INDUCTIVE CORE WITH LOW MAGNETIC LOSSES

Description

TECHNICAL FIELD AND STATE OF THE PRIOR ART

[0001] The present invention relates to an inductance core for producing inductances, particularly for manufacturing passive components in the field of power electronics, in particular at high frequencies, for example between 100 kHz and 10 MHz.

[0002] An inductor consists of a core and an electrical conductor arranged in n turns around a part of the core. The core is made of a ferromagnetic material characterized by a relative magnetic permeability µ. In operation, the turns carry an alternating electric current generating a magnetic induction of the same frequency in the core.

[0003] Such an inductance is for example used in a power converter, which is an electronic device whose function is to adapt the voltage and current delivered by an electrical power source to supply, according to specifications, a distribution network or a given electrical system.

[0004] The converter consists of electronic components that function as switches (active components) that switch at a given frequency. In the case of DC / DC converters, for example, the active components are transistors that are used to "cut" the input voltage according to regular cycles. In order to deliver a direct voltage at the output, inductors are used to store and release electrical energy over each cycle and to smooth the output voltage to its average value. These so-called "passive" elements are essential in the operation of converters, but they can represent up to 40% of the volume and cost of the converter.

[0005] Converters operating at high frequencies, for example above 1 MHz, can be produced using GaN material, which allows transistors capable of switching at very high frequencies to be produced. In theory, increasing the frequency is particularly interesting because it would reduce the volume of the passive components of the converters and therefore their size, mass and cost of these devices. Indeed, by increasing the switching frequency, the number of electrical cycles increases and thus the energy transferred by the magnetic core in a given time increases in the same proportion. Since the converter power remains constant, the volume of the magnetic inductances can theoretically be reduced inversely proportional to the frequency.

[0006] However, inductors compatible with operation at frequencies between 100 kHz and 10 MHz have inductance values between 1 µH and 10 mH. The most suitable inductors are monolithic inductors made of ferromagnetic material. This material is characterized by a relative magnetic permeability µ r > 50 and an induction Bs > 100 mT.

[0007] Ferrite-type oxide materials with a spinel crystallographic structure exhibit stable permeability values at high frequencies. For this reason, they are widely used as inductance cores, particularly for high-frequency operations between 100 kHz and 10 MHz. The most common formulations are (Mn1-xZnxFe2O4) and (Ni1-xZnxFe2O4). These materials are also characterized by high electrical resistivity values, limiting losses due to induced currents.

[0008] These ferromagnetic materials are the site of energy dissipation processes also called magnetic losses. These magnetic losses are dissipated in the form of heat at every point in the volume of the core.

[0009] Furthermore, a current in the turns creates a magnetic field and a variable induction of the same frequency as that of the current comprising a continuous component and a variable component.

[0010] The peak value of the variable induction can be written:

[0011] With B DC the continuous component and ΔB / 2 is the average between the two extrema of the variable component.

[0012] However, magnetic losses increase with frequency and with the peak value of magnetic induction.

[0013] One technique to reduce magnetic losses is then to reduce the peak value of the magnetic induction.

[0014] A first solution consists of generating a magnetic polarization by circulating a direct current around the core. The intensity of the direct current is determined by applying Ampere's theorem so as to create a constant induction value of opposite sign to the direct component B DC set by the converter. Such a solution is described in document US 6,388,896. This solution presents a certain bulk and a certain additional cost. For example, for small cores, space is not always available to make the additional winding.

[0015] A second solution is to generate magnetic polarization by means of magnets inserted into an area of the core or arranged against a face of the core. The magnets are arranged so as to cause the magnetic flux in the core to flow in the direction opposite to the magnetic flux corresponding to the direct component B DC .

[0016] Documents EP 1187150 and EP 1187151 A1 describe such a solution. The magnet(s) generate a magnetomotive force allowing the circulation of the magnetic flux throughout the magnetic circuit.

[0017] This solution is effective for inductors operating at low frequency and materials with high relative magnetic permeability, for example greater than 500. In this case, the entire magnetic flux produced by the magnet remains confined in the core and flux losses are low.

[0018] On the other hand, magnetic materials that can operate at frequencies above 1 MHz, such as NiZn ferrites, are characterized by permeability values below 100. In this case, the magnetic circuit is subject to magnetic leakage at the magnets, a part of the flux lines produced by each magnet loops directly from one pole to the other of the magnet by crossing the surrounding medium without traveling through the entire magnetic circuit. The efficiency of the magnetic polarization is therefore impaired and the value of the DC component of the induction is not effectively reduced. In addition, the magnetic flux lines radiate into the environment of the core, which can affect the operation of other components of the converter. Other documents relevant to the present application are: JP 2003 338414 A, JP H04 97315 U and US 2002 / 030574 A1. STATEMENT OF THE INVENTION

[0019] The aim of the present invention is therefore to provide an inductance core suitable for producing inductances capable of operating at high frequency, for example > 1 MHz, and having reduced magnetic losses.

[0020] The above stated object is achieved by an inductance core as defined by claim 1 and the manufacturing methods as defined in claims 10 and 13.

[0021] When a current flows through the winding, the core is the seat of two magnetic circuits, in one circulate the magnetic flux lines produced by the winding and in the other circulate the magnetic flux lines generated by the magnet(s). The flux lines circulate in opposite directions.

[0022] In other words, the ferromagnetic material is placed as close as possible to the magnet between its poles on the natural path of the magnetic flux lines produced by the magnet as they loop from the north pole to the south pole. This makes it easy to "collect" the flux lines. This creates the shortest path for the magnetic flux lines produced by the magnet between the north pole and the south pole, which produce a homogeneous magnetic flux in the ferromagnetic material. Since the magnetic flux produced by the magnet loops directly back into the material. magnetic flux produced by the magnet between the north pole and the south pole, which produce a homogeneous magnetic flux in the ferromagnetic material. Since the magnetic flux produced by the magnet loops directly back into the ferromagnetic material, it does not radiate or radiates little to the outside, so the operation of other components is little or not at all disturbed.The invention is therefore suitable for implementation in inductors whose ferromagnetic material has a low magnetic permeability, for example less than 100, and is particularly suitable for operation at high frequency.

[0023] Embodiments of the present invention are defined in the dependent claims of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The present invention will be better understood on the basis of the following description and the appended drawings in which: there Figure 1A is a longitudinal sectional view of an inductor core according to an example not forming part of the present invention, the Figure 1B is a cross-sectional view of the core of the Figure 1A , there Figure 2A is a schematically represented top view of an inductor implementing an inductor core according to an embodiment of the invention, the Figure 2Bis a perspective view of a half-core of type E, the Figure 3 is a perspective view of an inductor core of the Figure 2A , THE Figures 4A and 4B are graphical representations of the evolution of the magnetic induction B in mT for a state-of-the-art inductance core and the inductance core of the Figure 3 respectively as a function of time t in ms, the Figure 5 is a schematic representation of a state-of-the-art EE-type core and the magnetic flux lines passing through it, the flux lines being generated by a current flowing in a conductor wound around the central bar. DETAILED DESCRIPTION OF SPECIFIC EMBODIMENTS

[0025] The inductance core according to the invention uses one or more permanent magnets, but for the sake of simplicity the remainder of the description will only use the term “magnet” to designate a permanent magnet.

[0026] On the Figures 1A and 1B, we can see an inductance core N1 comprising a body 2 of cylindrical shape with longitudinal axis X and circular section, and a magnet 6. The body 2 comprises a ferromagnetic material 4. The body has an annular section and delimits within it a cavity 8 with longitudinal axis X. The shape and section of the core are not limiting, for example a body of square section falls within the scope of the present invention.

[0027] The core is advantageously monolithic, i.e. molded in a single piece.

[0028] The magnet 6 extends longitudinally along the X axis and has a circular cross-section. The south poles S and north poles N of the magnet are located at the longitudinal ends of the magnet 6. The outer diameter of the magnet 6 corresponds to the inner diameter of the cavity 8, so that the magnet can be arranged in the cavity 8 and is in contact with the ferromagnetic material 6. The length l1 of the magnet is at least equal to the length l2 of the ferromagnetic material. In the example shown, the length l1 of the magnet is substantially equal to the length l2 of the ferromagnetic material.

[0029] It will be noted that, in this case, the magnetic flux reversal zones naturally located in line with the poles of the magnet are outside the ferromagnetic material so as to allow a rectilinear flow of the flux in the core. The ferromagnetic material 4 then surrounds the magnet 6 over its entire length and its entire circumference. Furthermore, in the example shown, the magnet is in contact with the magnet over its entire circumference. However, an embodiment in which the magnet would not be in contact with the ferromagnetic material does not depart from the scope of the present invention.

[0030] The magnet produces magnetic flux lines Fm. Due to the relative arrangement of the magnet poles and the ferromagnetic material, the magnetic flux lines flow from the south pole S to the north pole N in the magnet 6 and then, thanks to the ferromagnetic material surrounding the magnet and extending between the S pole and the N pole, they loop back in the ferromagnetic material towards the S pole. The direction of the magnetic flux lines in the ferromagnetic material is opposite to that of the flux lines in the magnet.

[0031] All ferromagnetic material is then polarized uniformly by the magnet.

[0032] When the core N1 is used to make an inductance, a conductor (not shown) is wound around the core. The conductor is for example made of copper and has for example n turns with a longitudinal axis X.

[0033] A current flows in the conductor, which generates a magnetic field in the core and therefore magnetic flux lines.

[0034] By choosing either the direction of current flow of the conductor or the orientation of the polarity of the magnet, the magnetic flux lines generated by the magnet and those generated by the conductor flow in opposite directions. By also choosing the value of the magnetic field of the magnet, it generates a polarization that will reduce and advantageously cancel the continuous component of the induction generated by the current flowing in the conductor.

[0035] The peak value of induction is written: B ^ = B DC + Δ B 2

[0036] With B DC the continuous component and ΔB / 2 is the average between the two extrema of the variable component.

[0037] By canceling B DC thanks to the magnet, the peak value is then equal to ΔB / 2, its value is therefore reduced.

[0038] However, since magnetic losses are proportional to the peak value of induction, these are reduced as well as thermal losses.

[0039] The structure of the core, in particular the relative arrangement of the ferromagnetic material and the magnet, makes it possible to ensure a looping of the magnetic flux lines in the ferromagnetic material even in the case where the ferromagnetic material has a low permeability, for example less than 100. Indeed, the ferromagnetic material is arranged around the magnet on the natural passage of the magnetic flux lines produced by the magnet and looping from the north pole to the south pole. Thus, the polarization of the ferromagnetic material by the magnetic flux does not require a specific device, for example pole pieces, acting on the flux lines to guide them in the ferromagnetic material. These loop from the north pole to the south pole of the magnet over the entire length of the ferromagnetic material and this in a homogeneous manner, even with materials having a low permeability.

[0040] Furthermore, in the example shown, the ferromagnetic material advantageously surrounds the entire magnet, the magnetic flux lines looping symmetrically around the axis of the magnet, the majority of the magnetic flux lines are confined in the ferromagnetic material and the ferromagnetic material is homogeneously polarized.

[0041] Alternatively, it could be provided that the ferromagnetic material does not completely surround the magnet and extends, for example, only over an angular portion of the lateral surface of the magnet between the two poles. The ferromagnetic material of the core would then still be fully uniformly polarized, the peak value would then be reduced. However, a fraction of the magnetic flux of the magnet could leak into the surrounding medium.

[0042] On the Figures 2A And 2B , we can see an example of a core for an N2 inductor of type EE. This type of core has a high compactness.

[0043] The N2 core, seen from above on the Figure 2A , comprises a frame 10 of rectangular shape and a central bar 12 of longitudinal axis X' extending perpendicularly to the sides of the frame of greater length substantially in their middle. This central bar 12 is intended to be surrounded by the turns of a conductor (not shown). The bar 12 is in the example shown formed of two half-bars separated by an air gap 14.

[0044] The N2 core can be formed by assembling two half-cores 15 of type E as shown in the Figure 2B or be made directly from a single piece. Alternatively, it can be formed by assembling an E-shaped part and an I-shaped part or a U-shaped part and a complementary part.

[0045] The sides of the frame and the central bar then delimit two magnetic circuits C1 and C2 symmetrical with respect to a plane passing through the axis X of the central bar 12 and perpendicular to a mean plane of the frame. The two circuits are rectangular in shape. The magnetic circuits C1 and C2 are intended to be traversed by magnetic flux lines generated by the circulation of the current in the conductor 11, looping back at the air gap. The magnetic flux lines are designated by FM3 on the Figure 5 .

[0046] The core N2 also comprises magnets A1, A2, A3, A4, A5, A6, A7, A8 arranged in each of the magnetic circuits C1 and C2. The magnets A1 and A5 are located in the central bar 12 and are common to both magnetic circuits.

[0047] Both magnetic circuits are of similar structures, only circuit C1 will be described in detail.

[0048] The magnetic circuit C1 comprises straight portions 16.1, 16.2, 16.3, 16.4, 16.5. The portions 16.1 and 16.5 are formed by the two half-bars of the central bar 12. The magnets have, in the example shown, the shape of a rectangular parallelepiped extending over the entire thickness of the core, the thickness of the core being considered in a direction perpendicular to the mean plane of the core.

[0049] Magnet A2 extends almost the entire length of portion 16.2.

[0050] Magnet A3 extends almost the entire length of portion 16.3. Magnet A4 extends almost the entire length of portion 16.4.

[0051] Magnets A1 and A5 extend almost the entire length of portions 16.1 and 16.5 respectively.

[0052] Magnets A1 to A5 have an outer side face and an inner side face, the inside and outside being considered relative to the inside and outside of the magnetic circuit C1.

[0053] Alternatively, multiple aligned magnets could be implemented instead of a single magnet in each portion.

[0054] The magnets also form an open frame only at the air gap.

[0055] In the example shown, the magnets are arranged in the ferromagnetic material so that the ferromagnetic material covers the inner and outer faces of the magnets, and extends continuously between the N poles and S pole of two successive magnets. The magnets, in the example shown and preferably, extend throughout the thickness of the core and are flush with the front and rear faces of the core, the front and rear faces of the core being the faces parallel to the mid-plane of the core. As will be described later, the core may be made by molding a ferromagnetic material, cavities for the magnets being provided during the molding.

[0056] In the example shown, the width of magnetic material considered in the direction of the X axis for the portions 16.2 and 16.4 on the side of the inner faces of the magnets is greater than that on the side of the outer faces, but this is not limiting, the same thickness could be provided. This arrangement of the non-symmetrical magnets makes it possible to transfer the connection zones between magnets to the level of the deflectors, in the corners of the frame. The looping of the flux on each magnet takes place in a low-active zone of the inductance and does not affect its operation.

[0057] Furthermore, the magnets are arranged relative to each other so that the N pole of one magnet is opposite or close to an S pole of a following magnet.

[0058] Furthermore, the magnetic circuit C1 advantageously comprises deflectors between the poles of the successive magnets to guide the magnetic flux from one magnet to another, and to isolate the magnetic flux circulating in the magnets from that circulating in the ferromagnetic material.

[0059] The deflectors comprise, for example, non-magnetic zones 18 located near two poles of two successive magnets, more particularly they are in contact with the two successive magnets within a frame defined by the magnets.

[0060] The zones 18 advantageously comprise cavities 19 made in the thickness of the core and opening into the two faces of the core parallel to the mean plane of the core. The cavities 19 can be left empty and contain air, allowing heat to be evacuated to the outside of the core. In a particularly advantageous embodiment, the cavities 19 are filled with a non-magnetic, non-electrically conductive material offering good thermal conductivity, this material draining the heat to the outside of the core. The cavities are for example filled with AIN.

[0061] Preferably, the baffles have at least the same dimension as the thickness of the magnets.

[0062] The effect of the presence of magnets on the magnetic circuit C1 will now be described.

[0063] A magnetic flux FM1 flows in the magnet A1 from the S pole to the N pole, the flux leaves the magnet A1 through the N pole. Due to the presence of a non-magnetic zone 18, part of the magnetic flux enters the magnet A2 through the S pole after having circulated in the ferromagnetic material. Indeed, the cavity 19 prevents the magnetic flux lines from looping directly back to the S pole of the magnet A1 in the ferromagnetic material of the portion 16.1 and contributes to the homogeneity of the flux.

[0064] The magnetic flux then flows in magnet A2 towards the N pole, joins the S pole of magnet A3, in particular due to cavity 19, then magnet A4 and finally through magnet A5, exits through its N pole and due to the air gap which forms a non-magnetic deflector, the magnetic flux then flows in the opposite direction in portions 16.5, 16.4, 16.3, 16.2 and 16.1 and closes the circuit at the S pole of magnet A1. The magnetic flux circulating in the ferromagnetic material is designated FM2. Thanks to cavities 19, magnetic flux FM2 cannot loop back onto magnets A5, A4, A3, A2.

[0065] C1 magnetic circuits have two magnetic branches, one formed by the network of magnets and the other by the ferromagnetic material along the magnets.

[0066] In this advantageous embodiment, the magnetic flux generated by the magnets and circulating in the magnetic material FM2 is continuous along the entire magnetic path of the core. In addition, since the magnets extend throughout the entire thickness of the ferromagnetic material, the magnetic flux is homogeneous throughout the entire thickness of the ferromagnetic material. This results in a homogeneous polarization of the magnetic circuit C1. It could be provided that the magnets do not extend over the entire thickness of the core; the polarization would be less homogeneous, but the continuous component of the induction would nevertheless be reduced.

[0067] It should be noted that part of the magnetic flux leaving the N pole loops back directly to the south pole of the same magnet through the outer ferromagnetic material. This part of the flux that loops back through the outside of the magnet is directed in the same direction as the flux in the inner part, so it participates in the continuous polarization of the outer part.

[0068] In the example shown, the cavities have a square or rectangular section but it could be provided that they have another shape, for example an arc-shaped section extending between two successive magnets.

[0069] Alternatively, all the magnets could be replaced by a single, one-piece magnet forming an open frame at the air gap, which would eliminate the need for non-magnetic cavities. Alternatively, only some of the magnets could be made from a single piece, for example magnets A2 and A3 or A2, A3 and A4, etc.

[0070] A magnetic flux circulation FM2 is established in the same way in the magnetic circuit C2.

[0071] A magnetic flux is therefore generated homogeneously throughout the core.

[0072] In the example shown, magnets A1 and A5 are common to both magnetic circuits, but it could be planned to have magnets dedicated to the first magnetic circuit C1 and magnets dedicated to the second magnetic circuit C2.

[0073] When a current flows in the conductor 11 surrounding the central bar 12, a magnetic field FM3 is generated, a magnetic flux flows in the two magnetic circuits and generates a variable induction having a continuous component and a variable component (relation I).

[0074] By choosing and orienting the magnets so that the generated magnetic flux cancels the DC component of the induction generated by the conductor in the core, the peak value of the induction generated in the core and the magnetic losses, and therefore the heating of the core, can be reduced. The orientation of the magnets and the flow of current in the conductor are such that the magnetic flux FM2 and the magnetic flux FM3 (dotted lines on the Figure 2A ) generated by the driver have opposite directions.

[0075] The present invention applies to any shape of inductance core, for example it could have a U shape, the magnets extending into the bottom of the U and into the two branches of the U, the magnetic flux FM2 looping at the free ends of the branches of the U.

[0076] Preferably, the magnets are made of electrically non-conductive materials to reduce the risk of coupling and the occurrence of high-frequency eddy currents which would cause heating of the core.

[0077] Advantageously, the magnets are bonded or plastomagnets. For example, the magnets comprise magnetic powders dispersed in a polymer matrix or an electrically insulating resin. They can be advantageously molded into complex shapes. These magnets then have high electrical resistivity. The bonded magnets can be of the NdFeB type with a value of BHmax = 10 MGOe. Alternatively, the magnets could be made of SmCo, ferrite, or SmFeN.

[0078] According to a variant of the core of the Figure 1A, magnet 6 could be replaced by several magnets aligned so that the N pole of one magnet is opposite the S pole of the other magnet. In addition, deflectors would be provided at the level of the opposite poles to prevent the magnetic flux lines leaving the N pole of a magnet from looping directly back to the S pole of the magnet instead of joining the opposite S pole.

[0079] A sizing example will now be given.

[0080] On the Figure 3 , we can see the core of the Figure 2A in perspective. We consider a core containing NiZ as ferromagnetic material. The core has an external length I equal to 46 mm, an external width L equal to 30 mm, a thickness equal to 11 mm. The sides of the frame have a width equal to 6 mm, the central bar 12 has a width equal to 12 mm and the air gap is equal to 3 mm.

[0081] The magnets are parallelepiped and all have a thickness of 11 mm. Magnets A1 and A5 are 10 mm long and 2.4 mm wide. Magnets A3 and A7 are 23 mm long and 1 mm wide. Magnets A2, A4, A6 and A8 are 17 mm long and 1 mm wide.

[0082] The eight cavities 19 have a square section of 1 mm × 1 mm and a height of 11 mm and are filled with air.

[0083] This core allows for example to produce a step-up chopper converter having the following characteristics: P = 1 kW, F = 5 MHz, D = 0.5, Ve = 200 V, r = 0.4; Ve being the input voltage of the converter, D the duty cycle of the converter (fraction of the cycle where the switch is closed) and r the ripple rate of the DI / Idc current.

[0084] For the magnet, the remanent induction is Br = 0.7 T and for the current the continuous average value Idc = 5A and the ripple DI = 2 A.

[0085] On the Figures 4A , we can see the variation of the magnetic induction B in mT generated by the current flowing in the conductor during a cycle as a function of time t in ns in a core of type EE of the state of the art, i.e. without magnet, in NiZn and having the same dimensions as the core of the Figure 3

[0086] On the Figure 4B , we can see the variation of the magnetic induction B in mT resulting from the polarization by the magnets in a core of the Figure 3 during a cycle as a function of time t in ns.

[0087] On the Figure 4B , we see that the continuous component BDC is equal to 0, whereas without polarization this continuous component is worth 55 mT ( Figure 4A). The variable component varies in both cases by 22 mT. The peak value of the induction is therefore reduced by 55 mT in the core of the invention, which makes it possible to significantly reduce the heating of the core. For example, in the case of a NiZn type core the losses dissipated per unit volume of the core P d are reduced by a factor of 10 and the dissipated power can be evacuated by simple natural convection from the surface of the core.

[0088] An example of a method for producing a core according to the invention will now be described.

[0089] The inductance cores according to the invention can be very advantageously produced by powder injection molding (or PIM for Powder Injection Molding in English terminology).

[0090] In a PIM process, the first step is to obtain a masterbatch (or "feedstock" in English terminology) adapted to the intended application. Masterbatches consist of a mixture of organic material (or polymeric binder) and inorganic powders (metallic or ceramic) which will constitute the final part. Then, the masterbatch is injected like a thermoplastic material in an injection molding machine using technology known to those skilled in the art. Molding melts the injected polymers with the powder in a cavity and gives the mixture the desired shape. During cooling, the mixture solidifies and retains the shape given by the mold.

[0091] After demolding, the part is subjected to various thermal or chemical treatments to remove the organic phases. The elimination of the organic phase during this step, called debinding, leaves a porosity of 30% to 50% in the blank.

[0092] An example of a method for preparing a masterbatch and debinding in the case of PIM manufacturing is described in US 8940816 B2.

[0093] After debinding, the porous blank contains only the powders of the inorganic material. This blank is then densified to form the final dense part. The consolidation of the porous blanks is achieved by high-temperature sintering, preferably at a temperature above 1000°C, carried out in furnaces operating in an atmosphere adapted to the type of material used. When the optimal density is reached, the part is cooled to room temperature.

[0094] Preferably, to produce the cores according to the invention, NiZn or MnZn spinel ferrite powders are used in a mixture with the organic material to produce the masterbatch. The ferrite powders are, for example, produced by solid-state or chemical synthesis. The solid-state synthesis comprises the steps of grinding precursor oxides and synthesizing the spinel phase by heat treatment of the ground powders between 800°C and 100°C. The powders are again ground and sieved to obtain a particle size of the order of 10 µm to 20 µm. For NiZn and MnZn spinel ferrites, sintering can be carried out in air according to operating conditions well known to those skilled in the art for this type of material.

[0095] Alternatively, other soft ferromagnetic materials can be used to make the masterbatch. These materials are, for example, shaped by powder metallurgy, such as Fe-based magnetic alloys (Fe-Si, Fe-Co, Fe-Ni).

[0096] After preparing the masterbatch, it is shaped in a mold.

[0097] To realize the core of the Figure 3 , the mold is such that it forms the cavities 18 and the cavities intended to house the magnets.

[0098] Preferably, the EE type core is made of two or more symmetrical parts molded separately and then assembled. The mold has removable inserts which are positioned in the mold so as to create, on the molded part, the through cavities for the magnets and to form the deflectors.

[0099] After molding the master batch and cooling the green part, a step of debinding the organic material takes place. This takes place, for example, in an oven, maintaining the temperature between, for example, 400°C and 700°C during the temperature rise.

[0100] Sintering to densify the core then takes place, advantageously in the furnace used for debinding. Thus, sintering can be carried out directly after debinding by continuing to increase the temperature to the value recommended for the magnetic phase in question. Debinding takes place, for example, at 1220°C.

[0101] In a next step, the magnets are introduced into the cavities. The magnets can be bonded magnets manufactured beforehand. For example, they are molded and magnetized according to the dimensions adapted to the polarization of the core. The bonded magnets can be of any type, for example NdFeB, SmCo, SmFeN, hexaferrites. The polymer matrix, in which the magnetic powders are dispersed, is chosen so as to be compatible with the operating temperature of the inductor, for example, this is between 100°C and 150°C. The magnets can be held in the cavities by means of an adhesive capable of withstanding the operating temperature.

[0102] In a subsequent step, it may be provided to fill the cavities 16 with a non-magnetic, non-electrically conductive and good thermal conductor material, such as AIN. For example, the filling material is previously shaped by extrusion or molding and then introduced into the cavities 16 in a manner similar to the mounting of the magnets. This step of filling the cavities 16 may not take place, the air-filled cavities being retained.

[0103] AIN can also be held in the cavities using an adhesive capable of withstanding the operating temperature.

[0104] According to another example of a method, the inductance core can be produced by overmolding the ferromagnetic material around the magnets and possibly the elements forming the non-magnetic zones. The sintering step can be omitted. Advantageously, the ferromagnetic material can also be overmolded onto the n-turn conductor.

Claims

1. Inductor core for magnetic inductor, comprising a body comprising : - a body comprising a ferromagnetic material, and - several magnets (6, A1, A2, A3, A4, A5), each magnet comprising an exterior lateral face between the south pole and the north pole, said magnets being arranged relative to the others with the north pole of a magnet that is near to a south pole of a successive magnet to form a first path for circulating magnetic flux lines produced by the magnets (6, A1, A2, A3, A4, A5), said first path comprising at one end a south pole (S), and at another end a north pole (N), and in which the ferromagnetic material forms a second path for circulating said magnetic flux lines, intermediate non-magnetic zone (18) non-magnetic zone, arranged between the poles of two successive magnets to guide the magnetic field from a magnet to another and isolate the magnetic field flowing in the magnets from the one flowing in the ferromagnetic material, said ferromagnetic material extending continuously from the south pole (S) to the north pole (N) along the magnets (6, A1, A2, A3, A4, A5), and comprising, facing the end north pole (N), a first non-magnetic zone forcing the magnetic flux lines coming out of the end north pole to take the second path and, facing the end south pole (S), a second non-magnetic zone forcing the magnetic flux lines coming out of the second path to loop back on the end south pole, said body comprising several cavities, each cavity receiving a magnet, said cavities being such that the lateral face of each magnet is at least partly surrounded by the ferromagnetic material of the body.

2. Inductor core according to claim 1, in which the ferromagnetic material is in contact with at least a part of the exterior lateral surface of each magnet.

3. Inductor core according to claim 1, in which each intermediate non-magnetic zone (18) comprises an additional cavity (19), said additional cavity (19) advantageously emergng in opposite exterior faces of the body, and saif additional cavity (19) being for example filled with a heat conducting and electrically insulating material, for example AIN.

4. Inductor core according to claim3, in which the body comprises a given thickness, said magnets (A1, A2, A3, A4, A5) extending over the entire thickness of the body (10).

5. Inductor core according to claim 3 or 4, in which the body comprises a rectangular frame (10) and a central bar (12) arranged transversally with respect to the sides of the frame of longest length and parallel to the sides of the frame of smallest length, and in which two first paths are delimited in the frame (10) and in the central bar (12) in a symmetrical manner with respect to a plane of symmetry passing through the central bar (12) and perpendicular to a mean plane of the frame, and two second paths are delimited in the frame and in the central bar in a symmetrical manner with respect to said plane of symmetry and in which the central bar comprises an air gap.

6. Inductor core according to one of claims 3 to 5, in which the air gap is arranged between the end south pole and the end north pole and forming the end non-magnetic zones.

7. Inductor core according to one of preceding claims, in which the magnets are of bonded type comprising at least one powder magnetic material dispersed in a matrix made of electrically insulating material and / or the ferromagnetic material has a permeability less than 100, the ferromagnetic material being for example a spinel ferrite selected from NiZn and MnZn.

8. Inductor comprising an inductor core according to one of preceding claims and a conductor wound around at least one part of the core.

9. Converter comprising at least one electronic component and at least one inductor according to claim 8.

10. Method for manufacturing an inductor core according to one of claims 1 to 7, comprising the steps of: a) Supplying at least one magnet, b) Manufacturing a body made of ferromagnetic material by injection moulding from a feedstock comprising at least one ferromagnetic powder and organic matter, so as to arrange at least one cavity for the mounting of the magnet in the body, c) Mounting the magnet in the cavity.

11. Method for manufacturing according to claim 10, in which during step b), at least one cavity is produced to form a non-magnetic zone, said method comprising for example a step of putting in place a non-magnetic, non-electrically conducting and heat conducting material in the cavity forming the non-magnetic zone.

12. Method for manufacturing according to claim 10 or 11, in which step b) comprises a sub-step of moulding the feedstock, a sub-step of debinding and a sub-step of heat treatment, the sub-step of heat treatment advantageously taking place directly after the sub-step of debinding by increasing the temperature with respect to that of the debinding.

13. Method for manufacturing an inductor core according to one of claims 1 to 8, comprising the steps of: a') Supplying at least one magnet, b') Manufacturing a body made of ferromagnetic material by over-moulding on the magnet.