Voltage regulator, circuit module, electronic equipment and coupling inductor structure

Abstract

According to the voltage regulator, the circuit module, the electronic equipment and the coupling inductor structure, the three windings are integrated on the three-column magnetic core, so that magnetic coupling and electric coupling between different inductor windings in a two-phase voltage reduction circuit are achieved, and the size of a VRM is reduced while the performance of the VRM is optimized. The VRM can serve as a power supply solution to be deployed around a main chip of ICT related hardware products (such as a server or AI inference platform equipment).

Description

Technical Field

[0001] This application relates to the field of power electronics technology, and in particular to a voltage regulator, circuit module, electronic device and coupled inductor structure. Background Technology

[0002] As chips continue to evolve, their clock frequencies are getting faster, their operating voltages are getting lower, and their required operating currents are getting higher. Furthermore, the operating current is constantly being dynamically adjusted according to changes in the operating business. As a result, the application of voltage regulator modules (VRMs) that power chips is becoming more and more widespread.

[0003] However, as the layout space allocated to power supplies on single boards continues to decrease, the requirements for VRM performance (such as steady-state performance and / or dynamic performance) are constantly increasing. How to optimize VRM performance while reducing its size has become a research hotspot. Summary of the Invention

[0004] Achieving magnetic and electrical coupling between different inductor windings in a two-phase buck converter using a coupled inductor structure with three windings is beneficial for optimizing VRM performance. However, as more windings need to be wound on the core, the number of core posts increases accordingly. For example, three windings need to be integrated on a five-post core to achieve magnetic and electrical coupling between different inductor windings in the two-phase buck converter. This results in a larger core size, which in turn increases the size of the VRM.

[0005] To address this, this application provides a voltage regulator, circuit module, electronic device, and coupled inductor structure. By integrating three windings on a three-pillar magnetic core, magnetic and electrical coupling between different inductor windings in a two-phase step-down circuit is achieved, which helps to optimize VRM performance while reducing VRM size. The solution provided in this application is described below.

[0006] Firstly, this application provides a power supply device, which can be a voltage regulator or the aforementioned VRM. The following description uses a VRM as an example. The VRM includes a capacitor and at least one circuit unit, the circuit unit including a coupled inductor structure and two switching circuits. The coupled inductor structure includes a magnetic core and multiple windings integrated on the magnetic core, wherein the magnetic core has three core legs, and the multiple windings include secondary windings and two primary windings, and the multiple windings are respectively wound on different core legs. The input terminals of the two switching circuits are respectively used to connect to a power supply, the output terminals of the two switching circuits are respectively connected to the input terminals of the two primary windings, and the output terminals of the two primary windings are respectively used to connect to a capacitor and a load.

[0007] For ease of description, the two switching circuits are referred to as the first switching circuit and the second switching circuit, respectively, and the two primary windings are referred to as the first primary winding and the second primary winding, respectively. The output terminal of the first switching circuit is connected to the first primary winding, and the output terminal of the second switching circuit is connected to the second primary winding. A filter circuit can be formed by connecting a capacitor to the output terminal of the primary winding. Therefore, the two switching circuits and the two primary windings can be used to form a two-phase buck circuit in a VRM. For ease of description, this application refers to these two-phase buck circuits as the first buck circuit and the second buck circuit, respectively, wherein the first buck circuit includes the first switching circuit and the first primary winding, and the second buck circuit includes the second switching circuit and the second primary winding.

[0008] By integrating the three windings onto a three-column core, it is beneficial to achieve magnetic and electrical coupling between different inductor windings (i.e., the first primary winding and the second primary winding) in a two-phase step-down circuit. This not only helps to optimize the performance of the VRM, but also helps to reduce the size of the core by reducing the number of core columns, thereby reducing the size of the VRM and improving the power supply efficiency of the VRM.

[0009] Optionally, when there is negative coupling between the two primary windings, the two primary windings are positively coupled to the secondary windings respectively.

[0010] A portion of the magnetic flux generated by the primary winding (called the first magnetic flux) can be used to couple with another primary winding. Another portion of the magnetic flux generated by the primary winding (called the second magnetic flux) can be used to couple with the secondary winding. A third portion of the magnetic flux generated by the primary winding (called the third magnetic flux) is neither used to couple with another primary winding nor with the secondary winding. Therefore, the third magnetic flux can be equivalent to leakage inductance.

[0011] This application also provides a method for increasing leakage inductance in a coupled inductor structure with three windings integrated on a three-pillar magnetic core. By increasing leakage inductance, it is beneficial to obtain an equivalent secondary inductance (denoted as Lc) in the VRM circuit topology that meets the dynamic requirements of the VRM, thereby avoiding the need for an additional Lc in series with the secondary winding in the VRM, and thus reducing the physical size of the VRM. The magnetic core column with the secondary winding wound among the three core columns is referred to as the first core column. The possible methods for increasing leakage inductance are described below.

[0012] In the first method of increasing leakage inductance, the effective cross-sectional area Ae of the secondary winding on the first core post can be smaller than the cross-sectional area of ​​the first core post; in other words, the secondary winding is wound around a portion of the cross-section of the first core post. This allows only a portion of the magnetic flux generated by the primary winding in the first core post to couple with the secondary winding, thereby increasing the leakage inductance of the primary winding.

[0013] In the second method of increasing leakage inductance, a material with low permeability can be selected to prepare the magnetic core. For example, the material of the magnetic core structure is a metallic magnetic material, and / or the permeability of the material of the magnetic core structure does not exceed 100 henries / meter. This facilitates the leakage of a portion of the magnetic flux generated by the primary winding to the surrounding environment outside the magnetic core, thereby increasing the leakage inductance of the primary winding. It also helps to reduce the processing difficulty of the magnetic core and facilitates mass production. In addition, compared with ferrite materials, metallic magnetic materials (or magnetic powder core materials) have lower permeability, thus having a higher saturation magnetic flux density, which helps to reduce the volume of the magnetic core and further reduce the volume of the VRM, thus expanding the application scenarios of the VRM. For example, it can also be applied to application scenarios with strong requirements for size and height. In the second method of increasing leakage inductance, the Ae of the secondary winding on the first magnetic core column can be equal to the cross-sectional area of ​​the first magnetic core column, or in other words, the secondary winding is wound around the entire cross-section of the first magnetic core column.

[0014] In the third way of increasing leakage inductance, the Ae of the secondary winding on the first magnetic core post can be smaller than the cross-sectional area of ​​the first magnetic core post, and the material of the magnetic core structure is a metallic magnetic material and / or the permeability of the material of the magnetic core structure does not exceed 100 henry / meter. This is beneficial to further increase the leakage inductance of the primary winding.

[0015] This application is not limited to the method of achieving a cross-sectional area (Ae) of the secondary winding on the first magnetic core post being less than that of the first magnetic core post in the coupled inductor structure. Optionally, the first magnetic core post has a groove, the depth direction of which can be perpendicular to the cross-sectional direction of the first magnetic core post or perpendicular to the plane where the secondary winding is located. All or part of the secondary winding can be installed in the groove, thereby achieving a cross-sectional area (Ae) of the secondary winding on the first magnetic core post being less than that of the first magnetic core post.

[0016] The previous section introduced the mounting method of the secondary winding on the magnetic core. The following section will refer to the two core posts other than the first core post as the second core post and the third core post, respectively. The mounting method of the primary winding on the magnetic core is described below.

[0017] In this application, the portion of the primary winding located within the magnetic column gap is referred to as the target side. The magnetic column gap can be the gap between the second magnetic core column and the first magnetic core column (referred to as the first magnetic column gap), or the gap between the third magnetic core column and the first magnetic core column (referred to as the second magnetic column gap). The target side of the first primary winding can be located within the first magnetic column gap, and the target side of the second primary winding can be located within the second magnetic column gap.

[0018] In this application, the portion of the secondary winding located in the first magnetic column gap is referred to as the first side, and the portion of the secondary winding located in the second magnetic column gap is referred to as the second side. The target side of the first primary winding can be used to couple with the first side of the secondary winding, and the target side of the second primary winding can be used to couple with the second side of the secondary winding.

[0019] In the first installation method of the primary winding, the primary winding includes a portion of winding (called other sides) besides the target side. That is, the target side of the primary winding is used for coupling with the secondary winding, while the other sides are not used for coupling with the secondary winding. Optionally, the other sides of the first primary winding are wound on the second core post, and the other sides of the second primary winding are wound on the third core post. Based on the second or third method of increasing leakage inductance, it is beneficial to increase the magnetic flux leaking from the other sides of the primary winding into the surrounding environment outside the core, thereby further increasing the leakage inductance of the primary winding.

[0020] In the second installation method of the primary winding, the two ends of the target side can be the input and output ends of the corresponding primary winding, respectively. Alternatively, the entire primary winding can be used for coupling with the secondary winding, or the entire primary winding can be located within the first or second magnetic column gap. For example, the two ends of the target side of the first primary winding can be directly connected to the input and output ends of the first primary winding, respectively, and the two ends of the target side of the second primary winding can be directly connected to the input and output ends of the second primary winding, respectively. This allows for a shorter primary winding length while maintaining the coupling coefficient between the primary and secondary windings, thereby reducing the loss of the drive signal in the primary winding and improving the power supply efficiency of the VRM to the load.

[0021] The following describes the possible mounting methods of the coupled inductor structure on the circuit board.

[0022] Optionally, the coupled inductor structure is used to be fixed on the surface of the circuit board or embedded inside the circuit board.

[0023] In the first mounting method, at least one winding of the coupled inductor structure is perpendicular to the surface of the circuit board, which helps to reduce the area occupied by the coupled inductor structure on the surface of the circuit board, thereby helping to reduce the area of ​​the circuit board surface.

[0024] In the second mounting method, the multiple windings in the coupled inductor structure are parallel to the surface of the circuit board, or in other words, the extension direction of the windings in the coupled inductor structure is parallel to the surface of the circuit board. This ensures that even if the input and output terminals of the windings are on different surfaces of the coupled inductor structure—for example, in the second mounting method for the primary winding, the input and output terminals might be located on opposite surfaces of the coupled inductor structure—it still helps to ensure that the input and output terminals of the windings are at the same height relative to the circuit board. This helps to shorten the length of the lead wires between the input terminals and the PCB, reducing the loss of the drive signal in the primary winding and thus improving the power supply efficiency of the VRM to the load. Furthermore, having multiple windings parallel to the circuit board also allows the coupled inductor structure to be embedded within the circuit board, using the traces on the circuit board as windings, thereby further reducing the height of the VRM on the board.

[0025] As mentioned earlier, the target side of the first primary winding can be coupled with the first side of the secondary winding, and the target side of the second primary winding can be coupled with the second side of the secondary winding.

[0026] This application does not limit the deployment orientation of the target side of the first primary winding and the surface (referred to as the first coupling surface) that is opposite to (or closest to) the first side of the secondary winding. (e.g., the angle between the coupling surface and the plane where the magnetic core column section / secondary winding is located) is located, nor does it limit the deployment orientation of the target side of the second primary winding and the surface (referred to as the second coupling surface) that is opposite to (or closest to) the second side of the secondary winding.

[0027] To increase the coupling coefficient, the area of ​​the first coupling surface can be larger than the area of ​​the other adjacent surfaces, and the area of ​​the second coupling surface can be larger than the area of ​​the other adjacent surfaces.

[0028] In the first deployment orientation of the coupling surface, the coupling surface can be perpendicular to the cross-section of the core post or the plane where the secondary winding is located. The coupling surface can be a first coupling surface and / or a second coupling surface. This facilitates laying the winding flat along a plane perpendicular to the spacing between the core posts or perpendicular to the arrangement direction of the multiple core posts, thereby reducing the thickness of the winding along the direction of the core post spacing or the arrangement direction of the multiple core posts, and consequently reducing the spacing between adjacent core posts and the volume of the core.

[0029] In the second deployment orientation of the coupling surface, the coupling surface can be parallel to the cross-section of the magnetic core column or the plane where the secondary winding is located. This allows the winding to be laid flat on a plane parallel to the winding, reducing the thickness of the winding along the direction perpendicular to that plane, thereby reducing the height of the magnetic core column (i.e., perpendicular to its cross-section). Thus, mounting the coupled inductor structure on the circuit board according to this second mounting method helps reduce the dimension of the coupled inductor structure along the thickness direction of the circuit board, thereby reducing the thickness of the single board.

[0030] The cross-section of the magnetic core column mentioned in this application may refer to the cross-section of the effective cross-section of the magnetic core parallel to the secondary winding, or the cross-section of the magnetic core column perpendicular to the direction of the magnetic field generated by the winding.

[0031] This application does not limit the function of the switching circuit. For example, the switching circuit can be used to adjust the phase shift angle of the drive signal output by the power supply, and / or adjust the amplitude of the drive signal. This application does not limit the type of drive signal. For example, the drive signal can be a voltage signal or a current signal.

[0032] This application does not limit the type of switching circuit. Optionally, the two switching circuits each include a first switching transistor and a second switching transistor, the first terminal of the first switching transistor is connected to the input terminal of the switching circuit, the second terminal of the first switching transistor is connected to the first terminal of the second switching transistor and the output terminal of the switching circuit, and the second terminal of the second switching transistor is grounded.

[0033] Optionally, the VRM further includes a controller for controlling the two switching circuits respectively. The controller controls the switching states of the first and second switching transistors respectively, thereby facilitating the separate control of the two switching circuits to perform phase shifting of the drive signal output by the power supply at different angles, and / or adjusting the amplitude of the drive signal.

[0034] Optionally, the VRM further includes a drive circuit, the input of which is connected to the controller, and the output of which is connected to the control terminals of the first switch and the second switch, respectively. The drive circuit is used to control the first switch and the second switch according to the control signal output by the controller, thereby controlling the two switch circuits to perform phase shifting of the drive signal output by the power supply at different angles, and / or to adjust the amplitude of the drive signal.

[0035] This application does not limit the specific structure of the magnetic core. For example, the coupled inductor can be a three-pillar EE type magnetic core or a three-pillar EI type magnetic core. Alternatively, the magnetic core in the coupled inductor can be obtained by slotting the magnetic core column in the middle of the three-pillar EE type magnetic core or the three-pillar EI type magnetic core.

[0036] In a second aspect, this application provides a circuit module, the circuit module including a circuit board, one or more chips integrated on the circuit board, and one or more VRMs integrated on the circuit board, the VRMs being the VRMs described in the first aspect or any implementation thereof, the one or more VRMs being used to power the one or more chips.

[0037] A single VRM can be used to power a single chip, or a single VRM can be used to power multiple chips, or multiple VRMs can be used to power a single chip.

[0038] The output of the VRM can be connected (or directly connected) to the power pin of the corresponding chip. The VRM can be deployed around the main chip on the circuit board as a power supply solution.

[0039] Thirdly, this application provides an electronic device including a housing and one or more circuit modules installed inside the housing, which can be the circuit modules described in the second aspect.

[0040] This application does not limit the type of electronic device. For example, the electronic device may be an information and communications technology (ICT) related hardware product, such as a server or an artificial intelligence (AI) inference platform device.

[0041] Fourthly, this application provides a coupled inductor structure, the coupled inductor structure including a magnetic core and a plurality of windings integrated on the magnetic core, wherein the magnetic core has three core posts, the plurality of windings include a secondary winding and two primary windings, and when the two primary windings are negatively coupled to each other, the two primary windings are positively coupled to the secondary windings respectively.

[0042] Applying the coupled inductor structure provided in this application to a VRM, with the two primary windings serving as inductors in two buck circuits respectively, is beneficial for achieving magnetic and electrical coupling between different inductor windings in the two-phase buck circuit. This not only helps optimize the performance of the VRM, but also helps reduce the volume of the magnetic core by reducing the number of magnetic core pillars, thereby reducing the volume of the VRM and improving the power supply efficiency of the VRM.

[0043] Optionally, the secondary winding is wound on the first of the three magnetic core pillars; and the effective cross-sectional area Ae of the secondary winding on the first magnetic core pillar is smaller than the cross-sectional area of ​​the first magnetic core pillar, and / or the material of the magnetic core structure is a metallic magnetic material, and / or the permeability of the material of the magnetic core structure does not exceed 100 henry / meter.

[0044] Optionally, the two primary windings each include a target side, and the target sides of the two primary windings are respectively located in the first magnetic column gap and the second magnetic column gap, wherein the first magnetic column gap is located between the second magnetic column and the first magnetic column among the three magnetic core columns, and the second magnetic column gap is located between the third magnetic column and the first magnetic core column among the three magnetic core columns.

[0045] The two primary windings each include other sides besides the target side, and the other sides of the two primary windings are respectively wound on the second magnetic core post and the third magnetic core post, or the two ends of the target side are the input end and the output end of the corresponding primary winding.

[0046] Optionally, the coupled inductor structure is used to be fixed on the surface of the circuit board or embedded inside the circuit board, and the plurality of windings are respectively parallel to the circuit board.

[0047] Optionally, the surface on the target side opposite to the secondary winding is parallel to the circuit board.

[0048] The details and beneficial effects of the various possible implementation methods of the coupled inductor structure in the fourth aspect can be understood by referring to the relevant content in the first aspect, and will not be repeated here. Attached Figure Description

[0049] Figure 1 This schematic illustrates one possible circuit topology for a VRM;

[0050] Figure 2 The circuit topology of the VRM corresponding to Solution 1 is illustrated schematically;

[0051] Figure 3 The circuit topology of the VRM corresponding to Solution 2 is illustrated schematically;

[0052] Figure 4 The circuit topology of the VRM corresponding to solution 4 is illustrated schematically;

[0053] Figures 5-1 to 5-3 The possible structures of the coupled inductor provided in this application are illustrated schematically.

[0054] Figure 6 The circuit topology of the VRM provided in this application is illustrated schematically.

[0055] Figures 7-1 to 7-4 The application scenarios of the VRM provided in this application are illustrated respectively. Detailed Implementation

[0056] As chips continue to evolve, their area and the types of power supplies they require are increasing, and their power consumption is also growing. The usable area for power supplies on a single board is becoming increasingly smaller, making it crucial to improve power supply density. The types of chips discussed in this application include, but are not limited to, central processing units (CPUs), graphics processing units (GPUs), field-programmable gate arrays (FPGAs), and neural network processing units (NPUs).

[0057] Furthermore, chip clock frequencies are getting faster, operating voltages are getting lower, and the required operating current is getting higher, with the operating current constantly undergoing large dynamic adjustments as the operating business changes. For example, the operating voltage of some current CPU cores has been reduced to below 1V, and the operating current can reach 500A to 1000A or even higher, with the dynamic change rate of the operating current reaching over 1000A / μs.

[0058] To meet the above operating conditions of the chip, a voltage regulator module (VRM) is generally used to power the chip. Furthermore, VRMs often employ multiphase buck circuits to reduce output voltage ripple and improve dynamic response speed. The circuit topology of a VRM is generally as follows: Figure 1 As shown.

[0059] In a Virtual Resonant Module (VRM), the inductor, as a passive component, is a key component. It not only has a significant impact on the steady-state and dynamic performance of the circuit, but also accounts for a large proportion of the total size of the VRM because inductors are generally implemented through magnetic cores and windings.

[0060] To meet the increasingly demanding dynamic requirements of chip loads, the filter inductance in VRM circuits is gradually decreasing. However, a smaller filter inductance leads to increased current ripple and reduced efficiency. Therefore, how to reduce current ripple and improve the dynamic rate of change of operating current while reducing the inductor structure size has become an urgent problem to be solved.

[0061] Figure 2 This schematically illustrates one solution (referred to as Solution 1). For example... Figure 2As shown, by integrating the inductor windings of two buck circuits onto the same magnetic core, magnetic coupling is formed, resulting in a coupled inductor structure (or simply coupled inductor). Replacing the discrete inductors in the VRM (i.e., a single winding wound on a single magnetic core) with coupled inductors not only reduces the volume occupied by inductors in multiphase buck circuits, thus reducing the volume of the VRM on the board, but also improves transient performance while ensuring the same steady-state ripple, as the windings of the two buck circuits are negatively coupled through the same magnetic core. Magnetic coupling can refer to the physical phenomenon of different current-carrying coils (or windings) being interconnected through their respective magnetic fields. In VRM, it specifically refers to the inductor windings of multiphase buck circuits being wound on the same magnetic core in various forms, thereby coupling with each other in the magnetic circuit.

[0062] However, for magnetically coupled VRMs, when the coupling coefficient is small, the steady-state and dynamic inductance changes are not significant, and performance optimization is not obvious. Only when the coupling coefficient is large, i.e., the two-phase coupling is strong, is the steady-state and dynamic improvement effect more significant. Increasing the coupling coefficient means stronger two-phase coupling, increased mutual influence between the two phases, increased control difficulty, and worsened process consistency when multiple phases are coupled. Furthermore, while maintaining the same steady-state inductance, the adjustment range of the coupling coefficient is limited, and the adjustment range of the transient inductance is also relatively limited.

[0063] Figure 3 The circuit topology of a transient inductor voltage regulator (TLVR) is schematically shown. The VRM uses the TLVR as an alternative solution to the aforementioned problem (referred to as Solution 2). Figure 3 As shown, the inductor winding of the multiphase buck circuit serves as the primary winding of the transformer. Inductors from different buck circuits are indirectly electrically coupled through the secondary winding of the transformer. This electrical coupling achieves the coordinated operation between phases in the traditional multiphase interleaved reduction circuit, improving the circuit's transient response. When the primary current of one phase changes, the currents of other phases also change accordingly through the coupling on the secondary side of the transformer, thus significantly increasing the total output current I. L The rate of change is optimized for transient response. In this application, the winding can also be referred to as a coil. Electrical coupling can refer to the direct connection of various parts of a circuit through wires to coordinate their operation. In this application, it can refer to the secondary winding of the VRM (or TLVR) transformer being connected in series through wires to achieve the desired effect. Figure 1 The step-down circuits of the VRM shown are connected together and work in coordination to achieve the effect that when the current of one phase changes, the current of the other phases follows suit.

[0064] However, TLVR transforms the inductance of VRM into a transformer. This not only requires adding a secondary inductor (e.g., Li0, where i is a positive integer less than or equal to n) to each inductor in VRM, but also requires adding an additional inductor L in series with the secondary inductors to achieve the corresponding coupling (e.g., Mi0). c These factors make the losses and size of TLVR larger than those of traditional multiphase interleaved buck circuits.

[0065] To reduce the size of the TLVR, the structure of the transformer in the TLVR can be improved (referred to as Solution 3). In the improved transformer, the core has three core pillars. Both the primary and secondary coils are wound around the middle core pillar (or the middle pillar core). However, the middle pillar core is divided into two parts, and the primary or secondary coil only winds around a portion of the middle pillar core. This results in a portion of the magnetic flux of the primary or secondary coil not coupling with other coils, forming leakage inductance. The leakage inductance in each step-down circuit can be equivalent to the inductor L connected in series with the secondary inductor as described earlier. c (or TLVR secondary-side coupling inductor or secondary-side coupling inductor L) c This structure allows for flexible adjustment of the turns ratio even when there is only one turn. This core structure enables quantitative control of the leakage inductance of the TLVR transformer, equivalent to coupling the secondary side inductance L... c The secondary coupling inductors L are distributed and integrated into each transformer, which helps to avoid the need to install additional secondary coupling inductors L outside of each transformer (or coupling inductor). c This helps to reduce the overall size and loss of magnetic components in TLVR.

[0066] However, this method results in a more complex core structure, requires higher machining precision for the central core, and presents challenges in manufacturing consistency. Furthermore, using this core only eliminates the secondary-side coupling inductance L. c The volume of the original transformer cannot be reduced, and the overall volume reduction is limited.

[0067] Considering the advantages of combining the magnetically coupled VRM of Solution 1 and the electrically coupled VRM of Solution 2 (i.e., TLVR), namely, reducing the size of magnetic components through magnetic coupling and improving transient performance while maintaining the same steady-state ripple, thus reducing losses, designing a hybrid TLVR combining magnetic and electrical coupling is of great significance for improving the dynamic characteristics, reducing size, and increasing efficiency of traditional interleaved buck converters.

[0068] Figure 4 The schematic diagram illustrates the structure of the hybrid coupled inductor used in Solution 4, which can form both magnetic and electrical coupling. For example... Figure 4As shown, Solution 4 uses a five-pillar EI-type core instead of the three-pillar EE-type core used in Solution 3. Two primary windings (i.e., the inductor windings in the buck circuit) and one secondary winding are wound on the three middle core pillars. Of the three middle core pillars, the two outer pillars are used to wind the primary windings of the two buck circuits of the TLVR, while the middle pillar is used to wind the secondary windings of the TLVR. The primary-secondary coupling coefficient and leakage inductance can be adjusted by changing the air gap of each core pillar and the width of the two outer pillars (i.e., the unwound core pillars). Then, the secondary windings of all the magnetic components (i.e., the inductor windings in the buck circuit) are connected in series to form the secondary side of the TLVR. This structure achieves the integration of the inductor and secondary inductor Lc of the two-phase buck circuit, forming magnetic and electrical coupling, resulting in high integration and reduced size.

[0069] However, to adjust the coupling coefficient, the five core posts may require different air gap lengths, which is difficult to control in actual processing and results in high manufacturing difficulty. When a high primary-secondary coupling coefficient is required, the magnetic flux of the two outer posts (i.e., the core posts without windings) is small, and the thickness of the outer posts may be very thin, further increasing the difficulty of the manufacturing process. In addition, the outer posts are prone to damage, reducing the reliability of TLVR.

[0070] To further reduce the size of the magnetic components in a VRM, this application provides a coupled inductor that integrates two primary windings and one secondary winding on a three-limb magnetic core. When the two primary windings of this coupled inductor are used as the inductor in a two-phase step-down circuit of a VRM, the advantages of a two-phase magnetically coupled VRM and an electrically coupled TLVR can be combined to achieve hybrid magneto-electric coupling. This reduces the core size, improves circuit efficiency, and significantly enhances the transient response performance of the circuit.

[0071] The structure of the coupled inductor provided in this application is described below with reference to the accompanying drawings.

[0072] Figure 5-1 This schematically illustrates one possible diagram of a coupled inductor mounted on a circuit board. (Example:) Figure 5-1 As shown, the coupled inductor includes three windings: primary winding 1, primary winding 2, and secondary winding. Figure 5-1 In the coupled inductor structure shown, the part other than the three windings is the magnetic core. The magnetic core includes two cover plates and three magnetic core pillars located between the two cover plates, which are respectively denoted as magnetic core pillar 1 to magnetic core pillar 3.

[0073] like Figure 5-1As shown, the cross-sectional area of ​​the magnetic core column is the area of ​​its cross-section in the xz plane. Primary winding 1 and primary winding 2 are wound on magnetic core columns 2 and 3, respectively. Magnetic core column 1 has a groove along the xy plane. The secondary winding is installed in the groove and the gap between the different magnetic core columns; in other words, the secondary winding is wound on a portion of the cross-section of magnetic core column 1. For example... Figure 5-1 As shown, the cross-sectional area of ​​the magnetic core column is divided into section 1 and section 2 by the groove. The effective cross-sectional area Ae of the secondary winding on the magnetic core column 1 is the area of ​​section 1, which is smaller than the cross-sectional area of ​​the first magnetic core column. Since the magnetic flux of the primary winding in the magnetic core column 1 includes the magnetic flux flowing through section 1 and the magnetic flux flowing through section 2, it is beneficial to increase the leakage inductance of the primary winding 1 and the primary winding 2. Figure 5-1 The coupled inductor shown can be understood as using the first method of increasing leakage inductance introduced earlier. Optionally, to further increase the leakage inductance, Figure 5-1 The coupled inductor shown can also utilize the second method for increasing leakage inductance described above, namely, by selecting a material with low permeability to fabricate the magnetic core. For example, the material of the magnetic core structure is a metallic magnetic material, and / or the permeability of the material of the magnetic core structure does not exceed 100 Hen / meter. This application does not limit the type of metallic magnetic material; for example, the metallic magnetic material may include iron and / or nickel.

[0074] The target side of the primary winding 1 introduced earlier can be understood as... Figure 5-1 The primary winding 1 is the winding portion located between core column 1 and core column 2. The target side of the primary winding 2 can be understood as... Figure 5-1 The primary winding 2 is the section of the winding located between core column 1 and core column 3. The other sides of primary winding 1 and primary winding 2 can be understood as... Figure 5-1 The corresponding primary winding portion excluding the target winding. Therefore, Figure 5-1 The coupled inductor shown can be understood as using the first mounting method of the primary winding described earlier. Optional, Figure 5-1 The coupled inductor shown can also be installed using the second mounting method for the primary winding described above.

[0075] The first side of the secondary winding introduced earlier can be understood as... Figure 5-1 The secondary winding is the section of the winding located between core column 1 and core column 2. The second side of the secondary winding can be understood as... Figure 5-1 The secondary winding is located in the section of the winding between core column 1 and core column 3. For example... Figure 5-1As shown, the surfaces on the primary winding 1 and the first side of the secondary winding that are opposite to each other (or closest to each other) (i.e., the first coupling surface described above) are parallel to the yz plane, that is, perpendicular to the cross-section of the magnetic core column or the plane where the secondary winding is located. Similarly, the surfaces on the primary winding 2 and the second side of the secondary winding that are opposite to each other (or closest to each other) (i.e., the second coupling surface described above) are parallel to the yz plane, that is, perpendicular to the cross-section of the magnetic core column or the plane where the secondary winding is located. Therefore, Figure 5-1 The coupled inductor shown can be understood as using the first deployment orientation of the coupling surface described above. This facilitates the laying of the winding along a plane perpendicular to the spacing between the magnetic cores or a plane perpendicular to the arrangement direction of the multiple cores (i.e., the yz plane), thereby reducing the thickness of the winding along the direction of the spacing between the magnetic cores or the arrangement direction of the multiple cores (i.e., along the x-direction), and consequently reducing the spacing between adjacent cores (e.g., the spacing between core 1 and core 2 and / or the spacing between core 1 and core 3), thus reducing the core volume. Optionally, Figure 5-1 The coupled inductor shown can also use the second deployment orientation of the coupled surface described above.

[0076] like Figure 5-1 As shown, the three windings are perpendicular to the xy plane, that is, perpendicular to the surface of the circuit board. Therefore, Figure 5-1 The coupled inductor shown can be understood as using the first mounting method described above. Optional, Figure 5-1 The coupled inductor shown can also be installed using the first mounting method described above.

[0077] Figure 5-2 This schematically illustrates another possible diagram of a coupled inductor mounted on a circuit board. Figure 5-1 The coupling inductors shown are the same. Figure 5-2 The coupled inductor shown includes a three-pillar core and three windings, using the first mounting method for the primary winding, the first deployment orientation of the coupling surface, and the first mounting method on the circuit board described above. Figure 5-1 The coupling inductances shown are different. Figure 5-2 In the coupled inductor shown, the effective cross-sectional area Ae of the secondary winding on the core column 1 is equal to the cross-sectional area of ​​the first core column.

[0078] Optional, to increase leakage inductance, Figure 5-2The coupled inductor shown can utilize the second method for increasing leakage inductance described earlier, namely, using a material with low permeability to fabricate the core. For example, the core structure is made of a metallic magnetic material, and / or the permeability of the core structure material does not exceed 100 henries / meter. Due to the low permeability of the core material, the leakage flux (or leakage magnetic flux) of the primary winding in the surrounding environment outside the core is increased. This leakage flux is sufficient to form a sufficiently large leakage inductance, thereby enabling the circuit to achieve suitable transient and steady-state performance. Furthermore, by using a material with low permeability (e.g., magnetic powder core material), leakage inductance can be constructed through the leakage flux in the surrounding environment outside the core, allowing the secondary winding to completely surround the middle column. This greatly simplifies the structure of the core and winding; for example, a conventional three-column EE core or a three-column EI core can be used, significantly reducing processing difficulty, improving process consistency, facilitating manufacturing and use, and enabling mass production.

[0079] Optional, Figure 5-2 The coupled inductor shown can also use the second mounting method of the primary winding described above, and / or the second deployment orientation of the coupling surface and the second mounting method on the circuit board described above.

[0080] Figure 5-3 This schematic illustrates another possible diagram of a coupled inductor mounted on a circuit board. (Example) Figure 5-3 As shown, the coupled inductor includes three windings: primary winding 1, primary winding 2, and secondary winding. The portion or region of the coupled inductor other than the windings represents the magnetic core. The region within the magnetic core located between the two primary windings corresponds to... Figure 5-1 and Figure 5-2 The magnetic core column 1 shown has a region located in the -x direction (i.e., the opposite direction to the x direction) of the primary winding 1, which is equivalent to... Figure 5-1 and Figure 5-2 The magnetic core column 2 shown has a region in the primary winding 2 located in the x-direction that corresponds to... Figure 5-1 and Figure 5-2 The magnetic core column 3 shown has its cross-section or the plane containing the secondary winding located in the xy plane. Therefore, and Figure 5-1 The coupling inductors shown are the same. Figure 5-3 The coupled inductor shown includes a three-pillar core and three windings. This coupled inductor uses the first method described earlier for increasing leakage inductance, where the secondary winding winds wind a portion of the cross-section of the core pillar between the two primary windings. Optionally, to further increase leakage inductance, Figure 5-3 The coupled inductor shown can also utilize the second method for increasing leakage inductance described above, namely, selecting a material with low permeability to fabricate the magnetic core. For example, the material of the magnetic core structure is a metallic magnetic material, and / or the permeability of the material of the magnetic core structure does not exceed 100 henries / meter. Optionally, Figure 5-3In the coupled inductor shown, the secondary winding can also wind the entire cross section of the magnetic core column between the two primary windings.

[0081] and Figure 5-1 and Figure 5-2 The coupling inductors shown are all different. Figure 5-3 In the coupled inductor shown, both primary windings are located entirely between two adjacent core pillars. In other words, the two primary windings are the target sides, excluding the other sides described earlier; or, the two ends of the target sides are the input and output terminals of the primary windings, respectively. Therefore, Figure 5-3 The coupled inductor shown can be understood as using the second mounting method of the primary winding described earlier. This helps to shorten the length of the primary winding while maintaining the coupling coefficient between the primary and secondary windings, thereby reducing the loss of the drive signal output by the power supply in the primary winding and improving the power supply efficiency of the VRM to the load. Figure 5-3 The coupled inductor shown can also use the first mounting method of the primary winding described above, that is, the primary winding also includes other sides wound on the outer magnetic core post. For example... Figure 5-3 As shown, the current I1 in the primary winding 1 and the current I2 in the primary winding 2 are in opposite directions. The secondary winding is U-shaped and placed above the primary winding, thus forming a positive coupling structure with both primary windings. In this way, the coupled inductor structure can make the most of the core area.

[0082] and Figure 5-1 and Figure 5-2 The coupling inductors shown are all different. Figure 5-3 In the coupled inductor shown, multiple windings are parallel to the xy plane, that is, parallel to the surface of the circuit board; or in other words, the extension direction of the windings in the coupled inductor structure is parallel to the surface of the circuit board. Therefore, Figure 5-3 The coupled inductor shown can be understood as using the second mounting method of the coupled inductor described above. In this way, even if the input and output terminals of the windings are on different surfaces of the coupled inductor structure—for example, the input and output terminals of primary winding 1 and primary winding 2 are located on the surface of the coupled inductor facing the y-direction and the surface facing the -y-direction (i.e., the opposite direction to the y-direction), respectively—it is still beneficial to ensure that the input and output terminals of the windings are at the same height relative to the circuit board. This helps to shorten the length of the lead wires between the winding input and the PCB, reduce the loss of the drive signal in the primary winding, and thus improve the power supply efficiency of the VRM to the load. Furthermore, having multiple windings parallel to the circuit board also allows the coupled inductor structure to be embedded within the circuit board, using the traces in the circuit board as windings, thereby further reducing the height of the VRM on the board. For example, the magnetic core can be constructed by embedding it. Figure 5-3 The coupled inductor structure is shown, and then layers are added and interconnected on the magnetic core using PCB technology.

[0083] and Figure 5-1 and Figure 5-2 The coupling inductors shown are all different. Figure 5-3 In the coupled inductor shown, the surfaces (i.e., the first coupling surface) on which the target side of the primary winding 1 is positioned opposite (or closest to) the first side of the secondary winding, and the surfaces (i.e., the second coupling surface) on which the target side of the primary winding 2 is positioned opposite (or closest to) the second side of the secondary winding, are respectively parallel to the xy plane. That is, the first and second coupling surfaces are respectively parallel to the cross-section of the magnetic core column or the plane containing the secondary winding. This allows the winding to be laid flat on a plane parallel to the winding (such as the xy plane), reducing the thickness of the winding along a direction perpendicular to this plane (such as the z-direction), thereby reducing the height of the magnetic core column (i.e., perpendicular to its cross-section, such as the z-direction). Thus, as... Figure 5-3 As shown, mounting the coupled inductor structure on the circuit board using the second mounting method helps reduce the size of the coupled inductor structure along the thickness direction of the circuit board (i.e., the z-direction), thereby reducing the thickness of the single board. Optionally, Figure 5-3 The coupled inductor shown can also be mounted on the circuit board using the first mounting method for coupled inductors described above.

[0084] This application does not limit the relative size relationship between the cross-sectional areas of the three core pillars. For example, the cross-sectional areas of core pillar 2 and core pillar 3 can be the same. For example, the cross-sectional area of ​​core pillar 1 can be greater than the cross-sectional areas of core pillar 2 and core pillar 3, respectively. Optionally, the cross-sectional area of ​​core pillar 1 can be 1.5 times or 2 times the cross-sectional area of ​​core pillar 2, and similarly, the cross-sectional area of ​​core pillar 1 can be 1.5 times or 2 times the cross-sectional area of ​​core pillar 3. Alternatively, the cross-sectional area of ​​core pillar 1 can be equal to the cross-sectional area of ​​core pillar 2, and the cross-sectional area of ​​core pillar 1 can be equal to the cross-sectional area of ​​core pillar 3.

[0085] The above example uses a magnetic core made of a material with low permeability. Since materials with low permeability have higher saturation magnetic flux density, they are better suited for applications where size and height are critical. For applications with high efficiency requirements, coupled inductor structures can also use materials such as ferrite with high permeability and low loss to fabricate the magnetic core.

[0086] Figures 5-1 to 5-3 Taking the example of three magnetic core pillars located in the same plane, or in other words, the three magnetic core pillars arranged sequentially along the x-direction, magnetic core pillar 1 can be called the middle magnetic core pillar, and magnetic core pillars 2 and 3 can be called the outer magnetic core pillars. Optionally, the three magnetic core pillars do not have to be located in the same plane.

[0087] like Figures 5-1 to 5-3The coupled inductor shown in this application integrates two primary windings and one secondary winding on a three-limb magnetic core. When the two primary windings of this coupled inductor are used as inductors in a two-phase buck circuit in a VRM, it not only facilitates the magnetic coupling between the inductors in the two-phase buck circuit implemented in Solution 1, but also achieves the electrical coupling between the two-phase buck circuits implemented in Solution 2 through the secondary winding of the coupled inductor. This facilitates the realization of hybrid magneto-electric coupling, which, compared to Solutions 1 and 2, significantly improves the transient response performance and increases circuit efficiency. Furthermore, compared to Solution 2, reducing the number of secondary windings—for example, by half—not only reduces the core size, thereby reducing the layout space occupied by the VRM on the board, but also reduces the loss of induced current in the secondary windings, improving the power conversion efficiency of the VRM.

[0088] Different application scenarios have different performance requirements, and the different implementation methods of the coupled inductors introduced above can be combined arbitrarily to adapt to various application scenarios.

[0089] Coupled inductors can be discrete inductors that are mounted on the surface of a circuit board. Alternatively, in some examples, coupled inductors can be embedded inside the circuit board and can be fabricated during the circuit board manufacturing process.

[0090] This application does not limit the shape or form of the magnetic core material. For example, the magnetic core can be made of one or more types of magnetic core materials, including at least one of paste, film, and powder. This application does not limit the manufacturing method of the winding. For example, the winding can be enameled flat wire or PCB winding, etc.

[0091] The three windings of the coupled inductor provided in this application are wound on three different magnetic core pillars. Compared with Solution 3 introduced above, by avoiding winding different windings on the same magnetic core pillar, it is beneficial to reduce the requirements for the processing accuracy of the magnetic core.

[0092] The coupled inductor provided in this application, by using the first and / or second methods to increase leakage inductance, can increase the leakage inductance of the primary winding in a three-pillar magnetic core. This not only helps avoid the need for an additional series Lc for the secondary winding in the VRM, but also helps reduce the physical size of the VRM and improve the integration density of the circuit board. Compared with Solution 4, by reducing the number of core pillars in the magnetic core, it is beneficial to reduce the size of the coupled inductor structure, reduce the number of air gaps that need to be processed, and reduce the processing difficulty; in terms of the core structure, it eliminates the need for additional Lc for the secondary winding in the VRM. Figure 4The two outer pillars of the magnetic core shown are replaced with a section of the core pillar (or middle pillar) where the secondary winding is wound. This can be essentially regarded as converting the two outer pillars of the magnetic core corresponding to Solution 4 into the part of the middle pillar that is not surrounded by the secondary winding (e.g., Figure 5-1 As shown in S2), on the one hand, under the same coupling coefficient, according to the first method of increasing leakage inductance, the cross-sectional area of ​​the region on the central column not surrounded by the secondary winding (i.e., S2) is greater than that of the secondary winding. Figure 4 The cross-sectional area of ​​the outer column of the single magnetic core shown is easier to process. On the other hand, according to the first method of increasing leakage inductance, the coupling coefficient is adjusted by adjusting the cross-sectional area of ​​the secondary winding surrounding the middle column. There is no need to change the air gap size between the magnetic core column and the cover plate. Therefore, the air gap size between different magnetic core columns and the cover plate can be the same, which reduces the processing difficulty and increases the processing feasibility.

[0093] The structure of the coupled inductor provided in this application was introduced above. The VRM using this coupled inductor is described below. This VRM can also be called a transient inductor VRM based on a hybrid coupled inductor.

[0094] Figure 6 The circuit topology of a VRM using the coupled inductor provided in this application is illustrated schematically. For example... Figure 6 As shown, the VRM can include n-phase buck converter circuits, where n is a positive integer greater than 1. The input of each buck converter circuit is connected to a power supply, and its output is connected to a chip. The current input from the n-phase buck converter circuit to the chip is I. L The chip can be replaced with other types of loads as needed. Each buck circuit includes a switching circuit and a filtering circuit. The filtering circuit includes a capacitor and an inductor. The switching circuit of buck circuit i includes switches Ki_1 and Ki_2, where i is a positive integer less than or equal to n. The two switches are connected in a half-bridge configuration. The top and bottom of the half-bridge are connected to the input voltage of the power supply and ground, respectively. The middle of the half-bridge arm is connected to the first end of an inductor. The second end of the inductor is connected to one end of the output capacitor C and the power supply pin of the chip. The other end of the capacitor is grounded.

[0095] Taking an even number of n as an example, the VRM may also include n / 2 inductors connected in series (referred to as secondary inductors Lc), and the n / 2 inductors connected in series can be grounded at both ends. Furthermore, the inductors in the two buck circuits and one secondary inductor correspond to one coupling inductor provided in this application, and accordingly, the VRM includes n / 2 coupling inductors.

[0096] Alternatively, the VRM can be considered to consist of n / 2 circuit units. Figure 6 A circuit unit in the VRM is marked with a dashed box. This circuit unit is described below. Figure 6 As shown, the circuit unit includes a coupled inductor, which can be any coupled inductor provided in this application, for example... Figures 5-1 to 5-3 The coupled inductor shown in any of the attached figures. Figure 6 In this context, Np1 and Np2 can represent the two primary windings in the coupled inductor, Ns can represent the secondary winding in the coupled inductor, M10 represents the coupling between the primary winding Np1 and the secondary winding Ns, M20 represents the coupling between the primary winding Np2 and the secondary winding Ns, and M represents the coupling between the primary winding Np1 and the primary winding Np2.

[0097] like Figure 6 As shown, the circuit unit also includes two switching circuits, the outputs of which are connected to the inputs of the primary winding Np1 and the primary winding Np2, respectively. The switching circuits connected to the primary winding Np1 include switches K1_1 and K1_2 in a half-bridge configuration, and the switching circuits connected to the primary winding Np2 include switches K2_1 and K2_2 in a half-bridge configuration. The top of the half-bridge is the input terminal of the switching circuit, and the bridge arm of the half-bridge is the output terminal of the switching circuit.

[0098] like Figure 6 As shown, the output terminals of the primary winding Np1 and Np2 are connected to the output capacitor C and the chip, respectively. Optionally, the circuit unit can be considered to also include capacitor C or a portion of capacitor C.

[0099] The VRM also includes a controller for controlling the switching state of the switches in the switching circuit. This application does not limit the type of switch in the VRM; for example, the switch can be an insulated gate bipolar transistor (IGBT), a metal-oxide-semiconductor field-effect transistor (MOSFET), or a silicon carbide metal-oxide-semiconductor field-effect transistor (SiC MOSFET), etc.

[0100] The controller in this application embodiment is an application-specific integrated circuit (ASIC), a programmable logic device (PLD), a digital signal processor (DSP), or a combination thereof. The PLD can be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof; this application embodiment does not specifically limit the type of PLD.

[0101] The controller can control the phase shift angle and / or amplitude of the drive signal input to Np1 by controlling the switching states of switches K1_1 and K1_2 respectively. In one example, the controller can control the drive signals input to the primary windings of different buck circuits to have different phase shift angles. For example, the phase shift angle of the drive signal of the i-th phase buck circuit is (i*360° / n). Optionally, the phase shift angles of the drive signals input to two primary windings belonging to the same circuit unit or corresponding to the same coupled inductor may differ or interleave by 180°. As an example, assume n is 8. Figure 6 In the circuit, the step-down circuit with current I1 can be the first phase step-down circuit, the step-down circuit with current I2 can be the fifth phase step-down circuit, the step-down circuit with current In-1 can be the fourth phase step-down circuit, and the step-down circuit with current In can be the eighth phase step-down circuit.

[0102] exist Figure 6 In the same coupled inductor of the VRM shown, Np1 and Np2 can form negative coupling, Np1 and Ns can form positive coupling, and Np2 and Ns can form positive coupling, which helps to reduce current ripple and improve dynamic response.

[0103] Furthermore, by using the first method to increase leakage inductance and / or the second method to increase leakage inductance in the coupled inductor, it is beneficial to increase the leakage inductance of Np1 and Np2, thereby achieving the desired effect. Figure 6 As shown, this helps to avoid adding other inductors besides the coupling inductor in the series circuit of Ns.

[0104] exist Figure 5-1 and Figure 5-3 In the coupled inductor shown, the secondary winding N sThe ratio between the cross-sectional area surrounding the magnetic core column 1 and the total cross-sectional area of ​​the magnetic core column 1 can be adjusted to adjust the coupling coefficient between the primary and secondary windings or to adjust the leakage inductance of the primary winding, thereby adjusting the steady-state and transient characteristics of the VRM. When the secondary winding N s The larger the cross-sectional area surrounding the magnetic core post 1, the larger the primary-secondary coupling coefficient and the smaller the leakage inductance. Therefore, the stronger the electrical coupling between the primary windings of the two-phase step-down circuit in the VRM, the larger the steady-state ripple and the stronger the transient response characteristics of the corresponding circuit. Conversely, the smaller the cross-sectional area of ​​the secondary winding Ns surrounding the magnetic core post 1, the smaller the primary-secondary coupling coefficient and the larger the leakage inductance. The weaker the electrical coupling between the primary windings of the two-phase step-down circuit in the VRM, the smaller the steady-state ripple and the weaker the transient response characteristics of the corresponding circuit. When the secondary winding does not surround the magnetic core post 1 at all, the primary-secondary coupling coefficient is approximately 0. In this case, the VRM is equivalent to a two-phase magnetically coupled VRM. The circuit topology in this case can be found by referring to... Figure 2 .

[0105] The above example uses a two-phase coupled inductor with three windings. In some examples, based on the same concept, more windings can be integrated into the coupled inductor as needed, and the number of core posts can be increased accordingly, for example, to form a three-phase coupled inductor. Accordingly, a single circuit unit of the VRM can include more components from the step-down circuit.

[0106] This application does not limit the application scenario of the VRM. For example, the VRM can be integrated with one or more chips on a circuit board to power one or more chips, or multiple VRMs can power one chip. The output terminal of the VRM can be connected (or directly connected) to the power supply pin of the corresponding chip. The VRM can be deployed around the chip on the circuit board as a power supply solution. The VRM provided in this application can be applied to low-voltage, high-current chip power supply applications, and is particularly suitable for chip power supply scenarios with high power density requirements and high chip current dynamics.

[0107] A circuit board integrating a VRM and a chip (referred to as a circuit module) can be installed in an electronic device. One or more of these circuit modules can be installed inside the casing of the electronic device. This application does not limit the type of electronic device; for example, the electronic device can be an ICT-related hardware product, such as a server or data center device, or an artificial intelligence (AI) inference platform device.

[0108] Currently, data center rack power supply system solutions mainly include two types: 12V and 48V bus voltage. Figure 7-1The 12V bus voltage architecture is a traditional power supply system scheme. In the attached diagram, AC / DC converters are used to convert alternating current (AC) to direct current (DC), DC / AC converters are used to convert DC to AC, and DC / DC converters are used to convert DC to DC, for example, converting a fixed DC voltage to a variable DC voltage. Figure 7-1 As shown, the server rack may include AC / DC converters, DC / DC converters, multiple VRMs, and multiple chips. The AC / DC and DC / DC converters are used to convert the AC power input from the uninterruptible power system (UPS) into 12V DC power, and the multiple voltage regulators (VRs) are used to convert the 12V bus voltage to a lower voltage that meets the requirements of the connected chips. Figure 7-1 In this context, different VR devices can be integrated on the same or different circuit boards. Figure 7-1 Taking the VR connection to a CPU, memory, and chipset as an example, this application does not limit the type of chip connected to the VR, nor does it limit the number of VRMs in the server rack. The VRM provided in this application can be applied to at least one VR.

[0109] Figure 7-2 The rack power supply boosts the bus voltage from 12V to 48V, while the bulky online uninterruptible power supply (UPS) is replaced by a local compact DC 48V UPS located close to the load side. This results in lower bus copper losses (due to increased bus voltage and reduced current) and fewer power conversion stages, contributing to improved overall efficiency. The 48V powered VR circuit is divided into two stages (e.g., Figure 7-3 (as shown) and single-level (such as) Figure 7-4 As shown, there are two structures. Figure 7-4 As shown, a single-stage VR is used to convert a voltage of 40V to 60V to a voltage of 0.8V to 1.8V. For example... Figure 7-4 The two-stage VR structure first uses an intermediate bus converter (IBC) to transform 48V into 5-12V, and then uses a second stage to step down to 0.8-1.8V. The first stage (or first-stage VR) mainly achieves a large voltage drop, while the second stage (or second-stage VR) mainly meets the low voltage, high current, and high dynamic response requirements of the load chip, offering the advantage of structural flexibility. Regardless of the bus voltage, the second stage is always based on a voltage regulator (VR) architecture, meaning that all server chips use VR power supply.

[0110] The VRM provided in this application can be directly used as the VR section of a 12V bus voltage structure, or as the downstream section of a two-stage VR structure corresponding to a 48V bus voltage architecture (e.g., 3.3V-VR and / or 5V-VR). This invention can also be used in other application scenarios for multiphase buck converters of other power and voltage levels to improve circuit transient performance and reduce circuit size.

[0111] The terms "first," "second," etc., used in this application description are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. It should be understood that in this application, "at least one" means one or more, and "more than one" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can indicate three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can mean: a, b, c, "a and b," "a and c," "b and c," or "a and b and c," where a, b, and c can be single or multiple.

[0112] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A voltage regulator, characterized in that, The voltage regulator includes a capacitor and at least one circuit unit, the circuit unit including a coupled inductor structure and two switching circuits; The coupled inductor structure includes a magnetic core and multiple windings integrated on the magnetic core, wherein the magnetic core has three core pillars, and the multiple windings include secondary windings and two primary windings, and the multiple windings are respectively wound on different core pillars; The input terminals of the two switching circuits are respectively used to connect to the power supply, the output terminals of the two switching circuits are respectively connected to the input terminals of the two primary windings, the output terminals of the two primary windings are respectively connected to the capacitor, and the output terminals of the two primary windings are respectively used to connect to the load.

2. The voltage regulator according to claim 1, characterized in that, The secondary winding is wound on the first of the three magnetic core pillars; and the effective cross-sectional area Ae of the secondary winding on the first magnetic core pillar is smaller than the cross-sectional area of ​​the first magnetic core pillar, and / or the material of the magnetic core structure is a metallic magnetic material, and / or the permeability of the material of the magnetic core structure does not exceed 100 henry / meter.

3. The voltage regulator according to claim 1 or 2, characterized in that, The two primary windings each include a target side, and the target sides of the two primary windings are respectively located in the first magnetic column gap and the second magnetic column gap, wherein the first magnetic column gap is located between the second magnetic column and the first magnetic column among the three magnetic core columns, and the second magnetic column gap is located between the third magnetic column and the first magnetic core column among the three magnetic core columns. The two primary windings each include other sides besides the target side, and the other sides of the two primary windings are respectively wound on the second magnetic core post and the third magnetic core post, or the two ends of the target side are the input end and the output end of the corresponding primary winding.

4. The voltage regulator according to claim 3, characterized in that, The coupled inductor structure is used to fix it on the surface of the circuit board or to embed it inside the circuit board, and the plurality of windings are respectively parallel to the circuit board.

5. The voltage regulator according to claim 4, characterized in that, The surface on the target side opposite to the secondary winding is parallel to the circuit board.

6. The voltage regulator according to any one of claims 1-5, characterized in that, The two switching circuits each include a first switching transistor and a second switching transistor; The first terminal of the first switching transistor is connected to the input terminal of the switching circuit, the second terminal of the first switching transistor is connected to the first terminal of the second switching transistor and the output terminal of the switching circuit, and the second terminal of the second switching transistor is grounded.

7. The voltage regulator according to any one of claims 1-6, characterized in that, The voltage regulator also includes a controller for controlling the two switching circuits respectively.

8. A circuit module, characterized in that, The circuit module includes a circuit board, one or more chips integrated on the circuit board, and one or more voltage regulators as described in any one of claims 1-7 integrated on the circuit board, wherein the one or more voltage regulators are used to power the one or more chips.

9. An electronic device, characterized in that, The electronic device includes a housing and one or more circuit modules as described in claim 8, which are installed inside the housing.

10. A coupled inductor structure, characterized in that, The coupled inductor structure includes a magnetic core and multiple windings integrated on the magnetic core. The magnetic core has three core posts, and the multiple windings include secondary windings and two primary windings. The multiple windings are respectively wound on different core posts.

11. The coupled inductor structure according to claim 10, characterized in that, The secondary winding is wound on the first of the three magnetic core pillars; and the effective cross-sectional area Ae of the secondary winding on the first magnetic core pillar is smaller than the cross-sectional area of ​​the first magnetic core pillar, and / or the material of the magnetic core structure is a metallic magnetic material, and / or the permeability of the material of the magnetic core structure does not exceed 100 henry / meter.

12. The coupled inductor structure according to claim 10 or 11, characterized in that, The two primary windings each include a target side, and the target sides of the two primary windings are respectively located in the first magnetic column gap and the second magnetic column gap, wherein the first magnetic column gap is located between the second magnetic column and the first magnetic column among the three magnetic core columns, and the second magnetic column gap is located between the third magnetic column and the first magnetic core column among the three magnetic core columns. The two primary windings each include other sides besides the target side, and the other sides of the two primary windings are respectively wound on the second magnetic core post and the third magnetic core post, or the two ends of the target side are the input end and the output end of the corresponding primary winding.

13. The coupled inductor structure according to claim 12, characterized in that, The coupled inductor structure is used to fix it on the surface of the circuit board or to embed it inside the circuit board, and the plurality of windings are respectively parallel to the circuit board.

14. The coupled inductor structure according to claim 13, characterized in that, The surface on the target side opposite to the secondary winding is parallel to the circuit board.

Images