A design method and apparatus for a matrix transformer

Abstract

The application provides a design method and device of a matrix transformer, and belongs to the technical field of transformer design. The method provided by the application comprises the following steps: determining the number of split units according to the power of the transformer; determining the unit row-column arrangement mode according to the number of split units and the packaging length-width ratio; determining the winding connection mode according to the unit row-column arrangement mode; calculating the cross-sectional area of a magnetic core according to Faraday's law of electromagnetic induction, and determining the cross-sectional area of a winding according to the cross-sectional area of the magnetic core; constructing an initial numerical model of the transformer based on the unit row-column arrangement mode, the connection mode, the cross-sectional area of the magnetic core and the cross-sectional area of the winding, simulating the initial numerical model, and optimizing and adjusting the cross-sectional area of the magnetic core and the cross-sectional area of the winding according to the iron loss and the copper loss in the simulation result, so as to complete the design of the matrix transformer. The design method and device of the matrix transformer are used to improve the working efficiency of the transformer, optimize heat dissipation, and ensure high power density and high reliability under the constraint of a given height packaging space.

Description

Technical Field

[0001] This application relates to the field of transformer design technology, and in particular to a design method and apparatus for a matrix transformer. Background Technology

[0002] As a key magnetic component suitable for high-frequency, high-power-density power converters, matrix transformers are becoming increasingly important in modern power supply design, especially in server power supplies, communication equipment power supplies, and on-board chargers for new energy vehicles.

[0003] Current matrix transformer designs mostly employ traditional fixed-specification magnetic cores (such as PQ and RM types). The design process primarily relies on trial and error based on engineers' experience, or only involves local adjustments to single parameters such as the air gap. The design process typically begins by determining the core structure and dimensions, followed by configuring winding parameters and connection methods. Some designs utilize simulation software to verify magnetic flux density and losses to meet basic electrical performance and packaging requirements.

[0004] Traditional fixed magnetic cores have limited adjustable space for the number, cross-sectional area, and arrangement of magnetic pillars, making it difficult to achieve optimal matching with a specific packaging space. This easily leads to uneven magnetic flux distribution and excessively high local magnetic flux density, resulting in increased iron losses and low space utilization. At the same time, existing designs lack a systematic method for simultaneously and collaboratively optimizing the core geometry, spatial arrangement, and winding series-parallel relationships from the perspective of packaging constraints. This results in long design cycles, high iteration costs, and an inability to achieve a globally optimal design for total losses. Summary of the Invention

[0005] In view of this, this application provides a design method and apparatus for a matrix transformer to solve the problems of low space utilization, long design cycle and poor loss performance in transformer design.

[0006] Specifically, this application is implemented through the following technical solution:

[0007] The first aspect of this application provides a design method for a matrix transformer, the method comprising:

[0008] The number of split units is determined based on the transformer's power.

[0009] The row and column arrangement of the units is determined based on the number of split units and the aspect ratio of the package.

[0010] The winding connection method is determined based on the unit row and column arrangement.

[0011] The cross-sectional area of ​​the magnetic core is calculated according to Faraday's law of electromagnetic induction, and the cross-sectional area of ​​the winding is determined based on the cross-sectional area of ​​the magnetic core.

[0012] An initial digital model of the transformer is constructed based on the unit row and column arrangement, the connection method, the core cross-sectional area, and the winding cross-sectional area. The initial digital model is simulated, and the core cross-sectional area and the winding cross-sectional area are optimized and adjusted according to the iron loss and copper loss in the simulation results to complete the design of the matrix transformer.

[0013] A second aspect of this application provides a design apparatus for a matrix transformer, the apparatus comprising a processing module and a design module;

[0014] The processing module is used to determine the number of splitting units based on the power of the transformer;

[0015] The processing module is also used to determine the row and column arrangement of the units based on the number of split units and the aspect ratio of the package.

[0016] The processing module is also used to determine the winding connection method according to the unit row and column arrangement;

[0017] The processing module is also used to calculate the cross-sectional area of ​​the magnetic core according to Faraday's law of electromagnetic induction, and to determine the cross-sectional area of ​​the winding according to the cross-sectional area of ​​the magnetic core.

[0018] The design module is used to construct an initial digital model of the transformer based on the unit row and column arrangement, the connection method, the core cross-sectional area, and the winding cross-sectional area, to simulate the initial digital model, and to optimize and adjust the core cross-sectional area and the winding cross-sectional area according to the iron loss and copper loss in the simulation results, thereby completing the design of the matrix transformer.

[0019] The matrix transformer design method and apparatus provided in this application determine the number of magnetic pillars of the transformer core, the geometric dimensions and spatial arrangement of each magnetic pillar through collaborative optimization, and simultaneously determine the winding method of the primary winding and secondary winding on each magnetic pillar and the series and parallel connection relationship between them, so as to minimize the total loss of the transformer, thereby achieving the purpose of improving efficiency, optimizing heat dissipation, and ensuring high power density and high reliability. Specifically, by introducing the dual constraints of the optimal power range for a single unit and the requirement that the total number of turns on the primary side must be divisible by the number of units when determining the number of split units, power distribution and magnetic circuit symmetry are decoupled. This avoids flux imbalance and increased leakage inductance caused by uneven turns from the source, making the split parameter design more accurate and more in line with the physical nature of high-frequency matrix transformers. By jointly deciding the unit row and column arrangement method based on the number of split units and the package aspect ratio, the empirical limitations of traditional fixed magnetic core arrangement are broken, achieving synchronous optimal matching of space utilization and magnetic circuit symmetry. This effectively eliminates excessive local magnetic flux density caused by improper layout and optimizes the uniformity of magnetic flux distribution. By establishing a winding connection rule of fixed series connection on the primary side and flexible selection of series and parallel connection on the secondary side according to the arrangement and turns ratio, current balance, voltage matching, and convenient wiring are solidified into a reusable design paradigm. This avoids local saturation caused by parallel flow imbalance and significantly reduces leakage inductance and circulating current loss. This approach improves the accuracy and reliability of winding design. By following a step-by-step design process—first calculating the core cross-sectional area using Faraday's law of electromagnetic induction, then determining the window space based on the core cross-sectional area, and finally calibrating the winding cross-sectional area using current density and window space—electromagnetic constraints and structural constraints are decoupled and sequentially coupled. This ensures that the core does not experience iron loss control due to excessive magnetic flux density, while also ensuring that the winding achieves optimal current-carrying capacity within a limited window. This allows the selection of core and winding cross-sectional areas to precisely match the actual physical boundaries, optimizing space utilization and power density. By constructing a simulation iteration closed loop based on iron and copper losses, the core and winding cross-sectional areas are optimized collaboratively. This innovative approach dynamically balances core and winding losses with the unified goal of minimizing total loss, avoiding the performance trade-offs caused by local adjustments of a single parameter in traditional trial-and-error methods. This results in more accurate final design parameters, lower total loss, and optimized overall transformer efficiency and heat dissipation performance. Attached Figure Description

[0020] Figure 1 A flowchart of an embodiment of the matrix transformer design method provided in this application;

[0021] Figure 2 A schematic diagram showing the split of a matrix transformer unit as an exemplary embodiment of this application;

[0022] Figure 3 A schematic diagram of the initial simulation structure shown for an exemplary embodiment of this application;

[0023] Figure 4A schematic diagram of a simulation structure shown as an exemplary embodiment of this application;

[0024] Figure 5 This is a schematic diagram of the structure of a second embodiment of the matrix transformer design device provided in this application. Detailed Implementation

[0025] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0026] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used herein are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0027] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0028] The following specific embodiments are given to illustrate the technical solution of this application in detail.

[0029] Figure 1 This is a flowchart of an embodiment of the matrix transformer design method provided in this application. Please refer to... Figure 1 The method provided in this embodiment may include:

[0030] S101. Determine the number of split units based on the power of the transformer.

[0031] Specifically, the processing power of a single unit is determined based on the rated power of the transformer, and the number of split units is determined based on the ratio of the rated power to the processing power of a single unit; the processing power of a single unit is kept within the range of 1000W-3000W, the total number of primary turns is divisible by the number of split units, and the number of split units does not exceed 6.

[0032] Optionally, the power of a single unit is greater than or equal to 1000W and less than or equal to 1500W.

[0033] Furthermore, with the goals of power distribution, magnetic flux balance, and structural feasibility, power distribution requires that the total rated power of the transformer be evenly distributed among the various split units, ensuring that the power handled by each unit falls within the preferred range of 1000W to 1500W. This guarantees consistent heat generation across all units, preventing uneven distribution where some units experience localized overheating while others have idle cooling capacity. Magnetic flux balance requires that the total number of turns in the primary winding be divisible by the number of split units, ensuring that the number of primary turns connected in series with each magnetic post is identical. This generates a uniformly distributed magnetic flux in the symmetrical magnetic circuit, fundamentally eliminating magnetic flux imbalance, localized saturation, and increased leakage inductance caused by differences in the number of turns. Structural feasibility requires that the final number of split units not exceed 6, and that they be arranged in a symmetrical row and column matrix according to the aspect ratio of the packaging space. This avoids isolated units that cannot be incorporated into the regular layout, ensuring simple winding interconnections, feasible assembly processes, and maximized space utilization. These three objectives are mutually constraining and indispensable; only when all three are met can a number of modular units that satisfy both electrical performance requirements and physical manufacturability be determined. Due to limitations in heat dissipation and manufacturing processes, high-power PCB transformers should not exceed 3000W per column; therefore, a preferred range of 1000-1500W is used for modularization. Simultaneously, the number of primary-side turns must be divisible by the number of modular units to avoid flux imbalance and increased leakage inductance across multiple columns; the number of units should not exceed 6 to ensure a compact structure and feasible assembly. It should be noted that the number of primary-side turns must be divisible by the number of modular units; different units cannot have different numbers of primary-side turns. Based on Faraday's law of electromagnetic induction and Ohm's law for magnetic circuits, the unbalanced flux induced in the circuit when the number of winding turns in each unit is unequal is calculated. The constraint that the core's working flux density caused by the unbalanced flux must be lower than the saturation flux density limit is used to determine the mandatory design criterion that the total number of primary-side turns must be divisible by the number of modular units.

[0034] Furthermore, the transformer's power rating refers to the apparent or active power that the transformer must continuously transmit under rated operating conditions. Its value is determined by the power system architecture, load characteristics, and converter topology, and is explicitly given as a design input. Design requirements include not only this rated power value but also peak power, short-time overload multiple, maximum ambient temperature, heat dissipation method (such as natural air cooling or liquid cooling), and absolute height limitations of the packaging space. These requirements collectively constitute the boundary conditions for subsequent verification, with packaging height being the most stringent constraint because it directly limits the maximum thickness of the magnetic core and the usable window height of the windings.

[0035] Furthermore, the peak power or short-term overload condition that the transformer needs to withstand is obtained, which is defined by system-level requirements. Based on the peak power or overload multiple, the transient power that each split unit needs to handle under abnormal conditions is recalculated. This transient power is compared with the upper limit of the preferred power range for a single unit set in step one. If the transient power exceeds the upper limit, it is determined that the current number of split units is insufficient, and the power stress per column needs to be reduced by increasing the number of split units. Based on the adjusted number of split units, the total number of turns on the original side is re-verified to ensure divisibility, until an optimal number of split units is obtained that simultaneously meets the requirements of rated operating condition thermal balance, abnormal operating condition anti-saturation capability, and magnetic circuit symmetry.

[0036] By incorporating the passability verification for abnormal operating conditions into the parametric design stage and forming a closed loop with the decision on the number of split units, peak power is proactively used as an independent driving variable to determine the number of split units. By strictly constraining the power of a single column under abnormal operating conditions within the safety boundary, the core is fundamentally guaranteed to stay away from the saturation region under any permissible load conditions, eliminating the risk of iron loss control and thermal collapse caused by excessive magnetic flux density. At the same time, the empirical concept of "power distribution margin" is solidified into a quantifiable and iterative design step: by proactively increasing the number of units to distribute overload stress, it is equivalent to injecting inherent overload robustness into the transformer, provided that physical space allows.

[0037] Furthermore, the steps for determining the number of splitting units based on the transformer's power include:

[0038] (1) Determine the power of a single unit based on design requirements;

[0039] Specifically, the overall rated power, target operating frequency, and geometric constraints of the package space of the matrix transformer are obtained; based on the geometric constraints of the package space, combined with the current carrying capacity and heat dissipation conditions of the PCB windings, the rated power of a single split unit is set.

[0040] Furthermore, obtaining overall rated power and package size constraints is to clarify the electrical and physical boundaries of the entire design task. Within the highly constrained flat package space, traditional single large magnetic core structures are difficult to match and must be broken down into multiple flat small units. When setting the power of a single unit, the upper limit of 1500W is determined by heat dissipation capacity. When the power handled by a single magnetic core is too high, the heat generated by the iron losses of the core and the copper losses of the windings cannot be effectively dissipated in the compact space, forming local hot spots. This leads to a decrease in core permeability, an increase in winding resistance, and accelerated aging of insulation materials, potentially causing thermal failure in severe cases. The lower limit of 1000W is determined by structural efficiency and power density: if the unit power is too small, the number of units must be increased to meet the total power requirement. Too many units will lead to a decrease in the total cross-sectional area utilization of the core, excessively long copper foil traces for interconnection between units, a significant increase in AC losses and leakage inductance, and a sharp increase in assembly complexity. Therefore, anchoring the power of a single unit within this range is the optimal engineering balance point found between thermal management, electrical performance, power density, and manufacturing complexity.

[0041] (2) Determine the number of split units based on the ratio of the power of the transformer to the power of the individual unit.

[0042] Specifically, the overall rated power of the transformer is divided by the rated power of a single splitting unit to calculate the theoretical ratio; the theoretical ratio is rounded down to obtain an initial number of splitting units; the total number of turns of the transformer's primary winding is obtained, and it is checked whether the total number of turns is divisible by the initial number of splitting units; if it is not divisible, the initial number of splitting units is adjusted by increasing or decreasing the value until an integer that is divisible by the total number of turns of the primary winding is found; finally, the integer that satisfies the divisibility relationship is determined as the number of splitting units.

[0043] Furthermore, the overall rated power of the transformer is divided by the power of the determined individual unit to obtain a theoretical ratio. If this ratio is not an integer, it needs to be rounded up or down. A key electrical constraint is introduced during this process: the total number of primary winding turns must be divisible by the final determined number of split units. This is to ensure the symmetry of the multi-pillar magnetic circuit system. According to the principle of magnetic circuits, when the primary windings of the matrix transformer are connected in series on each pillar and have the same number of turns, the magnetomotive force generated by each pillar during excitation will be completely consistent, thus ensuring the uniform distribution of magnetic flux throughout the core structure. If a unit number that cannot be divided by the total number of primary winding turns is chosen to accommodate the power ratio, the number of turns in the primary windings of different units will not be absolutely equal. This difference in the number of turns will generate magnetic flux components that cannot be completely canceled out in parallel or series magnetic circuits, leading to higher magnetic flux density in some pillars and saturation, while significantly increasing leakage inductance and severely degrading transformer performance. Therefore, when adjusting the number of units after rounding, it is necessary to repeatedly verify its divisibility with the total number of turns on the original side, and ensure that the final number of units does not exceed the upper limit of structural feasibility, such as six. The final number of split units is an optimized parameter that simultaneously satisfies the requirements of power thermal management range, magnetic circuit symmetry principle, and structural compactness.

[0044] S102. Determine the row and column arrangement of the units based on the number of split units and the aspect ratio of the package.

[0045] Specifically, the number of rows and columns of the units is determined based on the number of subdivided units, and the arrangement of the unit rows and columns is determined based on the aspect ratio of the package. The arrangement of the unit rows and columns refers to the combination of the number of rows and columns of each subdivided unit in the matrix transformer on the plane, which is used to determine the layout of all magnetic column units within the package space.

[0046] Furthermore, the layout needs to match the proportion of the packaging space to maximize space utilization. A single-row layout is suitable for long and narrow spaces; 2×2 / 2×3 is suitable for near-square or flat spaces. When the length is significantly longer, use 2×3 to reduce the height, improve heat dissipation and magnetic flux uniformity, and avoid excessively high local magnetic flux density.

[0047] Furthermore, the steps for determining the row and column arrangement of the units based on the number of split units and the aspect ratio of the package include:

[0048] When the number of split units is a first value, the unit row and column arrangement is determined to be a row arrangement;

[0049] When the number of split units is the second value, the unit row and column arrangement is determined to be one row of four columns or two rows of two columns;

[0050] When the number of split units is the third value, the arrangement of the unit rows and columns is determined to be either one row of six columns or two rows of three columns.

[0051] Specifically, since the first value of units is difficult to divide into a regular matrix with multiple rows and columns, or dividing it into multiple rows would lead to a decrease in package space utilization and asymmetry in magnetic flux paths, a single-row arrangement is uniformly adopted. That is, all units are arranged in a row along the length direction, which simplifies the structure, ensures uniform magnetic flux, facilitates winding routing, and can adapt to narrow and elongated package spaces. The second value refers to the case where the number of units is 4. The 4 units can be arranged in a narrow form of one row and four columns (1×4), or in a square form of two rows and two columns (2×2). The specific choice depends on the length-to-width ratio of the package: when the package length is approximately twice the width, one row and four columns are used; when the package length and width are close, two rows and two columns are used. Both arrangements can ensure unit symmetry, uniform magnetic flux, and optimal window space utilization, while meeting the layout and heat dissipation requirements of the PCB windings.

[0052] Furthermore, the third value refers to the case where the number of split units is 6. The 6 units can be arranged in a neat row of six columns (1×6) or two rows of three columns (2×3): when the package length is much greater than the width (approximately 3 times the width or more), a row of six columns is used; when the package is a flat rectangle with a length approximately twice the width, two rows of three columns are used. The two rows of three columns can achieve a more uniform magnetic flux distribution, lower leakage inductance, and better heat dissipation within a limited height, and is the most commonly used layout for high-power matrix transformers. Figure 2 A schematic diagram showing the disassembled matrix transformer unit as an exemplary embodiment of this application is provided below. Figure 2 When the number of split units is 6, the magnetic column units are arranged in a two-row, three-column configuration.

[0053] It should be noted that the first and second values ​​are less than the third value, and the first value can be greater than or less than the second value. For example, in one embodiment, the first value is 1, 2, 3, and 5, the second value is 4, and the third value is 6.

[0054] Furthermore, when the number of split units is 1-3 or 5, they are all arranged in one row without any other special layout. When the number of split units is 4, they can be set in one or two rows, i.e., 2*2 or 1*4. The more suitable layout can be selected according to the package size. If the length and width are close, choose 2*2; if the length is about twice the width, consider a 1*4 layout. When the number of split units is 6, they can be set in one or two rows, i.e., 2*3 or 1*6. The more suitable layout can be selected according to the package size. If the length and width are close, choose 2*2; if the length is about three times the width, consider a 1*6 layout. Based on the current package size requirements, the length of the transformer is more than three times the width, so a 2*3 layout is considered.

[0055] When determining the cell row and column arrangement, the number of rows is determined based on the aspect ratio constraint of the package space. Using the geometric dimensions of the package space as input, the space utilization and magnetic circuit symmetry indices under different row number arrangements are calculated. The goal is to maximize space utilization and minimize the difference in magnetic path length among the magnetic pillars. By comparing the fill rate and magnetic flux path symmetry within the same package boundary between a one-row arrangement and a two-row arrangement, the optimal number of rows is determined. When the package length is significantly greater than the width, exhibiting a long and narrow shape, the space constraint forces the optimal number of rows to be one. When the package length and width are close, exhibiting a square or flat rectangle shape, a two-row arrangement achieves a higher space fill rate and a more uniform magnetic flux distribution; therefore, the optimal number of rows is two. After determining the optimal number of rows, the total number of split cells is a known value, and the number of columns is directly obtained by dividing the total number by the number of rows, without requiring independent decision-making. Thus, the cell row and column arrangement is completely determined as an optimal matrix form that satisfies the triple constraints of space matching, magnetic circuit symmetry, and structural regularity.

[0056] S103. Determine the winding connection method according to the unit row and column arrangement.

[0057] Specifically, a matrix transformer is composed of multiple independent, structurally identical magnetic core units arranged in a regular row-column matrix. Each unit contains a magnetic column and primary and secondary windings wound around it. All units share the same encapsulation space, but their magnetic circuits are independent of each other. The electromagnetic relationship inside the transformer is a multi-circuit discrete coupling. The excitation state, induced voltage, and current distribution of each unit depend on the selection of the external electrical connection method between the windings of each unit. The winding connection method directly affects three key performance aspects: firstly, it affects current distribution and magnetic flux balance; the series and parallel connection method of the primary winding determines the current flow. First, the current in each magnet winding must be forced to be equal. If parallel connection is used, the impedance difference between units may lead to imbalance in power distribution, magnetic flux bias in some magnets, or even saturation. Second, it affects voltage superposition and power synthesis. The series-parallel combination of the secondary windings determines whether the induced electromotive force of each unit is accumulated or shunt, directly affecting whether the output voltage level and load capacity match the design specifications. Third, it affects loss distribution and heat dissipation uniformity. An unreasonable connection method will cause uneven distribution of copper and iron losses in each unit, resulting in local overheating and insufficient utilization of the overall heat dissipation capacity, reducing the reliability of the transformer. Therefore, after determining the physical layout of the unit rows and columns, it is necessary to design a matching winding connection method to achieve balanced and controllable current, voltage, and losses among multiple discrete units, ensuring that the matrix transformer can still operate stably and efficiently like an ideal single transformer in a distributed structure.

[0058] Furthermore, when determining the layout and winding connection methods, the number of split units and the aspect ratio of the encapsulation space are used as input variables. All feasible and regular row and column layout schemes are enumerated. For each layout scheme, two indicators are calculated: space utilization and magnetic circuit symmetry. Space utilization refers to the proportion of the sum of the cross-sectional areas of each magnetic column in the projected area of ​​the encapsulation plane, reflecting the compactness of the structure. Magnetic circuit symmetry is calculated by calculating the variance of the distance from the center of each magnetic column to the geometric center of the transformer. The smaller the variance, the more uniform the magnetic circuit distribution. With the goal of maximizing space utilization and optimizing magnetic circuit symmetry, the optimal layout scheme is selected from the enumerated schemes. After the layout scheme is determined, the superposition result of the induced electromotive force of each branch and the balance of current distribution are further calculated when different series and parallel topologies are used between the units in the secondary winding. The induced electromotive force superposition result is obtained by vector summation of the induced voltages on the secondary sides of each unit according to the series-parallel relationship. It is necessary to ensure that the synthesized output voltage is consistent with the design target value. The current distribution balance is calculated by constructing an equivalent circuit model of the winding resistance and leakage inductance of each unit, calculating the current deviation value of each parallel branch under different connection topologies, and determining the final secondary winding connection topology with the minimum current deviation value as the objective. In this way, the determination of the arrangement and winding connection method is no longer a simple matching of preset rules, but rather the optimal values ​​obtained through quantitative calculations with the objectives of spatial-magnetic circuit joint optimization and voltage-current balance optimization, respectively.

[0059] Furthermore, the primary windings are connected in series between each unit; when arranged in one row, if the number of turns of the secondary winding is less than the number of units, they are connected in parallel, and if they are greater, they are connected in series; when arranged in two rows, the secondary windings of units in the same column are connected in series, and those in different columns are connected in parallel.

[0060] Furthermore, the primary windings must be connected in series to avoid imbalances in voltage distribution, uneven magnetic flux, and local saturation caused by parallel connections. The secondary windings are connected in series or parallel according to their arrangement and number of turns to ensure voltage balance, simple wiring, and minimal circulating current. A two-row structure with series followed by parallel connection improves withstand voltage, balances current, and is suitable for high-power output.

[0061] Furthermore, the winding connection method is determined based on the unit row and column arrangement, including:

[0062] The primary windings are connected in series between each unit;

[0063] When the unit row and column arrangement is in a row, the number of turns of the secondary winding and the number of split units are compared. If the number of turns is less than the number of split units, the secondary winding is connected in parallel; if the number of turns is greater than the number of split units, the secondary winding is connected in series.

[0064] When the unit is arranged in two rows, the secondary windings of the units in the same column are connected in series, and the secondary windings of the units in different columns are connected in parallel.

[0065] Specifically, the primary windings of the matrix transformer are connected in series between each split unit. This series connection forces the current flowing through each unit to be completely uniform, preventing current unevenness and magnetic flux imbalance caused by differences in the impedance of each core / winding. This, in turn, prevents local core saturation, increased leakage inductance, and increased losses. The series structure ensures magnetic circuit symmetry and uniform magnetic flux, making it the most stable primary winding connection method for multi-unit matrix transformers.

[0066] Furthermore, when the units are arranged in a row, the number of turns in the secondary winding is compared with the number of split units: if the number of turns is less than the number of units, the secondary winding is connected in parallel; if the number of turns is greater than the number of units, the secondary winding is connected in series. When the number of turns is less than the number of units, parallel connection can reduce single-path current, improve current sharing, and simplify wiring, making it suitable for high-current, low-voltage output; when the number of turns is greater than the number of units, series connection can increase output voltage and reduce current stress, making it suitable for high-voltage output. In addition, voltage and current levels can be flexibly matched according to electrical specifications to ensure simple wiring and low circulating current.

[0067] Furthermore, when the unit rows and columns are arranged in two rows, the secondary windings of the two units in the same column are connected in series, and the series groups between different columns are then connected in parallel. Series connection in the same column allows the voltages of the upper and lower units to be superimposed, matching the required output voltage; parallel connection in different columns allows for balanced current distribution, improving load capacity and heat dissipation uniformity. This connection method has optimal symmetry and best magnetic flux balance, which can significantly reduce leakage inductance and ripple, making it suitable for high-power, high-reliability applications.

[0068] Using the number of split units and the aspect ratio of the packaged space as dynamic input variables, all geometrically feasible row and column arrangement schemes under the current conditions are enumerated to form a candidate set of arrangement schemes. For each arrangement in this candidate set, three quantifiable evaluation indicators are calculated: space utilization, average trace length from each magnetic post to the winding lead-out end, and magnetic circuit symmetry. Among them, space utilization represents the compactness of the structure, average trace length directly affects AC loss, and magnetic circuit symmetry is defined by the standard deviation of the magnetic circuit length of each magnetic post; the smaller the standard deviation, the more uniform the magnetic flux distribution. These three indicators are weighted and normalized into a comprehensive fitness function. With the goal of maximizing the comprehensive fitness function, the optimal row and column arrangement is searched and determined from the candidate set. Based on this, using the optimal arrangement as the structural boundary condition, the series and parallel topology between the secondary winding units is further abstracted into a variable electrical connection diagram. For all feasible connection topology schemes, the deviation between the total output voltage of the secondary side and the design target value, as well as the imbalance of the current in each parallel branch, are calculated for each topology. The current imbalance is calculated by establishing an equivalent circuit model that takes into account differences in winding resistance and leakage inductance coupling, and then solving for the root mean square deviation of the current in each branch. With the joint optimization objective of minimizing the absolute value of the voltage deviation and the current imbalance, the optimal winding connection method matching the current arrangement is dynamically selected from the candidate connection topologies. Thus, based on real-time input variables, and through calculation using quantifiable indicators and multi-objective optimization, dynamic optimization of the winding connection method is achieved.

[0069] Furthermore, the primary side needs to be connected in series in each unit. According to Kirchhoff's current law and Ohm's law, if the primary side is connected in parallel, the different impedances in different units may lead to a certain degree of current sharing problem, and even affect the magnetic flux distribution at different positions of the magnetic core, which may lead to local saturation. If the units are arranged in one row and the number of turns on the secondary side is less than the number of split units, the secondary side is connected in parallel in each unit. At this time, the secondary side of each unit does not need to be connected, and the wiring method is more convenient. If the units are arranged in one row and the number of turns on the secondary side is greater than the number of split units, the secondary side is connected in series in each unit. If the units are arranged in two rows, the secondary side needs to be connected in series with the same number of turns in each column, and then connected in parallel in each row. For example, if the secondary side has 2 turns and the number of split units is 6, then all six units have 3 turns. The first row and first column are connected in series with the second row and first column to form 6 turns, and then the three columns are connected in parallel.

[0070] By connecting the primary windings in series among the units, problems such as loss balancing and uneven flux distribution caused by impedance differences among units can be avoided, preventing local core saturation and reducing leakage inductance. For secondary windings arranged in a single row, series or parallel connection can be selected according to the relationship between the number of turns and the number of units, which simplifies wiring and improves the convenience of outgoing lines while ensuring matching of output electrical characteristics. For two-row arrangements, a connection method of series connection of secondary windings in the same column and parallel connection of secondary windings in different columns can achieve balanced current distribution and reasonable voltage matching, further improving the electrical symmetry and heat dissipation uniformity of the windings. The overall connection strategy and unit arrangement are synergistically adapted to effectively reduce transformer losses, improve power density and operational reliability.

[0071] S104. Calculate the cross-sectional area of ​​the magnetic core according to Faraday's law of electromagnetic induction, and determine the cross-sectional area of ​​the winding according to the cross-sectional area of ​​the magnetic core.

[0072] Optionally, when calculating the cross-sectional area of ​​the magnetic core, the waveform coefficient is 4.44 and the B value is 0.2-0.25T.

[0073] Furthermore, the maximum voltage amplitude that the transformer primary winding must withstand under normal operating conditions is obtained. This maximum voltage amplitude is determined by the converter topology and the DC bus voltage. Simultaneously, the transformer's operating frequency and the number of turns in the primary winding are determined. The operating frequency is the system switching frequency, and the number of turns is derived from the previously determined number of split units and the total number of turns constraint. Furthermore, the waveform coefficient is used to characterize the influence of the excitation voltage waveform on the rate of change of magnetic flux. Its value depends on the converter's operating waveform type. When a standard sine wave voltage is applied to the transformer primary, the waveform coefficient is 4.44, derived from the product of the ratio of the effective value to the average value of the sine wave and the frequency. When the primary excitation is a square wave or rectangular wave, the waveform coefficient is 4.0, corresponding to the ratio of the effective value to the average value of the square wave. The waveform coefficient is directly substituted into the calculation formula of Faraday's law of electromagnetic induction, allowing the relationship between voltage and the rate of change of magnetic flux to be accurately calculated based on the actual excitation waveform.

[0074] Furthermore, the B-value refers to the maximum operating magnetic flux density of the magnetic core, expressed in Tesla (T). It is taken from the safe operating range below the saturation flux density limit of the core material, typically set within the range of 0.2T to 0.25T based on the characteristics of the selected core material. The B-value, as the upper limit constraint of the core's operating point, is placed in the denominator of the calculation formula. Its physical meaning is to ensure that, under rated voltage and frequency conditions, the peak value of the alternating magnetic flux density inside the core does not enter the nonlinear saturation region of the material. Substituting the voltage amplitude, waveform coefficient, operating frequency, number of primary turns, and B-value into the derivation of Faraday's law of electromagnetic induction, the minimum cross-sectional area required for the core can be calculated. The calculated minimum cross-sectional area directly determines the geometric dimensions of the magnetic core, thus affecting the iron loss level, window space, and the available range of subsequent winding cross-sectional areas.

[0075] Furthermore, while the theoretical value of the magnetic core cross-sectional area is calculated using Faraday's law of electromagnetic induction, its actual effective working capacity is also affected by the loss characteristics of the core material at high frequencies, the constraints of heat dissipation on temperature rise, and the non-uniform magnetic flux distribution effect caused by the core geometry. As the operating frequency increases, the eddy current loss and residual loss of the core material increase non-linearly. Even if the peak magnetic flux density does not exceed the B-value limit, excessively high loss density may lead to thermal failure. Therefore, the cross-sectional area needs to be corrected according to the loss curve of the selected core material to keep the unit volume loss within the allowable range. Furthermore, within the highly confined packaging space, the heat dissipation of the magnetic core... The relationship between area and cross-sectional area is not a simple linear one. While increasing the cross-sectional area can reduce magnetic flux density and thus reduce iron loss, it will also reduce the surface area ratio available for air convection or thermal interface material contact. This may lead to an increase in the internal temperature of the magnetic core and a further reduction in the B-value limit, forming a negative feedback loop. The actual cross-sectional shape of the magnetic core will result in the magnetic flux not being completely uniformly distributed within the cross-section. Local magnetic flux density may appear at corners or edges. The peak value of this local magnetic flux density is often 10% to 20% higher than the average magnetic flux density. If only the average magnetic flux density is designed, these local areas will saturate first. Therefore, a magnetic flux distribution non-uniformity coefficient needs to be introduced to correct the theoretical cross-sectional area.

[0076] Furthermore, to accurately calculate the final core cross-sectional area, the theoretical initial cross-sectional area can be calculated by substituting the waveform coefficient and B value based on Faraday's law of electromagnetic induction. A three-dimensional electromagnetic field finite element model of the core is constructed, and the theoretical cross-sectional area is used as the initial geometric parameter input to the model. The magnetic flux density cloud map distribution inside the core is obtained through simulation, and the ratio of the maximum local magnetic flux density to the average magnetic flux density is extracted as the non-uniformity correction coefficient. At the same time, the predicted core temperature rise under this cross-sectional area is calculated based on the loss density curve of the core material and the encapsulation thermal resistance network to determine whether the temperature rise exceeds the allowable limit. If the local magnetic flux density exceeds the upper limit of the B value or the temperature rise exceeds the limit, the cross-sectional area is adjusted according to the correction coefficient and the simulation is repeated. After iterative convergence, the accurate core cross-sectional area that simultaneously satisfies the constraints of average magnetic flux density, local magnetic flux density unsaturation, and iron loss temperature rise is obtained.

[0077] Specifically, Figure 3 This is a schematic diagram of the initial simulation structure shown as an exemplary embodiment of this application. Figure 4 The simulation structure diagram shown is an exemplary embodiment of this application. Please refer to... Figure 3 and Figure 4 . Figure 3 contrast Figure 4 The winding section is simplified, resulting in a simple model and short simulation time during finite element simulation. Furthermore, the simplified winding has minimal impact on the core's operating state, which can be ignored, allowing for rapid simulation and iterative optimization. Figure 4To ensure that the digital model is drawn entirely based on the final PCB layout design and to accurately reflect the actual loss situation, the final simulation is performed after the basic design parameters are determined, and the effectiveness of comparing the simulation with the measured loss is verified.

[0078] The core cross-sectional area is calculated based on Faraday's law of electromagnetic induction, operating frequency, number of turns, and magnetic flux density. The winding window space is determined by the core cross-sectional area and package size. The initial winding cross-sectional area is calculated based on the operating current and PCB winding current density. The winding cross-sectional area is then adjusted using the window space to obtain the final winding cross-sectional area.

[0079] Furthermore, the cross-sectional area of ​​the magnetic core is calculated and determined in advance by the electromagnetic constraints, which determines the withstand voltage, magnetic flux density, and iron loss. The cross-sectional area of ​​the winding depends on the window space after the magnetic core is determined and cannot be calculated independently in advance. The flat structure of the PCB has good heat dissipation, and the current density can be taken as 10A / mm². The copper area is reasonably allocated within the window to control copper loss and temperature rise.

[0080] Furthermore, the core cross-sectional area was calculated according to Faraday's law of electromagnetic induction, with a waveform coefficient of 4.44 and a B value of 0.2-0.25T. The previously mentioned turn count requirements and arrangement were also considered. A preliminary digital model of the core was drawn based on the calculated core cross-sectional area and the required package. Ignoring current density, a preliminary finished digital model was completed based on the designed winding arrangement. Ansys Maxwell was used to simulate the preliminary design, focusing on the overall magnetic flux density distribution of the core. The magnetic flux density could be adjusted by changing the core cross-sectional area at corresponding positions. The initial AE of the transformer was initially selected as 450 mm². However, simulation iterations revealed that the magnetic flux density was too high, and the losses could not be met. Therefore, the core cross-sectional area was further adjusted to a reasonable range.

[0081] Furthermore, determining the winding cross-sectional area based on the magnetic core cross-sectional area includes:

[0082] (1) Determine the window space of the winding based on the magnetic core area;

[0083] Specifically, the cross-sectional area of ​​the magnetic core determines the space occupied by the magnetic column body. Under a fixed package size, once the magnetic core area is determined, the remaining area that can be used to arrange the winding copper foil is the winding window space. The window space is the maximum physical range that the winding can be arranged, which directly limits the upper limit of the winding cross-sectional area.

[0084] (2) Calculate the initial cross-sectional area of ​​the winding based on the transformer operating current and winding current density;

[0085] Specifically, the initial cross-sectional area of ​​the winding is calculated jointly by the operating current and the allowable current density. The purpose is to ensure that the temperature rise of the winding is reasonable and the copper loss is controllable under the rated current. The current density value is determined based on the heat dissipation capacity of the PCB's flat structure, providing an electrical constraint on the winding cross-sectional area. For details on the calculation process of the initial cross-sectional area of ​​the winding, please refer to the description in the relevant technical documents; it will not be repeated here.

[0086] (3) Adjust the initial cross-sectional area of ​​the winding according to the window space to obtain the cross-sectional area of ​​the winding.

[0087] Specifically, the initial cross-sectional area calculated theoretically is matched and calibrated with the actual available window space. The final winding cross-sectional area is determined without exceeding the window space, so that the winding can meet the requirements of current carrying and heat dissipation, and can be reasonably arranged within the package, thus achieving a coordinated match between electrical performance and structural space.

[0088] Furthermore, within the fixed package outline, after deducting the core volume occupied by all magnetic pillars from the total available volume, the remaining three-dimensional area is the window space. Its geometry directly defines the maximum cross-sectional area boundary of the winding copper foil. The initial cross-sectional area of ​​the winding is calculated based on the rated operating current of the transformer and the allowable current density of the PCB winding. The operating current comes from the output power and voltage specifications of the converter, and the allowable current density is determined based on the PCB copper thickness, heat dissipation airflow, and allowable temperature rise. The two variables mentioned above represent structural space constraints and electrical current-carrying constraints, respectively. The initial cross-sectional area is compared with the maximum winding cross-sectional area that the window space can accommodate. If the initial cross-sectional area exceeds the capacity of the window space, the cross-sectional area must be compressed and adjusted. However, compression will lead to increased copper losses and temperature rise. At this time, it is necessary to recheck whether the current density exceeds the safety threshold. If the initial cross-sectional area is much smaller than the window space, it means that the space is not fully utilized. The winding cross-sectional area can be appropriately increased to reduce copper losses. However, after increasing the area, it is necessary to simultaneously check whether the insulation creepage distance or ventilation and heat dissipation channel is encroached. After closed-loop iteration, the obtained winding cross-sectional area is a precise value that achieves a balance under the four constraints of space boundary, current-carrying capacity, temperature rise limit and insulation safety.

[0089] By first determining the window space based on the core area, then calculating the initial cross-sectional area according to electrical parameters, and finally calibrating and adjusting in conjunction with the physical space, the winding cross-sectional area can achieve an optimal balance between electrical and structural constraints. This ensures the winding's current-carrying capacity, reduces copper losses and temperature rise, and fully utilizes the packaging space to increase the transformer's power density.

[0090] S105. Based on the unit row and column arrangement, the connection method, the core cross-sectional area, and the winding cross-sectional area, construct an initial digital model of the transformer, simulate the initial digital model, and optimize and adjust the core cross-sectional area and the winding cross-sectional area according to the iron loss and copper loss in the simulation results to complete the design of the matrix transformer.

[0091] Specifically, an initial digital model is constructed based on the arrangement, connection method, core cross-sectional area, and winding cross-sectional area; electromagnetic field simulation is performed to obtain iron loss and copper loss. If the iron loss is too large, the core cross-sectional area is increased; if the copper loss is too large, the winding cross-sectional area is increased; iterative optimization is performed until the total loss is minimized, thus completing the matrix transformer design.

[0092] Furthermore, a complete 3D digital model was constructed, and simulations were used to verify the distribution of magnetic flux, current, and losses. Iron losses are dominated by the cross-sectional area of ​​the magnetic core, while copper losses are dominated by the cross-sectional area of ​​the winding; the two are inversely related. The total loss was minimized through iteration, ensuring optimal efficiency, heat dissipation, and reliability under strict packaging constraints.

[0093] Furthermore, the initial digital model is simulated, and the cross-sectional area of ​​the magnetic core and the cross-sectional area of ​​the winding are optimized and adjusted based on the iron loss and copper loss in the simulation results, including:

[0094] (1) Calculate the total simulated loss based on iron loss and copper loss;

[0095] Specifically, the iron loss and copper loss of the initial digital model are obtained through electromagnetic field simulation, and the two are added together to obtain the total transformer loss. The iron loss is determined by the cross-sectional area of ​​the magnetic core and the magnetic flux density, while the copper loss is determined by the cross-sectional area of ​​the winding and the current density. The total loss is used to intuitively evaluate the merits of the current design scheme.

[0096] Furthermore, iron loss refers to the sum of hysteresis loss and eddy current loss generated by the transformer core material under the action of an alternating magnetic field. It is mainly determined by the core cross-sectional area, magnetic flux density, and operating frequency. The smaller the core cross-sectional area and the higher the magnetic flux density, the greater the iron loss, which is one of the main factors leading to transformer heating and reduced efficiency. Copper loss refers to the Joule loss generated by the winding's own resistance when current flows through the transformer winding. It is mainly related to the winding cross-sectional area and the magnitude of the operating current. The smaller the winding cross-sectional area and the higher the current density, the greater the copper loss, which is also a key source of transformer temperature rise and power loss.

[0097] (2) Adjust the cross-sectional area of ​​the magnetic core according to the iron loss, and adjust the cross-sectional area of ​​the winding according to the copper loss;

[0098] Specifically, when the iron loss is too high in the simulation results, the cross-sectional area of ​​the magnetic core is increased to reduce the magnetic flux density, thereby reducing the iron loss; when the copper loss is too high, the cross-sectional area of ​​the winding is increased to reduce the current density, thereby reducing the copper loss.

[0099] (3) Recalculate the total simulation loss based on the adjusted core cross-sectional area and winding cross-sectional area, and iteratively optimize the total simulation loss until the loss condition is met.

[0100] Specifically, the updated core cross-sectional area and winding cross-sectional area are substituted into the digital model, and simulation is performed again to calculate the new total loss. The process of "simulation-judgment-adjustment-resimulation" is repeated until the total loss drops to the preset target value or reaches the minimum stable value. Through iterative optimization, the global optimum of the total loss can be achieved under the packaging constraints.

[0101] Furthermore, the simulation space utilizes the periodic symmetry of the matrix transformer unit structure to construct only a sub-model of the smallest symmetric unit for simulation, significantly reducing the solution domain from the full model to the sub-domain. This significantly reduces the number of finite element meshes and the degrees of freedom, directly improving the simulation speed. The simulation objective is clearly to minimize the total loss. Iron loss is obtained by substituting the magnetic flux density amplitude of each mesh unit in the core sub-domain into the Steinmetz equation for volume integral summation. Copper loss is calculated by volume integral of the product of the square of the current density and resistivity in the winding region. The simulation accuracy is significantly improved by using a loss integral calculation method based on the actual field distribution rather than an engineering estimation formula. In each simulation iteration, the adjustment amount of the core cross-sectional area and the winding cross-sectional area is calculated by scaling the total loss at the current design point according to the gradient sensitivity of the cross-sectional area. Parameters with higher sensitivity are given smaller step sizes for fine optimization, while parameters with lower sensitivity are allowed larger step sizes to quickly cross flat regions. This allows the iteration path to converge to the vicinity of the optimal value as quickly as possible.

[0102] By optimizing the closed-loop iterative process of iron loss, copper loss and total loss, dynamic matching of core cross-sectional area and winding cross-sectional area is achieved, which can quickly converge to the optimal structural parameters, effectively reduce the overall loss of transformer, improve conversion efficiency, avoid the blindness of traditional experience design, significantly shorten the design cycle and improve the reliability of the scheme.

[0103] Furthermore, the method provided in this embodiment also includes:

[0104] (1) Construct a magnetic core digital model and simulate the magnetic core digital model;

[0105] Specifically, a three-dimensional digital model of the magnetic core is established based on the calculated cross-sectional area, number of units, and row and column arrangement. The magnetic core digital model is then simulated and analyzed using electromagnetic field simulation software to obtain data on the magnetic flux and magnetic flux density distribution inside the magnetic core, providing a basis for subsequent optimization.

[0106] (2) Determine the magnetic flux density distribution based on the simulation results, adjust the cross-sectional area of ​​the magnetic core according to the magnetic flux density distribution, and return to the step of simulating the magnetic core digital model until the magnetic flux density distribution is uniform.

[0107] Specifically, the simulation results are used to determine whether the magnetic flux density is uniform across the core. If localized areas of excessively high or unevenly distributed magnetic flux density are observed, the cross-sectional area of ​​the core at that location is adjusted, and the simulation is repeated. This process is repeated until the overall magnetic flux density of the core becomes uniform, with no obvious localized high points, thus completing the core structure optimization. It should be noted that the uniformity of the overall magnetic flux density is an engineering uniformity criterion guided by electrical performance and loss control. On all main magnetic circuit working sections of the core, the distribution differences in magnetic flux density are controlled within an allowable deviation band, and there are no abnormal concentration points that would cause localized magnetic flux density to exceed the allowable upper limit. Specifically, uniformity focuses not on non-primary working areas such as the edges and corners of the core, but rather on the effective cross-sectional area of ​​each core column and the transition areas at the connections between the core columns and the upper and lower yokes. These areas are the main channels for magnetic flux transmission, and the uniformity of their magnetic flux density distribution directly determines the overall iron loss level and saturation margin of the core. During the evaluation process, the average magnetic flux density on the effective cross-section of each magnetic column in the simulation cloud diagram is used as the benchmark. It is required that the magnetic flux density of more than 90% of the grid cells on this cross-section deviates from the average value by no more than ±10%, and the maximum local magnetic flux density on the cross-section does not exceed the set upper limit of the B value (e.g., 0.25T). Simultaneously, the average magnetic flux density difference between magnetic columns with the same function (e.g., all primary and secondary magnetic columns) should also be less than 5% to ensure the symmetrical operation of each cell. If the above conditions are met, it can be determined that the overall magnetic flux density of the core has achieved engineering uniformity, realizing the unity of minimizing iron loss and anti-saturation capability.

[0108] By using closed-loop iterative optimization of magnetic core numerical simulation and magnetic flux density distribution, the problems of excessive local magnetic flux density and uneven magnetic flux distribution in the magnetic core can be effectively eliminated, hysteresis loss and eddy current loss can be reduced, local saturation of the magnetic core can be avoided, and the symmetry of the magnetic circuit and operational reliability can be improved, laying the foundation for low-loss and high-efficiency operation of the transformer.

[0109] The matrix transformer design method provided in this embodiment achieves synergistic optimization of structural parameters and electrical performance under packaging constraints through a systematic and streamlined matrix transformer design approach, demonstrating significant technical advantages. By rationally dividing the number of units according to power, the power of each unit can be placed within the optimal range, adapting to the heat dissipation and process limitations of the PCB windings, while ensuring that the number of primary turns is divisible by the number of units, thus avoiding magnetic flux imbalance and increased leakage inductance. By matching the number of units with the aspect ratio of the package to the row and column arrangement, space can be fully utilized to form a symmetrical and uniform layout, improving space utilization, ensuring magnetic circuit symmetry and uniform magnetic flux distribution, and reducing the problem of increased iron loss caused by excessive local magnetic flux density. Furthermore, the winding connection method and arrangement are strictly matched. The series connection of the primary side can ensure consistent current and balanced magnetic flux in each unit, preventing local saturation of the magnetic core. The secondary side can flexibly choose series and parallel connection according to the number of turns and arrangement, taking into account voltage matching, current balance and convenient wiring, effectively reducing circulating current and losses. First, the cross-sectional area of ​​the magnetic core is calculated by electromagnetic laws, and then the cross-sectional area of ​​the winding is determined by combining the window space to achieve a balance between electrical constraints and structural space. While meeting the current carrying and heat dissipation requirements, the power density is improved. Finally, through initial numerical simulation and iterative optimization of iron loss and copper loss, the cross-sectional area of ​​the magnetic core and winding can be adjusted in a targeted manner to quickly converge to the optimal solution with the minimum total loss, greatly shortening the design cycle and avoiding the blindness of trial and error. With the special optimization of magnetic flux density distribution, the problem of excessive local magnetic flux density is further eliminated, and the core loss is reduced. Under strict high-encapsulation constraints, the transformer conversion efficiency is significantly improved, heat dissipation is optimized, and high power density and operational reliability are guaranteed.

[0110] Corresponding to the aforementioned embodiment of a matrix transformer design method, this application also provides an embodiment of a matrix transformer design apparatus.

[0111] Figure 5 This is a schematic diagram of the structure of Embodiment 2 of the matrix transformer design device provided in this application. Please refer to... Figure 5 The apparatus provided in this embodiment includes a processing module 510 and a design module 520;

[0112] The processing module 510 is used to determine the number of splitting units based on the power of the transformer;

[0113] The processing module 510 is also used to determine the row and column arrangement of the units based on the number of split units and the aspect ratio of the package.

[0114] The processing module 510 is also used to determine the winding connection method according to the unit row and column arrangement;

[0115] The processing module 510 is also used to calculate the cross-sectional area of ​​the magnetic core according to Faraday's law of electromagnetic induction, and to determine the cross-sectional area of ​​the winding according to the cross-sectional area of ​​the magnetic core.

[0116] The design module 520 is used to construct an initial digital model of the transformer based on the unit row and column arrangement, the connection method, the core cross-sectional area, and the winding cross-sectional area, to simulate the initial digital model, and to optimize and adjust the core cross-sectional area and the winding cross-sectional area according to the iron loss and copper loss in the simulation results, thereby completing the design of the matrix transformer.

[0117] The apparatus of this embodiment can be used to perform... Figure 1 The steps of the method embodiment shown are similar in principle and process, and will not be repeated here.

[0118] The specific implementation process of the functions and roles of each unit in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0119] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0120] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method of designing a matrix transformer, characterized by, The method includes: The number of units to be split is determined based on the transformer's power; among which, the power of a single unit is determined based on design requirements. The number of splitting units is determined based on the ratio of the transformer's power to the power of a single unit; The unit row and column arrangement is determined based on the number of split units and the package aspect ratio; wherein, when the number of split units is a first value, the unit row and column arrangement is determined to be a single row arrangement; When the number of split units is the second value, the unit row and column arrangement is determined to be one row of four columns or two rows of two columns; When the number of split units is the third value, the row and column arrangement of the units is determined to be either one row of six columns or two rows of three columns; The winding connection method is determined based on the unit row and column arrangement. The cross-sectional area of ​​the magnetic core is calculated according to Faraday's law of electromagnetic induction, and the cross-sectional area of ​​the winding is determined based on the cross-sectional area of ​​the magnetic core. Based on the unit row and column arrangement, the connection method, the core cross-sectional area, and the winding cross-sectional area, an initial digital model of the transformer is constructed. The initial digital model is simulated, and the core cross-sectional area and the winding cross-sectional area are optimized and adjusted according to the iron loss and copper loss in the simulation results to complete the design of the matrix transformer. The process includes simulating the initial digital model and optimizing the core cross-sectional area and the winding cross-sectional area based on the iron loss and copper loss results, including: The total simulated loss is calculated based on iron loss and copper loss. The cross-sectional area of ​​the magnetic core is adjusted according to the iron loss, and the cross-sectional area of ​​the winding is adjusted according to the copper loss; The total simulation loss is recalculated based on the adjusted core cross-sectional area and winding cross-sectional area, and the total simulation loss is iteratively optimized until the loss condition is met. The winding includes a primary winding and a secondary winding. Determining the winding connection method based on the unit row and column arrangement includes: The primary windings are connected in series between each unit; When the unit row and column arrangement is in a row, the number of turns of the secondary winding and the number of split units are compared. If the number of turns is less than the number of split units, the secondary winding is connected in parallel; if the number of turns is greater than the number of split units, the secondary winding is connected in series. When the unit is arranged in two rows, the secondary windings of the units in the same column are connected in series, and the secondary windings of the units in different columns are connected in parallel.

2. The method of claim 1, wherein, Determining the winding cross-sectional area based on the magnetic core cross-sectional area includes: The window space of the winding is determined based on the cross-sectional area of ​​the magnetic core. Calculate the initial cross-sectional area of ​​the winding based on the transformer operating current and winding current density; The initial cross-sectional area of ​​the winding is adjusted according to the window space to obtain the winding cross-sectional area.

3. The method of claim 1, wherein, The method further includes: Construct a digital model of the magnetic core and simulate the digital model of the magnetic core; The magnetic flux density distribution is determined based on the simulation results. The cross-sectional area of ​​the magnetic core is adjusted according to the magnetic flux density distribution. The process of simulating the digital model of the magnetic core is repeated until the magnetic flux density distribution is uniform.

4. The method of claim 1, wherein, The power of a single unit is greater than or equal to 1000W and less than or equal to 1500W.

5. The method of claim 1, wherein, When calculating the cross-sectional area of ​​the magnetic core, the waveform coefficient is 4.44 and the B value is 0.2-0.25T.

6. A design device for a matrix transformer, characterized in that, The device is implemented based on any one of the methods of claims 1-5 above, and the device includes a processing module and a design module; The processing module is used to determine the number of splitting units based on the power of the transformer; The processing module is also used to determine the row and column arrangement of the units based on the number of split units and the aspect ratio of the package. The processing module is also used to determine the winding connection method according to the row and column arrangement of the unit; The processing module is also used to calculate the cross-sectional area of ​​the magnetic core according to Faraday's law of electromagnetic induction, and to determine the cross-sectional area of ​​the winding according to the cross-sectional area of ​​the magnetic core. The design module is used to construct an initial digital model of the transformer based on the unit row and column arrangement, the connection method, the core cross-sectional area, and the winding cross-sectional area, to simulate the initial digital model, and to optimize and adjust the core cross-sectional area and the winding cross-sectional area according to the iron loss and copper loss in the simulation results, thereby completing the design of the matrix transformer.

Images