System, method and controller for overcurrent protection
By measuring the secondary inductor current in a multiphase trans-inductor voltage regulator and generating an overcurrent fault signal, the problem of increasing inductor size in the prior art is solved, and overcurrent protection with high current throughput is achieved.
Patent Information
- Application Number
- CN202411293652.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-09-18
- Filing Date
- 2024-09-14
- Publication Date
- 2025-03-18
AI Technical Summary
The prior art is difficult to provide a high current throughput overcurrent protection solution for multiphase cross-inductor voltage regulators without increasing inductor size.
The current evaluation of the magnetized inductor supported by the primary inductor is achieved by measuring the current flowing through the secondary inductor using a first sensor device in a multiphase trans-inductor voltage regulator and generating an overcurrent fault signal by the controller based on the current.
This solution provides overcurrent protection with high current throughput without increasing inductor size, accurately assessing the current supported by the primary inductor magnetization inductor, thereby reducing the possibility of overcurrent failure.
Smart Images

Figure CN119651496A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a system and method for overcurrent protection of a power converter, and more particularly to overcurrent protection of a multi-phase trans-inductor voltage regulator. Background Art
[0002] Many modern microprocessors require very high currents. For example, in AI applications, the hardware accelerators used require high power and therefore consume high currents for short periods of time. The voltage regulators used for such accelerators must be able to handle high instantaneous peak power / currents for short periods of time.
[0003] The thermal impact on the voltage regulator is determined by the thermal design current (TDC) operating point. However, in many ASICs / GPUs, the required peak current may be twice the TDC, meaning that the regulator's inductor needs to be larger to avoid saturation. Therefore, the dependency between peak current requirements and inductor size needs to be mitigated.
[0004] Recently, multiphase trans-inductor voltage regulators (TLVRs) have been proposed for delivering power from an input voltage to an output (e.g., a point of load). Instead of using uncoupled individual inductors in each stage of a multiphase converter, each stage uses a transformer in which the primary inductor is coupled between the half-bridge and the point of load, and all secondary inductors are connected in series and referenced to ground. The coupling between the primary and secondary inductors is close to perfect, with a coupling coefficient close to unity.
[0005] Compared to conventional multiphase converters, in a TLVR, a load jump that causes an increase in the equivalent duty cycle in one phase (and therefore an increase in current in that phase) results in an increase in output current in all phases. This inherent feature of the TLVR enables a faster transient response corresponding to the significantly reduced effective output inductance of the system.
[0006] The inventors have determined that there is a need for an overcurrent protection scheme for TLVR that can achieve high current throughput without increasing the size of the inductor. Summary of the invention
[0007] According to one aspect of the present invention, a system for overcurrent protection of a multi-phase trans-inductor voltage regulator is provided. The multi-phase trans-inductor voltage regulator includes a plurality of transformers, each transformer including a primary inductor and a secondary inductor, wherein the primary inductor of each transformer is arranged between a corresponding input node and a common output node, the secondary inductors of the plurality of transformers are connected in series, and wherein each input node is connected to a corresponding phase input of a plurality of phase inputs. The system includes:
[0008] a first sensor device configured to measure a current flowing through the secondary inductor; and
[0009] A controller is configured to generate an over-current fault signal based at least in part on the measured current flowing through the secondary inductor.
[0010] As an illustration, when the current required by the load on the output node of the trans-inductor voltage regulator (TLVR) increases, the controller of the TLVR compensates for the output load (in the case of a constant frequency controller) by increasing the duty cycle of the phase input, thereby increasing the supply current to the output node. For a multi-phase TLVR, this means that the secondary inductor (i.e., the secondary winding or the inductor on the secondary side) is energized. Initially, part of the load current is supported by the leakage current of each transformer. This means that the current flowing through the secondary inductor (i.e., present on the secondary side) enables the primary inductor to (temporarily) withstand a load current that is higher than the rated inductor saturation current. This can be seen in the following equation, which is always valid:
[0011]
[0012] Among them, i ph,i (t) is the phase input current for transformer i, is the current flowing through the series connected secondary inductor / secondary winding, and is the current supported by the magnetizing inductance of the primary inductor of transformer i.
[0013] It is worth noting that after the phase current input increases, the current flowing through the secondary side decreases with time because the secondary side exhibits resistive behavior. Therefore, the current decreases to zero according to the following decay constant:
[0014]
[0015] Among them, DCR tr is the combined DCR impedance of the inductor connected on the secondary side and the distributed impedance of all secondary inductor interconnections and connections to ground, L k2,i is the leakage inductance of the secondary inductor of transformer i, L m2,i is the magnetizing inductance of the secondary inductor of transformer i, and L tr is the inductance of any other inductor on the secondary side (eg, compensation inductor).
[0016] Therefore, the controller generates an over-current fault signal based on the current flowing through the secondary inductor. The over-current fault signal may indicate a risk of an over-current fault.
[0017] For example, in the case where the current flowing through the secondary inductor (i.e., the current on the secondary side) is high, this indicates that the current built into the magnetizing inductance of each primary inductor is low. This is because the phase current through each primary inductor is partially supported by the high leakage current of each transformer. Essentially, this means that the primary inductors are each capable of handling currents greater than the rated inductor saturation current due to the high leakage current of the transformer. Therefore, in the case where the current flowing through the secondary inductor is high, an overcurrent fault signal may not be generated or is unlikely to be generated.
[0018] Conversely, in the case where there is little or no current flowing through the secondary inductor, this indicates that there is little or no leakage current in each transformer, and therefore the current supplied by the phase input is fully supported by the magnetizing inductance of the primary inductor. A high current supported by the magnetizing inductance of the primary inductor may cause an overcurrent fault. Therefore, in the case where the current flowing through the secondary inductor is low, an overcurrent fault signal may be generated or is more likely to be generated.
[0019] The output node may be connected to a load. The load may be a variable load that requires different currents over time. Thus, as the current required by the load increases, the overcurrent protection system may be able to protect the multi-phase trans-inductor voltage regulator.
[0020] The first sensor device may be any device capable of measuring the current flowing through the secondary inductor. This may include generating a voltage proportional to the current on the secondary side by the first sensor device, converting the voltage into a numerical value, and converting the numerical value into an absolute or relative current reading.
[0021] The first sensor device may be configured to measure current, and the controller is configured to generate an overcurrent fault signal in response to an increase in current provided by one or more phase inputs of the multi-phase converter.
[0022] The controller may control one or more switching elements connected to the input voltage to provide each phase input. That is, for each of the plurality of phase inputs, the controller may control one or more switching elements to provide the phase input at the corresponding input node. For example, one or more switching elements may be in the form of a half-bridge, wherein each switching element is associated with a gate driver controlled by the controller to determine a switching frequency. The switching frequency and duty cycle may determine the phase current provided to each input node.
[0023] The system may further include a second sensor device configured to measure, for at least two of the plurality of phase inputs, a phase current provided by the phase input. In this case, the controller may be configured to generate an overcurrent fault signal further based on at least one of the measured phase currents.
[0024] The measurement may be performed by a typical implementation of a controller for a multi-phase regulator, and the second sensor device hardware may form part of the controller. However, it is worth noting that in some examples the second sensor device may be provided separately from the controller.
[0025] Therefore, the controller can use both the measured current flowing through the secondary inductor and the phase current (for each measured phase current) to determine the generation of an overcurrent fault signal. That is, both the measured current flowing through the secondary conductor and the measured phase current can be used to determine whether to generate an overcurrent fault signal and / or determine the type of overcurrent fault generated.
[0026] The second sensor device may measure, for each of at least two of the phase inputs, a phase current provided to a primary inductor connected to the corresponding phase input via an input node. In some cases, the second sensor device may measure each phase current provided to each of the primary inductors. Each of the measured phase currents may then be used individually to generate an overcurrent fault signal. For example, an overcurrent fault signal may be generated if any of the measured phase currents exceeds a certain threshold.
[0027] For example, during transient operation, the phase current provided to the first primary inductor of the plurality of primary inductors may be low and thus may not cause the generation of an overcurrent fault signal. However, the phase current provided to the second primary inductor of the plurality of primary inductors may be high and thus an overcurrent fault signal is generated on this basis.
[0028] In other words, an overcurrent fault signal may be generated based on at least one of the measured phase currents satisfying a certain criterion rather than based on all (or even most) of the phase currents satisfying the criterion. This means that each measured phase current may be used individually to determine the generation of overcurrent fault information.
[0029] In some implementations, the controller may be configured to generate the over-current fault signal further based on a static over-current protection threshold.
[0030] The static overcurrent protection threshold may define a target maximum current (e.g., a rated current) through each primary inductor. For example, the static overcurrent protection threshold may reflect the amount of current that can flow through the primary inductor before the primary inductor saturates. This may indicate the amount of current that can be supported by the magnetizing inductance of the primary inductor before saturation is reached.
[0031] The static overcurrent protection threshold may be predetermined (ie, set by the manufacturer) or may be programmed into the controller.
[0032] The static overcurrent protection threshold may be equal to a rated inductor saturation current of at least one of the primary inductors.
[0033] In a case where the controller is configured to generate an overcurrent fault signal further based on the measured phase current and at least one of the static overcurrent protection thresholds, the controller can be configured to generate a fault signal in response to at least one of the measured phase currents exceeding the sum of the static overcurrent protection threshold and the measured current flowing through the secondary inductor.
[0034] That is, the current flowing through the secondary inductor / secondary winding (indicative of the load current supported by the leakage current of each transformer) enables the current provided by the phase input to exceed the static overcurrent protection threshold without triggering the overcurrent fault signal.
[0035] Essentially, this can be formulated as generating an overcurrent fault signal (i.e., an indication that the controller output inductor may be saturated or may soon be saturated) in response to the following conditions:
[0036]
[0037] Among them, i static is the static overcurrent protection threshold. Similarly, the comparison can be made based on the following conditions:
[0038]
[0039] In all cases, the presence of current flowing through the second inductor allows the phase current supplied to the primary inductor to exceed the static threshold without triggering a fault signal.
[0040] The exemplary controller may be configured to determine a dynamic overcurrent protection threshold based on the sum of the measured current flowing through the secondary inductor and the static overcurrent protection threshold. An overcurrent fault signal may then be generated in response to at least one of the measured phase currents exceeding the dynamic overcurrent protection threshold.
[0041] That is, a time varying threshold may be generated below which the current through the primary inductor should be maintained to avoid saturation. This dynamic threshold may be used to adjust the point at which an overcurrent fault signal is generated.
[0042] In some embodiments, the first sensor device may include a DC impedance current sensing circuit configured to measure a current flowing through the secondary inductor.
[0043] A DC impedance (DCR) current sensing circuit may include a series connected resistor and capacitor, which are themselves placed in parallel with the secondary inductor to measure the current flowing through the secondary inductor.
[0044] Alternatively, in case the multi-phase trans-inductor voltage regulator further comprises a compensation inductor connected in series with the secondary inductor, the first sensor device may comprise a DCR current sensing circuit configured to measure the current flowing through the secondary inductor and the compensation inductor.
[0045] In this case, a DC impedance current sensing circuit is provided across the compensation inductor or across the compensation inductor and the secondary inductor.
[0046] That is, the DCR current sensing circuit may be disposed in parallel with the compensation inductor or both the compensation inductor and the plurality of secondary inductors.
[0047] A compensation inductor may be added in series with the secondary inductor in order to adjust and more finely / precisely control the current slope in the secondary inductor. Therefore, whether to include a compensation inductor will depend on the various design requirements of the TLVR. However, by providing a DCR sensing circuit across the compensation inductor, the compensation inductor can be used to determine the current flowing through the secondary side.
[0048] If a compensating inductor is not provided, it may be desirable to provide a DCR sensing circuit across the entire trace on which the secondary inductor is provided to ensure that the current flowing through the secondary side is reliably determined.
[0049] If the controller requires a voltage signal for processing, the current measured by the DCR current sensing circuit can be output in the form of a voltage.
[0050] In an exemplary embodiment, the first sensor device may further include an amplifier to amplify the output of the DC impedance current sensing circuit. The amplifier may be an instrumentation amplifier.
[0051] As an alternative, the first sensor device may comprise a transformer comprising a primary side connected in series with the secondary inductor and a secondary side terminated by a resistor.The first sensor device may comprise a voltage sensor configured to sense a voltage across the secondary side resistor.
[0052] That is, the primary inductor may be placed in series with the secondary inductor and the primary inductor is coupled to the secondary inductor. The primary inductor is terminated by a resistor to provide a readable voltage indicative of the current through the secondary inductor of the TLVR (eg, where the controller requires a voltage signal rather than current).
[0053] As another alternative, the first sensor arrangement may include a shunt resistor connected in series with the second inductor and a voltage sensor configured to sense a voltage across the shunt resistor.
[0054] In this case, the first sensor device may further include an amplifier configured to amplify the voltage across the shunt resistor. The amplifier may be an instrumentation amplifier.
[0055] Although the first sensor arrangement including the shunt resistor is lossy, it can be more accurate than the transformer and DCR sensing arrangement described above.The shunt resistor also helps to increase the decay constant with which the current flowing through the secondary inductor drops to zero.
[0056] In a case where the first sensor device is configured to sense a voltage (e.g., a voltage across a shunt resistor, or a current across a resistor on the secondary side of a transformer), the controller can be configured to receive the sensed voltage and determine a current flowing through the second inductor based on the sensed voltage.
[0057] In this case, the controller may include an analog-to-digital converter configured to convert the received sensed voltage into a numerical value describing the sensed voltage. To this end, the controller may include hardware suitable for converting the sensed voltage into a numerical value and then determining the current based on the numerical value.
[0058] A multi-phase trans-inductor voltage regulator is also provided, comprising:
[0059] a plurality of transformers, each transformer comprising a primary inductor and a secondary inductor, wherein the primary inductor of each transformer is disposed between a corresponding input node and a common output node, the secondary inductors of the plurality of transformers are connected in series, and wherein each input node is connected to a corresponding phase input of the plurality of phase inputs; and
[0060] A system for overcurrent protection as summarized above.
[0061] Therefore, the TLVR provided by the present invention can have high current throughput without requiring increased inductor size. That is, by generating an overcurrent fault signal based (in part) on the current flowing through the secondary inductor, the current supported by the magnetizing inductance of the primary inductor can be accurately evaluated.
[0062] In case the current through the secondary inductor is high, this means that the load current supported by the leakage inductance of the transformer is high, causing the current supported by the magnetizing inductance of the primary inductor to decrease. Since the current supported by the magnetizing inductance is the relevant parameter for overcurrent faults, a low current supported by this inductance means saturation and therefore a low probability of an overcurrent fault.
[0063] The multiphase trans-inductor voltage regulator may be any converter having an inductor-based topology, wherein current regulation is achieved by flux balancing across the inductor. For example, the multiphase trans-inductor voltage regulator may have a topology based on a half-bridge, a full-bridge, or a half-bridge with a current doubler rectifier. Suitable topologies may include both isolated and non-isolated topologies.
[0064] More specifically, the multi-phase trans-inductor voltage regulator may be a buck converter. In other words, the multi-phase trans-inductor voltage regulator may have a buck-derived topology.
[0065] A method for overcurrent protection of a multi-phase trans-inductor converter is also provided. The multi-phase trans-inductor converter includes a plurality of transformers, each transformer including a primary inductor and a secondary inductor. The primary inductor of each transformer is arranged between a corresponding input node and a common output node, and the secondary inductors of the plurality of transformers are connected in series. Each input node is connected to a corresponding phase input of a plurality of phase inputs. The method includes:
[0066] measuring the current flowing through the secondary inductor; and
[0067] An overcurrent fault signal is generated based at least in part on the measured current flowing through the secondary inductor.
[0068] In addition, a controller for overcurrent protection of a multi-phase trans-inductor voltage regulator is provided, the multi-phase trans-inductor voltage regulator comprising a plurality of transformers, each transformer comprising a primary inductor and a secondary inductor, wherein the primary inductor of each transformer is arranged between a corresponding input node and a common output node, the secondary inductors of the plurality of transformers are connected in series, and wherein each input node is connected to a corresponding phase input of a plurality of phase inputs, the controller being configured to:
[0069] obtaining a measured current flowing through the secondary inductor; and
[0070] An overcurrent fault signal is generated based at least in part on the measured current flowing through the secondary inductor. BRIEF DESCRIPTION OF THE DRAWINGS
[0071] The present invention will now be described by way of example with reference to the accompanying drawings, in which:
[0072] Figure 1 is a circuit diagram showing a multi-phase trans-inductor voltage regulator (TLVR) according to a general example;
[0073] Figure 2 yes Figure 1 An equivalent circuit diagram of one of the transformers illustrated in;
[0074] Figure 3 It shows that the Figure 2The equivalent circuit of Figure 1 Circuit diagram of TLVR;
[0075] Figure 4 Yes Description Figure 3 A diagram of waveforms of a single phase of a circuit during steady-state operation;
[0076] Figure 5 Is a description Figure 3 A diagram of waveforms of a single phase of a circuit during transient operation;
[0077] Figure 6 is a simplified circuit diagram of the operation of the controller for overcurrent protection of TLVR;
[0078] Figure 7 is a circuit diagram showing a TLVR and a first sensor device according to an example;
[0079] Figure 8 is a circuit diagram showing a TLVR and a first sensor device according to another example;
[0080] Fig. 9 Explained Figure 8 Different options for sensing points of the first sensor device.
[0081] Fig.10 is a circuit diagram showing a TLVR and a first sensor device according to yet another example;
[0082] Fig.11 is a circuit diagram showing N-phase of a multi-phase buck converter with TLVR; and
[0083] Fig.12 is a generic block diagram of a TLVR for any inductor-based converter topology configured for multi-phase operation.
[0084] It should be noted that these drawings are schematic and not drawn to scale. For the sake of clarity and convenience of drawing, the relative dimensions and proportions of the various parts of these drawings have been shown exaggerated or reduced in size. DETAILED DESCRIPTION
[0085] In order to understand the present disclosure, it is important to understand the operation of a multi-phase trans-inductor voltage regulator (TLVR). Figure 1 This is the circuit diagram of TLVR 100-1.
[0086] TLVR 100-1 includes a plurality of transformers 110-1, ..., 110-N. Each transformer 110-1, ..., 110-N includes a primary inductor 112-1, ..., 112-N on the primary side of TLVR 100-1 and a secondary inductor 114-1, ..., 114-N on the secondary side of TLVR 100-1. Each of the primary inductors 112-1, ..., 112-N is disposed between a corresponding input node 120-1, ..., 120-N (connected to a corresponding phase input) and a common output node 130. That is, the primary inductors 112-1, ..., 112-N are each connected to a different phase input, but are connected to a common output. Each of the secondary inductors 114-1, ..., 114-N is connected / arranged in series with the other secondary inductors, and the secondary inductors are commonly referenced to ground.
[0087] Accordingly, the TLVR 100-1 delivers power from the input voltage of the multi-phase supply to a common output 130, which can be considered a load point. Instead of using uncoupled individual inductors connected to each phase input (as in conventional voltage regulators), each phase input is input to a transformer 110-1, ..., 110-N. The coupling coefficient between each primary inductor / winding 112-1, ..., 112-N and each secondary inductor / winding 114-1, ..., 114-N in the same transformer 110-1, ..., 110-N can be close to one, nearly perfect.
[0088] In some cases, each input node 120-1, ..., 120-N can be connected to one or more switches. Where multi-phase operation is used, all phase inputs are coupled to enhance current conversion rate during changes in the root mean square (RMS) voltage on each input node 120-1, ..., 120-N.
[0089] A controller of the TLVR 100-1 may be provided that controls the switching frequency and / or duty cycle of the switches connected to each input node 120-1, ..., 120-N. The controller may output control signals to the drive circuits of the switches to control the phase input to each input node 120-1, ..., 120-N.
[0090] In the presence of a load jump (i.e., an increase in current / power required by the load), a pulse width modulation (PWM) duty cycle associated with one or more of the phase inputs (i.e., the phase inputs of one or more switches connected to input nodes 120-1, ..., 120-N) may be increased. This may be performed by the controller mentioned above.
[0091] For TLVR 100-1, since all phases are indirectly connected by the secondary inductors 114-1, ..., 114-N arranged in series, an increase in current provided in one phase input causes an increase in output current in all phases. This feature of TLVR 100-1 enables a faster transient response corresponding to the significantly reduced effective output inductance of the system.
[0092] Optionally, the TLVR 100-1 may further include a compensation inductor 140 disposed / connected in series with the secondary inductors 114-1, ..., 114-N. The compensation inductor 140 may be designed to adjust the current slope in the secondary inductors 114-1, ..., 114-N.
[0093] Now refer to Figure 2 , each of the transformers 110-1, ..., 110-N (i.e., across the inductor) can be described by the illustrated equivalent circuit / model 10. The equivalent circuit 10 can be considered as having a certain coupling (between the primary and secondary windings / inductors) and having a magnetizing inductance L mi and leakage inductance L k1 The classic transformer.
[0094] By considering the equivalent circuit 10, the following equation can be established:
[0095]
[0096] L k1 =(1-k 1,2 )L1 [6
[0097] L m2 =k 1,2 L2 [7]
[0098] L k2 =(1-k 1,2 )L2 [8]
[0099] In k i-1,i >0, the turns ratio of the transformer has the following relationship:
[0100]
[0101] After model analysis using the above equations, a voltage proportional to the voltage applied by all phases is induced at each input node. In other words, each input to one transformer 110-1...110-N steps up the voltage of the other transformers 110-1,...,110-N of the TLVR 100-1.
[0102] Figure 3 Shown with Figure 2 The equivalent model of 10 combined Figure 1 The equivalent model 100-2 of the TLVR 100-1 assumes that the L of each transformer 110-1, ..., 110-N is k < <L m .
[0103] Figure 3 Each secondary inductor 114-1, ..., 114-N in the equivalent model 100-2 shown in can be regarded as an induced voltage source. When the voltage on the input node 120-1, ..., 120-N supplied by the phase input is pulled to the input voltage potential (for example, by turning on the high-side switch of the buck converter connected to the input node 120-1, ..., 120-N), there is a positive voltage applied to the magnetizing inductance of the primary inductor 112-1, ..., 112-N. When the input node 120-1, ..., 120-N is connected to the GND potential (for example, by turning off the high-side switch of the buck converter connected to the input node 120 and turning on the low-side switch of the buck converter), the applied voltage is negative. On each secondary inductor 114-1, ..., 114-N, each voltage causes a current that compensates for the change in the inductor 140, shown as In case of changes in the secondary side current, this will also cause the current supplied by each phase input to trend up or down.
[0104] Considering Figure 4 and Figure 5 The operation can be further understood in the case of Figure 4 and Figure 5 The current waveforms flowing through different parts of the TLVR operating in steady state (i.e., constant DC load) and transient conditions (i.e., when a load transition occurs) are shown. Specifically, the current magnitude supported by the magnetizing inductance of the primary inductor 112-1, ..., 112-N is shown. The current through the compensation inductor 140 and the current I on the input nodes 120 - 1 , . . . , 120 -N ph .
[0105] like Figure 4 As seen in FIG. 1 , with the TLVR in steady-state operation, the primary inductor 112 - 1 , . . . , 112 -N of each transformer carries a DC load current The AC current component flowing into each input node 120-1, ..., 120-N depends on the primary inductor current ripple and the current ripple of the compensation inductor 140. During steady-state operation, the compensation inductor 140 and the secondary inductors 114-1, ..., 114-N do not have any DC current flowing therethrough.
[0106] like Figure 5As seen in FIG. 1 , in the presence of a load change (i.e., a change in the current required by the load), the controller of the switch connected to the input node 120-1, ..., 120-N of each primary inductor 112-1, ..., 112-N can change the duty cycle of the switch to compensate for the output voltage change, thereby increasing I ph The reflected voltage on each transformer 110-1, ..., 110-N has a significant effect on the current on the secondary side. The charging is performed with a slope depending on the inductance of the compensation inductor 140 and the sum of all secondary inductor voltages reflected from the primary side / primary inductors 112-1, ..., 112-N. Then, the changing current flowing through the series-connected secondary inductors 114-1, ..., 114-N also causes the current on each input node 120-1, ..., 120-N of the primary inductors 112-1, ..., 112-N to increase.
[0107] During this time interval, the current flowing to the output node 130 is supported by the leakage current of each of the transformers 110-1, ..., 110-N. In fact, when the current built into the secondary inductor 114-1, ..., 114-N increases rapidly, this compensates for a portion of the DC flux entering the primary side magnetizing inductance necessary to carry the actual load current.
[0108] However, because the secondary inductors 114-1, ..., 114-N and the associated physical connections are subject to the distributed impedance of the inductor DCR impedance and the interconnection of all secondary inductors 114-1, ..., 114-N to the ground path (at Figure 3 It is represented as DCR tr ) and exhibits resistance behavior, so the current through the secondary inductor 114-1, ..., 114-N is It decreases to zero with a certain decay constant. The decay constant depends on the following formula:
[0109]
[0110] in, is the magnetizing inductance on each secondary inductor, is the secondary inductor leakage inductance, and is the peak current on the secondary side / through the secondary inductor during a load transition.
[0111] According to this formula, the decay constant of the current flowing through the secondary inductor after the load jump is considered to be:
[0112]
[0113] By knowing this decay constant, an appropriate overcurrent protection strategy can be developed.
[0114] In addition, and taking into account Figure 3 The equivalent TLVR circuit in the following equation is always valid:
[0115]
[0116] This is Figure 5 Essentially, the current built into the magnetizing inductance of the primary inductors 112-1, ..., 112-N Depending on the load jump and the above-mentioned decay constant. Therefore, during a load jump, each individual primary inductor 112-1, ..., 112-N of TLVR 100-2 may be subjected to a current higher than the inductor saturation current for a period of time that depends on the decay constant and the magnitude of the load jump.
[0117] For some applications, it may be necessary to operate the TLVR 100-2 at a higher than rated current (particularly, the rated current of one or more of the primary inductors) for a short period of time due to a load step. Typically, the TLVR 100-2 will only operate at a higher than rated current for a short period of time, where the amount of time is generally inversely proportional to the magnitude of the load current.
[0118] For example, the Open Compute Project (OCP) Accelerator Module (OAM) 2.0 specification states that a voltage regulator must be able to handle / withstand instantaneous peak power that is higher than the average power absorbed by the digital load. Therefore, in the case where the TLVR 100-2 is used as a voltage regulator in accordance with an OAM-compliant module, it may be necessary to operate the TLVR 100-2 at 2 times the thermal design power for less than 500 μs, operate the TLVR 100-2 at 1.6 times the thermal design power for less than 2 ms, and operate the TLVR 100-2 at 1.2 times the thermal design power for less than 10 ms.
[0119] While the TLVR 100-2 facilitates these temporary increases in peak power, it is important to note that the time that the rated current is exceeded should be carefully controlled based on the decay constant of the current flowing through the secondary inductors 114-1, ..., 114-N. Figure 5 As seen in , the amount of current flowing through the secondary inductors 114-1, ..., 114-N decays according to the decay constant τ. This reflects the amount of time the input to the TLVR 100-2 can remain above the rated current.
[0120] In other words, the TLVR 100-2 enables a temporary excursion of the load current above the rated current because the load current is supported by the leakage inductance in each of the transformers 110-1, ..., 110-N. However, since the current flowing through the secondary inductors 114-1, ..., 114-N decays according to the decay constant τ, this increase in load current must be temporary. Therefore, the decay constant should be designed to keep the current flowing through the secondary inductors 114-1, ..., 114-N above the desired value at least in the presence of transients caused by load steps.
[0121] Examples according to the present disclosure involve utilizing the physical behavior described above during transient operation of the TLVR 100. By sensing the current on the secondary side / flowing through the secondary inductors 114-1, ..., 114-N, an appropriate overcurrent protection strategy can be provided (e.g., generating an appropriate overcurrent fault signal). In some implementations, in addition to the current flowing through the secondary inductors 114-1, ..., 114-N, the magnetizing current of each primary inductor 112-1, ..., 112-N can be reconstructed by sensing the actual phase current on the input node 120.
[0122] Typically, in an inductor-based (i.e., non-trans-inductor-based) voltage regulator, each phase is protected from overcurrent to avoid inductor saturation and / or silicon failure. This is typically achieved by a static overcurrent protection threshold (depending on the inductor and power MOSFET used to implement the regulator). A typical controller senses the current provided by each phase input to ensure current sharing and output voltage control. However, a controller according to the present disclosure additionally senses or otherwise obtains a measured current flowing through the secondary inductors 114-1, ..., 114-N, and generates an overcurrent fault signal based at least in part on the measured current.
[0123] Examples according to the present disclosure also utilize knowledge of the current on the secondary side / flowing through the secondary inductors 114 - 1 , ..., 114 -N. Thus, in contrast to controllers of typical conventional multi-phase voltage regulators, examples according to the present disclosure also obtain a measurement of the current on the secondary side.
[0124] One aspect of configuring a controller for generating an overcurrent fault signal may be based on Figure 6 The circuit shown in Fig.10In this case, a static overcurrent protection threshold and the measured (variable) current flowing through the secondary inductor 114-1, ..., 114-N are added to determine a dynamic overcurrent protection threshold. This dynamic overcurrent protection threshold is then compared with the phase current provided by one of the phase inputs (i.e., the current provided at the input nodes 120-1, ..., 120-N of the primary inductor 112-1, ..., 112-N). If the phase current exceeds the dynamic overcurrent protection threshold, an overcurrent fault signal is generated.
[0125] This may be accomplished for each transformer 110 - 1 , . . . , 110 -N of the TLVR. That is, each of the phase currents provided by the phase inputs may be individually compared to a dynamic overcurrent protection threshold to determine whether to generate an overcurrent fault signal.
[0126] The controller for generating the overcurrent fault signal may be the same as the controller for controlling the switching frequency and / or duty cycle of the switches connected to each input node 120-1, ..., 120-N. Alternatively, the controller for generating the overcurrent fault signal may be implemented using hardware separate from the switching controller. For example, the controller for generating the overcurrent fault signal may be implemented in the analog domain, while the controller for switching is typically digital.
[0127] Next, Figures 7 to 10 Various options and concepts are shown related to sensing / measuring current flowing through the secondary inductors 114-1, ..., 114-N of the TLVR 100-1. In each illustrated case, the first sensor device 200-1, 200-2, 200-3 outputs a voltage signal (representative of the measured current) to the controller 300 for generating an overcurrent fault signal based at least in part on the measured current.
[0128] Specifically, Figure 7 A first sensor device 200-1 is shown, which uses a current transformer 210 in series with a plurality of secondary inductors 114-1, ..., 114-N (and an optional compensation inductor 140). The primary side of the transformer 210 (i.e., the primary inductor) is connected in series with the secondary inductors 114-1, ..., 114-N. A resistor 220 may be used to terminate the secondary side of the transformer 210 to convert the current information of the secondary side of the transformer 210 into voltage information. This may be desirable, for example, where a voltage signal is required by the controller.
[0129] Using this method, the output voltage is proportional to the current flowing through the secondary inductors 114-1, ..., 114-N according to the following equation:
[0130] V S =N·Rs I Ltr
[13]
[0131] Figure 8 Another example of a first sensor device 200-2 is shown, which uses a (lossless) DC impedance (DCR) current sensing circuit. All of the components on the secondary side of the TLVR 100-1, whether it is the secondary inductor 114-1, ..., 114-N, the compensation inductor 140 or any traces connecting the various components have an inherent impedance. This inherent impedance is used to measure the current flowing through the secondary inductor 114-1, ..., 114-N.
[0132] In one implementation, the DCR sensing circuit uses a capacitor 240 and a resistor 230 arranged in series and connected across (at least a portion of) the inherent impedance of the secondary side of the TLVR 100-1. This enables sensing across the capacitor 240 a voltage signal proportional to the current flowing through the secondary inductors 114-1, ..., 114-N.
[0133] An amplifier 250 (eg, an instrumentation amplifier) may be used across capacitor 240 to amplify the signal.
[0134] Therefore, if the capacitance of capacitor 240 and the impedance of resistor 230 are properly selected, a voltage signal can be generated according to the following formula:
[0135] V s =DCR tr ·G d I Ltr
[14]
[0136] Correct selection means that the capacitance and impedance must be chosen to cancel the voltage generated by the parasitic impedance so that only the voltage based on the current is measured.
[0137] In the TLVR circuit 100-1, there is usually a long physical connection between the secondary inductors 114-1, ..., 114-N and the ground. Fig. 9 As shown in , a DCR current sensing circuit can be applied across different parts of the connection (ie, in parallel with different parts of the connection).
[0138] Where the secondary side includes a compensating inductor 140, as shown in (i), the DCR sensing circuit may be applied across the entire trace (i.e., physical connection) and the compensating inductor 140. This may give the most accurate readings since the DCR sensing circuit is applied across the maximum available DC impedance. However, it may be necessary to measure the parasitic impedance of the trace and the compensating inductor 140 to ensure accurate results.
[0139] Alternatively, as shown in (ii), a DCR sensing circuit may be provided only across the compensation inductor 140. This may be desirable to ensure that the capacitance and impedance of the DCR sensing circuit are selected directly, since the DC impedance of the physical compensation inductor 140 may be well known (compared to the DCR of the entire range of the physical connection).
[0140] In some examples, TLVR 100-1 may not have a physical compensation inductor 140. In this case, it may be desirable to place the DCR sensing circuit across as much of the secondary side of TLVR 100-1 as possible to maximize the parasitic impedance provided by the DCR sensing circuit. This situation is described in (iii).
[0141] Fig.10 An alternative first sensor device 200-3 is shown which uses a shunt resistor 260 arranged in series with the secondary inductors 114-1 ... 114-N and the (optional) compensation inductor 140. This sensor device 200-3 has losses due to the shunt resistor 260, but ensures the highest accuracy of the described sensor devices 200-1, 200-2, 200-3.
[0142] Additionally, the shunt resistor 260 also affects the decay constant associated with the decay of the current flowing through the secondary inductors 114-1 ... 114-N. Essentially, this means that the sensor device 200-3 changes the behavior of the TLVR circuit 100-1.
[0143] An amplifier 270 (eg, an instrumentation amplifier) may be used to assist in sensing the voltage across the shunt resistor 260 according to the following equation:
[0144] V s =R shunt ·G s I Ltr
[15]
[0145] Of course, the current sensing strategies outlined above should not be considered exhaustive or limiting.Any means capable of determining the current flowing through the secondary inductors 114-1, ..., 114-N may be employed.
[0146] The proposed concept for overcurrent protection can be applied to all trans-inductor based converters. Fig.11 The circuit shown in illustrates one example of a TLVR 100 - 3 that may be suitable for use in high current demand applications such as data centers.
[0147] Specifically, Fig.111 is a circuit diagram showing an N-phase multiphase buck converter with a TLVR 100-3. As shown, the input node 120-1 of each respective transformer 110-1 ... 110-N ... 120-N is connected to a switch node between two switches. One switch is a low-side switch 174-1 ... 174-N connected to ground, and the other switch is a high-side switch 172-1, ..., 172-N connected to a power supply voltage. The switch (in this example, provided by a MOSFET) can be driven by a controller so that a phase input is provided at an input node 120-1 ... 120-N. The output node 120-1 ... 120-N of each transformer 110-1 ... 110-N is connected to a common output node 130, which can be connected to a load.
[0148] Transients experienced at the input nodes 120-1 ... 120-N of only one transformer 110-1 ... 110-N (i.e., applied in only one phase) are magnetically reflected to the secondary inductor / winding 114-1 ... 114-N and through it coupled to the other transformers / phases 110-1 ... 110-N. This implements the overcurrent protection method described above because the transient current is initially sustained by the leakage inductance in each transformer 110-1 ... 110-N.
[0149] However, the above concept can be applied to any voltage regulator including a trans-inductor. Fig.12 is a general diagram of a TLVR 100-4 to which the described concepts for overcurrent fault protection may be applied. Any converter topology using an inductor may be able to use the trans-inductor concept, and any converter using the trans-inductor concept may be able to apply a dynamic / variable overcurrent protection threshold using the described overcurrent protection method.
[0150] The first sensor devices 200-1, 200-2, 200-3 (for example, see Figures 7 to 10 The sensor device described) can be used with Fig.12 The TLVR 100-4 is arranged together to sense current flowing through the secondary inductors 114-1 ... 114-N of the transformer included in the inductor-based converter. The output from the first sensor device is sent to the controller 300 for determining the generation of an over-current fault signal.
[0151] In the claims, any reference numerals placed between brackets shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements and steps other than those listed in the claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The embodiment may be implemented by hardware comprising several different elements. In a device claim listing several means, several of these means may be implemented by the same item of hardware. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Furthermore, in the appended claim list, "at least one of A, B and C" shall be interpreted as (A and / or B) and / or C.
[0152] The embodiments discussed herein may be practiced in various components such as integrated circuit modules. The design of integrated circuits is generally a highly automated process. Complex and powerful software tools are available to convert logic-level designs into semiconductor circuit designs that are etched and formed on semiconductor substrates.
[0153] Programs such as those provided by Synopsys, Inc. of Mountain View, Calif., and Cadence Design, Inc. of San Jose, Calif., use well-established design rules and a library of pre-stored design modules to automatically route conductors and position components on a semiconductor chip. Once the design of a semiconductor circuit is complete, the resulting design in a standardized electronic format (e.g., Opus, GDSII, etc.) can be transmitted to a semiconductor manufacturing facility or "fab" for fabrication.
[0154] The following embodiments are disclosed:
[0155] 1. A system (1000-1, 1000-2, 1000-3) for overcurrent protection of a multi-phase trans-inductor voltage regulator (100-1, 100-2, 100-3, 100-4), the multi-phase trans-inductor voltage regulator comprising a plurality of transformers (110-1, 110-2, 100-N), each transformer comprising a primary inductor (112-1, 112-2, 112-N) and a secondary inductor (114-1, 114-2, 114-N), wherein the primary inductor of each transformer is arranged between a corresponding input node (120-1, 120-2, 120-N) and a common output node (130), the secondary inductors of the plurality of transformers are connected in series, and wherein each input node is connected to a corresponding phase input of a plurality of phase inputs, the system comprising:
[0156] a first sensor device (200-1, 200-2, 200-3) configured to measure a current flowing through the secondary inductor; and
[0157] A controller (300) is configured to generate an over-current fault signal based at least in part on the measured current flowing through the secondary inductor.
[0158] 2. The system according to embodiment 1 further includes a second sensor device configured to measure a phase current provided by a phase input for each of at least two phase inputs of a plurality of phase inputs, and wherein the controller (300) is configured to generate an overcurrent fault signal further based on at least one of the measured phase currents.
[0159] 3. The system according to embodiment 1 or 2, wherein the controller (300) is configured to generate an overcurrent fault signal further based on a static overcurrent protection threshold.
[0160] 4. A system according to embodiment 2, wherein the controller (300) is configured to generate an overcurrent fault signal further based on a static overcurrent protection threshold, and wherein the controller is configured to generate the overcurrent fault signal in response to at least one of the measured phase currents exceeding the sum of the static overcurrent protection threshold and the measured current flowing through the secondary inductor (114-1, 114-2, 114-N).
[0161] 5. The system of embodiment 4, wherein the controller (300) is configured to:
[0162] determining a dynamic overcurrent protection threshold based on a sum of a measured current flowing through the secondary inductor (114-1, 114-2, 114-N) and a static overcurrent protection threshold; and
[0163] An overcurrent fault signal is generated in response to at least one of the measured phase currents exceeding a dynamic overcurrent protection threshold.
[0164] 6. A system according to any one of embodiments 1 to 5, wherein the first sensor device (200-1) includes a DC impedance current sensing circuit, which is configured to measure the current flowing through the secondary inductor (114-1, 114-2, 114-N).
[0165] 7. The system of any one of embodiments 1 to 5, wherein the multiphase trans-inductor voltage regulator (100-1, 100-2, 100-3, 100-4) further comprises a compensation inductor (140) connected in series with the secondary inductor (114-1, 114-2, 114-N), and wherein the first sensor device (200-2) comprises:
[0166] A DC impedance current sensing circuit is configured to measure current flowing through the secondary inductor and the compensation inductor.
[0167] 8. The system of embodiment 7, wherein the DC impedance current sensing circuit is disposed across the compensation inductor (140).
[0168] 9. The system of embodiment 7, wherein the DC impedance current sensing circuit is disposed across the compensation inductor (140) and the secondary inductor (114-1, 114-2, 114-N).
[0169] 10. The system according to any one of embodiments 6 to 9, wherein the first sensor device (200-2) further comprises an amplifier (250) to amplify the output of the DC impedance current sensing circuit.
[0170] 11. The system according to any one of embodiments 1 to 5, wherein the first sensor device (200-1) comprises:
[0171] a current transformer (210) comprising a primary side connected in series with a secondary inductor (114-1, 114-2, 114-N) and a secondary side terminated by a resistor (220); and
[0172] A voltage sensor is configured to sense a voltage across the resistor.
[0173] 12. The system according to any one of embodiments 1 to 5, wherein the first sensor device (200-3) comprises:
[0174] A shunt resistor (260) connected in series with the secondary inductor (114-1, 114-2, 114-N);
[0175] A voltage sensor is configured to sense a voltage across the shunt resistor.
[0176] 13. The system of embodiment 12, wherein the first sensor device (200-3) further comprises an amplifier (270) configured to amplify a voltage across the shunt resistor (260).
[0177] 14. The system of embodiments 10 or 13, wherein the amplifier (250, 270) is an instrumentation amplifier.
[0178] 15. The system according to any one of embodiments 11 to 14, wherein the controller (300) is configured to:
[0179] receiving the sensed voltage; and
[0180] A current flowing through the secondary inductor (114-1, 114-2, 114-N) is determined based on the sensed voltage.
[0181] 16. The system of any one of embodiments 1 to 15, wherein a static overcurrent protection threshold is equal to a rated inductor saturation current of at least one of the primary inductors (112-1, 112-2, 112-N).
[0182] 17. A multi-phase trans-inductor voltage regulator, comprising:
[0183] a plurality of transformers (110-1, 110-2, 110-N), each transformer comprising a primary inductor (112-1, 112-2, 112-N) and a secondary inductor (114-1, 114-2, 114-N), wherein the primary inductor of each transformer is disposed between a corresponding input node (120-1, 120-2, 120-N) and a common output node (130), the secondary inductors of the plurality of transformers are connected in series, and wherein each input node is connected to a corresponding phase input of a plurality of phase inputs; and
[0184] A system (1000-1, 1000-2, 1000-3) according to any one of embodiments 1 to 16.
[0185] 18. The multi-phase trans-inductor voltage regulator (100-1, 100-2, 100-3, 100-4) according to embodiment 17, wherein the multi-phase trans-inductor voltage regulator is a buck converter.
[0186] 19. A method for overcurrent protection of a multi-phase trans-inductor voltage regulator (100-1, 100-2, 1003, 100-4), the multi-phase trans-inductor voltage regulator comprising a plurality of transformers (110-1, 110-2, 110-N), each transformer comprising a primary inductor (112-1, 112-2, 112-N) and a secondary inductor (114-1, 114-2, 114-N), wherein the primary inductor of each transformer is arranged between a corresponding input node (120-1, 120-2, 120-N) and a common output node (130), the secondary inductors of the plurality of transformers are connected in series, and wherein each input node is connected to a corresponding phase input of a plurality of phase inputs, the method comprising:
[0187] measuring the current flowing through the secondary inductor; and
[0188] An overcurrent fault signal is generated based at least in part on the measured current flowing through the secondary inductor.
[0189] 20. A controller (300) for overcurrent protection of a multi-phase trans-inductor voltage regulator (100-1, 100-2, 1003, 100-4), the multi-phase trans-inductor voltage regulator comprising a plurality of transformers (110-1, 110-2, 110-N), each transformer comprising a primary inductor (112-1, 112-2, 112-N) and a secondary inductor (114-1, 114-2, 114-N), wherein the primary inductor of each transformer is arranged between a corresponding input node (120-1, 120-2, 120-N) and a common output node (130), the secondary inductors of the plurality of transformers are connected in series, and wherein each input node is connected to a corresponding phase input of a plurality of phase inputs, the controller being configured to:
[0190] obtaining a measured current flowing through the secondary inductor; and
[0191] An overcurrent fault signal is generated based at least in part on the measured current flowing through the secondary inductor.
Claims
1. A system (1000-1, 1000-2, 1000-3) for overcurrent protection of a multi-phase trans-inductor voltage regulator (100-1, 100-2, 100-3, 100-4), the multi-phase trans-inductor voltage regulator comprising a plurality of transformers (110-1, 110-2, 100-N), each transformer comprising a primary inductor (112-1, 112-2, 112-N) and a secondary inductor (114-1, 114-2, 114-N), wherein: The primary inductor of each transformer is arranged between a corresponding input node (120-1, 120-2, 120-N) and a common output node (130), the secondary inductors of the plurality of transformers are connected in series, and wherein each input node is connected to a corresponding phase input of a plurality of phase inputs, the system comprising: a first sensor device (200-1, 200-2, 200-3) configured to measure a current flowing through the secondary inductor; and A controller (300) is configured to generate an over-current fault signal based at least in part on the measured current flowing through the secondary inductor.
2. The system of claim 1 , further comprising a second sensor device configured to measure, for at least two of the plurality of phase inputs, phase currents provided by the respective phase inputs, and wherein, The controller (300) is configured to generate the overcurrent fault signal further based on at least one of the measured phase currents.
3. The system according to claim 1 or 2, wherein: The controller (300) is configured to generate the overcurrent fault signal further based on a static overcurrent protection threshold.
4. The system according to claim 2, wherein: The controller (300) is configured to generate the overcurrent fault signal further based on a static overcurrent protection threshold, and wherein the controller is configured to generate the overcurrent fault signal in response to at least one of the measured phase currents exceeding the sum of the static overcurrent protection threshold and the measured current flowing through the secondary inductor (114-1, 114-2, 114-N).
5. The system according to claim 4, wherein: The controller (300) is configured to: determining a dynamic overcurrent protection threshold based on a sum of a measured current flowing through the secondary inductor (114-1, 114-2, 114-N) and the static overcurrent protection threshold; and The overcurrent fault signal is generated in response to at least one of the measured phase currents exceeding the dynamic overcurrent protection threshold.
6. The system according to any one of claims 1 to 5, wherein: The first sensor device (200-1, 200-2, 200-3) includes a DC impedance current sensing circuit configured to measure a current flowing through the secondary inductor (114-1, 114-2, 114-N).
7. The system according to any one of claims 1 to 5, wherein: The multi-phase trans-inductor voltage regulator (100-1, 100-2, 100-3, 100-4) further comprises a compensation inductor (140) connected in series with the secondary inductor (114-1, 114-2, 114-N), and wherein the first sensor device (200-1, 200-2, 200-3) comprises: A DC impedance current sensing circuit is configured to measure current flowing through the secondary inductor and the compensating inductor.
8. The system according to claim 7, wherein: The DC impedance current sensing circuit is disposed across the compensation inductor (140).
9. The system according to claim 7, wherein: The DC impedance current sensing circuit is disposed across the compensation inductor (140) and the secondary inductor (114-1, 114-2, 114-N).
10. The system according to any one of claims 1 to 5, wherein: The first sensor device (200-1, 200-2, 200-3) comprises: a current transformer (210) comprising a primary side connected in series with the secondary inductor (114-1, 114-2, 114-N) and a secondary side terminated by a resistor (220); and A voltage sensor is configured to sense a voltage across the resistor.
11. The system according to any one of claims 1 to 5, wherein: The first sensor device (200-1, 200-2, 200-3) comprises: A shunt resistor (260) connected in series with the secondary inductor (114-1, 114-2, 114-N); A voltage sensor is configured to sense a voltage across the shunt resistor.
12. The system according to claim 10 or 11, wherein: The controller (300) is configured to: receiving the sensed voltage; and A current flowing through the secondary inductor (114-1, 114-2, 114-N) is determined based on the sensed voltage.
13. A multi-phase trans-inductor voltage regulator, comprising: a plurality of transformers (110-1, 110-2, 110-N), each transformer comprising a primary inductor (112-1, 112-2, 112-N) and a secondary inductor (114-1, 114-2, 114-N), wherein the primary inductor of each transformer is arranged between a corresponding input node (120-1, 120-2, 120-N) and a common output node (130), the secondary inductors of the plurality of transformers are connected in series, and wherein each input node is connected to a corresponding phase input of a plurality of phase inputs; and A system (1000-1, 1000-2, 1000-3) according to any one of claims 1 to 12.
14. A method for overcurrent protection of a multi-phase trans-inductor voltage regulator (100-1, 100-2, 100-3, 100-4), the multi-phase trans-inductor voltage regulator comprising a plurality of transformers (110-1, 110-2, 110-N), each transformer comprising a primary inductor (112-1, 112-2, 112-N) and a secondary inductor (114-1, 114-2, 114-N), wherein: The primary inductor of each transformer is arranged between a corresponding input node (120-1, 120-2, 120-N) and a common output node (130), the secondary inductors of the plurality of transformers are connected in series, and wherein each input node is connected to a corresponding phase input of a plurality of phase inputs, the method comprising: measuring a current flowing through the secondary inductor; and An overcurrent fault signal is generated based at least in part on the measured current flowing through the secondary inductor.
15. A controller (300) for overcurrent protection of a multi-phase trans-inductor voltage regulator (100-1, 100-2, 100-3, 100-4), the multi-phase trans-inductor voltage regulator comprising a plurality of transformers (110-1, 110-2, 110-N), each transformer comprising a primary inductor (112-1, 112-2, 112-N) and a secondary inductor (114-1, 114-2, 114-N), wherein: The primary inductor of each transformer is arranged between a corresponding input node (120-1, 120-2, 120-N) and a common output node (130), the secondary inductors of the plurality of transformers are connected in series, and wherein each input node is connected to a corresponding phase input of a plurality of phase inputs, and the controller is configured to: obtaining a measured current flowing through the secondary inductor; and An overcurrent fault signal is generated based at least in part on the measured current flowing through the secondary inductor.