Matrix transformer for an isolated high-frequency power converter

Abstract

The present invention relates to a transformer comprising an array of at least three sheets of magnetic material, each pair of adjacent magnetic sheets being separated by at least two pillars of magnetic material, and a primary and secondary winding each having one or more winding turns formed on at least one of said pillars of magnetic material. The total number of winding turns of the primary winding on said pillars between a first pair of adjacent magnetic sheets differs from the total number of winding turns of the primary winding on said pillars between a second pair of adjacent magnetic sheets, said second pair being different from the first pair, and / or the total number of winding turns of the secondary winding on said pillars between the first pair of adjacent magnetic sheets differs from the total number of winding turns of the secondary winding on said pillars between the second pair of adjacent magnetic sheets.

Description

Field of the invention

[0001] The present invention is generally related to the field of isolated DC / DC converters for electric vehicles.Background of the invention

[0002] Electric vehicles (EVs) are increasingly considered a promising alternative to those powered by internal combustion engines. Progress in the field of battery technologies has led to a significant increase in the battery size of EVs. However, to maintain convenient charging times and ensure availability, charging solutions have become the critical bottleneck for widespread EV adoption.

[0003] Given that context, onboard chargers (OBCs), which offer an AC charging solution to be built into EVs, have already gained considerable attention. The OBC charging hardware present in most EVs provides convenient charging through an AC grid connection at home, at work, in a public parking, etc. Hence, using the EV's power converter makes the required charging infrastructure minimal and cost-effective. Fig.1 illustrates an OBC in the electric powertrain of an EV. Typical OBCs comprise a two-stage power converter where the first stage is responsible for the rectification and power factor correction (PFC), ensuring that grid power quality standards are met. The second stage is a galvanic isolated DC-DC converter that provides the battery with the requested charging voltage and current. Offboard DC chargers bypass the OBC and connect directly to the HV battery, as depicted in Fig.1.

[0004] Bidirectional chargers grant EV owners access to the large battery capacity by expanding the functionality of an EV beyond transportation. The trend towards bidirectional chargers in the EV market poses new challenges on the OBC design. The automotive industry is notoriously strict when it comes to size, weight, and packaging requirements. Meanwhile, frequent charging and discharging of the EV relies on high-efficiency conversion to avoid excessive power losses. Furthermore, bidirectional converter topologies often imply the use of extra components or different component selections, for instance, favouring transistors over unidirectional diodes. Thus, the design of cost-competitive bidirectional chargers is paramount to ensure their adoption in the EV market. To meet the strict demands of governmental policies and the automotive industry, there is a clear need for OBCs to keep evolving towards higher power density, higher efficiency, lower cost, and better reliability.

[0005] As the need for DC-DC converters for, for example, OBCs and DC chargers grows, the bottleneck towards higher efficiencies and power densities lies in the magnetic components. Magnetic components are often the main contributor to the size, weight, and cost of the device they are part of. While advancements in wide-bandgap (WBG) semiconductor devices and soft-switching topologies push the trend towards lower switching losses and high-frequency operation, magnetic devices become more challenging to design. Moreover, combining multiple components in a single core and winding structure is known to be beneficial when targeting compact and efficient magnetics. Major power electronic applications commonly require isolated DC-DC converters for various safety and regulatory reasons. Examples can be found in EV chargers, data centre virtual regulator modules, consumer electronics devices, and in the automotive and telecom industry.

[0006] Integrated magnetic components describe various types of solutions where multiple magnetic structures with distinct functions are integrated into a single device. This ranges from combining various inductor types with transformers to merging multiple transformers into a single one. Magnetic integration often offers synergy with other power electronic design innovations, such as, for example, a) high-frequency WBG devices, as they enable high-frequency operation when increased power density is targeted, and b) planar magnetics with printed circuit board windings for their design flexibility.

[0007] High-frequency converter operation allows for the design of smaller passive components, but their downsizing is typically thermally limited by the increased magnetic losses that occur at high frequencies. Consequently, the heat from the losses in the windings and in the magnetic core must be extracted; otherwise, it would limit the converter's power. Litz and foil windings have traditionally been used to reduce winding losses. The thin cross-sectional areas alleviate skin effect losses, while twisted parallel strands in Litz wire reduce the proximity effect. However, as switching frequencies and power levels of chargers increase, the required strand diameter becomes challenging to manufacture, while the required high number of strands results in excessive proximity losses and a low copper fill factor.

[0008] Further, functionalities not conventionally performed by magnetic devices may be integrated into a magnetic device as well. For instance, magnetic devices can be realised in the converter's printed circuit board (PCB), electromagnetic interference (EMI)-filtering features can be performed by a transformer, etc. Developments in the field of printed circuit board technology have led to PCB windings and a low-profile core structure, reducing the size of magnetic components while offering improved reliability through automated manufacturing processes and good control of the parasitic components. WBG devices allow increasing the switching frequency of power converters to several hundreds kHz or even several MHz. The core size can so be shrunk, and the required number of turns can be reduced.

[0009] In the paper "High-Frequency PCB Winding Transformer with Integrated Inductors for a Bi-Directional Resonant Converter" (Bin Li et al., IEEE Trans. Power Electronics, vol.34, no.7, July 2019, pp.6123-6135) a PCB winding based magnetic structure is proposed to integrate the inductor and the transformer into one component. To reduce the number of layers required in the PCB the conventional EE-core structure of Fig.2 can be split up into two discrete UI-transformers, obtaining a split-transformer structure as indicated in Fig.3, where the primary and secondary windings of the first and the second transformer are connected in series. Combining these two series-connected transformers yields a solution with half the number of layers in the PCB. When merging these two cores, the same functionality can still be obtained. Having winding turns on adjacent pillars of the core being wound in opposite directions results in flux cancellation in the centre leg. Therefore, the centre leg can be omitted, and the transformer can be designed using a simple UI-core structure with windings on the external legs. The transformer proposed in the paper builds further on this and introduces the so-called matrix transformer concept, wherein AC winding losses are minimized through a perfect interleave structure that balances efficiency with practical manufacturing constraints.

[0010] The matrix transformer is extensively discussed in US10910140 B1. A matrix transformer is therein defined as a transformer structure divided into several ideally identical transformers, for example four, connected to function as a single transformer. However, the construction of transformers based on the matrix transformer concept cannot, as a practical matter, be made to be electrically or magnetically identical. US10910140 B1 presents a matrix transformer comprising first and second sheets of magnetic material, pillars of magnetic material extending between the first and second sheets of magnetic material in an array extending in perpendicular directions, and winding turns of a primary winding and a secondary winding on respective pillars wherein a winding direction of a respective primary winding is opposite a winding direction of a primary winding on an adjacent pillar in each of the perpendicular directions. The transformer core is usually formed in two or more pieces, so that the windings can be fabricated separately and then installed on a portion of the core. A pair of core elements, typically of complementary shapes, then form together a UI-shaped core. The term 'pillars' refers to the upstanding legs of the UI-shaped transformer core. For the remainder of this document, both terms "pillar" and "legs" will be used.

[0011] Interleaving between primary and secondary windings is a common practice to reduce the high-frequency AC winding loss and is widely used in planar transformers with PCB windings. Alternating winding turns of the primary and the secondary winding results in a limited magnetomotive force (MMF) surrounding the windings, which reduces the induced eddy currents. Thus, the proximity losses dominating the high-frequency AC resistance of the windings are reduced. However, alternating winding turns also lead to a very high coupling between the primary and secondary windings, resulting in very small and non-controllable leakage inductance. For most isolated DC-DC converters, like an LLC or CLLC resonant converter or a dual-active bridge (DAB) converter, one or more inductors are needed to be connected in series with the high-frequency transformer. For low-power applications (below 100 W), compact chip inductors may be used, but in higher power converters, sizeable additional inductors are needed. To reduce the number of magnetic components and reduce the size and losses of the converter, it is preferred to utilize the leakage inductance of the transformer to perform the functionality of this additional inductor. However, this requires a well-controlled leakage inductance, which is typically achieved with a discrete, non-integrated inductor. Ideally, the transformer and series inductors are combined in a single structure without the need for additional windings or magnetic materials, enabling a more compact design and / or lower losses. In the literature, for such integrated structures a volume reduction has been indicated of up to 60% and loss reductions of up to 40% compared to discrete transformers and inductors, as illustrated in Fig.4.

[0012] Fig.5 illustrates the drastic reduction of the magnetic field surrounding the wires achieved by interleaving, i.e., alternating the primary and secondary windings in the direction perpendicular to the plane of the PCB layers. Fig.5 depicts the magnetomotive force surrounding the windings, which causes eddy current losses. Non-interleaved structures reach an MMF equal to the product N p I p of the number of primary turns N p and the primary current I p . Fully interleaved structures reduce this by a factor N p by alternating turns of the primary and secondary windings. By doing so, the AC resistance and winding losses are drastically reduced, as indicated by Fig.6. Especially at high-frequency operation the difference escalates as interleaving drastically reduces proximity losses.

[0013] Analytically, the high-frequency winding losses can be modelled by means of 1D models such as Dowell's equation. It accommodates computationally efficient loss estimations for diverse winding geometries and interleaving techniques, including both skin effect and proximity effect losses. Furthermore, it provides flexibility in computing the losses for different winding geometries and interleaving methods. The eddy effects depend on the relationship between the thickness h of the PCB tracks and the skin depth δ at the operating frequency. This can be expressed as the ratio ξ = h / δ. Furthermore, the skin depth can be computed using the following equation : δ = ρ π f μ 0 where ρ is the resistivity of copper and µ 0 the permeability of vacuum. Based on this, the ratio between AC and DC resistance F r = R ac / R dc according to Dowell's model is provided by the expression F r = ξ 2 sinh ξ + sin ξ cosh ξ − cos ξ + 2 m − 1 2 sinh ξ − sin ξ cosh ξ + cos ξ Here, the first term represents the skin effect factor, and the second term is the proximity effect factor. Importantly, the parameter m represents the total magnetomotive force (MMF) at a certain turn, divided by its increase over this specific turn, as described by the expression m = mmf h mmf h − mmf 0 Thus, the higher the grade of interleaving, the lower the average m factor, and the lower the high-frequency losses. Naturally, an m factor of 1 reached by full-interleaving has been considered an effective measure of reducing high-frequency loss.

[0014] These models and their implementation have been extensively described in the literature. Fig.6(a) shows the significant impact of the m factor on the AC resistance. Thus, comparing the non-interleaved windings with full-interleaved windings, interleaving methods can drastically reduce the AC resistance, as depicted in the comparison of Fig.6(b).

[0015] In the paper "The Analysis and Comparison of Leakage Inductance in Different Winding Arrangements for Planar Transformer" (Z. Ouyang et al., 2009 Int'l Conf. on Power Electronics and Drive Systems (PEDS), pp.1143-1148) leakage inductance is analysed based on magnetomotive force and energy distribution. Various winding arrangements are investigated, which show the significant advantages an interleaving structure offers.

[0016] In the paper an interleaving method based on a minimized MMF interleaving structure is proposed to reduce the leakage inductance further. An m factor equal to 0.5 is achieved by splitting the first primary turn into two parallel layers positioned at the top and bottom, respectively, of the winding structure. In doing so, the MMF-profile still varies with a peak-to-peak variation of 1.I p , but it is centred around zero. I.e., the MMF only reaches 0.5.I p , i.e. half of the MMF at full-interleaving. This is illustrated in the right hand part of Fig.5. However, the minimised MMF interleaving structure has proven to be impractical to implement. For the interleaving structure to work, exactly 50% of the current needs to pass through the first and last layers, so perfect current sharing is required. Practically, this is impossible to achieve in a repeatable and reproducible manner using split-windings because of the termination, interference by magnetic fields, winding imperfections, geometric asymmetries, etc.

[0017] Magnetic loss comprise losses in the windings, as well as in the magnetic cores. Numerous models have been proposed to model core losses with varying degrees of complexity. For example, the improved Generalised Steinmetz Equation (iGSE) model can be adopted. It captures the dependency of core losses on the flux density and frequency while being compatible with arbitrary voltage waveforms. The iGSE equation is provided by the following equations : P v = f ∫ 0 t k i dB dT α Δ B β − α dt k i = k 2 π α − 1 ∫ 0 2 π cosθ α 2 β − α d θ Here, P v represents the volumetric core losses, f the frequency, and B the flux density. Furthermore, α, β, and k are the Steinmetz constants, which can be obtained by curve fitting P v (f) or P v (B) curves from experiments or from a datasheet using the original Steinmetz equation P v = k i f α< B β< . While offering a sufficiently accurate approximation, the iGSE model assumes a homogeneous flux distribution. This simplifies the field distribution.

[0018] Collectively, core and winding losses account for the magnetic losses, which commonly constitute 20 to 60% of total losses in modern high-frequency converters. Design optimisations are commonly used to optimise the magnetic components' parameters and balance all the complex trade-offs in the inductors and transformers. For example, lowering the operating frequency often results in reduced winding losses, albeit at the cost of increased core losses due to the increased flux density. Similarly, interleaving may drastically lower AC winding resistance but also reduces the leakage inductance to levels that prevent the integration of series inductors into the transformer design.

[0019] Hence, there is a need for a transformer wherein one or more of the above-mentioned limitations are addressed.Summary of the invention

[0020] It is an object of embodiments of the present invention to provide for a compact and efficient transformer for isolated power converters that offers an increased level of integration of magnetic components while allowing a high degree of interleaved windings to be employed.

[0021] The above objective is accomplished by the solution according to the present invention.

[0022] In a first aspect the invention relates to a transformer comprising an array of at least three sheets of magnetic material. Each pair of adjacent sheets of the at least three sheets of magnetic material is separated by at least two pillars of magnetic material and has a primary and secondary winding each with one or more winding turns formed on at least one of the at least two pillars of magnetic material. The total number of winding turns of the primary winding on the pillars between a first pair of adjacent magnetic sheets differs from the total number of winding turns of the primary winding on the at least two pillars between a second pair of the adjacent magnetic sheets, said second pair being different from the first pair, and / or the total number of winding turns of the secondary winding on the at least two pillars between the first pair of adjacent magnetic sheets differs from the total number of winding turns of the secondary winding on the at least two pillars between the second pair of adjacent magnetic sheets.

[0023] The proposed solution presents a three-dimensional matrix transformer (i.e. a TENSOR transformer) implementation that comes with various advantages. No external components are required, which allows for a more compact design. The at least three sheets of magnetic material define at least three different magnetic paths between two nodes. Due to the difference in number of winding turns between the first and the second pair of sheets, either in the primary winding or in the secondary winding or in both the primary winding and the secondary winding, a leakage flux is created which can be controlled by the number of winding turns and the amount of imbalance so obtained. The total magnetic loss and size can so be reduced.

[0024] It is an advantage of the transformer of this invention that the inductors are integrated into the design. More precisely, it is an advantage of the transformer that the integrated inductors are realized without the need for additional winding turns or a significant increase of the core volume.

[0025] In preferred embodiments the winding turns on the at least one pillar of magnetic material are formed in layers of a printed circuit board. In some embodiments multiple winding turns are formed in a single layer of the printed circuit board. Such printed circuit boards can be constructed from a variety of materials, including but not limited to FR4, aluminium, Rogers, polyimide, and other advanced composites or substrates.

[0026] Advantageously at least some of the winding turns in the layers of the printed circuit board are so arranged that winding turns of the primary and the secondary winding alternate, i.e. they are configured in an interleaved structure.

[0027] In preferred embodiments of the transformer the winding turns of the primary and secondary winding in the layers of the printed circuit board are configured in a shape resembling an '8', indicative of an infinity symbol.

[0028] In some embodiments the magnetic sheets of magnetic material are integrated in the printed circuit board.

[0029] In a particular embodiment the printed circuit board is flexible. The printed circuit board may then be so folded that it comprises two substantially parallel end portions interconnected by means of flexible interconnectors establishing a bent connection between the two end portions.

[0030] In a preferred embodiment the winding turns of the primary winding and the secondary winding of the first pair are formed in layers of a first printed circuit board and the winding turns of the primary and secondary winding of the second pair are formed in layers of a second printed circuit board.

[0031] Preferably, the sheets of magnetic material are positioned substantially in parallel. The pillars then preferably extend in substantially perpendicular direction from the sheets of magnetic material.

[0032] In preferred embodiments the transformer has a core with at least one U-shaped part and an I-shaped part.

[0033] Typically, the first and second secondary windings are provided with output ports.

[0034] In some embodiments the winding turns of at least one of the first and second primary or secondary windings are in parallel.

[0035] In further embodiments the winding turns of at least one of said first and second pair of adjacent sheets are divided into a plurality of subsets formed on each of the at least two pillars so that on each pillar a stack is obtained of said subsets, wherein subsets of winding turns of the primary and of the secondary winding alternate between consecutive layers of each stack. A first and / or a last layer of said stacks accommodate a current path split into parallel paths, one per pillar, with each of said parallel current paths carrying a fraction of a total current, while winding turns in layers between said first and last layer carry double said fraction of the total current.

[0036] The transformer is in preferred embodiments arranged to be excited by a multi-phase input voltage.

[0037] In another aspect the invention relates to a DC-DC converter comprising a transformer as described previously.

[0038] In another aspect the invention relates to a transformer comprising two sheets of magnetic material separated by at least two pillars made of magnetic material and having a primary and secondary winding each with a plurality of winding turns. The transformer is characterized in that the winding turns are divided into a plurality of subsets formed on each of said at least two pillars so that on each pillar a stack is obtained of said subsets, wherein subsets of winding turns of said primary and of said secondary winding alternate between consecutive layers of each stack. A first and / or a last layer of said stacks accommodate a current path split into parallel paths, one per pillar, with each of said parallel current paths carrying a fraction of a total current, while winding turns in layers between said first and last layer carry double said fraction of the total current.

[0039] The specific construction of a transformer according to this aspect of the invention allows achieving a minimized MMF profile, with a maximum amplitude equal to a fraction of the total current. Subsets of winding turns of the primary and secondary winding are created on each of the pillars. These subsets of winding turns form on each pillar a stack of winding turns, in which layers of primary and secondary winding turns are used alternatingly. A distinction is made between the first and / or the last layer of the stack on the one hand and the intermediate layers in between the first layer and the last layer on the other hand. At least one of the first and the last layer have a current path split into two or more parallel connected paths, whereby there is one path per pillar. Each of the split current paths carries a fraction of the total current. In the aforementioned intermediate layers, the current path forms a loop around multiple pillars, carrying twice the current around each of the pillars compared to the first and / or last layers.

[0040] In some embodiments the transformer comprises more than two sheets of magnetic material, whereby at least two pillars are provided between each pair of adjacent sheets

[0041] In preferred embodiments the sheets of magnetic material are positioned substantially in parallel. The pillars preferably extend in substantially orthogonal direction from the sheets.

[0042] In a preferred embodiment the winding turns of the primary and secondary winding are formed in layers of a printed circuit board. Advantageously, one winding turn per PCB layer is formed around each of the pillars in the respective layer.

[0043] Preferably the various parallel paths each carry an equal fraction of the total current. In this way each path has a same impedance.

[0044] In advantageous embodiments the parallel paths are symmetrical. A configuration with symmetrical parallel paths is beneficial for ensuring an equal impedance.

[0045] For purposes of summarizing the invention and the advantages achieved over the prior art, certain objects and advantages of the invention have been described herein above. Of course, it is to be understood that not necessarily all such objects or advantages may be achieved in accordance with any particular embodiment of the invention. Thus, for example, those skilled in the art will recognize that the invention may be embodied or carried out in a manner that achieves or optimizes one advantage or group of advantages as taught herein without necessarily achieving other objects or advantages as may be taught or suggested herein.

[0046] The above and other aspects of the invention will be apparent from and elucidated with reference to the embodiment(s) described hereinafter.Brief description of the drawings

[0047] The invention will now be described further, by way of example, with reference to the accompanying drawings, wherein like reference numerals refer to like elements in the various figures. Fig.1 illustrates an onboard charger in an electric powertrain with its key subsystems. Fig.2 illustrates a transformer with a conventional EE-core and PCB-windings structure as known in the prior art. Fig.3 illustrates a transformer with a split-transformer structure as known in the art, wherein magnetic flux in the centre leg is cancelled. Fig.4 illustrates the improvements in volume and loss decrease achieved by implementing an integrated magnetic structure, as reported in literature, whereby each datapoint represents a reported result. Fig.5 illustrates the stack configuration and related MMF-profiles for respectively non-interleaving, full-interleaving, 0.5-MMF interleaving as in the prior art. Fig.6 illustrates the beneficial effect of interleaving on the AC resistance for various mmf ratios (a) and interleaving structures (b). Fig.7 illustrates schematically an embodiment of a transformer of this invention comprising three sheets of magnetic material, each pair connected by two pillars with primary (P) and secondary (S) windings formed on each pillar. Fig.8 illustrates a transformer with balanced windings as known in the art, resulting in leakage flux cancellation through the middle sheet, so bringing the leakage flux close to zero. Fig.9 illustrates an embodiment of a transformer according to the present invention wherein the condition of imbalanced windings is met, and thus, an integrated leakage inductance is created as a result of the net flux through the middle sheet. Fig.10 illustrates another embodiment of a transformer according to the present invention wherein the condition of imbalanced windings is met, albeit with turns formed on only some of the magnetic pillars in said structure. Fig.11 illustrates an embodiment of a transformer from this invention, in a 3D view as well as in a schematic cross-sectional view. Fig.12A illustrates an embodiment of a transformer according to this invention whereby three pillars connect the different sheets of magnetic material, with winding turns formed around each of the pillars. Fig.12B illustrates the flux induced by the bottom center pillar, formed in parallel flux paths through the two sheets and each of the pillars. Fig.12C illustrates an embodiment of a transformer of this invention comprising four sheets of magnetic material, each pair connected by four magnetic pillars, with winding turns formed around each of the pillars formed by three PCB boards. Fig.13 illustrates two windings, primary and secondary, wherein the figure-8 structure is realized while a single PCB layer features multiple turns around each pillar. Fig.14 illustrates a possible PCB-winding layout of a split-turn and figure-8 turn. Fig.15 illustrates a possible winding structure for a 7:7 transformer according to this invention, with the proposed integrated transformer structure in two PCB boards as well as the interleaving structure that achieves a minimized 0.5-MMF profile with maximum amplitude of half the primary current, featuring split-windings in the first and last layer, and figure-8 windings in the intermediate layers. Fig.16 illustrates a 3-phase CLLC resonant converter, having three separate transformers with respective resonant inductors, which can be replaced by a single 3-phase matrix transformer with integrated inductors. Fig.17 illustrates the T-model of a transformer with leakage inductance L k and magnetizing inductance L m , as well as its parasitic components. Fig.18 illustrates a model of a transformer as described employing a primary and secondary winding each divided into two subwindings (on the left side) and a reluctance model of the transformer of the invention (on the right side). Fig.19 illustrates a 3D finite elements simulation wherein balanced winding are used, proving the occurrence of flux cancellation in the middle sheet of magnetic material. Fig.20 illustrates a 3D finite elements simulations wherein imbalanced windings are used, as described in this invention, indicating that some flux is present in the centre leg, which is establishes a leakage inductance. Fig.21 illustrates a hardware implementation of an embodiment of the transformer of the invention. Fig.22 illustrates a winding structure featuring a minimized MMF interleaving solution as presented in this invention. Fig.23 illustrates a PCB winding layout for a split-winding (first and / or last layer of a stack). Fig.24 illustrates a winding structure for a transformer with three magnetic pillars, featuring a minimised-MMF interleaving solution as presented in this invention, whereby each first and last layer are excited by an MMF equal to ⅓ the current, and intermediate layers 2 3 the current. Fig.25 illustrates a winding structure for a transformer in which the different stacks are excited by a different MMF, while each stack still features the minimised-MMF interleaving solution as presented in this invention. Detailed description of illustrative embodiments

[0048] The present invention will be described with respect to particular embodiments and with reference to certain drawings but the invention is not limited thereto but only by the claims.

[0049] Furthermore, the terms first, second and the like in the description and in the claims, are used for distinguishing between similar elements and not necessarily for describing a sequence, either temporally, spatially, in ranking or in any other manner. It is to be understood that the terms so used are interchangeable under appropriate circumstances and that the embodiments of the invention described herein are capable of operation in other sequences than described or illustrated herein.

[0050] It is to be noticed that the term "comprising", used in the claims, should not be interpreted as being restricted to the means listed thereafter; it does not exclude other elements or steps. It is thus to be interpreted as specifying the presence of the stated features, integers, steps or components as referred to, but does not preclude the presence or addition of one or more other features, integers, steps or components, or groups thereof. Thus, the scope of the expression "a device comprising means A and B" should not be limited to devices consisting only of components A and B. It means that with respect to the present invention, the only relevant components of the device are A and B.

[0051] Reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment, but may. Furthermore, the particular features, structures or characteristics may be combined in any suitable manner, as would be apparent to one of ordinary skill in the art from this disclosure, in one or more embodiments.

[0052] Similarly it should be appreciated that in the description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. This method of disclosure, however, is not to be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.

[0053] Furthermore, while some embodiments described herein include some but not other features included in other embodiments, combinations of features of different embodiments are meant to be within the scope of the invention, and form different embodiments, as would be understood by those in the art. For example, in the following claims, any of the claimed embodiments can be used in any combination.

[0054] It should be noted that the use of particular terminology when describing certain features or aspects of the invention should not be taken to imply that the terminology is being redefined herein to be restricted to include any specific characteristics of the features or aspects of the invention with which that terminology is associated.

[0055] In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the invention may be practiced without these specific details. In other instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.

[0056] In a first aspect the present invention discloses a transformer wherein there are between two nodes at least three magnetic paths created by means of at least three sheets of magnetic material, whereby each pair of adjacent magnetic sheets is separated by two or more pillars of magnetic material. A primary and a secondary winding each have one or more winding turns formed around one or more of the pillars between a pair of magnetic sheets. In some embodiments the total number of winding turns of the primary winding on the at least one pillar between a first pair of adjacent magnetic sheets is different from the total number of winding turns of the primary winding on the at least one pillar between a second pair of adjacent magnetic sheets. The second pair of magnetic sheets is different from the first pair, although the two pairs may have one magnetic sheet in common. In other embodiments the total number of winding turns of the secondary winding on the at least one pillar between the first pair of adjacent magnetic sheets is different from the total number of winding turns of the secondary winding on the at least one pillar between the second pair of adjacent magnetic sheets. It is repeated that the second pair of magnetic sheets is different from the first pair. In yet other embodiments the two previously described options are combined, that is to say, both the total number of winding turns of the primary winding and of the secondary winding between the first pair of adjacent magnetic sheets is different from the total number of winding turns of the primary winding and of the secondary winding on the at least one pillar between the second pair of adjacent magnetic sheets.

[0057] Fig.7 provides an illustration of an embodiment of the 3D-transformer with exactly three magnetic sheets, wherein the top magnetic sheet (11) and the centre magnetic sheet (12) form a first pair and the bottom magnetic sheet (13) forms together with the centre magnetic sheet (12) the second pair. The magnetic sheets in the embodiment of Fig.7 are positioned substantially in parallel. In this particular example, for each pair two pillars of magnetic material are provided between the adjacent magnetic sheets. In other embodiments there may be more than two pillars. In Fig.7 the pillars extend in a substantially perpendicular direction from the magnetic sheets. In this way three magnetic flux paths are created between the two nodes (A,B), that is, one along each of the magnetic sheets. The primary winding and the secondary winding for the first pair and the second pair, respectively, are distributed among sub-windings P1, P2, S1, and S2, each additionally divided into a left (L) and right (R) side part, on the first and second flux paths. The respective sub-windings are connected in series. The winding direction of the respective sub-windings is such that the flux through the third path (i.e., the path along the centre magnetic sheet) flows in opposite directions depending on whether also the first path or the second path are used to form a complete loop. If the turns of the primary and secondary windings are distributed equally among their respective sub-windings (i.e., P1L+P1R=P2L+P2R and S1L+S1R=S2L+S2R), the magnetic flux induced in the first path and flowing through the third path is completely cancelled. In Fig.8 such a prior art solution is shown. However, upon any imbalanced distribution of turns (where P1L+P1R≠P2L+P2R and S1L+S1R≠S2L+S2R), as shown in Fig.9, which is the case in the present invention, a leakage flux is obtained. This flux can be tuned by the number of turns, the imbalance, and the reluctance in each flux path.

[0058] Another embodiment depicted in Fig.10 similarly comprises three magnetic sheets with two pillars between each pair of adjacent magnetic sheets, whereby primary (P) and secondary (S) winding turns are formed around the left-side (L) and right-side (R) pillars. In this specific embodiment, one of the pillars does not contain any turns of a specific winding, namely the top right pillar is devoid of secondary winding turns. Herein, the same condition with respect to the number of winding turns as in Fig.7 applies, as any imbalance of primary and / or secondary turns among the respective sub-windings (P1L+P1R≠P2L+P2R and S1L+S1R≠S2L+S2R) causes a leakage flux to be established in the middle sheet. Thus, one such alternative embodiment is depicted in Fig.10.

[0059] Fig.11 illustrates another embodiment of the transformer in 3D view as well as in a schematic cross-sectional view. In the top PCB (PCB 1), a total of 8 primary (P) and 6 secondary turns is applied, distributed among the left- and right-side pillars. On the bottom PCB (PCB 2), 6 primary and 8 secondary turns are formed. This imbalance results in a net-leakage flux in the middle leg, whereby the series-inductors are integrated in the transformer due to its increased leakage inductance.

[0060] In embodiments with three sheets (as, e.g., in Figs.7, 9, 10 and 11), the magnetic core may comprise two U-shaped parts with a I-shaped part in between. As already mentioned, the sub-windings P1, S1, P2, S2 are split into a left and right part. Each part is wound around a pillar of magnetic material formed by a leg of the U-shape. The sheets of magnetic material constitute the flat parts of the U-shapes and the I-shape, respectively.

[0061] In some embodiments the winding turns of the primary and / or secondary winding on at least one pillar of magnetic material between two magnetic sheets are formed in layers of a printed circuit board (PCB). In Fig.9 and Fig.11 an embodiment of a transformer of this invention is depicted comprising two PCBs. As explained above, an imbalance is created in the distribution of the winding turns of the primary and secondary windings on the top versus bottom part of the transformer, whereby a leakage flux is induced in the central leg. The primary and secondary windings in each PCB are split into a left and right part. In the top part of the transformer (i.e., the upper PCB) there is a primary winding with one winding turn, and a secondary winding has three winding turns. On the contrary, in the bottom part (the lower PCB) the primary winding has three winding turns and the secondary winding only one winding turn.

[0062] The magnetic core in the example embodiments of Fig.9 and Fig.11 comprises two U-shaped parts with a I-shaped part in between. The U-shaped parts comprise a sheet of magnetic material and on each side a pillar of magnetic material. Each part is wound around the pillar formed by a leg of the U-shape. The pillars so separate the magnetic paths, which each comprise a magnetic sheet. The sheets of magnetic material then constitute in these embodiments the flat parts of the U-shapes and the I-shape, respectively.

[0063] Although Figs.7, 9, 10 and 11 depict embodiments with three magnetic sheets defining three magnetic paths, the solution according to the invention is definitely not limited thereto : in other embodiments there may be more than three magnetic paths created by more than three magnetic sheets and corresponding pillars in between. An embodiment wherein more than two pillars are used to connect two adjacent sheets of magnetic material, is depicted in Fig.12A. In Fig.12A a core structure with three sheets of magnetic material is constructed, with three substantially perpendicular pillars extending between each pair of adjacent sheets. The transformer in Fig.12A comprises two PCB boards with winding turns of the primary and secondary windings wound around each of the core pillars.

[0064] In one possible embodiment three-phase primary and secondary ports are implemented. Each phase of the primary and secondary windings is split among the top and bottom pillars. The flux induced by the middle pillar of the bottom pair of sheets is shown in Fig.12B. As indicated, the flux is split into several parallel paths through the other pillars, before returning through the bottom sheet. The winding direction is such that the flux through the centre leg originating from the top windings is opposite the flux direction from the bottom windings. Similarly, by introducing an imbalance between the number of turns primary and / or secondary turns formed around the top pillars compared to the bottom pillars, a net leakage flux is established. The cross-sectional area of the core, the number of turns, number of imbalanced turns, and the length of the flux path can be designed to achieve the desired leakage inductances.

[0065] In another embodiment, depicted in Fig.12C, the core structure comprises four sheets of magnetic material, each pair connected by four pillars, with PCB windings formed around each of the pillars. Herein, the primary and secondary windings can thus be distributed around twelve pillars. This yields a flux distribution in the magnetic cores dictated by the distribution of the windings. An imbalanced distribution of the primary and secondary windings, results in a decoupling between said windings due to leakage flux in the magnetic sheets.

[0066] In some embodiments multiple winding turns are formed in a single layer of the printed circuit board. Fig.13 illustrates a top-view of the winding in an embodiment in which multiple turns are applied per layer, while maintaining a figure-8 winding structure. This indicates the ability to deploy an arbitrary number of winding turns per layer.

[0067] Advantageously some interleaving is applied, whereby for at least some of the winding turns in the layers of the PCB alternatingly winding turns of the primary winding and winding turns of the secondary winding are used.

[0068] In advantageous embodiments the winding turns of the primary and secondary winding in the layers of at least one printed circuit board have an infinity shape, alike the number '8', as illustrated in Fig.14 and Fig.15 (top PCB layers 2-4 and bottom PCB layers 1-3).

[0069] In some embodiments the magnetic sheets of magnetic material are integrated in the printed circuit board.

[0070] In advantageous embodiments the printed circuit board is flexible, made of materials such as PI (polyimide) film, PET (polyester) film, PEN (polyethylene naphthalate) film, or PTFE (polytetrafluoroethylene) film. In such case the printed circuit board can be foldable so that the board can be seen as comprising two substantially parallel end portions. In between flexible interconnectors can be used to make a bent connection between the two end portions.

[0071] Multi-phase transformers are popular in high-power converters, wherein a number of phase-shifted input voltages are applied to separate windings. As such, the 3-phase CLLC resonant converter depicted in Fig.16 can employ a single matrix transformer, rather than requiring three discreet transformers, each with two series inductors. Hereby, 3-phase transformers with three primary and three secondary windings are used to transmit 3-phase AC power. Similarly, popular converter topologies such as multi-phase DABs or resonant converters use pulse width modulation techniques to provide such multi-phase high-frequency signals to a transformer. Some embodiments of the transformer according to the invention, therefore, accommodate multiple primary and secondary windings to transmit these multi-phase signals with integrated leakage inductances.

[0072] The benefits the transformer of the invention brings over prior art solutions are manifold. Firstly, magnetic scaling laws dictate that larger, combined magnetic components can achieve a higher power density and efficiency compared to a series of small discrete magnetics, underlining the potential of integrated transformers. By integrating the series inductors into the transformer structure by means of a controlled leakage flux, three components are compressed into a single device, reducing the total size, loss, cost, manufacturing requirements, and carbon footprint. Moreover, the winding configuration of the transformer in the invention is such that in specific parts of the magnetic core, a certain amount of flux cancellation occurs. This technique reduces the total flux density in areas where it is undesired, leading to significantly lower core losses. Furthermore, the integration of the series inductors is realised without the need for additional windings, as the leakage is created by an imbalanced distribution of the existing transformer turns. Thus, it offers a distinctive solution to integrated magnetics, wherein typically additional windings and magnetic material are needed. This saves copper, reduces the components' size, and the winding losses. Overall, these benefits lead to a magnetic device and DC-DC converter that can achieve ultra-high efficiency and power density.

[0073] As already mentioned, the required number of layers per PCB can be reduced by a factor two compared to the split-core structure presented in the background section or even a factor four compared to the conventional EE-core as in Fig.2. Using PCBs with the same number of layers, two times as many turns can be applied compared to the split-core configuration. Thus, for example, an 8:8 transformer can be designed using only 4-layer PCBs or even a 12:12 transformer using 6-layer PCBs.

[0074] Alternatively, the many available layers can be employed to provide parallel turns. This is often used in high-output current applications such as data centre VRMs. For example, a 400 / 48 V supply of 3 kW has an output current of 62.5 A. Thus, it could be designed using an 8:1 transformer with 4 parallel secondary windings, needing only 4-layer PCBs.

[0075] The proposed structure is advantageous in that better interleaving can be obtained due to the increased number of degrees of freedom for the interleaving.

[0076] Parasitic capacitance is reduced by having the windings split in at least four parts and geometrically spaced out from each other. Less interfaces between the windings result in a reduced parasitic capacitance.

[0077] Another benefit relates to a better heat dissipation, especially through (forced) air cooling of the PCB windings with few turns.

[0078] A low number of layers in the PCB boards further results in a cheaper manufacturing.

[0079] Since each flux path has two air gaps, their length is halved, which reduces fringing losses, which play a crucial role in high-frequency AC winding losses.

[0080] Fringing losses are reduced further since UI-core structures are employed. Having the air gaps at the end of a winding rather than in the centre reduces fringing loss and even allows for the windings to be spaced away from the gap if these losses are critical.

[0081] Two connectors are required between the top and bottom PCB boards, and the height is increased compared to the split-transformer, while the number of layers per PCB is reduced.

[0082] Now a more mathematical explanation is provided based on a model of the integrated magnetic component.

[0083] A configuration is thereby assumed with three sheets of magnetic material, with at least two pillars between each pair of adjacent sheets. The transformer comprises a first primary winding P1 and a first secondary winding S1 between the first pair of sheets and a second primary winding P2 and second secondary winding S2 between the second pair of sheets. The number of winding turns can vary between these various windings. A leakage inductance is obtained by unequal distribution of the turns of the primary windings (P1, P2) and / or secondary windings (S1, S2) as set out above. The voltage v over each winding can be expressed by the inductance matrix multiplied by the rate of change in current through the respective winding. Hereby, the inductance matrix contains the self-inductances (L ii ) and mutual inductances (L ij ) of each sub-winding P1, P2, S1, S2. This is expressed as v p 1 v p 2 v s 1 v s 2 = L p 1 p 1 L p 1 p 2 L p 1 s 1 L p 1 s 2 L p 1 p 2 L p 2 p 2 L p 2 s 1 L p 2 s 2 L p 1 s 1 L p 2 s 1 L s 1 s 1 L s 1 s 2 L p 1 s 2 L p 2 s 2 L s 1 s 2 L s 2 s 2 ⋅ di p 1 dt di p 2 dt di s 1 dt di s 2 dt Since the primary and secondary sub-windings are connected in series, i p = i p1 = i p2 and i s = i s1 = i s2 , whereby this can be expressed as: v p 1 v p 2 v s 1 v s 2 = L p 1 p 1 + L p 1 p 2 L p 1 s 1 + L p 1 s 2 L p 1 p 2 + L p 2 p 2 L p 2 s 1 + L p 2 s 2 L p 1 s 1 + L p 2 s 1 L s 1 s 1 + L s 1 s 2 L p 1 s 2 + L p 2 s 2 L s 1 s 2 + L s 2 s 2 ⋅ di p dt di s dt

[0084] Since the primary and secondary voltages are distributed among the windings, the expressions v p = v p1 + v p2 and v s = v s1 + v s2 further simplify the matrix equations to: v p v s = L pp ′ M ′ M ′ L ss ′ ⋅ di p dt di s dt where L pp ′ = L p 1 p 1 + 2 L p 1 p 2 + L p 2 p 2 L ss ′ = L s 1 s 1 + 2 L s 1 s 2 + L s 2 s 2 M ′ = L p 1 s 1 + L p 1 s 2 + L p 2 s 1 + L p 2 s 2

[0085] Thus, the leakage inductance and magnetizing inductance can be determined from this derived equivalent inductance matrix, representing the complete transformer in its T-model equivalent as in Fig.17. L kp = L pp ′ − M ′ N p 1 + N p 2 N s 1 + N s 2 L ks = L ss ′ − M ′ N s 1 + N s 2 N p 1 + N p 2 L m = M ′ N p 1 + N p 2 N s 1 + N s 2 This provides a general expression that comprehensively characterises the transformer's behaviour. To implement it, the values for the self- and mutual inductances can be derived from the schematic visualisation of the transformer in the left hand side part of Fig.18. Therein three sheets of magnetic material are connected by two pillars between each set of adjacent pillars. The total number of primary turns formed around the first set of pillars are denoted by P1, the total number of respective secondary turns is S1. Similarly, all primary and secondary turns on the second set of pillars are denoted as P2 and S2, respectively. The total airgap reluctance between two sheets is combined in R g1 and R g2 . As depicted, the U-I-U cores are denoted as A, B, and C, respectively. The flux can follow three paths: AB, AC, and / or BC. Note that windings are only applied to flux paths A and C, while path B is the leakage path. The total of primary and secondary turns around flux path A is denoted as p1 and s1, respectively, while the turns on path C are denoted with the suffix 2. Remember that the condition for leakage flux to be generated in the B path is that N p1 ≠ N p2 and / or N s1 ≠ N s2 . Furthermore, the total air gaps of the top and bottom transformers are R g1 and R g2 , respectively. The self-inductances can be estimated by the reluctance model in the right-hand side of Fig.18. Depending on the length of the airgap and the desired accuracy, the model can be simplified by only considering the airgap reluctance due to the high permeability of the air gaps.

[0086] The proposed core structure is validated in 3D finite elements analysis (FEA) simulations in Ansys Maxwell, using both Eddy Current and Transient solvers with sinusoidal and square wave voltage excitations. Fig.19 shows the resulting maximal flux density distribution when balanced windings are used in the top and bottom PCB boards. As such, one can see that indeed flux cancellation occurs in the middle sheet of magnetic material, as the flux density is approximately zero. This is confirmed by the extremely small simulated leakage inductance of this configuration, being equal to 0.02 µH.

[0087] Furthermore, Fig.20 depicts the flux density distribution with imbalanced windings, where the top PCB contains 5 primary and 3 secondary turns, while the bottom PCB contains 3 primary and 5 secondary turns (5:3+3:5). Here, the flux distribution varies at different stages of the period, as depicted by the distribution at 45° on the left-hand side and at 135° on the right-hand of Fig.20. On these occasions, the centre path conducts the respective windings' maximal primary and secondary leakage flux. According to the FEA simulation, these leakage inductances are equal to 12 µH, which validates the mathematical models. Hence, it achieves an increase of factor 600 compared to the imbalanced windings. Hereby, the principal operation is confirmed.

[0088] A hardware prototype of an embodiment of the patent is presented in Fig.21, according to the parameters previously described in Fig.11, using two 64 / 50 / 10 planar U-cores and a 65 / 50 / 5 planer I-core of 3F36 material.

[0089] In another aspect the present invention discloses an interleaving method based on an alternative minimized MMF interleaving structure in order to reduce the winding losses due to high-frequency eddy currents. The maximum MMF that occurs is reduced to a fraction of the current by splitting the current path for the current applied to the first and / or the last primary turn into two or more parallel paths with each of the parallel current paths carrying a fraction of a total current, while winding turns in layers between the first and last layer carry double said fraction of the total current.

[0090] In its most general form a transformer is considered comprising two or more sheets of magnetic material separated by at least two pillars made of magnetic material. A magnetic path is so created between a first node A at one end of the sheets and a second node B at an opposite end. A primary winding and a secondary winding each have a plurality of winding turns on the pillars. The winding turns of the primary and the secondary winding on each of the pillars are divided into two or more subsets. The subsets of winding turns of the primary and of the secondary winding alternate between consecutive layers of each stack. A first and / or a last layer of said stacks accommodates a current path split into parallel paths, one per pillar, with each of the parallel current paths carrying a fraction of a total current, while winding turns in layers between the first and last layer carry double the fraction of the total current.

[0091] An example is provided in Fig.11, wherein the first and last layers are split into parallel-connected current paths. The winding configuration with minimised MMF is specified further in Fig.22, which shows the cross-section of a 7:7 transformer as in Fig.11, being divided into two stacks. The winding direction is opposite in the two subsets: current for each respective winding flows in winding turns of one subset in clockwise direction and in winding turns of the other one in counterclockwise direction. This ensures that the flux in the outer sheets of magnetic material, coupling the different subsets of primary and secondary windings, flows in equal direction. Distinction is made between the first and the last layer of the stack on the one hand and the layers in between on the other hand. The first and / or last layers are shared by both stacks, splitting the current into two parallel paths within said layer, each directing 50% of the current around one of the pillars. Fig.23 illustrates a winding layout to achieve this design. The layers in between consist of alternating primary (P) and secondary (S) windings, configured in a figure-8 pattern (as already shown in Fig.14) to ensure that the full current flows around each pillar.

[0092] Fig.22 and Fig.24 show embodiments wherein each parallel path receives an equal portion of the total current fed to the transformer. In Fig.22, the transformer features two pillars. The winding turns around each of the pillars carry half of the total current. In contrast, the transformer in Fig.24 features three pillars, each receiving one third of the current. Furthermore, both implementations comprise intermediate layers in which the induced MMF is double that of the first and last layer. Such fractional MMF-excitations as in Fig.24 can be practically achieved by applying multiple turns per layer and / or implementing parallel paths. Thereby, it achieves the required MMFs and minimises the MMF profile as described in this disclosure.

[0093] In other embodiments, however, there may be a difference between the amounts of current in the various paths, as long as the MMF induced in subsequent intermediate layers of the same stack is double the MMF of the first and / or last layer. This is exemplified in the winding configuration of the embodiment in Fig.25. This figure shows an advanced winding structure split into three stacks, wherein the first and last layer of stack 1 receives half of the total current, while stacks 2 and 3 receive a quarter of the total current. Still, in each respective subsequent layer, twice the current of the first layer is formed around each pillar. I.e., while the total current flows in stack 1, this is split into parallel paths forming turns in stack 2 and 3. As a result, the MMF-profile for each stack is again minimized.

[0094] In some embodiments symmetrical parallel paths are provided to achieve an equal impedance. This ensures that the current distribution is as intended by the design. An example is shown in the current path of Fig.23, which is implemented in the transformer of Fig.21.

[0095] The minimized MMF concept as set out above can also be applied to a transformer as described earlier in this disclosure, wherein the total number of winding turns of the primary winding on the at least two pillars between a first pair of adjacent sheets of magnetic material differs from the total number of winding turns of the primary winding on the at least two pillars between a second pair of adjacent sheets, and / or the total number of winding turns of the secondary winding on the pillars between the first pair of adjacent sheets differs from the total number of winding turns of the secondary winding on the pillars between the second pair of adjacent sheets.

[0096] In such embodiments of the transformer the winding turns of at least one of the first and second pair of adjacent sheets are divided into at least two subsets formed on each of the pillars so that on each pillar a stack of subsets is obtained. Subsets of winding turns of the primary and of the secondary winding alternate between consecutive layers of each stack. A first and / or a last layer of the stacks accommodates a current path split into parallel paths, one per pillar, with each of the parallel current paths carrying a fraction of a total current, while winding turns in layers between said first and last layer carry double said fraction of the total current. In this way the AC resistance of the windings can be reduced drastically, while still featuring sufficient leakage inductance to integrate the series inductors.

[0097] Thus, in this embodiment the advantages of improved integration are combined with the advantages offered by the specific way of realising the interleaving. The interleaved windings with minimised MMF ensure that very low winding losses are obtained, resulting in an efficient and compact transformer design. While in conventional transformers, this high level of interleaving results in a low leakage inductance, the proposed transformer structure of this invention also allows for the integrated of a leakage inductor. Thus, magnetic integration and interleaving are successfully combined.

[0098] In advantageous embodiments the parallel paths are arranged symmetrically. Preferably the total current applied to the transformer is divided in equal portions over the various parallel paths.

[0099] While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive. The foregoing description details certain embodiments of the invention. It will be appreciated, however, that no matter how detailed the foregoing appears in text, the invention may be practiced in many ways. The invention is not limited to the disclosed embodiments.

[0100] Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. A single processor or other unit may fulfil the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. A computer program may be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims should not be construed as limiting the scope.

Claims

1. A transformer comprising an array of at least three sheets of magnetic material, each pair of adjacent sheets of said at least three sheets of magnetic material being separated by at least two pillars made of magnetic material and having a primary and secondary winding each with one or more winding turns formed on at least one of said at least two pillars, characterised in that the total number of winding turns of said primary winding on said at least two pillars between a first pair of said adjacent sheets differs from the total number of winding turns of said primary winding on said at least two pillars between a second pair of said adjacent sheets, and / or the total number of winding turns of said secondary winding on said at least two pillars between said first pair of said adjacent sheets differs from the total number of winding turns of said secondary winding on said at least two pillars between said second pair of said adjacent sheets.

2. The transformer as in claim 1, wherein said winding turns formed on said at least one pillar of magnetic material are formed in layers of a printed circuit board.

3. The transformer as in claim 2, wherein multiple winding turns are formed in a single layer of said printed circuit board.

4. The transformer as in claims 2 or 3, wherein at least some of said winding turns in said layers of the printed circuit board are alternatingly winding turns of the primary and the secondary winding.

5. The transformer as in any of claims 2 to 4, wherein said winding turns of said primary and secondary winding in said layers of said printed circuit board have a figure-8 shape.

6. The transformer as in any of claims 2 to 5, wherein said magnetic sheets of magnetic material are integrated in said printed circuit board.

7. The transformer as in any of claims 2 to 6, wherein said printed circuit board is flexible.

8. The transformer as in claim 7, wherein said printed circuit board is so folded that it comprises two substantially parallel end portions interconnected by means of flexible interconnectors.

9. The transformer as in any of the claims 2 to 6, wherein said winding turns of said primary winding and said secondary winding of said first pair are formed in layers of a first printed circuit board and said winding turns said primary winding and said secondary winding of said second pair are formed in layers of a second printed circuit board.

10. The transformer as in any of the previous claims, wherein said at least three sheets of magnetic material are positioned substantially in parallel.

11. The transformer as in any of the previous claims, wherein said primary and / or secondary windings are split into multiple parallel windings.

12. The transformer as in any of the previous claims, wherein said pillars extend in substantially perpendicular direction from said at least three sheets of magnetic material.

13. The transformer as in any of the previous claims, having a core with at least one U-shaped part and an I-shaped part.

14. The transformer as in any of the previous claims, wherein said winding turns of at least one of said first and second pair of adjacent sheets are divided into a plurality of subsets formed on each of the at least two pillars so that on each pillar a stack is obtained of said subsets and wherein subsets of winding turns of the primary and of the secondary winding alternate between consecutive layers of each stack, and a first and / or a last layer of said stacks accommodates a current path split into symmetrical parallel paths, one per pillar, with each of said symmetrical parallel current paths carrying an equal fraction of a total current, while winding turns in layers between said first and last layer carry double said fraction of the total current.

15. The transformer in any of the previous claims, arranged to be fed with a multi-phase input voltage.

16. DC-DC converter comprising a transformer as in any of the previous claims.

Images