator as claimed in any one of the preceding claims, wherein the ends of the at least one conductive portion overlap substantially completely on either side of a dielectric portion.
10. 10. A resonator as claimed in any one of the preceding claims, wherein at least one of the dielectric portions comprises a dielectric having a relative permittivity of at least 5, preferably at least 7.5.
11. 11. A resonator according to any preceding claim wherein the dielectric portions have a breakdown strength of at least 6 kV/mm.
12. 12 A resonator according to any preceding claim wherein at least one of the dielectric portions is substantially formed of sapphire.
13. 13 A resonator as claimed in any one of the preceding claims, wherein the continuous path is substantially circular.
14. 14. A resonator as claimed in any one of the preceding claims, further comprising means to adjust a resonant frequency of the resonator by adjusting the relative location of at least one conducting portion and at least one insulating portion.
15. 15. A resonator as claimed in claim 14, wherein the means to adjust a resonant frequency of the resonator comprise means for moving at least part of at least one dielectric portion substantially perpendicular to a conductive portion.
16. 16 A resonator as claimed in any one of the preceding claims, which resonator is dimensioned to have a resonant frequency of between 27 and 300 MHz, substantially 40 MHz.
17. 17. A resonator as claimed in any one of the preceding claims, further comprising a conductive cavity arranged at least partially around the resonator.
18. 18. A resonator according to any preceding claim including a liquid dielectric, preferably pentadecane.
19. 19 A resonator according to Claim 18 wherein the liquid dielectric is provided in conjunction with a solid dielectric to fill the interface between the or each solid dielectric portion and the or each conducting portion.
20. A resonator according to any preceding claim wherein the size of the gap at the interface between the solid dielectric and the conducting portion comprises a thin film and/or wherein the size of the gap at the interface between the solid dielectric and the conducting portion is between 0.01 mm and 1 mm, preferably about 0.1 mm.
21. 21 A resonator according to any preceding claim arranged for transmitting a power of at least 50 kW in a frequency range of 27 to 90 MHz comprising a total insulating portion of 10% or less and preferably 5% or less of the path of the resonator.
22. 22 A resonator as claimed in any one of the preceding claims, further comprising an electrical coupler for exchanging electromagnetic energy with the resonator, the electrical coupler arranged in proximity to the resonator.
23. 23 A resonator as claimed in claim 22, wherein the electrical coupler has a diameter of one half or less than the diameter of the resonator, preferably having a diameter of one fifth or less than the diameter of the resonator.
24. 24 A resonator as claimed in claim 22 or claim 23, wherein the resonator and electrical coupler are arranged within a conductive cavity, the electrical coupler being arranged closer to a wall of the conductive cavity than the resonator.
25. 25. A resonator as claimed in claim 24, wherein the conductive cavity is arranged to have a diameter of less than 200% of the diameter of the resonator, preferably less than 150% of the diameter of the resonator and still further preferably less than 110% of the diameter of the resonator.
26. 26 A system for wireless transfer of power comprising a first resonator structure as claimed in claim 21, and a second resonator structure, the system further comprising means for introducing energy into the electrical coupler of the first resonator structure, and means for retrieving energy from the second resonator structure.
27. 27 A system as claimed in claim 26, wherein the second resonator structure also comprises a resonator as claimed in claim 17.
28. 28 A wireless power transmitter comprising a resonator structure as claimed in claim 21, further comprising a source of alternating current coupled to the electrical coupler.
29. 29. A wireless power transmitter as claimed in claim 26, further comprising sensing means for detecting changes in the quality factor of the electrical coupler.A wireless power receiver comprising a resonator structure as claimed in claim 21, further comprising rectifying means coupled to the electrical coupler to provide a DC power output.