Method for measuring high-frequency loss of thin film magnetic core

Abstract

The invention belongs to the field of magnetic core loss measurement, and particularly relates to a thin film magnetic core high-frequency loss measurement method. According to the technical scheme, an initial magnetization curve and hysteresis loops under different external magnetic field intensities are measured by using a vibration sample magnetometer (VSM), then a two-winding method measurement platform is built, and high-frequency phase deviation is calibrated by using an air core transformer; after the loss curve of the laminated thin film magnetic core is obtained through calculation, a finite element model containing the thin film magnetic core is established and corrected; and finally, determining the complex permeability of the laminated thin film magnetic core, substituting the complex permeability into the finite element model, eliminating the influence of the air core coupling main magnetic flux on the magnetic induction intensity measurement value in the thin film magnetic core, carrying out iterative updating on the loss curve, and taking the loss curve obtained by the last iterative updating as output. According to the technical scheme, the thin film magnetic core loss suitable for magnetic integration under high frequency can be accurately measured.

Description

Technical Field

[0001] The invention relates to the field of magnetic core loss measurement, and in particular to a method for measuring high-frequency loss of a thin-film magnetic core. Background Art

[0002] In recent years, with the increase of switching frequency, the use of co-packaged passive devices and power management ICs, integrated power modules have become possible. Below 100MHz, the magnetic components required for switching power supplies must use magnetic cores to increase the inductance and power density of the components in order to facilitate the integration of magnetic components. The magnetic cores used for integrated magnetic components should have superior performance and a preparation method compatible with CMOS technology. Therefore, thin film cores are the most suitable and have triggered extensive research. For this reason, it is necessary to accurately measure the high-frequency loss of thin film cores in order to analyze and design integrated magnetic components.

[0003] The most relevant core loss measurement method for this solution is the two-winding method (also known as the AC power method, two-winding method) widely used in block core loss measurement. Its classic measurement circuit is as follows: Figure 1 As shown. This method uses a toroidal magnetic core, on which a magnetic field intensity induction coil (also used as the primary excitation winding) and a magnetic induction intensity induction coil (secondary winding) are wound. Therefore, in terms of the equivalent circuit, the two-winding system in this measurement scheme is equivalent to a transformer. Under this measurement scheme, the instantaneous magnetic core loss power is equal to the winding open circuit voltage u measured by the magnetic induction intensity induction coil B The current i flowing through the magnetic field strength induction coil H The product of the average power loss P of the toroidal core under the assumption of periodic excitation. core for:

[0004]

[0005] Where T is the excitation period. The advantage of this measurement method is that it can eliminate the influence of AC winding loss (copper loss) in the measurement system. However, as the applied excitation frequency increases, the phase error caused by the parasitic parameters of the current sampling resistor or the voltage / current probe mismatch will lead to significant loss measurement errors. Therefore, based on this classic measurement structure, a series of improved methods have been proposed to improve the measurement accuracy of core loss at high frequencies. One of the typical methods is to take a capacitor C in parallel with the excitation winding so that it is parallel to the excitation inductance L. m The resonance is induced at the frequency of interest, thereby reducing the sensitivity of the phase error to the core loss. However, due to the addition of capacitance, the voltage on both sides of the capacitance must be considered during measurement, which introduces winding loss and causes additional errors. To this end, Mu proposed Figure 2 The improved solution shown in the figure. In this case, a series capacitor is used to measure the port voltage v3 To replace the port voltage v 2 To eliminate winding loss. Due to the resonant capacitor C and the magnetizing inductor L m Series connection, such as Figure 2 As shown, the instantaneous core loss power is v 3 with i R The disadvantage of this method is that it is only applicable to sinusoidal excitation. In order to extend the excitation to arbitrary waveforms, we can further Figure 2 The scheme shown is improved to obtain Figure 3 The circuit shown uses an air core transformer to replace the capacitor C. Different voltage polarities are obtained by using the same and different terminals to eliminate v 3 The inductive voltage drop at different frequencies reduces the sensitivity of the core loss measurement to phase offset errors. Summary of the invention

[0006] In view of the shortcomings of the existing high-frequency core loss measurement technology, the present invention proposes a method for measuring the high-frequency loss of thin-film magnetic cores. The measurement is based on the classic two-winding method. The thin-film magnetic core sample to be tested is formed of a laminated composite structure of a magnetic film with high magnetic permeability and an insulating layer. After being made into a ring shape, primary and secondary windings made of enameled wire are wound on it. The primary winding is used to provide excitation current and to measure the magnetic field intensity, and the secondary winding is used to measure the magnetic induction intensity in the magnetic core. This structure can more simply and accurately measure the core loss of thin-film inductors, but in the measurement process, in addition to dealing with the phase offset problem caused by high-frequency measurement, it is also necessary to eliminate the main magnetic flux coupling effect of the air core through a certain procedure. To this end, the present invention proposes to use a standard air core transformer on the one hand to calibrate the phase offset within a wider frequency range, and on the other hand, by using the quasi-static magnetization curve of the thin-film magnetic core measured by a vibrating sample magnetometer (VSM), the finite element modeling of the thin-film magnetic core and the measurement results of the above two-winding method, the present invention proposes an iterative algorithm for eliminating the main magnetic flux coupling effect of the air core. Finally, the measurement accuracy of the high-frequency loss of the thin-film magnetic core is effectively improved.

[0007] The present invention provides a method for measuring high-frequency loss of a thin film magnetic core, comprising the following steps:

[0008] S1 obtains a laminated thin-film magnetic core to be tested, wherein the laminated thin-film magnetic core to be tested is annular, and uses VSM to measure the initial magnetization curve of a sample of the same magnetic material as the laminated thin-film magnetic core to be tested and the hysteresis loop under different external magnetic field strengths; based on the initial magnetization curve, extracts the initial magnetic permeability of the sample close to the magnetic neutrality; and obtains an equivalent quasi-static magnetization curve by sequentially connecting the vertices of each hysteresis loop;

[0009] S2 builds a two-winding measurement platform:

[0010] The two-winding method measurement platform comprises: a replaceable annular magnetic core and a magnetic field intensity induction coil and a magnetic induction intensity induction coil uniformly wound on the replaceable annular magnetic core; the shape and size of the replaceable annular magnetic core are the same as those of the laminated thin film magnetic core to be measured; the current flowing through the magnetic field intensity induction coil is converted into a voltage signal via a sampling resistor connected in series thereto, and is measured by one channel of an oscilloscope, while the voltage signals on both sides of the magnetic induction intensity induction coil are measured by another channel of the oscilloscope;

[0011] The magnetic field intensity induction coil is used as an excitation winding, and its feed source is composed of a signal generator connected to a power amplifier;

[0012] The S3 uses an air core transformer to calibrate high frequency phase deviation:

[0013] Get a standard air core transformer of the same shape and size as the replaceable toroidal core,

[0014] The standard air core transformer is used to replace the replaceable annular magnetic core in the two-winding method measurement platform to obtain the phase deviation required to be compensated for the high-frequency measurement of the thin film core at different frequencies; and a calibrated two-winding method measurement platform is obtained;

[0015] The specific method for obtaining the phase deviation that needs to be compensated includes the following steps:

[0016] In the actual measurement of the secondary voltage u of the magnetic induction intensity induction coil B The primary excitation current i of the magnetic field intensity induction coil H Finally, the average air core loss P in one cycle is calculated using the two-winding method. core , the calculation formula is as follows:

[0017]

[0018] Because the air core loss should be zero, the air core loss calculated according to formula (1) is approximately zero by adding a time delay to the current curve; the added time delay corresponds to the phase angle required to be adjusted, which is used to calibrate the phase error measured by the system;

[0019] S4 obtains the loss curve of the laminated thin film core:

[0020] Loading the laminated thin film magnetic core to be tested obtained in step S1 onto a calibrated two-winding method measurement platform, replacing the standard air core transformer;

[0021] Use the two synchronous channels of the oscilloscope to measure the induced voltage on both sides of the magnetic induction coil and the excitation current flowing through the magnetic field intensity induction coil, respectively, as the secondary voltage u B With the primary excitation current i H; Use the phase angle required to adjust the excitation current i obtained by S3 H The phase of the laminated thin film core to be tested is calculated using formula (1);

[0022] Secondly, use the secondary voltage u B According to Faraday's induction law, the magnetic induction intensity amplitude B in the magnetic core is calculated using formula (2): m for:

[0023]

[0024] where n B is the number of turns of the magnetic induction coil, S is the cross-sectional area of ​​the laminated film core; according to the average power loss P core and the magnetic induction intensity amplitude B in the core m Obtain loss curve B m ~P core ;

[0025] S5 establishes a finite element model including a thin film core and calibrates it:

[0026] The calibrated two-winding method measurement platform is modeled using finite element analysis software to obtain a finite element model; the finite element model models the magnetic field intensity induction coil, the magnetic induction intensity induction coil and the laminated thin film magnetic core to be measured in step S1, wherein the geometric dimensions and electrical parameters of each component are consistent with the corresponding components in the calibrated two-winding method measurement platform;

[0027] After the modeling is completed, the finite element model is calibrated, including the following steps:

[0028] On the one hand, the DC inductance value of the inductor composed of the magnetic field intensity induction coil and the thin film core is measured by the step-voltage method after the magnetic induction intensity induction coil is open-circuited, as the measurement value of the equivalent DC inductance; during the measurement, the step voltage value should be small enough to avoid saturation, so that the magnetic permeability of the laminated thin film core is equal to the initial magnetic permeability;

[0029] On the other hand, a DC excitation is applied to the finite element model, and the magnetic permeability of the laminated thin film core is taken as the initial magnetic permeability; the calculated value of the equivalent DC inductance is extracted through finite element simulation;

[0030] Compare the measured value and calculated value of the equivalent DC inductance, and adjust the geometric dimensions of the winding in the finite element model so that the measured value and the calculated value match within the preset engineering accuracy, thereby obtaining a calibrated finite element model;

[0031] S6 determines the complex magnetic permeability of the laminated thin film core, substitutes the complex magnetic permeability into the finite element model proposed and calibrated in S5, eliminates the influence of the air core coupling main magnetic flux on the measured value of the magnetic induction intensity in the thin film core, and replaces the magnetic induction intensity amplitude B in the core obtained in step S4 m , update the loss curve;

[0032] S7 iteratively updates the loss curve until the difference between the loss curve obtained after the current iterative update and the loss curve before the current iterative update is no greater than a preset threshold, stops the iterative update and outputs the loss curve obtained by the last iterative update.

[0033] Preferably, in step S5, the DC inductance is measured using a step voltage method, which specifically includes the following steps:

[0034] Constructing a measurement circuit for measuring DC inductance using a step voltage method, the circuit comprising a DC voltage source, a capacitor C connected between two poles of the DC voltage source, a switch, a small sampling resistor R and an equivalent DC inductor L connected in series between two plates of the capacitor C;

[0035] A step voltage is applied to the equivalent inductor composed of the magnetic field intensity induction coil and the thin film core to be tested. At this time, the magnetic induction intensity induction coil is open; first, the capacitor is charged by a DC voltage source; secondly, after the charging is completed, the switch is closed to apply a specified step voltage excitation to the equivalent DC inductor L; the current i flowing through the equivalent DC inductor L B The small sampling resistor R is used to convert the measured primary voltage into a voltage signal for measurement. The flux λ(i B ), the equivalent DC inductance is calculated as follows:

[0036]

[0037] The step S6 specifically comprises the following steps:

[0038] According to the equivalent quasi-static magnetization curve and loss curve measured by S1, the complex magnetic permeability μ is obtained, and the formula is as follows:

[0039] μ=μ′-jμ″ (4)

[0040] The real part μ′′ is obtained using the equivalent quasi-static magnetization curve;

[0041] Based on the following formula, the loss curve B measured by S4 is m ~P core Find the imaginary part μ′”

[0042]

[0043] The obtained complex magnetic permeability μ is substituted into the finite element model proposed and calibrated by S5;

[0044] By using the finite element model, the actual excitation main magnetic flux flowing through the thin film magnetic core is extracted to eliminate the influence of the air core coupling main magnetic flux on the measured value of the magnetic induction intensity in the thin film magnetic core, and the magnetic induction intensity amplitude B in the magnetic core obtained in step S4 is replaced. m , and update the loss curve.

[0045] Preferably, the step S7 specifically includes the following steps:

[0046] Compare the loss curve obtained after this iteration update with the loss curve before this iteration update. If the difference between the two is greater than the preset threshold, the loss curve is updated by under-relaxation iteration, that is, the magnetic induction intensity is adjusted according to the following formula:

[0047]

[0048] in is the magnetic induction intensity amplitude in the core used to obtain the loss curve before the kth update, is the magnetic induction intensity amplitude in the core used to obtain the loss curve after the kth update, ω r is the relaxation factor used for iteration, 1<ω r <2, while is the magnetic induction intensity amplitude in the magnetic core used to obtain the loss curve before the k+1th update;

[0049] If the difference between the loss curve obtained after this iterative update and the loss curve before this iterative update is greater than a preset threshold, steps S6 to S7 are repeated; otherwise, the iteration is terminated and the loss curve obtained by the last iterative update is output.

[0050] Preferably, in step S7, the difference between the two is specifically the difference between the two in the least squares sense.

[0051] Preferably, in step S7, the preset threshold is no greater than 5%.

[0052] The present invention is based on the two-winding measurement method, firstly abandons the block-shaped annular magnetic core and replaces it with a laminated film structure, which is also made into an annular shape, such as Figure 4 This structure is used, firstly, to suppress the eddy current loss in the magnetic core under high-frequency excitation; secondly, this magnetic core is used for thin-film integrated magnetic components. If too many layers are stacked, or even a block magnetic core is made, the process will be too complicated; finally, if a block magnetic core is used, the magnetization performance is very different from that of an actual thin-film magnetic core, and it cannot be used to analyze and design the performance of integrated magnetic components.

[0053] Secondly, in order to solve the phase error caused by the parasitic parameters of the current sampling resistor or the mismatch of the voltage / current probe under a wider frequency band, the present invention uses a standard air core inductor to calibrate the measurement error. The advantage of this method is that the phase offset required for calibration under a wider frequency band can be obtained without selecting a suitable critical capacitor or air core transformer.

[0054] Finally, the magnetic induction intensity flowing through the annular sample measured by the two-winding method depends on the main magnetic flux flowing through the primary and secondary windings at the same time, and it is assumed that most of the main magnetic flux flows through the annular core to achieve the coupling of the primary and secondary windings. However, due to the high excitation frequency and the thin film core thickness, part of the main magnetic flux in the measured two-winding structure will not pass through the thin film core to complete the coupling between the magnetic field intensity induction coil and the magnetic induction intensity induction coil, which will affect the loss measurement results and overestimate the magnetic induction intensity flowing through the annular core. For this reason, it is necessary to eliminate the above-mentioned air core main magnetic flux coupling effect based on the two-winding method to ensure the accuracy of the thin film core loss measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the specific implementation of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the specific implementation or the prior art description. Some specific embodiments of the present invention will be described in detail in an exemplary but not restrictive manner with reference to the drawings. The same reference numerals in the drawings indicate the same or similar components or parts. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings:

[0056] Figure 1 It is the classic measurement circuit of the two-winding method;

[0057] The two-winding method system can be equivalent to a two-winding transformer, where V 1 , V 2 are the primary voltage and secondary voltage of the two windings, R 1 , L w1 is the primary winding resistance and leakage inductance, R 2 , L w2 is the secondary winding resistance and leakage inductance, R core With L m is the iron loss resistance and magnetizing inductance, and V m is the excitation voltage.

[0058] Figure 2 Improved two-winding method measurement circuit for series resonant capacitor;

[0059] In the improved two-winding system, V 1 , V 2 , R 1 , L w1 , R 2 , Lw2 The meaning and Figure 1 Same, as and It is still the iron loss resistance and excitation inductance, but the method of calculating loss and stored magnetic energy is the same as R core and L m Different from R, it is introduced to facilitate mathematical processing. s For current using resistance, V R The voltage on both sides of it, i R For the flow through R s The current is 0, while C is the capacitor used for resonance. V3 is the voltage introduced to facilitate the extraction of core loss.

[0060] Figure 3 Improved two-winding method measurement circuit for series air core transformer;

[0061] Compared to Figure 2 system, Figure 3 The improvement of the system shown is mainly to replace the resonant capacitor with an air core transformer. Therefore, in addition to the original secondary winding resistance R σ And the total leakage inductance L σ The remaining parameters are Figure 2 The same as shown.

[0062] Figure 4 This is a product picture of a laminated thin-film core.

[0063] Figure 5 It is a schematic diagram of the coupling effect of the main magnetic flux of the air core, in which the magnetic flux flowing through the film core in the turn link primary and secondary windings is the main magnetic flux of the magnetic core, and the magnetic flux flowing through the air in the turn link primary and secondary windings is the main magnetic flux of the air core.

[0064] Figure 6 Schematic diagram of the two-winding method measurement platform.

[0065] Figure 7 Measurement circuit for measuring DC inductance using the step voltage method.

[0066] Figure 8 This is a flow chart of the thin film core loss measurement in Example 1.

[0067] Fig. 9 It is the geometric entity of the two-winding method finite element modeling in Example 1.

[0068] Fig.10 The test loss curve, the simulation loss curve and the corrected loss curve in Example 1. DETAILED DESCRIPTION

[0069] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0070] 1) Prepare the laminated thin film core sample and measure its magnetization curve

[0071] In order to facilitate the preparation of test samples for the two-winding test system, the thin film core is prepared into a ring shape to apply the winding and eliminate the demagnetization field. At the same time, due to the high-frequency measurement, a laminated structure should be used to eliminate the influence of eddy current loss as much as possible. The specific preparation instructions are as follows:

[0072] Electroplating magnetic film: First, a seed layer is formed on the bottom silicon wafer substrate by backsputtering and sputtering, and then a magnetic film of the required size is chemically grown on the seed layer through an electroplating machine with certain process parameters.

[0073] Laminated composite: The required annular magnetic core is cut out of the entire magnetic film by machine pressing, and each layer of the magnetic core is isolated by high electrical insulation material.

[0074] The initial magnetization curve of the same magnetic material sample (such as an ellipsoid) and the hysteresis loop under different external magnetic field strengths are measured by a vibrating sample magnetometer. The former can be used to extract the initial magnetic permeability, and the latter can be used to obtain the equivalent quasi-static magnetization curve by sequentially connecting the vertices of each hysteresis loop.

[0075] 2) Build a two-winding measurement platform

[0076] like Figure 6 As shown, take the thin film magnetic core prepared above and the air core of the same shape and size, and apply the magnetic field intensity induction coil and the magnetic induction intensity induction coil on them. The two sets of windings need to be wound carefully to ensure their uniformity as much as possible. The magnetic field intensity induction coil is used as the excitation winding, and its feed source is composed of a signal generator connected to a power amplifier. The measurement can be made with the help of an oscilloscope with two synchronous channels to measure the induced voltage u on both sides of the magnetic induction intensity induction coil. B The excitation current i flowing through the magnetic field strength induction coil H In specific implementation, the excitation current is connected to the primary side sampling resistor R H Note that due to high frequency measurement, the feed line and measurement lead in the figure can be regarded as transmission lines, and their length should be minimized to suppress the influence of distributed parameters.

[0077] 3) Using air core transformer to calibrate high frequency phase deviation

[0078] A standard air core transformer is inserted into the two-winding measurement device to replace the toroidal core, so as to obtain the phase deviation that needs to be compensated for loss measurement at different frequencies. The specific method is that after actually measuring the voltage curve and current curve, the core loss can be obtained using (1):

[0079]

[0080] Because the core loss of an air core inductor should be zero, the core loss obtained according to (1) can be approximately zero by artificially adding a time delay to the current curve. The phase angle (time delay) that needs to be adjusted is exactly the phase error measured by the system.

[0081] 4) Measure the loss curve of the thin film core

[0082] The method for calculating the core loss is: using the secondary voltage u measured by the two-winding method B With the primary excitation current i H After that, the phase of the excitation current is adjusted using the phase angle required, and then the average power loss of the thin film core in one cycle T can be calculated according to (1). In addition, using u B According to Faraday's induction law, the magnetic induction intensity amplitude B in the magnetic core can be obtained m for:

[0083]

[0084] where n B is the number of turns of the magnetic induction coil, and S is the cross-sectional area of ​​the laminated film core. Based on this, the loss curve B can be obtained. m ~P core .

[0085] 5) Determine the complex permeability of the thin film core

[0086] According to the quasi-static equivalent magnetization curve measured in 1) and the loss curve measured in 4), the complex magnetic permeability can be introduced as follows:

[0087] μ=μ′-jμ″ (4)

[0088] The real part μ' is obtained using the equivalent magnetization curve extracted in 1).

[0089]

[0090] According to this, the loss curve B measured by claim 4) m ~P core μ” can be obtained.

[0091] 6) Establish a finite element model including thin film core and calibrate it

[0092] In the two-winding method measurement system, as far as we know, the electrical measurement method cannot directly separate the main magnetic flux flowing through the magnetic core and the main magnetic flux coupling flowing through the air core. The magnetic field distribution in the magnetic core must be known before the two can be specifically separated. The magnetic field distribution in the magnetic core can be calculated by the electromagnetic field numerical calculation method. Among the many electromagnetic field numerical calculation methods, the finite element method is known for its flexible geometric modeling, easy to deal with nonlinear problems, high calculation efficiency, and easy program standardization. This method first divides the area of ​​the magnetic field distribution to be calculated into discrete grids. Then, relying on the grid, the Maxwell equations are transformed into a set of algebraic equations using the variational minimum principle or the Galerkin method, and the magnetic field distribution sampled on the grid is obtained using numerical algebra methods. At present, there are many commercial / open source software that can realize this calculation. The general process is to divide the area of ​​the magnetic field distribution to be calculated into discrete grids, and then assign the material properties of the corresponding physical entity to each grid (for magnetic field analysis, that is, assign magnetic permeability and electrical conductivity). After applying current excitation to the current source area and defining the boundary conditions, the magnetic field distribution of the area to be calculated can be realized. Specifically for the present invention, firstly, a geometric entity including a magnetic field intensity induction coil, a magnetic induction intensity induction coil and a thin film magnetic core is established, and attention is paid to ensuring that its geometric dimensions are consistent with the corresponding parts of the measuring device. Then, these structures and the adjacent air area are discretized, and the material properties of the corresponding materials are assigned according to the position of the grid. Secondly, an excitation current of a given frequency and effective value is applied to the magnetic field intensity induction coil. As for the application of boundary conditions, attention is paid to ensuring a sufficiently large adjacent air area and applying a magnetic field parallel boundary condition at the boundary of the area. At this point, finite element modeling is achieved.

[0093] To calibrate the finite element model, the permeability of each layer in the laminated thin film core is first set to the initial permeability measured in 1). The calibration process is as follows:

[0094] On the one hand, using Figure 7 The equivalent DC inductance is measured by the step voltage method test circuit shown in the figure. Note that the step voltage value should be small enough to avoid saturation, so that the magnetic permeability of the thin film core is equal to the initial magnetic permeability of the magnetic material; on the other hand, DC excitation is applied to the finite element model, and the magnetic permeability of the core is taken as the initial magnetic permeability. The calculated value of the equivalent DC inductance is extracted through finite element simulation. Finally, the measured value and calculated value of the DC inductance are compared, and the geometric parameters of the finite element model are adjusted to make the measured value and the calculated value match with the required engineering accuracy, thereby completing the calibration of the finite element model.

[0095] Among them, the steps of measuring DC inductance by step voltage method are:

[0096] like Figure 7To apply a step voltage to the inductor (secondary open-circuit transformer), the capacitor can be charged by a DC voltage source. Secondly, after the charging is completed, the switch S is closed to apply a specified step voltage excitation to the measured inductor. The current flowing through the inductor is converted into a voltage signal by the sampling resistor and measured. The flux λ(i B ), so the equivalent DC inductance is:

[0097]

[0098] 7) Correction of loss curve

[0099] Substitute the complex permeability extracted in 5) into the finite element model proposed and calibrated in 6). Using this finite element model, the air core coupling main magnetic flux and the actual excitation main magnetic flux flowing through the thin film core can be obtained respectively. In this way, the air core coupling main magnetic flux can be eliminated to obtain the actual corresponding core loss P core The magnetic induction intensity amplitude B m , the loss curve is calibrated.

[0100] Compare the loss curve obtained from the test or the last time with the loss curve obtained from this calibration. If the two cannot meet the accuracy requirements in the least squares sense, use under-relaxation iteration to obtain a new corrected loss curve, that is, for a certain core loss, the magnetic induction intensity is adjusted as follows:

[0101]

[0102] In the above formula is the magnetic induction intensity used in the loss curve before the kth correction, is the magnetic induction intensity used in the loss curve after the kth correction (removing the main magnetic flux coupling effect of the air core), ω r is the relaxation factor used for iteration, that is, to ensure that 1<ω r <2, while is the magnetic induction intensity that will eventually participate in the next iteration. By repeating 5) to 7), the iteration is repeated until the correction of the loss curve is no longer obvious. If the correction amount of the magnetic induction intensity is less than 5%, the measurement process ends. It is worth noting that the finite element model needs to be calibrated before the first iteration, but in subsequent iterations, only the complex magnetic permeability needs to be introduced.

[0103] The entire iterative process of measurement and correction is as follows: Figure 8 shown.

[0104] Example 1 Ring Sample Loss Measurement

[0105] Example 1 Test Ni 44 Fe 55Soft film core, for this purpose, 4 groups of 20 layers of magnetic ring laminated samples with an inner diameter of 10mm and an outer diameter of 15mm were prepared, with a thickness of 6μm. The winding adopts a double-wire winding method, using a copper wire with a diameter of 0.25mm, and the number of turns of the primary and secondary sides is N p =N s =15.

[0106] The test equipment includes Tektronix's AFG3152C multi-channel signal generator, P6022 current probe, TPP0201 voltage probe, TDS2024C four-channel oscilloscope (sampling 1000 points in one cycle), ATA-1200A amplifier from Aigtek, and E4991B impedance analyzer from KEYSIGHT. During the test, 10MHz, 20MHz and 30MHz sinusoidal currents were input to the primary side, and four groups of different current amplitudes were used to test the voltage curve of the secondary side, and then the power loss density curve was calculated and drawn.

[0107] In addition, by calibrating with an air core transformer, the time delay of the current probe relative to the voltage probe under 10MHz, 20MHz, and 30MHz excitations is 5.0ns, 5.2ns, and 5.2ns, respectively.

[0108] According to the geometric parameters of the samples used in the actual measurement, the finite element modeling was realized using Ansoft Maxwell software, such as Fig. 9 shown.

[0109] The loss curve B measured using the two-winding method m ~P core After that, the loss curve is corrected by calibrating the time delay of the current probe relative to the voltage probe. The results of 10MHz are listed in Table 1 below. Correspondingly, for the calibrated finite element model, the loss curve obtained by simulation after substituting the measured complex permeability is also listed in Table 1. At this time, the main magnetic flux coupling effect of the air core has not been eliminated from the finite element model.

[0110] Table 1 Simulation and test B m , P cv Compare

[0111]

[0112] 4 sets of measurement values ​​are obtained based on 4 sets of current excitations with different amplitudes and With four sets of simulation values and The error in the calculation is basically below 1%, which is ideal. Based on this finite element model, the correction to eliminate the main magnetic flux coupling effect of the air core is carried out. The final result is as follows Fig. 9As shown, the magnetic induction intensity is reduced at the same loss due to the removal of the air core coupling effect.

[0113] The present invention proposes a high-frequency loss measurement method suitable for thin-film magnetic cores. Based on the two-winding method, this method uses a high-permeability magnetic film and an insulating layer to form a laminated composite annular structure as a measured magnetic core sample. This new type of magnetic core sample is closer to the magnetic core form and loss characteristics used in actual integrated magnetic components at high frequencies, and can avoid the loss of bulk magnetic material measurement samples being greater than that of thin-film magnetic cores at high frequencies. It is also easy to prepare and has a simple process. On the other hand, by using a reference air core transformer as a calibration element, the phase offset error of the two-winding measurement method can be accurately calibrated within a wider frequency band. Finally, the influence of the main magnetic flux coupling effect of the air core on the loss curve measurement is eliminated by using finite element simulation and super-relaxation iterative algorithm.

[0114] The above descriptions are only some specific implementation methods of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any person familiar with the art within the technical scope disclosed in the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for measuring high-frequency loss of a thin film magnetic core, characterized in that: The following steps are involved: S1 obtains a laminated thin film magnetic core to be tested, wherein the laminated thin film magnetic core to be tested is annular, and uses a vibrating sample magnetometer to measure an initial magnetization curve of a sample of the same magnetic material as the laminated thin film magnetic core to be tested and a hysteresis loop under different external magnetic field strengths; based on the initial magnetization curve, extracts the initial magnetic permeability of the sample close to magnetic neutrality; By connecting the vertices of each hysteresis loop in sequence, an equivalent quasi-static magnetization curve is obtained; S2 builds a two-winding measurement platform: The two-winding method measurement platform comprises: a replaceable annular magnetic core and a magnetic field intensity induction coil and a magnetic induction intensity induction coil uniformly wound on the replaceable annular magnetic core; the shape and size of the replaceable annular magnetic core are the same as those of the laminated thin film magnetic core to be measured; the current flowing through the magnetic field intensity induction coil is converted into a voltage signal via a sampling resistor connected in series thereto, and is measured by one channel of an oscilloscope, while the voltage signals on both sides of the magnetic induction intensity induction coil are measured by another channel of the oscilloscope; The magnetic field intensity induction coil is used as an excitation winding, and its feed source is composed of a signal generator connected to a power amplifier; The S3 uses an air core transformer to calibrate high frequency phase deviation: Get a standard air core transformer of the same shape and size as the replaceable toroidal core, The standard air core transformer is used to replace the replaceable annular magnetic core in the two-winding method measurement platform to obtain the phase deviation required to be compensated for the high-frequency measurement of the thin film core at different frequencies; and a calibrated two-winding method measurement platform is obtained; The specific method for obtaining the phase deviation that needs to be compensated includes the following steps: In the actual measurement of the secondary voltage u of the magnetic induction intensity induction coil B The primary excitation current i of the magnetic field intensity induction coil H Finally, the average air core loss P in one cycle is calculated using the two-winding method. core , the calculation formula is as follows: Because the air core loss should be zero, the air core loss calculated according to formula (1) is approximately zero by adding a time delay to the current curve; the added time delay corresponds to the phase angle required to be adjusted, which is used to calibrate the phase error measured by the system; S4 obtains the loss curve of the laminated thin film core: Loading the laminated thin film magnetic core to be tested obtained in step S1 onto a calibrated two-winding method measurement platform, replacing the standard air core transformer; Use the two synchronous channels of the oscilloscope to measure the induced voltage on both sides of the magnetic induction coil and the excitation current flowing through the magnetic field intensity induction coil, respectively, as the secondary voltage u B With the primary excitation current i H ; Use the phase angle required to adjust the excitation current i obtained by S3 H The phase of the laminated thin film core to be tested is calculated using formula (1); Secondly, use the secondary voltage u B According to Faraday's induction law, the magnetic induction intensity amplitude B in the core is calculated using formula (2): m for: where n B is the number of turns of the magnetic induction coil, S is the cross-sectional area of ​​the laminated film core; according to the average power loss P core and the magnetic induction intensity amplitude B in the core m Obtain loss curve B m ~P core ; S5 builds the finite element model and calibrates it: The calibrated two-winding method measurement platform is modeled using finite element analysis software to obtain a finite element model; the finite element model models the magnetic field intensity induction coil, the magnetic induction intensity induction coil and the laminated thin film magnetic core to be measured in step S1, wherein the geometric dimensions and electrical parameters of each component are consistent with the corresponding components in the calibrated two-winding method measurement platform; After the modeling is completed, the finite element model is calibrated, including the following steps: On the one hand, the DC inductance value of the inductor composed of the magnetic field intensity induction coil and the thin film core is measured by the step voltage method after the magnetic field intensity induction coil is open-circuited, which is used as the measurement value of the equivalent DC inductance; On the other hand, a DC excitation is applied to the finite element model, and the magnetic permeability of the laminated thin film core is taken as the initial magnetic permeability; the calculated value of the equivalent DC inductance is extracted through finite element simulation; Compare the measured value and calculated value of the equivalent DC inductance, and adjust the geometric dimensions of the winding in the finite element model so that the measured value and the calculated value match within the preset engineering accuracy, thereby obtaining a calibrated finite element model; S6 determines the complex magnetic permeability of the laminated thin film core, substitutes the complex magnetic permeability into the finite element model proposed and calibrated in S5, eliminates the influence of the air core coupling main magnetic flux on the measured value of the magnetic induction intensity in the thin film core, and replaces the magnetic induction intensity amplitude B in the core obtained in step S4 m , update the loss curve; S7 iteratively updates the loss curve until the difference between the loss curve obtained after the current iterative update and the loss curve before the current iterative update is no greater than a preset threshold, stops the iterative update and outputs the loss curve obtained by the last iterative update.

2. A method for measuring high-frequency loss of a thin film magnetic core as claimed in claim 1, characterized in that: In step S5, the DC inductance is measured using a step voltage method, which specifically includes the following steps: Constructing a measurement circuit for measuring DC inductance using a step voltage method, the circuit comprising a DC voltage source, a capacitor C connected between two poles of the DC voltage source, a switch, a small sampling resistor R and an equivalent DC inductor L connected in series between two plates of the capacitor C; A step voltage is applied to the equivalent inductor composed of the magnetic field intensity induction coil and the thin film core to be tested. At this time, the magnetic induction intensity induction coil is open; first, the capacitor is charged by a DC voltage source; secondly, after the charging is completed, the switch is closed to apply a specified step voltage excitation to the equivalent DC inductor L; the current i flowing through the equivalent DC inductor L B The small sampling resistor R is used to convert the measured primary voltage into a voltage signal for measurement. The flux λ(i B ), the equivalent DC inductance is calculated as follows:

3. A method for measuring high-frequency loss of a thin film magnetic core as claimed in claim 2, characterized in that: The step S6 specifically comprises the following steps: According to the equivalent quasi-static magnetization curve and loss curve measured by S1, the complex magnetic permeability μ is obtained, and the formula is as follows: μ=μ′-jμ″ (4) The real part μ' is obtained using the equivalent quasi-static magnetization curve; Based on the following formula, the loss curve B measured by S4 is m ~P core Find the imaginary part μ" The obtained complex magnetic permeability μ is substituted into the finite element model proposed and calibrated by S5; By using the finite element model, the actual excitation main magnetic flux flowing through the thin film magnetic core is extracted to eliminate the influence of the air core coupling main magnetic flux on the measured value of the magnetic induction intensity in the thin film magnetic core, and the magnetic induction intensity amplitude B in the magnetic core obtained in step S4 is replaced. m , and update the loss curve.

4. A method for measuring high-frequency loss of a thin film core as claimed in claim 3, characterized in that: The step S7 specifically includes the following steps: Compare the loss curve obtained after this iteration update with the loss curve before this iteration update. If the difference between the two is greater than the preset threshold, the loss curve is updated by under-relaxation iteration, that is, the magnetic induction intensity is adjusted according to the following formula: in is the magnetic induction intensity amplitude in the core used to obtain the loss curve before the kth update, is the magnetic induction intensity amplitude in the core used to obtain the loss curve after the kth update, ω r is the relaxation factor used for iteration, 1<ω r <2, while is the magnetic induction intensity amplitude in the magnetic core used to obtain the loss curve before the k+1th update; If the difference between the loss curve obtained after this iterative update and the loss curve before this iterative update is greater than a preset threshold, steps S6 to S7 are repeated; otherwise, the iteration is terminated and the loss curve obtained by the last iterative update is output.

5. A method for measuring high-frequency loss of a thin film magnetic core as claimed in claim 4, characterized in that: In step S7, the difference between the two is specifically the difference between the two in the sense of least squares.

6. A method for measuring high-frequency loss of a thin film magnetic core as claimed in claim 5, characterized in that: In step S7, The preset threshold is no greater than 5%.

Images